John E. Repetski1, James D. Loch2, and Raymond L. Ethington3

1U.S. Geological Survey, MS 926A National Center, Reston, VA 20192;
2Earth Sciences Department, Central Missouri State University, Warrensburg, MO 64093; and
3Geological Sciences Department, University of Missouri-Columbia, Columbia, MO 65211

Abstract—Conodonts from exposures in and near the Ozark National Scenic Riverways, southeastern Missouri, clarify the biostratigraphic setting of the Lower Ordovician Roubidoux Formation. The underlying Gasconade Dolomite ranges from at least as low as the Rossodus manitouensis Zone in its lower and middle parts to the "Low Diversity Interval" in its upper member. Conodonts of the "Low Diversity Interval" continue through the lower part of the Roubidoux. Macerodus dianae Zone conodonts appear approximately one-third to halfway through the Roubidoux and the biozone is represented through most or all of the remainder of the formation. Faunas characteristic of the Acodus deltatus - Oneotodus costatus Zone first occur either in the uppermost part of the Roubidoux or in the lowermost part of the overlying Jefferson City Dolomite.


Figure 1—Map showing location of Ozark National Scenic Riverways and the watershed areas of the Current and Jacks Fork Rivers in southeastern Missouri. Triangle indicates location of Jacks Fork crossing section.

The Ozark Plateau region (Figure 1), comprising basically that part of Missouri south of the Missouri River and the part of Arkansas north of the Buffalo and White Rivers, contains one of the largest areas of exposed Lower Ordovician, or Ibexian Series, rocks in the United States. Due to its inboard location on the trailing margin of Laurentia, in tropical to subtropical latitudes during that time of deposition, this area accumulated chiefly shallow-marine carbonate sediments. Most of this section is now dolostone, much of it primary or early diagenetic. However, much of it also is coarsely-grained dolostone, often chert-rich, reflecting one or more periods of subsequent diagenetic episodes, such as the periods of massive diagenesis related to subsurface hydrothermal fluids that moved through the section in the Pennsylvanian and Permian associated with the Ouachita orogeny (e.g., see Leach and Rowan, 1986; Leach and others, 1997).

The section also contains many sandstones; nearly all are quartz sands, and most of them probably are multiply-reworked, as the region was hundreds of miles from the low- lying exposed sources of Precambrian and Cambrian rocks in the upper Midwest. Most of these Ozarks sands are thin lenses, stringers, and channels, with only local lateral extent, but some horizons or intervals contain apparently laterally-persistent sands that have been used for both subdividing the section into formations and for local and long-distance correlations. Numerous hills and pinnacles of Precambrian volcanic rocks of the St. Francois Mountains complex apparently were emergent periodically during the Late Cambrian and Early Ordovician, and these certainly affected the local sediment distribution patterns. However, these Ozarks volcanics did not contribute significant volumes of sand to the Ibexian units. Some of the sand intervals have been incorporated into sequence stratigraphic schemes of various scales, but without rigorous control of their positions in a stratigraphic framework. Our work on conodonts (Figure2) and trilobites in the Ibexian of the Ozarks is only in its early stages, but we are able to make some refinements in the age control in some of the stratigraphic marker horizons.

Very little biostratigraphic control exists for this succession of rock, especially considering its areal extent. The depositional environments for most of the units were probably somewhat hypersaline and with restricted circulation, and were not conducive for development of diverse faunas. In addition, the pervasive secondary dolomitization wiped out much of the original shelly record. Most of the macrofauna known from the Ibexian of the Ozarks is preserved in chert. Presumably, local replacement of the carbonate rocks by secondary chert took place before pervasive dolomitization had destroyed the invertebrates, as must have happened where chert did not form. Because much of the record is from chert float blocks, precise occurrence and stratigraphic range data are seriously lacking. Mollusks are the most abundant and diverse of these macrofossils; trilobites are known from relatively few intervals. Most of these faunas were described in a very few works, in the 1930's to 1950's (e.g., Bridge, 1930; Heller, 1954). This study is part of a cooperative U.S. Geological Survey-National Park Service project of bedrock mapping in the Ozarks of Missouri.

Figure 2—Scanning electron microscope (SEM) photomicrographs of some representative conodont elements from the Gasconade, Roubidoux, and Jefferson City formations in southeastern Missouri. Illustrated specimens are reposited in the type collections of the Paleobiology Department, U.S. National Museum (USNM), Washington, D.C. 20560. A_B, Juanognathus? felicitii (Ji and Barnes); posterolateral views of two specimens from sample RC-149 at Roubidoux Creek section, X 85, USNM 498496 and 498497. C, Drepanodus sp., inner lateral view of drepanodontiform element, from sample RC-149, Roubidoux Creek section, X 85, USNM 498498. D_E, Colaptoconus quadraplicatus (Branson and Mehl); posterolateral (D) and lateral (E) views of triplicatiform and quadraplicatiform elements, respectively; D, from sample JF-H, Jacks Fork section, X 110, USNM 498499; E, from sample JC-J, lower part of Jefferson City Dolomite at Jim's Creek section, X 64, USNM 498500. F, cf. Colaptoconus quadraplicatus (Branson and Mehl); shallowly grooved specimen, from sample JF-J at Jacks Fork section, X 90, USNM 498501. G, Ulrichodina deflexa Furnish; posterolateral view of immature(?) specimen from lower part of Jefferson City Dolomite at Jim's Creek section, sample JC-J, X 110, USNM 498502. H, Paroistodus? sp.; inner lateral view of scandodontiform element, same sample and location as G, X 110, USNM 498503. I_L, Histiodella donnae Repetski; posterior views of blade-like elements (I, K, L) and inner lateral view of coniform element (J), from samples JF-H (K) and JF-J (I, K, L) at Jacks Fork section, I and J -X 90, K -X 170, L -X 110, USNM 498504_507. M. Laurentoscandodus? n. sp.; inner posterolateral view of short-based element, from sample JF-J at Jacks Fork section, X 70, USNM 498508. N, Drepanoistodus sp.; inner lateral view of drepanodontiform element, from sample JF-J at Jacks Fork section, X 90, USNM 498509. O, Juanognathus? n. sp.; posterior view of nearly symmetrical element, from sample JF-J at Jacks Fork section, X 90, USNM 498510. P_Q, Striatodontus? prolificus Ji and Barnes; posterolateral views, P from sample JF-J at Jacks Fork section, X 90; Q from upper part of Gasconade Dolomite (sample RC-02) at Roubidoux Creek section, X 95, USNM 498511 and 498512. R_S, Oneotodus aff. O. simplex (Furnish); posterior and lateral views of two specimens from lower part of upper Gasconade Dol. at Phillips Quarry, Bartlett 7-1/2 minute quadrangle, Shannon Co., MO, X 65, USNM 498513 and 498514. T, Chosonodina herfurthi Müller; posterior view of specimen from upper part of middle Gasconade Dol. at Phillips Quarry, X 75, USNM 498515. U, Rossodus manitouensis Repetski and Ethington; inner lateral view of coniform element; same sample as T, X 50, USNM 498516. V, Loxodus bransoni Furnish; inner lateral view; specimen broke during preparation; from upper part of middle Gasconade Dol. at a section near Rolla, MO, X 45, USNM 498517. W, Acanthodus uncinatus Furnish; lateral view of non-serrate suberectiform element, from same sample as T, X 50, USNM 498518. X, Variabiloconus bassleri (Furnish); inner lateral view, X 60, same sample as T, USNM 498519. Y, Oneotodus simplex (Furnish); lateral view, X 75, USNM 498520. Z, Scolopodus sulcatus Furnish; inner lateral view of scandodontiform element, X 75, same sample as T, USNM 498521.

The Ozark National Scenic Riverways is a National Park Service unit along the Current and Jacks Fork Rivers (Figure 1). Large tracts in the Ozarks are part of the Mark Twain National Forest. Fort Leonard Wood, a major U.S. Army reservation, is nearby. Associated with these public lands, and the region as a whole, are a variety of land-use issues for which accurate geologic maps are vital. This is a world-famous karst region, with vast cave systems, thousands of sinkholes and springs, and all the hydrogeological problems associated with those regions. This region also is host to the world's largest known lead deposits, hosted mainly in Upper Cambrian carbonate rocks. Lead sulfide exploration is continuing, and accurate knowledge of the geologic framework is needed, especially as it pertains to the public lands tracts, for the permitting process.

From a few preliminary studies (Kurtz, 1981) conodonts were known to occur, at least sparsely, in some of the formations of the Ozarks. We therefore began a systematic examination of these dolostones, involving sampling of key mea
sured sections for conodonts as well as other fossils, to try to establish a refined biostratigraphic framework that can be applied to the mapping and other studies in these units and to better correlate the Ozark succession with other regions.

The Ibexian section in the south-central Missouri Ozarks is up to more than a thousand feet thick. The units are, in ascending order: the Eminence Dolomite, only the upper part of which is Ibexian; the Gasconade Dolomite, about 300 feet thick; the Roubidoux Formation, 100 to 300 feet thick; Jefferson City Dolomite, 125 to 350 feet thick; and the Cotter/Powell formations, about 100 to more than 300 feet thick (Thompson, 1991). Because of mapping needs, we began our biostratigraphic work with the Roubidoux Formation, including its position relative to the underlying Gasconade Dolomite and to the lower part of the overlying Jefferson City Dolomite.

We located and measured several sections of the Roubidoux, including the type section, and concentrated on sections which contain formational boundaries and other marker horizons. For this work we collect and process samples of 4- to 6-kilograms from the estimated best lithologies, and concentrate on bracketing these boundaries and markers. The large samples have proven necessary, as our productive samples average only a few conodont elements per kg. A fine sieve size also is necessary; we use a 200-mesh bottom sieve. In most samples there are few accessory minerals in the heavy residues; following magnetic separation, the heavy residue is almost exclusively snowy-white diagenetic dolomite rhombs. The color alteration index (CAI) of the conodonts is 1 to 1-1/2, indicating only low levels of long-term post-burial heating, in the range of less than 50° to about 90° C. Preservation of the conodont elements tends to be quite good.

As with artifacts and living plants and animals, the fossils of this and all other National Parks can be collected only with formal permission from the appropriate Park Superintendent.

Gasconade Dolomite

Previously, Kurtz (1981) showed, from a few samples near Camdenton, MO, some tens of miles north-northwest of our study area, that the lower part of the Gasconade Dolomite is lower Ibexian, rather than Upper Cambrian as earlier studies and maps had assumed. Our initial sampling from the middle Gasconade near Rolla, MO (Repetski and others, 1993; Repetski and others, in press) yielded species typical of the Rossodus manitouensis Zone, including the diagnostic species Loxodus bransoni (Figure 2:V) Furnish, Scolopodus sulcatus (Figure 2:Z), and R. manitouensis itself (Figure 2:U). Subsequently we have sampled several additional sections that expose the upper part of the Gasconade.

Somewhat surprising was finding that the uppermost Gasconade has the fauna of the so-called "Low Diversity Interval" of Ethington and Clark (1981). This interval, characterized by presence of only a few coniform taxa of rather simple morphologies, mostly assignable to species of Oneotodus, Teridontus, Striatodontus?, and Eucharodus (e.g., Figure 2:P-S), follows the abrupt disappearance of most of the taxa of the taxi of the subjacent R. manitouensis Zone. This abrupt faunal changeover was documented by Ethington, Engel, and Elliott (1987) to occur almost precisely at the Mackenzie Hill - Cool Creek formation contact in Oklahoma, and similarly at the House - Fillmore formation boundary in Utah. Recently, Ji and Barnes (1993) also discussed this same turnover within the Boat Harbour Formation of the St. George Group of western Newfoundland, Canada. What surprised us is that this faunal changeover in the Ozarks does not coincide with the Gasconade - Roubidoux contact, which is recognized at a regionally significant influx of sand, but it occurs somewhat lower, at a level near or at the boundary between the cherty middle part and the non-cherty upper part of the Gasconade.

The rather diverse fauna that is typical for the R. manitouensis Zone occurs, in low numbers of specimens per kg of rock, beginning low in the Gasconade (Kurtz, 1981). It contains mostly shallow-water Laurentian species such as Loxodus bransoni, Scolopodus sulcatus, and Variabiloconus bassleri (Figure 2:V,Z,X). Uncommonly, cosmopolitan species such as Chosonodina herfurthi (Figure 2:T) also occur in this part of the Gasconade.

The uppermost member of the Gasconade contains mostly morphologically "simple" coniform elements of species similar to, and presumed related to, Oneotodus simplex (Figure 2:Y), which occurs in the underlying R. manitouensis Zone. The most significant appearance in this interval is that of Striatodontus? prolificus Ji and Barnes (Figure 2:P,Q), which continues upward as one of the numerically dominant species of the Roubidoux Formation.

Roubidoux Formation

Figure 3—Top surface of a large chert float block from middle part of the Roubidoux Formation, preserving molds of the gastropod Lecanospira. This moldic mode of preservation is typical for mollusks found in the Roubidoux. Block collected in steep slope immediately east of Pike Creek; SW 1/4 of section 22, T. 27N., R. 3W., Low Wassie 7-1/2 minute quadrangle, Shannon Co., MO; USGS fossil locality number 11527-CO; USNM 498495. Scale bar is 2 inches long.

The Roubidoux Formation is separated from the Gasconade by its higher content of quartz sand, including numerous sandstone beds especially in the upper half, and generally thinner bedding (see Thompson, 1991, for history of nomenclature and usage). The quartz sand content also serves to distinguish the Roubidoux from the overlying Jefferson City Dolomite. Partly because of the increased permeability due to the sandstones and partly due to the thin- to medium-bedded nature of its carbonate beds, the Roubidoux weathers back rapidly and the resulting slopes are characterized by loose blocks of sandstone. Chert float blocks from the middle part of the formation are common in some areas; sometimes they preserve the molds of gastropods (Figure 3), nautiloid cephalopods, and, rarely, of trilobites. Significant exposed thicknesses of the Roubidoux are extremely rare, leading to difficulties in studying the actual succession of rocks and fossils within the formation, and thereby hindering attempts to biostratigraphically subdivide the unit.

The designated type section of the Roubidoux (Heller, 1954) is a cut-bank cliff along Roubidoux Creek, Texas County, MO. It affords reasonably good exposure of the entire formation, including both lower and upper contacts. Conodont samples from the Roubidoux type section confirm again that the Low Diversity Interval begins well below the base of the Roubidoux (Figure 4). We can document here that the base of the next biozone, the Macerodus dianae Zone, falls in the lower part of the middle Roubidoux, indicated by the appearance of Histiodella donnae (Figure 2I_L) and Ulrichodina n. sp. 1 of Repetski (1982). The few samples near the top were not diagnostic of a zonal call near the base of the Jefferson City.

Figure 4 also shows the distribution of conodonts and some trilobites that we collected from Heller's (1954) section at Ava, Douglas Co., approximately 50 miles west of the National Park. This section does not expose either base or top of the Roubidoux, but it preserves the interval from the Low Diversity Interval to the M. dianae Zone and it allows some calibration of the conodont and trilobite ranges for this region.

Jacks Fork Crossing Section

We sampled another section that exposes most of the Roubidoux and more than 50 feet of the upper Gasconade, where Highway 17 crosses the Jacks Fork River, within the boundary of the Ozark National Scenic Riverways. From the level of the river below the bridge to approximately the level of the highway at the north end of the bridge, the thick-bedded dolomite of the uppermost part of the Gasconade forms a cliffy exposure. The Roubidoux drops back in profile because of weathering of the sandy dolostone beds. The sandstone beds tend to stand out and they become more prominent higher in the section. This section was described in detail by Muilenberg and Beveridge (1954), and their section description is repeated in Thompson (1991)

Figure 4—Conodont and trilobite faunal distribution charts for the Ava and Roubidoux Creek sections.

Figure 5.—Conodont faunal distribution chart for the section at Jacks Fork crossing section.

Figure 6—Conodont zonation for the Ibexian Series (Lower Ordovician) of the North American Midcontinent faunal realm, following Ross and others (1997), and showing the biostratigraphic range of conodonts recovered from the Roubidoux Formation in southeastern Missouri.

Figure 5 shows the conodont distribution for the Jacks Fork crossing section. A little more than 100 feet of the Roubidoux is exposed here; we estimate that about 50 feet is covered above. All of the exposed Gasconade is in the "Low Diversity Interval." Histiodella donnae (Figure 2:I-L), and probably also Ulrichodina n. sp. 1 of Repetski (1982), appearing in the middle Roubidoux marks the Macerodus dianae Zone. It is clear now that the widespread influx of sand marking the base of the Roubidoux does not coincide with any notable change in the conodont fauna. However, the thicker and more prominent sandstone beds that begin in the middle part of the Roubidoux do grossly coincide with the base of the M. dianae Zone within the current resolution of our faunal control.

Lithologic and biostratigraphic relations at the boundary between the Roubidoux and overlying Jefferson City Dolomite are difficult to assess because that contact is so rarely exposed. For mapping purposes, that contact is usually drawn at 25 to 35 feet below the base of the "Quarry Ledge" of the Jefferson City, a 12- to 15-foot thick widespread marker bed that is more often exposed than the beds beneath it, due to its characteristic thick to massive bedding.

The Quarry Ledge also is significant because its top and immediately overlying beds have produced most of the trilobites known from the lower Jefferson City. We sampled a short section in the lower part of the Jefferson City Dolomite near Vichy, MO, about 40 miles north of the National Park, that yielded both trilobites and conodonts. This interval not only yielded a number of trilobite taxa, essentially marking the Jeffersonian Stage assemblage, but the conodonts allow recognition of the Acodus deltatus - Oneotodus costatus Biozone at this level, from the appearance of Ulrichodina deflexa (Figure 2:G) and Eucharodus toomeyi.

Another of our Jefferson City sections, Jim's Creek, yields conodont data that suggest that the boundary between the Macerodus dianae Zone and the Acodus deltatus-Oneotodus costatus Zone may well fall near or at the base of the "Quarry Ledge."


Even though much more work remains to be done in refining the Ibexian framework for the Ozark region, the work thus far shows that the conodonts, especially, are both present and biostratigraphically useful in these strata. They are already allowing a firming up of some of the formational and marker bed age constraints. Figure 6 represents our latest age assignment for the Roubidoux Formation against the current Ibexian conodont biozonal framework for the North American Midcontinent faunal realm (Ross et al., 1997).

On a broader scale, some recent sequence stratigraphic interpretations within this interval have been made assuming age-equivalency for certain levels essentially based only on physical stratigraphic grounds. For example, some workers
have correlated the coeval stratigraphic break at the House-Fillmore contact in Utah and the McKenzie Hill-Cool Creek in Oklahoma with the base of the Rockdale Run Formation and Nittany Dolomite in the central Appalachians (Goldhammer and others, 1993). Conodonts show that the Stonehenge-Rockdale Run and Stonehenge-Nittany contacts in Maryland and central Pennsylvania, respectively, occur within the Rossodus manitouensis Zone, and thus are demonstrably older than either the base of the Roubidoux, base of the Fillmore, or base of the Cool Creek. Likewise, the fossils demonstrate that the influx of sand at the base of the Roubidoux is younger than the House-Fillmore and McKenzie Hill-Cool Creek contacts. Biostratigraphy can be a powerful ally to sequence stratigraphers; certainly it should not be ignored.

Ongoing and planned paleontological work in the Ozark National Scenic Riverways includes: 1) continued biostratigraphic support for the geologic mapping efforts; 2) refinement of the placement of the individual biozonal bases within the Ibexian formations relative to formation contacts and other marker beds or intervals; 3) testing of the lateral continuity, i.e., the reliability to mapping, of these marker beds; and 4) documentation of the fossil successions through the other formations in this region.


We thank our colleagues of the National Park Service, Ozark National Scenic Riverways, for their continued cooperation and interest in all phases of this project, the numerous landowners who kindly gave us access to their properties, and R.C. McDowell and D.J. Weary, for helpful reviews of this manuscript. Ellis L. Yochelson, USGS-Emeritus, confirmed the identification of Lecanospira on the illustrated chert block, and D.J. Weary, USGS, helped with the graphics.


Bridge, J. 1930. Geology of the Eminence and Cardareva quadrangles: Missouri Bureau of Geology and Mines, ser. 2, v. 24, 228 p.

Ethington R. L., and D. L. Clark. 1981. Lower Ordovician conodonts in North America, in Sweet, W.C., and Bergstrom, S.M. (eds.), Symposium on conodont biostratigraphy: Geological Society of America, Memoir 127, p. 63-82.

———, K. M. Engel, and K. L. Elliott. 1987. An abrupt change in conodont faunas in the Lower Ordovician of the Midcontinent Province, in, Aldridge, R.J. (ed.), Palaeobiology of conodonts: Ellis Horwood, Ltd., Chichester, p. 111-127.

Goldhammer, R.K., P. J. Lehmann, and P. A. Dunn. 1993. The origin of high-frequency platform carbonate cycles and third-order sequences (Lower Ordovician El Paso Gp, West Texas): constraints from outcrop data and stratigraphic modeling: Journal of Sedimentary Petrology, v. 63, no. 3, p. 318-359.

Heller, R.L. 1954. Stratigraphy and paleontology of the Roubidoux formation of Missouri: Missouri Geological Survey and Water Resources, 2nd Series, v. 35, 118 p.

Ji, Z. and C. R. Barnes. 1993. A major conodont extinction event during the Early Ordovician within the Midcontinent Realm: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 104, p. 37-47.

Kurtz, V. E. 1981. The Cambrian-Ordovician boundary in Missouri as determined by conodonts, in, Taylor, M.E. (ed.), Short pa

repetski et al.—ozar, ordovician conodonts
pers for the Second International Symposium on the Cambrian System: U.S. Geological Survey, Open-File Report 81-743, p. 115-117.

Leach, D. L., and E. L. Rowan. 1986. Genetic link between Ouachita foldbelt tectonism and the Mississippi Valley-type lead zinc deposits of the Ozarks: Geology, v. 19, p. 931-935.

———, L. E. Apodaca, J. E. Repetski, J. W. Powell, and E. L. Rowan. 1997. Evidence for hot Mississippi Valley-type brines in the Reelfoot rift complex, south-central United States, in late Pennsylvanian - early Permian: U.S. Geological Survey, Professional Paper 1577, 36 p.

Muilenberg, G. A., and T. R. Beveridge. 1954. Guidebook, 17th regional field conference, Kanss Geological Society, southeastern and south-central Missouri: Missouri Geological Survey and Water Resources Report of Investigations 17, 63 p.

Repetski, J. E. 1982. Conodonts from El Paso Group (Lower Ordovician) of westernmost Texas and southern New Mexico: New Mexico Bureau of Mines and Mineral Resources, Memoir 40, 121 p.

———, R. L. Ethington, W. M. Furnish, and D. J. Kennedy. 1993. Conodonts from the Oneota and Gasconade Dolomites (Lower
Ordovician) of the central midcontinent, U.S.A. [abs.]: Geological Society of America, Abstracts with Programs, North-Central Section, v. 25(3), p. 74, 75.

———, J. D. Loch, R. L. Ethington, and R. I. Dresbach. In press. A preliminary re-evaluation of the stratigraphy of the Roubidoux Formation of Missouri and correlative Lower Ordovician units in the southern midcontinent, U.S.A.: Oklahoma Geological Survey, Circular.

Ross, R.J., Jr., L. F. Hintze, R. L. Ethington, J. F. Miller, M. E. Taylor, and J. E. Repetski. 1997. The Ibexian, lowermost Series in the North American Ordovician, with a section on Echinoderm biostratigraphy, by J. Sprinkle and T.E., Guensberg, 1997, In Taylor, M.E. (ed.), Early Paleozoic biochronology of the Great Basin, western United States: U.S. Geological Survey, Professional Paper 1579, p. 1-50, + 1 oversize plate.

Thompson, T. L. 1991. Paleozoic succession in Missouri, Part 2, Ordovician System: Missouri Department of Natural Resources, Report of Investigations No. 70, 282 p.