The Upper Silurian Salina Group of eastern North America is well known for its thick evaporite successions and hydrocarbon resources. These strata have been assigned to numerous chronostratigraphic schemes within Ohio and Michigan and are currently identified by varying subsurface and outcrop nomenclatural schemes. These chronostratigraphic challenges have persisted for over 50 yr and dramatically inhibit the correlation of events recorded in the Silurian section of eastern North America with the global record of Silurian biogeochemical events. To help resolve the chronostratigraphic correlation of these units, we provide new high-resolution δ13Ccarb chemostratigraphic analyses of a core located in central Ohio for strata assigned to the Greenfield and Tymochtee Formations and integrate existing biostratigraphic, chemostratigraphic, and subsurface geophysical data in western, southern, and eastern Ohio. The new data presented here, integrated for the first time with basinwide subsurface geophysical data, demonstrate a mid-late Homerian Stage global sea-level lowstand, suggest a short interval of tectonic stability within the study area at the beginning of “Salina B–G” deposition, during which accommodation was occupied by the Greenfield Formation and laterally equivalent strata, and provide chronostratigraphic constraints for basin flexure and potential forebulge migration associated with renewed tectonic activity. The new chronostratigraphic correlation of these strata provides a broader picture of Silurian environmental change across the eastern half of the Laurentian paleocontinent.

Upper Silurian strata assigned to the Greenfield and Tymochtee Formations of the Salina Group in Ohio have undergone extensive biostratigraphic and lithostratigraphic analyses for well over a century (Newberry, 1870; Orton, 1871, 1888; Winchell, 1873; Whitfield, 1893; Grabau, 1898, 1900, 1906, 1909; Prosser, 1903; Lane et al., 1907; Stauffer, 1908; Sherzer and Grabau, 1909; Carman, 1927; Landes, 1945; Miller, 1955; Alling and Briggs, 1961; Ulteig, 1964; Sparling, 1970; Clifford, 1973; Kahle and Floyd, 1971; Janssens, 1977; Cramer, 2009; Kleffner, 2011; Kleffner et al., 2012a, 2012b). The type localities for these lithostratigraphic units reside in either southern (Greenfield) or northern (Tymochtee) Ohio and are geographically separated by over 160 km (100 mi). Bedrock assigned to this interval is often covered by thick glacial drift, inhibiting precise chronostratigraphic correlation of individual and often small outcrops that rarely contain contacts between lithostratigraphic units (Carman, 1927). Therefore, previous studies relied heavily upon correlation of the thick argillaceous-, gypsum-, and evaporite-bearing strata of the Salina Group and subsequent chronostratigraphic constraints from underlying (Greenfield Formation at the base of the Salina Group) and overlying (Bass Islands Group above the Salina Group) dolostones and inclusive (long-ranging) shelly faunas. Unfortunately, the type localities for the Greenfield and Tymochtee Formations reside in regions of Ohio with limited hydrocarbon exploration, leading to little subsurface data for correlation and uncertainty in the precise chronostratigraphic positions and correlative nature of these strata.

Extensive analyses of subsurface geophysical data throughout northern Ohio (Landes, 1945; Ulteig, 1964; Clifford, 1973; Janssens, 1977) refined the definition of the Tymochtee Formation and established subsurface correlations—confining the formation to a position correlative with the Salina Group A Unit. These conclusions are, however, in stark disagreement with the chronostratigraphic position of the Tymochtee Formation as it is utilized in southern Ohio, where strata assigned to the unit were correlated with subsurface B through G Units of the Salina Group by Horvath (1964). The discrepancies between the Tymochtee definitions of northern Ohio and usage in southern Ohio, proximal to the type locality of the Greenfield Formation, have never been sufficiently addressed, and adequate data have been only recently produced to permit the unit to be further assessed (Swift, 2011; Caruthers et al., 2018; Danielsen et al., 2019; Oborny et al., 2020a, 2020b; Rine et al., 2020).

Here, we contribute a new δ13Ccarb isotopic record from a drill core in Union County, central Ohio (i.e., core OW-11), where the Greenfield and Tymochtee Formations reach their maximum thicknesses, and we integrate subsurface geophysical data with existing biostratigraphic and δ13Ccarb chemostratigraphic data from western, southern, and eastern Ohio. Subsurface nomenclature and identification of the A through G Units are well established in northeastern Ohio, thereby permitting the correlation of these stratigraphic units elsewhere within the Appalachian, Michigan, and Illinois Basins, where the chronostratigraphic position and correlation of the evaporite strata of the Salina Group are similarly well established.

Geologic Setting

Silurian strata analyzed in this study were deposited along the western margin of the Appalachian Basin at ~25°S–30°S latitude on the paleocontinent of Laurentia (Cocks and Torsvik, 2011). Drill core and geophysical data utilized in subsurface correlations either parallel or lie perpendicular to the structural and depositional strike of the basin’s forebulge (i.e., combined Cincinnati, Findlay, and Algonquin arches; Figs. 1 and 2), a structural feature that developed during the Ordovician–Silurian Taconic orogeny due to basin flexure and crustal loading, and that was later reactivated during the Wenlockian–Pridolian(?) Salinic orogeny (Ettensohn and Brett, 1998; Dimitrov, 2007; Ettensohn, 2008). The Salinic orogeny was associated with significant changes in regional basin dynamics (Sparling, 1970; Brett et al., 1990), during which localized uplifts and/or bathymetric highs promoted increased rates of carbonate production (e.g., Cincinnati, Findley, and Algonquin arches). These structural and depositional highs invariably contributed to the restriction of water masses in the Appalachian and Michigan Basins and deposition of evaporitic strata. These strata are regionally assigned to the Salina Group (Slucher, 2004; Swezey, 2008), and in Michigan and northeastern Ohio, where evaporites are most prevalent, are referred to as the subsurface A through G Units, which are represented by over 700 m (2300 ft) of interbedded salts and dolostones (Catacosinos et al., 2001).

Nomenclatural History

The use of the terms “Greenfield” and “Tymochtee” in Ohio and their regional correlations historically have caused a great deal of confusion due to their variable usage, rank, assignment, and correlation. For example, both units at one time or another have been placed within the Bass Islands stratigraphic unit (itself with variable rank as a series, formation, and group) and Salina stratigraphic unit (as a formation and group). Further compounding this issue, there have never been clearly defined upper and lower stratigraphic limits for the Tymochtee Formation (modern use). Numerous studies have assessed strata assigned to these units in various locations; however, a full synopsis of these localized works is beyond the scope of this report, and we direct the reader to previous studies in which those works were discussed in more detail (Miller, 1955; Ulteig, 1964; Horvath, 1964; Sparling, 1970; Janssens, 1977). We instead provide here a brief regional summary of the establishment, usage, and subsurface correlations of these terms (Fig. 3). The modern terms “Greenfield” and “Tymochtee” are also intimately related to the historical terms “Waterlime” and “Monroe,” which we discuss in further detail below.

Greenfield Formation

The term “Greenfield” was first utilized by Orton (1871, p. 290–291) in southern Ohio near the town of Greenfield in Highland County, and the unit was named the Greenfield “stone.” The term was later extended to western New York by Grabau (1898), who referred to the unit as the Greenfield “limestone” of the Ohio Waterlime Formation and assigned the term to strata at the top of the Waterlime Formation based on shared shelly fauna immediately underlying the Silurian–Devonian unconformity. The author later determined this faunally based correlation to be in error, though the implications of his initial work remained problematic for the following decade (see below).

Historically, “Waterlime” was utilized to define strata that overlie the gypsum- and salt-bearing strata assigned to the Salina Group (modern rank), and the use of the term Waterlime in Ohio was first discussed in a general way by Newberry (1870, p. 15). Complexities quickly arose in northern Ohio due to an inability to readily distinguish between dolomitic units in that area (i.e., Ottawa County), and the term Waterlime became variably used (Winchell, 1873; Orton, 1888). Ultimately, the definition of the Waterlime stratigraphic unit (see discussion in Sparling, 1970) came to include the evaporitic strata and underlying dolostone in Ohio, in contrast to the lithological definition and stratigraphic position it was originally assigned. The complexities that arose from these previous works were later resolved by abandonment of the term Waterlime in Ohio and assignment of the problematic Waterlime strata to the lower Monroe Formation of the Bass Islands Series (Fig. 3, leftmost column). At that time, the Monroe Formation was thought to reside at a stratigraphic position above salt-bearing strata assigned to the Salina Group (modern use) in Michigan and New York (Prosser, 1903; Lane et al., 1907). With the establishment of the Monroe Formation, the authors also established the Greenfield Dolomite as the basal “formation” of the Bass Islands Series in northwestern Ohio, owing to similarities in stratigraphic position and shelly fauna to those observed in the type Greenfield Formation (modern use) of southern Ohio and the supposed Greenfield Formation–equivalent strata of New York (Whitfield, 1893; Grabau, 1898, 1900). Shortly after this nomenclatural revision, Sherzer and Grabau (1909) confirmed that the shelly fauna of northwestern Ohio was similar to that of southern Ohio. However, Grabau (1909) determined that, upon more critical comparison of the Greenfield faunas in Ohio to those in New York, the two regions had very little in common, and the initial correlations of the Greenfield strata between these two states were in error. Unfortunately, this observation appears to have had no effect in the literature on the stratigraphic position of the Greenfield Formation overlying the Salina Group in Ohio, as the extension of previous interpretations and usage of these terms in New York had become entrenched.

This regional interpretation, in which the Greenfield Dolomite (Greenfield Formation of today) was utilized as the base of the Bass Islands Series, persisted for the next four decades (Fig. 3; with revision to the Bass Islands stratigraphic unit ranking as either a series, formation, or group), at which point subsurface correlations from Michigan into northern Ohio determined that the subsurface A–G Units of the Salina Formation (Salina Group of today) of the Michigan Basin were in fact correlative to the Greenfield and Tymochtee “beds” of northern Ohio (Fig. 3; Landes, 1945), and these units (Greenfield and Tymochtee) were, in turn, reassigned to the underlying Salina Formation. Additionally, Landes (1945) determined that the A Unit of the Michigan Basin was correlative to strata previously identified as Greenfield Dolomite in northwestern Ohio. This surface-to-subsurface correlation was the accepted interpretation for the following 30 yr, at which point detailed subsurface analyses of northwestern Ohio by Janssens (1977) determined that the Greenfield Formation (as then recognized in rank), as defined in northern Ohio, was restricted entirely to the lower division of the A Unit (Fig. 3; modern A1 through A2 anhydrite).

The first and only detailed outcrop-to-outcrop analyses of the Greenfield Formation of southern Ohio northward to the Greenfield Formation of northern Ohio (study spanned over 240 km [150 mi]) were conducted by Carman (1927), who determined that physical correlation of the Greenfield Formation (then Greenfield Dolomite) northward was not possible due to the extensive glacial drift that covers the northwest half of Ohio, and that study concluded that the numerous outcrops available for analyses were often too small, preventing any meaningful correlation of the Greenfield Formation or distinction from the overlying lithologies of the Tymochtee Formation. Carman also noted that the contact between the Greenfield and overlying Tymochtee stratigraphic units was likely gradational everywhere and referenced suggestions from unnamed sources that considered the Tymochtee Formation to be a lateral, contemporaneous facies of the Greenfield Formation. This observation was also made by Miller (1955) in his analysis of the Greenfield-Tymochtee contact in southern Ohio, where he described a gradational trend northward along Paint Creek from Highland to Fayette Counties in which the typical upper porous, vuggy, brecciated facies of the Greenfield Formation transitioned into the basal tabular dolostones and shaly facies of the Tymochtee Formation. Given the limitations in these surface correlations, others have reverted to subsurface analyses of geophysical data, which have demonstrated that the type Greenfield Formation of southern Ohio does in fact correlate to the basal Salina Group of northern Ohio (Fig. 3; Horvath, 1964; Janssens, 1977), in particular, to the subsurface A Unit (Horvath, 1964).

Tymochtee Formation

The term “Tymochtee” was first utilized by Winchell (1873) for a creek exposure in Wyandot County, Ohio, to define 7.3 m (24 ft) of argillaceous to silty dolostone, referring to the unit as the Tymochtee “slate.” This newly defined unit was representative of the lower portion of an ~26-m-thick (~85-ft-thick) exposure assigned to the Waterlime Formation by Newberry (1870), a term that was eventually discontinued in favor of the Bass Islands Series (see discussion above). During the substantial nomenclatural revisions, in which the term Waterlime was discontinued, the Tymochtee stratigraphic unit was first assigned as a member by Prosser (1903) and subsequently raised in rank by Lane et al. (1907), who referred to the unit as the Tymochtee shales and limestones “formation” and defined it as encompassing all strata between the Greenfield and Put-in-Bay dolostone “formations” (Fig. 3). The work of Prosser (1903), with intention, redefined the Tymochtee strata at the type locality due to Winchell’s (1873) unclear definition of the “slate’s” lower and upper limits. This revised definition expanded the use of “Tymochtee” to encompass the entirety of strata initially assigned to the “Waterlime” at that locality, then consisting of over 30 m (~100 ft) of section, predominantly composed of shales and thin-bedded carbonates. Several more recent publications overlooked the work of Prosser (1903) in this reclassification of the Tymochtee “formation” and as such concluded that this revision was made by Lane et al. (1907), who provided an insufficient synopsis of the unit’s history. These subsequent works therefore determined that the inclusion of additional strata at the type locality was done in error. This uncertainty by later authors as to whether there was intent in changing the definition of the lower and upper limits of the Tymochtee “formation” is arguably the source of much of the succeeding confusion in the use of this term and has been discussed in literature on at least two occasions (Sparling, 1970; Kahle and Floyd, 1971).

As with many units, the Tymochtee Formation is plagued with a long history of stratigraphic revision (see discussions in Carman, 1927; Ulteig, 1964; Sparling, 1970; Kahle and Floyd, 1971; Janssens, 1977), largely due to the unit’s limited exposure and inability of older studies to trace strata for any meaningful distance—both of which resulted in repeated speculation as to the precise stratigraphic position of individual outcrops. A thorough discussion of this history is beyond the scope of this report. We do, however, feel the need to highlight that the formation-level status of the Tymochtee stratigraphic unit is often erroneously attributed to Stauffer (1908), though Lane et al. (1907) clearly referred to the unit as a “formation” within the lower Monroe Formation of the Bass Islands Series, a designation utilized by Stauffer (1908) the following year, with guidance from Dr. C.S. Prosser (co-author of Lane et al., 1907), in describing the Tymochtee “formation” in his work. This problematic issue of a “formation within a formation” was later resolved by Carman (1927), who redefined the Greenfield, Tymochtee, Put-In-Bay, and Raisin River stratigraphic units as members within the Bass Islands Formation and described the lower Monroe Formation as a “division” (modern Bass Islands Group), as it could not be defined as a “series” or “formation” (Fig. 3).

The first detailed study in which the Tymochtee Formation was correlated into southern and northern reaches of Ohio, extending from Lucas County in the north to Fayette County in the south, was conducted by Carman (1927; see also Greenfield discussion above). The product of that work served as the foundation for all subsequent outcrop studies of the Tymochtee and Greenfield stratigraphic units for the entirety of Ohio for the next 50 yr. In northern Ohio, outcrops assigned to the Tymochtee Formation (modern rank) in Lucas and Wood Counties by Carman (1927) were further subdivided into four members by Kahle and Floyd (1971) (see Fig. 3 herein), a subdivision that appears limited to that work but is recognized in the U.S. Geological Survey’s GEOLEX (USGS, 2019 [online]). In southern Ohio, Miller (1955) extended the Tymochtee correlations of Carman (1927) southward into Adams County and determined that the unit was overlain by Upper Devonian strata of the “Ohio Shale.” Therefore, Miller (1955) did not define an upper limit for the Tymochtee Formation in southern Ohio but merely stated that the unit was overlain by Devonian strata. Horvath (1964) later studied the Tymochtee strata of southern Ohio in more detail and correlated the unit into the subsurface of eastern Ohio, concluding that the Tymochtee Formation of southern Ohio correlated to the subsurface B through G Units in the eastern part of the state (Fig. 3).

Subsurface correlations for the Tymochtee Formation in northern Ohio predate those of Horvath (1964) and, in fact, originate in southern Michigan. There, it was postulated by Landes (1945), who utilized the lithostratigraphic division of Lane et al. (1907), that the Tymochtee “beds” of Ohio were correlative to the subsurface C through G Units of the Michigan Basin (Landes determined that the B Unit was absent in Ohio). This interpretation persisted in the literature with minor modification for over 30 yr within studies dissociated from the type locality of the Tymochtee Formation (e.g., Wilmarth, 1938; Stout, 1941; Ulteig, 1964; Kahle and Floyd, 1971). This history is, however, problematic, because Landes (1945) stated that correlation of strata overlying the C Unit was nondeterminant due to erosion of these overlying units in northern Ohio, and he clearly shows in his subsurface correlations that only the A Unit crops out in the area of the type Tymochtee Formation in Wyandot County, Ohio. Alling and Briggs (1961, p. 524) later confined the correlation of the Tymochtee Formation to the C Unit through the analyses of well cuttings and referred to the unit as the Tymochtee “shale” formation. This correlation was also considered by Ulteig (1964) in his subsurface analyses of northeastern Ohio as a secondary hypothesis.

Later subsurface work by Sparling (1970) in Ottawa County, Ohio, concluded that Lane et al. (1907) misrepresented the true meaning of “Tymochtee” as intended by Winchell (1873; see Greenfield discussion above). Due to the ambiguity of the Tymochtee “slate’s” upper contact, and until such time that it received additional analyses, he chose to retain the nomenclatural division of Lane et al. (1907) and determined that the basal beds of the Tymochtee Formation extended into the underlying A Unit (i.e., the subsurface A2). Subsequent and extensive subsurface analyses of Silurian strata in northwestern Ohio by Janssens (1977), in which over 500 localities were analyzed, reverted to the original definition of Winchell (1873; i.e., Tymochtee “slate,” representing 7.3 m [24 ft] of strata) and referred to the unit as the Tymochtee Dolomite Formation. The findings of that work determined that the Tymochtee stratigraphic unit was correlative to the upper part of the A Unit of the Michigan Basin and restricted specifically to a position within the A2 carbonate, and that all strata between the upper Tymochtee Formation and basal Bass Islands Group (modern use of these terms) were undefined, and therefore described the subsurface B through G Units as “undifferentiated Salina” (Fig. 3). Janssens (1977) continued by stating that the true lithological character of the Tymochtee stratigraphic unit could not be correlated beyond ~20 km (12 mi) from the unit’s type locality due to facies changes. There are, however, geophysical markers (Janssens, 1977, p. 24) that permit the subsurface correlation of Tymochtee strata throughout the region, which, when compared to older yet similar studies from northeastern Ohio (Ulteig, 1964; Clifford, 1973), do verify that the Tymochtee Formation of Janssens (1977) stratigraphically underlies the salts of the B Unit as defined by Ulteig (1964) and Clifford (1973). This detailed analysis by Janssens (1977) later provided the foundation for the Ohio Division of Geological Survey’s current correlation of the type Tymochtee locality to the upper portion of the subsurface A2 Unit for northern Ohio and restriction of the Greenfield Formation to the remainder of underlying strata assigned to the A Unit (Fig. 3; Ohio Division of Geological Survey, 2004).

In summary of the above discussion, it is critical for the reader to take note of the following: The type Tymochtee “slate” was originally defined to encompass 7.3 m (24 ft) of strata by Winchell (1873) and later redefined by Prosser (1903) to encompass the ~30 m (100+ ft) of strata exposed at that locality. The definitions and objectives of the revisions of Prosser (1903) were not acknowledged in later work and therefore were overlooked in the literature; however, the expanded definition of the Tymochtee stratigraphic units persisted in the literature for ~70 yr, at which point confusion ensued, and the term’s use reverted back (Janssens, 1977) to the original definition as established by Winchell (1873) in which the Tymochtee “slate” consisted of 7.3 m (24 ft) of strata. Therefore, literature published prior to the 1970s, or studies that cite literature prior to the 1970s, will likely utilize the definition of Prosser (1903) in defining the upper and lower limits of the Tymochtee stratigraphic unit, whereas the opposite is true for works published after the 1970s, in which the definition of Winchell (1873) will likely be utilized. Currently, the Ohio Department of Natural Resources Division of Geological Survey (Slucher, 2004) utilizes the interpretation of Janssens (1977) and therefore the definition of Winchell (1873; ~24 ft of strata). We maintain this definition for the study herein and note that the lower and upper contacts of the Tymochtee stratigraphic unit, at its type locality, remain poorly defined.

Biostratigraphy and Chemostratigraphy

Several δ13Ccarb chemostratigraphic analyses of Silurian strata have been conducted within southern and western Ohio (Cramer, 2009; Kleffner, 2011; Swift, 2011; Kleffner et al., 2012a, 2012b; McLaughlin et al., 2012; Sullivan et al., 2016; Oborny et al., 2020a) and within the Michigan Basin (O’Shea et al., 1988; McLaughlin et al., 2013, 2019; Rine et al., 2020). Unfortunately, most of these studies were conducted in locations geographically beyond the reach of the current investigation and/or sampled intervals assigned to strata chronostratigraphically underlying the Salina Group. The works of Swift (2011) and Danielsen et al. (2019) produced δ13Ccarb and/or conodont biostratigraphic data in Ohio proximal to the present study area in Logan, Auglaize, and Mercer Counties, due west of the OW-11 core in Union County (Fig. 2). These authors identified an excursion termed the “Mulde” and recovered conodont elements (Swift, 2011) of Pseudooneotodus linguicornis and Ozarkodina confluens densidentata(?). These species are observed in the Baltic and eastern Appalachian basins and have respective last and first appearances within the upper Homerian Stage that are closely tied to the onset of the Mulde excursion (Jeppsson, 1969; Helfrich, 1975; Jeppsson et al., 1995; Calner and Jeppsson, 2003; Bancroft and Cramer, 2020). The studies of Cramer (2009) and Oborny et al. (2020a) sampled the Greenfield Formation and Greenfield through the Silurian-Devonian boundary, respectively, in southern Ohio near the type Greenfield locality and, as such, are both directly applicable to the current investigation. These studies (i.e., Cramer, 2009; Oborny et al., 2020a) determined that the upper Homerian carbon isotope excursion (Mulde) was contained within the Greenfield Formation and that overlying isotopic data, though highly erratic, could be correlated to the eastern margin of the Appalachian Basin (Oborny et al., 2020a, 2020b). Through these correlations, it was determined by Oborny et al. (2020a) that Upper Silurian strata (underlying the Silurian-Devonian boundary) in Scioto County, southern Ohio, extended into the Pridoli Series.

Recent studies within in the Michigan Basin (Caruthers et al., 2018; Rine et al., 2020) utilized δ13Ccarb or combined δ13Ccarb13Corg chemostratigraphic analyses within the subsurface A Unit. Both of these works identified the Mulde excursion within the A Unit. Of these, Caruthers et al. (2018) extended their analyses into the basal salt lithology of the overlying B Unit and assigned an isotopic perturbation in δ13Corg, observed within the basal salt of the B Unit, to an excursion termed the “Linde” excursion. At present, there exist no δ13Ccarb chemostratigraphic data from Upper Silurian (Ludlow–Pridoli) strata in northern or eastern Ohio. There are, however, limited conodont biostratigraphic data from the Raisin River Formation of the Bass Islands Group in northern Ohio (Kleffner and Larsen, 2003). That work identified elements of Zieglerodina remscheidensis, Belodella, Panderodus sp., and Wurmiella excavata. An assemblage similar to this was observed within the Tonoloway Limestone Formation in Virginia on the eastern side of the Appalachian Basin (Bancroft and Cramer, 2020; Oborny et al., 2020b) and would suggest a position somewhere within the Pridoli Epoch (Viira, 1999).

Subsurface Geophysical Signatures of the Salina A–G Units

The evaporites contained within the Salina Group A through G Units in Ohio are restricted to the northeastern quadrant of the state (Clifford, 1973). There, the basal contacts of salts and anhydrites were utilized by Clifford (1973; designations used in subsurface correlations by this study) in the identification of unit boundaries (with the exception of the G Unit). Thick successions of salt within these units are considered to be representative of deep-water precipitates, forming from dense, hypersaline water near the base of the water column (see summary in Clifford, 1973). Within the Michigan Basin, Rine et al. (2020) considered these episodes of hypersalinity to be driven by an overall diminished volume in water mass, and, in the case of the A Unit, they were considered to be regressive in nature (i.e., basinwide fall in relative sea level; Rine et al., 2020). Therefore, based on the sequence stratigraphic interpretations of Rine et al. (2020), the basal contacts of salt units that were established by Clifford (1973) in northeastern Ohio are within a regressive phase (falling stage). This contrasts with common practice in the mixed carbonate-clastic setting of southern Ohio, where formational boundaries are typically established at the basal contacts of transgressive systems tracts (i.e., base of carbonate formations; e.g., McLaughlin et al., 2008; Oborny et al., 2020a). With this framework, the subsurface nomenclature of Clifford (1973) will not align precisely with outcropping formational boundaries, and the geophysical signal observed in the central and southwestern parts of the state, where evaporites are largely absent, will vary from those reported in the northeast.

Despite regional changes in facies, there remain several correlative intervals within geophysical logs associated with repetitive stacking of laterally continuous dolostone and shale units. Reliable correlation of the basal A Unit contact into central and southern Ohio in previous works was, however, hindered by lateral changes in facies owing to the absence of salts and anhydrites (Horvath, 1964; Ulteig, 1964; Clifford, 1973). For the sake of brevity, we direct the reader to these previous works for discussion of the upper B Unit through basal Devonian boundary correlations and turn our attention to the lower A and B Unit boundaries because they are critical for the work presented below.

In distal locations in northeastern Ohio, the base of the A Unit is established at the bottom of an anhydrite, providing a distinguishable geophysical horizon for correlation (characterized by low gamma-ray and low porosity values; Clifford, 1973). A thin, radioactive shale preceding an interval of increased porosity is often observed immediately overlying this anhydrite. Southward toward Muskingum County, this anhydrite thins and is frequently missing in locations to the southwest, resulting in the absence of its accompanying geophysical marker. The radioactive shale and interval of increased porosity do, however, remain prevalent and are correlative southward (see discussion below). Similarly, the base of the B Unit is also correlative into southern Ohio. Subsurface studies in northeastern Ohio utilized the base of a salt in the establishment of this boundary (Clifford, 1973). Southward toward Muskingum County, this salt is replaced by a correlative dolostone and/or anhydritic dolostone (with low gamma-ray and low porosity values; see below).

For this study, we conducted lithostratigraphic, geophysical, and carbonate carbon isotopic (δ13Ccarb) chemostratigraphic analyses of the OW-11 core for strata tentatively assigned to the Greenfield and Tymochtee Formations dolostones by Carman (1927). The OW-11 core was drilled by Reynolds, Inc., near the Marysville Reservoir in Union County, Ohio (40.267655°N, 83.398439°S), and later generously donated to the Paleozoic Earth History Group at the University of Iowa for the purpose of our analyses. The OW-11 core is registered under the International Generic Sample Number (IGSN) IESCO0001 within the System of Earth Sample Registration (SESAR) which can be accessed at https://www.igsn.org. The Ohio Department of Natural Resources (DNR) Division of Geological Survey (OGS) kindly provided spectral gamma-ray scans for the core. These spectral data are available from the OGS upon request of “DNR Well Log 2033346/Core OW-11.”

Whole-rock samples for chemostratigraphic analyses were collected at a 30 cm (~12 in.) resolution using the standard sampling procedure of Saltzman (2002) to minimize contamination. Samples were then processed at the W.M. Keck Paleoenvironmental and Environmental Stable Isotope Laboratory (KPESIL) at the University of Kansas using a Thermo Finnigan MAT 253 dual-inlet stable isotope ratio mass spectrometer system coupled with an UltraTrace gas chromatograph and KIEL III carbonate device in continuous-flow mode. Corrections were calculated via NBS-18, NBS-19, and NIST LSVEC standards and reported in parts per mil (‰) relative to Vienna Peedee belemnite (VPDB). Standard deviations for δ13Ccarb and δ18Ocarb were ±0.05‰ and ±0.20‰, respectively (1σ).

Geophysical data utilized in this study primarily included gamma-ray and porosity logs. Spectral gamma-ray scans of the OW-11 core and OGS core 3409 (see Franklin Furnace core in McLaughlin et al., 2012; Oborny et al., 2020a) provided a means for direct comparison of geophysical and lithological data and assisted in accurate alignment (compensating for cable stretch through comparison to downhole geophysical data) of the δ13Ccarb chemostratigraphic analyses of Oborny et al. (2020a) to downhole geophysical data acquired at the Franklin Furnace borehole. We utilized the lithostratigraphic unit designations applied by Oborny et al. (2020a) for the Franklin Furnace core in Scioto County, Ohio, and extended this nomenclature when appropriate throughout our transects.

Lithostratigraphic and δ13Ccarb chemostratigraphic analyses of the OW-11 core are presented in Figure 4. Isotopic data are variable over the length of the core, ranging between −3.3‰ and +3.7‰. An increase from +0.5‰ to +3.7‰ is observed at the base of the core (82.9 m to 78.3 m [272.0–257.0 ft]). Overlying isotopic values decrease abruptly at 78.3 m (257.0 ft), returning to between 0.0‰ and 0.5‰; these values persist until 56.4 m (185.0 ft) in the core. However, two minor negative inflections of at least −1.0‰ exist within this interval (Fig. 3). At 56.4 m (~185.0 ft), isotopic values decrease abruptly to −3.3‰. Values then increase until 41.7 m (137.0 ft), at which point isotopic data again reach a plateau at peak values of ~−0.3‰. This plateau continues through the remainder of the interval, until a minor inflection of −1.0‰ is observed, and isotope data return to levels similar to those of underlying baseline values for the remainder of the core (Fig. 4; purple coloration).

Several lithologies indicative of either highstand (black shales) or lowstand (rip-up clasts, tempestites) systems tracts (McLaughlin et al., 2008) were observed in the OW-11 core and are reflected within the stratigraphic column (Fig. 4). Our coupled subsurface geophysical correlations and δ13Ccarb chemostratigraphic data indicate that a majority of strata in the OW-11 core, which comprises the upper 82.9 m (272.0 ft) of bedrock geology of Union County, Ohio, are laterally equivalent to the Greenfield Formation of southern Ohio (Plates S1–S61). Geophysical correlations indicate that the Peebles-Greenfield Formation contact, as identified in the Franklin Furnace core by Oborny et al. (2020a), correlates to the ascending limb of increasing isotopic values at the base of the OW-11 core (80.54 m [264.25 ft]). This horizon is readily identified within geophysical logs in our north-south transect (Plates S2, S4, and S6; transect B–B′) underlying an interval of increased porosity, and in Union County, it correlates to the bottom of a tan stromatolitic and skeletal dolowacke to dolo pack stone, overlying dark dolostones with black shale partings.

The upper contact of the Greenfield Formation, as defined in Scioto County, southern Ohio, by Oborny et al. (2020a) is similarly correlative within geophysical logs. The geophysical signals attributed to this horizon in our north-south transect vary moderately. However, this boundary is discernible from one location to the next and can be further stratigraphically constrained by overlying and underlying intervals of increased clay content (i.e., increase in gamma). With these correlations, we establish the upper contact of the Greenfield Forma tion at 15.3 m (50.33 ft) in the OW-11 core at the base of a shell lag (i.e., sequence boundary) underlying thrombolitic, green, argillaceous dolostone, which grades into an overlying clean dolo-calcilutite (i.e., transgression) before returning to shallow-marine facies and collapse breccia at ~11.7 m (~38.5 ft). The establishment of the upper Greenfield Formation contact confines the overlying 4 m (13.33 ft) of bedrock (top of OW-11 core is at 11.3 m [37.00 ft]) to an interval laterally equivalent to the Tymochtee Formation as it is defined in southern Ohio (Horvath, 1964; Oborny et al., 2020a).

In correlating these formational boundaries eastward to the subsurface in Muskingum County, Ohio, we note that the Peebles-Greenfield Formation contact aligns well with the base of the subsurface A Unit established by previous works (Ulteig, 1964; Clifford, 1973). Given the geophysical characteristics of the radioactive shale and interval of increased porosity (discussed above), which are correlative in our data, we extend the base of the A Unit to Union and Scioto Counties and demonstrate that the Peebles-Greenfield Formation contact is at the same chronostratigraphic position as the basal contact of the A Unit. In establishing these correlations, we note that we do not have lithological data for our subsurface transects and are limited in interpretation to the geophysical signal. There are locations in our transects where the basal boundary of the A Unit may be accompanied by a thin (<1.5 m [5.0 ft] thick) anhydrite (e.g., see A–A′ and B–B′ intercept, base of A Unit); however, such a facies is beyond the known geographic range as mapped by previous studies (Clifford, 1973), and we have no means of eliminating the possibility that the geophysical signals are representative of a sandstone or other dense lithology underlying the A Unit. We therefore maintain our correlations of the basal A Unit throughout our transects at the base of the radioactive shale and interval of increased porosity, which align with the sequence boundary at the base of the Greenfield Formation in southern Ohio.

The overlying A-B Unit contact is readily correlated within our east-west transect. Previous studies identified a dolostone and/or anhydritic dolostone (Clifford, 1973) at the base of the B Unit in central and eastern Ohio. This unit is represented by a distinguishable geophysical marker (low gamma, low porosity) that correlates to a position of ~21.1 m (69.4 ft) in the OW-11 core, ~5.8 m (~19.0 ft) below the upper Greenfield Formation contact, at a horizon separating underlying dolo-calcilutite from overlying weakly laminated and mottled argillaceous dolostones. Evidence of salt precipitate is observed at, and immediately overlying, the 21.1 m (69.4 ft) contact, and it would appear to be consistent with the presence of rip-up clasts and shallow-water facies indicators (e.g., bird’s-eye, scour surfaces, etc.) in overlying strata extending to the top of the Greenfield Formation, indicating the 21.1 m (69.4 ft) contact marks the onset of a regression. The contact’s geophysical signal is less well defined in our north-south transect; however, this boundary is discernible from one location to the next and can be further chronostratigraphically constrained by overlying and underlying intervals of increased clay content (Plates S2, S4, and S6, see footnote 1). These correlations demonstrate that the basal boundary of the B Unit aligns precisely with the top of the dual-peaked isotopic excursion identified by Oborny et al. (2020a, orange excursion) in the Franklin Furnace core.

Driller logs for the OW-11 core indicate that an 11.3 m (37.0 ft) interval of glacial till and/or drift overlies the core. These strata were not retained for study, but driller observations are in agreement with regional glacial drift maps (Ohio Division of Geological Survey, 2004). Therefore, the remainder of the OW-11 core could not be directly correlated to subsurface units, as designated by nomenclature of drillers, although it is clear that all bedrock overlying 21.15 m (69.4 ft) correlates to a position within the subsurface B Unit.

Given the laterally continuous nature of lithologies within the upper B through G Units, and overlying Devonian strata, we correlate and establish boundaries for the base of the C, D/E, F, and G Units and Devonian strata in the Franklin Furnace core at ~520 m (~1706.0 ft), ~511 m (~1676.0 ft), ~501 m (~1644.0 ft), ~470 m (~1543.0 ft), and ~455.4 m (~1494.0 ft), respectively, and we note that the establishment of these boundaries aligns well with either ascending or descending limbs of isotopic perturbations observed by Oborny et al. (2020a). Additionally, we tentatively identify a thin, 2.74-m-thick (9 ft) interval of the Bass Islands Group and establish the base of the unit at 458 m (~1503 ft) in the Franklin Furnace core, immediately underlying the Silurian-Devonian boundary. In assessment of the Silurian-Devonian boundary, we assign the immediately overlying Devonian strata to the Columbus Limestone Formation (i.e., equivalent to the Onondago Limestone Group; see combined Slucher, 2004; Baranoski and Riley, 2013) in our transects and within the Franklin Furnace core. Given the ~35 m.y. of missing strata at this boundary and variable geophysical signal associated with the unconformity, it was at times difficult to determine the horizon’s exact position.

Revision to Nomenclature

Strata tentatively assigned to the Tymochtee Formation in Union County, central Ohio, by Carman (1927) are correlative to the Greenfield Formation of southern Ohio (Plates S1–S6, see footnote 1). Geophysical correlations also demonstrate that the Greenfield Formation is correlative with the subsurface A Unit. The basal boundary of the A Unit correlates precisely to the Peebles-Greenfield Formation contact and is coincident with the ascending limb of the Mulde excursion identified in the OW-11 and Franklin Furnace cores (Oborny et al., 2020a). The upper contact of the A Unit correlates to a position within the Greenfield Formation. As such, the upper contact of the Greenfield Formation is at a chronostratigraphic position within the lowermost part of the subsurface B Unit. These observations demonstrate that the base of the overlying Tymochtee Formation as defined in southern Ohio (Horvath, 1964; Oborny et al., 2020a) is at a higher chronostratigraphic position than the upper boundary of the Tymochtee Formation as it is defined at the type locality in Wyandot County, northern Ohio (Winchell, 1873; Janssens, 1977). There, the Tymochtee Formation has been correlated to the subsurface A2 carbonate of northeastern Ohio and southern Michigan (Landes, 1945; Ulteig, 1964; Clifford, 1973; Janssens 1977). Based upon our carbon isotope data and subsurface geophysical correlations presented herein, it is clear that the Tymochtee Formation of northern Ohio, as defined by Winchell (1873; 7.3 m [24 ft] of strata), is equivalent to a position within the Greenfield Formation of southern Ohio. Given these observations, the best course of action is to utilize the senior nomenclature of Orton (1871), and in doing so, we propose abandonment of the lithostratigraphic term “Tymochtee.” This term is, however, historically entrenched, and it may be more practical to retain both Greenfield and Tymochtee with a stated caveat.

We herein informally refer to strata overlying the Greenfield Formation and underlying the Bass Islands Group as the “Salina B–G” interval. Given that there is currently no definable type locality for this interval, we refrain from assigning a hierarchal designation, and we do not refer to this interval as “Salina undifferentiated,” as it is clearly differentiable within the subsurface and also clearly does not include the A Unit.

Historically, the upper Greenfield Formation contact has not been well defined (Carman, 1927; Miller, 1955; Horvath, 1964). We therefore propose that the first sequence bounding surface overlying the salt at the base of the B Unit be utilized as the Greenfield–“Salina B–G” contact (see Fig. 4). In southern Ohio, this contact is placed at the base of a stromatolitic and stromatoporoid-bearing interval underlying green shale (Oborny et al., 2020a). This green facies is similarly observed immediately above the upper Greenfield Formation contact within Union County, and it is reported near the base of the B Unit in West Virginia (Horvath, 1964) and parts of northern Ohio (Ulteig, 1964; Janssens, 1977). We do, however, caution the reader that these former works also described green intervals within the subsurface A, C, and G Units. Outcrops of this contact can be found at the Plum Run and Ralph Rodgers Quarries in Adams and Pike Counties, Ohio (Miller, 1955; Horvath, 1964).

Subsurface Correlations and Depositional Motif

In large part, these transects demonstrate that much of the interval studied is composed of strata that thicken in a general eastward to northeastward direction (Plates S1–S6, see footnote 1) and overlie thickened sites of apparent mounding and/or reef development in the Bisher and Lilley Formations (Plates S5 and S6). This pattern records filling of accommodation by the Greenfield Formation and laterally equivalent strata (Plates S3 and S4).

The Greenfield Formation is observed onlapping underlying strata in our north-south transect accompanied by intermittent thinning in the underlying Peebles Formation. An example of this can be observed between Jackson and Scioto Counties in southern Ohio (Plate S4). Given that the Greenfield Formation contains the Mulde excursion and that strata showing this excursion overlie an unconformity recording the mid-late Homerian global sea-level lowstand (Calner, 1999, 2002; Calner and Jeppsson, 2003; Danielsen et al., 2019; Oborny et al., 2020b), we suggest that the intermittent thinning of the Peebles Formation and onlapping of the Greenfield Formation may be evidence of this eustatic event. In this case, the Peebles Formation was removed by wave energy (coastal degradation) during sea-level rise, which would explain the disconformity and limonite-stained shales underlying the Greenfield Formation in parts of southern Ohio (Miller, 1955). Furthermore, several highstand intervals characterized by black shales are observed within the Greenfield Formation in the OW-11 core, indicating significant and rapid sea-level fluctuation during this short interval of time.

Well-defined and uniform thickening of strata is observed overlying the B Unit in our east-west transect (most prominent in the D–F Units). Given the uniformity in thickening of the B Unit and overlying units, and variable thickness of the underlying A Unit, it would appear that sites of sediment accommodation were largely absent by the end of Greenfield Formation (A Unit) deposition and may indicate a short interval of tectonic stability within this region in which accommodation was filled prior to the deposition of overlying strata (Plates S3 and S4, see footnote 1).

Uniform thickening of strata overlying the Greenfield Formation (overlying the B Unit) is less prominent in our north-south analyses, indicating that the basin depocenter was to the east-northeast during deposition of “Salina B–G” strata. This interpretation is in agreement with subsurface studies, in which isopach maps show the thickest intervals of salt in the northeastern quarter of Ohio during deposition of strata overlying the B Unit, approximately paralleling the state line and our north-south transect (Clifford, 1973).

Detailed discussion of Devonian strata is beyond the scope of this report. Identification of the Silurian-Devonian boundary, however, indicates that strata assigned to the Columbus Limestone Formation immediately overlie Silurian strata in our north-south transect and that rapid subsidence was occurring to the east during deposition of Devonian strata. The analysis of this boundary permitted thorough assessment of the Bass Islands Group, and we determined that a small part of this unit may be present within the Franklin Furnace core; we, however, state this without full confidence. Should these strata in fact be the Bass Islands Group, it would be the first identification of this unit in southern Ohio.

Overall, the observed depositional motif verifies the conclusions drawn by Oborny et al. (2020a), who stated that the primary site of decreased bathymetry and thinnest deposition was located in northeastern Kentucky along the Appalachian Basin’s southwestern margin during Homerian time. There, the Peebles and Greenfield Formations are observed to thicken to the north, indicating that the primary site of sediment accommodation was no longer to the east of our study area, as was the case in underlying strata. The causal mechanism behind this regional change in bathymetry remains unclear, but it could be related to previously unrecorded tectonic activity or the development of a carbonate platform to the south.

A thin tongue of detrital siltstone is observed in the upper Estill Shale in our north-south transect (Plate S4, see footnote 1). The only known clastic source at that time was located on the southeastern margin of the basin. Therefore, this tongue may represent a far-reaching and early distal component of the “Eagle Rock Sandstone” observed in Virginia, a unit that first appears at the same chronostratigraphic level and persists through the basal Pridoli section. Onset of this sand progradation would be equivalent to the upper Estill Shale (Haynes et al., 2015; Oborny et al., 2020a, 2020b). Should this be the case, it may suggest that clastic influx along the basin’s southern margin contributed to the basin’s restriction. It is important to note that the δ13Ccarb isotopic record throughout much of the basin shifted significantly toward more negative values during the late Telychian (i.e., upper Estill Shale) and early Sheinwoodian time (Oborny et al., 2020a, 2020b), indicating a substantial change in basin chemistry, potentially due to the basin’s southern margin being closed and/or restricted, therein limiting water-mass mixing between the Appalachian Basin and Rheic Ocean. Such an interpretation would also explain the sudden transition from open-marine to restricted facies within the basin, in which evaporites became prevalent during deposition of “Salina B–G” strata (post-Greenfield deposition). The exact causal mechanism for this restriction remains unclear; however, it was likely a combination of Salinic-related orogenesis, resulting in crustal flexure and removal of accommodation space, and eustasy.

Uniform thickening of strata overlying the Greenfield Formation toward a depocenter in the northeast is suggestive of crustal flexure in that area. The timing at which this depocenter developed aligns well with tectonic evidence in New Brunswick at the northern end of the Appalachian Basin. There, northwest-southeast axial folds are first reported within the South Charlo Forma tion (Dimitrov, 2007), a unit that overlies an unconformity termed “Salinic B” (Wilson and Kamo, 2012). Rhyolites immediately overlying the “Salinic B” unconformity provide an age of 422.3 ± 0.3 Ma (post–Lau excursion age), thereby confining the northwest-southeast axial folds to latest Ludlow or early Pridoli time. This age, in conjunction with fossil evidence underlying the “Salinic B” unconformity, indicates a >5 m.y. hiatus (i.e., Ludlow Series is absent). These observations align with those of Oborny et al. (2020a, 2020b), who determined through biochemostratigraphic analyses that the Lau excursion and much of the Ludlow Series are similarly absent on the eastern and southwestern margins of the Appalachian Basin. Our correlations between southwestern and northeastern Ohio permit chronostratigraphic constraint for these thickening strata and indicate that the base of the Ozarkodina crispa zone (immediately overlying the Lau excursion), as established by Bancroft and Cramer (2020) and Oborny et al. (2020a, 2020b) elsewhere within the basin, correlates to a position within the subsurface D–E Units. Coincidently, northeastward thickening in our subsurface transects becomes prominent at this interval. Therefore, our findings along the western margin of the basin are in alignment with findings along the basin’s northern, eastern, and southwestern margins (Dimitrov, 2007; Wilson and Kamo, 2012; Bancroft and Cramer, 2020; Oborny et al., 2020a, 2020b). Combined, these works provide significant evidence for a substantial regional hiatus during the Ludlow Epoch.

Assessment of Geochemical Data

Geophysical correlations between the OW-11 and Franklin Furnace cores indicate that strata containing the Mulde excursion, as identified in the Franklin Furnace core, thicken significantly northward into Union County, where the interval is encompassed by roughly 61 m (200 ft) of strata. Therefore, the δ13Ccarb isotopic record observed in the lower three quarters of the OW-11 core (Fig. 4; the combined blue, yellow, and orange coloration in figure) in central Ohio is representative of an expanded section of the Mulde excursion (Fig. 5). The position of the basal B Unit within the ascending limb of the overlying excursion in the OW-11 core (Fig. 4, purple coloration) is at the same chronostratigraphic position as the B boundary as identified in Michigan. There, the boundary is established within the ascending limb of an excursion that encompasses the basal salt of the B Unit, identified by Caruthers et al. (2018) as the Linde excursion. We interpret these data to demonstrate that these excursions are correlative between the basins. However, the isotopic perturbations in this interval correlate to the Ozarkodina bohemica interval zone on the eastern margin of the Appalachian Basin (Bancroft and Cramer, 2020; Oborny et al., 2020a, 2020b) and suggest that this interval may be within the upper Homerian or lowermost Gorstian Stages. Globally, the ascending limb of the Linde excursion has been utilized in a general way to mark the Gorstian-Ludfordian Stage boundary (Cramer et al., 2011a, 2011b; McAdams et al., 2019). With these observations, we believe it to be unlikely that this excursion is the Linde excursion, and we suggest that identification of the isotopic signal await further biostratigraphic information.

The isotopic data from the OW-11 core compare well to the isotopic record observed and compiled by Swift (2011) in west-central Ohio (Fig. 5) and indicate, based upon the combined biostratigraphic and chemostratigraphic records of the Con Ag and Duff Quarries (Auglaize and Logan Counties, respectively), that these sections span the lower peak of the Mulde excursion as identified in the Franklin Furnace core by Oborny et al. (2020a). More specifically, the Cedarville-Greenfield “transition zone” observed at the Con Ag Quarry (Swift, 2011) is chronostratigraphically correlative to the basal A Unit, and the lithological transition observed near the top of the Duff Quarry is chronostratigraphically correlative to the argillaceous interval between 50.3 m and 56.4 m (165 and 185 ft) in the OW-11 core (Fig. 5). With the establishment of these boundaries, the entirety of strata exposed at both sections, overlying the “transition zone,” is representative of the Greenfield Formation. Additionally, the biostratigraphic data acquired by Swift (2011) at these sections are in alignment with the recent findings on the eastern margin of the basin (Bancroft and Cramer, 2020; Oborny et al., 2020b) and indicate that the interval overlying the “transition zone” resides within the Oz. bohemica interval zone.

The chemostratigraphic record observed by Danielsen et al. (2019) near the town of Celina in Mercer County, Ohio, and within Indiana locations of that study compares well to our data (Fig. 5) and indicates that the Waldron Shale of the Pleasant Mills Formation of eastern Indiana and western Ohio is at a chronostratigraphic position equivalent to the lower A1 salt unit (see Fig. 3) and therefore is correlative to the lower Greenfield Formation in southern Ohio, where the shale facies is largely absent. Regional analyses by Droste and Shaver (1982) similarly concluded that the Waldron Shale correlated to a position somewhere within the lower A Unit and are therefore in agreement with our findings. These authors also determined that the Pleasant Mills–Wabash formational contact in central and northern Indiana and northwestern Ohio was correlative to a position at or near the base of the B salt unit of the Salina Group. Though we cannot directly verify the precise chronostratigraphic position of the upper Pleasant Mills contact, there is a strong lithological similarity between the Mississinewa Shale (i.e., member at the base of the Wabash Formation) and the predominantly argillaceous shales and evaporites of the “Salina B–G” interval of this study. Therefore, we retain the correlations of Droste and Shaver (1982) and tentatively assign the Pleasant Mills–Wabash contact (i.e., a sequence bounding surface) at a chronostratigraphic position equivalent to the upper Greenfield Formation contact (see earlier discussions of this contact in the OW-11 core) and confine the overlying Mississinewa Shale Member of western Ohio and eastern Indiana to a chronostratigraphic position equivalent to the predominantly shaly and evaporitic strata of the “Salina B–G” interval of this study (Fig. 6). In doing so, we further revise the regional lithostratigraphic and sequence stratigraphic correlations of Oborny et al. (2020a).

The chemostratigraphic and subsurface geophysical data presented here provide significant insights into the chronostratigraphic correlations of strata assigned to the Salina Group within the Appalachian, Michigan, and Illinois Basins, and in particular, to strata assigned to the Greenfield and Tymochtee Formations in Ohio. Our identification of an expanded stratigraphic record for the Mulde isotopic excursion in central Ohio, in conjunction with regional works by other authors in Michigan and west-central and southern Ohio, demonstrates that the Tymochtee Formation of northern Ohio, as correlated from its type area by prior works, is in fact correlative with the type Greenfield Formation of southern Ohio. These data demonstrate that the upper Greenfield Formation contact is chronostratigraphically positioned at the first sequence bounding surface within the base of the subsurface B Unit. As such, we utilize the senior Greenfield Formation nomenclature (see combined Orton, 1871; Winchell, 1873) in redefining these strata and abandon the historical and often problematic use of the term “Tymochtee” in southern Ohio and suggest it may be beneficial to abandon the term “Tymochtee” altogether. In doing so, we informally define strata overlying the Greenfield Formation, spanning the first sequence bounding surface at the base of the B Unit through the basal Bass Islands Group contact, as “Salina B–G.”

Through our subsurface geophysical analyses, we demonstrate that “Salina B–G” strata are correlative throughout the study area. The depositional motif of these strata provides supporting evidence for the mid-late Homerian global sea-level lowstand immediately prior to the deposition of Greenfield Formation strata. The Greenfield Formation sediments filled in around thickened sites of apparent reef development and/or mounding in the underlying Bisher and Lilley Formations. Well-defined and uniform thickening of “Salina B–G” strata toward the east-northeast indicates that sediment accommodation was relatively low at the end of Greenfield Formation deposition within southern Ohio, suggesting a short interval of tectonic stability within this region prior to subsidence in and around the Ohio-Pennsylvania State line during deposition of upper “Salina B–G” strata.

Extension of chemostratigraphic data from central Ohio to similar regional works by other authors in west-central Ohio demonstrates that strata studied in those localities are at a chronostratigraphic position within the A1 salt unit and correlative to the lower Greenfield Formation and that the biostratigraphic and chemostratigraphic data of that region are representative of the Oz. bohemica interval zone and Mulde excursion (in part), respectively. Though we cannot directly assess the Pleasant Mills–Wabash formational contact, there does exist a strong lithological similarity between the Mississinewa Shale Member, at the base of the Wabash Formation, and “Salina B–G” strata. These units are also at similar stratigraphic positions, and, as such, we have no reason to file disagreement with previous interpretations. Therefore, we conclude that the upper contact of the Pleasant Mills Formation is correlative, or approximately correlative, with the upper contact of the Greenfield Formation, which confines the overlying shales of the Mississinewa Member and Wabash Formation of eastern Indiana and western Ohio to a chronostratigraphic position within the “Salina B–G” units.

We would like to thank the core librarian, Jeffrey Deisher, at the Ohio Division of Geological Survey (OGS) for providing spectral gamma-ray scans of the OW-11 core for Union County, Ohio. We would also like to thank Christopher Waid at the OGS for delivery of the OW-11 core to our laboratory at the University of Iowa and for providing unfettered access to subsurface geophysical data. Furthermore, we would like to thank Christopher Waid for verifying the identification of Devonian strata in our subsurface transects to ensure that our use of lithostratigraphic units was in agreement to those currently utilized by the OGS. Mark Kleffner also read an early version of our nomenclatural history, and we greatly appreciate his input. We would also like to thank Mark Kleffner and an anonymous reviewer for their reviews and comments, which improved this manuscript. This work was partially supported by the American Chemical Society–Petroleum Research Fund Doctoral New Investigator grant 53196-DNI8 and National Science Foundation Career grant 1455030 to B.D. Cramer. Journal fees were paid for by the Kansas Geological Survey.

1Supplemental Material. Plates S1-S6: Geophysical cross-sections of Upper Silurian–Lower Devonian strata in Ohio. Table S1: Isotopic data. Please visit https://doi.org/10.1130/GEOS.S.20477271 to access the supplemental material, and contact editing@geosociety.org with any questions.
Science Editor: Christopher J. Spencer
Associate Editor: Andrea Fildani
Gold Open Access: This paper is published under the terms of the CC-BY-NC license.