Late Viséan ( late Mississippian ) ammonoids from the Barnett Shale , Sierra Diablo Escarpment , Culberson County , Texas , USA

Mid-Carboniferous strata of the Barnett Shale in the Sierra Diablo region are deep water, offshore sediments deposited in the Marathon Foreland Basin. These strata contain a remarkably complete ammonoid record spanning from the Late Viséan to the middle Atokan (Moscovian). Late Viséan strata are referred to as the “Folks Member” of the Barnett Shale and locally yield numerous ammonoids. Three assemblages can be recognized, which contain, from oldest to youngest, Goniatites eganensis and Girtyoceras hamiltonense (Goniatites eganensis Zone), Goniatites multiliratus and Girtyoceras meslerianum (Goniatites multiliratus Zone), and Choctawites cumminsi and Pachylyroceras cloudi (Choctawites cumminsi Zone). We erect the new genera Choctawites and Uralyroceras to accommodate, respectively, the North American species “Goniatites choctawensis Shumard, 1863”, “G. kentuckiensis Miller, 1889” and “G. cumminsi Hyatt, 1893”, and Uralian species formerly attributed to Pachylyroceras. For the material of “Pachylyroceras cloudi” of the South Urals, the new species name Uralyroceras arquatum is proposed. ZooBank: http://zoobank.org/References/6C6A1411F88F-45C2-BA4A-D97C4CD4B415

Although mentioned repeatedly in the literature and tabulated in a 50-year-old faunal list (King, 1965), the Late Viséan ammonoids of the Barnett Shale (Helms Formation of most older papers) in west Texas have never been previously described.Intrigued by this repeated mention in the where the only significant Mississippian ammonoid bearing strata crop out (Fig. 1).It quickly became obvious that the Figure 2 Ranch section preserves one of the most important ammonoid-bearing late Mississippian to early Pennsylvanian sections in North America, and possibly the world.Subsequently to A. L. Titus' work on the Figure 2 Ranch, L. L. Lambert initiated field studies on the Figure 2 Mississippian section with J. E. Harrell in 2003 and2004, measuring sections and making extensive collections of both Mississippian and Pennsylvanian ammonoids.
Unfortunately, since the Figure 2 Ranch was sold in 2005, researchers have no longer had access.Harrell (2007) presented an overview of the ammonoids they had recovered as part of the stratigraphic revision they were recommending for the Carboniferous of the region.This paper describes for the first time the complete Late Viséan ammonoid succession of the Barnett Shale at the Figure 2 Ranch, needed now more than ever since there will be no access to this locality in the foreseeable future.

Previous work
Although Pennsylvanian (Atokan) ammonoids have been extensively described and figured from the Sierra Diablo (Plummer and Scott, 1937;Miller andFurnish, 1940b, 1958), no systematic analysis of Viséan ammonoids from this region has ever been published.Prior to this work, only two Viséan specimens had been figured from the area; a "Girtyoceras meslerianum" (referred to by us as G. hamiltonense) by McCaleb et al. (1964, pl. 2, figs. 8, 9) and a specimen of Pachylyroceras cloudi in Furnish and Saunders (1971, pl. 2, fig. 5).The next year, King (1965) listed eight Viséan species from the Barnett Shale; "Goniatites crenistria Phillips", "Goniatites choctawensis Shumard", "Goniatites sp.", "Girtyoceras sp.cf.G. meslerianum (Girty)", "Neoglyphioceras cloudi (Miller and Youngquist)", "Neoglyphioceras sp.", and "Lyrogoniatites sp.cf.L. newsomi (Smith)" based on collections made by J. B. Knight, A. K. Miller, and W. M. Furnish in the 1930s and 1940s.We have not examined the collections that the faunal list was based upon, but it is clear they are stratigraphically mixed; "Goniatites crenistria" and "Girtyoceras sp.cf.G. meslerianum" are from our lowest recognized ammonoid zone (Goniatites eganensis Biozone) and the remainder of material from our newly established Choctawites cumminsi Zone.Because King's collections were actually tied to a published measured section, we consider this the only other significant report of Viséan ammonoids from the Sierra Diablo region.In contrast, there are several mentions that Mississippian ammonoids occur in the Sierra Diablo within faunal descriptions of material from other localities (Miller and Furnish, 1940a;Miller and Youngquist, 1948;Youngquist, 1949).

Geological overview
The Carboniferous stratigraphy of the Sierra Diablo was recently summarized by Harrell (2007).Virtually every major stratigraphic unit from the Viséan to the Bashkirian has yielded ammonoids in this region (Harrell, 2007), with the Viséan to late Serpukhovian section being particularly fossiliferous and rich in ammonoids.All of the Viséan material in this report came from the "Folks Member" of the Mississippian-Pennsylvanian Barnett Shale, which is exposed in only three places in the northeast portion of the Sierra Diablo below an imposing escarpment formed out of the Permian Hueco Formation (King, 1965), entirely within the Figure 2 Ranch property.Although generally homoclinal and south-dipping, the Folks Member is locally faulted, folded, and intruded by small dikes and sills which caused local contact metamorphism.The entire pre-Permian section was folded, faulted, and eroded into a peneplain during the Ouachita Orogeny before being covered by the sediments of the Permian Paw Paw Formation, which along with the Hueco Formation is the source for most of the erosional debris that covers the non-resistant Folks Member.The Folks Member is comprised mostly of mudstone and siltstone, with prominent carbonate concretions, phosphatic pebble-rich lags, and thin-bedded pelagic micritic limestone that all point towards a largely sediment-starved depositional regime during the Late Viséan and early Serpukhovian (Ti- tus, 1999).This character is generally similar to that of timeequivalent strata of the Barnett Shale exposed in the Llano Uplift, which is even more condensed than the Sierra Diablo section.The Folks Member is abundantly fossiliferous and yields a largely pelagic fauna, but also a dwarf, low diversity benthic assemblage of gastropods and bivalves in the phosphate pebble lag horizons.Well-preserved, uncrushed ammonoids occur through much of the unit, preserved either as testiferous moulds in carbonate concretions or beds, or as phosphatic steinkerns in the pebble lag rich shales.Although there are three sections now known to yield Viséan ammonoids in the northern Sierra Diablo (Harrell, 2007) only one of these, herein referred to as King's section, had been studied and sampled extensively enough to allow for detailed analysis when all access to the Figure 2 Ranch to researchers was ended.Fortunately the other two sections do not appear to preserve any fauna or significant horizons not seen at King's section.

King's section
Our King's section (Fig. 2) is the same as Section 11 published by King (1965) and from which all of the Mississippian ammonoids in his Barnett Shale faunal list came from.The beds at King's section dip northeast about 15  W).The covered interval just below our bed 1 has scattered pieces of large ex situ ammonoid-bearing concretions.These concretion fragments were derived from our beds 9, 11, and 13.The only ammonoids we collected from these loose pieces were specimens of Goniatites eganensis Korn and Titus, 2011.The lowest 7 m of exposed section did not yield in situ threedimensional ammonoids, only rare leiorhynchid and lingulid brachiopods and Nereites-type trace fossils.
Seven metres above the base of the section are three separate horizons of 1-1.5 m diameter concretions less than a half metre apart from each other stratigraphically.The lowest of these three concretionary horizons (bed 9) yielded an abundant ammonoid fauna dominated by the crenistriate species G. eganensis, but also containing Girtyoceras hamiltonense Korn and Titus, 2011 (samples 00TXCU-21a and 00TXCU-21b).The next concretion horizon (bed 11) yielded uncommon specimens of G. eganensis (sample 00TXUT-22), while the highest concretionary horizon (bed 13) in this immediate grouping did not yield any diagnostic material.
Approximately 2 m higher is another single band of elliptical 0.3 m diameter concretions (bed 15) that we obtained no diagnostic specimens from.Above the bed 15 concretions is 2.5 m of non-concretionary shale/siltstone, which ends at beds 17-19, two closely spaced concretionary horizons separated by 40 cm of siltstone.The concretions of bed 17 yielded an assemblage of Goniatites multiliratus Gordon, 1962 and Girtyoceras meslerianum (Girty, 1909) (sample 00TXCU-25).Although we cannot make out the suture lines on the Goniatites specimens from this horizon, we are confident of its assignment because its ornament is highly diagnostic given other biostratigraphic constraints on the sample, which include the co-occurring Girtyoceras.The upper concretionary horizon (bed 19), which is highly septarian, did not yield diagnostic material.
The next 4 m of section is a highly condensed, fossiliferous succession of glauconitic nodular phosphatic mudstone punctuated by a thin oxidized silty/sandy bed (bed 21a) and three limonitic iron bands (beds 22,24,26).The phosphatic intervals are sediment-starved hardgrounds, some of which exhibit extensive recycling.Phosphatic-glauconitic steinkerns of Choctawites cumminsi (Hyatt, 1893) and Pachylyroceras cloudi (Miller and Youngquist, 1948) are common (sample 00TXCU-27), and occur with an impoverished benthic molluscan fauna that was not collected.The highest iron band hardground (bed 26) is overlain by 1.3 m of phosphatic shale, which yielded no fossils.
Two metres higher is a couplet of concretionary horizons bearing striate goniatitids that look virtually identical to beds 17 and 19, but in reverse stratigraphic order, with the septarian horizon (bed 35) being the lower of the two.Although it was not detected in 2000 when the section was measured, the ammonoid evidence, coupled with the repetition of the distinctive yellow siltstone horizon and concretionary beds, seems to suggest there is a small-scale structural repetition of this portion of the section.
The highest ammonoids recovered from King's section are from bed 37, which probably due to structural repetition, contains the same assemblage found in bed 17.Because we did not observe an obvious structural break in the exposed section, there is a small chance that beds 35-37 are in correct stratigraphic position and therefore Serpukhovian in age.If this is the case, then the lirate goniatitids probably belong to Lusitanoceras, which are known from lower Serpukhovian strata of the Antler and Marathon-Ouachita foreland basin systems (e.g.Drahovzal and Quinn, 1972).Unfortunately, without sutural evidence we cannot assign specimens from bed 37 to a genus with certainty and our bias, given the lithologic repetition and lack of associated cravenoceratids or glaphyritids, is the structural hypothesis.Beds 43-45 are not seen in any of the other Sierra Diablo sections and the remaining 22 m of section up to the angular unconformity with the overlying Paw Paw Formation is clearly from shallower water facies.Aside from some sparse crinozoan debris, this upper interval lacks obvious macrofossils.

Biostratigraphic interpretation
Late Viséan.-The ammonoids Goniatites eganensis and Girtyoceras hamiltonense were discovered from the base of the measured section, as well as other localities to the south, indicating that the Goniatites eganensis Biozone, which was first delineated in the Chainman Shale of Utah and Nevada (Korn and Titus, 2011), is the lowest in the Sierra Diablo Barnett section that can be diagnosed with in situ material (Fig. 2).Specimens of Goniatites multiliratus from bed 17 indicate that the zone for that species can also be diagnosed in that general part of the section.This same general succession can be observed entirely, or in part, in the Chainman Shale of Nevada and Utah, the Caney Shale of Oklahoma, and Moorefield Formation of Arkansas (Korn and Titus, 2011).

M I S S I S S I P P I A N
nosus" (now referred to as Lusitanoceras Pereira de Sousa, 1923) Biozone (Gordon, 1965;Saunders et al., 1977;Webster et al., 1984).It is now clear to us that both the taxonomy and biostratigraphy of this interval in North America is much more complicated and can be subdivided in the eastern USA using species of the characteristic genus Choctawites.As Ch. cumminsi is particularly common in Texas, we use it to name a new ammonoid biozone, whose reference locality is the Chappel Hill section of the Barnett Shale, San Saba, Texas.It is defined as the interval between the first occurrence of the nominate taxon and the first appearance of Cravenoceras.Associated ammonoids in Texas include Sulcogirtyoceras ornatissimum, Pachylyroceras sp., Lusitanites sp., and various dimorphoceratids.At present, no unequivocal younger Late Viséan ammonoid zones can be defined in North America (Fig. 3), although the position of Lyrogoniatites georgiensis would presumably fall into this interval.
Although not discernible in the Sierra Diablo section, we also erect the Choctawites kentuckiensis Biozone with the Buffalo Wallow reference section (Moorefield Shale; east of Batesville, Arkansas; Malinky and Mapes, 1982) for the interval between the first appearance of that taxon and the appearance of Choctawites choctawensis.Key North American first occurrences of ammonoids in this zone include Sulcogirtyoceras (as S. limatum), Lusitanoceras, Neoglyphioceras, Lusitanites, and Ruddellites (Gordon, 1965;Malinky and Mapes, 1982).The interval between the Ch.kentuckiensis and Ch.cumminsi biozones contains Choctaw-ites choctawensis, Neoglyphioceras caneyanum, Lusitanites sp., Sulcogirtyoceras cf. S. ornatissimum, and dimorphoceratids.Based on this assemblage, we establish the Choctawites choctawensis Biozone with Girty's station 2078 as the reference section (Caney Shale; Pittsburg County, Oklahoma; Branson et al., 1959), defined as the interval between that taxon's first appearance (Fig. 3) and the appearance of Choctawites cumminsi.
Serpukhovian.-Although not the focus of this paper, Serpukhovian ammonoids are abundant in the Sierra Diablo sections and it is clear that a succession of multiple zones could be recognized at least as high as the Reticuloceras Genus zone (Harrell, 2007).Earliest Serpukhovian ammonoids are discussed here to give the Late Viséan succession a better context.However, no high-resolution zonation for lower Serpukhovian ammonoids equivalent to the Pendleian Stage of the British Isles currently exists for North America and establishing one is beyond the scope of this work.The oldest Serpukhovian ammonoids collected from the Sierra Diablo (Tumulites, Paracravenoceras, Fayettevillea, Beyrichoceratoides, Eosyngastrioceras) are from beds 28-29 and are almost identical to those found in the middle and upper Fayetteville Shale of northern Arkansas.Collectively, they are diagnosed as the "Tumulites varians-Cravenoceras fayettevillae" zone of Saunders et al. (1977), probably equivalent to the British E 1 b.As might be expected in a deeper-water offshore section, earliest Serpukhovian assemblages (equaling British E 1 a) characterized elsewhere in North America by A. L. Titus et al.: Late Viséan (late Mississippian) ammonoids from the Barnett Shale a distinctive Girtyoceras-Cravenoceras sensu stricto assemblage are not preserved as three-dimensional material in the Sierra Diablo area.

Global correlations
High-resolution global correlations for the Late Viséan (Asbian and Brigantian in the British subdivision scheme) ammonoids of North America were proposed by Korn and Titus (2011).Our work on the Sierra Diablo sections has not substantively changed this earlier interpretation and no further discussion on correlation of pre-Choctawites kentuckiensis Biozone assemblages is warranted.
Choctawites cumminsi Biozone -Because the genus Choctawites is endemic to North America, direct correlations are difficult.However, Sulcogirtyoceras associated with Choctawites cumminsi at San Saba (S. ornatissimum), Texas, and Choctawites choctawensis (S. cf. S. ornatissimum) in the Caney Shale of Oklahoma are very similar to the Rhenish Mountain species S. burhennei, which co-occurs with Paraglyphioceras rotundum.In both Europe and North America, the earliest occurrence of Neoglyphioceras is in the immediately underlying assemblage.We think the similar pattern of occurrence for these two distinctive morphotypes forms a better basis for correlation than the species or genus ranges of various goniatitid taxa such as Lusitanoceras and Dombarites, which appear to occur diachronously in various basins.As a result, we equate the occurrence of Neoglyphioceras in the Choctawites kentuckiensis assemblage of the Moorefield Formation with the Neoglyphioceras spirale zone of western Europe (Fig. 3).The overlying Choctawites choctawensis and Choctawites cumminsi zones thus correlate with the Paraglyphioceras rotundum and Neoglyphioceras suerlandense-Lusitanoceras postriatum intervals of the Rhenish Mountains sections (Fig. 3).We believe this correlation is the most parsimonious solution as it synchronizes the first appearance of S. burhennei-type Sulcogirtyoceras, Neoglyphioceras, and advanced forms of more openly umbilicate neoglyphioceratids like N. suerlandense (very similar to the more evolute North American genus Pachylyroceras).Unfortunately Choctawites is very rare in the US cordilleran region.Specimens tentatively referable to Ch. cumminsi are known from the Chainman Shale in west Utah, occurring in the upper portion of the range of Lusitanoceras, associated with Lusitanoceras, Lusitanites, Sulcogirtyoceras ornatissimum, and Pachylyroceras utahense.

Systematic descriptions (D. Korn, A. L. Titus)
Descriptive terminology for conch morphology is patterned after Korn (2010).Abbreviations of conch dimensions (Fig. 4) are conch diameter (dm), whorl width (ww), whorl height (wh), umbilical width (uw = dm 1 − wh 1 − wh 2 ) and aperture height (ah).The whorl expansion rate (WER) is cal- Order Goniatitida Hyatt, 1884 Suborder Goniatitina Hyatt, 1884 Superfamily Girtyocerataceae Wedekind, 1918 Family Girtyoceratidae Wedekind, 1918 Family diagnosis: Girtyocerataceae with ontogenetically changing conch shape; inner whorls evolute, umbilicus closing during ontogeny, adult stage often discoidal and oxyconic.Ornament with biconvex growth lines.Suture line with V-shaped external lobe and rounded ventrolateral saddle; adventive lobe V-shaped with slightly sinuous flanks.Remark: For the genus composition of the family and a discussion of the characteristics and limits, see Korn and Titus (2011).Remark: For the species composition of the genus and a discussion of the characteristics and limits, see Korn and Titus (2011).
Girtyoceras hamiltonense Korn and Titus, 2011 (Fig. 5)  (2011).Description: The largest specimen, nearly 22 mm conch diameter (Table 1), is almost fully ornamented and has a thinly discoidal, subinvolute conch shape (ww / dm = 0.42; uw / dm = 0.18).At the largest diameter, it possesses an angular umbilical margin, subparallel flanks, and a narrowly rounded venter.The fine growth lines are strongly biconvex with a prominent ventrolateral projection and a deep ventral sinus.The last volution displays four shell constrictions extending parallel to the growth lines.They originate just outside the umbilical margin on the inner flank area; they are deepest on the mid-flank but continue across the venter (Fig. 5).
The next smaller specimen (NPL 68362) has a conch diameter of nearly 15 mm with similar conch proportions (ww / dm = 0.45; uw / dm = 0.25).It is a partly testiferous internal mould with predepositional breakage to the living chamber.There are four constrictions, which are, as in all specimens, visible externally.The fine rib density is about 12 per millimetre.
The smallest specimen (UTSA 07062) is approximately 10 mm in diameter, thickly discoidal, and broken obliquely, but nearly in half, with the other half missing.The umbilical shoulders are angular and about 90 • .The umbilical width index is entirely within a narrow range.Ribs and constrictions are moderately sinuous, and just start to develop a ventral sinus at about 8 mm diameter.There are three constrictions.
Discussion: G. hamiltonense is the stratigraphically oldest Girtyoceras species currently known from the Sierra Diablo section.It resembles G. gordoni Korn andTitus, 2011, G. welleri Gordon, 1965, andG. meslerianum (Girty, 1909).However, G. gordoni has less sinuous ribs and constrictions at comparable diameters, while G. welleri has a slightly narrower umbilicus between 6 and 20 mm conch diameter and a more narrowly rounded ventral profile at diameters between 6 and 15 mm.G. hamiltonense is now known from Nevada and Utah (Chainman Shale) and the Sierra Diablo (bed 9).The near time-equivalent species G. welleri occurs in the lower portion of Moorefield Formation near Batesville, Arkansas, and the Caney Shale of the Arbuckle Mountains region.In every case, these eastern occurrences are the oldest known horizons with a Goniatites assemblage at their respective localities.
The better preserved of the two larger specimens (NPL 68364) is nearly 10 mm in diameter (Fig. 6).By 8 mm diameter the umbilical margin forms a strong right angle and the ribs and constrictions (which number four per whorl, are moderately deep and visible on both the internal moulds and external shell) are very strongly deflected anteriorly.The ribs and constrictions also form a distinct ventral sulcus at this stage.Faint longitudinal lines can be seen in the ventral area.
The smallest specimen (NPL 68365) is transitioning from the widely umbilicate acutely marginated form to the almost parallel-sided thickly discoidal form with a 90 • umbilical shoulder.The flanks and venter are gently rounded.Fine thread-like riblets originate from the umbilical margin and deflect anteriorly at the ventrolateral margin.The ribs are straight across the venter.Discussion: Girtyoceras meslerianum is a rare taxon in the Sierra Diablo sections mostly because well-preserved threedimensional material is difficult to obtain from the concretions that bear its assemblage.Although interpreted rather broadly in the past, in our view, G. meslerianum is a morphological intermediate between G. hamiltonense and the stratigraphically younger taxon Sulcogirtyoceras limatum.It bears the more compressed conch form, strongly developed anterior deflection of the ribs and constrictions on the ventrolateral shoulder, and wider umbilicus of the latter (uw / dm = 0.31), but has the weaker ornament and lack of ventrolateral sulcus of the former.Thus G. meslerianum might be confused with G. hamiltonense and small specimens (smaller than 12 mm) of S. limatum.However, G. meslerianum can be readily differentiated from G. hamiltonense because the latter has a consistently narrower umbilicus (uw / dm = 0.25 at 16 mm diameter).Early members of the genus Sulcogirtyoceras develop a prominent ventrolateral sulcus at approximately 12-15 mm diameter, but also have slightly stronger ribs and constrictions at comparable stages (less than 20 mm diameter).Smaller specimens of more advanced species of Sulcogirtyoceras have more depressed whorls.
Species that are stratigraphically older than G. hamiltonense in North America have less sinuous constrictions and narrower umbilici at comparable diameters (e.g.G. gordoni), while younger Serpukhovian species (e.g.Girtyoceras jasperense Gordon) are more narrowly umbilicate and more robust (higher-value ww / wh ratios).Girtyoceras meslerianum appears to be limited to the Goniatites multiliratus Biozone in both the mid-continent USA and the western Cordillera (Chainman Shale).It is rather uncommon outside  (Girty, 1909); based on material from Jackfork Creek, Oklahoma (Fig. 7).(Gordon, 1965).
Superfamily Goniatitaceae de Haan, 1825 Family Goniatitidae de Haan, 1825 For the subfamily composition of the family and a discussion of the characteristics and limits, see Korn and Ebbighausen (2008) and Korn et al. (2010).
Subfamily Goniatitinae de Haan, 1825 For the genus composition of the subfamily and a discussion of the characteristics and limits, see Korn and Titus (2011).
Diagnosis: Goniatitinae with subinvolute inner whorls; the umbilicus becomes closed in early ontogeny.External lobe usually V-shaped and rarely Y-shaped, very narrow or narrow, usually with slightly curved flanks.

Goniatites eganensis
Diagnosis: Goniatites with a globular conch between 2 and 8 mm diameter (ww / dm 0.85-0.90)and thickly pachyconic conch (ww / dm 0.75-0.80)at 20 mm diameter.Umbilicus moderate in early ontogeny (uw / dm = 0.20-0.30at 2 mm diameter) and very narrow in all stages larger than 4 mm diameter (uw / dm 0.02-0.10).Umbilical wall convexly rounded in all stages.Aperture low or moderately high (WER = 1.70-1.80 at 20 mm diameter).Ornament with crenulated, slightly biconvex, and rectiradiate growth lines; external sinus shallow.Suture line at 25 mm conch diameter with narrow external lobe (0.55 of the external lobe depth; 1.25 of the ventrolateral saddle width), and moderately low median saddle (0.40 of the external lobe depth).External lobe V-shaped with narrow, V-shaped prongs; ventrolateral saddle subangular.Description: Four specimens are selected to outline the morphology of the material from the Sierra Diablo.Unfortunately, inner whorls are largely recrystallized and thus do not contain solid information about the conch ontogeny.
Specimen NPL 68431 is a fully chambered specimen of 47 mm conch diameter; it is representative for a larger growth stage (Fig. 8a).The pachyconic conch (ww / dm = 0.67; uw / dm = 0.06) is very narrowly umbilicate and possesses a broadly rounded, wide venter and a steep umbilical wall.Parts of the specimen are covered with shell remains.These show fine and crenulated growth lines, which extend with weakly biconvex course across flanks and venter, forming equally high dorsolateral and ventrolateral projections and shallow lateral and ventral sinuses.
The smaller specimen NPL 68471 (38 mm dm) shows very similar conch proportions but differs in the slightly coarser, irregularly spaced growth lines (Fig. 8b).Specimens NPL 68493 (27 mm dm) and NPL 68389 (18.4 mm dm) are thickly pachyconic conchs with stronger depressed whorl cross section (Fig. 8c, d).The larger of the two possesses wellpreserved shell ornament with biconvex, strongly crenulated growth lines.These form a shallow lateral sinus and a very shallow ventral sinus.The smaller of the two possesses weakly crenulated growth lines standing in wide distances.Holotype: Specimen USNM 119499; illustrated by Gordon (1965, pl. 18, figs. 9-12).Type locality and horizon: USGS locality 6619A, 2 km south of Buckhorn (Murray County, Oklahoma); Caney Shale, Goniatites multiliratus Biozone.Material: Three specimens (NPL 68494 through NPL 68496 from sample 00TXCU-25).Diagnosis: Goniatites with a globular conch between 2 and 8 mm diameter (ww / dm 0.85-1.10)and thickly pachyconic to globular conch (ww / dm 0.80-0.85)at 20 mm diameter.Umbilicus moderately wide in early ontogeny (uw / dm = 0.30-0.35at 2 mm diameter) and very narrow in all stages larger than 4 mm diameter (uw / dm 0.02-0.10).Umbilical wall convexly rounded in all stages.Aperture low or moderately high (WER = 1.60-1.80 at 20 mm diameter).Ornament with more than 200 spiral lines and crenulated, slightly biwww.foss-rec.net/18/81/2015/Foss.Rec., 18, 81-104, 2015 Progoniatites pilus convex, and rectiradiate growth lines; external sinus shallow.Suture line at 25 mm conch diameter with narrow external lobe (0.55 of the external lobe depth; 1.00 of the ventrolateral saddle width), and moderately low median saddle (0.45 of the external lobe depth).External lobe Y-shaped with very narrow, V-shaped prongs; ventrolateral saddle subacute.Description: The largest specimen (NPL 68494) has a partly crushed body chamber, but the phragmocone appears to be uncrushed (Fig. 9).The conch is, at 30 mm diameter, thinly pachyconic (ww / dm = 0.70) with a very narrow umbilicus.Shell fragments are visible at some places.They show fine spiral lines, which in accompany with crenulated growth lines form a spider web-like pattern.The steinkern is smooth except for wide, shallow constrictions.The smallest specimen (NPL 68495) is partly crushed and has a diameter of 15 mm.Faint traces of spiral ornament show that this feature has developed by this diameter.Like both of the other specimens, the phragmocone is extensively recrystallized.The medium sized specimen was difficult to extract and the only useful data that can be obtained is that it shows strong spiral lines, with a density of 13 per 2 mm, at an estimated diameter of 25 mm.Discussion: This species can easily be confused with the stratigraphically older species G. deceptus, which occurs in Nevada and Utah sections of Chainman Shale.The main differences between the two species are the generally wider conch in G. multiliratus (ww / dm = 0.80-0.85 at 20 mm diameter in G. multiliratus in contrast to 0.65-0.70 in G. deceptus) and the wider external lobe with higher median saddle in G. multiliratus.
G. multiliratus is the stratigraphically youngest species of Goniatites known from North America, and certain features of its conch ontogeny (prolonged subevolute juvenile stage) and suture line (relatively high median saddle and angular flanks of the adventive lobe) may foreshadow the evolution of a Choctawites-like form, although there would clearly be other intermediate stages.Family Delepinoceratidae Ruzhencev, 1957 Family diagnosis: Representatives of the superfamily Goniatitaceae with advanced suture lines possessing a Y-shaped external lobe and a tectiform ventrolateral saddle.The adventive lobe and also the prongs of the external lobe show the tendency to become tridentate.Discussion: There exist three diverging concepts about the classification of the advanced goniatitid species and genera.Ruzhencev andBogoslovskaya (1971, 1978) separated the families Delepinoceratidae (including Platygoniatites and Delepinoceras) and Agathiceratidae (including Dombarites, Pericleites, Proshumardites, and Agathiceras) on the basis of the subdivision mode of the external lobe, thus delineating two independent evolutionary lineages.Leonova (2002Leonova ( , 2011) ) doubted the assignment of Agathiceras in the order Goniatitida and transferred it into the order Tornoceratida (which has usually only been accepted as a suborder within the Goniatitida).Finally, Kullmann et al. (2007) and Kullmann (2009a, b) separated Agathiceras and closely related forms in the independent superfamily Agathiceratoidea within the suborder Goniatitina.In Kullmann's concept, the genera Dombarites and Proshumardites are grouped in the subfamily Dombaritinae within the family Delepinoceratidae.
If Agathiceras is excluded from the discussion of the phylogeny of the family Goniatitidae and its descendants (for instance, if the central or sub-central position of the siphuncle in the juvenile stage is been taken as a reason for separation on higher taxonomic levels), then the phylogenetic analysis (see below) speaks for the extension of the family Delepinoceratidae to also include the subfamily Dombaritinae (Kullmann, 2009a).This subfamily, however, also requires re-definition to include the genera Revilloceras, Lusitanoceras, and Choctawites.
For resolving the phylogenetic relationships of the goniatitid ammonoid genera, we performed a cladistic analysis based on the character matrix of Korn (1997a).In contrast to that analysis, we used Progoniatites uncus Korn, Bockwinkel and Ebbighausen, 2010 as the outgroup species and included three more ammonoid species in the new analysis, Choctawites choctawensis, Revilloceras granofalcatum, and A. L. Titus et al.: Late Viséan (late Mississippian) ammonoids from the Barnett Shale Dombarites falcatoides.We included a new character, the presence of shell constrictions, in the matrix.
The heuristic cladistics analysis performed with PAST (Hammer et al., 2009) led to 35 most parsimonious trees of 35 steps length, of which a majority consensus tree was formed (Fig. 10).This tree shows some unresolved branches but demonstrates that a crown group is well established with the genera Choctawites, Lusitanoceras, Revilloceras, Dombarites, Proshumardites, Platygoniatites, and Delepinoceras, which are all characterized by the apomorphies of a Y-shaped external lobe and a tendency towards a tectiform ventrolateral saddle.This clearly defined clade can best be described as the family Delepinoceratidae.One has, however, to take into account that a similar tendency of sutural development also occurs in the more conservative lineage leading to Neogoniatites.
The cladogram also shows that Choctawites is more plesiomorphic than the other of the genera, which are all defined by the presence of outer shell constrictions in the juvenile stage.The crown group is subdivided into two independent clades, one represented by the advanced Dombarites species D. tectus and Proshumardites and the other represented by the genera Platygoniatites and Delepinoceras.
Subfamily Dombaritinae (in Kullmann et al., 2007) Subfamily diagnosis: Delepinoceratidae with advanced, moderately wide external lobes (usually between 0.60 and 0.90, measured at half height, of the external lobe depth), and median saddles with a height of 0.50 to 0.75 of the external lobe depth.External lobe Y-shaped and undivided.Included genera: Choctawites n. gen.Lusitanoceras Pereira de Sousa, 1923Revilloceras Wagner-Gentis, 1980Dombarites Librovitch, 1957Proshumardites Rauser-Tschernoussowa, 1928Deleshumardites Kullmann, 2007 Discussion: Distinctive Late Viséan to early Serpukhovian goniatitids with advanced lobes (especially with broader, more deeply divided external lobe) have an almost cosmopolitan distribution (Librovitch, 1957;Gordon, 1965;Ruzhencev andBogoslovskaya, 1970, 1971;Drahovzal and Quinn, 1972;Wagner-Gentis, 1980;Webster et al., 1984;Korn et al., 1999;Nikolaeva and Konovalova, 2005;Klug et al., 2006).Until recently, it was thought that this morphological complex formed an essentially monophyletic clade that dispersed during the Late Viséan, evolving from the advanced goniatitid Lusitanoceras (Korn, 1988(Korn, , 1997b;;Kullmann, 2009a).Cross sections of various specimens belonging to Lusitanoceras were shown by Kullmann and Pitz (1980), Kusina (1987), Korn (1988Korn ( , 1997b)), Kusina and Yatskov (1999) and Nikolaeva and Konovalova (2005); all sections possess a rather long subevolute juvenile stage.Nikolaeva and Konovalova (2005) figured cross sections of a series of specimens from the Urals that were all referred to as the genus Dombarites by Ruzhencev and Bogoslovskaya (1971) and concluded that the shapes of the early whorls in this are very different to those of the genus Lusitanoceras.Consequently, Nikolaeva and Konovalova (2005) postulated that Dombarites sensu stricto did not evolve from Lusitanoceras, but directly from the genus Goniatites.Indeed, the figured cross sections of Dombarites with subinvolute inner whorls suggest that the umbilical width in the inner whorls are more similar to Goniatites; however, one has to keep in mind that there is no reason to assume that evolutionary pathways always proceed in the same trend (e.g. from involute to evolute inner whorls); Dombarites, with its less widely umbilicate inner whorls may also have been derived from more evolute forms.Furthermore, the suture lines of all of the genera assembled in the subfamily Dombaritinae are very similar in great detail, speaking for very close relationships and obviously ruling out homoplasy.
A more detailed subdivision of the various forms often assembled in the genus Dombarites requires the separation of genera on the base of morphology of the inner whorls, ornament and suture line details (Table 6).For this reason we describe the new genus Choctawites to accommodate the North American species "Goniatites choctawensis Shumard, 1863", "G. kentuckiensis Miller, 1889" and "G. cumminsi Hyatt, 1893" and revive the genus Revilloceras Wagner-Gentis, 1980 for superficially similar species known from the Cantabrian Mountains of Spain and the Anti-Atlas of Morocco.

Choctawites n. gen.
Type species: Goniatites choctawensis Shumard, 1863.Genus diagnosis: Dombaritinae with subevolute to evolute inner whorls; the umbilicus becomes closed in early ontogeny.Whorl profile with parallel or subparallel flanks, conch in juveniles widest at some distance from the umbilicus.Externally expressed constrictions extremely rare or absent.Growth lines slightly biconvex and rectiradiate; external sinus shallow.External lobe Y-shaped, very narrow or narrow, usually with slightly curved flanks.Included species: choctawensis: Goniatites choctawensis Shumard, 1863, p. 109;Oklahoma. cumminsi: Glyphioceras cumminsi Hyatt, 1893, p. 467;Texas. kentuckiensis: Goniatites kentuckiensis Miller, 1889, p. 439;Kentucky.Discussion: All Late Viséan-early Serpukhovian goniatitid ammonoids with prominent adult spiral ornament found in North America can be readily divided into two major groups (Gordon, 1965), (1) those with external shell constrictions and ( 2) those in which constrictions are developed as internal shell thickenings and thus only visible on the internal mould.This subdivision largely parallels the whorl cross sections, of which the first group possesses more or less trapezoidal early whorl profiles (in which the conch is widest at some distance from the umbilicus), and the second group shows "normal" whorl profiles with flanks converging from the umbilical wall towards the venter.
Examinations of collections from the Moorefield Formation, Caney Shale, and Barnett Shale show that the two morphological groups have very little morphological or ornamental overlap.Trapezoidal forms are consistently crenistriate or cross lirate to about 15 mm diameter.In contrast, the normal forms develop strong spiral ornament already by 10 mm conch diameter and can exhibit sub-triangular coiling.Furthermore, at any given locality or horizon in the eastern USA, either trapezoidal or rounded forms can totally dominate a sample.
Three possible hypotheses could explain this dimorphic pattern: (1) broad intraspecific variation; (2) sexual dimorphism, and (3) two different taxonomic groups are present in North America.Given the lack of morphological interme-diates between trapezoidal and normal forms, we reject the first hypothesis.Regarding the second hypothesis, the stratigraphically lower normal forms associated with Sulcogirtyoceras limatum (Miller and Faber, 1892) are readily referable to Lusitanoceras because they possess the characteristic early ontogeny, sutural shape, and ornament (particularly with granulated spiral lines) of that genus.In fact, these forms are so close to the European species of Lusitanoceras that they have been referred to as "Goniatites granosus Portlock, 1843" since the early 1960s (Gordon, 1965).While we doubt the species assignment, the referral of the North American material to the genus Lusitanoceras is correct.If the trapezoidal forms are simply sexual dimorphs of Lusitanoceras, it would be expected that this same dichotomy would exist in the European and North American occurrences.Instead, the trapezoidal forms are found almost exclusively in Marathon-Ouachita-Appalachian foreland basin system, being unknown from Europe and extremely rare in the Antler Foreland Basin.This pushes us to accept that the trapezoidal forms are a distinct taxonomic entity endemic to Laurentia (North America).Since they cannot be accommodated in any existing genus, we erect the new genus Choctawites for them.
Lectotype: Specimen UT 12031 (designated by Cloud and Barnes, 1948); illustrated by Gordon (1965, pl. 19, figs. 1-6).Type locality and horizon: 5 miles west of Lampasas (Lampasas County, Texas); Barnett Shale.Diagnosis: Choctawites with a thickly pachyconic to globular conch between 2 and 12 mm diameter (ww / dm = 0.85-0.95) and thickly pachyconic conch (ww / dm = 0.70) at 20 mm diameter.Umbilicus moderately narrow to moderately wide in early ontogeny (uw / dm = 0.25-0.35at 1.5-2 mm diameter) and very narrow in all stages larger than 4 mm diameter (uw / dm 0.05-0.08).Umbilical wall convexly rounded in all stages, very short early juvenile stage with slightly trapezoidal whorl cross section.Aperture low (WER = 1.65-1.70throughout ontogeny).Ornament with about 120 spiral lines and crenulated, slightly biconvex, and rectiradiate growth lines; external sinus shallow.Falcate ornament begins at 40 mm dm.Material: 43 specimens (NPL 68497 through NPL 68539), all from sample 00TXCU-27.The material consists of brown to black coloured phosphatic steinkerns, which have varying amounts of shell preserved on them.Shell occasionally exhibits iridescent colour in small areas.Most of the inner phragmocones (diameter less than 10 mm) were not filled with minerals and weathered out rapidly or were not preserved, leaving a hollow space in the specimen.Some of these small interior phragmocones are exceptionally well preserved when freshly exposed and show the fine details of the septa and shell.The largest intact specimens were about 20 mm in diameter.Only fragments of larger specimens were collected.Description: The conch cross section of the small specimen NPL 68499 (10 mm conch diameter) shows a globular shape (ww / dm = 0.86) and a very narrow umbilicus (uw / dm = 0.06) with broadly rounded flanks and venter (Fig. 12e).The inner whorls show crescent-shaped whorl profiles and a moderately wide umbilicus; the highest uw / dm ratio with 0.34 occurs at 1.2 mm conch diameter.In the growth interval between 5 and 10 mm conch diameter, the conch is widest at some distance from the umbilicus.
The conch cross section from the Sierra Diablo specimen is very similar to four cross sections (specimens 26-13, 26-2, 26b, 26-10) produced by R. Kant (Tübingen) of material from San Saba (original specimens identified by M. Gordon in the USNM; acetate peels in the collections of the GPI Tübingen).The San Saba topotypes only differ in the slightly more pronounced trapezoidal whorl profile and the more subparallel flanks (Fig. 12a-d).
Conchs are thickly pachyconic at 20 mm (e.g.specimen NPL 68497; Fig. 11a), with a very narrow umbilicus and moderately depressed whorls (ww / dm = 0.74; ww / wh = 1.36).Flanks and venter are continuously rounded and the aperture is low (WER = 1.64).The specimen is an internal mould without shell remains; the body chamber has nearly the length of a volution.The internal mould shows shallow constrictions with irregular distances; they extend linearly across the flanks and turn forward for a low and wide ventral projection.
The slightly smaller specimen NPL 68497 (17 mm dm) largely resembles the previous specimen.The internal mould has four prominent constrictions, which probably follow the growth lines in their course, are almost rectiradiate, bearing only a slight forward inflection or are salient over the ventral area (Fig. 11b).
The smallest specimens (e.g.specimen NPL 68501) are 7 mm in diameter.By that stage, the conch is starting to change from sub-trapezoidal (strongly recurved flanks) to rounded whorl profiles.As a result, the trapezoidal phase, which is only weakly developed in this taxon, was only seen on cross-sectioned conchs.Specimen NPL 68500 shows the suture line of a specimen in the intermediate growth stage (at 14.4 mm dm).It possesses a Y-shaped external lobe with sinuous flanks.The external lobe has, measured at half depth, 0.64 of the external lobe depth and is 1.25 times wider than the ventrolateral saddle.The median saddle reaches a height of 0.42 of the external lobe depth.The ventrolateral saddle is tectiform and subacute (Fig. 12f).
Ornament is only commonly preserved on the larger whorl fragments, but smaller specimens show crenistriate ornament up to diameters of approximately 15 mm, after which it changes to strong longitudinal lirae numbering about 19-20 per 5 mm.Discussion: Forms we refer to Choctawites have historically been attributed to three species; "Goniatites choctawensis Shumard, 1863", "Goniatites cumminsi Hyatt, 1893", and "Goniatites kentuckiensis Miller, 1889". Gordon (1965) regarded G. cumminsi as a junior synonym of G. choctawensis, a view supported in nearly every subsequent paper.Recently, Work and Mason (2009a) synonymized G. kentuckiensis with G. choctawensis and followed Ruzhencev and Bogoslovskaya (1971) in assigning G. choctawensis to Dombarites, which left G. choctawensis as the only valid species in this group.After reviewing the literature and numerous specimens, we have concluded that all three established species of the "trapezoidal" group (G.kentuckiensis, G. choctawensis, and G. cumminsi) are valid and represent a stratigraphic and possibly evolutionary succession.
The stratigraphically oldest species is Choctawites kentuckiensis, which is associated with Sulcogirtyoceras lima-  (Hyatt, 1893); mainly based on specimens from San Saba (Elias Coll; cross sections by R. Kant, Tübingen).Ch. kentuckiensis are similar, but the former has a much stronger expression of the trapezoidal shape.These two criteria can be used to readily differentiate between the two species.There is an apparent facies difference in the occurrence pattern of the two species, with Choctawites kentuckiensis being known from shallower water deposits includ-ing a carbonate platform and pro-deltaic mudstone facies in the Appalachian and Ouachita foreland basins (Gordon, 1965;Work and Mason, 2009a), and Ch.choctawensis occurring in the deeper water sequences of the Caney and Barnett shales.The exclusion of the latter from shallow water facies may have resulted from the sea level fall recorded by the Batesville Sandstone wedge in Arkansas.
The stratigraphically next species we can recognize in succession is the type species, Ch. choctawensis, which has the unique combination of early appearance of falcations and a strongly developed trapezoidal phase (Fig. 14a-c).Ch. choctawensis is associated with a more advanced species of Sulcogirtyoceras, S.. cf. S. ornatissimum.It is known widely from the Caney Shale of Oklahoma, e.g.specimen USNM 119505 from USGS locality 2078A (Gordon, 1965, fig. 44G, H) and specimen Kant71 (Elias Collection) from 6 miles SE of Ada along Highway 99 (Pontotoc County, Oklahoma), the Barnett Shale of central Texas (specimen Kant26-1), and tentatively, the Batesville Sandstone of Arkansas (Gordon, 1965).
The stratigraphically youngest species, Ch. cumminsi, is a form whose distinct, but reduced, trapezoidal phase (Fig. 12a-e), more narrowly umbilicate early whorls, and early appearance of falcations are highly definitive.The species can still be differentiated from the other species of Choctawites using the brief or even absent trapezoidal stage.Ch. cumminsi is known mostly from the Barnett Shale of Texas, but almost certainly occurs in the Caney Shale of Oklahoma and Arkansas (e.g.specimen USNM 119502; Gordon 1965, fig. 44e, f).This species is associated with Sulcogirytoceras ornatissimum (Miller and Youngquist, 1948) and species of Pachylyroceras, which together comprise one of the youngest Viséan ammonoid assemblages we can recognize in the eastern USA.Since the Sierra Diablo specimens all exhibit the characteristic reduced trapezoidal stage, we refer to them as Ch.cumminsi.
Superfamily Neoglyphiocerataceae Plummer and Scott, 1937 Family Neoglyphioceratidae Plummer andScott, 1937 Pachylyroceras Ruzhencev andBogoslovskaya, 1971 Type species: Lyrogoniatites cloudi Miller and Youngquist, 1948 [original designation].Genus diagnosis: Representatives of the family Neoglyphioceratidae with subevolute conch, which becomes more evolute in the adult stage.20-30 coarse spiral lines, growth lines, and shell constrictions concavo-convex with shallow ventral sinus.Suture line with rather narrow external lobe, usually with slightly sinuous flanks.Included species: cloudi: Lyrogoniatites cloudi Miller and Youngquist, 1948, p. 660;Texas. utahense: Lyrogoniatites utahensis Miller, Youngquist and Nielsen, 1952, p. 154;Utah.Discussion: When they introduced the genus Pachylyroceras, Ruzhencev and Bogoslovskaya (1971) assumed that certain Uralian specimens and the types for Lyrogoniatites cloudi from the Barnett Shale were conspecific.As a result, they based Pachylyroceras on the types of L. cloudi from Texas, rather than the Uralian material they had in their collections and erected three new species, P. angustum, P. consequens, and P. constrictum.We see significant differences in the course of growth lines in all of the Uralian specimens (ventral salient) versus the Texas types of Pachylyroceras (ventral sulcus), and conclude that the genus, which is closely related to Neoglyphioceras, is currently known only from North America.
This excludes all of the Uralian material from Pachylyroceras, and given that there is no existing genus to accommodate these widely umbilicate Neoglyphioceras-like forms with ventral salient in their growth lines and constrictions, we erect the new genus Uralyroceras for the Uralian species (see below).
According to its conch morphology, suture line, and ornament, Pachylyroceras has an intermediate position between Neoglyphioceras and Lyrogoniatites.It possesses the type of shell constrictions with ventral sinus of Neoglyphioceras but the subevolute conch morphology of Lyrogoniatites.In the suture line, Pachylyroceras shows the sinuous flanks of the external lobe like Neoglyphioceras, but at the same time, the external lobe is rather narrow with weakly diverging flanks like in Lyrogoniatites.(Miller and Youngquist, 1948) (Figs.15a, b and 16c)
The suture line of specimen NPL 68542 (phragmocone whorl width of 13.2 mm, whorl height of 8.0 mm) shows a Vshaped external lobe with sinuous flanks and slightly asymmetric prongs.The median saddle has a height of one-third of the external lobe depth.On the flanks follow a bell-shaped ventrolateral saddle and a symmetric adventive lobe with slightly convex flanks (Fig. 16c).Discussion: The Sierra Diablo material of Pachylyroceras appears to represent only a single species, with all specimens being very close to each other in ornament and morphology.
They are also close, in their conch width/diameter and umbilical width ratios, to several of the paratypes in the type lot from San Saba.As has been pointed out by Gordon (1965), the holotype is atypical for the type material in that it possesses five fewer spiral lines and a much wider umbilicus, rather approaching the morphology of "Neoglyphiceras hyatti Gordon, 1960" than the other types.However, the holotype is the largest specimen in the type lot by about 4 mm, and these differences may be the result of later ontogenetic changes or strong variation rather than of taxonomic value.Unfortunately, because the type lot was collected from ex situ nodules and has no detailed stratigraphic context, it is possible that the type lot is from a different horizon.Occurrence: Ruzhencev and Bogoslovskaya (1971) reported the species from the South Urals, but we consider this assignment erroneous because the conch shape, suture, and ornament of the Uralian forms are fundamentally different from the type of P. cloudi.We propose the new name Uralyroceras arquatum (named after the arched course of the constrictions) for the Uralian species (see below).Pachylyroceras cloudi appears to be characteristic of the Choctawites cumminsi Biozone of the eastern USA and equivalent age strata in Nevada and Utah.

Conclusions
The Mississippian to Pennsylvanian section in the Sierra Diablo is remarkable in that it preserves in one place a Late Viséan through Moscovian ammonoid record that rivals any other known from North America.The strata were deposited on the north side of the Marathon Foredeep and represent deeper water offshore facies rich in pelagic fauna.For the pre-Pennsylvanian portion of this record, three-dimensional material appears to be limited to facies representing relative low stands in sea level (e.g.Titus, 1999), indicating that sea level was especially low during the Viséan Goniatites eganensis, Goniatites multiliratus, and Choctawites cumminsi biozones, as well as the succeeding Serpukhovian Tumulites varians Biozone.Key events for correlation of North Amer-ican and western European ammonoid faunas appear to be the first appearance of Neoglyphioceras and advanced Sulcogirtyoceras bearing angular juvenile whorl edges (e.g. S. burhennei).The Sierra Diablo ammonoid assemblages are low diversity in both the Goniatites and Choctawites intervals, with only Goniatites and Girtyoceras known from the former and Choctawites and Pachylyroceras from the latter.However, it must be pointed out that the sample sizes are relatively small.The collections described herein along with historic collections made over 50 years ago will be the only ones available for the foreseeable future even though more field work is required at this scientifically significant locality to establish the exact stratigraphic relationships of the Serpukhovian and Bashkirian ammonoid assemblages.The new genus Choctawites, which dominates the youngest Sierra Diablo Viséan assemblage, appears to be largely confined to the Appalachian-Ouachita-Marathon foreland basin system and is only rarely found in timeequivalent strata (Chainman Shale) of Nevada and Utah.

AFigure 1 .
Figure 1.Reference map showing location of the Figure 2 Ranch King section, Sierra Diablo, Culberson County, Texas.

Figure 2 .
Figure 2. Columnar King's section on the Figure 2 Ranch, Sierra Diablo, Texas, with the position of fossil samples and proposed biostratigraphic attribution.

Figure 4 .
Figure 4. Descriptive terms for the conch morphology and suture lines of the ammonoids described here.

Figure 10 .
Figure 10.Cladogram of selected genera, represented by well-known species, of the superfamily Goniatitaceae.

FFigure 12 .
Figure 12.Choctawites cumminsi (Hyatt, 1893) from San Saba (a-d) and the Figure 2 Ranch section (e, f).(a) Cross section of specimen 26-13 from San Saba; × 2.5.(b) Cross section of specimen 26-2 from San Saba; × 2.5.(c) Cross section of specimen 26b from San Saba; × 2.5.(d) Cross section of specimen 26-10 from San Saba; × 2.5.(e) Cross section of specimen NPL 68499 from bed 25 (sample 00TXCU27) of Figure 2 Ranch; × 2.5.(f) Suture line of specimen NPL 68500 from bed 25 (sample 00TXCU27) of Figure 2 Ranch, at 14.4 mm diameter, 11.3 mm ww, 7.2 mm wh; × 5.0.(g-i),Ontogenetic development of the conch width index (ww / dm), umbilical width index (uw / dm), and WER of all available specimens (the cross sections of specimens (a-d) from San Saba were produced by R. Kant; the original specimens are stored in the collections of the USNM, acetate peels are stored in the collections of the GPI Tübingen).

Figure 13 .CFigure 14 .
Figure13.Choctawites kentuckiensis(Miller, 1889).(a) Cross section of specimen 57a from Moorefield, Arkansas; × 2.5.(b) Cross section of specimen MB.C.25470 from Moorefield, Arkansas; × 2.5.(c) Suture line of specimen MB.C.25471 from Moorefield, Arkansas, at 14.1 mm diameter, 10.2 mm ww, 6.6 mm wh; × 5.0.(d-f)Ontogenetic development of the conch width index (ww / dm), umbilical width index (uw / dm), and WER of all available specimens (cross section (a) was produced by R. Kant; the original specimen is stored in the collections of the University of Iowa, acetate peels are stored in the collections of the GPI Tübingen).
3482258N (31.47504 • N, 104.86844 • .Exposures through most of the approximate 46 m of Barnett Shale section here are excellent.The start of our measured section is near the slope break, in a drainage located at UTM coordinates (NAD27): 13R 0512497E; Global correlation chart for upper Viséan and lowest Serpukhovian ammonoid zones; after
ah ww / dm ww / wh uw / dm WER coidal or pachyconic with subangular or rounded umbilical margin.The intermediate stage has rounded flanks and venter and the adult stage is lenticular and oxyconic.Without or with very faint ventrolateral grooves.Suture line with Vshaped external lobe, moderately high median saddle, narrowly rounded ventrolateral saddle, and V-shaped adventive lobe.

Table 4 .
Korn and Titus, 2011)millimetres) and proportions for reference specimens of Goniatites eganensisKorn and Titus, 2011from the Sierra Diablo.area, with the only other definitive occurrences we know of at San Saba, Texas in the base of the Barnett Formation and in the Chainman Shale at the classic Duckwater locality (not mentioned byKorn and Titus, 2011).It should occur in the Goniatites multiliratus Biozone of the upper Pool Creek Member of the Moorefield Formation, but specimens from this interval are all crushed

Table 6 .
Characteristics of genera within the subfamily Dombaritinae.

Table 8 .
(Hyatt, 1893)ons and ratios for Choctawites cumminsi(Hyatt, 1893)from the type area and the Sierra Diablo.

Table 9 .
Conch dimensions and ratios for Pachylyroceras cloudi from the type area and the Sierra Diablo.
B C A Figure 16.Suture lines of representatives of Pachylyroceras from localities in Utah, Nevada, and Texas; all × 6.0.(a) Pachylyroceras utahensis