PALAIOS, 2011, v. 26, p. 639–657 Research Article DOI: 10.2110/palo.2010.p10-121r

FISHES AND TETRAPODS IN THE UPPER PENNSYLVANIAN (KASIMOVIAN) COHN COAL MEMBER OF THE MATTOON FORMATION OF ILLINOIS, UNITED STATES: SYSTEMATICS, PALEOECOLOGY, AND PALEOENVIRONMENTS

DAVID CARPENTER,1 HOWARD J. FALCON-LANG,2* MICHAEL J. BENTON,1 and W. JOHN NELSON 3 1Department of Earth Sciences, University of Bristol, Bristol BS8 1RJ, UK, [email protected], [email protected]; 2Department of Earth Sciences, Royal Holloway, University of London, Egham, Surrey TW20 0EX, UK, [email protected]; 3Illinois State Geological Survey, Champaign, Illinois 61820, USA, [email protected]

ABSTRACT sedimentation are dominant in Pennsylvanian successions in Euramer- ica (Wanless and Weller, 1932), reflecting coupled fluctuations in A newly discovered vertebrate assemblage is reported from the Upper climate and sea level (Tandon and Gibling, 1994) linked to Gondwanan Pennsylvanian (mid- to upper Kasimovian) Cohn Coal Member of the glacial–interglacial cycles (Fielding et al., 2008a, 2008b; Falcon-Lang et Mattoon Formation of southeast Illinois, United States. Teeth, scales, al., 2009). With paleoenvironments rapidly oscillating between fresh, and spines of xenacanth (Dicentrodus, Orthacanthus, Triodus, Xena- brackish, and marine conditions in space and time (e.g., Hampson et canthus) and euselachian (Sphenacanthus) sharks dominate the assem- al., 1999; Falcon-Lang, 2004, 2005), unusually precise geologic control blage. Less common are the teeth, scales, and centra of holocephalan of assemblages is required to determine habitat salinity. Notable (Helodus) and actinopterygian fishes, together with rare tetrapod (mainly examples of such analyses include the classic studies on faunal pelycosaur) phalanges and centra. The assemblage occurs within a broad, distribution in vertical profiles through British marine bands (Calver, shallow channel incised into a prominent Vertisol. The channel is 1968), or the identification of a salinity gradient from brackish interpreted as having been cut during a seasonally dry glacial phase nearshore to marine offshore assemblages in the Mazon Creek estuary when sea level was low, but filled during a subsequent transgression (Baird, 1997a, 1997b, 1997c). triggered by deglaciation. We interpret this as a brackish water In this paper, a new assemblage of Pennsylvanian fishes, together (estuarine) assemblage, based on the co-occurrence of the vertebrate with some rare tetrapods, is reported from the Friendsville Mine of material with spirorbids (putative microconchids) and paleoecological Illinois, United States (Fig. 1). In addition to describing vertebrate inferences gleaned from a critical analysis of the literature dealing with systematics, an analysis of the facies context, sequence stratigraphy, Pennsylvanian fish ecology. This interpretation is broadly consistent with taphonomy, and 87Sr/86Sr isotope geochemistry of the vertebrate 87 86 taphonomic data and the results of Sr/ Sr isotope analysis of shark assemblage is presented. These data, combined with a critical review material. The pelycosaur material may have been reworked from the of the literature concerning the geological context of Pennsylvanian lowstand Vertisol, however, and these animals occupied dryland niches fishes, allow assessment of paleoenvironments and paleosalinity, in that developed during glacial phases. particular, and provide a firm basis for interpretation of fish paleobiology and paleoecology. INTRODUCTION

Fish assemblages have been widely reported from the Pennsylvanian GEOLOGICAL SETTING of Euramerica (e.g., Agassiz, 1837–1843; Traquair, 1881; Fitsch, 1889; The vertebrate fossil assemblage from the Friendsville Mine comes Woodward, 1891; Zidek, 1978; Zangerl, 1981; Schultze, 1985; Soler- from an incised channel below the Cohn Coal, a member of the middle Gijo´n, 1997; Calder, 1998; Hampe, 2003; Johnson and Thayer, 2009), Missourian to Virgilian Mattoon Formation (Nance and Treworgy, but in many cases relatively little attention has been given to their 1981). In earlier reports, this coal was termed the McCleary’s Bluff Coal paleoenvironmental context and paleoecology (see Zangerl and (Kosanke et al., 1960). Regional correlation based on borehole cores, Richardson, 1963; Baird et al., 1985a, 1985b for notable exceptions). however, shows that it is equivalent to the type Cohn Coal in Clark Traditionally, Pennsylvanian fishes have been classified either as marine County, Illinois, 110 km north of the Friendsville Mine (W.J. Nelson, or nonmarine (Calder, 1998), the latter often taken to mean freshwater unpublished data, 2009). Although the two coals are no longer (Dick, 1998). In very many cases, the mere absence of an associated continuous following erosion across the La Salle Anticlinorium, stenohaline fauna and the co-occurrence of plant fossils have been cited correlation is beyond reasonable doubt because the prominent as sure evidence for a freshwater habitat (Hook and Baird, 1988; Hook Livingston Limestone and Eudora Shale bracket the coals in both and Ferm, 1988; Sander, 1989). Many such analyses, however, overlook areas (Heckel, 2008). As the name Cohn Coal (Newton and Weller, the importance and widespread distribution of Pennsylvanian brackish 1937) has priority over McCleary’s Bluff Coal (Kosanke et al., 1960) we water settings, e.g., commonly encountered in estuaries, interdistribu- use the former stratigraphic term. tary bays, and even certain epicontinental seas (Buatois et al., 1998; The Cohn Coal can be tied precisely to global Carboniferous Schultze, 1998; Falcon-Lang, 2005). Advances in the understanding of stratigraphy, based on the biostratigraphy of bracketing marine units tidal deposits (Kvale et al., 1989), invertebrate salinity tolerances (Fig. 2; Gradstein et al., 2004; Heckel, 2008; Davydov et al., 2010; (Calver, 1968; Bennett, 2008), and ichnology (Archer et al., 1995; Falcon-Lang et al., 2011). As noted above, the Cohn Coal occurs just Buatois et al., 2005) now permit recognition of such brackish facies in above the widespread marine transgressive unit termed the Livingston the rock record (Schultze, 2009). Limestone (Willman et al., 1975). This unit is positioned two major Nonetheless, determining paleosalinity for Pennsylvanian fish 400 ka cyclothem groups (Iola Group) below the top of the Missourian assemblages is particularly challenging. Cyclothemic patterns of regional stage in North America (Heckel, 2008). The Livingston Limestone and its lateral equivalents can be correlated, based on * Corresponding author. conodont biostratigraphy, throughout the Midcontinent, Illinois, and

Copyright G 2011, SEPM (Society for Sedimentary Geology) 0883-1351/11/0026-0639/$3.00 640 CARPENTER ET AL. PALAIOS

overlying Cohn Coal interval is of mid- to upper Kasimovian age (,304 Ma) and may be broadly time-equivalent to the mid-Stephanian (Barruelian) of western Europe.

GEOLOGY OF FRIENDSVILLE MINE, ILLINOIS

The Friendsville Mine is located 9 km west northwest of Mount Carmel in Wabash County, southeast Illinois, United States (Latitude 38u269280N; Longitude 87u519470W; Section 9, T1S, R13W, Wabash County, Mount Carmel 7.59 quadrangle map). Operated by the Vigo Coal Company, it comprises a 1.8-km-long highwall oriented N–S and advancing to the east. We studied the geology in the course of three visits on 14 June 2005, 19 August 2008, and 26 January 2009. In the course of the latter visit the vertebrate remains were discovered.

Sedimentary Facies

In the Friendsville Mine, a 19.5-m-thick succession is accessible in multiple benches and comprises part of the lower Mattoon Formation. Five main depositional units were recognizable during our site visits (Fig. 3). Unit 1: The lowermost unit currently exposed is the 0.7–1.1-m-thick Friendsville Coal, a low pyrite, dull-banded, semibituminous coal with a stigmarian seat earth. This comprises the Vigo Coal Company’s main target coal and forms the floor of the strip mine. FIGURE 1—Location of Friendsville Mine in United States (inset map) and Wabash Unit 2: Sharply overlying the Friendsville Coal is a complex body of County, southeast Illinois, near the border with Kentucky and Indiana, United States laterally variable facies about 9 m thick. This succession is dominated (modified from Falcon-Lang et al., 2009). The outcrop belt of Pennsylvanian rock is over much of the length of the highwall by medium- to dark-gray, shown (gray). heterolithic, rhythmically laminated shale, which coarsens upward. Features include flaser lamination, symmetrical ripples, siderite Appalachian basins of North America (Heckel, 2008) and into the nodules, and a rich ichnofauna of Cochlichnus, Arenicolites, and Moscow and Donets basins of eastern Europe (Falcon-Lang et al., Lockeia. Along strike, especially within scoured surfaces, however, 2011), where the global stratotypes are defined (Davydov et al., 2010). sediments may grade up into very fine to fine-grained, thinly bedded, This correlation suggests that the Livingston Limestone and the sandstone showing cross ripple lamination, soft-sediment deformation,

FIGURE 2—Stratigraphic position of the Missourian (mid- to upper Kasimovian) Cohn Coal in relation to regional and global stratigraphy (after Heckel et al., 2007; Heckel, 2008; Davydov et al., 2010; Falcon-Lang et al., 2011). Dotted boundaries between European units reflect that these are dominantly terrestrial (t) sections that cannot be correlated, with certainty, to the marine global stratotypes in eastern Europe. PALAIOS PENNSYLVANIAN FISH AND TETRAPOD FAUNA FROM ILLINOIS, USA 641

FIGURE 3—Geological context of study site. A) Upper Pennsylvanian Stratigraphy of Illinois (after Willman et al., 1975; modified using stratigraphic revisions in Heckel, 2008). B) The sedimentary succession in Friendsville Mine showing the context of the fossil vertebrate assemblage. Abbreviations: Grain size, Cl 5 clay, S 5 silt, Fs 5 fine- grained sandstone, Ms 5 medium-grained sandstone, Cs 5 coarse-grained sandstone.

Lockeia, and locally, lateral accretion bedforms, 1–2 m thick. showing a blocky structure of peds with clay skins, irregular siderite- Elsewhere, the Friendsville coal may be overlain by dark-gray fossil- and calcite-cemented roots, pervasive slickensides, pseudoanticlines, rich calcareous shale with channels filled with productid- or crinoid- pedogenic calcite nodules, #40 mm diameter, and an undulating upper dominated packstone and grainstone lenses. surface. The calcite nodules are in greatest abundance in the upper Unit 3: The uppermost 1.1 m of Unit 2 marks an abrupt change in 0.25–0.5 m of the unit. facies. Here, the shale exhibits siderite-filled desiccation cracks, Unit 4: Draping the irregular top of Unit 3 is a #0.6-m-thick zone of networks of tiny fractures, and masses of intergrown calcite and laminated dark gray, carbonaceous shale containing abundant plant siderite nodules, #0.3 m diameter. This hummocky unit grades up into compressions (dominantly Macroneuropteris scheuchzeri pinnules and an olive green to medium-gray claystone paleosol, 1.4–2.6 m thick, axes with rare Neuropteris ovata, Alethopteris sp., Pecopteris sp., 642 CARPENTER ET AL. PALAIOS

FIGURE 4—Interpreted photomosaic of part of the incised channel fill below the Cohn Coal, in the area from which the vertebrate fossils were collected (shown by stars).

Calamites suckowi, and Cordaites sp.) with spirorbids (putative Paleoenvironmental Interpretation microconchids) locally encrusting foliage in abundance. These shaley beds contain common, thin coaly intervals, 10–30 mm thick, and are The Friendsville Coal (Unit 1) is interpreted as the deposits of long- capped by the ,0.3-m-thick Cohn Coal, a pyrite-rich, bright-banded, lived coastal peat mires formed under perhumid tropical conditions semibituminous coal. Locally, the Cohn Coal is cut, from base to top, (DiMichele et al., 2001, 2007). The overlying complex of rhythmically by a series of claystone dykes. laminated shales, sandstone channels with lateral accretion and Unit 5: The uppermost strata in the strip mine are #6 m thick, but channels infilled with crinoid and brachiopod debris (Unit 2) represents not well exposed. In some areas, they may comprise beds of calcareous a variety of tidal flat, estuarine, and coastal marine deposits formed shale with thin limestone lenses of brachiopod- and crinoid-dominated following rapid marine transgression (Kvale et al., 1989). The thick wackestone. Elsewhere, they may consist of medium-gray siltstone, claystone paleosol (Unit 3) developed on top is interpreted as a Vertisol, showing upward coarsening, convolute lamination, and a prominent a fossil soil formed under seasonally dry conditions (Driese and Ober, layer of ball-and-pillow soft sediment deformation structures immedi- 2005). The associated shallow channel is interpreted as an alluvial ately above the Cohn Coal. drainage incised into this landscape during base-level fall (Falcon-Lang et al., 2009). The channel must have been originally at least 5 m deep, Incised Channel Fill taking into account a minimal 32 sediment compaction. A conglom- erate of pedogenic carbonate nodules in the base of the channel was Near the southeast corner of the strip mine, a shallowly incised channel reworked from the adjacent dryland soil. Dark shales with brackish is present, cutting down from a horizon within Unit 3, the claystone spirorbids (Schultze, 2009), higher in the channel fill, record the early paleosol below the Cohn Coal, and shallowly eroding into the top of Unit stages of transgression that started to flood the channel and form an 2. The channel body is 2.1–2.6 m thick at its deepest point and estuary. The thin lenses of pedogenic carbonate conglomerate approximately 45 m wide—estimated perpendicular to the approximate- containing most of the vertebrate remains probably represent a ly N–S oriented channel margins (Fig. 4). The Cohn Coal shows no thickness change where it extends over the top of the channel fill. The basal channel fill contains a matrix-supported, granular to pebble conglomerate, 0.6 m thick, composed of reworked pedogenic carbonate nodules (presumably derived from Unit 3) and shale clasts, together with a few quartz grains, within a lime mud matrix. Vertebrate skeletal fragments are present in low numbers. Higher in the channel fill are units of dark-gray, laminated shale containing common plant compressions—especially Cordaites sp., Macroneuropteris scheuchzeri, rare calamiteans, and indeterminate axes—encrusted with spirorbids (Taylor and Vinn, 2006) and a few freshwater to brackish water bivalves. The shale beds thicken toward the deepest part of the channel (Fig. 4). Several thin lenses, #0.1 m thick, of conglomerate identical to that lower in the channel fill are interbedded with the dark shale. These lenses contain highly abundant vertebrate skeletal fragments, including elements #0.14 m long. Also present in the conglomerate are abundant specimens of Valvisisporites auritus, the megaspore of the small FIGURE 5—Megaspores and microconchids in the channel-fill. A) Valvisisporites lycopsid, Chaloneria (Figs. 5A, C; Gastaldo, 1981; Pigg and Rothwell, auritus, proximal polar view. B) Triletes sp. C) Valvisisporites auritus, distal polar 1983), other sparse megaspores (Fig. 5B), and two morphotypes of view. D) ?Microconchus sp. 1. E) Microconchus sp. 2. F) Shell microstructure of spirorbids (Figs. 5D–F). ?Microconchus sp. 1. Scale bars: 1 mm (A–E), 50 mm (F). PALAIOS PENNSYLVANIAN FISH AND TETRAPOD FAUNA FROM ILLINOIS, USA 643 ravinement surface that reworked channel-floor sediments during transgression to produce a lag deposit rich in vertebrate remains. The Cohn Coal that blankets this surface represented a renewed phase of coal formation in a coastal setting following a switch back to perhumid climatic conditions (Unit 4). Finally, shallow marine facies of Unit 5 represent further transgression and development of shallow shelf conditions. The presence of a prominent ball-and-pillow soft sediment deformation layer just above the Cohn Coal may imply that coseismic subsidence contributed to rapid flooding in this coastal setting (Gastaldo et al., 2004).

Cyclothems, Sea Level, and Climate

The cyclic patterns of marine transgression and regression (cy- clothems) seen in our study area are generally interpreted as the response of coastal regions to coupled fluctuations in climate and sea level (Tandon and Gibling, 1994) inferred to represent the far-field effects of Gondwanan glacial–interglacial cycles (Fielding et al., 2008a, 2008b). The Vertisol (Unit 3) and its associated incised channel likely started to form as sea level was dropping toward, and following, lowstand (glacial maximum) when the region was seasonally dry (Feldman et al., 2005) and coniferopsid vegetation dominated FIGURE 6—Histogram showing the absolute abundance of fish and tetrapod (DiMichele et al., 2009, 2010; Falcon-Lang and DiMichele, 2010). elements in the Friendsville Mine assemblage (n 5 1399). The upper part of the channel, which is filled with brackish spirorbid- bearing shale and contains the vertebrate remains, however, marks a turnaround with sea level rising once again during deglaciation. Being with the rare remains of chimaeroid fish (n 5 5; 0.4%), other dominated by the megaspores of Chaloneria-type herbaceous lycopsids, indeterminate bony fish (n 5 20; 1.4%), and tetrapods (n 5 17; the palynoflora of this interval is similar to that seen in early 1.2%) (Fig. 6). All figured specimens, and additional material, are deglaciation deposits from other Pennsylvanian cyclothems (Dolby et stored in the Vertebrate Paleontology collection of the National al., 2011). Museum of Natural History, Smithsonian Institution, Washington, D.C., United States (USNM 542530–542560). MATERIAL AND METHODS SHARKS (FIGS. 7A–F, H–L) Approximately 15 kg of bulk samples of rock were obtained from Class Chondrichthyes Huxley, 1880 conglomerate lenses associated by dark-gray shales in the upper part of Subclass Elasmobranchii Bonaparte, 1832–1841 the incised channel fill in the Friendsville Mine. In the laboratory, Sharks are by far the most abundant and diverse component of the samples were disaggregated by immersion in 5% acetic acid solution, fish fauna comprising 73.5% of all the vertebrate specimens (n 5 1399), buffered with tri-calcium di-orthophosphate and calcium acetate (spent representing at least five recognizable genera. acid); the latter was produced by adding a small quantity of sodium carbonate to each freshly mixed batch of acid before the samples were Order Xenacanthiformes Berg, 1937 introduced. Each acid cycle lasted 72 hours followed by a 144-hour The identification of xenacanthid species using teeth alone is rinse cycle to neutralize residual acid. The rinse cycle involved problematic, but is often the only option available since preservation immersion in a mixture of water, detergent and sodium carbonate. of cartilaginous elements is rare (Johnson, 1999). Xenacanths are During each acid cycle, the samples were suspended in a plastic sieve to known to possess graduated, monognathic heterodont dentition aid acid circulation. After each rinse cycle, any fossils that were visibly (Compagno, 1970), and the morphological variation—particularly the protruding $5 mm were removed and allowed to dry. These fossils were angle of divergence of the lateral cusps, and the numbers of coronal and then coated in a weak, ethanol-based adhesive (Mowital), to prevent basal nutritive foramina—can be as great within individual dentitions breakage during subsequent manual extraction. The residue from each as between different species (Johnson, 1999). Ontogenetic variation can acid cycle was passed through a series of sieves (aperture sizes: 1 mm, also be an issue, for example in the development of serrations in 0.5 mm, 0.25 mm) and decanted onto filter paper. After 48 hours drying Orthacanthus (Soler-Gijo´n, 1997). Ultimately, there is often no single time, remaining fossils were manually picked under a transmitted light reliable character to differentiate between closely related species, and a microscope. suite of characters must be used in combination. If the specimens are A total of 7642 vertebrate specimens were obtained through this incomplete or poorly preserved, this may be impossible. Xenacanthid process with a size range of 0.25–15 mm (typically 0.5–3 mm). Most tooth components are described here using the terminology of Hampe specimens either lacked any diagnostic features or were too abraded to (2003, fig. 2). In the following, length refers to the distance between permit identification. The 1399 specimens (18.31%) that could be anteromedial and posterolateral margins of the base, and is perpendic- identified to some taxonomic level were subdivided into morphotypes ular to width, which refers to the distance between the labial and lingual with the aid of a transmitted light microscope and picked into plastic margins (Johnson and Thayer, 2009). containers. The best specimens of each taxon were then mounted onto Family Xenacanthidae Fritsch, 1889 10 mm aluminum stubs, gold coated using a Bio-Rad SC650 sputter Triodus sp. (Fig. 7K–L) coater, and imaged using a Hitachi S-3500N scanning electron microscope. Genus: Triodus Jordan, 1849. The vertebrate assemblage is described in systematic detail, Type species: Triodus sessilis Jordan, 1849. dominated by a rich diversity of xenacanth and euselachian sharks (n Material: 105 teeth. Figured specimen (USNM 542530), other 5 1028; 73.5%) and actinopterygian fish (n 5 329; 23.5%), together material (USNM 542546). 644 CARPENTER ET AL. PALAIOS

FIGURE 7—Sharks and chimaeroid fish. A) Xenacanthus cf. ossiani, oblique lingual view, USNM 542531. B) Xenacanthus cf. ossiani, occlusal view, USNM 542531. C) Elasmobranchii indet., lateral view, USNM 542537. D) Elasmobranchii indet., dorsal view, USNM 542537. E) Orthacanthus cf. compressus, oblique lingual view, USNM 542532. F) Orthacanthus cf. compressus, occlusal view, USNM 542532. G) Holocephali indet., oblique dorsal view, USNM 542536. H) Sphenacanthus sp., oblique lingual view, USNM 542534. I) Sphenacanthus sp., occlusal view, USNM 542534. J) Dicentrodus cf. alatus, lateral view, USNM 542533. K) Triodus sp., oblique lingual view, USNM 542530. L) Triodus sp., occlusal view, USNM 542530. M) Helodus sp., labial view, USNM 542535. N) Helodus sp., occlusal view, USNM 542535. Scale bars for all images: 1 mm (A–N). PALAIOS PENNSYLVANIAN FISH AND TETRAPOD FAUNA FROM ILLINOIS, USA 645

Diagnosis: Teeth tricuspid, cusps polygonal in cross section and of the lateral cusps. Coronal button isolated from cusps and flush with equipped with straight vertical cristae on labial and lingual surfaces, lingual margin of the base, which is typically but not always longer than which may split dichotomously. wide and varies greatly in thickness between specimens. Variable Description: The teeth range in length from 0.50 to 2.25 mm. The numbers of coronal and basal nutritive foramina are present, most median cusp is slender and approximately two thirds of the length of commonly two and four respectively. the lateral cusps; this size relationship is difficult to judge, however, as Remarks: The genera Orthacanthus and Lebachacanthus are the median cusp is incomplete in most specimens. A median foramen is indistinguishable based on dental characteristics alone (Soler-Gijo´n, always present, and variable numbers of coronal and basal nutritive 1997, 2000). The tentative assignment of the specimens described here foramina are observed, most commonly 3 and 4, respectively. The to Orthacanthus rests solely on the fact that Lebachacanthus has thus coronal button is prominent and rounded, and the basal tubercle far only been reported from lower Permian strata in the Saar Nahe depressed. Basin, Germany (Soler-Gijo´n, 1997), whereas Orthacanthus is known Remarks: The teeth are distinguished from those of the otherwise very from Pennsylvanian of North America (Johnson and Thayer, 2009). similar bransonelliforms Bransonella and Barbclabornia and the xena- Our material is closely similar to Orthacanthus compressus, which is canthid Plicatodus by the pattern of cristae. These form characteristic the most common Orthacanthus species found in the Pennsylvanian nested, inverted V shapes in all known bransonelliforms (Johnson, 2003, of North America. There is considerable morphological overlap figs. 8A–N; Hampe and Ivanov, 2007, figs. 1A–H), and in Plicatodus among the four known North American Orthacanthus species, follow an ‘‘undulating hybodont-like pattern’’ (Hampe, 2003, p. 225), however, i.e., the variation between teeth from a single heterodont neither of which patterns are observed here. Most Triodus species are dentition can be as great as the variation between species, so the known from the Saar-Nahe Basin, Germany; the only other known value of specific determination may be limited (Johnson and Thayer, occurrence of Triodus in North America is that of Triodus elpia from the 2009). Lower Pennsylvanian (upper Bashkirian) of Arizona (Johnson and Thayer, 2009), so our new specimen extends the stratigraphic range of Dicentrodus cf. alatus (Fig. 7J) this genus in North America by approximately 10 Ma. Hampe (2003) notes that the reason this genus is so uncommon in many areas may Genus: Dicentrodus Traquair, 1888. simply be the result of collector bias, as the teeth are typically very small Type species: Dicentrodus bicuspidatus Traquair, 1881. and easily overlooked. Material: Two spine fragments (USNM 542533, USNM 542548) Diagnosis: Xenacanthid spine with two rows of denticles oriented Xenacanthus cf. ossiani Johnson, 1999 (Figs. 7A–B) ventro-laterally. Description: Straight spines with an elliptical cross section, flattened Genus: Xenacanthus Beyrich, 1848. ventrally, and narrowing distally. Two ventrolateral rows of denticles Type species: Xenacanthus decheni Goldfuss, 1847. with slender, pointed cusps are present; the surface of one of the spines Material: 93 teeth. Figure specimen (USNM 542531), other material is striated, while the other is smooth. The denticle cusps are directed (USNM 542547). proximally. Diagnosis: Teeth tricuspid, cusps lacking serrations and lanceolate in Remarks: This material was original placed in Anodontacanthus cross section. Basal surface concave, basal tubercle depressed. Davis, 1881 but this taxon was recently synonymized with Dicentrodus Description: Length ranges from 0.5 to 3 mm. The median cusp is (Ginter et al., 2010). The level of striation varies between proximal approximately a third of the length and width of the lateral cusps. and distal portions of individual spines, as well as between species Variable numbers of coronal and basal nutritive foramina are present, (Hampe, 2003, figs. 12a–d); therefore, whether the two incomplete most commonly three and four respectively. Aside from cases where specimens presented here represent different species is not possible to obscuring matrix or incompleteness prevents assessment, a prominent determine. median foramen is present in all specimens. The base is longer than wide, with a flat or slightly concave basal surface. The basal tubercle is Xenacanthiformes indet. depressed, and a shaft extends to the center, although this is not a Material: 232 teeth (USNM 542550). prominent feature. The coronal button is very prominent and extends Description: Moderately to severely abraded and pitted tricuspid beyond the lingual margin, but is isolated from the cusps and medial teeth of length (where complete), 0.5–5 mm, with two lateral cusps, foramen. The cusps are sometimes spatulate at their tips, and though flanking one or sometimes two smaller median cusps. The cusps (where divergent are generally of similar size. present) may be carinated or smooth. In most cases the cusps are Remarks: The teeth have many features in common with those of broken near the base, and the base itself is also frequently incomplete. Xenacanthus ossiani (Johnson, 1999, fig. 24)—the base in particular is The base extends lingually, and possesses a coronal button, basal extremely similar—but lack the labial inclination of the cusps observed tubercle, and variable numbers of nutritive foramina on its basal and in all known specimens of that species. Johnson (1999) speculated that upper surfaces. X. ossiani might exhibit a heterodont dentition; if so, this could explain Remarks: From incompleteness and poor preservation, these the differences seen here. specimens lack the full suite of diagnostic characters required for Family Diplodoselachidae Dick, 1981 further identification. ?Orthacanthus cf. compressus (Figs. 7E–F) Elasmobranchii Indet. (Figs. 7C–D) Genus: Orthacanthus Agassiz, 1843. Material: 318 scales. Figured specimen (USNM 542537), other Type species: Orthacanthus cylindricus Agassiz, 1843. material (USNM 542553). Material: 216 teeth. Figured specimen (USNM 542532), other Description: Scales comprise distinct base and crown components. material (USNM 542548). The crown is composed of 1–5 recurved lepidomoria-type cusps, Diagnosis: Teeth tricuspid, with lateral cusps much larger than equipped with strong vertical ridges that tend to converge distally. The median cusp. Lateral cusps serrated, median foramen always present. base is circular or ovoid, with a gently to moderately convex basal Basal tubercle undepressed. surface in which one or two small foramina are occasionally present; Description: Teeth range in length from 0.5 to 4 mm. The median these do not follow any regular pattern. The scales range in diameter cusp is frequently minute, and never more than a quarter of the length from 0.5 to 3 mm across the widest part of the base. 646 CARPENTER ET AL. PALAIOS

Remarks: Scales consisting of aggregates of lepidomoria sharing a Remarks: Some 72 species of Helodus have been named over the last common base are a common feature of Paleozoic chondrichthyans 150 years (Stahl, 1999), many of them based on isolated teeth. The (Zangerl, 1981). In addition, several chondrichthyan taxa are known in complete dentition is known only from the type species H. simplex, but which different scale types are found on different parts of the body, e.g., the available evidence suggests that at least some Helodus species were Diplodoselache (Hampe, 2003, fig. 5), Heterodontus (Reif, 1974), and heterodont, with multiple tooth types in one organism; the discovery of Orodus (Moy-Thomas and Miles, 1971). Scales are of little taxonomic associated dentition of Psephodus (Traquair, 1885) showed that teeth value because of these factors: Zangerl (1981), on the subject of species assigned to up to four different species belonged to a single fish (Stahl, specificity, notes that: ‘‘many [scales] have been described, but the 1999). Smith and Patterson (1988) felt that many teeth referred to the variation, even in a single individual, is so great that this [species family Helodontidae represented generalized bradyodont teeth. Pend- determination] is a hopeless proposition and merely burdens the ing the discovery of more complete dentitions, the taxonomy of Helodus nomenclature with useless names.’’ An identical scale was described must remain uncertain. by Lebedev (1996, fig. 3b) from the Lower Carboniferous of the Northern Urals, Russia, which he also attributed to Elasmobranchii Holocephali Indet. (Fig. 7G) indet. Material: One scale. Figured specimen (USNM 542536). Euselachii Incertae Sedis Description: A wide, ovoid base is surmounted by a single, narrow Family Sphenacanthidae Maisey, 1982 conical cusp, broken at its tip. The base is 2 mm across at its widest Sphenacanthus sp. (Fig. 7H–I) point, and folded into gentle ridges radiating from the cusp. Several prominent foramina penetrate the base, but do not appear to follow Genus: Sphenacanthus Agassiz, 1837. any regular pattern. Type species: Sphenacanthus serrulatus Agassiz, 1837. Remarks: The specimen appears quite similar to one of the scales of Material: 62 teeth. Figure specimen (USNM 542534), other material Helodus simplex illustrated by Stahl (1999, fig. 1A). (USNM 542551). Diagnosis: Labio-lingually compressed, multi-cusped teeth compris- RAY-FINNED FISHES (Figs. 8A, C–E) ing a large median cusp and one to four pairs of smaller lateral Subclass Actinopterygii Cope, 1887 cuspules, attached to a distinct base. Cusps and cuspules bear Ray-finned fishes (Actinopterygii) comprise the second most prominent vertical ridges. abundant component of the vertebrate assemblage comprising 329 Description: The base is gently arched along its anteromedial– specimens (23.5% of the total) and two morphotypes. Taxonomic posterolateral axis, is slightly overhung by the crown on its labial determination is only possible at a very general level, however. margin and develops a prominent flat shelf on the lingual margin. Remarks: Janvier (1996) remarks that, ‘‘the determination of isolated This shelf contains numerous, strongly developed foramina on its actinopterygian remains is extremely difficult’’. In particular, there has ventral surface, but there is little consistency in the pattern or never been a comprehensive study of the various forms of vertebrae that number of these. The basal surface is sharply divided between a occur within this group (Arratia et al., 2001). smooth section underlying the lingual shelf, and a recessed groove penetrated by multiple foramina beneath the crown. The cusps or Actinopterygii Indet. A (Figs. 8C–E) cuspules are often inclined lingually, and in larger teeth are slightly Material: 241 teeth. Figured specimen (USNM 542538), other recurved. material (USNM 542554). Remarks: As with many Paleozoic and Mesozoic shark groups, the Diagnosis: Teeth equipped with apical caps of acrodin. dentition of Sphenacanthus was heterodont; larger teeth with promi- Description: Slender, conical teeth with a gently sigmoidal shape. The nent, recurved cusps are likely to have been anteriorly situated, while translucent apical caps contrast with shafts exhibiting a distinctive smaller teeth with cusps less well developed probably represent cross-hatched surface texture. posterior teeth from closer to the jaw articulation (Dick, 1998). Remarks: The teeth of many actinopterygians are equipped with an apical cap of hypermineralized tissue (acrodin), which is not seen in any CHIMAEROID FISH (Figs. 7M–N) other group (Ørvig, 1978). This substance, first noted by Peyer (1968) Class Chondrichthyes Huxley, 1880 and subsequently named by Ørvig (1973), is therefore a clear indicator Subclass Subterbranchialia Zangerl, 1979 of actinopterygian affinity, although there are several cases in which it Superorder Holocephali Bonaparte, 1832–1841 is absent, e.g., Cheirolepis (Patterson, 1982), Andreolepis (Janvier, Chimaeroid fishes are a very rare component of the fish fauna 1978), the pachycormids, and some teleosts (Gardiner et al., 2005). This comprising 0.4% of all the vertebrate specimens (n 5 5), representing cap is clearly recognizable as distinct from the tooth shaft, with a only one confirmed family, Helodontidae. smooth surface texture and translucent appearance. The specimens described here appear identical to a single tooth reported by Robb Order Helodontiformes Patterson, 1965 (1992, figs. 2A–B) from the Wathena Shale Member, upper Lawrence Family Helodontidae Patterson, 1965 Formation (Douglas Group, Virgilian), which he assigns without Helodus sp. (Figs. 7M–N) evidence to Palaeoniscoidea indet. Genus: Helodus Agassiz, 1838. Actinopterygii Indet. B (Fig. 8A) Type species: Helodus simplex Agassiz, 1838. Material: 88 centra. Figured material (USNM 542540), other Material: Four teeth. Figured specimen (USNM 542535), other material (USNM 542556) material (USNM 542552). Description: Centrally constricted hourglass–shaped, vertebral centra, Description: Teeth labiolingually compressed, with a mesiodistally 0.5–11 mm long. A gently concave, frequently pitted dorsal surface is elongated crown, forming a rounded cone at its midpoint. The base is flanked on either side by longitudinally elongated neural arch sockets; convex, and extends on its labial side to form a prominent shelf the ventral surface is equipped with a single deep channel, also oriented containing strongly developed foramina. The lingual portion of the longitudinally. No attachment points for a hemal arch are evident. basal surface contains a wide, shallow groove, again containing Remarks: The dorsal and ventral surfaces are distinguished following strongly developed foramina. The mesiodistal length—estimated for the guidelines given by Berman (1970). The centra bear a close the incomplete specimens—ranges from 1.5 to 5 mm. resemblance to those described by Schultze and Chorn (1986, fig. 4), PALAIOS PENNSYLVANIAN FISH AND TETRAPOD FAUNA FROM ILLINOIS, USA 647

FIGURE 8—Actinopterygian fish. A) Actinopterygii indet. centrum, USNM 542540. B) Rhomboid scale base, USNM 542539. C) Actinopterygian tooth, USNM 542538. D– E) Enlarged views of C, USNM 542538. Scale bars: 1 mm (A–B), 500 mm (C), 150 mm(D),50mm(E). which they assign to Palaeoniscoidea indet. Arratia et al. (2001), in their TETRAPODS (Figs. 9A–D; Figs. 10A–G) brief overview of actinopterygian vertebrae, state that vertebrae of this Infraclass Tetrapoda Broili, 1913 type only occur in ‘‘some Late Paleozoic palaeonisciformes… and Supercohort Amniota Haeckel, 1866 probably in polypteriformes,’’ apparently referring to the vertebral Cohort Synapsida Osborn, 1903 centra described by Schultze and Chorn (1986). Tetrapods are a rare component (n 5 17) of the assemblage (1.2 ). INDETERMINATE BONY FISHES (Fig. 8B) % Class Osteichthyes Huxley, 1880 Osteichthyes Indet. Suborder Eupelycosauria Kemp, 1982 Material: Twenty rhomboid scales. Figured specimen (USNM Remarks: Only 13 pelycosaur localities have been reported from the 542539), other material (USNM 542555) Pennsylvanian of North America to date (Kissel and Lehman, 2002). Diagnosis: Rhomboid scales with distinct concentric growth lines. This includes material described as Milosaurus mccordi from the Description: The scales, where complete, are #3 mm along the Mattoon Formation of Illinois (DeMar, 1970), from which our own diagonal long axis and are lacking any kind of superficial covering of material originated. That species, however, is clearly dissimilar to the the bony base; it seems likely that any enameloid layer has been lost specimens described here; the wide, flattened cross-sectional shape of through abrasion the terminal phalanges of M. mccordi, typical of more primitive Remarks: Rhomboid scales are found in both the Sarcopterygii and synapsids, contrasts strongly with the triangular cross section seen in Actinopterygii during the Pennsylvanian, and thus the state of our specimens, which is a feature of more advanced sphenacodontid preservation of the specimens described here precludes more detailed groups. identification. Family Sphenacodontidae Williston, 1912 648 CARPENTER ET AL. PALAIOS

FIGURE 9—Sphenacodontid amniote, terminal phalange, USNM 542543. A) Ventral view. B) Dorsal view. C–D) Lateral views. Scale bars for all images: 5 mm.

Sphenacodontidae Indet. (Figs. 9A–D) Tetrapoda Indet. A (Figs. 10A–D) Material: One complete, and twelve incomplete, terminal phalanges. Material: Two incomplete centra (USNM 543545), other material Figured specimen (USNM 542543), other material (USNM 542558). (USNM 542559). Description: Specimens whose complete length can be measured Description: The first centrum is amphicoelous, longitudinally range from 7 to 14 mm; the features can best be seen in the largest compressed and distorted, but appears to have been smoothly specimen (Figs. 9A–D). The overall shape is similar to that depicted by cylindrical in its original state. The second is less complete, retaining Maddin and Reisz (2007, fig. 4E) for Dimetrodon limbatus; the dorsal only one articular end (which is also concave), but has a prominent surface is wide proximally and narrows to a point distally; in lateral notch in its outer edge, and part of a process (neural or hemal arch) is view, the downward curvature is moderate; in distal view, there is a also preserved. The articulating surfaces are 12–15 mm in diameter at marked asymmetry, with the lateral edge protruding further on one their widest points. side; in cross section, the shape is roughly triangular. Blood vessel channels run down each lateral edge, moving to a more ventral position Tetrapoda Indet. B (Fig. 10G) proximally; the channel on the more protruding edge is more Material: One caudal vertebra (USNM 543546). prominent. Numerous foramina occupy the area in and around the Description: Laterally compressed, amphicoelous centrum, 14 mm blood vessel channels, and also appear sporadically on the ventral long, with no visible rib attachment points and a gently concave dorsal surface; small, faint, undulating channels are also observed, which may process running along its length. record the presence of small blood vessels. The flexor tubercle is well developed; 1.25 mm ventral projection, 4.5 mm distal placement as measured from the proximal edge of the phalange. ABRASION INDEX Remarks: Maddin and Reisz (2007) showed that the morphology of In order to assess the taphonomic history of the vertebrate material, Permian–Carboniferous synapsid terminal phalanges follows a clear the level of abrasion for each specimen was assessed following the evolutionary trend. Based on their descriptions and figures (Maddin abrasion index developed by Behrensmeyer (1978) and modified by and Reisz, 2007, fig. 4), the current specimens look most like the claws Cook (1995). In this scheme, specimens are scored on a scale of 0–4 as of derived sphenacodontids such as Dimetrodon. This is uncertain, follows: however, as Dimetrodon is known exclusively from the lower Permian, specifically the lower Asselian–upper Kungurian, 3–4 myr younger than Stage 0. Angular: bone edges and processes, well defined and sharp. the Friendsville deposits. Among sphenacodontids, such haptodontines Stage 1. Subangular: bone edges and processes, slightly abraded as Haptodus were relatively common in the Pennsylvanian (Reisz, and polished. 1986). Stage 2. Subrounded: bone edges, well rounded; processes recognizable. Stage 3. Rounded: bone edges, high degree of rounding; processes, Amniota Indet. (Figs. 10E–F) generally remnant. Material: One incomplete vertebra (USNM 542544). Stage 4. Highly rounded: fragments often show a high degree of Description: A largely incomplete centrum retaining a partial neural sphericity. arch with transverse processes, identified as diapophyses based on the rib attachment point visible on the (left) lateral surface. Only those specimens (n 5 913) whose pristine morphology could be Remarks: The specimen is insufficiently complete to identify to inferred were measured. This is because determining the level of ordinal level, but it should be noted that there are no features that abrasion relies on knowing the original shape of the element. The preclude a sphenacodontid affinity for this specimen; given the presence analysis is somewhat subjective as it rests on judgments of whether, for of other sphenacodontid material in the assemblage, this would not be example, the bluntness of tooth tips is a natural feature of the tooth, a unexpected. result of wear through use—although this would typically result in PALAIOS PENNSYLVANIAN FISH AND TETRAPOD FAUNA FROM ILLINOIS, USA 649

FIGURE 10—Other tetrapod material. A–D) Tetrapoda indet., USNM 542545. E–F) Amniota indet. USNM 542544. G) Tetrapoda indet. (caudal vertebra), USNM 542546. Scale bars for all images: 5 mm. distinctive wear facets—or a result of postmortem abrasion. Thus, one remains show similar abrasion scores to the fish material, the dataset is major bias is that it is easier to detect abrasion of elements that have a very limited, and it remains an open question as to whether these complex morphology; e.g., fine surface ornamentation. terrestrial-based animals were living in and around the estuary (Laurin and Soler-Gijo´n, 2010), or whether their remains were reworked from the seasonally dry Vertisol into which the channel is incised. Results

Analysis of the vertebrate assemblage indicates wide variation in the STRONTIUM ISOTOPE ANALYSES level of abrasion (Fig. 11). While some specimens are not noticeably abraded, others are so rounded as to be hardly recognizable. Overall, To further test the hypothesis that the fishes were residents of the data suggest rather minimal abrasion, most specimens either showing brackish estuary, we conducted 87Sr/86Sr analysis of skeletal remains to no abrasion (23%; Stage 0) or only light abrasion (46%; Stage 1). In determine habitat paleosalinity (Schmitz et al., 1991). We did not addition, when analyzed taxonomically, there is no statistical difference perform similar tests on the tetrapods because the 87Sr/86Sr ratio of between taxa (range 0.58–1.36; statistical methods performed using their skeletal remains would be related to diet rather than paleoenvi- Microsoft ExcelH), with the exception of a single outlier, Sphenacanthus ronment. (mean 2.82). This outlier, however, is probably explained by the The 87Sr/86Sr paleosalinity technique relies on the difference morphological complexity of Sphenacanthus (prominently ridged between the 87Sr/86Sr isotopic composition of seawater and river cusps), which would make it more susceptible to abrasion. water. The 87Sr/86Sr isotopic ratio of modern riverine environments is highly variable, controlled by the bedrock geology of the drainage Taphonomic Interpretation area (Schmitz et al., 1991); water passing through old, Rb rich granitic rocks and siliciclastic sedimentary deposits will have a much The unimodal abrasion data suggests that our vertebrate material higher ratio than water passing through calcareous or young basaltic obtained from the ravinement surface originated from a single source. If strata (Schmitz et al., 1991). In contrast, the Sr isotopic composition the material under investigation was from multiple sources, a bimodal of open oceans is globally uniform at any given time, determined by or multimodal distribution of abrasion levels might be expected, the relative contribution of oceanic and terrestrial sources (Denison et reflecting the differing transport histories of specimens from different al., 1993). The slow changes in oceanic Sr composition that have areas; this is not seen in our data. As the abrasion score of most occurred during the Phanerozoic are well constrained (Prokoph et al., elements is low, this further implies that the assemblage has undergone 2008). Thus, fossils recording Sr isotopic ratios that are anomalous relatively little reworking and that the environment of deposition— with respect to contemporaneous oceanic values must logically be brackish estuary formed during transgression—was probably the from environments with some level of freshwater input (Dasch and environment in which the fishes were living. Although the rare tetrapod Campbell, 1970). 650 CARPENTER ET AL. PALAIOS

elutant was dried, and loaded into the ion exchange column using a few mlof10% HNO3, onto a single rhenium filament preconditioned with 1ml TaCl5 solution and 1ml10% H3PO4. The isotopic analysis was conducted with a Thermo Finnegan Triton Thermal Ionization Mass Spectrometer. Values were corrected to 0.710248 for the NBS 987 standard. Reproducibility did not exceed 0.000013.

Results

All five samples yielded nearly identical 87Sr/86Sr ratios in the narrow range from 0.71027 to 0.71033.

Interpretation

The 87Sr/86Sr ratios are substantially higher than Late Pennsylvanian (Kasimovian) marine values, which are in the range of 0.7080–0.7085 (Veizer et al., 1999; Prokoph et al., 2008). They are lower than those seen in rivers draining granitic cratons such as the Laurentian craton (0.725–0.740), however, that formed the Pennsylvanian hinterland for the Illinois Basin (Aberg and Wickman, 1987). Nonetheless, given that the Sr content of marine water is typically around 100 times greater than that of freshwater, the 87Sr/86Sr ratios obtained for our samples imply a strong freshwater influence (Schmitz et al., 1991), assuming that they have not reequilibrated. One important caveat to this interpretation is that the Illinois Basin comprised part of a large, epeiric sea that was poorly mixed and dysaerobic (Heckel, 2008). In such analogous present-day environments as the Baltic Sea where salinity is near marine to brackish, 87Sr/86Sr ratios are also dominated by freshwater sources. Ratios rise from FIGURE 11—Chart showing the mean abrasion index for each vertebrate element in the Friendsville Mine assemblage (horizontal bars plot the standard deviation). The 0.7092 at the near marine (35.29%o salinity) end of the system to 0.7097 mean abrasion value (1.21) for the whole assemblage is shown. at the near freshwater end (2.46%o salinity), and attain values comparable to our specimens within brackish estuaries (Andersson et al., 1992, 1994). Thus, the ratios we have obtained for the xenacanth sharks are more supportive of a brackish salinity rather than There are a number of uncertainties with the application of the fish freshwater. paleosalinity technique, however. Schmitz et al. (1991) noted that 87 86 postdepositional isotopic exchange might cause Sr/ Sr ratios of fossil FISH SALINITY TOLERANCE INFERRED FROM FACIES material to reequilibrate with pore waters. Later, Schmitz et al. (1997) confirmed that apatite and dentine are indeed subject to reequilibration, Further verification that the fishes were residents of brackish habitats but also showed that more inert substances such as tooth enamel are comes from a critical analysis of their facies context and co-occurring much more resistant to this process. Further work by Becker et al. invertebrate assemblages. (2008) on the teeth of Upper Cretaceous sharks supported this conclusion; isotopic analysis of dentine and enameloid from individual Xenacanth Sharks teeth gave conflicting results, and when these signatures were used for radiometric dating only the enameloid gave biostratigraphically Although xenacanths have been traditionally considered to be consistent results. freshwater sharks (Zangerl, 1981; Masson and Rust, 1984; Gray, 1988), recent studies have questioned this interpretation (Calder, 1998; Methods Soler-Gijo´n, 1999; Schultze and Soler-Gijo´n, 2004; Schultze, 2009). The concept of freshwater xenacanths seems to have arisen from their We analyzed five shark samples for 87Sr/86Sr composition as follows: association with Pennsylvanian plant-bearing Coal Measures (Hampe, Orthacanthus (n 5 1), Sphenacanthus (n 5 2) and elasmobranch scales 2003), especially in basins lacking such stenohaline fauna as Saar-Nahe, (n 5 2). There was insufficient material to separate enameloid from Germany (Boy, 1998) and Joggins, Nova Scotia (Calder, 1998). New dentine and apatite—Orthacanthus does not possess enameloid in any data, however, indicate that shark-bearing facies at Joggins were case—so bulk samples were submitted for analysis. This weakens the brackish, based on trace fossil assemblages and foraminifera (Archer et significance of our measurements because we cannot exclude the al., 1995), or even near-stenohaline based on the co-occurrence of the possibility that samples have been reequilibriated by groundwater. bivalve, Curvirimula (Calver, 1968; Falcon-Lang, 2005; Falcon-Lang et Samples were first cleaned with ultrapure (Milli-Q) water, then al., 2006), and the recent report of brachiopods and echinoids (Grey et ultrasonicated for 3 minutes; this cycle was repeated twice more. After al., 2011), while marine bands with glauconite have now been cleaning, the samples were dissolved in 3 ml 7N HNO3, dried, and recognized in shark-bearing units in the Saar-Nahe basin (Kra¨tschmer redissolved in 2 ml 3N HNO3. This solution was ultrasonicated for and Forst, 2005; Schultze, 2009). 15 minutes, then transferred to a centrifuge for 5 minutes. An aliquot of In addition, many reports have shown that xenacanths are the solution representing 3 mg of solid was removed, and made up to intimately associated with brackish and marine fauna (e.g., Bardack, 0.5 ml using 3N HNO3. Separation of Sr was achieved with standard 1979; Schultze, 1985, 1995; Thayer, 1985; Calder, 1998; Soler-Gijo´n, ion exchange chromatography using 70 ml of Eichrom Sr specific resin 1999; Schultze and Soler-Gijo´n, 2004). For example, 75% of (50–100 mm). The Sr was eluted with 1.5 ml of Milli-Q water. The Xenacanthus and 35% of Orthacanthus—two taxa that we report PALAIOS PENNSYLVANIAN FISH AND TETRAPOD FAUNA FROM ILLINOIS, USA 651 from Friendsville, Illinois—occur in facies association with stenoha- osmotic status at a hyperosmotic level relative to seawater through urea line marine invertebrates according to data compilations (Schultze retention, and this is linked to protein function. Thus, when they and Chorn, 1997). Further, investigations of growth increments in encounter freshwater, they must synthesize urea to counteract water specimens of Orthacanthus from Spain suggest that this shark lived in absorption. a coastal environment subject to tidal cyclicity (Soler-Gijo´n, 1999), Ballantyne and Robinson (2010) identify three hypothetical stages which is consistent with brackish to stenohaline salinity, though not in the colonization of freshwater and brackish water environments by indicative. elasmobranchs: (1) euryhaline forms that periodically entered brackish Thus, the interpretation that best fits these data is that xenacanths to freshwater settings, (2) permanent freshwater and brackish water were euryhaline, capable of living in marine, brackish, and perhaps, forms that retained the capability to live in marine settings, and (3) freshwater environments. Ironically, given the former dominance of the obligate freshwater forms. This transition involves, first, the freshwater xenacanth concept, the weakest support is for unequivocally reduction, and then, the elimination of the urea retention osmoreg- freshwater habitats, in part, because it is difficult to positively identify ulatory strategy. All the evidence suggests that Paleozoic xenacanth such settings in the geological record (Grey, 1988). Euryhaline and euselachian sharks were euryhaline forms (Stage 1); certain capability would explain the cosmopolitan distribution of certain species of Orthacanthus may have been permanent brackish water xenacanth groups (Hampe and Ivanov, 2007), and the close similarities residents (Stage 2), though if they were, they clearly inhabited the that exist between some geographically distant species, e.g., Ortha- tidal zone (Soler-Gijo´n, 1999). canthus compressus and O. bohemicus (Johnson, 1999). If xenacanths were exclusively freshwater taxa, one would expect to see a high degree Other Fishes of endemism—and this is not the case. In his review, Hampe (2003) showed that most xenacanth taxa (e.g., The other fishes in the Friendsville assemblage are generally not Anodontacanthus-Dicentrodus, Triodus and half of Orthacanthus and known in sufficient systematic detail to allow for analysis of salinity Xenacanthus species) disappeared from the British Coal Measures tolerance. Possible exceptions are rare chimaeroid fishes that include during the Bolsovian substage. This stratigraphic level, which helodontids and other indeterminate holocephalans. Holocephalans coincides with the last stenohaline marine incursion in the British have traditionally been considered a strictly marine group (Moy- Pennsylvanian (Waters and Davies, 2006), also witnessed the local Thomas and Miles, 1971). Helodontids were durophagous bottom extirpation of such near-stenohaline bivalves as Curvirimula (Calver, dwellers, and upper Paleozoic examples are known from shallow 1968). This implies that these xenacanth species were tolerant only of marine deposits (Falcon-Lang, 1998; Trinajstic and George, 2009), waters of near-stenohaline salinity. The only taxon that diversified including an occurrence in crinoidal limestone in the Lower Missourian after the Bolsovian was Orthacanthus, suggesting that it tolerated Bond Formation of Illinois (Brusatte, 2007), of very similar age and salinities closer to the freshwater end of the spectrum. This tolerance paleogeographic context to our material. for fresh- to brackish water is supported, to a limited degree, by the facies context of various Orthacanthus species in North America and PALEOECOLOGICAL SYNTHESIS Europe (Hanson, 1986; Sander, 1989; Johnson, 1999), for example, O. platypternus is associated with both marine facies (Olson, 1989) and Drawing together all these data from systematics, taphonomy, presumably fresh to brackish alluvial channel deposits (Johnson, 87Sr/86Sr ratios, and critical literature review, building an understanding 1999) in the Permian of Texas. of the paleoenvironments and paleoecology of the Friendsville vertebrate assemblage is possible. Euselachian Sharks Paleoenvironments The salinity tolerance of Sphenacanthus, the only euselachian shark found in our vertebrate assemblage, is less well documented. Dick (1998) Sedimentary facies analysis reveals that the incised channel, which described specimens from the Mississippian Oil Shales of the Midland contains the vertebrate assemblage, and its associated interfluve Valley of Scotland, and interpreted the habitat as freshwater or brackish. Vertisol originated during a glacial lowstand when the tropics were Subsequent research on the invertebrate fauna and oxygen isotopes, seasonally dry and sea level was low. The incised channel was infilled however, has shown that this unit was deposited under brackish to near- during subsequent deglaciation when sea level rose and the channel stenohaline conditions, at least in part (Guirdham et al., 2003; Williams flooded, forming a broad estuary (cf. Falcon-Lang et al., 2009). The et al., 2006). Very similar Mississippian oil shale facies in the Horton vertebrate assemblage contained in the channel is possibly of mixed Group of Nova Scotia and New Brunswick, which also contains taphonomic origin with some components derived from seasonally dry, euselachian sharks (Calder, 1998), are now considered brackish water terrestrial settings (reworked from the lowstand Vertisol) and others deposits (Tibert and Scott, 1999; Rygel et al., 2006). Thus, the likelihood from brackish coastal waters (transgressive shales containing brackish is that these euselachians were euryhaline sharks, though generally spirorbids). The fishes almost certainly inhabited the brackish estuary preferring lower levels of salinity than xenacanths (Maisey, 1982). formed during transgression, based on literature review, abrasion data, and 87Sr/86Sr ratios. What is less certain is whether the tetrapods also Elasmobranch Salinity Tolerance lived in, and around, the estuary or whether their remains have been reworked from the lowstand Vertisol. Consequently we show pelyco- Today, there are only 43 species of elasmobranchs in ten genera and saurs on both lowstand and transgressive reconstructions (Fig. 12) four families that have a reported tolerance to brackish and freshwater (5% of all elasmobranch species; Helfman et al., 1997), and only a Community Ecology proportion of these are permanent residents of freshwater and brackish water environments (Martin, 2005). This compares to 41% of teleost Although sharks and actinopterygian fish dominate our assemblage fishes that have an exclusively freshwater mode of life. The (97.0% of specimens), with rare chimaeroids and tetrapods, it is osmoregulatory strategy employed by elasmobranchs is cited as the unlikely that this taphocoenosis accurately reflects original community main barrier that has prevented their more extensive invasion of composition, as a result of taphonomic bias (Korth, 1979; Maas, 1985). freshwater and brackish water environments (Ballantyne and Robin- In addition, there may be collector bias (250 mm sieve size), resulting in son, 2010). In contrast to teleosts, elasmobranchs maintain their undersampling of smaller members of the fish community—e.g., small 652 CARPENTER ET AL. PALAIOS

Kansas; Essex, Illinois; Braidwood, Illinois; Linton, Ohio; Montceau- les-Mines, France; Ny´rˇany, Czech Republic; Lawrence, Kansas; summarized in Schultze and Maples, 1992; Robb, 1992) shares most similarities with the Upper Pennsylvanian Hamilton Quarry Lager- sta¨tte, Kansas, United States, which also contains many xenacanth and euselachian sharks, actinopterygian fishes, and rare tetrapods (Schultze et al., 1993) and comprises a similar geological setting—an incised paleovalley drowned by eustatic sea-level rise during deglaci- ation (Cunningham et al., 1993).

Food Webs

Within the limits imposed by taphonomy, the Friendsville Coal food web can be roughly reconstructed (Fig. 13). The actinopterygians would have mostly fed on such invertebrate prey as arthropods and molluscs, and some perhaps on water plants. The mid- (Triodus)to large-sized (e.g., Orthacanthus, Xenacanthus) sharks, #1–2 m long, would have likely fed on such smaller fishes as the actinopterygians (Williams, 1972; Hampe, 1988). Olson (1971) suggested that, in addition to fishes, Orthacanthus might have preyed opportunistically on amphibians and reptiles that came down to drink at the water’s edge. Further, the dramatic discovery of a three-step trophic chain, a lower Permian xenacanth shark, containing in its stomach two temnospondyl amphibians, one of which had devoured an acanthodian fish (Kriwet et al., 2008), confirms that these sharks had catholic tastes. These predatory fishes, however, did not eat exclusively pelagic prey: coprolites and gut contents of fully marine upper Paleozoic holoce- phalans, sharks, and other fishes also contain fragmented brachiopods and crinoid ossicles (Brett and Walker, 2002), and there is no reason that shallow marine fishes of these kinds would not also have dredged benthic organisms as part of their diet. The holocephalan Helodus, with broad tooth plates, is widely regarded as a shell crusher (Brett and Walker, 2002). The rare tetrapods from Friendsville, including, or exclusively, such sphenacodontid pelycosaurs (amniotes) as Haptodus or Dimetrodon, were fully terrestrial animals (e.g., Olson, 1952, 1971; Olson and Vaughn, 1970). Early amniotes, in general, were dryland adapted as implied by footprint records in red beds (Falcon-Lang et al., 2007, 2010). Further, the origin of amniotes is associated with a break from dependence on the water for egg-laying, and phylogenetic reconstructions of ancestral states suggest low osmotic tolerance, and so dryland and freshwater FIGURE 12—Schematic reconstruction of the incised channel and its flora and fauna adaptation, at the base of the clade (Garland et al., 1997). This is during A, lowstand and B, transgressive phases (modified from Falcon-Lang et al., consistent with the facies distribution of basal synapsid eupelycosaurs 2009). Vertical scale (incised channel) is exaggerated 32 for clarity and adjusted for and sphenacodontids, which are interpreted largely as terrestrial red bed the effects of 32 sediment compaction. deposits, both in the Pennsylvanian and lower Permian of North America and Europe (Reisz, 1986). These inferences lend some support to the hypothesis that our tetrapods may have been reworked from the lowstand Vertisol and inhabited the seasonal drylands that developed actinopterygians and acanthodians, which may not have been identified during glacial maxima (Falcon-Lang et al., 2009). because of the minute size of their teeth and scales. Thus, Friendsville assemblage cannot be demonstrated to be an accurate sample of a single Evolutionary Context of Pelycosaurs original ecosystem. This means that detailed community reconstruction, comparison Importantly, the finding of likely pelycosaur remains at Friendsville with other assemblages, and application of quantitative paleoecolog- coincides with a dramatic upheaval in continental ecosystems. Olson ical measures—e.g., Margalef diversity, a diversity, dominance, (1952, 1971) long identified a multistep process of full terrestrialization relative abundance—are not meaningful for our assemblage (Eberth, of Permian–Carboniferous tetrapod ecosystems, with the increasing 1990; Schultze, 1995, Baird and Maples, 1997). To directly compare significance of amniotes (reptiles) as climates and floras became ever the quantitative composition of our assemblage with other localities more dryland adapted. The first step occurred in the early would require a taphonomic investigation of each site to evaluate the Kasimovian, toward the end of the Pennsylvanian, when sphenaco- degree of isotaphonomy (Blob and Fiorillo, 1997; Moore and dontids became relatively abundant, and further steps toward modern- Norman, 2009) and our fragmentary material is not amenable to style terrestrial ecosystems happened in the early Permian. This such an assessment. Nonetheless, at a qualitative level, the faunal important series of events has now been tied firmly to the rapid assemblage from Friendsville, when compared with those of ten other disappearance of the luxuriant Coal Forests of the Pennsylvanian contemporary Pennsylvanian sites (Kinney Brick Company Quarry, (Sahney et al., 2010), and places the Friendsville deposit at a critical New Mexico; Hamilton Quarry, Kansas; Robinson, Kansas; Garnett, time in ecosystem evolution. PALAIOS PENNSYLVANIAN FISH AND TETRAPOD FAUNA FROM ILLINOIS, USA 653

FIGURE 13—Paleoecological reconstruction of the incised channel during transgressive phase when it was flooded with brackish water (modified from Falcon-Lang et al., 2006). Key to flora and fauna: 1 5 Macroneuropteris tree with mangrove habit; 2 5 Cordaitalean tree with mangrove habit; 3 5 calamiteans; 4 5 herbaceous lycopsids; 5 5 spirorbids encrusting living and dead plant matter; 6 5 bivalve shoals; 7 5 euselachian shark; 8 5 palaeoniscoid fish; 9 5 Dimetrodon-like reptiles. The relative abundance of taxa cannot be determined for our fragmentary assemblage; the intention of the reconstruction is to illustrate the general setting rather than capture details of the community ecology.

CONCLUSIONS illustrate the importance of precise facies context for paleoecological studies. 1. A new vertebrate assemblage is reported from the Upper Pennsylvanian (mid- to upper Kasimovian) Cohn Coal Member of ACKNOWLEDGMENTS the Mattoon Formation of Illinois, United States. 2. The assemblage occurs within a conglomerate-filled channel, We thank Vigo Coal Company for generous access to the Friendsville incised into a lowstand vertic paleosol. The paleosol and channel was Mine, Illinois, over several years. HFL gratefully acknowledges a formed during a glacial phase, but the vertebrate assemblage occurs in a Natural Environment Research Council (NERC) Advanced Fellowship transgressive fill, interpreted as a ravinement deposit, formed as (NE/F014120/2) and DC acknowledges financial support from the Bob deglaciation-driven, sea-level rise created a broad estuary. Savage Fund of the University of Bristol. We thank Remmert Schouten 3. Xenacanth and euselachian sharks overwhelmingly dominate our and Pedro Viegas (fossil preparation), Stuart Kearns (SEM), Tim vertebrate assemblage (73.5%), which also includes common actinop- Elliott, Claudia Gerling and Carolyn Taylor (Sr analysis), and Simon terygian (23.5%) and rare chimaeroids and tetrapods, especially Powell (photography). The following kindly gave specialist advice on pelycosaurs (together totalling only 1.6%). fossil identification: Gilles Cuny, Michael Coates, and Michel Ginter 4. Taphonomic and isotopic data, combined with a critical literature, (actinopterygians, chondrichthyans), Oliver Hampe and Alexander suggest that the fish-dominated assemblage occupied brackish water Ivanov (chondrichthyans, xenacanthids), Matt Friedman (sarcopter- habitats in the estuary. The presumed ecological tolerance of rare ygians), Robert Reisz (pelycosaurs), Stuart Sumida, David Berman, pelycosaurs, however, implies that these specimens may have inhabited Marcello Ruta, Jenny Clack, and Johannes Mu¨ller (tetrapods), Olev seasonal drylands represented by the lowstand paleosol. Vinn and Paul Taylor (spirorbids), and Cortland Eble (megaspores). 5. The findings improve knowledge of the paleoecology of Pennsyl- We especially thank Bill DiMichele and the Smithsonian Institution, vanian vertebrate ecosystems during an evolutionary phase and Washington, United States for the generous loan of material used in 654 CARPENTER ET AL. PALAIOS this study. We also thank Steve Brussatte and an anonymous reviewer BRETT, C.E., and WALKER, S.E., 2002, Predators and predation in Paleozoic marine for their detailed and constructive reviews, and Steve Hasiotis, for his environments: Paleontological Society Papers, v. 8, p. 93–118. meticulous editorial work. BROILI, F., 1913, The skull elements of the Permian Tetrapoda in the American Museum of Natural History: Bulletin of the American Museum of Natural History, v. 32, p. 315–386. BRUSATTE, S.L., 2007, Pennsylvanian (Late Carboniferous) chondrichthyans from the REFERENCES LaSalle Limestone Member (Bond Formation) of Illinois, USA: Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Abhandlungen, v. 244, p. 1–8. 87 86 ABERG, F., and WICKMAN, F.E., 1987, Variations of Sr/ Sr in water from streams BUATOIS, L.A., MANGANO, M.G., MAPLES, C.G., and LANIER, W.P., 1998, Ichnology discharging into the Bothnian Bay, Baltic Sea: Nordic Hydrology, v. 18, p. 33–42. of an Upper Carboniferous fluvio-estuarine paleovalley: The Tonganoxie AGASSIZ, L., 1837–1843, Recherches sur les Poissons fossiles, Tome III. Contenant Sandstone, Buildex Quarry, Eastern Kansas, USA: Journal of Paleontology, v. l’Histoire de l’Ordre des Placoı¨des: Petitpierre, Neuchaˆtel et Soleure, Switzerland, 72, p. 152–180. 390 p. BUATOIS, L.A., GINGRAS, M.K., MACEACHERN, J., MANGANO, M.G., ZONNEVELD, J.P., ANDERSSON, P.S., WASSERBURG, G.J., and INGRI, J., 1992, The sources and transport of PEMBERTON, S.G., NETTO, R.G., and MARTIN, A., 2005, Colonization of brackish- Sr and Nd isotopes in the Baltic Sea: Earth and Planetary Science Letters, v. 113, p. water systems through time: Evidence from the trace-fossil record: PALAIOS, v. 459–472. 20, p. 321–347. ANDERSSON, P.S., WASSERBURG, G.J., INGRI, J., and STORDAL, M.C., 1994, Strontium, CALDER, J.H., 1998, The Carboniferous evolution of Nova Scotia, in Blundell, D.W., dissolved and particulate loads in fresh and brackish waters: The Baltic Sea and and Scott, A.C., eds., Lyell: The Past is the Key to the Present: Geological Society Mississippi Delta: Earth and Planetary Science Letters, v. 124, p. 195–210. of London, Special Publication, v. 143, p. 261–302. ARCHER, A.W., CALDER, J.H., GIBLING, M.R., NAYLOR, R.D., REID, D.R., and CALVER, M.A., 1968, Invertebrate faunas, in Murchison, D., and Westoll, T.S., eds., WIGHTMAN, W.G., 1995, Invertebrate trace fossils and agglutinated foraminifera as Coal and Coal-bearing Strata: Oliver & Boyd, Edinburgh, p. 147–177. indicators of marine influences within the classic Carboniferous section at Joggins, COMPAGNO, L.J.V., 1970, Systematics of the genus Henritriakis (Selachii; Carcha- Nova Scotia, Canada: Canadian Journal of Earth Sciences, v. 32, p. 2027–2039. rinidae) and related genera: California Academy of Sciences Proceedings, v. 38, p. ARRATIA, G., SCHULTZE, H.-P., and CASCIOTTA, J., 2001, Vertebral column and 63–98. associated elements in dipnoans and comparison with other fishes: Development COOK, E., 1995, Taphonomy of two Lower Cretaceous bone accumulations from and homology: Journal of Morphology, v. 250, p. 101–172. Southeastern England: Palaeogeography, Palaeoclimatoogy, Palaeoecology, v. 116, BAIRD, G.C., 1997a, Geologic setting of the Mazon Creek area fossil deposit, in p. 263–270. Shabica, C.W., and Hay, A.A., eds., Richardson’s Guide to the Fossil Fauna of COPE, E.D., 1887, Geology and Paleontology: American Naturalist, v. 21, p. 1014– Mazon Creek: Northeastern University, Chicago, p. 16–20. 1019. BAIRD, G.C., 1997b, Fossil distribution and fossil associations, in Shabica, C.W., and CUNNINGHAM, C.R., FELDMAN, H.R., FRANSEEN, E.K., GASTALDO, R.A., MAPES,G., Hay, A.A., eds., Richardson’s Guide to the Fossil Fauna of Mazon Creek: MAPLES, C.G., and SCHULTZE, H.-P., 1993, The Upper Carboniferous Hamilton Northeastern University, Chicago, p. 21–26. Fossil-Lagersta¨tte in Kansas: A valley-fill, tidally influenced deposit: Lethaia, v. 26, p. 225–236. BAIRD, G.C., 1997c, Paleoenvironmental setting of the Mazon Creek biota, in Shabica, C.W., and Hay, A.A., eds., Richardson’s Guide to the Fossil Fauna of DASCH, E.J., and CAMPBELL, K.S.W., 1970, Strontium-isotope evidence for marine or Mazon Creek: Northeastern University, Chicago, p. 35–51. freshwater origin of fossil Dipnoans and Arthrodires: Nature, v. 227, p. 1159. DAVIS, J.W., 1881, On Anodontocanthus, a new genus of fossil fishes from the coal- BAIRD, G.C., and MAPLES, C.G., 1997, Regional, temporal, and evolutionary significance of the Mazon Creek biota, in Shabica, C.W., and Hay, A.A., eds., measures: With descriptions of three new species: Quarterly Journal of the Richardson’s Guide to the Fossil Fauna of Mazon Creek: Northeastern University, Geological Society of London, v. 37, p. 427–429. Chicago, p. 52–72. DAVYDOV, V.I., CROWLEY, J.L., SCHMITZ, M.D., and POLETAEV, V.I., 2010, High- precision U-Pb zircon age calibration of the global Carboniferous time scale and BAIRD, G.C., SHABICA, C.W., ANDERSON, J.L., and RICHARDSON, E.S., JR., 1985a, Milankovitch-band cyclicity in the Donets Basin, eastern Ukraine: Geochemistry, Biota of a Pennsylvanian muddy coast: Habitats within the Mazonian delta Geophysics, Geosystems, v. 11, p. 1–22. complex, northeastern Illinois: Journal of Paleontology, v. 59, p. 253–281. DEMAR, R., 1970, A primitive pelycosaur from the Pennsylvanian of Illinois: Journal BAIRD, G.C., SROKA, S.D., SHABICA, C.W., and BEARD, T.L., 1985b, Mazon Creek- of Paleontology, v. 44, p. 154–163. type fossil assemblages in the U.S. midcontinent Pennsylvanian: Their recurrent DENISON, R.E., KOEPNICK, R.B., BURKE, W.H., HETHERINGTON, A., and FLETCHER,A., character and paleoenvironmental significance: Philosophical Transactions of the 1993, Construction of the Mississippian, Pennsylvanian and Permian seawater Royal Society of London, B311, p. 87–99. 87Sr/86Sr curve: Chemical Geology (Isotope Geoscience Section), v. 112, p. 145–167. BALLANTYNE, J.S., and ROBINSON, J.W., 2010, Freshwater elasmobranchs: A review of DICK, J.R.F., 1981, Diploselache woodi gen. et sp. nov., an Early Carboniferous shark their physiology and biochemistry: Journal of Comparative Physiology B: from the Midland Valley of Scotland: Transactions of the Royal Society of Biochemical, Systemic and Environmental Physiology, v. 180, p. 475–493. Edinburgh Earth Sciences, v. 72, p. 99–113. BARDACK, D., 1979, Fishes of the Mazon Creek fauna, in Nitecki, M.H., ed., Mazon DICK, J.R.F., 1998, Sphenacanthus, a Palaeozoic freshwater shark: Zoological Journal Creek Fossils: Academic Press, New York, p. 501–528. of the Linnean Society, v. 122, p. 9–25. BECKER, M.A., SEIDEMANN, D.E., CHAMBERLAIN, J.A., BUHL, D., and SLATTERY,W., DIMICHELE, W.A., PFEFFERKORN, H.W., and GASTALDO, R.A., 2001, Response of Late 2008, Strontium isotopic signatures in the enameloid and dentine of upper Carboniferous and Early Permian plant communities to climate change: Annual Cretaceous shark teeth from western Alabama: Paleoecologic and geochronologic Review of Earth Planetary Science, v. 29, p. 461–487. implications: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 264, p. 188–194. DIMICHELE, W.A., FALCON-LANG, H.J., NELSON, J., ELRICK, S., and AMES, P., 2007, BEHRENSMEYER, A.K., 1978, Taphonomic and ecologic information from bone Ecological gradients within a Pennsylvanian mire forest: Geology, v. 35, p. 415–418. weathering: Paleobiology, v. 4, p. 150–162. DIMICHELE, W.A., MONTAN˜ EZ, I.P., POULSEN, C.J., and TABOR, N.J., 2009, Climate BENNETT, C., 2008, A review of the Carboniferous colonization of non-marine and vegetational regime shifts in the Late Paleozoic ice age earth: Geobiology, v. 7, environments by ostracods: Senckenbergiana Lethaea, v. 88, p. 37–46 p. 200–226. BERG, L.S., 1937, A classification of fish-like vertebrates: Bulletin of the Academy of DIMICHELE, W.A., CECIL, B., MONTANEZ, I., and FALCON-LANG, H.J., 2010, Cyclic Science of the URSS, Biology Series, v. 20, p. 1277–1280. changes in Pennsylvanian paleoclimate and it effects on floristic dynamics in BERMAN, D.S., 1970, Vertebrate fossils from the Lueders Formation, Lower Permian tropical Pangaea: International Journal of Coal Geology, v. 83, p. 329–344. of North-Central Texas: University of California Publications in Geological DOLBY, G., FALCON-LANG, H.J., and GIBLING, M.R., 2011, A conifer-dominated Sciences, v. 86, 1–61. palynological assemblage from Pennsylvanian (late Moscovian) alluvial drylands in BEYRICH, E., 1848, U¨ ber Xenacanthus decheni und Holacanthus gracilis, zwei Fische Atlantic Canada: Implications for the vegetation of tropical lowlands during aus der Formation des Rothliegenden in Norddeutschland: Berichte der glacial phases: Journal of the Geological Society, v. 168, p. 571–585. Ko¨niglichen Preussischen Akademie der Wissenschaften zu Berlin, 1848, p. 24–33. DRIESE, S.G., and OBER, E.G., 2005, Paleopedologic and paleohydrologic records of BLOB, R.W., and FIORILLO, A.R., 1997, The significance of vertebrate microfossil size precipitation seasonality from Early Pennsylvanian ‘‘underclay’’ paleosols, U.S.A.: and shape distribution for faunal abundance reconstructions: A Late Cretaceous Journal of Sedimentary Research, v. 75, p. 997–1010. example: Paleobiology, v. 22, p. 422–435. EBERTH, D.A., 1990, Stratigraphy and sedimentology of vertebrate microfossil sites in BONAPARTE, C.L.P., 1832–1841, Iconografia della fauna Italiana per le quattro classi the uppermost Judith River Formation (Campanian), Dinosaur Provincial Park, degli Animali Vertebrati, Volume 3: Salviucci, Rome. 475 p. Alberta, Canada: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 78, p. 1–36. BOY, J.A., 1998, Mo¨glichkeiten und Grenzen einer O¨ kosystem-Rekonstruktion am FALCON-LANG, H.J., 1998, The impact of wildfire on an Early Carboniferous coastal Beispiel des spa¨tpala¨ozoischen lakustrinen pala¨o-O¨ kosystems 1. Theoretische und system, North Mayo, Ireland: Palaeogeography, Palaeoclimatology, Palaeoecol- methodische Grundlagen: Pala¨ontologische Zeitschrift, v. 72, p. 207–240. ogy, v. 139, p. 121–138. PALAIOS PENNSYLVANIAN FISH AND TETRAPOD FAUNA FROM ILLINOIS, USA 655

FALCON-LANG, H.J., 2004, Pennsylvanian tropical rainforests responded to glacial- HAMPE, O., 2003, Revision of the Xenacanthida (Chondrichthyes: Elasmobranchii) interglacial rhythms: Geology, v. 32, p. 689–692. from the Carboniferous of the British Isles: Transactions of the Royal Society of FALCON-LANG, H.J., 2005, Small cordaitalean trees in a marine-influenced coastal Edinburgh: Earth Sciences, v. 93, p. 191–237. habitat in the Pennsylvanian Joggins Formation, Nova Scotia, Canada: Journal of HAMPE, O., and IVANOV, A., 2007, Bransonelliformes: A new order of the the Geological Society of London, v. 162, p. 485–500. Xenacanthimorpha (Chondrichthyes, Elasmobranchii): Fossil Record, v. 10, p. FALCON-LANG, H.J., and DIMICHELE, W.A., 2010, What happened to the Coal Forests 190–194. during Pennsylvanian glacial phases?: PALAIOS, v. 25, p. 611–617. HAMPSON, G., STOLLHOFEN, H., and FLINT, S., 1999, A sequence stratigraphic model FALCON-LANG, H.J., BENTON, M.J., BRADDY, S.J., and DAVIES, S.J., 2006, The for the Lower Coal Measures (Upper Carboniferous) of the Ruhr district, Pennsylvanian tropical biome reconstructed from the Joggins Formation of northwest Germany: Sedimentology, v. 46, p. 1199–1231. Canada: Journal of the Geological Society of London, v. 163, p. 561–576. HANSON, M.C., 1986, Microscopic chondrichthyan remains from Pennsylvanian FALCON-LANG, H.J., BENTON, M.J., and STIMSON, M., 2007, Ecology of early reptiles marine rocks of Ohio and adjacent areas: Unpublished Ph.D. thesis, Ohio State inferred from Lower Pennsylvanian trackways: Journal of the Geological Society University, Columbus, Ohio, 536 p. of London, v. 164, p. 1113–1118. HECKEL, P.H., 2008, Pennsylvanian cyclothems in Midcontinent North America as FALCON-LANG, H.J., NELSON, J., ELRICK, S., LOOY, C., AMES, P., and DIMICHELE, far-field effects of waxing and waning of Gondwana ice sheets, in Fielding, C.R., W.A., 2009, Incised channel fills containing conifers indicate that seasonally- Frank, T.D., and Isbell, J.L., eds., Resolving the Late Paleozoic Ice Age in Time dry vegetation dominated Pennsylvanian tropics lowlands: Geology, v. 37, p. 923– and Space: Geological Society of America Special Paper, v. 441, p. 275–289. 926. HECKEL, P.H., ALEKSEEV, A.S., BARRICK, J.E., BOARDMAN, D.R., GOREVA, N.V., FALCON-LANG, H.J., GIBLING, M.R., BENTON, M.J., MILLER, R.F., and BASHFORTH, NEMYROVSKA, T.I., UENO, K., VILLA, E., and WORK, D.M., 2007, Cyclothem A.R., 2010, Diverse tetrapod trackways in the Lower Pennsylvanian Tynemouth [‘‘digital’’] correlation and biostratigraphy across the global Moscovian-Kasimo- Creek Formation, southern New Brunswick, Canada: Palaeogeography, Palaeo- vian-Gzhelian stage boundary interval (Middle-Upper Pennsylvanian) in North climatology, Palaeoecology, v. 296, p. 1–13. America and Eastern Europe: Geology, v. 35, p. 607–610. FALCON-LANG, H.J., HECKEL, P., DIMICHELE, W.A., BLAKE, B.M., EASTERDAY,C., HELFMAN, G.S., COLLETTE, B.B., and FACEY, D.E., 1997, The Diversity of Fishes: EBLE, C., ELRICK, S., GASTALDO, R.A., GREB, S.F., MARTINO, R.L., NELSON, W.J., Blackwell, Malden, Massachusetts, 528 p. FEFFERKORN HILLIPS OSSCOE P , H.W., P , T.L., and R , S.J., 2011, No evidence for a HOOK, R.W., and BAIRD, D., 1988, An overview of the Upper Carboniferous fossil major unconformity at the Desmoinesian-Missourian boundary in North America: deposit at Linton, Ohio: Ohio Journal of Science, v. 88, p. 55–60. PALAIOS, v. 26, p. 125–139. HOOK, R.W., and FERM, J.C., 1988, Paleoenvironmental controls on vertebrate- FELDMAN, H.R., FRANSEEN, E.K., and JOECKEL, R.M., 2005, Impact of longer-term bearing abandoned channels in the Upper Carboniferous: Palaeogeography, modest climate shifts on architecture of high-frequency sequences (cyclothems) in Palaeoclimatology, Palaeoecology, v. 63, p. 159–181. the Pennsylvanian of midcontinent USA: Journal of Sedimentary Research, v. 75, HUXLEY, T.H., 1880, On the application of the laws of evolution to the arrangement p. 350–368. of the vertebrate and more particularly the Mammalia: Proceedings of the FIELDING, C.R., FRANK, T.D., BIRGENHEIER, L.P., RYGEL, M.C., JONES, A.T., and Scientific Meetings of the Zoological Society of London, v. 188, p. 649–662. ROBERTS, J., 2008a, Stratigraphic imprint of the Late Palaeozoic Ice Age in eastern JANVIER, P., 1978, On the oldest known teleostome fish Andreolepis hedei Gross Australia: A record of alternating glacial and non-glacial climate regime: Journal (Ludlow of Gotland), and the systematic position of the lophosteids: Eesti NSV of the Geological Society of London, v. 165, p. 129–140. Teaduste Akadeemia Toimetis, v. 27, p. 88–95. FIELDING, C.R., FRANK, T.D., and ISBELL, J.L., 2008b, The late Paleozoic ice age: A JANVIER, P., 1996, Early Vertebrates: Oxford University Press, New York, 393 p. review of current understanding and synthesis of global climate patterns, in JOHNSON, G.D., 1999, Dentitions of Late Paleozoic Orthacanthus species and new Fielding, C.R., Frank, T.D., and Isbell, J.L., eds., Resolving the Late Paleozoic Ice species of ?Xenacanthus (Chondrichthyes: Xenacanthiformes) from North Amer- Age in Time and Space: Boulder, Colorado, Geological Society of America Special ica: Acta Geologica Polonica, v. 49, p. 215–266. Publication, v. 441, p. 343–354. JOHNSON, G.D., 2003, Dentitions of Barbclabornia (new genus, Chondrichthyes, FRITSCH, A., 1889, Fauna des Gaskohle und der Kalksteine der Permformation Xenacanthiformes) from the Upper Palaeozoic of North America: Mitteilungen Bo¨hmens, vol. 2: Rˇ ivna´cˇ, Prague, p. 93–114. der Museum der Naturkunde, Berlin, Geowissenschaftliche Reihe, v. 6, p. 125–160. GARDINER,B.G.,SCHAEFFER,B.,andMASSERIE, J.A., 2005, A review of lower acti- nopterygian phylogeny: Zoological Journal of the Linnean Society, v. 144, p. 511–525. JOHNSON, G.D., and THAYER, D.W., 2009, Early Pennsylvanian xenacanth chon- drichthyans from the Swisshelm Mountains, Arizona, USA: Acta Palaeontologica GARLAND, T., JR., MARTIN, K.L.M., and DI´AZ-URIARTE, R., 1997, Reconstructing ancestral trait values using squared-change parsimony: Plasma osmolarity at the Polonica, v. 54, p. 649–668. origin of amniotes, in Sumida, S., and Martin, K., eds., Amniote Origins: JORDAN, H., 1849, Triodus sessilis ein neuer Fisch der Kohlenformation von Lebach: Completing the Transition to Land: Academic Press, London, p. 425–501. Neues Jahrbuch fu¨r Minoralogie, Geologie und Pala¨ontologie, 1849, p. 843. KEMP, T.S., 1982, Mammal-like Reptiles and the Origin of Mammals: Academic GASTALDO, R.A., 1981, An ultrastructural and taxonomic study of Valvisporites auritus (Zerndt) Bhardwaj, a lycopsid megaspore from the Middle Pennsylvanian Press, London, 363 p. of southern Illinois: Micropaleontology, v. 27, p. 84–93. KISSEL, R.A., and LEHMAN, T.M., 2002, Upper Pennsylvanian tetrapods from the Ada GASTALDO,R.A,STEVANOVIC-WALLS, I.M., and WARE, W.N., 2004, Erect forests are Formation of Seminole County, Oklahoma: Journal of Paleontology, v. 76, p. 529– evidence for coseismic base-level changes in Pennsylvanian cyclothems of the Black 545. Warrior Basin, USA, in Pashin, J.C., and Gastaldo, R.A., eds., Sequence KORTH, W.W., 1979, Taphonomy of microvertebrate assemblages: Annals of the Stratigraphy, Paleoclimate, and Tectonics of Coal-Bearing Strata: AAPG (American Carnegie Museum, v. 48, p. 235–285. Association of Petroleum Geologists) Studies in Geology, v. 51, p. 219–238. KOSANKE, J.A., SIMON, H.R., WANLESS, H.B., and WILLMAN, H., 1960, Classification GINTER, M., HAMPE, O., DUFFIN, C., and SCHULTZE, H.-P., eds., 2010, Paleozoic of the Pennsylvanian strata of Illinois: Illinois State Geological Survey, Report of Elasmobranchii: Teeth: Handbook of Paleoichthyology, vol. 3D: Verlag Dr. Investigation, v. 214, p. 5–84. Friedrich Pfeil, Mu¨nich, Germany, p. 1–168. KRA¨ TSCHMER,K.,andFORST, M.H., 2005, Ein Neufund von assoziierten Skelettele- GOLDFUSS, G.A., 1847, Beitra¨ge zur vorweltlichen Fauna des Steinkohlengebirges: menten von Sphenacanthus cf. carbonarius (Giebel 1848) aus dem Top Oberkarbon Henry and Cohen, Bonn, Germany, 28 p. (Breitenbach-Fm., Stefan C) des Saar–Nahe Beckens (SW-Deutschland) mit GRADSTEIN, F., OGG, J., and SMITH, A., 2004, A Geologic Time Scale 2004: Cambridge Anmerkungen zur Pala¨oo¨kologie der Vertebraten-Lokalita¨ten von Altenkirchen: University Press, Cambridge, UK, 610 p. Geowissenschaftliche Beitra¨ge zum Saarpfa¨lzischen Rotliegenden, v. 3, p. 29–37. GRAY, J., 1988, Evolution of the freshwater ecosystem: The fossil record: KRIWET, J., WITZMANN, F., KLUG,S.,andHEIDTKE, U.H., 2008, First direct evidence Palaeogeography, Palaeoclimatology, Palaeoecology, v. 62, p. 1–214. of a vertebrate three-level trophic chain in the fossil record: Proceedings of the GREY, M., PUFAHL, P.K., and AZIZ, A.A., 2011, Using multiple environmental proxies Royal Society, Biological Sciences B, v. 275, p. 181–186. to determine degree of marine influence and paleogeographic position of the KVALE, E.P., ARCHER, A.W., and JOHNSON, H.R., 1989, Daily, monthly, and yearly Joggins Fossil Cliffs, UNESCO World Heritage Site: PALAIOS, v. 26, p. 265–263. tidal cycles within laminated siltstones of the Mansfield Formation (Pennsylva- GUIRDHAM, C., ANDREWS, J.E., BROWNE, M.A.E., and DEAN, M.T., 2003, nian) of Indiana: Geology, v. 17, p. 365–368. Stratigraphic and palaeoenvironmental significance of microbial carbonates in LAURIN, M., and SOLER-GIJO´ N, R., 2010, Osmotic tolerance and habitat of early the Asbian Sandy Craig Formation of Fife: Scottish Journal of Geology, v. 39, p. stegocephalians: Indirect evidence from parsimony, taphonomy, palaeobiogeogra- 151–168. phy, physiology and morphology: Geological Society of London, Special HAECKEL, E., 1866, Generelle Morphologie der Organismen, Zweiter Band: Publication, v. 339, p. 151–179. Allgemeine Entwickelungsgeschichte: Georg Reiner, Berlin, 574 p. LEBEDEV, O.A., 1996, Fish assemblages in the Tournaisian-Vise´an environments of HAMPE, O., 1988, U¨ ber die Bezahnung des Orthacanthus (Chondrichthyes: the east European Platform, in Strogen, P., Somerville, I.D., and Jones, G.L., eds., Xenacanthida; Oberkarbon-Unterperm): Pala¨ontologische Zeitschrift, v. 62, p. Recent Advances in Lower Carboniferous Geology: Geological Society of London, 285–296. Special Publication, v. 107, p. 387–417. 656 CARPENTER ET AL. PALAIOS

MAAS, M.C., 1985, Taphonomy of a Late Eocene microvertebrate locality, Wind skeletal apatite in late Paleocene–early Eocene near-coastal sediments, Mississippi: River Basin, Wyoming (USA): Palaeogeography, Palaeoclimatology, Palaeoecol- Chemical Geology, v. 140, p. 275–287. ogy, v. 52, p. 123–142. SCHULTZE, H.-P., 1985, Marine to onshore vertebrates in the Lower Permian of MADDIN, H.C., and REISZ, R.R., 2007, The morphology of the terminal phalanges in Kansas and their paleoenvironmental implications: University of Kansas Permo-Carboniferous synapsids: An evolutionary perspective: Canadian Journal Paleontological Contributions, Paper 113, p. 1–18. of Earth Sciences, v. 44, p. 267–274. SCHULTZE, H.-P., 1995, Terrestrial biota in coastal marine deposits: Fossil-lagersta¨tten MAISEY, J.G., 1982, Studies on the Paleozoic selachian genus Ctenacanthus Agassiz. in the Pennsylvanian of Kansas, USA: Palaeogeography, Palaeoclimatology, No. 2. Bythiacanthus St. John and Worthen, Amelacanthus, New Genus, Palaeoecology, v. 119, p. 255–273. Eunemacanthus St. John and Worthen, Sphenacanthus Agassiz, and Wodnika SCHULTZE, H.-P., 1998, The Fossil Record of the Intertidal Zone, in Horn, M.H., Mu¨nster: American Museum Novitates, v. 2722, p. 1–24. Martin K.L.M., and Chotkowski, M.A., eds., Intertidal Fishes. Life in Two MARTIN, R.A., 2005, Conservation of freshwater and euryhaline elasmobranchs: Worlds: Academic Press, San Diego, p. 373–392. Journal of the Marine Biology Association, v. 85, p. 1049–1073. SCHULTZE, H.-P., 2009, Interpretation of marine and freshwater palaeoenvironments MASSON, A.G., and RUST, B.R., 1984, Freshwater shark teeth as environmental in Permo-Carboniferous deposits: Palaeogeography, Palaeoclimatology, Palaeoe- indicators in the Upper Pennsylvanian coal-bearing succession near Sydney, Nova cology, v. 281, p. 126–136. Scotia: Canadian Journal of Earth Sciences, v. 21, p. 1151–1155. SCHULTZE, H.-P., and CHORN, J., 1986, Palaeoniscoid (Actinopterygii, Pisces) MOORE, J.R., and NORMAN, D.B., 2009, Quantitatively evaluating the sources of vertebrae from the Late Paleozoic of Central North America: Journal of taphonomic biasing of skeletal element abundances in fossil assemblages: Paleontology, v. 60, p. 744–757. PALAIOS, v. 24, p. 591–602. SCHULTZE, H.-P., and CHORN, J., 1997, The Permo-Carboniferous genus Sagenodus MOY-THOMAS, J.A., and MILES, R.S., 1971, Palaeozoic Fishes, 2nd ed.: Chapman and and the beginning of modern lungfish: Contributions to Zoology, v. 67, p. 9–70. Hall, London, 259 p. SCHULTZE, H.-P., and MAPLES, C. G., 1992, Comparison of the Late Pennsylvanian NANCE, R.B., and TREWORGY, C.G., 1981, Strippable Coal Resources of Illinois. Part faunal assemblage of Kinney Brick Company Quarry, New Mexico, with other 8: Central and Southeastern Counties: Illinois Geological Survey, Circular 515, Late Pennsylvanian Lagersta¨tten: New Mexico Bureau of Mines and Mineral 32 p. Resources Bulletin, v. 138, p. 231–242. ´ NEWTON, W.A., and WELLER, J.M., 1937, Stratigraphic studies of Pennsylvanian SCHULTZE, H.-P., and SOLER-GIJON, R., 2004, A xenacanth clasper from the outcrops in part of southeastern Illinois: Illinois Geological Survey Report of ?uppermost Carboniferous-Lower Permian of Buxie´res-les-Mines (Massif Central, Investigations, v. 45, 31 p. France) and the palaeoecology of the European Permo-Carboniferous basins: Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Abhandlungen, v. 232, p. 325–363. OLSON, E.C., 1952, The evolution of a Permian vertebrate chronofauna: Evolution, v. 6, p. 181–196. SCHULTZE, H.-P., MAPLES, C.G., and CUNNINGHAM, C.R., 1993, The Hamilton Quarry Konservat-Lagersta¨tte: Stephanian terrestrial biota in a marginal-marine setting: OLSON, E.C., 1971, Vertebrate Paleozoology: John Wiley and Sons, New York, 839 p. Transactions of the Royal Society of Edinburgh: Earth Sciences, v. 84, p. 443–451. OLSON, E.C., 1989, The Arroyo Formation (Leonardian, Lower Permian) and its vertebrate fossils: Texas Memorial Museum Bulletin, v. 35, p. 1–25. SMITH, A.B., and PATTERSON, C., 1988, The influence of taxonomic method on the perception of patterns of evolution: Evolutionary Biology, v. 23, p. 127–216. OLSON, E.C., and VAUGHN, P.P., 1970, The changes of terrestrial vertebrates during SOLER-GIJO´ N, R., 1997, New discoveries of xenacanth sharks from the Late the Permian of North America: Forma et Functio, v. 3, p. 113–138. Carboniferous of Spain (Puertollano Basin) and Early Permian of Germany ØRVIG, T., 1973, Fossilized fish teeth under the scanning electron microscope: Old (Saar-Nahe basin): Implications for the phylogeny of xenacanthiform and questions under new light: Fauna och Flora Naturhistoriska Riksmuseet, v. 68, p. anacanthous sharks: Neues Jahrbuch fu¨r Geologie und Pala¨ontologie, Abhandlun- 166–173. gen, v. 205, p. 1–31. ØRVIG, T., 1978, Microstructure and growth of the dermal skeleton in fossil SOLER-GIJO´ N, R., 1999, Occipital spine of Orthacanthus (Xenacanthidae, Elasmo- actinopterygian fishes: Nephrotus and Colobodus, with remarks on the dentition in branchii): Structure and growth: Journal of Morphology, v. 242, p. 1–42. other forms: Zoologica Scripta, v. 7, p. 297–326. SOLER-GIJO´ N, R., 2000, Phylogenetic relationships of Lebachacanthidae Soler-Gijo´n, OSBORN, H.F., 1903, The reptilian subclasses Diapsida and Synapsida, and the early 1997 (Xenacanthiformes, Elasmobranchii): Pala¨ontologische Zeitschrift, v. 74, p. history of the Diaptosauria: Memoirs of the American Museum of Natural 636–377. History, v. 1, p. 265–270. STAHL, B.J., 1999, Holocephali, in Schultze, H.-P., ed., Handbook of Paleoichthyol- PATTERSON, C., 1965, The phylogeny of the chimaeroids: Philosophical Transactions ogy 4, Gustav Fischer Verlag, Stuttgart, p. 1–164. of the Royal Society, London, Series B, Biological Sciences, v. 249, p. 101–226. TANDON, S.K., and GIBLING, M.R., 1994, Calcrete and coal in Late Carboniferous PATTERSON, C., 1982, Morphology and interrelationships of primitive Actinopterygian cyclothems of Nova Scotia, Canada: Climate and sea-level changes linked: fishes: American Zoologist, v. 22, p. 41–259. Geology, v. 22, p. 755–758. PEYER, B., 1968, Comparative Odontology: University of Chicago Press, Chicago, 347 p. TAYLOR, P.D., and VINN, O., 2006, Convergent morphology in small spiral worm PIGG, K.B., and ROTHWELL, G.W., 1983, Chaloneria gen. nov.: Heterosporous tubes (‘‘Spirorbis’’) and its palaeoenvironmental implications: Journal of the lycophytes from the Pennsylvanian of North America: Botanical Gazette, v. 144, p. Geological Society of London, v. 163, p. 225–228. 132–147. THAYER, D.W., 1985, New Pennsylvanian lepospondyl amphibians from the PROKOPH, A., SHIELDS, G.A., and VEIZER, J., 2008, Compilation and time-series Swisshelm Mountains, Arizona: Journal of Paleontology, v. 59, p. 684–700. 18 13 87 86 34 analysis of marine carbonate d O, d C, Sr/ Sr and d S databases through TIBERT, N.E., and SCOTT, D.B., 1999, Ostracodes and agglutinated foraminifera as Earth history: Earth Science Reviews, v. 87, p. 113–134. indicators of paleoenvironmental change in an Early Carboniferous brackish bay, REIF, W.E., 1974, Morphogenese und Musterbildung des Hautza¨hnchen: Skelettes Atlantic Canada: PALAIOS, v. 14, p. 246–260. von Heterodontus: Lethaia, v. 7, p. 25–42. TRAQUAIR, R.H., 1881, Notice of new fish remains from the Blackband Ironstone of REISZ, R.R., 1986, Pelycosauria, in Wellnhofer, P., ed., Handbook of Paleoherpetol- Borough Lee near Edinburgh: Geological Magazine, Decade II, v. 8, p. 34–37. ogy, Part 17A: Gustav Fischer Verlag, Stuttgart. 102 p. TRAQUAIR, R.H., 1885, On a specimen of Psephodus magnus from the Carboniferous ROBB, A.J., 1992, Vertebrate fossils from the Lawrence Formation (Douglas Group, limestone of East Kilbride, Lanarkshire: Geological Magazine, Decade II, v. 2, p. Virgilian) in Northeastern Kansas: Transactions of the Kansas Academy of 338–344. Science, v. 95, p. 129–138. TRAQUAIR, R.H., 1888, Notes on Carboniferous Selachii: Proceedings of the Royal RYGEL, M.C., CALDER, J.H., GIBLING, M.R., GINGRAS, M., and MELROSE, C.S.A., Physical Society of Edinburgh, v. 9. 412–421. 2006, Forested Tournaisian swamps in the Horton Group of Atlantic Canada, in TRINAJSTIC, K., and GEORGE, A.D., 2009, Microvertebrate biostratigraphy of Upper Greb, S., DiMichele, W.A., eds., Wetlands Through Time: Geological Society of Devonian (Frasnian) carbonate rocks in the Canning and Carnavon Basins of America Special Paper, v. 399, p. 103–126. Western Australia: Palaeontology, v. 52, p. 641–661. SAHNEY, S., BENTON, M.J., and FALCON-LANG, H.J., 2010, Rainforest collapse VEIZER, J., ALA, D., AZMY, K., BRUCKSCHEN, P., BUHL, D., BRUHN, F., CARDEN, triggered Carboniferous tetrapod diversification in Euramerica: Geology, v. 38, p. G.A.F., DIENER, A., EBNETH, S., GODDERIS, Y., JASPER, T., KORTE, C., PAWELLEK, 1079–1082. F., PODLAHA, O.G., and STRAUSS, H., 1999, 87Sr/86Sr, d13C and d18O evolution of SANDER, P.M., 1989, Early Permian depositional environments and pond bonebeds in Phanerozoic seawater: Chemical Geology, v. 161, p. 59–88. central Archer County, Texas: Palaeogeography, Palaeoclimatology, Palaeoecol- WANLESS, H.R., and WELLER, J.M., 1932, Correlation and extent of Pennsylvanian ogy, v. 69, p. 1–21. cyclothems: Geological Society of America Bulletin, v. 43, p. 1003–1016. SCHMITZ, B., ABERG, G., WERDLIN,L.FOREY, P., and BENDIX-ALMGREEN, S.E., 1991, WATERS, C.N., and DAVIES, S.J., 2006, Carboniferous: Extensional basins, advancing 87Sr/86Sr, Na, F, Sr, and La in skeletal fish debris as a measure of the paleosalinity deltas and coal swamps, in Brenchley, P.J., and Rawson, P.F., eds., The Geology of of fossil–fish habitats: Geological Society of America Bulletin, v. 103, p. 786–794. England: Geological Society, London, p. 173–223. SCHMITZ, B., INGRAIN, L.S., DOCKERY, T.D., and ABERG, G., 1997, Testing 87Sr/86Sr as WILLIAMS, M.E., 1972, The origin of ‘‘spiral coprolites’’: The University of Kansas a paleosalinity indicator on mixed marine, brackish-water and terrestrial vertebrate Paleontological Contributions, v. 59, p. 1–19. PALAIOS PENNSYLVANIAN FISH AND TETRAPOD FAUNA FROM ILLINOIS, USA 657

WILLIAMS, M., LENG, M.J., STEPHENSON, M.H., ANDREWS, J.E., WILKINSON, I.P., ZANGERL, R., 1981, Chondrichthyes 1: Paleozoic Elasmobranchii, in Schultze, H.-P., SIVETER, D.J., HORNE, D.J., and VANNIER, J.M.C., 2006, Evidence that Early ed., Handbook of Paleoichthyology 3A, Gustav Fischer Verlag, Stuttgart, New Carboniferous ostracods colonised coastal flood plain brackish water environ- York, 115 p. ments: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 230, p. 299–318. ZANGERL,R.,andRICHARDSON, E.S., JR., 1963, The paleoecological history of two WILLISTON, S.W., 1912, Primitive reptiles: Journal of Morphology, v. 23, p. 637–666. Pennsylvanian Black Shales: Fieldiana, Geology Memoir, v. 4, p. 1–352. WILLMAN, H.B., ATHERTON, E., BUSCHBACH, T.C., COLLINSON, C.W., FREY, J.C., ZIDEK, J., 1978, New chondrichthyan spines from the Late Paleozoic of Oklahoma: HOPKINS, M.E., LINEBACK, J.A., and SIMON, J.A., 1975, Handbook of Illinois Journal of Paleontology, v. 52, p. 1070–1078. Stratigraphy: Illinois State Geological Survey Bulletin, vol. 95, 261 p. WOODWARD, A.S., 1891, Catalogue of the Fossil Fishes in the British Museum (Natural History), Part II: British Museum (Natural History), London, 574 p. ZANGERL, R., 1979, New chondrichthyans from the Mazon Creek fauna (Pennsyl- vanian) of Illinois, in Nitecki, M.H., ed., Mazon Creek Fossils: Academic Press, New York, p. 449–500. ACCEPTED JULY 27, 2011