Panoplosaurus

Panoplosaurus is a genus of armoured dinosaur from the Late Cretaceous of Alberta, Canada. Few specimens of the genus are known, all from the middle Campanian of the Dinosaur Park Formation, roughly 76 to 75 million years ago. It was first discovered in 1917, and named in 1919 by Lawrence Lambe, named for its extensive armour, meaning "well-armoured lizard". Panoplosaurus has at times been considered the proper name for material otherwise referred to as Edmontonia, complicating its phylogenetic and ecological interpretations, at one point being considered to have existed across Alberta, New Mexico and Texas, with specimens in institutions from Canada and the United States. The skull and skeleton of Panoplosaurus are similar to its relatives, but have a few significant differences, such as the lumpy form of the skull osteoderms, a completely fused shoulder blade, and regularly shaped plates on its neck and body lacking prominent spines. It was a quadrupedal animal, roughly 5 m long and 1600 kg in weight. The skull has a short snout, with a very domed surface, and bony plates directly covering the cheek. The neck had circular groups of plates arranged around the top surface, both the forelimb and hindlimb were about the same length, and the hand may have only included three fingers. Almost the entire surface of the body was covered in plates, osteoderms and scutes of varying sizes, ranging from large elements along the skull and neck, to smaller, round bones underneath the chin and body, to small ossicles that filled in the spaces between other, larger osteoderms.

Panoplosaurus was originally classified as a stegosaur related to the similarly armoured form Ankylosaurus, a group that was later divided with ankylosaurs becoming their own group. It was then considered close to Edmontonia in the subfamily Panoplosaurinae, but then moved into a general placement in Nodosauridae. Edmontonia was for a time considered the same taxon as Panoplosaurus, making it the only nodosaur from the Campanian of North America, but this was quickly disputed and they are now considered separate. Following consistent placements in phylogenetic analyses close to Edmontonia and the American taxon Animantarx, Panoplosaurus was placed into the clade Panoplosaurini, related but not close to Nodosaurus or Struthiosaurus, which it was considered close to around when it was named. Panoplosaurus is from deposits slightly younger than Edmontonia rugosidens, and existed alongside hadrosaurids like Corythosaurus and Lambeosaurus, ceratopsids like Centrosaurus, and the tyrannosaurid Gorgosaurus, as well as other small dinosaurs like Stegoceras, Dromaeosaurus and Ornithomimus, and various fishes, amphibians, crocodiles and pterosaurs.

Discovery
In 1917, Canadian paleontologist Charles M. Sternberg of the Geological Survey of Canada discovered a complete skull and significant amount of the skeleton of an armoured dinosaur in the sandstone layers of Quarry 69 of the Belly River Group, 210 ft above sea level. The specimen, designated by the Canadian Museum of Nature accession number CMN 2759, and excavated 2.75 mi south of the mouth of the Little Sandhill tributary of the Red Deer River in Alberta, includes a nearly complete skull in articulation, most or all of the cervical vertebrae and the front dorsal vertebrae, and armour plates covering them, a majority of the disarticulated forelimb and three articulated fingers, a fragment of the pelvis and partial sacrum, a few bones of the foot, and multiple hundreds of osteoderms and dermal ossicles. This material was being described by Canadian paleontologist Lawrence M. Lambe of the Geological Society, who completed the description of the skull and osteoderms prior to his death in March of 1919. Lambe's work was published posthumously by the Royal Society of Canada, where he named the new material as Panoplosaurus mirus, and a supplemental description of the vertebrae and limb material, which Lambe had not yet gotten to, was published by Sternberg in 1921. The name Panoplosaurus derived from the hoplon of Greece, translating as "well -armoured lizard". While the beds of discovery of the holotype, CMN 2759, were originally described as the Belly River beds, they are now considered part of the upper level of the Dinosaur Park Formation, pertaining to the late middle Campanian, 75 to 76 million years ago. Many additional specimens have since been referred to the genus, including Royal Ontario Museum (ROM) 1215 and Royal Tyrrell Museum of Paleontology (RTMP) 83.25.2 from the Dinosaur Park Formation of Alberta, and an unnumbered Oklahoma Museum of Natural History (OMNH) scapulocoracoid from the Aguja Formation of Texas. A scapula from the Naashoibito member of the Kirtland Formation in New Mexico was referred to Panoplosaurus in 1981, but as it is from a different age and location from other specimens, may instead represent the ankylosaurid taxon Nodocephalosaurus, although this is uncertain due to lack of overlapping material. 18 other specimens in the ROM, CMN, American Museum of Natural History, and Princeton University were referred to the genus by Walter P. Coombs in 1978, but these were referred on the assumption that Edmontonia was a synonym of Panoplosaurus, which was considered unjustified by Kenneth Carpenter in a 1990 review on Late Cretaceous nodosaurids. The referral of ROM 1215 to Panoplosaurus was questioned by Coombs in 1990, and Roland A. Gangloff referred it and the Alaskan specimen DPMWA 90–25 to Edmontonia in 1995 based on the anatomy of the lack of diagnostic traits, but Victoria Arbour and colleagues retained ROM 1215 in Panoplosaurus in 2009, limiting the genus to it, the holotype CMN 2759, and specimen AMNH 3072, all from the upper Dinosaur Park Formation. Arbour and Philip J. Currie further restricted Panoplosaurus to just the holotype in 2016, as a result of uncertainties surrounding the referrals of various specimens between it, Edmontonia rugosidens and Edmontonia longiceps, all from the Dinosaur Park Formation and similar deposits.

Description
Panoplosaurus was a rather large animal at 1600 kg, a comparable size to other ankylosaurs from the same location, and heavier than or approximately equivalent to all predators it coexisted with. Panoplosaurus was about 5 m long, and was armoured like most ankylosaurs, but lacked prominent spikes anywhere on its body. The characteristics of its armour are one of the features that distinguishes Panoplosaurus from its relative and contemporary Edmontonia, along with a tapering snout, lumpy cranial armour, a swollen vomer bone in the palate, tall neural arches and neural spines in the vertebrae, a small and round coracoid fused to the scapula, and a hand that may have only had three digits. There is also a prominent armour plate covering the cheek in the type specimen of Panoplosaurus, which may be unique feature of the taxon, or individual, depending on what additional skulls are referred to P. mirus.

Skull
The skull of Panoplosaurus is broad and depressed, narrowing towards the end of the blunt snout to form a triangular shape. At a total length of 335 mm in a straight line, the skull is curved in a way that across the the same length is 530 mm. Behind the, the skull reaches its maximum width of 295 mm, giving it a very short and broad appearance. As other nodosaurid specimens from the same location that may be referred to Panoplosaurus lack the distinctively short snout of the holotype, it may be that its unique appearance is due to it being a different age or sex than other individuals. The orbits of Panoplosaurus are very small, and placed far from the snout, which is very squared and bearing laterally open nostrils. Bony plates are present across the entire exterior of the skull, including large osteoderms on all the sides of the skull and mandible, small scutes within the nostrils, and underneath the head in the form of an aggregation of small ossicles covering the surface between the two lower jaws. The, where the skull articulates with the neck, is short and thick, facing nearly directly downwards, which would have meant the head was held with the snout down in life, about 20 degrees below the horizontal. Unlike in Edmontonia, the groove separating cranial osteoderms in Panoplosaurus never disappear, which show that there is a unique narrow scute across the entire rear of the skull.

Due to fusion and the covering of osteoderms, the only individual bone of the 310 mm long mandible than can be identified is the. The rami diverge strongly towards the read of the skull, where they curve inwards at the jaw articulation, and towards the front they bend inwards slightly where the predentary articulates with the, with the thin predentaries meeting at the midline of the jaw. The mandibles are deepest near the rear of the skull, approximately maintaining their depth along the tooth row before narrowing sharply at the front where the predentaries are. Predentaries, which make up the lower portion of the snout, are somewhat horseshoe-shaped, form a sharp beak that fits within the overhang of the in the upper jaw. Both the predentaries and their premaxillary counterparts lack any teeth, a derived feature among nodosaurids where premaxillary teeth are sometimes present. The teeth of the mandible are hidden by armour on the right side, and on the left side of the skull where the cheek plate is not in place, the mandibular teeth are hidden by the teeth of the in the upper jaw. Eight maxillary teeth are preserved, and though it is not certain that was the full tooth count there is not room for many more in the jaw. All the teeth are similar to those of Edmontonia and Palaeoscincus, with a mild expansion of the crown above the root, and formed by prominent ridges on both the front and rear edges of the crown, though there are more on the front than rear edge of the tooth. There are not significant differences between the teeth of Panoplosaurus and those of other armoured dinosaurs.

Postcranial skeleton
The number of vertebrae in Panoplosaurus is unknown, as the dorsal and caudal series' are incomplete, and the cervical vertebrae are obscured by armour. The first three cervicals after the skull are fused together into one single ossification, a feature unique to Panoplosaurus and Edmontonia among ankylosaurs. Following the fused atlas-axis-third cervical complex, the next three cervicals are all identifiable, but very morphologically conservative, without significant differences along the series, all vertebrae being. Dorsal vertebrae are poorly preserved where present, but the, but the synsacrum, as in other nodosaurids, is partially complete. Sternberg identified that the ilium of Panoplosaurus was supported by at least six vertebrae in 1921, which he identified as a dorsosacral, four true sacral vertebrae, and a caudosacral. The identification of four sacrals was questioned by Carpenter in 1990, as Edmontonia only has three true sacrals, but he was not able to definitively identify the first sacral of Sternberg as a dorsosacral. The neural spines of all four sacrals are fused into a single plate-like process. Few caudals are known in Panoplosaurus, but where preserved they are very similar to the corresponding elements of Ankylosaurus.

In Panoplosaurus the scapula and coracoid are completely fused together, with the only indication of the bone separation being a slight thickening along what would be the suture. The scapular region of the bone is relatively short, but is concave following the curvature of the body and curves down towards its distal end. Because of the curvature of the blade, the coracoid in Panoplosaurus sat at the front of the chest, as in Stegosaurus, Triceratops and Hadrosaurus. The top margin of the scapula forms a shelf as it approaches the coracoid, terminating in a prominent that projects outwards from the animal, and directly overhangs a large  attachment area for the deltoid muscle. The scapula itself is 410 mm long, and the coracoid is 255 mm. The complete fusion of the scapulocoracoid is unique to Panoplosaurus. The humerus is a robust, 430 mm long bone, with the shape and projection of the head suggesting the upper arm of Panoplosaurus was held in a flexed position in life. The only other portions of the forelimb known are three well-preserved, articulated digits of the hand, which may represent the complete hand suggesting Panoplosaurus had fewer fingers than its relatives. The was splayed, with two  on the first digit, and three on the second and third. All bones were hoof-shaped, lacking a point and bearing a flat bottom.

The only portions of the pelvis and hindlimb that are known in Panoplosaurus are the tibia, fibula, and a few isolated bones from the foot, the remainder being unpreserved. The right tibia and fibula were found still in articulation, with the 385 mm tibia being noticeably longer than the 310 mm fibula. As in ankylosaurids, the tibia of Panoplosaurus was shorter than the humerus, which contrasts with Stegosaurus and basal taxa like Scelidosaurus. This suggests that the forelimb of Panoplosaurus was only slightly shorter than the forelimbs. Both the metatarsal and phalanges of the foot resemble the corresponding elements of the hand, but are slightly larger and more robust. Though the tibia and fibula of Panoplosaurus are crushed, they appear to be straighter than the elements in Edmontonia, which are curved along their length.

Armour
Armour from the neck and trunk of Panoplosaurus, some of it still in articulation with bones, is known. A gradient of sizes exist from plate-like paired elements through to indistinct ossicles. Where they have a distinct shape, the osteoderms are keeled, with the strength of the keel dependent on location. Lambe identified 7 different categories of osteoderms in his 1919 description of Panoplosaurus. The first kind was large, paired elements with a low keel, which formed bands around the neck leading from the head. On the sides of the neck to the back was a second kind, individual elements that were slightly smaller, suboval, and had a strong keel. Small, keeled scutes with a thick base were identified as the fourth osteoderm type, occurring on the underside of the base of the neck forwards to the chin. A fifth kind of osteoderms was identified as small, polygonal elements that fit together along the underside of Panoplosaurus, slowly grading into the larger rectangular elements of category two on the sides. Small irregular scutes lacking a keel were identified as a sixth form, and were suggested to have been from the limbs, though this was not definitive. The final form of scute were small ossicles, which occurred all over the animal filling in gaps between the larger osteoderms. The scutes along the top and sides of the neck in Panoplosaurus are the most distinct form, differing significantly from the corresponding elements in Edmontonia. Three bands of cervical osteoderms were present in both genera, consisting of rounder plates that united on the midline of the animal, and one narrower element on each side with a sharp keel. In Panoplosaurus both the first and second bands of neck osteoderms had a third pair, lower on the side of the animal, again possessing a sharper keel than the elements on the top of the neck. While Edmontonia possesses lateral spines on the rear neck and shoulders, these are absent in Panoplosaurus. The arrangement of the armour on the torso and tail of Panoplosaurus is unknown, as no elements were found in articulation or association with this region of the skeleton.

Classification
Panoplosaurus was originally named simply as a genus of armoured dinosaur by Lambe in 1919, within the group Stegosauria, though it was subsequently specified by Sternberg in 1921 that within Stegosauria it was closest to taxa like Ankylosaurus and Euoplocephalus, and so within Ankylosauridae. The description of the new genus Edmontonia in 1928 by Sternberg identified it as a very close relative of Panoplosaurus, almost close enough to be considered the same genus, both being closely related to Palaeoscincus. Following further description of the armoured taxon Struthiosaurus by Franz Nopcsa in 1929, a different classification war proposed. Thyreophora replaced Stegosauria as the clade encompassing multiple families, with Panoplosaurus falling alongside Edmontonia, Dyoplosaurus, Hierosaurus, Scolosaurus and Stegopelta within the new subfamily Panoplosaurinae, which formed the family Nodosauridae with Nodosaurinae, Ankylosaurinae and Acanthopholinae. Nopcsa's classification of Panoplosaurinae was amended slightly by Charles W. Gilmore in 1930, who moved Palaeoscincus into the subfamily and removed Dyoplosaurus, which was discovered to have the skull of an ankylosaurine. Panoplosaurus was then placed in the new subfamily Edmontoniinae by Loris S. Russell in 1940, along with Edmontonia and Palaeoscincus, an equivalent classification to Gilmore's where Panoplosaurinae was used. Disagreeing with the classifications based on the work of Nopcsa, Evgeny Maleev moved Panoplosaurus into the family Ankylosauridae, which he considered to contain all ankylosaurs except Syrmosaurus. Coombs reviewed and revised the classifications of Ankylosauria in 1978, which he used as the group to encompass all heavily armoured ornithischians in a similar usage to Nodosauridae of Nopcsa. Within this suborder, he divided all taxa into two families, Ankylosauridae and Nodosauridae, distinguished by cranial and postcranial anatomy. Within Nodosauridae, which he considered a senior synonym of Acanthopholinae, Panoplosaurinae, Edmontoniinae, and other equivalent names, Coombs synonymized Panoplosaurus with Edmontonia, making Panoplosaurus the only Late Cretaceous nodosaurid from North America. Panoplosaurinae was then revived in 1986 by Paul Sereno, who used it to unite Panoplosaurus, Sauropelta, Silvisaurus and Struthiosaurus within Nodosauridae, excluding Hylaeosaurus and Polacanthus. Robert Bakker then redescribed specimens from the Late Cretaceous of North America in 1988, prompting him to separate Edmontonia rugosidens into the new subgenus Chassternbergia, remove a specimen from Edmontonia longiceps making it the new genus Denversaurus, and identify a former specimen of Panoplosaurus mirus as P. sp.. To distinguish the Panoplosaurus species from the Edmontonia complex, Bakker placed Panoplosaurus in Panoplosaurinae, and Edmontonia, Chassternbergia and Denversaurus in Edmontoniinae, uniting the two subfamilies into the family Edmontoniidae, which he placed in Nodosauroidea, the family Nodosauridae of Coombs' use elevated to superfamily rank. Carpenter further reviewed Late Cretaceous North American nodosaurids in 1990, concluding that Bakker's justification for separation was lacking; placing Chassternbergia and Denversaurus as junior synonyms of Edmontonia, and only narrowing down the classification of Panoplosaurus and Edmontonia to Nodosauridae. In 1998, Sereno modified his earlier position, using the name Nodosaurinae for the group of derived nodosaurids including Panoplosaurus, and informally defining the subfamily as all taxa closer to Panoplosaurus than Sarcolestes or Hylaeosaurus. Tracy Ford published a novel classification of Panoplosaurus in 2000, using osteoderm anatomy to divide Nodosauridae into Nodosaurinae including only Nodosaurus, Sauropeltinae including Sauropelta and Silvisaurus, Edmontoniinae including Edmontonia and a distinct Chassternbergia, and Panoplosaurinae including only Panoplosaurus.

The first phylogenetic analysis to include Panoplosaurus was that of Yuong-Nam Lee in 1996, where Panoplosaurus mirus, including both CMN 2759 and ROM 1215 nested next to Edmontonia, which included both E. longiceps and E. rugosidens, to the exclusion of all other nodosaurids. A similar union of Panoplosaurus and Edmontonia was recovered in 1998 by Jim Kirkland and colleagues, but the 2001 analysis of Carpenter instead recovered Edmontonia closest to Animantarx, while Panoplosaurus nested next to a clade of Sauropelta and Silvisaurus. The 2003 analysis of Robert Hill and colleagues resolved a group of Edmontonia and Animantarx, with Panoplosaurus as the sister taxon, and the analysis of Matthew Vickaryous et al. recovered Panoplosaurus and Edmontonia as sisters, though Animantarx was not analysed. Expansions on the analysis of Vickaryous and colleagues resolved the same clade of Panoplosaurus and Edmontonia with the additions of the nodosaurid Hungarosaurus, the ankylosaurid Crichtonsaurus benxiensis, the nodosaurid Struthiosaurus, and the supposed ankylosaurid Tatankacephalus. The next novel analysis was that of Richard Thompson and colleagues in 2012, combining previous ankylosaurian analyses into a single one to analyse both Nodosauridae and Ankylosauridae. Panoplosaurus was resolved next to Edmontonia, deep within an unresolved polytomy of all nodosaurids more derived than Animantarx, which included Niobrarasaurus, Nodosaurus, Pawpawsaurus, Sauropelta, Silvisaurus, Stegopelta, and Texasetes. In 2016, the phylogenetic analysis of Arbour and Currie initially meant to test the relationships of Ankylosauridae was expanded to include many of the nodosaurids known at the time, with Panoplosaurus limited to the holotype due to a lack of consensus about referred specimens. Following further modifications and expansions, Panoplosaurus was continually found within a group including Edmontonia and also at times Animantarx, Texasetes, Denversaurus (if considered separate from Edmontonia) and ‘’Patagopelta’’. As a result of this consistent support, Daniel Madzia and colleagues decided to name the clade uniting all taxa closer to Panoplosaurus than Nodosaurus or Struthiosaurus, giving it the formal name Panoplosaurini, modifying the suffix -inae from previous uses as it was continually nested within the clade Nodosaurinae. The reference phylogeny for Panoplosaurini designated by Madzia and colleagues was that of Rivera-Sylva et al. (2018), which is a modified version of the Arbour and Currie analysis expanded to include the Mexican taxon Acantholipan. Their results can be seen below.

Feeding
Ankylosaurs were traditionally viewed as having a generic method of feeding due to their simple teeth, stiff skulls, and unspecialized musculature, comparable to modern iguanids. However, unlike iguanids ankylosaurs frequently have significant tooth wear, sometimes all the way down to the base of the crown. Nodosaurids normally have oblique wear facets on teeth, in Panoplosaurus having a single wear facet per tooth which changes angulation across the tooth row, going from sub-vertical at the snout to nearly horizontal at the back of the mouth. This differs from ankylosaurids, but may be due to the alignment of teeth between the upper and lower jaws, instead of a difference in form of mastication. While the shape of teeth in Panoplosaurus and other ankylosaurs suggests a simple, soft food diet, their variability implies a less restrictive feeding strategy. Nodosaurid teeth are more blade-shaped than those of ankylosaurids, a possible evolutionary response to a tougher, fibrous diet. However, microwear on the teeth of ankylosaurids and Panoplosaurus instead show that their diets were not significantly different, with regular pitting and abrasion showing a diet consisting occasionally of fruit as well as more abundant foliage. Stomach contents of the closely related taxon Borealopelta were identified amongst gastroliths, showing that at least it had a diet regularly consisting of almost 85% fern material, along with 3.7% cycad matter, trace elements of conifers, and 11.4% undiagnostic plant remains. Gastroliths may have been found with Panoplosaurus as well, but their identification is uncertain as they were not originally mentioned among the material found as part of the specimen.

Airways and senses
While nodosaurids were traditionally thought to have had simply sinuses, lacking complicated cavities and paranasal sinuses. While this can be seen in some taxa like Edmontonia, the nasal system of Panoplosaurus is far more complex than previously thought. The complete nasal passage of Panoplosaurus undergoes two complete 360 degree loops in different planes along its length, before entering the olfactory recess for scent processing. However, while the shape of the nasal passages is more complicated, Panoplosaurus does lack the additional parallel sinus tracts that can be found in ankylosaurids like Euoplocephalus. It is possible that the function of these complicated sinuses was to warm incoming air as it passed through the skull. Inflowing air in Panoplosaurus was simulated to undergo a raise of between 17.9 and 18.2 C, primarily in the elongate nasal passage, while saturating the air with moisture. This is less heat efficient than the more complicated sinuses of Euoplocephalus, but still shows that the sinus cavities of Panoplosaurus increased the recovery of lost heat and moisture by over 60%.

The brain of Panoplosaurus takes up 33% of the length of the skull, similar to the nodosaurid Pawpawsaurus where the value is 30%, a higher value than in ankylosaurids. Panoplosaurus had a similar sense of smell to both Pawpawsaurus and Euoplocephalus, with the ratio between the length of the olfactory bulb and cerebral hemisphere being 44.0, 46.2 and 54.0 respectively. However, the size of the region of the brain devoted to the sense of smell is smaller in Panoplosaurus than expected for an animal of its size.

Paleoecology
The Dinosaur Park Formation deposits are a 70 m layer of sediments exposed in the badlands of Alberta, lying on top of the older Oldman Formation and underneath the younger Bearpaw Formation. Panoplosaurus was found about 20 to 30 m above boundary with the Oldman Formation, which is 76.5 million years old, within the "sandy zone" of the formation deposited by alluvial paleochannels. All the deposits of the Dinosaur Park Formation were lain during the beginning of the last significant sea level rise of the Western Interior Seaway. Sediments deposited during the formation show that the alluvial channels were flowing to the east or southeast, towards the Bearpaw Sea of the Western Interior Seaway, creating meandering paths as the flow of water changed down and up during tidal cycles across the floodplain was balancing out. The climate of the environment in the Cretaceous was much different than in present day, being warmer and more temperate, with wet and dry seasons allowing for a large variety of flora and fauna. Periodic areas of standing water would occur in the low-lying plains. Where there was dense vegetation, conifers were the dominant plants, with an understory composed of ferns and tree ferns, and flowering herbs and shrubs. The forest floor was covered by decaying vegetative matter, small ferns, lycopods and angiosperms, mosses, lichens, and fungi, with plentiful algae where there was standing water.

The constant presence of water in the Dinosaur Park Formation led many forms of freshwater or marine animals to enter the otherwise predominantly terrestrial ecosystem. In the lower Dinosaur Park Formation, assemblages of crevasse sites show that mollusks were commonly dominated by the freshwater clam Sphaerium, which occurred with abundant gastropods of the genera Goniobasis and Lioplacodes. This sphaeriid-dominated diversity was present during all of the duration of the formation and its over- and underlying deposits, but the mussel and brackish-water assemblages that can be seen in older and younger layers are absent in the range of 10 to 35 m above the Dinosaur Park Formation-Oldman Formation contact. A variety of forms of fish are present in the fluvial beds of the Dinosaur Park Formation, including chondrichthyans, teleosts and other ray-finned fishes. The ray Myledaphus is characteristic of the formation and similar deposits, and lived alongside the less common shark Hybodus montanensis, intermediate paddlefish and sturgeons, the long, slender fish Belonostomus, the gar Lepisosteus, bowfins, and an abundance of small teleosts including Paratarpon and Cretophareodus. At least nine forms of amphibians were present in the Dinosaur Park Formation, including the salamander-like Albanerpeton, two unnamed forms of frog, and salamanders from the genera Scapherpeton, Lisserpeton, Opisthotriton, Habrosaurus, and two or three unnamed forms. Turtles are commonly well-preserved in the formation, with many taxa known from many specimens. Three genera in the primitive turtle family Baenidae are known, Plesiobaena, Boremys and Neurankylus. The low-domed macrobaenid Judithemys was named for a nearly complete skeleton and skull from the deposits, an unnamed but new form of snapping turtle is known, Adocus and Basilemys represent the diversity of primitive trionychoids, and the softshelled turtles Apalone, Aspideretoides, and a new form, are also present. Fragmentary material from the choristodere Cteniogenys, and many well-preserved skulls and skeletons of its relative Champsosaurus, is present among the fauna of the Dinosaur Park Formation, representing an extinct group of semi-aquatic animals with crocodilian features that spanned the globe for much of the Mesozoic and Cenozoic. Mosasaurs and plesiosaurs are both known from the Dinosaur Park Formation, though mosasaurs are only represented by the teeth of Plioplatecarpus from the upper deposits, and apart from the relatively complete elasmosaurid Fluvionectes, plesiosaurs are only known from isolated and incomplete elements. Two or three true crocodilians are known, including the named genera Leidyosuchus and Albertochampsa. A rich and diverse vertebrate assemblage is known from the Dinosaur Park Formation, with the lower region, excluding the Lethbridge Coal Zone, being formed by terrestrial and coastal deposits. Microfossil sites are common, but a taphonomic bias is present in the formation towards the better preservation of large-bodied animals like Panoplosaurus, which is 63% complete. Six small lizards are known, representing five different families, with the teiids Socognathus and Glyptogenys, the xenosaurid Exostinus, the helodermatid Labrodioctes, the necrosaurid Parasaniwa, and the varanid Palaeosaniwa. No snakes are known, which, while it contrasts with the modern diversity of the group, does correspond with the rarity of the group during the Cretaceous. Two pterosaurs have been found, the azhdarchid Cryodrakon and an unnamed pterodactyloid. All three major types of mammal present in the Cretaceous have been found in the Dinosaur Park Formation, Multituberculata, Marsupialia and Placentalia. The named multituberculates include Cimexomys, Cimolodon, Cimolomys, Meniscoessus, and Mesodma. Marsupials are represented by the taxa Alphadon, Eodelphis, Pediomys, and Turgidodon, and placentals by Cimolestes, Gypsonictops, and Paranyctoides, though all the material represents isolated partial skull or skeletal bones, or isolated teeth. While no other dinosaurs were recovered at the same sites as any Panoplosaurus specimens, correlation and comparisons of localities throughout the Dinosaur Park Formation show it coexisted with a large variety of animals. The herbivorous fauna of the Dinosaur Park Formation can be divided into two Megafaunal Assemblage Zones, defined as the lowest 28 m directly above the Oldman Formation, and the deposits above those, where the common ceratopsid Centrosaurus and hadrosaurid Parasaurolophus are absent. Panoplosaurus, which was found 24.9 m above the contact, is from layers just younger than those bearing Edmontonia, but still corresponding to the upper portion of the older MAZ-1, and would have coexisted with the ankylosaurid Euoplocephalus, ceratopsids Centrosaurus and Chasmosaurus belli, and hadrosaurids Corythosaurus intermedius, Lambeosaurus lambei, Lambeosaurus clavinitialis, and Parasaurolophus. As well as herbivorous megafauna, unnamed ornithopods, the primitive ceratopsian Unescoceratops and the pachycephalosaurians Stegoceras validum, Stegoceras sternbergi and Stegoceras brevis are known from the Dinosaur Park Formation, though their stratigraphic correlations are uncertain. Among theropods, the dromaeosaurids Dromaeosaurus and Saurornitholestes are known, the former from a single specimen of uncertain provenance, and the latter from many specimens such as teeth throughout the entire formation, and troodontids from the MAZ-1 are limited to Stenonychosaurus. Three genera of oviraptorosaurs are known, all of which likely coexisted in the Dinosaur Park Formation, Caenagnathus, Chirostenotes and Citipes. It is possible that a therizinosaur was present based on a single frontal bone, and three ornithomimosaurs are preserved in the MAZ-1, Rativates, Ornithomimus, and Struthiomimus. The only large theropod to coexist with Panoplosaurus was the tyrannosaurid Gorgosaurus, although an unnamed species of Daspletosaurus was present in deposits slightly younger than known for Panoplosaurus.