3: Meramecian–Chesterian (Upper Viséan) Conodont Biostratigraphy and Revised Lithostratigraphy along the Southwestern Flank of the Ozark Uplift, Southern Midcontinent, U.S.A.
Published:January 01, 2019
- PDF LinkChapter PDF
Cory J. Godwin, Darwin R. Boardman, II, James O. Puckette, 2019. "Meramecian–Chesterian (Upper Viséan) Conodont Biostratigraphy and Revised Lithostratigraphy along the Southwestern Flank of the Ozark Uplift, Southern Midcontinent, U.S.A.", Mississippian Reservoirs of the Midcontinent, G. Michael Grammer, Jay M. Gregg, James Puckette, Priyank Jaiswal, S. J. Mazzullo, Matthew J. Pranter, Robert H. Goldstein
Download citation file:
Four conodont biozones, including three subzones, are interpreted within a revised lithostratigraphic framework for the upper Boone Group and Mayes Group in northeastern Oklahoma and adjacent parts of Missouri, Kansas, and Arkansas. Although revised lithostratigraphy is principally based on observed lithologic characteristics and stratigraphic relationships, conodont biostratigraphic data played an important role in correlation and final organization of units. Within the upper Boone Group, Biozone 1 (lower Meramecian) includes the Ritchey Formation and the Tahlequah limestone and Biozone 2 (middle Meramecian) includes the Moccasin Bend Formation and Quapaw Limestone. The Mayes Group spans Biozone 3 and Biozone 4. Biozone 3 (upper Meramecian) is represented by the Bayou Manard Member of the Pryor Creek Formation (new name). Biozone 4 marks the appearance of definitive Chesterian conodont fauna. The lower two subzones within Biozone 4 correspond to the Lindsey Bridge (Biozone 4L) and Ordnance Plant (Biozone 4M) members of the Pryor Creek Formation, whereas the upper subzone consists of the Hindsville Formation (Biozone 4U).
Documentation of conodont taxa and recognition of the proposed biozones provides relative time constraints for genetically meaningful interpretations of regional geology and subsequent evaluation of the Mayes Group and upper Boone Group within a broader interregional context.
Meramecian through Chesterian strata of the Mayes Group and upper Boone Group are exposed along the western edge of the Mississippian outcrop belt in northeastern Oklahoma and adjacent parts of Kansas, Missouri, and Arkansas. These rocks are underevaluated and poorly understood in terms of both their regional stratigraphic framework and their roles within southern midcontinent geology. They are also important because time-equivalent strata are potential oil and gas producers to the west within the subsurface of Oklahoma. Correlations involving the Mayes Group and upper Boone Group across the outcrop belt and into the subsurface are numerous, equally variable, and largely lithostratigraphic in nature (Aurin et al., 1921; Buchanan, 1927; Brant, 1934, 1957; Cline, 1934; Selk, 1948; Barker, 1950; Huffman, 1958; Ellzey, 1961; Harris, 1975; Boyd, 2008; Mazzullo et al., 2011, 2013).
The primary purpose of this investigation was the construction of a refined regional stratigraphic framework for the Mayes Group and upper Boone Group in northeastern Oklahoma through the integration of conodont biostratigraphy and standard lithostratigraphy. Results reported herein include: (1) a revised regional lithostratigraphy, (2) the identification of four informal conodont biozones (including three subzones) and their correlation with established conodont zonation schemes of the Upper Mississippi River Valley, (3) the preliminary construction of a temporally constrained stratigraphic framework within the study area, and (4) the evaluation of these strata within a broader interregional context through conodont-based correlations with time-equivalent strata in the southern midcontinent.
This study is not directly concerned with subsurface stratigraphic problems in hydrocarbon producing areas of Oklahoma and Kansas. The results presented herein, however, provide a foundation for continued study of equivalent strata within the Mississippian outcrop area, as well as within the subsurface considering the position of study area along the transition between the surface and subsurface.
STUDY AREA AND METHODOLOGY
The study area encompasses the Mississippian outcrop area along the southwestern flank of the Ozark uplift in northeastern Oklahoma, including the type Mayes Group area and Tri-State Mining District, the latter includes adjacent areas of southwestern Missouri and southeastern Kansas, and Arkansas (Figure 1). Also included within the study area are location 36 in Washington County, Arkansas, and location 37 in Okmulgee County, Oklahoma. The former is included because it represents an important reference section within the Hindsville Formation type area, whereas the latter is included for its relevance to potential correlations of the Mayes Group into the subsurface of Oklahoma.
For this sample-based study, 28 surface exposures and 9 subsurface cores were examined, measured, described, and selectively sampled (Figure 1). For biostratigraphic analysis, bulk samples of at least 2 kilograms were taken from each sampled bed. The coarsest sampling interval used was meter-scale, with higher-resolution sampling of decimeter-scale textural or lithologic changes. Higher-resolution sampling was applied at type and principal reference localities. Of the subsurface cores examined, only the PM-21 core in Cherokee County, Kansas, and the MODOT B-49-8 core in Jasper County, Missouri, were available for bulk sampling. Samples from cores were taken at regular intervals of decimeter-scale, while accounting for lithologic and textural boundaries. The processing of bulk samples for the recovery of conodonts followed the procedure of Collinson (1963).
PREVIOUS CONODONT STUDIES
Very little is published concerning the Meramecian through middle Chesterian conodont biostratigraphy within the study area. Branson and Mehl (1941a) described and illustrated the holotype of Lochriea commutata from the Hindsville Formation (reported by them as Pitkin Limestone) in Craig County, Oklahoma. Thompson (1972) examined conodonts from the Hindsville Formation, Fayetteville Shale, and Pitkin Limestone in Missouri, Arkansas, and Oklahoma. Grayson (1974, 1976) reported on conodont fauna of the Hindsville Formation in northern Arkansas. Goebel et al. (1968) described Mississippian conodont taxa from the Tri-State Mining District, although they questioned the presence of Meramecian Boone Group strata there. Goebel (1968), Goebel et al. (1968), and Thompson and Goebel (1969) summarized Mississippian taxa from across Kansas, including the Tri-State Mining District. In an unpublished thesis, Routh (1981) recovered conodonts from the Mayes Group at three locations in northeastern Oklahoma, including location 5 and location 14 of this report. Conodont specimens collected from surface exposures in northeastern Oklahoma and subsurface cores in northern Oklahoma during the 1960s by workers from the Amoco Research Center in Tulsa, Oklahoma, were referenced or briefly discussed by Ormiston (1966), Selk and Ciriacks (1968), Selk (1973), and Brenckle et al. (1974).
It is necessary at this point to address proposed revision to Mayes Group and upper Boone Group lithostratigraphy within the study area (Figure 2). Inclusion of these revisions early in this report serves to introduce terminology and refined stratigraphic relationships, thereby avoiding confusion resulting from converting midway through this paper. Most of the following lithostratigraphic descriptions are based on physical observations and are independent of conodont biostratigraphic data. Conodont data supported certain facets of the revised lithostratigraphy presented below through the application of relative time-constrained regional correlations and interpretations. The impact of conodont data will be addressed following the description of conodont recoveries and the proposed informal biozones.
Snider (1915) defined the term “Mayes” for rocks stratigraphically positioned between the Osagean Keokuk Formation and the Chesterian Fayetteville Shale within Mayes County, Oklahoma. Huffman (1958) formally defined the Mayes Group and divided it into the “Moorefield Formation” and overlying Hindsville Formation, terms derived from their type areas in northern Arkansas. The “Moorefield Formation” in Oklahoma was subdivided by Huffman into the Tahlequah, Bayou Manard, Lindsey Bridge, and Ordnance Plant members, in ascending order. We herein propose the term Pryor Creek Formation (new name) as a replacement for the “Moorefield Formation” in Oklahoma, and include within it the Bayou Manard, Lindsey Bridge, and Ordnance Plant members. The Pryor Creek Formation is present throughout much of the Mississippian outcrop area of northeastern Oklahoma, south of the Tri-State Mining District. The Pryor Creek Formation in not currently continuous with the type Moorefield Formation of northern Arkansas across the Mississippian outcrop area, and questions remain as to whether the two were contiguous during deposition (Garner, 1967). Additionally, important lithologic differences exist between the Pryor Creek Formation and Moorefield Formation. The Pryor Creek Formation within the type area is defined by a generalized vertical succession of light brown–gray to dark gray lime mudstone–wackestone of the Bayou Manard Member, fine to very coarse-grained bioclastic packstone–grainstone of the Lindsey Bridge Member, and shaly calcareous siltstone of the Ordnance Plant Member. The Moorefield Formation, however, comprises goniatite-bearing dark brown–gray–black shale with some lenses of calcareous siltstone and limestone (Gordon, 1944; Garner, 1967; Handford, 1995). Although dark brown–gray–black shale is present within the Pryor Creek Formation and becomes more prominent as the unit is traced southward and westward from the type Mayes Group area of central Mayes County, it more closely resembles parts of the Caney Shale of southern Oklahoma to which it is geographically closer and with which it is interpreted as continuous (Huffman, 1958). Application of “Moorefield Formation” in Oklahoma is therefore confusing and commonly requires clarification as to which “Moorefield” is beingdiscussed. The proposed type locality for the Pryor Creek Formation is the Pryor Quarry in central Mayes County (location 15) and the unit derives its name from the Pryor Creek tributary of the Grand River south of the town of Pryor (Figure 3). All or most of the unit is well exposed within several high-wall sections, including its lower and upper contacts, at location 15. Adjacent surface exposures and shallow subsurface cores in central Mayes County serve as valuable reference sections (locations 11–14 and 16–19). The base of the Mayes Group is a major unconformity, herein informally named the sub-Mayes unconformity. Where the Pryor Creek Formation is present, the unconformity is placed at the base of the Bayou Manard Member (Figure 4A), elsewhere the unconformity is placed at the base of the Hindsville Formation. The surface is sharp and irregular. Chert clasts derived from the Boone Group are distributed throughout the Mayes Group but are commonly concentrated at or near the base. Huffman (1958) interpreted the boundary between the Moorefield Formation (Pryor Creek Formation of this report) and overlying Hindsville Formation as an unconformity based on the apparent truncation of the Ordnance Plant Member northward from central Mayes County, as well as a single surface section in which clasts believed to be derived from the Ordnance Plant Member were incorporated within the basal Hindsville Formation. This investigation yielded no conclusive evidence of an unconformity between the Pryor Creek Formation and Hindsville Formation, and the contact is tentatively considered conformable (Figure 4B). An unconformity is herein interpreted between the Ordnance Plant and Lindsey Bridge members of the Pryor Creek Formation (Figure 4C), which differs from interpretations of earlier workers (Huffman, 1958; Turmelle, 1982). This contact is typically sharp and flat to irregular with iron and phosphate staining, clasts derived from the Lindsey Bridge Member incorporated into the basal Ordnance Plant Member, and truncation of the Lindsey Bridge Member. The basal Ordnance Plant Member is also characterized in some sections by an increased abundance of chert clasts derived from the Boone Group.
For reasons to be addressed later in this report, the Tahlequah limestone (Tahlequah Member of Huffman, 1958) is excluded from the Mayes Group and included within the Boone Group. Conodont data played a substantial role in the change proposed for the Tahlequah limestone, which displays a stronger faunal relationship to the upper Boone Group strata than to the Mayes Group. The principal reference section (location 3) for the Tahlequah limestone (Figure 4D) is situated approximately 800 feet (240 m) southeast of the now poorly exposed type locality defined by Huffman (1958). The Tahlequah limestone is abundantly glauconitic, thin to thick-bedded, fine to medium-grained, bioclastic packstone–grainstone. Prior to its inclusion within the Mayes Group by Huffman (1958), the Tahlequah limestone was informally referred to as the “glauconitic limestone member” of the Keokuk Formation (Bentonville Formation of this report; Degraffenreid, 1953). In Cherokee County, Oklahoma, the Tahlequah limestone unconformably overlies Osagean Boone Group and is unconformably overlain by the Mayes Group.
Mazzullo et al. (2013) proposed the Ritchey Formation for cherty limestone above the lithologically similar Bentonville Formation of the Boone Group. Within the Oklahoma portion of the Tri-State Mining District, the Ritchey Formation replaces the “Baxter Springs Member” of McKnight and Fischer (1970), who included it within their “Boone Formation.” At its type locality in Newton County, Missouri (location 34), as well as location 21 in Jasper County, Missouri, and location 30 in Cherokee County, Kansas, the Ritchey Formation is predominantly medium-bedded, very fine to coarse-grained bioclastic wackestone–packstone–grainstone, with lenses and discontinuous beds of light-colored chert. This description also applies to the Ritchey Formation at locations 30 and 31, as well as exposures of the unit in Boone County, Arkansas, which are not included within this report (Mazzullo et al., 2013). In Ottawa County, Oklahoma, and at locations 32 and 33 in Newton County, Missouri, however, two distinct lithologic phases within the Ritchey Formation are recognized in this study (Figure 5A). The “upper” phase of the Ritchey Formation is lithologically similar to typical Ritchey Formation to the north and east, as described above. In contrast, the “lower” phase of the Ritchey Formation is very cherty lime mudstone with lenses of bioclastic wackestone–packstone. In some instances, the “lower” Ritchey Formation consists entirely of chert, such as at location 26. A third lithologic phase of the Ritchey Formation is also recognized in Ottawa County, Oklahoma (Figure 5B). Informally termed the “Fairland facies,” it is named for exposures within a quarry east of the town of Fairland (location 22) and was included within the “K” bed (term of informal mining district usage) by McKnight and Fischer (1970) and considered by them to be correlative to rocks defined in this report as the “upper” phase of the Ritchey Formation. Although certain lithologic aspects of the “Fairland facies” are similar to those of the “upper” phase of the Ritchey Formation, there are significant differences. The “Fairland facies” at location 22 is 20–30 feet (6–9 m) of medium-bedded, massively cross-stratified, medium to very coarse, oolitic and bioclastic, packstone–grainstone, with abundant glauconite and siliceous sponge spicules. Lenses and discontinuous beds of chert, typically 2–6 inches (5–15 cm) thick, occur within the upper 5 feet (1.5 m) of the “Fairland facies.” The “lower” phase of the Ritchey Formation is absent at location 22. The Ritchey Formation unconformably overlies the Bentonville Formation (including Short Creek Oolite Member; Mazzullo et al., 2013, 2019). The base of the Ritchey Formation is commonly irregular, glauconitic, and mineralized (iron, silica, phosphate) (Figure 5C). At location 32 in Newton County, Missouri, the top of the Short Creek Oolite contains unlined burrows that appear to be passively filled during deposition of the Ritchey Formation. At this same location, pebble-sized clasts of Short Creek Oolite Member are present within the basal Ritchey Formation. At location 22 in Ottawa County, Oklahoma, the Ritchey Formation truncates the upper Bentonville Formation, locally removing all of the Short Creek Oolite Member, and pebble-sized clasts of Short Creek Oolite Member are again present within the basal Ritchey Formation (Figure 5D).
McKnight and Fischer (1970) also included within their “Boone Formation” the “Moccasin Bend Member,” which is herein raised to formation rank and included within the Boone Group of Mazzullo et al. (2013). The type locality was defined by McKnight and Fischer (1970) and comprises a series of east-facing bluffs along the Spring River 6 miles (9.6 km) east of Miami, Oklahoma (location 25) (Figure 6A). Here, as much as 45 feet (14 m) of the Moccasin Bend Formation is cumulatively exposed, including the contact with the Ritchey Formation. The Moccasin Bend Formation typically consists of thin- to medium-bedded lime mudstone and microbioclastic (silt-sized) to very fine-grained wackestone–packstone with minor to moderate glauconite and lenses to discontinuous beds of light to dark colored chert. Beds of white to light brown silicified limestone are common. Historically termed “cotton rock” and similar to tripolite elsewhere within the Boone Group (Mazzullo et al., 2013), these rocks are lightweight and contain oil-stained very fine moldic porosity (created from the dissolution of calcareous allochems) and microporosity within the silicified lime mudstone–wackestone matrix. The Moccasin Bend Formation is well exposed at several other localities along the Spring River (locations 24, 26, 28, and 29), in a roadcut at location 23 east of Wyandotte, and within a quarry east of Vinita in Craig County, Oklahoma (location 21). Contrary to the interpretation of McKnight and Fischer (1970), the base of the Moccasin Bend Formation is an unconformity along which a 2–18 inch (5–45 cm) thick zone of glauconitic- and phosphate-rich shaly limestone with abundant chert clasts is present and is herein interpreted to be equivalent to the “J” bed of previous informal use within the Tri-State Mining District (Figure 6B) (Huffman, 1958; McKnight and Fischer, 1970).
The Quapaw Limestone conformably overlies the Moccasin Bend Formation and was defined by McKnight and Fischer (1970) for a single surface exposure in Ottawa County and interpreted occurrences in underground lead and zinc mines to the west. The Quapaw Limestone, however, was not included within the “Boone Formation” of McKnight and Fischer (1970), presumably due a lack of chert and mineralization. We propose including the Quapaw Limestone within the Boone Group of Mazzullo et al. (2013) based on its conformable relationship with the Moccasin Bend Formation. The type locality is incomplete and poorly exposed, but the Quapaw Limestone is now well-exposed in a quarry (location 27) south of the town of Quapaw, which is herein designated as the principal reference locality (Figure 6C). Here, the Quapaw Limestone consists of 25 feet (8 m) of oil-stained, medium- to thick-bedded, cross-stratified, fine- to very coarse-grained, bioclastic (crinoidal) packstone–grainstone.
Platform (P1) elements were primarily used for this study, but we recognized the ultimate need for studies utilizing multielement taxonomy. More than 14,000 specimens representing at least 22 identifiable platform species were recovered from more than 740 samples taken from the Mayes Group and upper Boone Group, as well as additional nonplatform and unidentifiable specimens numbering more than 30,000. Platform species recovered include Cavusgnathus altus (Harris and Hollingsworth, 1933), C. charactus (Rexroad, 1957), C. convexa (Rexroad, 1957), C. regularis (Youngquist and Miller, 1949), C. unicornis (Youngquist and Miller, 1949), Gnathodus bilineatus (Roundy, 1926; morphotypes 1 and 2), G. girtyi girtyi (Hass, 1953), G. linguiformis (Branson and Mehl, 1941b), Gnathodus n. sp. 15 aff. punctatus (Boardman et al., 2013), G. pseudosemiglaber (Thompson and Fellows, 1970), Gnathodus sp. A, Hindeodontoides spiculus (Youngquist and Miller, 1949), Hindeodus cristula (Younquist and Miller, 1949), Lochriea commutata (Branson and Mehl, 1941a), L. homopunctatus (Ziegler, 1960; Atakul-Ozdemir et al., 2012), L. mononodosus (Rhodes et al., 1969), Lochriea sp. B, Lochriea sp. A, Rhachistognathus sp. B (morphotypes 1–3), Taphrognathus–Cavusgnathus transitional form, Taphrognathus varians (Branson and Mehl, 1941b), and Vogelgnathus campbelli (Rexroad, 1957). Examples of the more significant form species are illustrated in Plate 1.
Conodont elements were recovered from all lithostratigraphic units, albeit not from every sample taken. Samples from the Tahlequah limestone, Ritchey Formation, Moccasin Bend Formation, Quapaw Limestone, and Hindsville Formation consistently yielded stratigraphically useful specimens. Typical recoveries from the Pryor Creek Formation were sparse in comparison, and biozone definitions within the formation were based primarily on good recoveries throughout the Lindsey Bridge Member and near the bases of the Bayou Manard and Ordnance Plant members, as well as at the base of the Hindsville Formation. Recoveries from the “lower” and “upper” Ritchey Formation yielded an average of 8 P1 elements per kilogram of rock sample, whereas recoveries from the “Fairland facies” averaged 36 P1 elements per kilogram. Recoveries from the Tahlequah limestone yielded more than 500 P1 elements per kilogram. The Moccasin Bend Formation and Quapaw Limestone yielded averages of 23 and 12 P1 elements per kilogram, respectively. Recoveries from the Bayou Manard and Ordnance Plant members of the Pryor Creek Formation were commonly 0–5 P1 elements per kilogram, with better recoveries (as many as 255 P1 elements per kilogram) near the base of each unit. The Lindsey Bridge Member yielded an average of 9 P1 elements per kilogram. The Hindsville Formation yielded an average of 23 P1 elements per kilogram.
Four informal biozones, including three subzones, were defined for the upper Boone Group and Mayes Group (Figure 7). Selected stratigraphic sections including taxonomic data used to define these biozones are illustrated in Figures 8, 9. A strong correlation between the proposed biozones and current lithostratigraphic divisions is clearly evident in Figure 7. This is not surprising because most lithostratigraphic boundaries are disconformable and tend to introduce some degree of biostratigraphic bias through obstruction of the natural stratigraphic ranges of various taxa (Barrick and Mannik, 2005). This does not, however, diminish the utility of these biozones in terms of regional correlations, nor does it adversely affect comparisons of recovered taxa with those reported by other workers. It simply highlights the potential incompleteness of the stratigraphic record within the study area. Perhaps in more distal areas, where deposition was more continuous, variations of the ranges of selected taxa will be slightly different. Until this is explored in more detail, however, the proposed biozones below remain valid for two reasons. First, recoveries made during this investigation generally agree with those reported by previous workers within the study area. Second, an overall agreement exists between the observed ranges of important taxa recovered in this investigation and those reported in the established conodont zonation schemes of Collinson et al. (1971), Lane and Brenckle (2005), and Boardman et al. (2013)
Biozones 1 and Biozone 2
Biozone 1 includes the Ritchey Formation and Tahlequah limestone and is defined by the first and only observed occurrence of Gnathodus n. sp. 15 aff. punctatus (Plate 1, Figure C; see also Boardman et al., 2013, plate 15, figure 7) and Gnathodus sp. A (Plate 1, Figure A), as well as the first common occurrence of Taphrognathus varians (Plate 1, Figure D). Other species include G. pseudosemiglaber (Plate 1, Figure B), G. texanus, and G. linguiformis. Present in the Tahlequah limestone, but not observed within the Ritchey Formation, were specimens of Lochriea homopunctatus (Plate 1, Figure M), which is the oldest known occurrence of this species in North America (Brenckle et al., 1974). Specimens herein designated Gnathodus sp. A were recovered alongside morphologically distinct specimens assigned to G. pseudosemiglaber (Plate 1, Figure B) as defined by Thompson and Fellows (1970, p. 88, 89; plate 2, figures 6, 8, 9, 11–13) and Thompson (1979). Specimens resembling Gnathodus sp. A were illustrated as G. pseudosemiglaber by Lane et al. (1980, plate 4, figures 15–17 and 19; plate 5, figures 8–15), Belka and Groessens (1986, plate 7, figures 1–3), Haywa-Branch (1988, plate 5, figures 8 and 9), Perri and Spalletta (1998, plate 1, figure 14; plate 2, figure 12), and Blanco-Ferrera et al. (2005, p. 22, figure 6, number 27). Specimens similar to Gnathodus sp. A were interpreted by Belka and Groessens (1986, plate 7, figures 4 and 5) as transitional to G. girtyi, by Nemyrovska (2005, plate 6, figures 2, 3, 5, 6, and 8) as transitional between G. pseudosemiglaber and G. girtyi meischneri, and by Singh (2007, plate 6, figures 4–7) as primitive morphotypes of, or transitional to, G. bilineatus. The boundary between Biozones 1 and 2 is placed at the youngest observed occurrences of Gnathodus n. sp. 15 aff. punctatus, G. pseudosemiglaber, and Gnathodus sp. A, and the oldest observed occurrences of Hindeodus cristula and species of Cavusgnathus (Plate 1). Biozone 2 includes the Moccasin Bend Formation and Quapaw Limestone, and it is distinguished by the co-occurrence of Taphrognathus and Cavusgnathus. Other taxa recovered from Biozone 2 include G. texanus and rare L. homopunctatus. The top of Biozone 2 is defined by the youngest occurrence of Taphrognathus. Recoveries from the Moccasin Bend Formation were faunally more diverse than those from the Quapaw Limestone, the latter predominantly yielded specimens of Cavusgnathus and Taphrognathus.
Conodont taxa recovered from the upper Boone Group for this study largely confirm the age assignments of previous workers (Huffman, 1958; McKnight and Fischer, 1970), albeit with some important differences. Biozone 1 (Ritchey Formation and Tahlequah limestone) is identical to the upper texanus-Gnathodus n. sp. 15 aff. punctatus Zone of Boardman et al. (2013) in terms of its interpreted stratigraphic range, but important differences include the identification of a potentially new species, Gnathodus sp. A, and inclusion of L. homopunctatus in recoveries from the Tahlequah limestone. Recovery of L. homopunctatus from the Tahlequah limestone, a species that extends into the Moccasin Bend Formation and younger strata, may indicate that the Tahlequah limestone is slightly younger than the Ritchey Formation. Together, Biozones 1 and 2 are roughly equivalent to the Taphrognathus varians-Apatognathus Zone of Collinson et al. (1971) and the upper half of the texanus Zone of Lane and Brenckle (2005), and therefore provide a higher resolution division of otherwise long-ranging zones. Upper Boone Group strata are faunally similar to time-equivalent strata in the Upper Mississippi River Valley, Kansas, and Missouri (Rexroad and Collinson, 1963; Rexroad and Collinson, 1965; Goebel, 1968; Thompson and Goebel, 1969; Thompson and Fellows, 1970; Collinson et al., 1971; Thompson, 1986; Lane and Brenckle, 2005). The Ritchey Formation and Tahlequah limestone (Biozone 1) are interpreted as early Meramecian in age, potentially latest Osagean, and partially equivalent to the Warsaw Formation of the Upper Mississippi River Valley. This interpretation generally agrees with those of previous workers (Cline, 1934; Huffman, 1958; McKnight and Fischer, 1970). The Moccasin Bend Formation and Quapaw Limestone (Biozone 2) are both early–late Meramecian in age and may be equivalent to the lower St. Louis Limestone of the Upper Mississippi River Valley based upon the co-occurrence of species of Taphrognathus and Cavusgnathus (Collinson et al., 1962; Rexroad and Collinson, 1963). In contrast, McKnight and Fischer (1970) considered the Moccasin Bend Formation to be Warsaw-equivalent and Quapaw Limestone to be Warsaw or possibly Salem-equivalent. Of note, the upper Salem is considered a facies equivalent of the lower St. Louis Limestone (Heckel, 2005).
Biozone 3 includes only the Bayou Manard Member of the Pryor Creek Formation and is characterized by the occurrence of Cavusgnathus without Taphrognathus. Also marking the base of Biozone 3 is the oldest observed occurrence of Hindeodontoides spiculus (Plate 1, Figure G). Other taxa recovered include Hindeodus cristula, Lochriea homopunctatus, and Gnathodus texanus. Typical recoveries from the Bayou Manard Member yielded one to three specimens of G. texanus. The best recovery came from the lower 5 feet (1.5 m) of the Bayou Manard Member at locations in central Mayes County, including locations 13 and 14. Biozone 3 (Bayou Manard Member) is interpreted as roughly equivalent to the Apatognathus scalensus–Cavusgnathus Zone of Collinson et al. (1971) and the scitulus–scalensus Zone of Lane and Brenckle (2005). Absent from recoveries from the Bayou Manard Member were specimens of Apatognathus scalensus and Hindeodus scitulus, both of which are cited by Maples and Waters (1987) as diagnostic of the upper St. Louis Limestone. Therefore, correlation between Biozone 3 and established conodont zones is largely based upon the occurrence of Cavusgnathus without Taphrognathus and the first occurrence of Hindeodontoides spiculus, both of which are characteristic of the upper St. Louis Limestone of the Upper Mississippi River Valley and Kansas (Rexroad and Collinson, 1963; Goebel, 1968; Thompson and Goebel, 1968; Collinson et al., 1971; Lane and Brenckle, 2005).
Biozone 4 includes the Lindsey Bridge and Ordnance Plant members of the Pryor Creek Formation and the overlying Hindsville Formation, each of which corresponds to one of three subzones (Figure 9). The boundaries between the Biozone 4 subzones are tentatively defined by subtle faunal variations. As a whole, Biozone 4 is defined by the first observed occurrences of definitive Chesterian taxa including Gnathodus bilineatus, G. girtyi girtyi, and Lochriea commutata (Plate 1, Figures E, N, Q, R). Other taxa include G. texanus, Hindeodus cristula, Hindeodontoides spiculus, L. homopunctatus, and species of Cavusgnathus. Also defining Biozone 4 are occurrences of specimens resembling Rhachistognathus muricatus (Dunn, 1965). Defined for specimens recovered from the uppermost Mississippian (upper Chesterian, Serpukhovian) through lowermost Pennsylvanian (Morrowan, Bashkirian) strata of Nevada (Dunn, 1965, 1970), R. muricatus has since been documented in time-equivalent strata throughout the western United States (Webster, 1969; Tynan, 1980; Wilson, 1982; Abplanalp et al., 2009), southern Oklahoma (Dunn, 1970; Lane and Straka, 1974), and Alaska (Kurka, 1997). Morphologically similar specimens were recovered from lower to middle Chesterian (Viséan) strata in the western United States (Tynan, 1980 as Rhachistognathus sp. A) and in the study area (Thompson, 1972 as Spathognathodus muricatus; Routh, 1981 as Rhachistognathus lanei). In all instances, the older Chesterian specimens are stratigraphically separated, by a gap in observed occurrence, from younger specimens of R. muricatus sensu stricto recovered within the same areas, thereby leaving questions as to their taxonomic relationship due to the unclear evolutionary lineage (Lane and Straka, 1974; Tynan, 1980). All recovered rhachistognathid specimens from the Mayes Group are herein referred to as Rhachistognathus sp. B, as not to be confused with Rhachistognathus sp. A of Tynan (1980). At least two morphotypes of Rhachistognathus sp. B are recognized within the Mayes Group. Morphotype 1 (Plate 1, Figures K, L) includes those specimens whose carina is discontinuous and centrally located, whereas morphotype 2 (Plate 1, Figures I, J) includes those specimens whose carina appears to be continuous with the left margin. No clear distinction in the stratigraphic ranges of the two morphotypes was observed. Separation between the lower and middle subzones (4L and 4M), a boundary that corresponds to the base of the Ordnance Plant Member, is chiefly based on the first observed occurrences of the G. bilineatus (morphotype 1), which was not definitively recovered from the Lindsey Bridge Member during this investigation. Routh (1981), however, reported recovery of G. bilineatus from the Lindsey Bridge Member at the Lindsey Bridge type locality (location 14). First occurrences of Lochriea sp. A and Lochriea sp. B (Plate 1, Figures O, P) also define the boundary between subzones 4L and 4M, but were only recovered from the base of the Ordnance Plant Member (subzone 4M) at location 9 and location 7. Although both may simply represent morphologic variations within L. homopunctatus, future work could demonstrate their utility, so they are included here. The platform of Lochriea sp A. is very similar to that of L. homopunctatus in that it is mildly asymmetric and tapers posteriorly, but differs in that it is relatively unornamented except for one to three poorly developed nodes. Lochriea sp. B also possesses a mildly asymmetric platform, but with distinctive ornamentation consisting of rows of nodes on each side of the carina that are slightly angled inward posteriorly. The ornamentation on L. homopunctatus also angles inward posteriorly, but is less organized. A single specimen identified as L. mononodosus was recovered from a sample within the Ordnance Plant Member at location 5. The boundary between the middle subzone (4M) and upper subzone (4U) is less definitive due to sparse recoveries within the upper Ordnance Plant Member and their overall faunal similarities. Subzone 4U is therefore defined by the first observed occurrence of G. bilineatus morphotype 2 and the apparent absence of L. homopunctatus. Subzone 4U also includes the first observed occurrence of Vogelgnathus campbelli (Norby and Rexroad, 1985) near the middle of the Hindsville Formation at location 36 in Washington County, Arkansas.
Biozone 4 generally corresponds to the early to middle Chesterian conodont zones of Collinson et al. (1971) and Lane and Brenckle (2005) based on the first occurrences of G. bilineatus, G. girtyi girtyi, Rhachistognathus sp. B, and L. commutata. A Chesterian age for these strata is in general agreement with interpretations of previous workers (Huffman, 1958; Thompson, 1972; Selk, 1973). Huffman (1958) did, however, interpret the Lindsey Bridge Member and Ordnance Plant Member as Meramecian and correlative to the St. Louis and Ste. Genevieve limestones, respectively. In the case of the Ordnance Plant Member, conodont recoveries of this study confirm its correlation with the Ste. Genevieve Limestone. However, we follow the interpretation of Maples and Waters (1987) who placed the Meramecian–Chesterian boundary at the base of the Ste. Genevieve, rather than include it at the top of the Meramecian. Placement of this boundary by Maples and Waters (1987) coincides with a significant faunal shift, which is easily recognized in this study by the first occurrence of several conodont species within Biozone 4, which also demonstrate that the Lindsey Bridge Member should be included in correlations with the Ste. Genevieve Limestone or equivalent lower Chesterian strata.
CONODONT BIOZONE TEMPORAL RESOLUTION
Conodont biozones represent relative time-constrained divisions of the rock record at higher temporal resolution than that provided by both the North American (Meramecian, Chesterian) and international (Viséan) chronostratigraphic divisions. The span of time represented by the upper Boone Group and Mayes Group is approximately 11 million years (Menning et al., 2006). The average length of time represented by each of the four proposed conodont biozones is therefore 2.7 million years per zone, which is generally comparable to the resolutions provided by the zonation schemes of Collinson et al. (1971) and Lane and Brenckle (2005). In addition, two of the zones of Lane and Brenckle (2005) extend into overlying or underlying strata and two of the zones of Collinson et al. (1971) are subdivided by the proposed zonal scheme of this report. Inclusion of Biozone 4 subzones provides a potential resolution of 1.8 million years.
REGIONAL STRATIGRAPHIC FRAMEWORK
Important physical elements of the revised regional lithostratigraphy were outlined earlier in this report. Hence, the following discussion emphasizes the integration of biostratigraphy and lithostratigraphy and the role of conodont data in enhancing our understanding of the upper Boone Group and Mayes Group.
Two distinct biostratigraphicaly constrained stratigraphic successions are present within the study area and define a revised regional stratigraphic framework (Figure 10). Both the Hindsville Formation and lower Boone Group are generally present throughout the study area, but important differences exist with regard to the rocks between them. In the Tri-State Mining District, strata of the upper Boone Group strata are present between the Hindsville Formation and Bentonville Formation, although upper Boone Group strata are locally absent due to erosion below the sub-Mayes unconformity. In the Tri-State Mining District, the Hindsville Formation overlies the Bentonville Formation at location 35, the Ritchey Formation at location 22, the Moccasin Bend Formation at location 21, and the Quapaw Limestone at location 25. In contrast, the Pryor Creek Formation is absent in the Tri-State Mining District, but it is widely distributed south of Craig County where upper Boone Group strata are largely absent below the sub-Mayes unconformity. In these areas, the Pryor Creek Formation most often overlies the Reeds Spring Formation or Bentonville Formation. In parts of Cherokee and Sequoyah counties, the upper Boone Group is represented by the Tahlequah limestone. Inclusion of the Tahlequah limestone in the Boone Group is one of the more significant lithostratigraphic revisions presented earlier in this report. In addition to the physical evidence of an unconformity at the principal reference locality (location 3), conodont biostratigraphic data demonstrate separation of the Tahlequah limestone (Biozone 1) and Mayes Group (Biozones 3 and 4) by a gap in time representing at least Biozone 2, further supporting the exclusion from the Mayes Group of the Tahlequah limestone. Faunal similarities and correlation with the Ritchey Formation (Biozone 1) also support the inclusion of the Tahlequah limestone within the Boone Group.
Upper Boone Group
The “lower” and “upper” phases of the Ritchey Formation are conformable and yielded similar Biozone 1 conodont taxa, thus a genetic relationship between them is inferred, and they are interpreted as a minor shallowing-upward succession following the development of the sub-Ritchey unconformity (Mazzullo et al., 2019). The “lower” phase is absent to the north and east (locations 30, 31, and 34) due to the more proximal depositional positions of those sections, whereas sections containing both phases are in more distal positions within the Tri-State mining district, assuming a general north–northeast to south–southwest depositional dip direction similar to that of older Mississippian strata (Lane and De Keyser, 1980). Occurrences of the “Fairland facies” and Tahlequah limestone to the south–southwest of the two-phase Ritchey Formation are anomalous because both units display moderate- to high-energy depositional characteristics and lack definitive evidence of low-energy deposition. Precise correlations between the “Fairland facies,” Tahlequah limestone, and Ritchey Formation are below the current biostratigraphic resolution and an overall one-to-one correlation is therefore assumed. Based on the interpretation of McKnight and Fischer (1970), the “Fairland facies” exposed at location 22 remains tentatively correlated to the “upper” phase of the Ritchey Formation. Large-scale cross-stratification within the “Fairland facies” indicates a north–northeastward prograding depositional dip direction, opposite of that generally interpreted for Mississippian strata in this area. It is therefore possible that the “lower” phase of the Ritchey Formation represents low-energy back-barrier deposition in a relatively proximal position. The barrier in this case is a possible paleotopographic high, perhaps related to the Kanoka ridge of Mazzullo et al. (2019), along which the higher-energy “Fairland facies” of the Ritchey Formation was deposited. Lithologic comparison between the “Fairland facies” and Tahlequah limestone suggest that the latter is a more distal expression of the former. Conodont abundance and diversity within the “Fairland facies” is greater than that of both the “lower” and “upper” phases of the Ritchey Formation to the north and east, and the abundance and diversity within the Tahlequah limestone is greater still. Although this represents a very simplified biofacies model, abundance and diversity trends within the “Fairland facies” and Tahlequah limestone suggest the interpretation that they represent increasingly offshore and open marine conditions to the south, but without significant deepening, condensed sedimentation, or sediment starvation.
Together, the relatively low-energy Moccasin Bend Formation and high-energy Quapaw Limestone (Biozone 2) record a shallowing-upward succession and transition following the development of the sub-Moccasin Bend unconformity. Unlike the Ritchey Formation and Tahlequah limestone, the Moccasin Bend–Quapaw succession is limited to the Oklahoma portion of the Tri-State Mining District due to a combination of erosion below the sub-Mayes unconformity and removal by modern erosion. Lack of surface exposures therefore limit our ability to address this relationship more fully at this time. Comparison of the Moccasin Bend Formation in sections of the Tri-State Mining District and the section exposed farther to the southwest at location 21 in Craig County indicated no evidence of deepening between the two areas. Because the Moccasin Bend Formation at location 21 is overlain by the sub-Mayes unconformity and Hindsville Formation, the original thickness of the Moccasin Bend Formation and possible deposition of Quapaw Limestone southwest of the Tri-State Mining District remain unknown. Additionally, the differences in faunal diversity between the Moccasin Bend Formation and Quapaw Limestone conform to their inferred relative depositional settings. Diverse fauna of the Moccasin Bend Formation, including specimens of Lochriea homopunctatus, suggests a more offshore setting (Burchette and Wright, 1992), whereas the predominance of Cavusgnathus and Taphrognathus in recoveries from the Quapaw Limestone suggest deposition within a shallow marine setting (Klapper and Barrick, 1978; Austin and Davies, 1984; Davies et al., 1994; Krumhardt et al., 1996).
Identification of the sub-Moccasin Bend unconformity was initially based on physical evidence outlined earlier in this report, but conodont data was critical in its subsequent correlation and interpretation. Erosion below the sub-Moccasin Bend unconformity includes an anomalous north-to-south truncation of the Ritchey Formation in Ottawa County, Oklahoma (Figure 11). At location 23 in Figure 11, a single 12–18 inch (30–45 cm) chert bed and 2–6 inches (5–15 cm) of shaly limestone are attributed to the Ritchey Formation between the Short Creek Oolite Member and Moccasin Bend Formation. Although the poorly exposed beds above the sub-Moccasin Bend unconformity at location 23 may be easily misidentified as belonging to the “lower” phase of the Ritchey Formation, especially considering its vertical proximity to the top of the Short Creek Oolite Member. Conodont recoveries demonstrate that these strata are within Biozone 2 and are therefore the Moccasin Bend Formation.
Impact of Upper Boone Group Conodonts on Interpretations Involving the Lower Boone Group
Conodont recoveries from the upper Boone Group (Biozone 1 and Biozone 2) impact not only correlations of these strata but also interpretations involving the lower Boone Group. Lithologic differentiation of the Ritchey and Bentonville formations, for example, is difficult in most places and the intervening Short Creek Oolite Member, when present and exposed, serves as a valuable stratigraphic marker (Thompson, 1986; Mazzullo et al., 2019). This is especially true where only the “upper” phase of the Ritchey Formation is present. Thompson (1986) designated the top of the Short Creek Oolite as the top of the Burlington–Keokuk and also as the Osagean–Meramecian boundary, but often this boundary is considered chiefly lithostratigraphic in nature. In addition to lithologic similarities, the Bentonville Formation and Ritchey Formation (including Tahlequah limestone) are faunally similar based upon the observed ranges of forms attributed to Gnathodus linguiformis, G. pseudosemiglaber, G. texanus, and Taphrognathus varians (Rexroad and Collinson, 1965; Thompson and Fellows, 1970). Despite these similarities, a clear faunal distinction corresponding to the physical boundary between the Ritchey and Bentonville formations was recognized in this investigation and that of Boardman et al. (2013). Biozone 1 (Ritchey Formation) includes the first occurrences of Gnathodus n. sp. 15 aff. punctatus (Boardman et al., 2013) and Gnathodus sp. A of this report, as well as a significant increase in the occurrence of T. varians. Therefore, the contact between the Ritchey Formation and Bentonville Formation is more than lithostratigraphic in nature. It is a biostratigraphically definable boundary within the study area, and potentially into the subsurface of Oklahoma.
Misidentification of the Moccasin Bend Formation as the Reeds Spring Formation by Laudon (1939) and Zeller (1950) in Ottawa County near location 24 of this study and by the senior author of the present paper (as presented in Mazzullo et al., 2013) at location 21 of this study in Craig County is the result of lithologic similarities between the two units. In the latter example from Craig County, the original interpretation as Reeds Spring Formation was based on lithology alone. Subsequent recovery of fauna representing Biozone 2 demonstrated that these strata are Moccasin Bend Formation, and definitely not Osagean in age. Although these mistaken identifications may simply be isolated incidents, it brings into question lithostratigraphic interpretations of Reeds Spring Formation strata outcropping in northeastern Oklahoma in sections without exposed regionally recognized contacts.
Impact of Conodont Data on the Mayes Group of Northeastern Oklahoma
Within the Mayes Group, the greatest potential impact of conodont biostratigraphy is the refinement of intraformational correlations in the study area, especially away from the excellent surface and subsurface sections of central Mayes County and into areas where lithostratigraphic correlations break down due to combination of abundant incomplete sections and observed lithologic similarities between units of the Mayes Group. For example, lithologies that define the members of the Pryor Creek Formation are also present within the Hindsville Formation. An incomplete exposure consisting of lime mudstone, coarse-grained bioclastic packstone–grainstone, and calcareous siltstone may therefore represent the typical succession within the Pryor Creek Formation, but it may also represent the Hindsville Formation. Other examples include the occurrences of very fine-grained bioclastic packstone–grainstone (common to Lindsey Bridge Member) in the Bayou Manard and Ordnance Plant members, dark gray shaly lime mudstone–wackestone (similar to the Bayou Manard Member) in the Lindsey Bridge and Ordnance Plant members, and fine to coarse bioclastic wackestone–packstone–grainstone in the Ordnance Plant Member. Lithostratigraphic breakdown of this nature was described by Huffman (1958) and Turmelle (1982) as interfingering of facies but may represent depositional cyclicity within the Mayes Group (Godwin and Puckette, 2015). Differentiation between units within the Pryor Creek Formation also becomes difficult to the south and west of central Mayes County, where these rocks are predominantly shaly. Although further work is needed, especially the collection of larger bulk samples due to the typically poor recoveries per kilogram, conodont biostratigraphic data may prove useful in subdividing and correlating sections of undifferentiated Pryor Creek Formation, or refining previous lithostratigraphic interpretations. A specific example is the section at location 5. Here, Huffman (1958) attributed as much as 28 feet (9 m) of the shale-dominated section to the Ordnance Plant Member. Initial conodont recoveries suggest that the Ordnance Plant Member, or at least strata attributable to Biozone 4M, may be restricted to only the upper 6 feet (2 m) of the section, with the underlying part of the section being in the Bayou Manard Member (Biozone 3).
The sub-Mayes unconformity is the most significant stratigraphic surface within this study because it is the only surface across which relative time, measurable within the resolution of the current biostratigraphic data, is clearly missing. All other unconformities within this study separate units representing successive conodont biozones. The Bayou Manard Member (Biozone 3) most commonly overlies the Osagean Reeds Spring and Bentonville formations, but it locally overlies the Tahlequah limestone (Biozone 1) in Cherokee and Sequoyah counties and the Devonian Woodford Shale in southern Muskogee County (Huffman, 1958). In Mayes County core M-211 (location 10), the Pryor Creek Formation unconformably overlies Ordovician strata. Farther to the west, in Okmulgee County, the Pryor Creek Formation rests on the Lower Mississippian St. Joe Group in the Baker Hughes BH-1 core (location 37). In both the cores at location 10 and location 37, as well as the core at location 11, the Pryor Creek Formation is thicker than is typical across the western edge of the Mississippian outcrop belt. The Pryor Creek Formation is 229 feet (70 m) thick in the core at location 10 and 213 feet (65 m) thick in the core at location 37. In core M-207 (location 11), the Pryor Creek Formation rests unconformably on the Reeds Spring Formation and is 126.6 feet (38.6 m) thick. In contrast, the Pryor Creek Formation ranges from 0 to 95 feet (0–29 m) thick along its outcrop area in the northeastern Oklahoma. Expansion of the Pryor Creek Formation appears to occur primarily within the Bayou Manard Member. In parts of Adair and Cherokee counties in Oklahoma, the Pryor Creek Formation is also absent and the Hindsville Formation rests on Osagean Boone Group strata. Thickening of the Mayes Group at the expense of underlying Mississippian strata was discussed by previous workers in support of interpreted correlations between the Mayes Group and subsurface strata in Oklahoma informally known by as the “subsurface Mayes,” “Mississippi black limestone,” “Seminole Mayes,” or “Ada-Mayes” (Aurin et al., 1921; Buchanan, 1927; Brant, 1934; Cline, 1934; Laudon, 1948; Huffman and Barker, 1950; Huffman, 1958). Other workers believed the “subsurface Mayes” to represent a downdip facies of Kinderhookian or Osagean strata of Kansas and northern Oklahoma (Cram, 1930; Brant, 1934, 1957). Although restrictions prevented biostratigraphic sampling of the Mayes County shallow subsurface cores at locations 10, 11, and 16–19 and the Baker Hughes BH-1 core at location 37 in Okmulgee County, lithostratigraphic correlations appear to support correlation between the Mayes Group and “subsurface Mayes”—a section that comprises an expanded Pryor Creek Formation and distally thinning Hindsville Formation. Assuming the unconformities observed at the bases of the Pryor Creek and Hindsville formations are correlative, the Mayes Group displays a general south–southwest truncation of older strata and filling of post-Boone accommodation space by a basinward thickening Pryor Creek Formation (primarily the Bayou Manard Member).
INTERREGIONAL CORRELATIONS IN THE SOUTHERN MIDCONTINENT
Interregional correlations are discussed below and are the product of comparisons between recovered conodont taxa, proposed biozones, and interpretations of this study and those of previous workers. These biostratigraphically constrained comparisons suggest that the stratigraphic architecture of northeastern Oklahoma is an expression of a common theme throughout the southern midcontinent. In short, in many parts of the southern midcontinent, Meramecian and older strata are locally to regionally absent below a major unconformity (sub-Mayes unconformity) and are overlain by basinward-thickening successions.
Away from the study area, few strata time-equivalent to upper Boone Group (Biozone 1 and Biozone 2) are recognized within the southern midcontinent. A summary of these correlations are illustrated in Figure 13. In addition to this study, Warsaw through St. Louis strata are commonly recognized in both Kansas and Missouri and are considered to be generally contiguous with those of the Upper Mississippi River Valley (Goebel, 1968; Thompson and Geobel, 1969; Thompson 1986). Strata equivalent to Biozone 1 and Biozone 2 are also present in southern New Mexico and west Texas (Lane, 1974) and in the Texas panhandle (Ruppel and Lemmer, 1986), areas that are situated along the Lake Valley shelf and Chappel shelf, respectively, and represent southwestward extension of the Burlington shelf (Lane and De Keyser, 1980; Gutschick and Sandberg, 1983). Along the Llano uplift of central Texas, within the Chappel shelf area, the Barnett Shale (Chesterian) is underlain by the White’s Crossing limestone of Osagean to Meramecian age, which in turn rests on the Kinderhookian–Osagean Chappel Limestone (Hass, 1953, 1959; Turner, 1957; Grayson and Merrill, 1991). Conodont specimens illustrated by Singh (2007, plate 6, figures 4–7) from the White’s Crossing Limestone are very similar to Gnathodus sp. A of this study and suggest possible correlation to Biozone 1.
Unlike the upper Boone Group (Biozone 1 and Biozone 2), to which few identified strata within the southern midcontinent are correlative, biostratigraphically constrained correlations between the Mayes Group and a number of time-equivalent strata in the southern midcontinent are clearly evident. Based on taxa recovered during this investigation and evaluation of those reported by previous workers (Roundy, 1926; Hass, 1953; Schwartzapfel, 1990; Boardman and Puckette, 2006; Singh, 2007), the Hindsville Formation (Biozone 4U) is correlative to the Delaware Creek Member of the Caney Shale of southern Oklahoma and the lower Barnett Shale of central Texas. In addition, the Fayetteville Shale has been correlated to the Sand Branch Member of the Caney Shale and upper Barnett Shale (Thompson, 1986). Furthermore, ammonoid-based correlations between the Hindsville Formation, Fayetteville Shale, Caney Shale, and Barnett Shale generally support those based on conodont data (Saunders, 1973). Conodont recoveries from the Lindsey Bridge and Ordnance Plant embers of the Pryor Creek Formation (Biozones 4L and 4M) suggest they should be included in interregional correlations of the Hindsville Formation. Subzones proposed within Biozone 4, however, are not recognized outside of the current study area. Chesterian strata equivalent to Biozone 4 are interpreted within the Rancheria Formation of west Texas and southern New Mexico (Lane, 1974) and in Kansas (Goebel, 1968). Despite lack of published conodont recoveries from the Moorefield Formation of northern Arkansas, it is considered correlative to the Pryor Creek Formation based on historic correlations of benthic macrofauna (Girty, 1909, 1911, 1915; Gordon, 1944; Huffman, 1958) and ammonoid-based correlations of Moorefield Formation to the Caney Shale and Barnett Shale (Saunders, 1973), to which Lindsey Bridge and Ordnance Plant members (Biozones 4L and 4M) are herein correlated.
Unconformities and stratigraphic relationships comparable to the sub-Mayes unconformity are present elsewhere within the southern midcontinent, recording a common theme of expanded postunconformity sections at the expense of preunconformity section (Figure 12). In the subsurface of north–central Oklahoma, Selk (1973) and Selk and Ciriacks (1968) reported recoveries of St. Louis conodont fauna, interpreted as belonging to the Bayou Manard Member, from cores in Grant, Major, Noble, Osage, Pawnee, and Payne counties. These St. Louis conodont faunas were recovered from rocks overlying and underlying strata yielding Kinderhookian and Chesterian conodonts, respectively. In Boone County, Arkansas, the Moorefield Formation is absent and the Hindsville Formation rests unconformably on the Ritchey Formation of the Boone Group (Laudon, 1948; Mazzullo et al., 2013). To the east–southeast, the Moorefield Formation unconformably overlies the Boone Group, conformably underlies the Hindsville Formation and Batesville Sandstone, and thickens distally to the south and southeast (Handford, 1995). Furthermore, thinning of both the Moorefield Formation and the Pryor Creek Formation toward the Oklahoma–Arkansas state line suggest that area was possibly a positive feature during deposition of the two units and the two units were not depositionally contiguous. In southern New Mexico, an unconformity was interpreted between the Rancheria and Lake Valley formations, with the former thickening to the south at the expense of the latter (Lane, 1974; Greenwood et al., 1977; Bachtel and Dorobek, 1998). Along the Llano uplift, the Barnett Shale overlies the Chappel Limestone (Kinderhookian–Osagean) (Hass, 1953, 1959; Singh, 2007; Boardman et al., 2012). Previous workers have interpreted the contact between the Barnett Shale and Chappel Limestone as conformable (Zachry, 1969; Montgomery et al., 2005). Grayson and Merrill (1991, figure 21), who also interpreted a physically conformable relationship, but reported Chesterian taxa consistent with Biozone 4 of this study at the base of the Barnett Shale, directly above the Chappel Limestone containing Kinderhookian–Osagean taxa. Regardless of the physical expression of the unconformable contact, a clear biostratigraphically constrained time gap exists between the Barnett Shale and Chappel Limestone spanning much or all of the Meramecian and Osagean, without evidence of condensed sedimentation (Hass, 1953, 1959; Singh, 2007). Traced into the subsurface of the Fort Worth Basin, the Barnett Shale thickens and overlies Ordovician strata (Montgomery et al., 2005). In the northern Arbuckle uplift area, the Ahloso Member of the Caney Shale unconformably overlies the Kinderhookian–Osagean Welden Limestone, but to the south, it overlies the Woodford Shale (Elias, 1956). In the southern Arbuckle uplift area, the Ahloso Member is absent and the Sycamore Limestone unconformably overlies the Woodford Shale. Overlying the Sycamore Limestone are strata interpreted as the Delaware Creek Member of the Caney Shale (Elias, 1956; Haywa-Branch, 1988; Schwarzapfel, 1990). The Stanley Shale of southern Oklahoma also unconformably overlies the Woodford Shale (Hass, 1950, 1951; Laudon, 1959).
In addition to the basal unconformity, many of the postunconformity strata discussed above share similar conodont faunal characteristics. Recoveries from the Bayou Manard Member are typically dominated by Gnathodus texanus, and those at or near the base of the unit contain reworked taxa, including Osagean G. bulbosusThompson (1967) and G. pseudosemiglaber. Stratigraphic mixing due to unconformity-related reworking occurs within lowermost Sycamore Limestone (Ormiston and Lane, 1976; Schwartzapfel, 1990) and Ahloso Member of the Caney Shale (Haywa-Branch, 1988; Haywa-Branch and Barrick, 1990). The Ahloso Member is the least biostratigraphically constrained part of the Caney Shale, but, as part of ongoing research representing an extension of this investigation, two samples were taken from the Ahloso Member of the Caney Shale at the Hass “G” locality in Pontotoc County, Oklahoma. These samples were taken from two brachiopod-rich calcareous and shaly siltstone beds approximately 3.5 feet (1 m) and 7 feet (2 m) above the interpreted top of the Welden Limestone. Recoveries from the lower brachiopod bed were dominantly composed of specimens attributable to G. texanus. Recoveries from the upper brachiopod bed also included G. texanus, but also yielded G. bilineatus, Rhachistognathus sp. B, and Lochriea commutata. Based on these recoveries, the upper brachiopod bed is correlative with Biozone 4, whereas those of the lower brachiopod bed are similar to those of Biozone 3, if only due to the predominance of G. texanus. Poor recoveries, predominantly consisting of forms interpreted as G. texanus were also reported from the lowermost Stanley Group and subsequently overlain by strata containing G. bilineatus (Hass, 1950; 1951). Although the Barnett Shale is clearly Chesterian in age, some workers have interpreted an Osagean or Meramecian age, at least within the basal part (Ellison, 1989), possibly due to stratigraphic mixing. Above the unconformity at the base of the Las Cruces and Rancheria formations, Lane (1974) indicated that the oldest conodont recoveries consisted of G. texanus with no conclusively younger taxa, and were consequently considered to be late Osagean through early Meramecian. This zone was reportedly overlain at one locality by a zone consisting of G. texanus, Cavusgnathus altus, the Taphrognathus–Cavusgnathus transitional form, and reworked Kinderhookian–Osagean taxa, suggesting correlation with at least Biozone 2, if not Biozone 3 due to the lack of definitive specimens of Taphrognathus (Lane, 1974).
Interpretations of stratigraphic architecture within the southern midcontinent have been tied to early phases of Ouachita tectonism. To explain the absence of Kinderhookian and Osagean strata in parts of the southern midcontinent, Noble (1993) interpreted a lower Mississippian depositional hiatus stemming from sediment starvation and localized erosion associated with changing marine circulation during early phases of Ouachita tectonism. A similar lack of Meramecian strata (Biozones 1 and 2) within the southern midcontinent was noted by Noble (1993) and is evident from comparisons of conodont recoveries reported by others with those recovered from the upper Boone Group in this study. In recent work by Mazzullo et al. (2011), Boardman et al. (2013), and Mazzullo et al. (2019), anomalous Kinderhookian and Osagean stratigraphic architecture is attributed to periodic fore-bulge uplift and relaxation associated with early Ouachita tectonism and lack evidence indicative of sediment starvation and condensed sedimentation to the south and southwest of the Mississippian outcrop area. Likewise, development of the sub-Ritchey and sub-Moccasin Bend unconformities and the distribution of depositional facies within the upper Boone, interpreted within the constraints of provided by the proposed biozones, display some anomalous stratigraphic relationships and lack definitive deepening profiles into a starved basin area as would be expected in the model proposed by Noble. Thus, evidence from the upper Boone Group, including development of the high-energy Fairland facies along a possible paleotopographic high within the southwestern part of the Tri-State Mining District, erosional remnants of Biozone 1 (Tahlequah limestone) farther southward, and the southward truncation of the Ritchey Formation by the sub-Moccasin Bend unconformity, support the interpretation of syndepositional tectonism associated with fore-bulge flexure and relaxation.
Furthermore, high-energy facies of the Lindsey Bridge Member at locations 14 and 15 in central Mayes county displays northeastward progradation, which was first noted by Swinchatt (1967) in his evaluation of the unit at location 14. Northeastward progradation of the Lindsey Bridge Member is associated with observed thinning of both the Lindsey Bridge and Bayou Manard members across a paleotopographic high at location 13 at which a remnant of the Bentonville Formation is present above the Reeds Spring Formation (Figure 13). The sub-Mayes unconformity and its interregional correlative unconformities represent a significant event within the southern midcontinent, and it is below this unconformity that the widespread hiatus cited by Noble (1993) is commonly placed. The wide range of strata below the sub-Mayes unconformity, which range from Ordovician through Middle Mississippian (Meramecian) indicates that uplift and erosion played a more significant role in the development of the hiatus than did sediment starvation and erosion by geostrophic currents as suggested by Noble.
In the simplest terms, conodont biostratigraphy provides one method by which rocks are constrained in relative time and more accurately correlated, regardless of the inherent limitations of basic lithostratigraphy, including incomplete sections, complex or anomalous stratigraphic relationships, and lithologic similarities between temporally distinct strata. Although we recognize the potential for stratigraphic architecture below the resolution of current conodont biostratigraphic data, the proposed biozones still provide an improved set of temporally constrained boundaries between which those higher-resolution architecture and facies distributions may be interpreted. Conodont biozones proposed for the upper Boone Group and Mayes Group therefore improve lithostratigraphically based interpretations and correlations within the study area and allow for their evaluation within a broader interregional context.
During this investigation a number of key observations were made that potentially impact surface and subsurface correlations and geologic interpretations within the southern midcontinent:
Meramecian through middle Chesterian (post-Osagean–pre-Hindsville) rocks are present along the westernmost edge of the Mississippian outcrop belt and represent potentially important surface analogs for underevaluated components of the complex subsurface Mississippian section of Oklahoma.
Within the study area of northeastern Oklahoma, the regional stratigraphic framework includes important temporal and genetic distinctions between the stratigraphic succession of the Tri-State Mining District where upper Boone Group (Biozones 1 and 2) strata are present and the succession to the south where those strata are absent and the Pryor Creek Formation (Biozones 3, 4L, and 4M) is present below the Hindsville Formation.
The importance of the above geographic distinction within the study area is exemplified by relationship between the Bayou Manard Member (Pryor Creek Formation) and the Moccasin Bend Formation. Both are broadly considered equivalent to the St. Louis Limestone and are underlain by unconformities, yet the two units are faunally distinct. Therefore, a broad-brushed correlation of these units with the St. Louis Limestone is misleading. The Moccasin Bend Formation and Quapaw Limestone (Biozone 2) are equivalent to the lower St. Louis Limestone and upper Salem Limestone, whereas the Bayou Manard Member (Biozone 3) is equivalent to the upper St. Louis Limestone. These faunal differences are documented elsewhere, including within the Upper Mississippi River Valley.
Globally correlative sea-level falls were interpreted by Ross and Ross (1985) both before and after deposition of the lower St. Louis Limestone–upper Salem Limestone. Thus, the unconformity at the base of the Moccasin Bend is herein interpreted to be associated with pre-St. Louis sea-level fall and potentially correlative to the sub-“St. Louis” unconformity of Witzke et al. (1990), whereas the sub-Mayes unconformity appears to correspond to a mid-St. Louis sea-level fall.
Separation between the upper Boone Group section of the Tri-State mining district and the Mayes Group-dominated area to the south is marked by the sub-Mayes unconformity and expansion of parts of the Mayes Group at the expense of pre-Mayes strata.
Temporal and genetic separation between the Tahlequah limestone (Biozone 1) and the Pryor Creek Formation (Biozones 3 and 4) across the sub-Mayes unconformity, along with faunal correlation between the Tahlequah limestone and Ritchey Formation (Biozone 1), supports the removal of the Tahlequah limestone from the Mayes Group and its subsequent inclusion within the Boone Group. Occurrences of the Tahlequah limestone south of the Tri-State mining district represent erosional remnants below the sub-Mayes unconformity.
This paper would not be possible without the inspiration, contribution, and direction of Darwin R. Boardman II, who passed away in January of 2015.
Our understanding of the regional geology has also been greatly enhanced through the willingness of quarry operators in Oklahoma to grant access to their properties. These include: BuzziUnicem U.S.A (Pryor Quarry), Kemp Quarries (Pryor Quarry, Fairland Quarry, Neosho Quarry), Midwest Minerals (Quapaw Quarry), and APAC (Vinita Quarry). Parts of this study were supported by the Oklahoma State University Mississippian Carbonates Consortium. Conodont SEM photographs were taken in the Oklahoma State University Microscopy laboratory in Stillwater, Oklahoma. Access to the MODOT B-49-8 core from Jasper County, Missouri was provided by the Missouri Department of Natural Resources in Rolla, Missouri, with a special thanks to Dr. Tom Thompson. Access to the PM-21 core in Cherokee County, Kansas was provided by the Kansas Geological Survey in Lawrence, Kansas. Subsurface cores in Oklahoma were provided by the Oklahoma Geological Survey in Norman, Oklahoma.
We also thank S. J. Mazzullo and B. W. Wilhite for their critical review of the manuscript, as well for their input during open discussions in field concerning Mississippian strata in Oklahoma, Kansas, Missouri, and Arkansas, which were incredibly valuable throughout this investigation. We would also like to thank G. M. Grammer for his editorial input and direction in completion of this manuscript. I would also like to thank all the people at AAPG and on the editorial staff who also took the time to help find and fix many of our less obvious mistakes.
CONODONT PLATE 1
(All Specimens ×60; Scale bar in lower right-hand corner is 500 microns.)
All specimens held at the Paleontology Repository, Department of Earth and Environmental Sciences, University of Iowa.