Big Bend Paleo-Geo Journal

Big Bend Paleo-Geo Journal An Open Access Informal Publication from Mosasaur Ranch, Terlingua, Texas All rights reserved Copyright; Kenneth R. Barnes, 2014 New info and corrections in red 2 / 3 / 2015 mosasaurranch.com bigbendpaleogeo.com DOCUMENTATION OF INDETERMINAT CHASMOSAURINAE, (DINOSAURIA: CERATOPSIDAE), FOSSIL REMAINS FROM THE LOWER PART OF THE UPPER SHALE MEMBER OF THE AGUJA FORMATION, WEST OF THE BIG BEND NATIONAL PARK, TEXAS. Kenneth R. Barnes; (Ken Barnes) HC-65 Box 300, Alpine, Texas, 79830, e-mail kbarnes@bigbend.net. ABSTRACT These Chasmosaurinae fossil remains are from the lower part of the Upper Shale Member of the Aguja Formation. The most complete juvenile skull is indeterminate because of lack of knowledge of the possible ontogenetic changes into adulthood. The nearby scrappy associated adult skull elements are also indeterminate. The juvenile skull is most similar to Pentaceratops having a long slim squamosal that turns slightly down and then slightly posterodorsal not sharply posterodorsal as in the other ceratopsians reported from the Aguja Formation from the Big Bend area of Texas. Other workers, (co-authors?), speculate that during ontogeny into adulthood the frill may turn sharply up as it is in (Chasmosaurus, (Agujaceratops), mariscalensis). Keywords; Big Bend, Texas Ceratopsian Pentaceratops Aguja Formation ETHICS STATEMENT

These localities are on file with the Texas Memorial Museum Vertebrate Paleontology Laboratory, Austin, Texas. These specimens final home are in the Texas Memorial Museum collection at Austin Texas known as the Barnes/Gaddis Special Collection. The only permit required was the permission from the Gaddis family, the private land owners in Texas, to remove, study, display and report these fossils. This permission is documented in the agreement donating these fossils to the Texas Memorial Museum, Austin, Texas. See mosasaurranch.com/journals, or bigbendpaleogeo.com All necessary permits were obtained for the described study, which complied with all relevant regulations. INTRODUCTION Dinosaur fossils have been recovered from numerous localities from the Upper Cretaceous of the Big Bend area of West Texas. There are few young dinosaur fossils reported from this area. There are no publications detailing with young ceratopsian cranial elements from the Upper Cretaceous of this area other than limited in Lehman (1989). The focus of this research is on the well preserved associated juvenile ceratopsian cranial elements recovered from quarry locality number TMM 45922. This quarry has also produced abundant un-associated fossils of young to sub adult, (non- Kritosaurus Saurolophinae), hadrosaurs and several other taxa. The fossils in an adjacent quarry, locality number TMM 45921, is <50 meters NW of the above locality number TMM 45922 and contains abundant associated elements of one large adult, (non-kritosaurus Saurolophinae), hadrosaur and a few associated elements of an adult ceratopsian from the east end of the quarry. These adult ceratopsian elements are addressed below. These two localities are located at the same horizon in a blue-grey clay-stone layer that has abundant carbonized wood and other plant matter indicating a marsh environment. These localities are about 4 meters above a thin red sandstone layer with abundant small oysters and clams representing most likely the last shallow short lived transgression of the Western Interior Seaway at these localities. It is interesting that basically only Kritosaurus hadrosaurs have been reported from the Big Bend area but all of the hadrosaur fossils, (of which there are many), in this Barnes/Gaddis Special Collection are not Kritosaurus. This difference would indicate a different and most likely older time period and or environment for these quarries compared with previous reported quarries from the Big Bend area. These hadrosaur fossils are being studied and described in another paper for future documentation. ABREVIATIONS: TMM, Texas Memorial Museum, Austin Texas; JVP, Journal of Vertebrate Paleontology; MRM, Mosasaur Ranch Museum, Terlingua, Texas. MNA, Museum of Northern Arizona, Flagstaff, Arizona; NMMNH, Mew Mexico Museum of Natural History, Albuquerque, New Mexico; WPA, Works Progress Administration; RM,

Rattlesnake Mountain area; BBNP, Big Bend National Park; SDMNH, San Diego Museum of Natural History, San Diego, California. HISTORY AND GEOLOGIC SETTINGS There is ongoing confusion as concerns the nomenclature of ceratopsians from the Big Bend area of Texas. Chasmosaurus mariscalensis, (Lehman, 1989), from the WPA localities in the Southern area of Big Bend National Park has been reassigned by Lucas, et al, (2006), to a new genus, Agujaceratops mariscalensis. Lucas et al, (2006), included reassigning a complete skull, Forster et al,(1993), (TMM 43098 1), from the Rattlesnake Mountain, (RM), area of the Upper Shale Member of the Aguja Formation near the Western edge of Big Bend National Park, to this same new genus Agujaceratops mariscalensis. I find the cranial material in these two papers above to be separated in time and do not resemble each other and should be two separate genera. In Forster et al, (1993), (their (Fig. 2)), they illustrate TMM 43098 1 to be approximately 50 meters above the Terlingua Creek Sandstone Member of the Aguja Formation. The skull, TMM 43098-1, locality is located on the NE trailing edge of an anticline illustrated by Maxwell et al, (1967), Plate II, Big Bend National Park, Texas. The author, (K. B.), measured that section on the ground with an accurate GPS and Brunton Pocket Transit. Based on the measured sections, (Figures1, 2 and 3), of the stratigraphy at the locality in Forster et al, (1993), it was found to be more than 100 meters above the Terlingua Creek Sandstone Member of the Aguja Formation, (TCSSM). The specimens in this paper are 30 meters above the (TCSSM), (Figures 2 and 3), and only 8.8 km north of the TMM 43098 locality, (Figure 1). The skull, TMM 43098-1, is 70 meters of deposition younger in geologic time than the specimens in this paper. Chasmosaurus mariscalensis, Lehman, (1989), was described from the fossils recovered mostly from the WPA-1 locality. This WPA-1 locality, my observation, is located on the eroded NW to SE trending crest of the Mariscal Mt. anticline, just above a massive sill intrusion, and it is difficult to accurately measure the stratigraphic horizon. Lehman, (1989), places the WPA localities at, less than 70 meters above the second progradational sandstone of the Aguja (within the upper shale member of Lehman, 1985a). The WPA localities are approximately 40 Km SE of the localities this paper. The Western Interior Seaway was receding to the SE during these time periods with slow transgressions and regressions. These transgressions and regressions leave paralic deposits that are not necessarily the same age in all localities, (Maxwell et al, 1967). The thickness of each paralic deposit locality represents the length of time it took to deposit these paralic deposits as the shoreline of the seaway receded to the SE passing that locality. In some places these paralic deposits are thick and other places they are thin. The paralic deposits and shale deposits of the Aguja formation are older to the northwest than similar ones

observed to the southeast and would likely place the localities this paper in the Late Middle Campanian to Early Late Campanian. The Upper Shale Member of the Aguja Formation represents millions of years. Most other workers seem to bundle everything together in about the same time period. FIGURE 1 ---- Map of the Big Bend area of Texas; A) localities this paper; B) locality TMM 43098-1, (Forster et al, 1993); C) WPA localities, (Lehman, 1989)

5 FIGURE 2 ---- Stratigraphic section for localities this paper, (TMM 45921), and (TMM 45922).

6 FIGURE 3; Comparison of measured stratigraphic sections at SE Terlingua area, localities, TMM 45921and TMM 45922, with measured section at locality TMM 43098 1, Rattlesnake Mt. area. These sections were measured on the ground by author K.B. and no fault was observed as called for by Lehman.

7 Figure 3a --- My measured section on map from Maxwell et al. 1967

8 Figure 3b--- measured section on Google, all dip sites same elevation +/- 5 feet; Pen formation; BSS = Basal Sandstone Member; LSM = Lower Shale Member Aguja Fm.; TCSS = Terlingua Creek Sandstone Member; USM = Upper Shale Member; # 58 = TMM 43098-1 Rattlesnake Mt. ceratopsian skull

9 SYSTEMATIC PALEONTOLOGY Ornithischia, Seely 1888 Ceratopsidae, Marsh 1890 Chasmosaurinae, Lamb 1915 (Indeterminate) DIAGNOSIS The juvenile has a long slim squamosal that is laid back and does not turn sharply up. The squamosal has 8 9 places for episquamosals. The cranium from the tip of the rostral to the articulation of the quadrate with the squamosal is very elongate. The quadrate has a wide flat area along the distal one third of the lateral edge for the articulation of the quadratojugal. The quadrate does not have a narrow neck midway up and it does not have a knob like process for the quadratojugal. The maxilla has a tall and thin anterior dorsal process. The supraoccipital has no deep paired depressions for neck musculature. The adult squamosal this paper has long low undulations with a sharp edge and no episquamosals. HOLOTYPE Locality TMM 45922 young ceratopsian elements listed below in referred specimens.. REFERRED SPECIMENS OF THE JUVENILE CERATOPSIAN See (Figures 4 and 5) for composite of the elements. Also see (Figures 4-a, and 5-a, b, c). TMM 45922-10, well preserved left squamosal; TMM 45922-27, right maxilla; TMM 45922-28, right quadrate; TMM 45922-30, left dentary; TMM 45922-31, distal fragment of left quadrate; TMM 45922-32, postorbital with supraorbital horn core; TMM 45922-26, fragment of right Jugal to orbit-maxilla-lacrimal; TMM 45922-37, right premaxilla; TMM 45922-25, supraoccipital; TMM 45922-18, left femur; TMM 45922-24, right femur.

10 FIGURE 4 ---- Composite young juvenile skull, locality TMM 45922; sq) left squamosal; po) left postorbital and hc) horn core; d) left dentary; mirror images =, q) right quadrate; j) fragment of right jugal; l) right lacrimal groove in jugal; mx) right maxilla; pmx) right premaxilla; o) orbit. Figure 4-a ---- Composite skull made with 3-D prints on makeshift armature.

11 FIGURE 5 ---- Dorsal view of composite skull; so) supraoccipital articulated with quadrates; other abbreviations are the same as (Fig. 4)

12 Figure 5-a --- Left lateral display view Figure 5b --- lateral view with original fossils

13 Figure 5c--- Right lateral study view of found elements, no predentary or rostral found LOCALITY AND HORIZON These fossils were collected from locality TMM 45922, SE of Terlingua Ghost Town Brewster County, Texas. The exact locality is available at the Vertebrate Paleontology Laboratory, University of Texas at Austin. DISCREPTION LEFT SQUAMOSAL: (Figure 6). The left squamosal, (TMM 45922-10), is almost perfect. It is 360mm long / 133mm wide. The 6 8mm thick ventral edge of the blade has evidence for, 8 or 9 very low undulations for episquamosals. This number of episquamosals is between reported, 6 episquamosals on young to adult, (UTEP P.37.7.052;.052;.062;.045;.059; and (.046; and.066)), in Lehman, (1989), and 10 episquamosals on (43098-1), Forster et al, (1993). The lateral side of the anterior part of the blade is facing slightly dorsolateral and then has a twist laterally so the posterior end side is facing dorsolateral. There are no vascular grooves on this squamosal. There are a few roughened areas for muscle attachments. In dorsolateral view the ventral edge of the blade is convex on the anterior one half and then is straight to the posterior tip. In

14 Lehman, (1989) and Forster et al, (2006), this area is convex for the full length. It is concave along the 16mm wide dorsomedial edge to accept the lateral edge of the lateral parietal bar. The paroccipital buttress is well preserved and sigmodial in medial view with a 4mm sharp ridge oriented anteromedial on the posterior side of the round buttress grove. The groove or pit for the quadrate is 35+mm deep under a flange directed anterior from the paroccipital buttress. This flange is crushed in and should be directed medially and would lie between the posterior side of the quadrate and the anterior side of the exoccipital. The lateral edge of the quadrate would be between the above flange and a ventral directed triangular bone that forms the posterior side of the laterotemporal fenestra. The anterodorsal corner of the squamosal shows evidence that it would tongue into a groove on the posterodorsal corner of the postorbital. There is a small part of the anteroventral corner of the postorbital process of the squamosal missing. This missing part of the squamosal shows evidence of a groove to accept a tongue of the posteroventral corner of the postorbital, or a part of the dorsal edge of the jugal. This groove is present on the anteroventral corner of the right postorbital process of the squamosal of Pentaceratops MNA-V-1747, personal observation. This squamosal is the long and slim type and does not turn upward sharply as in, Lehman, (1989) or Forster et al, (1993); rather it turns downward slightly posterior to the horizontal anterior postorbital process of the squamosal and then turns gently upward. See (Figures 7, 8 and 9). It more closely resembles Pentaceratops.

15 FIGURE 6; left squamosal: A) dorsolateral; B); ventromedial view; par) paroccipital buttress; q) 35mm deep quadrate groove; ltf) laterotemporal fenestra; po) postorbital process; ma) roughened area for muscle attachment; m) smaller roughened areas, m lines on (A), show directions of small roughened areas and bone grain. FIGURE 7; Comparison of A), Composite young juvenile skull: B) reversed cast of TMM 43098-1.

16 FIGURE 8; Comparison of Chasmosaurinae right squamosals from the southwest with postorbital horizontal: A) reversed, TMM 45922 10, This paper; B) reversed, MNA V 1747, (MNA); C) TMM cast of, (RM) site, TMM 43089 1, (Forster et al 1993); D) 6 episquamosals, From the (WPA) sites, (Lehman 1989), UTEP P.37.7.066; and E).052,

17 FIGURE 9; Comparison of a cast of adult Pentaceratops, MNA V- 1747, at NMMNH, and this young juvenile skull from locality TMM 45922. COMPARISON OF THIS JUVENILE CERATOPSIAN SQUAMOSAL WITH PARTS OF A NEAR BY ADULT SQUAMOSSAL Locality TMM 45921, (TMM 45921-3A and -3B), adult squamosal and other elements listed below that are not figured. This juvenile left squamosal, (TMM 45922-10), is from Locality TMM 45922 and the adult left squamosal parts, (TMM 45921-3A and -3B), are from Locality TMM 45921. These two localities are <50 meters apart and at the same horizon. These recovered parts of the adult left squamosal, part of a descending lateral bar of the parietal, and a part of the nasal horncore, were found eroded out on the surface in many fragments associated with an in situ 250mm long ceratopsian cervical bar with an articulated rib, one cervical vertebra with rib, 2 dorsal vertebra and the left scapula. Some of the squamosal fragments above have been reassembled into two parts. Six fragments make up the distal tip, (45921-3A), and 16 fragments, (45921-3B), make up a large portion of the middle - anterior part of the blade of the squamosal, (Figures 10 and 11). Additional fragments of this frill area exist but cannot be reassembled at this time. The reassembled distal tip part, TMM 45921-3A is a triangle shape in dorsolateral view. This

18 distal tip piece in crossection is slightly concave on the dorsolateral side and slightly convex on the ventromedial side. The 125mm long dorsomedial edge is 25-30mm thick and has a rugose concave area for its length to accept the lateral side of the descending process of the posterolateral parietal bar, (Figure 10). From this point parts are missing along the dorsomedial edge. From the posterior sharp tip of this posterior part the ventrolateral edge has a small undulation that has a sharp edge that is rugose on the lateral side for 60mm by 10mm may be evidence of a missing small episquamosal, (Figure 10). This would be the only evidence of possible episquamosals on these squamosal parts. The next 90mm shows there was part of a long rounded undulation with a small sliver broken off the lateroventral edge. The broken anterior edge of this piece is 16mm thick in the middle. Along the ventrolateral edge parts are missing to the larger part, TMM 45921-3B. This larger part, in crossection, is also concave on the dorsolateral side and convex on the ventromedial side. The broken distal edge is only 8mm thick in the middle. From the posteroventral corner to the anteroventral corner of this larger part it is 224mm. There are most of two 12 cm long and rounded undulations along this ventrolateral edge. This edge is slightly rugose with vascular grooves and each rounded undulation having a sharp edge of the bone with rounded edge in between undulations with no evidence of co-ossified episquamosals. There are at least two and a half undulations missing between these two parts (Figure 12). This larger squamosal must be from an adult ceratopsian or one that has no episquamosals, or they have been absorbed into the border.

19 FIGURE 10;, Anterior and posterior parts of the large left squamosal; A - C) midanterior part, (TMM 45921 3B) ; A) part of posterior break; B) part of anterior break; C) all of the posterior break; D) and E) posterior tip, (TMM 45921 3A; D) medial rugose concave edge; E) dorsolateral views; rc) part of the rugose concave edge for the lateral parietal bar; es?) Possible small episquamosal facet; sl) missing sliver of bone; numbers) indicate thickness at that spot; s) sharp edge: r) rounded edge; apff) anterior parietal fan facet.

20 FIGURE 11; A) (21-3B) lateral, and B) medial views: arrows) indicates the direction of the bone grain, lateral rising, and thickening in that area; numbers) indicate thickness at that spot; s) sharp edge: R) rounded edge; apff) anterior parietal fan facet;

21 FIGURE 12; B) A possible reconstruction of the large left squamosal parts, (TMM 45921 3A), and TMM 45921 3B), compared to A) the left squamosal of (MNA V 1747); apf) anterior parietal fan; r) rounded edge; s) sharp edge; c) concave medial edge for the lateral parietal bar; white arrows) direction of the bone grain, rising up laterally, and thickening; numbers) indicate thickness in mm at that spot. DISCUSSION ON SQUAMOSALS: This juvenile ceratopsian squamosal and the adult squamosal are from the same horizon and within <50 meters of each other and are the same taxa. Ontogenetically a more mature ceratopsian squamosal would most likely grow longer and slimmer proportional to a juvenile squamosal. A juvenile squamosal may not have co-ossified episquamosals but large juveniles and adult usually do have them. However in Triceratops, (Horner and Goodwin, Proc. R. Soc. 2006), they show that Triceratops had episquamosals in juveniles and fused by sub adult and completely reabsorbed into the border of the squamosal in fully adult. This larger squamosal probably came from an adult. On the left squamosal in Pentaceratops, (Figure 12) (MNA-V-1747), 1/3 of the posterior ventrolateral edge is rounded and has no episquamosals. The squamosal of the reported Chasmosaurinae, (Figure 8)), from the Big Bend area do not resemble the squamosal from this young juvenile. The squamosal of MNA-1747 does resemble the squamosal of this young individual. LEFT POSTORBITAL WITH SUPRAORBITAL HORNCORE: (Figure 13). The left postorbital with supraorbital horncore, (TMM 45922-32), of this young ceratopsian is almost complete except for a few small fragments missing along the dorsomedial suture area and a lost small fragment on the distal side of the horncore. From the center of the medial side of the 40mm in diameter base of the horncore, along the fragmented edge

22 of the medial suture, it is 100mm to the posteromedial end. For most of this length the dorsal side is flat, 30mm wide, and horizontal. This blade like distal directed squamosal process part in cross section is an upside down L in shape. The dorsal 30mm wide horizontal side is directed laterally then curves ventrally. The 60mm lateral side is directed ventrolateral as it is in the squamosal. The posterior directed end is sub triangular with a groove in its posterodorsal end where most of the tip would underlay the squamosal and a smaller part of the groove would overlay the anterodorsal end of the squamosal. The posterior side of the orbit is vertical and straight for 40mm from the anteroventral corner of the postorbital. The dorsal side of the orbit is rounded for 50mm to the suture for the palpebral and prefrontal. For half this distance there is a lateral directed 15mm wide brow that blends into the posterolateral side of the horncore. The centerline of the base of the 40mm in diameter horncore is centered over the anterior 1/3 of the orbit. The 70mm, (from the roof of the orbit), tall, blunt, conical horn core is straight and angling off vertical 20 degrees to the anterolateral. This horncore has evidence of vascular groves. There is a depression, cornual sinus, entering the center of the horncore from the roof of the orbit. The ventral edge of the postorbit overlaps the jugal 15mm. FIGURE 13; Left postorbital with horncore, TMM 45922-32; A) lateral, B) medial, C)anterior, D), posterior, and E) ventral views; pal) palpebral; o) orbit; sq) groove for the anterior tongue of the squamosal; jsu) 15mm wide suture overlies the jugal here; pf) prefrontal; cs) cornual sinus; ro) roof of orbit.

23 A SMALL ANTERIOR FRAGMENT OF THE RIGHT JUJAL: (Figure 14).This small fragment of the right jugal, (TMM 45922-26), of this juvenile ceratopsian is significant in that it is the anterior part of the jugal. It has part of the ventral side of the orbit, most of the deep groove that straddles the posterior dorsal process of the maxilla, and has a shallow groove along its anterior edge for the lacrimal. This groove ends ventrally over a tri-lobate foramen rising out of the maxilla, (Figures 15 and 16). FIGURE 14; Right jugal fragment, TMM 45922 26; A) lateral, B) lingual, C) dorsal, and D) ventral views; mx ) jugal process of the maxilla; o) orbit; mxbj) maxilla buttress in jugal; l) lacrimal groove; smx) saddle for maxilla jugal process. RIGHT MAXILLA: (Figure 15 and 16) This right maxilla, (TMM 45922-27), of this juvenile ceratopsian is missing only the posterior 20mm of the posterior dorsal, (jugal),

24 process; a tiny dorsal tip of the anterior dorsal process; almost all of the teeth; and other minor fragments. The rostroventral 20mm tip is crushed slightly posterodorsaly. The tooth bearing area is 160mm long with 19 alveoli. There are 9-10 partial teeth. The alveoli are directed ventrolateral in such a manner that the teeth would meet the dorsolingual directed dentary teeth to produce a shearing action for mastication of plant matter. Between the lateral side of the posterolingual interior bottom of the antorbital fenestra and the medial side of the posterodorsal, (jugal), process and the anteroventral interior end of the jugal there is a tri-lobate foramen rising up from the interior of the maxilla. This foramen would be directly below and connected with the descending lacrimal groove in the anterior edge of the jugal and its juncture with the lacrimal and palatine. The anterodorsal process is very thin. There is little resemblance of this young ceratopsian maxilla with the maxillas from the WPA sites, (Figure 17 and 18), see (fig. 11, Lehman, 1989). The maxilla in, TMM 43089 1, is co-ossified with other elements and is of no help. The adult maxillas of, MNA V 1747 are not co-ossified and are very similar to this young ceratopsian maxilla, (Figure 19).

25 FIGURE 15; Right maxilla, TMM 45922 27; A) lingual, B) lateral, C) dorsal, and D) ventral views; j) jugal process of the maxilla; ec) ectopterygoid process; f) foramen; tf) tri-lobate foramen rising out of maxilla; aof) antorbital fenestra; fr) fracture, anterior tip pushed posterodorsaly; sf) special foramen; vt) very thin anterior dorsal process.

26 FIGURE 16; Right maxilla, TMM 45922 27, and jugal fragment, TMM 45922 26, in their natural position; A) anterior, B) dorsal, C) expanded insert; ec) ectopterygoid process; o) orbit; sy) symphysis; lg) lacrimal groove; j) jugal; paf) palatine facet; tf) trilobate foramen rising out of maxilla; lf) lacrimal facet; aof) antorbital fenestra.

27 FIGURE 17; Chasmosaurus mariscalensis maxilla, WPA sites, (Lehman, 1989), see descriptions (Figure 18). A) UTEP P.37.7.088; B) UTEP P.37.7.086.

28 FIGURE 18; Chasmosaurus mariscalensis maxilla, WPA sites, (Lehman, 1989), lateral views, A) UTEP P.37.7.088, B).086, C).087, D).082.: mjp) Maxilla-jugal process: adp) anterior dorsal process: ec) ectopterygoid process: Note: These WPA sites are not far above a massive Sill intrusion and the fossils are in very hard baked shale and hornfels and are not well preserved.

29 FIGURE 19; Comparison of medial view, A) this young juvenile maxilla, (TMM 45922 27), and B) adult maxilla from (MNA V 1747); j) jugal process; vt) very thin anterior dorsal process. RIGHT QUADRATE AND A DISTAL FRAGMENT OF THE LEFT QUADRATE: The right quadrate, (TMM 45922 28, (Figure 20), of this juvenile ceratopsian is almost complete and missing only the very thin parts in the pterygoid process area. It is 148mm tall from the lateral condyle to the top of the posterodorsal corner of the proximal blade that would insert 35mm into the quadrate groove in the squamosal. The medial side of the blade is flat. The posterolateral edge curves sharply down and has a small raised ridge that gets thicker distally for 50mm where it is 12mm thick. Medially 10mm from this posterolateral edge the 65mm wide dorsal blade- like process thins from 4 5 mm thick down to 1-2mm thick on its anterior edge. In anterolateral view this posterolateral edge is fairly straight down to 20mm above the lateral condyle where it flares slightly laterally down to the lateral condyle. This condyle is 15mm thick. The lateral, quadratojugal, side of this lateral edge is flat and 20mm wide for 40mm up from the condyle and then tapers down quickly to 8-10mm wide. This lateral 20mm wide and 40mm tall flat area is not present on the quadrate in C. mariscalensis described and figured in Lehman, (1989); rather he calls for a knob-like brace, (Figure 21), on the anterolateral edge about where the top of the 20mm wide flat area terminates on this specimen. This specimen does

30 have a small rugose area on the lateral flat side at that location. This 20mm wide area would be the quadratojugal process of the quadrate. The quadrates, (Figure 22), of MNA-V-1747 more closely resemble this young ceratopsian. These quadrates also have the wide, rugose, flat area for the quadratojugal. The right quadrate of MNA-V-1747, as illustrated in Rowe et al,(1981), (fig. 2.1 and 2.2), has a wide, thin loop of bone for the pterygoid process, this part must have been lost in preparation. The distal 45mm of the left quadrate, (TMM 45922 31), is identical to the distal end of the right quadrate except opposite. FIGURE 20; Right young juvenile quadrate TMM 45922 28; A) posterior, B) anterior C) medial and D) lateral views; E) expanded quadratojugal area; pt) pterygoid groove or buttress; qj) wide flat quadratojugal process; ar) articular; sq) squamosal; so) supraoccipital

31 FIGURE 21; Left quadrate, UTEP P.37.7.050, A) Lateral, and B) Medial views, my photos: narrow neck and; Knob) as per Lehman, (1989).

32 FIGURE 22; ---- Comparison of this right posterior view juvenile: A) (TMM 45922 28) quadrate with, B) right posterior and C) left posterior views of the quadrates of Pentaceratops, MNA-V-1747: pt) pterygoid groove or buttress; sq) squamosal; ar) articular; qj) flat and wide quadratojugal process with no knob so) supraoccipital LEFT DENTARY: The entire main body of the left dentary, (TMM 45922-30), of this juvenile ceratopsian is present. (Figure 23), There are no complete teeth or the thin sheath of bone on the lingual side of the teeth above the special foramen. None of the accessory parts for the articular area are present. This dentary is 210mm long by 110mm high at the coronoid process. The posterior 2/3 of the ventral edge is straight then the anterior 1/3 of the ventral edge turns up about 20⁰. The tooth bearing area is 160mm long, same as the maxilla. There are 17 alveoli. There are only 3-4 tiny emerging teeth.

33 There is another juvenile right dentary from this site, (TMM 45922 33), not figured. That dentary is 15% - 20% larger than TMM 45922 30, poorly preserved, and crushed. This would belong to a slightly larger juvenile individual than described above. FIGURE 23; Left dentary of young juvenile, TMM 45922 30; A) lingual, B) lateral, and C) dorsal; sp) splenial process. RIGHT PREMAXILLA; (Figure 24) This juvenile ceratopsian right premaxilla, (TMM 45922-37), (Figure24A), is missing a small part of the posterior maxillary process and the anterior edge that would surround the premaxilla fenestra. The exposed broken thin edge of this fenestra indicates it was

34 probably open. The premaxilla, (Figure 24B) (UTEP P.37.7.067), from the WPA site is similar. The adult premaxilla area, (Figure 25), of MNA-V-1747 is complete and a good example of anteropostero sliming and elongation of adult ceratopsian skull elements. FIGURE 24; Comparison of, A) Right premaxilla, TMM 45922 37, this juvenile, and B ) Right premaxilla, UTEP P.37.7.067, WPA site, sub adult, see (fig. 2, - B, Lehman 1989).

35 FIGURE 25; (C), Comparison of the left premaxilla area of MNA-V-1747 with figure 24; pmf) premaxilla fenestra; en) external nasal; nhc) nasal horn core; ros) rostral. SUPRAOCCIPITAL The supraoccipital, (TMM 45922-25), (Figure 26), would contribute only (?) 3-4mm to the dorsal edge of the foramen magnum. In all adult ceratopsids the paired exoccipitals push the supraoccipital up and join each other above the foramen magnum excluding the supraoccipital from the foreman magnum (Goodwin et al, 2006; Hatcher, 1907; Dodson and Currie, 1989; Forster, 1996b). This condition of the supraoccipital is discussed at length in Goodwin et al, 2006; therefore it is not necessary herein. The ventrolateral wings of the supraoccipital have a rugose suture area and a ventromedial groove that would overlie the dorsoanteromedial ends of the exoccipitals. The ventrolateral side of the supraoccipital above the exoccipitals there is a thin vertical oriented suture that would overlie the dorsomedial edge of the quadrate, see figure (26- a) and (26-b). Note: I have not found the above mentioned anywhere, I am sure they would be separated in older specimens. There are no paired depressions, with a thin septum of bone separating them, on the posterior surface for the neck musculature that is present on most ceratopsians. These paired depressions are not present on the adult Pentaceratops MNA-V-1747, (Figure 27). There is a roughened area on the posterodorsal side of this supraoccipital assumed for the attachment of the parietal. The dorsal part of this supraoccipital has a box like groove between two wings of bone that

36 would contain the postfrontal fontanelle. In life the dorsal part of supraoccipital would be leaning to the posterior. Figure 26; Supraoccipital TMM45922-25; A) anterior view; B) posterior view; C) ventral view; exo) exoccipitals; q) quadrate; fm) foramen magnum; P) parietal process; The clouded area on A and C is the attachment area for the exoccipitals. Figure 26-a; posteroventral view of articulation of quadrates with supraoccipital in this young ceratopsian

37 Figure 26-b: Dorsal view of above with actual supraoccipital with 3-D prints.. Figure 27; The underside of the occipital-parietal area of the adult Pentaceratops MNA- V-1747; bo) basioccipital; oc) occipital condyle; exo) exoccipitals; fm) foramen

38 magnum; p) parietal. No deep paired depressions on the supraoccipital for neck musculature. LEFT AND RIGHT JUVENILE CERATOPSIAN FEMURS: The right femur, (TMM 45922 26), (Figure 28, A, D and E) is 275mm tall from the top of head to the bottom of the medial condyle. From the top of the greater trochanter to the bottom of the lateral condyle it is 280mm tall. The lateromedial diameter at mid shaft is 42mm. The posteroanterior diameter at mid shaft is 37mm. On the proximal end it is 70mm across from the medial side of the head to the lateral side of the greater trochanter. The head of this femur has a rugose overhanging lip. There is a deep cleft between the head and the greater trochanter on the posterior side that extends down the shaft for some distance. The lesser trochanter is located on the anterior side of the greater trochanter and is rod like and terminates well below the greater trochanter. The distance across the distal condyles is 55mm, with the medial condyle being a little wider than the lateral one. The medial condyle is boot shaped and 70mm from the anterior to the posterior with the toe of the boot extending 25mm posterior to the shaft. There is gouged out area just above the anterior heel of this condyle that may be a bite mark. The intercondylar grove is open. The lateral condyle is missing the posterior toe of the boot. These distal condyles are twisted +/- 25 degrees to the posteromedial. The forth trochanter is located along the posteromedial edge of the shaft. The fourth trochanter is thin and +/- 75mm long and slowly rises distally from the shaft for 50mm to where it is most pendant, at 12mm high, and then slopes back down for 25mm to the shaft. The fourth trochanter is centered slightly below the mid-length of the shaft and there is no second ridge as mentioned in, (Lehman1989). Lehman also has the fourth trochanter centered above midline. There is a shallow cleft along the medial side of the fourth trochanter for its length. The shaft is slightly curved, concave on the medial side and convex on the lateral side. In lateral view the shaft is straight. This description of the femur fits well with Lehman, (1989), except as mentioned above. This femur, just above the fourth trochanter, has a fracture all the way around the shaft that indicates a crushing from proximal to distal. This crushing is at least 5+mm, (Figure 26, E), and shows possible partial healing. The left femur (TMM 45922-18), (Figure 28 B and C) is 290mm tall from the top of the head to the bottom of the medial condyle. From the top of the greater trochanter to the bottom of the lateral condyle it is 295mm tall. The lateromedial diameter at mid shaft is 42mm. The posteroanterior diameter is 37mm. These measurements are the same as above femur. On the proximal end it is 70mm from the medial side of the head to the lateral side of the greater trochanter, same as above. The head of this femur does not have an overhanging lip as above but it is missing part of this area. The medial condyle

39 has a small part of the posterior toe missing, but would be similar in size to the above femur, <70mm. The lateral condyle, as above, has a part missing. The distance across the distal condyle is 47mm wide compared to 55mm in above. The fourth trochanter of this specimen is centered at mid length of the shaft. Other than mentioned above this femur is identical to the above right femur. FIGURE 26; Young juvenile femurs; A), D), and E) right femur, (TMM 45922 24); B), and C) left femur, (TMM 45922 18); lt) lesser trochanter; frh) healing fracture; fro) open fracture, proximal end pushed distal into the bone; ft) fourth trochanter. SUMMATION The fossils reported herein do not resemble previously described ceratopsians from the Campanian of the Big Bend area of Texas. These fossils do resemble the well preserved adult Pentaceratops MNA-V-1747 specimen. All indications are that these ceratopsian fossils represent either Pentaceratops or a very similar new species unless there are unknown ontogenetic changes compared with below; A), previously described ceratopsians from the Big Bend Area of Texas and, B), Pentaceratops, (MNA V 1747), from New Mexico. A) Big Bend area ceratopsians, (WPA) localities of Lehman, (1989) and (RM) locality of Forster et al, (1993).

40 1) These specimens, localities (TMM 45922) and (TMM 45921), are older than the ceratopsian, (TMM 43098-1), specimen of Forster et al, (1993) and, (?), the WPA specimens of Lehman, (1989). 2) Key elements; Squamosal TMM 45922-10 is not similar to Lehman (1989), or Forster et al, (1993). They turn sharply upward after the postorbital, not down slightly then up slightly as it is in this specimen, (TMM 45922-10). There are 6 large episquamosals on all ages of the, (WPA), Lehman, (1989), specimens. There are 10 less pronounced episquamosals on the, (RM), Forster et al, (1993), specimen. There are 8 9 very small undulations on this young specimen. The two parts of the large squamosal, TMM 45921 locality, have long undulations with a sharp edge and no episquamosals. It would be difficult to reconstruct them to resemble the squamosals of the (WPA) or the (RM) element. 3) Maxilla, The poorly preserved and baked maxillas in the WPA specimens are more robust and the anterior dorsal process is not tall and thin as it is in this specimen. This element is co-ossified in the Forster et al, (1993), specimen and not of any help. 4) Quadrate, The quadrate in Lehman, (1989), has a narrow neck midway up that is not present on this specimen. There is a knob like process, (Lehman, 1989), for the quadratojugal. This specimen has no knob; rather it has a wide and flat area for the quadratojugal. The medial edge of the quadrate on this juvenile specimen underlay the lateral edge of the supraoccipital. The quadrate is obscure in Forster et al, 1993, and not of any help. 5) Horncore of this young specimen has a small cornual sinus entering into the center of its base on the anterior 1/3 of the roof of the orbit. The horncore of the WPA, Lehman, (1989), are solid. 6) Supraoccipital of this young specimen does not have the deep paired depressions for neck musculature that are very prominent in Chasmosaurus mariscalensis. B) Pentaceratops (MNA V 1747) 1) These young and adult elements are from Late Middle to Early Late Campanian which is the oldest time period for Pentaceratops. 2) Key elements, squamosal, maxilla, and quadrate are all very similar to this young specimen, TMM 45922, locality.

41 3) The squamosal of this young specimen, TMM 45922-10, is long and slim and not turned sharply up, it is more like the MNA-V-1747 Pentaceratops specimen. It also has the same number of undulations for episquamosals. 4) The maxilla, TMM 45922-27, is almost identical to adult Pentaceratops MNA-V- 1747, including the tall and very thin anterior dorsal process. 5) The quadrate, TMM 45922-28, is very similar to MNA - V -1747 in its morphology and having a wide flat area for the quadratojugal and no narrow neck. 6) The two parts of the adult squamosal, TMM 45921 locality, can be easily reconstructed to resemble the MNA - 1747 Pentaceratops specimen. 7) The supraoccipital of this young specimen does not have the prominent paired deep depressions for neck musculature; neither does adult Pentaceratops MNA-V- 1747. The lateral edge of the supraoccipital of this young specimen slightly overlay the medial edge of the quadrate that would most likely separate during ontogeny. At this time this juvenile ceratopsian skull is the only one known from the Aguja formation, (Late Middle to Early Late Campanian), Big Bend area of West, Texas. This paper should help in the cranial anatomy and ontogeny of ceratopsians. Colbert, (1961), discusses the paucity of young dinosaurs and the overwhelming number of adults described in the literature. It is evident this condition continues to the present due to the bias of collectors. This collector, (KB), on the other hand has collected many young dinosaur fossils. The only comprehensive literature on the ontogeny of the ceratopsian skull are, Goodwin et al, (2006), baby Triceratops skull, and Horner and Goodwin, (2006), baby through adult Triceratops ontogeny, are very informative. They have the privilege of having samples from stages of ontogeny from baby through adult. Their baby Triceratops skull is 38 cm long; this young ceratopsian skull is 70 cm long. At this time I am not aware of any other similar publications on any other young ceratopsians. Note: the frill is already turned up in the small juvenile Triceratops skull, (Horner and Goodwin, 2006), and the Tally Mountain juvenile ceratopsians from the Big Bend area, (Lehman, 1989). Goodwin et al, (2006), discuss at length the different suggestions about the purpose of the ornamentation, (horns and frills), on ceratopsid dinosaurs from young to adult; therefore it is not necessary that it be discussed herein. CONCLUSION

42 This juvenile ceratopsian skull from locality TMM 45922 would have to make significant ontogenetic changes, especially in the postorbital squamosal and quadrate area, to resemble any of the ceratopsians known from the Late Campanian, Big Bend area of Texas, These juvenile skull elements could make very minor ontogenetic changes to closely resemble adult Pentaceratops. MNA-V-1747 is a well preserved example of a complete Pentaceratops skull. Pentaceratops have only been recorded from NW New Mexico, except for the Williams Fork (WF) skull, (SDMNH 43470), from Colorado that has been attributed to Pentaceratops, Diem and Archibald, (2005); Lucas et al, 2006). The stratigraphic range of Pentaceratops spans the Late Campanian, (Lehman, 1993; Lucas et al, 2006). The ceratopsian described herein would be from Late Middle to Early Late Campanian and geologically older and +/- 900km south of any previously reported Pentaceratops. All indications are these specimens represent an indeterminate or new species. More young to adult associated specimens are needed for a conclusion to this problem. Acknowledgments I greatly acknowledge and thank my mentor and friend Dr. Wann Langston Jr. We sadly lost him, 4/7/2013, while this paper was being prepared. Thanks to Dr. Timothy Rowe, Texas Memorial Museum, for his help with me in getting this dinosaur fossil collection donated into Texas Memorial Museum collections. Many thanks to the Gaddis family for allowing me to excavate and study these fossils from their property and donating them to Texas Memorial Museum. Dr. David Gillett and his wife Janet, curators, at the Museum of Northern Arizona were of the greatest of help in my studding the Pentaceratops MNA-V-1747. I have enjoyed working with High School teachers Scott Clark, Larry Millar and others, from Abilene, Texas, for bringing many educational High School groups to participate in these projects since 1999. Also thanks to the many High School students with hopes they learned something. I thank Dr. Tom Lehman, Texas Tech University, for co-authoring a small paper with me on a Paleocene Champsosaurus that I recovered, and his knowledge of the Aguja Formation has been helpful. Steve Wick, Paleontological Technician, Big Bend National Park, and I have had many good discussions on the paleontology and geology of the Big Bend area. Tommy Diamond, (Fossil Preparator Technician), has helped to prepare several of the fossils in this collection. REFRENCES CITED: In alphabetic order 1) Diem and Archibald, 2005, RANGE EXTENSION OF SOUTHERN CHASMOSAURINE CERATOPSIAN DINOSAURS INTO NORTHWESTERN COLORADO, Journal Paleontology, 79(2), 2005, pp. 251-258.

43 2) Dodson and Currie, 1989, 3) Edwin H. Colbert, (Book); DINOSAURS, their discovery and their world, 1961. 4) Forster et al 1993, A COMPLETE SKULL OF CHAMOSAURUS MARISCALENSIS (DINOSAURIA: CERATOPSIDAE) FROM THE AGUJA FORMATION (LATE CAMPANIAN) OF WEST TEXAS, Journal of Vertebrate Paleontology 13(2) 161 170, June 1993 5) Forster, 1996b, 6) Goodwin, M. B., W. A. Clemens, J. R. Horner, and K. Padian. 2006. The smallest known Triceratops skull: New observations on ceratopsid cranial anatomy and ontogeny. Journal of Vertebrate Paleontology. 26(1):103:112. 7) Hatcher et al, 1907, Book, The Ceratopsia. 8) Horner and Goodwin, Proc. R. Soc. 2006, Major cranial changes during Triceratops ontogeny, Proc. R. Soc. B (2006) 273, 2757 2761doi:10.1098/rspb.2006.3643 Published online. 9) Lambe 1915, On Eoceratops canadensis gen. nov., with remarks on other genera of Cretaceous horned dinosaurs. Canadian Geological Survey, Museum Bulletin 12, Geology Series no, 24, 49pp. 10) Lehman 1989, CHASMOSAURUS MARISCALENSIS, SP.NOV., A NEW CERATOPSIAN DINOSAUR FROM TEXAS, Journal of Vertebrate Paleontology 9(2): 137 162, June 1989. 11) Lehman 1993, NEW DATA ON THE CERATOPSIAN DINOSAUR PENTACERATOPS STERNBERGII OSBORN FROM NEW MEXICO, The Paleontological Society. 12) Lucas et al 2006, Late Cretaceous vertebrates from the Western Interior. New Mexico Museum of Natural History and Science Bulletin 35. 13) Maxwell et al 1967, Geology of Big Bend National Park, Brewster County, Texas, BUREAU OF ECONOMIC GEOLOGY. 14) Osborn 1923, A NEW GENUS AND SPECIES OF CERTOPSIA FROM NEW MEXICO, PENTACERATOPS STERNBERGII, The American Museum of Natural History. 15) Sampson and Loewen, 2010, (Book), New perspectives on HORNED DINOSAURS, The Royal Tyrrell Museum Ceratopsian Symposium, Indiana University Press, 2010. 16) T. Rowe, E. Colbert, and J. Nations, 1981, The occurrence of Pentaceratops (Ornithschia: Ceratopsia) with a description of its frill, Advances in San Juan Basin Paleontology, University of New Mexico press, 1981. 17) Wick and Lehman, (2013), reported a new chasmosaurinae, Bravoceratops polyphemus, gen. et sp. nov. from the lowermost Javelina Formation, Big Bend National Park, Texas.

44

45

46