Introduction
The earliest record of glyptosaurines, an extinct group of armored lizards, is from the middle Paleocene (Torrejonian NALMA) of the United States, but the family achieved holarctic distribution during the Eocene (Meszoely et al., Reference Meszoely, Estes and Haubold1978; Sullivan, Reference Sullivan1979; Estes, Reference Estes1983; Holman,Reference Holman1998; Augé, Reference Augé2005; Smith, Reference Smith2009) and ultimately became extinct in the Oligocene (Augé and Smith, Reference Augé and Smith2009; Rage and Augé, Reference Rage and Augé2010). Purported post-Oligocene glyptosaur remains (Estes, Reference Estes1963; Holman, Reference Holman1976; Albright, Reference Albright1994) are considered to have been misidentified or reworked from older deposits (Wellstead, Reference Wellstead1982; Sullivan and Holman, Reference Sullivan and Holman1996). Outside of the United States, glyptosaur fossils have been found on Ellesmere Island in the Canadian high arctic (Estes and Hutchinson, Reference Estes and Hutchinson1980) and in Belgium (Sullivan et al., Reference Sullivan, Augé, Wille and Smith2012), England (Holman, Reference Holman1998; Klembara and Green, Reference Klembara and Green2010), France (Holman, Reference Holman1998; Augé, Reference Augé2003; Augé and Sullivan, Reference Augé and Sullivan2006), Germany (Sullivan, Reference Sullivan1979; Augé and Sullivan, Reference Augé and Sullivan2006), Spain (Bolet and Evans, Reference Bolet and Evans2013), Portugal (Rage and Augé, Reference Rage and Augé2003), Switzerland (Hoffstetter, Reference Hoffstetter1962), and Mongolia (Sullivan, Reference Sullivan1979).
In the United States, the fossil record shows that glyptosaurs were taxonomically diverse and geographically widespread west of the Missouri River. These lizards have been reported from Colorado (Sullivan, Reference Sullivan1989), Montana (Estes, Reference Estes1976), Nebraska (Sullivan, Reference Sullivan1974), North Dakota (Smith, Reference Smith2006), South Dakota (Lillegraven, Reference Lillegraven1970; Maddox and Wall, Reference Maddox and Wall1998), New Mexico (Sullivan, Reference Sullivan1981; Lucas et al., Reference Lucas, Sullivan and Logan1983; Sullivan and Lucas, Reference Sullivan and Lucas1988), Texas (Westgate, Reference Westgate1989), and especially Wyoming (Gauthier, Reference Gauthier1982; Sullivan, Reference Sullivan1982; Bartels, Reference Bartels1983; Gunnell et al., Reference Gunnell, Bartels, Gingerich and Torres1992; Zonneveld et al., Reference Zonneveld, Gunnell and Bartels2000; Smith, Reference Smith2009). To our knowledge, only one definitive glyptosaur occurrence has been reported from east of the Missouri River (Westgate, Reference Westgate1989; Estes, Reference Estes1963 vs. Sullivan and Holman, Reference Sullivan and Holman1996). Herein, we describe isolated glyptosaur osteoderms that were collected from an active limestone quarry in Berkeley County, South Carolina (Fig. 1). In addition, we comment on the age and paleoenvironment of the fossil deposit and discuss the paleobiogeographical significance of the fossils.
Paleogeographic map of the eastern United States showing the reconstructed late Thanetian (late Paleocene) shoreline. White area represents exposed landmass; gray portion is ocean. Collecting site is indicated by a closed circle. The estimated shoreline is based on stratigraphic correlations of time-equivalent shallow marine deposits identified from geologic maps and subsurface data, with some structural corrections. See the Methods section of this report for references used to reconstruct this shoreline.
Methods
The osteoderms were recovered from unconsolidated clay-rich, glauconitic, fine- to medium-grained quartz sand occurring in a Martin Marietta Aggregates limestone quarry located to the southeast of Jamestown in northern Berkeley County, South Carolina (Fig. 1). The fossiliferous deposit, which underlies the middle Eocene Santee Limestone, is below water level in the quarry and so is not directly available for observation. However, sediment is occasionally brought to the surface from the water-filled, actively mined portion of the quarry. We collected bulk matrix from spoil piles located directly adjacent to their excavation site and recovered vertebrate fossils by screen washing sediment in the laboratory down to 0.25 mm (#60 USA Standard Testing Sieve).
A calcareous nannofossil smear slide was made from some of this bulk matrix by spreading a thin veneer of raw material onto a coverslip and affixing this to a glass slide using Norland Optical Adhesive 61. The slide was scanned under cross-polarized light using a Zeiss Axioplan 2 petrographic microscope at x1250 magnification. Calcareous nannofossils were identified to the species level and assigned to the biozonation scheme of Martini (Reference Martini1971), as augmented by Aubry (Reference Aubry1999), and correlated to the timescale of Vandenberghe et al. (Reference Vandenberghe, Hilgen and Speijer2012). Calcareous nannofossils are common and exhibit good preservation throughout.
To create the regional late Paleocene shoreline shown in Figure 1, state geologic map, core, lithostratigraphic, and biostratigraphic data from Alabama, Georgia, South Carolina, North Carolina, Virginia, Maryland, Delaware, and New Jersey were compiled. Geospatial data for Alabama and Georgia are based on Lawton et al. (Reference Lawton1976) and Szabo et al. (Reference Szabo, Osborne, Copeland and Neathery1988). Information for the Atlantic Coastal Plain (Georgia to New Jersey) includes data from Gildersleeve (Reference Gildersleeve1942), Gohn et al. (Reference Gohn, Higgins, Smith and Owens1977), Hazel et al. (Reference Hazel1977), Scholle (Reference Scholle1979), Van Niewenhuise (Reference Van Niewenhuise1982), Houser (Reference Houser1985), Ward (Reference Ward1986), Colquhoun and Muthig (Reference Colquhoun and Muthig1991), Harris and Zullo (Reference Harris and Zullo1991), Nystrom et al. (Reference Nystrom, Willoughby and Price1991), Harris et al. (Reference Harris, Zullo and Laws1993), Miller et al. (Reference Miller, Browning, Pekar and Sugarman1997), Baldwin et al. (Reference Baldwin, Morton, Denny, Dadisman, Schwab, Gayes and Driscoll2004), Weems and Edwards (Reference Weems and Edwards2007), and Weems et al. (Reference Weems, Schindler and Lewis2010). Geologic map data for the Atlantic Coastal Plain states are based on Spoljaric and Jordan (Reference Spoljaric and Jordan1966), Cleaves et al. (Reference Cleaves, Edwards and Glaser1968), Lawton et al. (Reference Lawton1976), Rhodes and Conrad (Reference Rhodes and Conrad1985), Virginia Division of Mineral Resources (1993), Drake et al. (Reference Drake, Volkert, Monteverde, Herman, Houghton, Parker and Dalton1996), Horton and Dicken (Reference Horton and Dicken2001), and Newell et al. (Reference Newell, Prowell, Krantz, Powars, Mixon, Weems, Stone and Willard2007).
Geospatial geologic data from these sources were compiled within ArcGIS 10.1 geographic information systems software for spatial analyses and correlations. From the data compiled, the most age-equivalent units (most temporally equivalent to the fossiliferous deposit discussed herein) were identified. An estimated paleoshoreline was created by digitizing a best-fit line from the inland-most edge of the identified shallow marine deposits. For areas such as the Albemarle Embayment and the Cape Fear Arch, corrections were made for post-depositional tectonic uplift or downwarping that may have affected vertical-lateral placement. Deltas, rivers, and barrier islands are not included because more detailed stratigraphic data are needed to complete an accurate placement of these coastal features.
Systematic paleontology
Suborder Anguimorpha Fürbinger, Reference Fürbinger1900
Family Anguidae Gray, Reference Gray1825
Subfamily Glyptosaurinae McDowel and Bogert, Reference McDowel and Bogert1954
Genus and species indeterminate
Material examined
Specimens discussed in this report are housed at the South Carolina State Museum (SC) in Columbia. SC 2014.30.1, incomplete body osteoderm (Fig. 2.1); SC 2014.30.2, partial body osteoderm (Fig. 2.2); SC 2014.30.3, partial polygonal osteoderm (Fig. 2.3); SC 2014.30.4, partial body osteoderm (Fig. 2.4).
Late Paleocene glyptosaur osteoderms from Berkeley County, South Carolina. (1) Body osteoderm in (1a) external, (1b) internal, and (1c) profile views, SC 2014.30.1. (2) Body osteoderm in (2a) external, (2b) internal, and (2c) profile views, SC 2014.30.2. (3) Cephalic (?) osteoderm in external view, SC 2014.30.3. (4) Body osteoderm in external view, SC 2014.30.4. In (1a, 1b, 2a, 2b, 4) anterior margin of osteoderms at top. In (1c, 2c) anterior margin of osteoderms at left. In (3), scale bar=0.5 mm; in (1, 2, 4) scale bars=1 mm.
Occurrence
Late Thanetian (late Paleocene, subzone NP9a), Chicora Member of the Williamsburg Formation. Martin Marietta Aggregates quarry, approximately 4.8 km southeast of Jamestown (33.279496 N; -79.64374 W), Berkeley County, South Carolina.
Description
The most complete osteoderm in our sample is shown in Figure 2.1. It lacks the anterior edge, and some of the posterior end is missing. The osteoderm has a slightly curved rectangular outline and measures 5.2 mm long (as preserved) and slightly over 3 mm in width. The bone is thickest near the anterior end (0.8 mm) but tapers posteriorly, resulting in a wedge shape in lateral view. The anterior-most 1.5 mm of the exterior surface is smooth (gliding surface), with the remaining portion being ornamented with numerous tubercles and bearing an indistinct medial longitudinal ridge that extends to the broken posterior margin. The tubercles appear to be arranged in longitudinal rows that are subparallel to the medial ridge. Lateral surfaces are both beveled, and the bevels are somewhat concave and weakly rugose. The interior surface is smooth and concave. Both exterior and interior surfaces are perforated by several tiny foramina.
Figures 2.2, 2.4 shows two partial body osteoderms that measure approximately 3 mm in length (as preserved) and 0.7 mm in maximum thickness. An anterior gliding surface is visible on the exterior surface of both specimens (Fig. 2.2a, 2.4), and the area posterior to this surface is ornamented with a random tuberculated texture and pierced by tiny foramina. The specimen seen in Figure 2.2 tapers to a sharp edge anteriorly (Fig. 2.2c), and the one preserved lateral surface is beveled like the specimen shown in Figure 2.1. The interior surfaces are smooth and pierced by several foramina, and the broken posterior end of one osteoderm has a lamellar appearance (Fig. 2.2b). An incomplete polygonal osteoderm shows the effects of postmortem abrasion (Fig. 2.3). This specimen is slightly thicker than the body osteoderms previously described (1 mm), and lateral surfaces are rugose and vertical, not beveled. Eleven randomly located tubercles are preserved, whereas the interior surface is smooth.
Remarks
Osteoderms are common to all anguid lizards, but there are distinctive differences among the various subfamilies (Klembara and Green, Reference Klembara and Green2010). Two tribes are recognized within Glyptosaurinae, Melanosaurini and Glyptosaurini (Sullivan, Reference Sullivan1979), and synapomorphies characteristic of osteoderms include tubercular dorsal ornamentation and rugose lateral bevels (Gauthier, Reference Gauthier1982), both of which are present on the body osteoderms in our sample. Although glyptosaur osteoderms are not considered to be diagnostic below the tribe level (Sullivan, Reference Sullivan1979; Gauthier, Reference Gauthier1982; Sullivan and Augé, Reference Sullivan and Augé2006), certain morphological features allow us to meaningfully compare our sample to osteoderms of described species. For example, within Glyptosaurini, Eoglyptosaurus Sullivan, Reference Sullivan1979, Paraglyptosaurus Sullivan, Reference Sullivan1979, and Glyptosaurus Marsh, Reference Marsh1871 have keeled osteoderms, but tuberculation occurs in a concentric pattern (Sullivan, Reference Sullivan1979; Lucas et al., Reference Lucas, Sullivan and Logan1983), a feature not seen on any of the Jamestown material. Helodermoides Douglass, Reference Douglass1903 osteoderms lack concentrism and may be keeled, but they are subconical and the tuberculation is much more robust than on the South Carolina specimens (Lucas et al., Reference Lucas, Sullivan and Logan1983; Sullivan and Augé, Reference Sullivan and Augé2006). Within Melanosaurini, Proxestops Gauthier, Reference Gauthier1982 and Placosauriops Kuhn, Reference Kuhn1940 lack keeled osteoderms, and those of Proxestops have a vermiculate-tuberculate patterning. Xestops Cope, Reference Cope1873, Arpadosaurus Meszoely, Reference Meszoely1970, and Melanosaurus Gilmore, Reference Gilmore1928, have keeled osteoderms that lack concentrism as in the South Carolina taxon (Meszoely et al., Reference Meszoely, Estes and Haubold1978; Sullivan, Reference Sullivan1979; Gauthier, Reference Gauthier1982; Gunnell et al., Reference Gunnell, Bartels, Gingerich and Torres1992). The greater thickness and vertical, rugose lateral surfaces suggest that SC 2014.30.3 (Fig. 2.3) articulated with adjacent osteoderms, possibly forming more of a rigid outer shell around the head (Sullivan, Reference Sullivan1986). The lack of ventral rugosity indicates that it was not fused to an underlying bone and instead was embedded in the dermis, and its apparent hexagonal original outline suggests taxonomic placement within Glyptosaurini (Sullivan, Reference Sullivan1979, Gauthier, Reference Gauthier1982).
The body osteoderms have an anterior gliding surface, and the posterior end of the preceding osteoderm overlapped this area. Beveled margins of SC 2014.30.1 and SC 2014.30.2 (Fig. 2.1b, 2.2a, respectively) indicate that these osteoderms partially overlapped (and were overlapped by) adjacent in-column osteoderms and that there was at least some degree of flexibility to the armor covering (see Meszoely, Reference Meszoely1970; Gauthier, Reference Gauthier1982). The keeled osteoderm, SC 2014.30.1 (Fig. 2.1), occupied a position on the dorsal surface of the body (Sullivan, Reference Sullivan1979; Lucas et al., Reference Lucas, Sullivan and Logan1983), and although it could have come from one of a number of locations (Meszoely et al., Reference Meszoely, Estes and Haubold1978, fig. 4), the curvature of the beveled margins indicates it is from the left side. At this point, there is no reason to assume that our osteoderm sample represents anything other than normal morphological variation in the body armor elements of a single taxon.
Discussion
Age and stratigraphic occurrence of fossils
The elasmobranch species we have identified include, among others, Coupatezia woutersi Cappetta, Reference Cappetta1982, Jacquhermania duponti (Winkler, Reference Winkler1874), Heterotorpedo fowleri Ward, Reference Ward1983, Rhinobatos bruxelliensis Jaekel, Reference Jaekel1894, and Otodus obliquus Agassiz, Reference Agassiz1843. These are well-known species from upper Paleocene and lower Eocene strata elsewhere in Europe and the United States (Leriche, Reference Leriche1905, Reference Leriche1908, Reference Leriche1922; Casier, Reference Casier1946, Reference Casier1950, Reference Casier1966, Reference Casier1967; Case, Reference Case1994a, Reference Caseb; Kent, Reference Kent1999a, Reference Kentb; Weems, Reference Weems1999). The close similarity between the fossil fish assemblages of the United States, England, France, and Belgium has led some to conclude that the Jamestown, South Carolina, deposit was of early Ypresian age, possibly as old as NP10 (Knight et al., Reference Knight, Cicimurri and Purdy2007; Cicimurri and Knight, Reference Cicimurri and Knight2009; Cicimurri, Reference Cicimurri2010). Case et al. (Reference Case, Cook and Wilson2011) identified the Jamestown deposit as being part of the Fishburne Formation (Ypresian, NP11), probably on the basis of previously published age estimates. However, analysis of calcareous nannofossils from the sediment we collected has revealed that the fossiliferous deposit is slightly older and of latest Paleocene age (Thanetian Stage, subzone NP9a, equivalent to part of the Clarkforkian NALMA). This determination is based on the presence of Discoaster multiradiatus Bramlette and Riedel, Reference Bramlette and Ridel1954, which has its first occurrence at the base of NP9a, and the absence of Discoaster araneus Bukry, Reference Bukry1971, which marks the base of subzone NP9b, and thus constrains the sample to subzone NP9a. In addition, the presence of common Hornibrookina arca Bybell and Self-Trail, Reference Bybell and Self-Trail1995 and Prinsius bisulcus (Stradner, Reference Stradner1963), and the absence of Rhomboaster spp. and Bomolithus supremus Bown and Dunkley Jones, Reference Bown and Dunkley Jones2006, supports the conclusion that these sediments were deposited just prior to the Paleocene/Eocene boundary (Fig. 3.1; Self-Trail et al., Reference Self-Trail, Powars, Watkins and Wandless2012). There is no evidence of mixed-age assemblages, suggesting that reworking and/or contamination is not an issue.
(1) Late Paleocene to early Eocene geomagnetic polarity scale and calcareous nannofossil zonation. Calcareous nannofossil stratigraphy is based on the zonation of Martini (Reference Martini1971). The timescale is based on Vandenberghe et al. (Reference Vandenberghe, Hilgen and Speijer2012). Upward-pointing triangles indicate first-occurrence datums; downward-pointing triangles indicate last-occurrence datums. Dashed lines indicate possibly diachronous datums. (2) Paleocene strata within the Outer Coastal Plain of South Carolina. Gray areas represent unconformities. Modified from Weems and Bybell (Reference Weems and Bybell1998). The glyptosaur fossils occur within the upper part of the Chicora Member of the Williamsburg Formation (subzone NP9a, latest Paleocene).
The Jamestown deposit is therefore temporally equivalent to the Chicora Member (NP8 to NP9a) of the Williamsburg Formation as reported in northwestern Berkeley County (Fig. 3.2), and most of the lower vertebrates reported from the Williamsburg Formation (Hutchinson and Weems, Reference Hutchinson and Weems1998; Purdy, Reference Purdy1998; Weems, Reference Weems1998; Weems and Bybell, Reference Weems and Bybell1998) have been recovered in the Jamestown deposit. In Berkeley County, the Chicora Member may be disconformably overlain by Pleistocene strata, whereas in southern Dorchester County it is disconformably overlain by the Fishburne Formation (NP11; Harris and Zullo, Reference Harris and Zullo1991; Edwards et al., Reference Edwards, Gohn, Bybell, Chirico, Christopher, Frederiksen, Prowell, Self-Trail and Weems2000). Edwards et al. (Reference Edwards, Gohn, Bybell, Chirico, Christopher, Frederiksen, Prowell, Self-Trail and Weems2000) noted that the Chicora Member disconformably underlies the Santee Limestone in northern Dorchester County, and the Santee Limestone also overlies the fossiliferous deposit at the Jamestown site (fragments of Santee Limestone show that late Paleocene fossils were reworked into the Eocene deposit). The fossiliferous Jamestown deposit is lithologically and paleontologically similar, and temporally equivalent, to the Chicora Member, and at this time it is most parsimonious to identify the deposit as this lithostratigraphic unit.
Geographic significance of find
No major mountain range or, after regression of the Cannonball Sea by the middle Paleocene (Boyd and Lillegraven, Reference Boyd and Lillegraven2011; Weems, Reference Weems2014), epeiric sea posed a barrier to inhibit eastward migration of glyptosaurines in North America, and the South Carolina fossils show that these lizards were established along the southeastern coastal margin of the United States by the latest Paleocene (57 Ma). To our knowledge, the South Carolina specimens thus far represent the only unequivocal Cenozoic occurrence of these lizards from east of the Mississippi River.
McKenna (Reference McKenna1975) argued for an early Eocene continuity between North America and Europe, and two early Eocene routes linking these two regions are recognized (Hooker, Reference Hooker2007; Eberle and Greenwood, Reference Eberle and Greenwood2012). Discoveries of early Eocene glyptosaur fossils on Ellesmere Island (Estes and Hutchison, Reference Estes and Hutchinson1980), Canadian high arctic, provided evidence that this region was a potential dispersal route for these lizards into Europe during the Wasatchian NALMA (Sullivan, Reference Sullivan1979; Smith, Reference Smith2009). However, a slightly older dispersal route, following the Atlantic coast of the United States and Canada, now seems plausible and could explain the apparent sudden appearance of glyptosaurines in the earliest Eocene of Europe (Sullivan et al., Reference Sullivan, Augé, Wille and Smith2012).
Paleoecology
Glyptosaur fossils found in the western United States occur in strata representing fluvial and lacustrine environments (Sullivan, Reference Sullivan1979 and references cited therein), but the South Carolina taxon occurs in a stratum that represents a near-shore marine habitat. The diverse vertebrate assemblage recovered from the Jamestown deposit is dominated by marine species, including more than 60 taxa of elasmobranchs and osteichthyans (based on teeth and otoliths). Calcareous nannofossils indicate that deposition occurred in an inner to middle neritic environment. Although at least one salinity-tolerant turtle is present (Taphrosphys sp.), shell fragments of kinosternid, pelomedusid, and tryonichid turtles are common and indicative of fluvial or brackish habitats. The glyptosaur osteoderms, as well as crocodilian teeth, a mammal tooth, and fragments of lignite, also hint at the existence of these environments. In general, the fish otoliths represent species that inhabited inner- to middle-shelf-depth environments, but estuarine taxa are also present. Abundant oyster shells also point to the existence of an estuary. However, calcareous nannoplankton do not typically thrive in estuarine settings, and where they do, their assemblages tend to be mono- or oligospecific, which the Jamestown assemblage is not. A pro-delta environment would be deep enough for calcareous nannoplankton, but still receive an influx of sediment/vertebrate remains from the shore. We conclude that deposition of the fossiliferous deposit occurred in a warm-water environment deep enough for calcareous nannofossils but close enough to the shore for the remains of terrestrial/brackish-water organisms to be abundant. Middle Eocene (Bridgerian NALMA) glyptosaur osteoderms found in Texas occur in strata representing an estuarine, possibly mangrove-swamp environment (Westgate, Reference Westgate1989; Westgate and Gee, Reference Westgate and Gee1990).
Acknowledgments
H. Major and B. Palmer arranged access to the Martin Marietta quarry. V. McCollum assisted with field collections, and J. Ciegler sorted concentrated matrix using a binocular microscope. R. Sullivan confirmed our identification of the South Carolina osteoderms as belonging to a glyptosaurine lizard. R. Denton shared his knowledge of Cretaceous anguid lizards from New Jersey, and F. Swain kindly identified ostracodes. J. Ebersole drafted the stratigraphic column seen in Figure 3.2. The editorial suggestions of I. Burgess, E. Osborne, E. Seefelt, and R. Weems helped to improve an earlier version of this manuscript and are greatly appreciated.
