Introduction
During the Late Cretaceous, eastern and western North America were separated as the Western Interior Seaway, a shallow body of saline water, inundated the American interior. This isolated the vertebrate faunas of the west from the eastern landmass (Appalachia) formed from the continent. In contrast to the excellent, stratigraphically continuous record of terrestrial vertebrates known from the Upper Cretaceous of the American West, only a poor record exists for Appalachia (e.g., Langston, Reference Langston1960; Gallagher, Reference Gallagher1993, Reference Gallagher1997; Schwimmer et al., Reference Schwimmer, Williams, Dobie and Siesser1993; Weishampel and Young, Reference Weishampel and Young1996; Schwimmer, Reference Schwimmer, Wolberg and Stump1997, Reference Schwimmer2002; Weishampel et al., Reference Weishampel, Barrett, Coria, Loeuff, Xing, Xijin, Sahni, Gomani, Noto, Weishampel, Dodson and Osmólska2004; Carr et al., Reference Carr, Williamson and Schwimmer2005; Weishampel, Reference Weishampel2006; Brusatte et al., Reference Brusatte, Benson and Norell2011; Ebersole and King, Reference Ebersole and King2011; Longrich, Reference Longrich2016; Prieto-Márquez et al., Reference Prieto-Márquez, Erickson and Ebersole2016a, Reference Prieto-Márquez, Erickson and Ebersoleb). Appalachia seems to have harbored a distinctive endemic terrestrial vertebrate fauna, exemplified by the discovery of aberrant, relictual dinosaurs such as the large-handed Maastrichtian non-tyrannosauroid tyrannosauroid Dryptosaurus (e.g., Brusatte et al., Reference Brusatte, Benson and Norell2011) and the middle Campanian basal hadrosaurid Hadrosaurus foulkii Leidy, Reference Leidy1858 (Prieto-Márquez et al., Reference Prieto-Márquez, Weishampel and Horner2006, Reference Prieto-Márquez, Erickson and Ebersole2016a, Reference Prieto-Márquez, Erickson and Ebersoleb). However, little is known about the faunal transitions in Appalachia that took place during and following the closure of the Western Interior Seaway in the Maastrichtian. Although there is evidence that western dinosaur clades migrated eastward by the very end of this stage (Farke and Phillips, Reference Farke and Phillips2017), no distinctively western dinosaurs have been described from the Maastrichtian of the Atlantic Coastal Plain.
Reports of bones from the Atlantic Coastal Plain resembling those of lambeosaurines, a group of herbivorous dinosaurs in the nearly cosmopolitan clade Hadrosauridae that sported an array of bony head crests, have hinted at the possibility that some clearly foreign faunal components successfully migrated to the eastern coastline of North America by the Maastrichtian (e.g., Gallagher, Reference Gallagher1993, Reference Gallagher1997). However, these putative lambeosaurine occurrences have remained unsubstantiated. We describe the partial forelimb of a large hadrosaurid dinosaur from the New Egypt Formation, an uppermost Maastrichtian unit that directly underlies the K-Pg boundary in parts of New Jersey (e.g., Gallagher, Reference Gallagher1993; Miller et al., Reference Miller, Sugarman, Browning, Kominz, Olsson, Feigenson and Hernandez2004; Brusatte et al., Reference Brusatte, Benson and Norell2011). Baird and Horner (Reference Baird and Horner1976) and Gallagher (Reference Gallagher1993, Reference Gallagher1997) considered the forelimb, which was recovered from the same site as the holotypes of Dryptosaurus and “Hadrosaurus minor” Marsh, Reference Marsh1870 (e.g., Gallagher, Reference Gallagher1993, Reference Gallagher1997; Brusatte et al., Reference Brusatte, Benson and Norell2011), to be from an indeterminate lambeosaurine, but no description of the specimen was provided. If this referral is true, it would be of major biogeographic importance because the Lambeosaurinae are known exclusively from Eurasia and western North America (e.g., Weishampel et al., Reference Weishampel, Barrett, Coria, Loeuff, Xing, Xijin, Sahni, Gomani, Noto, Weishampel, Dodson and Osmólska2004; Prieto-Márquez et al., Reference Prieto-Márquez, Chiappe and Joshi2012). The forelimb shows several important similarities to lambeosaurine hadrosaurids, although the absence of synapomorphies of that clade preclude a confident referral at this time.
The partial forelimb, which consists of an associated radius and ulna, is littered with biological traces that show it was deposited along the sea floor of the Atlantic Ocean. These include circular traces identifiable as those of invertebrates, as well as deep, linear to ellipsoid grooves on the ulna characteristic of shark feeding traces on dinosaur and Mesozoic vertebrate bones (e.g., Everhart et al., Reference Everhart, Everhart and Shimada1995; Schwimmer, Reference Schwimmer, Wolberg and Stump1997; Schwimmer et al., Reference Schwimmer, Stewart and Williams1997; Everhart, Reference Everhart1999; Shimada and Everhart, Reference Shimada and Everhart2004; Shimada and Hooks, Reference Shimada and Hooks2004; Everhart and Ewell, Reference Everhart and Ewell2006; Schein and Poole, Reference Schein and Poole2014; Hill et al., Reference Hill, Roberts, Tapanila, Bouare, Sissoko and O'Leary2015). The forelimb provides another example of the bloat-and-float model for dinosaur deposition in oceans and seas, wherein the body would float out to sea and slowly decompose as it was scavenged by marine predators (e.g., Schwimmer, Reference Schwimmer, Wolberg and Stump1997, Reference Schwimmer2002; Davis and Briggs, Reference Davis and Briggs1998; Everhart and Ewell, Reference Everhart and Ewell2006; Syme and Salisbury, Reference Syme and Salisbury2014; Mallon et al., Reference Mallon, Henderson, McDonough and Loughry2018).
Geological setting
The bones described in this paper were recovered from the former site of the West Jersey Company marl pits (Figure 1), which has also yielded the remains of the tyrannosauroid Dryptosaurus aquilunguis Cope, Reference Cope1866 (e.g., Brusatte et al., Reference Brusatte, Benson and Norell2011) and small-bodied hadrosauroids (e.g., Gallagher, Reference Gallagher1993). All this material comes from the New Egypt Formation, an upper Maastrichtian unit in the Atlantic Coastal Plain that consists of dark gray and brown glauconitic clays and sands (Gallagher, Reference Gallagher1993). Quartz is also present in minor quantities (Gallagher, Reference Gallagher1993). The formation is ~35 feet thick in Monmouth County surface exposures and increases to 90 feet as it moves towards the Atlantic coast, where it comfortably overlies the Navesink Formation (Olsson, Reference Olsson1989) and gradually thins as it moves farther south (Miller, Reference Miller1956). In the upper section of the New Egypt Formation, sideritic mudstones and pockets of greensand are common (Olsson, Reference Olsson1989).
Macrofossils are diffuse in both the New Egypt and the Navesink formations, and usually consist of invertebrate material when present (Olsson, Reference Olsson1989; Gallagher, Reference Gallagher1993). The lower portion of the New Egypt Formation preserves an assemblage of marine vertebrate and invertebrate remains, including gastropods such as Lunatia, Pyropsis, and Turitella, as well as the bivalves Exogyra costata Say, Reference Say1820, Crassatellites vadose Morton, Reference Morton1834, Cucullaea vulgaris Morton, Reference Morton1830, Cardium sp., and Trachycardium (Gallagher, Reference Gallagher1993; Miller et al., Reference Miller, Sugarman, Browning, Kominz, Olsson, Feigenson and Hernandez2004). The ammonites Sphenodiscus lobatus (Tuomey, Reference Tuomey1856) and Eubaculites carinatus (Morton, Reference Morton1834) are also represented (Olsson, Reference Olsson1963). Fragments of the belemnite Belemnitella have been found in the lower New Egypt Formation, and Oleneothyris fragments are common in the upper half of the unit (Olsson Reference Olsson1963). Koch and Olsson (Reference Koch and Olsson1977) and Gallagher (Reference Gallagher1993) reported an assemblage of larger marine vertebrates, including the panchelonioid turtle Toxochelys atlantica (Zangerl, Reference Zangerl1953) and Mosasaurus maximus Cope, Reference Cope1869, from the New Egypt Formation. The small-bodied mosasaur Halisaurus platyspondylus Marsh, Reference Marsh1869 is also known from several specimens from the New Egypt Formation, including a well-preserved partial skull and jaws (Polycn and Lamb, Reference Polcyn and Lamb2012).
Material
Repositories and institutional abbreviation
YPM, Yale Peabody Museum, New Haven, CT, USA.
Systematic paleontology
Dinosauria Owen, Reference Owen1842
Ornithischia Seeley, Reference Seeley1888
Ornithopoda Marsh, Reference Marsh1881
Hadrosauroidea Cope, Reference Cope1870
Hadrosauridae Cope, Reference Cope1870
Hadrosauridae indet.
Figures 2, 3
Occurrence
West Jersey Co. Marl Pits, New Jersey, USA. The material is from the Maastrichtian (ca. 66 Ma) New Egypt Formation, which underlies the K-Pg transition (e.g., Gallagher, Reference Gallagher1993; Miller et al., Reference Miller, Sugarman, Browning, Kominz, Olsson, Feigenson and Hernandez2004).
Description
Radius
The radius (Fig. 2.1–2.4) is only known from a portion of the shaft, but was clearly elongate and straightened (Fig. 2.1, 2.4) as in most other hadrosaurids (e.g., Parks, Reference Parks1919; Pinna, Reference Pinna1979; Godefroit et al., Reference Godefroit, Zan and Jin2001, Reference Godefroit, Bolotsky and Van Itterbeeck2004, Reference Godefroit, Bolotsky and Bolotsky2012a; Prieto-Márquez et al., Reference Prieto-Márquez, Weishampel and Horner2006, Reference Prieto-Márquez, Dalla Vecchia, Gaete and Galobart2013; Evans and Reisz, Reference Evans and Reisz2007; Senter, Reference Senter2012; Prieto-Márquez, Reference Prieto-Márquez2014; Anné et al., Reference Anné, Hedrick and Schein2016; Kobayashi et al., Reference Kobayashi, Nishimura, Takasaki, Chiba, Fiorillo, Tanaka, Chinzorig, Sato and Sakurai2019). This bone is 339 mm long as preserved, with a midshaft circumference of 139 mm. The bone shaft (Fig. 2) clearly shows the radius was more elongate than in non-hadrosaurid hadrosauromorphs such as Gilmoreosaurus (Prieto-Márquez and Norell, Reference Prieto-Márquez and Norell2010), Gobihadros (Tsogtbaatar et al., Reference Tsogtbaatar, Weishampel, Evans and Watabe2019), Bactrosaurus (Gilmore, Reference Gilmore1933), or Eolambia (McDonald et al., Reference McDonald, Bird, Kirkland and Dodson2012). An elongate radius is found in most lambeosaurine dinosaurs (e.g., Brown, Reference Brown1913; Parks, Reference Parks1931; Godefroit et al., Reference Godefroit, Zan and Jin2001, Reference Godefroit, Bolotsky and Van Itterbeeck2004; Evans and Reisz, Reference Evans and Reisz2007; Senter, Reference Senter2012). Given the circumference of the bone, the length of the preserved shaft portion, which does not show proximal or distal expansion, suggests the radius was similar in robustness to the preserved ulna when complete (Figs. 2, 3). This condition contrasts with the antebrachia of most saurolophine hadrosaurids, in which the ulna is considerably more robust than the radius and widens proximally (e.g., Parks, Reference Parks1919; Pinna, Reference Pinna1979; Godefroit et al., Reference Godefroit, Bolotsky and Lauters2012b; Senter, Reference Senter2012; Prieto-Márquez, Reference Prieto-Márquez2014; Kobayashi et al., Reference Kobayashi, Nishimura, Takasaki, Chiba, Fiorillo, Tanaka, Chinzorig, Sato and Sakurai2019).
The radius is slightly more worn than the associated ulna (Figs. 2, 3), and preserves several circular and nearly circular depressions and scars along its shaft that may be attributable to biological trace-makers. It is well established in the paleontological literature that following post-mortem deposition at sea, dinosaur bones would often go through extensive biological weathering (e.g., Carpenter et al., Reference Carpenter, Dilkes and Weishampel1995; Everhart et al., Reference Everhart, Everhart and Shimada1995; Schwimmer, Reference Schwimmer, Wolberg and Stump1997; Schwimmer et al., Reference Schwimmer, Stewart and Williams1997; Everhart, Reference Everhart1999; Schein and Poole, Reference Schein and Poole2014). The traces that appear on the radius described in this contribution are identifiable as biological based on their clearly circular shape and deepened center, and are similar to “macroborings” found on other marine material (Wilson, Reference Wilson and Miller2007). These circular borings are especially concentrated along the center of the radius shaft, and indicate this surface of the bone was exposed to invertebrate trace-makers (e.g., Bromley and Heinberg, Reference Bromley and Heinberg2006). These traces also compare somewhat favorably with the circular, radially symmetrical barnacle scars present on the fossilized bones of many types of marine reptiles (e.g., Martill, Reference Martill1987; Buckeridge, Reference Buckeridge2011; Janssen et al., Reference Janssen, Baal and Schulp2013).
Ulna
The ulna (Fig. 3.1–3.6) is completely preserved and possesses a less-eroded bone surface than the radius. The ulna is elongate as in other hadrosaurids, but considerably more gracile than the corresponding element in other taxa (e.g., Brown, Reference Brown1913; Parks, Reference Parks1919, Reference Parks1922, Reference Parks1923; Pinna, Reference Pinna1979; Godefroit et al., Reference Godefroit, Zan and Jin2001, Reference Godefroit, Bolotsky and Van Itterbeeck2004, Reference Godefroit, Bolotsky and Bolotsky2012a, Reference Godefroit, Bolotsky and Lautersb; Prieto-Márquez et al., Reference Prieto-Márquez, Weishampel and Horner2006, Reference Prieto-Márquez, Dalla Vecchia, Gaete and Galobart2013; Evans and Reisz, Reference Evans and Reisz2007; Senter, Reference Senter2012; Prieto-Márquez, Reference Prieto-Márquez2014; Anné et al., Reference Anné, Hedrick and Schein2016; Kobayashi et al., Reference Kobayashi, Nishimura, Takasaki, Chiba, Fiorillo, Tanaka, Chinzorig, Sato and Sakurai2019), with the exception of many hadrosauromorphs (e.g., Gilmore, Reference Gilmore1933; Dalla Vecchia, Reference Dalla Vecchia2009; Prieto-Márquez and Norell, Reference Prieto-Márquez and Norell2010; McDonald et al., Reference McDonald, Bird, Kirkland and Dodson2012; Tsogtbaatar et al., Reference Tsogtbaatar, Weishampel, Evans and Watabe2019). The ulna gently expands proximally in anterior and posterior views. One lambeosaurine specimen from Europe is notable for having especially gracile ulnae over ten times as long as dorsoventrally wide (Conti et al., Reference Conti, Vila, Sellés, Galobart, Benton and Prieto-Márquez2020), which is the condition in YPM 3216.
This bone is 624 mm long, 107 mm wide mediolaterally, 81 mm wide anteroposteriorly at its proximal end, and 149 mm in midshaft circumference. The olecranon process is reduced, whereas the medial proximal process is heavily expanded to form a distinct, triangular expansion of bone that begins to diverge from the ulnar shaft at approximately one-third of the proximodistal run of the bone. Proximally, the medial process suddenly diverges more strongly from the main bone shaft. The divergence of the medial process from the ulnar shaft remains relatively constant in saurolophine hadrosaurids, where it is also more heavily developed (e.g., Brown, Reference Brown1913; Godefroit et al., Reference Godefroit, Bolotsky and Bolotsky2012a; Senter, Reference Senter2012; Campione, Reference Campione, Evans and Eberth2014; Prieto-Márquez, Reference Prieto-Márquez2014). The medial process does not expand more proximally than the lateral process. These processes form the borders of the articular facet for the radius, which is broadly arcuate in proximal view and produces a distinctively symmetrical, “T-shaped” proximal surface on the ulna. The “T-shaped” outline of the proximal ulna (Fig. 3.5) is less symmetrical in specimens of saurolophine hadrosaurids such as Edmontosaurus (Senter, Reference Senter2012), although the condition varies due to preservation (see illustrations in Senter, Reference Senter2012; Campione, Reference Campione, Evans and Eberth2014). The bone surface in the proximal articular facet for the radius contains numerous, developed striations that run parallel to the long axis of the ulna and form a ligament attachment site (e.g., Godefroit et al., Reference Godefroit, Bolotsky and Bolotsky2012a).
The ulnar shaft is straightened in anterior, posterior, lateral, and medial views (Fig. 3.1–3.4). This is similar to the condition in most lambeosaurinine lambeosaurines, wherein the ulna is elongate and straightened (e.g., Brown, Reference Brown1913; Godefroit et al., Reference Godefroit, Bolotsky and Lauters2012b; Fig. 4) rather than gently bowed as in saurolophines (Godefroit et al., Reference Godefroit, Bolotsky and Lauters2012b; Senter, Reference Senter2012; Campione, Reference Campione, Evans and Eberth2014; Prieto-Márquez, Reference Prieto-Márquez2014) or shortened and robust as in parasaurolophines (Parks, Reference Parks1922) and Amurosaurus (Godefroit et al., Reference Godefroit, Bolotsky and Van Itterbeeck2004). The buttressing surface for the radius on the distal ulna gently curves towards the medial margin of the bone, whereas the distal end of the ulna is slightly displaced laterally. This morphology matches the lambeosaurine condition more than the saurolophine one, where the radius and ulna are separated for most of their proximodistal run (e.g., Brown, Reference Brown1913; Senter, Reference Senter2012). The distal articular facet for the radius is slightly developed, as in other hadrosaurids (e.g., Brown, Reference Brown1913; Parks, Reference Parks1919, Reference Parks1922, Reference Parks1923; Pinna, Reference Pinna1979; Godefroit et al., Reference Godefroit, Zan and Jin2001, Reference Godefroit, Bolotsky and Van Itterbeeck2004, Reference Godefroit, Bolotsky and Bolotsky2012a, Reference Godefroit, Bolotsky and Lautersb; Prieto-Márquez et al., Reference Prieto-Márquez, Weishampel and Horner2006, Reference Prieto-Márquez, Chiappe and Joshi2012, Reference Prieto-Márquez, Dalla Vecchia, Gaete and Galobart2013; Senter, Reference Senter2012; Prieto-Márquez, Reference Prieto-Márquez2014; Anné et al., Reference Anné, Hedrick and Schein2016; Kobayashi et al., Reference Kobayashi, Nishimura, Takasaki, Chiba, Fiorillo, Tanaka, Chinzorig, Sato and Sakurai2019). In distal view, the ulna is subrectangular.
The presence of well-preserved ligament attachments on the ulna, along with the presence of smooth, original bone surface on the element, suggests this bone was minimally exposed to biotic and abiotic wear following its deposition at the bottom of the Atlantic. One intriguing feature present on the external surface of the bone is a series of nearly parallel linear to slightly ellipsoid and straight to gently arched deep scores penetrating the lateral face of the specimen. A total of nine of these can be easily seen, although more may actually be present (Fig. 2.6). Each was clearly made by an individual tooth or tooth cusp, and all scores are oriented parallel to one another. They vary very slightly in maximum length and diameter, the larger ones approaching 20 mm in the former dimension.
Previous studies have attributed closely similar grooves on other dinosaur and vertebrate fossils to sharks based on the presence of shed shark teeth in bones bearing these distinctive grooves (e.g., Everhart et al., Reference Everhart, Everhart and Shimada1995; Schwimmer, Reference Schwimmer, Wolberg and Stump1997; Shimada and Everhart, Reference Shimada and Everhart2004; Everhart and Ewell, Reference Everhart and Ewell2006; Schein and Poole, Reference Schein and Poole2014). The scores on the ulna, which are identical to definite shark feeding traces on other dinosaur bones from eastern North America (e.g., Schwimmer, Reference Schwimmer, Wolberg and Stump1997; Schwimmer et al., Reference Schwimmer, Stewart and Williams1997) do not display identifiable grooves running perpendicular to their long axes, suggesting the sharks that made the traces lacked serrated dentition and therefore were not individuals of the genus Squalicorax, which are commonly found in the Maastrichtian units of the Atlantic Coastal Plain (e.g., Gallagher, Reference Gallagher1993; Schein and Poole, Reference Schein and Poole2014). The goblin shark Scapanorhynchus texanus (Roemer, Reference Roemer1849) is also a less likely candidate for the identity of the trace-maker, given that the morphology of the grooves suggests the teeth of the trace-maker or trace-makers were blade-shaped and not tall and thin as in the mesial teeth of that taxon. Furthermore, the estimated sizes of individuals of the Scapanorhynchus and the shark Odontaspis (also known from uppermost Maastrichtian-lowermost Danian sites in New Jersey) (Gallagher et al., Reference Gallagher, Parris and Spamer1986; Gallagher, Reference Gallagher1993; Schein and Poole, Reference Schein and Poole2014) do not match with the size of the traces, which clearly came from large sharks based on their length and width. Among the species reported from the Maastrichtian of the Atlantic Coastal Plain, the taxon Cretalamna appendiculata Agassiz, Reference Agassiz1835 is considered the most likely candidate trace maker, although any precise identification is certainly tentative (see Schein and Poole, Reference Schein and Poole2014).
Material
YPM (Yale Peabody Museum) 3216, the associated complete ulna and partial radius of a large hadrosaurid dinosaur.
Morphometric analysis
In order to further assess the referral of YPM 3216 to a lambeosaurine rather than saurolophine hadrosauroid, we used a discriminant analysis assessing the similarity of the New Jersey specimen to other hadrosaurid forelimbs using the program P.A.S.T. v. 3.18 (Hammer et al., Reference Hammer, Harper and Ryan2001). A previously published dataset of hadrosauroid radius and ulna dimensions (Maidment et al., Reference Maidment, Linton, Upchurch and Barrett2012) was used for this analysis. The matrix of Maidment et al. (Reference Maidment, Linton, Upchurch and Barrett2012) was modified to include measurements of the radii and ulnae of only specimens assignable to particular hadrosauromorph taxa. We also removed the small juvenile Maiasaura specimen included in the original dataset to reduce the potentially confounding effects of ontogenetic differences in long bone dimensions on the results. The discriminant analysis of ulna proportions in members of the Hadrosauromorpha (N = 27) classified YPM 3216 with Hypacrosaurus (Fig. 5; Table 1), a Campanian–Maastrichtian age lambeosaurine from western North America (e.g., Weishampel et al., Reference Weishampel, Barrett, Coria, Loeuff, Xing, Xijin, Sahni, Gomani, Noto, Weishampel, Dodson and Osmólska2004).
Remarks
YPM 3216 is a hadrosaurid forelimb that resembles those of lambeosaurines more than saurolophines in the similar shaft circumferences of the radius and ulna, the strongly “T-shaped” outline of the proximal ulna, the close appression of the bones of the antebrachium, and the absence of a heavily developed medial process on the ulna that creates a distinct eminence along the shaft. A number of these features are apparent on the hadrosauromorph ulnae shown in Figure 3. The comparability of YPM 3216 to lambeosaurine forelimbs is also substantiated by the results of the phylogenetic and morphometric analyses conducted, which show YPM 3216 is referable to Hadrosauridae and compares most closely to lambeosaurines among hadrosaurids in the dimensions of the ulna included in that specimen.
Discussion
The results of firsthand comparisons with other hadrosauromorph forelimbs and the morphometric analysis conducted posit a hadrosaurid, and possibly lambeosaurine, identity for the radius and ulna. Previously, YPM 3216 was identified as a lambeosaurine forelimb by Gallagher (Reference Gallagher1993, Reference Gallagher1997), as were the humerus ANSP 15550 and the radius and ulna NJSM 11961. NJSM 11961 preserves extensive pathologies consistent with septic arthritis (Anné et al., Reference Anné, Hedrick and Schein2016). These pathologies heavily distort the morphology of the articular surfaces of the radius and ulna, meaning that comparison of these potentially taxonomically important areas in that specimen with the same areas on the radii and ulnae of other hadrosaurids is impossible. Therefore, we regard the identification of this specimen as the forelimb of a lambeosaurine by Gallagher (Reference Gallagher1993) as tenuous. We note that NJSM 11961 does include a clearly very elongate radius and ulna, which is closely comparable to the condition in YPM 3216. The humerus ANSP 15550 is similar to lambeosaurines in its expanded deltopectoral crest (Gallagher, Reference Gallagher1997), and was assigned to this clade by Weishampel et al. (Reference Weishampel, Sartin and Nabavizadeh2012).
Material compared with Lambeosaurus has been identified and reported from the Campanian–Maastrichtian of the Arctic in Nunavut, Canada (Vavrek et al., Reference Vavrek, Hills and Currie2014). However, most of the Nunavut hadrosaurid material is highly fragmentary, and so it is difficult to argue for its assignment to more inclusive clades than Hadrosauridae (e.g., Vavrek et al., Reference Vavrek, Hills and Currie2014). Only a single caudal centrum from the Nunavut collection has been the subject of detailed description (Vavrek et al., Reference Vavrek, Hills and Currie2014). This bone is assignable to Hadrosauridae, but Vavrek et al. (Reference Vavrek, Hills and Currie2014) did not assign it to an indeterminate lambeosaurine. Resultantly, YPM 3216 is the best-characterized possible lambeosaurine specimen from the Maastrichtian of eastern North America.
Given the lack of lambeosaurine material known from the Maastrichtian of the American West (e.g., Weishampel et al., Reference Weishampel, Barrett, Coria, Loeuff, Xing, Xijin, Sahni, Gomani, Noto, Weishampel, Dodson and Osmólska2004), the forelimb described here is also important for providing a potential record of this clade at a time when they were generally rare on the North American continent. Although Asia and Europe harbored a high diversity of lambeosaurines throughout the end of the Cretaceous (e.g., Godefroit et al., Reference Godefroit, Zan and Jin2001, Reference Godefroit, Bolotsky and Van Itterbeeck2004, Reference Godefroit, Bolotsky and Bolotsky2012a; Weishampel et al., Reference Weishampel, Barrett, Coria, Loeuff, Xing, Xijin, Sahni, Gomani, Noto, Weishampel, Dodson and Osmólska2004; Prieto-Márquez et al., Reference Prieto-Márquez, Dalla Vecchia, Gaete and Galobart2013), the record of this clade in the well-documented Maastrichtian units of the American West consists of a number of possible indeterminate specimens (e.g., Weishampel et al., Reference Weishampel, Barrett, Coria, Loeuff, Xing, Xijin, Sahni, Gomani, Noto, Weishampel, Dodson and Osmólska2004), with a notable absence of diagnostic material (Godefroit et al., Reference Godefroit, Zan and Jin2001). The derived lambeosaurine Hypacrosaurus altispinus, known from the Campanian–Maastrichtian Horseshoe Canyon Formation of Alberta (e.g., Weishampel et al., Reference Weishampel, Barrett, Coria, Loeuff, Xing, Xijin, Sahni, Gomani, Noto, Weishampel, Dodson and Osmólska2004) is the best-characterized member of this group known from the American West that survived into the last stage of the Cretaceous. The presence of possible lambeosaurine forelimb material from the Maastrichtian Atlantic Coastal Plain is therefore all the more important, suggesting that, while uncommon, this group might have persisted in the Americas up to the K-Pg boundary. The reduction in lambeosaurine diversity during the Maastrichtian corresponds to the extinction of the Centrosaurinae, the dominant clade of the horned ceratopsians, during the approximately the same period (e.g., Brown and Henderson, Reference Brown and Henderson2014). Although recognition of the Lambeosaurinae in eastern North America must wait for the recovery of more complete specimens, the forelimb described in this contribution suggests the occurrence of this group along the Maastrichtian Atlantic coastline must be considered when biogeographic scenarios for the Hadrosauridae are proposed.
The possibility that the partial hadrosaurid forelimb described in this contribution belongs to a lambeosaurine has drawn us to consider what evidence there is for cross-continental dinosaur migration in the end-Cretaceous of North America. Previously, a single fossilized tooth from the Maastrichtian Owl Creek Formation of Mississippi was referred to a derived ceratopsid ceratopsian by Farke and Phillips (Reference Farke and Phillips2017). This single element shows that large ceratopsians successfully dispersed into what would have been the eastern margin of the Western Interior Seaway by the Maastrichtian (Farke and Phillips, Reference Farke and Phillips2017). The presence of lambeosaurines on the eastern seaboard would suggest elements of western faunas were able to reach the edge of the continent, and would provide more evidence for large-scale, east-west dispersals than that given by the Mississippi ceratopsian tooth. Lambeosaurines themselves were a key component of the diverse megaherbivore assemblages during the Late Cretaceous in Asia, Europe, and western North America (e.g., Weishampel et al., Reference Weishampel, Barrett, Coria, Loeuff, Xing, Xijin, Sahni, Gomani, Noto, Weishampel, Dodson and Osmólska2004; Prieto-Márquez et al., Reference Prieto-Márquez, Dalla Vecchia, Gaete and Galobart2013), although they seem to have undergone a significant decrease in their diversity during the Maastrichtian in the American West (e.g., Weishampel et al., Reference Weishampel, Barrett, Coria, Loeuff, Xing, Xijin, Sahni, Gomani, Noto, Weishampel, Dodson and Osmólska2004) along with other dinosaur groups such as the centrosaurine ceratopsians (Brown and Henderson, Reference Brown and Henderson2014) and the albertosaurine tyrannosaurids (e.g., Brusatte and Carr, Reference Brusatte and Carr2016). Although the terrestrial fossil record of Appalachia is poor for the entirety of the Late Cretaceous, all diagnostic pre-Maastrichtian hadrosauromorph materials currently described from the landmass are from genera situated near the root of Hadrosauridae (e.g., Prieto-Márquez et al., Reference Prieto-Márquez, Weishampel and Horner2006, Reference Prieto-Márquez, Erickson and Ebersole2016a, Reference Prieto-Márquez, Erickson and Ebersoleb), and most indeterminate materials do not compare favorably enough with lambeosaurines to warrant consideration that they are potentially referable to this group (see the catalog of specimens in Gallagher, Reference Gallagher1993). Despite the relatively incomplete hadrosauromorph record from Appalachia, it seems likely that any lambeosaurine records from the latest Maastrichtian Atlantic Coastal Plain represent genuine evidence of a Late Cretaceous eastward migration of this clade.
The presence of many examples of biological traces on the radius and ulna described in this paper adds to the well-documented record of bio-eroded dinosaur bones from the American East. Previously, biological traces have been described from the remains of indeterminate theropods (e.g., Schwimmer, Reference Schwimmer2002; Brownstein, Reference Brownstein2018) and hadrosauroids (e.g., Carpenter et al., Reference Carpenter, Dilkes and Weishampel1995; Schwimmer, Reference Schwimmer, Wolberg and Stump1997; Schein and Poole, Reference Schein and Poole2014). The record described here includes some of the best examples of shark feeding traces on dinosaur bones from the Maastrichtian of North America. The bone surface of the ulna is exquisitely preserved compared to other Maastrichtian-age hadrosaurid material from the Atlantic Coastal Plain (e.g., Schein and Poole, Reference Schein and Poole2014), and shows that the bloat-and-float model of dinosaur bone preservation, wherein bones from the extremities of a bloated carcass floating on the surface of the water column would slowly come loose and fall to the sea floor, holds for some of the dinosaur material recovered from the West Jersey Marl Pit Company site. Given that the bones of Dryptosaurus (Brusatte et al., Reference Brusatte, Benson and Norell2011) and other hadrosauromorphs from the same site (Brownstein, Reference Brownstein2020) possess heavily abraded surfaces, the documentation of these definite biological traces on YPM 3216 is important for understanding the taphonomy of this unusually productive dinosaur site.
Conclusions
YPM 3216, a partial hadrosaurid forelimb consisting of a radius and ulna, is the best-characterized specimen of those potentially assignable to lambeosaurines from the eastern half of North America. The radius preserves biological traces assignable to marine invertebrates, whereas the ulna bears a multitude of deep scores identifiable as shark bite marks. The specimen supports the presence of a distinctive morphotype of long-armed, potentially lambeosaurine hadrosaurid along the latest Cretaceous Atlantic coastline.
Author contributions
C.D.B. examined and described the specimens, and wrote the introduction, results, and discussion sections, and conducted the morphometric analysis. I.B. researched the geological setting of the bones and wrote the geology section. C.D.B. and I.B. edited drafts of the paper.
Acknowledgments
We would like to thank D. Brinkman for granting C.D.B. access to the Yale Peabody Museum collections, and E. Lagos for help on performing a phylogenetic analysis included in a previous draft of this paper. Finally, we thank the editor H.-D. Sues, the associate editor D. Varricchio, and several anonymous reviewers for their comments, which greatly improved this paper.
Accessibility of supplemental data
Data available from the Dryad Digital Repository: https://doi.org/10.5061/dryad.t4b8gtj07