The Bois d'arc Formation of Southern Oklahoma PArt I

---

Authors Note: It is recommended that the article on the corals of the Birdsong Shale is read first as it will put some of the comments made in the following article into context.

The Bois d'Arc Formation of Southern Oklahoma was deposited in the western arm of the Lower Devonian Epeiric Sea. Deposition was across portions of the continental shelf in relatively calm waters at a depth often below those disturbed by effective wave action. Amsden (1958) stated, "The Bois d'Arc Formation is essentially a carbonate sequence with varying amounts of chert. The lower part (Cravatt Member) consists of argillaceous and silty calcilutites or marl stones, which in most areas grade upwards into relatively pure calcarenites. Beds of fossiliferous calcarenite up to 6 inches or so in thickness are present throughout the Cravatt... The almost complete absence of shallow water features such as cross bedding and channeling, combined with the fine grained character of the [rocks] suggest deposition in quiet waters." (p.10 and p.20)

The Bois d'Arc Formation is considered a facies of the Haragan Formation, representing the upper portion of the sequence. Like the rest of the Haragan it is richly fossiliferous and is dominated by brachiopods.

Based on the fauna, the Bois d'Arc has been correlated with the Birdsong Shale of Tennessee and to a lesser degree the New Scotland of New York. For this work the similarities between the Birdsong Shale and Bois d'Arc, specifically the Cravatt Member, are of primary focus due to the relative ease in which fossils from these areas can be obtained by the hobbyist.

References:

Amsden, T.W. 1958. Stratigraphy and Palaeontology of the Hunton Group in the Arbuckle Mountain Region - Part V - Bois d'Arc Brachiopods. Oklahoma Geological Survey Bulletin 82.

No single work exists in which the entire fauna of the Haragan and its facies has been described.  Works on the conodonts, ostracods, foraminifera, graptolites, brachiopods, and trilobites were  published by the Oklahoma Geological Survey during the 1950's to 1970's and the bryozoans were described in Oklahoma Geology Notes, but essentially all of the molluscs, echinoderms, corals and various minor groups remain unstudied. Compiling all these disparate groups into a single reference is beyond the scope of this work, but research by the author into both the Lower Devonian of Oklahoma and Tennessee does provide previously unpublished data that correlates the fauna in more detail than previously available.

A guide to the macrofossils of Cravatt Fauna will follow at a later date.

The coral fauna of the Cravatt is limited in diversity but large numbers of individuals are preserved in the fossiliferous beds. The two species that overwhelmingly dominate are Favosites conicus and Streptelasma sp., which combined account for over 90% of the coral fauna. This is an identical situation to that seen in the Birdsong Shale (which was deposited on what is termed the Ross shelf) and indicates the palaeoenvironment of the two locales was perfect for anthozoans - warm water, below regular wave activity, and where sedimentation rates were usually low.

Pleurodictyum lenticulare accounts for approximately 1% of the Cravatt fauna, which is  less than the Birdsong where Pleurodictyum (represented by two species) varies from 2 to 10% of the coral fauna. However, the species is widespread and has proven to be a useful in correlation.

The remaining coral types that have been recorded in the Cravatt are Favosites helderbergiae?; a Favosites species (right) that is characterised by a globular shaped colony with small corallite apertures (about one third that of F. conicus - 1mm vs 3mm); Aulopora schoharie, which occurs in the Birdsong and in New York; and Striatopora sp.,  (also known from a very limited number of Tennessee specimens) that appears to match specimens of Striatopora issa recorded in New York by Hall (1887).  All of these latter types are found only infrequently in numbers that are statistically insignificant but with the exception of the unusual globular Favosites they all occur across the entire Lower Devonian Epeiric Sea.

Reference:

Hall, J. 1887. Palaeontology of New York, Vol. 6. Corals and Bryozoa.

Favosites conicus is the most important part of the Cravatt coral fauna as it is also known from both New York and Tennessee. Truly cosmopolitan in nature the species ranged across the entire Lower Devonian Epeiric Sea and is one of the key species used in correlation of the three deposits. 

It has not been possible to examine any collections from New York to determine how abundant the coral was (and Hall (1887) states nothing about their average size), but comparison with specimens from Tennessee suggest that the species achieved greater success in the Cravatt where it could reach sizes considerably larger than those seen in Tennessee (on average 50% larger, often more). This larger size suggests many of the Oklahoman specimens were also able to survive for more than one year (see article of the Birdsong Shale for the theory that F. conicus was monocarpous) because the greater water depth meant catastrophic storms less frequently reached the ocean floor. From James Hall's illustrations of New York specimens of F. conicus, the species attained a larger size than in either Oklahoma or Tennessee, suggesting the deeper waters of the New Scotland Helderbergian deposits provided a calm enough environment for the species to survive multiple years. 

[Favosites foerstei, a mid sized polycarpous species found in Tennessee is absent in the Cravatt with larger specimens of F. conicus apparently filling the niche and growing to comparable sizes].

The images below illustrate some of the the physical differences between the F. conicus populations found in the Birdsong and Cravatt. (Click on the images for the appropriate caption).

Streptelasma, like F. conicus, was one of the dominant species in the Cravatt with specimens accounting for approximately half of all the coral specimens found. In the past these specimens were assigned to Duncancella rudis, and then Syringaxon. Although no professional study has been undertaken on the Cravatt specimens a comparison of some thin sections made from both Tennessee and Cravatt specimens and with reference to the original descriptions of Duncanella, Syringaxon and Streptelasma suggest the Cravatt specimens are in fact representatives of Streptelasma. This identification makes the Cravatt rugosans another indicator species for correlating the Oklahoma, Tennessee and New York faunas. [It will take professional study to determine exactly which species of Streptelasma is present in the Cravatt].

Below: Images of thin sections taken from Cravatt specimens with comparison to other genera. The presence of strongly developed intersepta that do not fuse to the major septa and the absence of an open central aulos are features that separate Streptelasma from Duncanella and Syringaxon.

Pleurodictyum is represented by a single species in the Cravatt - P. lenticulare - whereas in the Birdsong Shale two species are present - P. trifoliatum and P. lenticulare. P. trifoliatum is considered to be endemic to the Birdsong (descended from P. brownsportensis, which is found in the underlying Brownsport Formation of the late Silurian) and unlike P. lenticulare does not extend its geographical range into the Lower Devonian of New York. Because of its wide ranging geographical distribution P. lenticulare, like F. conicus, is an important index fossil for both correlation and dating (being restricted to the Lower Devonian).

P. trifoliatum was at one time reported from the Cravatt, but the specimens in question were actually juvenile colonies in which only the first three corallites had developed. Enough specimens have now been collected by the author that show the growth sequence of the corallites. After settling on the substrate, the polyp formed a single corallite. From this initial cell two other corallites budded to form a distinct trio that are encountered periodically as fossils. After the first three corallites had developed, additional ones were added one or two at a time until a total of 6 (sometimes 5  or 7) formed a ring around the original (now central) corallite. The images below show this ontogeny and variations in the number of outer ring corallites. 

Brachiopods overwhelming dominate the macrofauna of the Cravatt with ~40 species having been recorded. The warm, clear waters allowed these filter feeding animals to thrive even on an open shelf environment as the effects of wave activity was so diminished at that their depth that it did little more than bring in a plentiful supply of food particles. Sedimentation was low enough that the ocean floor could be covered with numerous species, and even with their inability to move there was little fear of inundation as storm activity mostly happened high above their shells. 

The fossiliferous beds of the Cravatt do periodically contain bedding plains littered with brachiopod shells and other fossil debris indicating at times current activity was strong enough to break up, scatter and sometimes concentrate the remains of dead animals. Rapid inundation from storm activity is also evident by the presence of so many enrolled trilobites and the fact that most brachiopods are still articulated - the shells never having had a chance to come apart.

With so many species being present in the Cravatt it is not possible to describe each one and assess its relevance to correlation, instead a few species are selected for discussion while the rest are dealt with as a whole. A distribution list for the entire brachiopod fauna is included on page 2 as this best presents the data as it is currently understood.

Many Lower Devonian Atrypid brachiopods have been lumped together in the species Atrypa reticularis, which originates  from the Silurian of Gotland and England. However, it is now known that Lower Devonian forms do not belong to this species, although work still has to be done to hammer out the finer details of which species occur where. Amsden & Boucot (1958) did investigate several of the North American  forms when describing the fauna of the Haragan Formation and they assigned all specimens from the deposit (and later the Cravatt) to the new species Atrypa oklahomensis. They stated that the species "is most like the Brownsport-Henryhouse species A. tennesseensis, being similar in outline and profile" (p.120) and noting that differences between the two are difficult to determine with the exception of the Oklahoma specimens having a fold and sulcus. Specimens from the Birdsong Shale, once assigned to A. reticularis, are also now placed in A. oklahomensis and likely evolved from A. tennesseensis.

Whether A. oklahomensis evolved from A. tennesseensis in Oklahoma and then migrated to Tennessee or vice versa is impossible to determine, however the species does not extend into New York and it is one of the species that indicates the western/central realms of the Lower Devonian Epeiric Sea had a brachiopod fauna that differed more greatly to the eastern realm than compared other groups  (e.g. the corals, which show great uniformity across the entire sea). There is little difference between the Tennessee and Oklahoma populations as in both localities the species is common. The Oklahoma specimens do, on average, attain a slightly larger size but it is considered to be simply a product of the more turbulent environment of the Ross Shelf limiting growth rather than a diagnostic feature.

In New York the Atrypid brachiopods are represented by Pseudsoatrypa and Spinatrypa.

Reference :

Amsden, T.W. & Boucot A.J. 1958. Stratigraphy and Palaeontology of the Hunton Group of the Arbuckle Mountains. Oklahoma Geological Survey Bulletin 78.

Below : Specimens of Atrypa oklahomensis from the Birdsong Shale

Originally described by James Hall as Orthis oblata based on specimens from the Helderbergian of New York, this species was redescribed as Rhipidomelloides oblata by Amsden and Boucot (1958) and later as Discomyorthis oblata by Boucot (1975). Common in the Cravatt and the Birdsong shale, this distinct brachiopod had a range that extended across the entire Epeiric Sea. Being readily found in Oklahoma, Tennessee and New York, the species clearly thrived in the environments of the Cravatt, Ross, and New Scotland shelves. Capable of attaining sizes of up to around 4 cm this was one of the larger species of brachiopod that inhabited the Lower Devonian seas.

The species is characterised  by a "typically subcircular or transversely elliptical outline, with a width in nearly all specimens great than length. The length of the hingeline is variable giving a subrounded [long hingline] to... slightly subtriangular [short hingeline] outline. In lateral profile the valves are equally biconvex... the beaks of both valves are small. Ornamentation multicostellate. Shell punctate." (Amsden and Boucot 1958 p. 55.)

Specimens from the Birdsong shale are identical in every respect to those from Oklahoma, with their being practically no differences in average size or relative abundance. Specimens from New York have not been available for comparison, but Amsden and Boucot (1958) reported they are often slightly larger.

Reference :

Amsden, T.W. & Boucot A.J. 1958. Stratigraphy and Palaeontology of the Hunton Group of the Arbuckle Mountains. Oklahoma Geological Survey Bulletin 78.

Boucot, A.J. 1975. Evolution and Extinction Rate Controls 

Below: Discomyorthis oblata from the Cravatt, except far right which is from the Birdsong Shale

Another example of a cosmopolitan species. Originally described from New York by James Hall (as Orthis subcarinata)  Amsden and Boucot (1958) believed the specimens in their collections from Oklahoma represented a subspecies as they were consistently smaller (by ~25%). Specimens from Tennessee measured by this author are more consistent with the Oklahoma fauna and in general can actually be slightly smaller. Why there was a reduction in size moving eastwards across the Epeiric Sea is likely due to the changing palaeoenvironment, with the increasingly turbulent waters directly effecting the maximum growth size of the species. [This is consistent with data discussed previously in the corals where the largest recorded specimens of F. conicus are in New York (which had the deepest, calmest waters), followed by Oklahoma and finally Tennessee].

Levenea subcarinata pumilis is readily identifiable due to the strong fold and sulcus that developed in the brachial and pedicle valves respectively (right) . 

Reference :

Amsden, T.W. & Boucot A.J. 1958. Stratigraphy and Palaeontology of the Hunton Group of the Arbuckle Mountains. Oklahoma Geological Survey Bulletin 78.

Below: Levenea subcarinata pumilis specimens showing the nature of the fold and sulcus of the valves. (Middle specimens is from the Birdsong Shale)

It is not just the larger brachiopod species that can be used in correlation and paleoenvironmental analysis. Atrypina hami is one of the smaller species known from the Epeiric Sea but it is encountered frequently enough to be of use in both these fields. 

Described as Atrypina imbricata by Reeds (1911) from specimens collected in the Hunton Group of Oklahoma with later reassignment to the new species A. hami by Amsden and Boucot (1958) , this species is also found in the Birdsong Shale. In the New Scotland of New York the more finely ornamented A. imbricata is found. Reasonably uncommon, A. hami can, however, be locally abundant (such as at the White Mound locality in the Haragan Formation, and certain outcrops in the Ross Formation of Decatur County, Tennessee) but it is strongly restricted to the Lower Devonian, particularly Helderbergian Rocks. This limited temporal range is important in correlating the faunas of Oklahoma, New York and Tennessee, although the difference in species does indicate evolution of the genus as it migrated across the Epeiric Sea - the thickening  of the costae seen in A. hami  an adaptation required to strengthen the shell in the higher turbity environment of the central and western realm.

Reference :

Amsden, T.W. & Boucot A.J. 1958. Stratigraphy and Palaeontology of the Hunton Group of the Arbuckle Mountains. Oklahoma Geological Survey Bulletin 78.

Reeds , C.A. 1911. The Hunton Formation of Oklahoma. American Journal of Science. Vo. 182. p. 256-268.

Below: Atrypina hami specimens from the Birdsong Shale (left and middle) and the Cravatt Formation (right).

For a continuation of this article please see page II