- PDF LinkChapter PDF
F. Jerry Lucia, 2012. "The Great Lower Ordovician Cavern System", Great American Carbonate Bank: The Geology and Economic Resources of the Cambrian—Ordovician Sauk Megasequence of Laurentia, James Derby, Richard Fritz, Susan Longacre, William Morgan, Charles Sternbach
Download citation file:
Karsting and collapse brecciation in the Lower Ordovician carbonates have been recognized for many years. However, the time of cavern formation and the geochemical hydrology responsible is debatable. In this chapter, I intend to review pertinent literature and evaluate the evidence of the presence of paleocaverns, the time of their formation, the history of cavern collapse to form collapse breccias, and the relationship of collapse breccias to structure. I will not review chemical hydrology issues because a discussion on the geochemical environment is pertinent only after the time of cavern development has been adequately resolved. The most robust data sets come from outcrop studies. Outcrops with extensive exposures reviewed here are the El Paso Group in the Franklin Mountains, west Texas; the Pogonip Group in Nopah Range, southeastern California; and the St. George Group in Newfoundland. The Lower Or-dovician outcrops in central Texas, the Mississippi Valley, Virginia, and the Arbuckle Mountains are also useful. Robust subsurface data sets include the Ellenburger Group of Texas; the Knox Group of Tennessee, Kentucky, and Ohio; and the Arbuckle Group of central Kansas. Core descriptions from the subsurface Arbuckle Group in Oklahoma and Arkansas are also helpful.
The most convincing evidence of cavern formation is roof collapse, that is, evidence that breccia blocks are below their stratigraphic positions. The time of cavern formation is more difficult to ascertain and most commonly is based on the source of the infilling sediment by comparing lithologies and, in places, using biostratigraphy. The history of the collapse can be determined only in the most extensive outcrops and mining operations, although modern three-dimensional (3-D) seismic volumes are useful. The relationship of collapse breccias to structure is basically a timing issue and can be resolved only by detailed geologic studies.
I conclude from reviewing published data that convincing evidence shows that an extensive cavern system existed in the Lower Ordovician carbonates at the time of the Sauk-Tippecanoe unconformity. In some areas, the unconformity surface is highly irregular and appears to represent karst terrain. Caverns located far below the unconformity were most likely formed in response to internal disconformities. Lower Ordovician fractures and faults can have a controlling influence on the location and geometry of the caverns. Collapse of these caverns produced the collapse breccia and fracturing of the cavern roof. In some instances, cavern collapse has produced structural sags similar to those produced by the expansion related to strike-slip faulting. In extreme cases, collapse of large caverns produced breccia pipes that extended more than 330 m (>1000 ft) into overlying Ordovician, Silurian, and possibly Devonian units.
Karsting and collapse brecciation in the Lower Ordovician carbonates beneath the Sauk-Tippecanoe unconformity have been observed since as far back as 1948 by Walters and Price. However, the origin of these features is controversial. Issues that are hotly debated include karst versus burial dissolution, solution collapse versus fault brecciation, meteoric versus hydro-thermal fluids, and the general timing of diagenetic events characteristic of these features. Wilson et al. (1992) described the subsurface occurrence of brecciation in the Cambrian-Ordovician of North America and suggested that most of the brecciation is related to solution collapse, and they summarized the many unsolved problems associated with these breccias. In this chapter, I summarize and evaluate pertinent observations bearing on the origin of the brecciation that have been reported in the literature. Only after the history of brecciation has been established can the nature of the fluids involved be logically addressed. The nature of the fluids involved and their hydrology will not be addressed in this chapter.
The scale of observation is a significant consideration in evaluating conclusions drawn from reported studies. The largest scale is found in outcrops of the Basin and Range province of the southwestern United States. In the Franklin Mountains of El Paso, Texas, about 500 m (∼1500 ft) of the Lower Ordovician El Paso Group is exposed. In the Nopah Range of eastern California, about 400 m (∼1300 ft) of the Lower Ordovician Pogonip Group is exposed. In contrast, the Lower Ordovician Prairie du Chien Group in the upper Mississippi Valley region is about 80 m (∼250 ft) thick, and the spotty exposures provide only a glimpse of the karst features.
In the subsurface, zinc mine workings in the Lower Ordovician Knox Group of Tennessee and Kentucky provide detailed descriptions of dissolution, brecciation, and fracturing. However, these descriptions are biased because the focus is on the distribution of ore bodies and not specifically on the geologic history of the breccia. Subsurface studies using core material provide some of the best descriptions of breccia fabrics and Stratigraphic relationships. Image logs from wellbores have been used successfully to identify and describe brecci-ated zones. These descriptions, however, are limited to a one-dimensional view, and interpreting the three-dimensional (3-D) distribution is commonly model driven. If a faulting model is used, then the breccias are most likely to be interpreted as fault breccia, and tectonics is the most likely driving force. If a solution-collapse model is used, breccias are most likely to be interpreted as karst, and gravity is the most likely driving force. Close attention to breccia fabric can be most helpful in discriminating between these two models (Loucks, 1999). Modern 3-D seismic volumes are useful for mapping structural features that can be linked to faulting and/or solution collapse. Interpretation of these structural features is once again based mostly ona geologic model, and close attention must be paid to details to ascertain the most correct model.
A basic problem in interpreting the origin of breccias, fractures, and faults is that these features can have numerous origins. Distinguishing between sedimentary and diagenetic breccias is normally not controversial; the principal controversy is distinguishing between fault breccias, karst breccias, and cavern collapse breccias (including hydrothermal breccias), or between breccias with a tectonic origin and breccias with a gravity origin. Four basic questions need to be answered to resolve this controversy: (1) Did a cavern exist? (2) What is the age of the cavern relative to the host stratigraphy?(3)What is the age of the collapse? (4) What is the relationship of the cavern to structural elements?
For gravity to be the principal driving mechanism in the originofabreccia, acavern hadtohave existed. Key criteria for identifying whether a cavern existed are (1) observing an intact cavern and (2) demonstrating that the breccia blocks have moved downward from their original positions. Once the presence of a cavern has been established, the next question is the age of dissolution and cavern collapse. Times of cavern formation and collapse can be inferred from the nature of the cavern fill, its relationship to stratigraphy (particularly to exposure surfaces), the crosscutting relationship of fractures, and mineral paragenesis.
In this chapter, I begin with descriptions of outcrop examples because they provide the most robust picture of the origin and geometry of collapse breccias. Next, I summarize subsurface data, with special attention to mineral deposits and petroleum reservoirs. Finally, I evaluate the data presented in terms of understanding the development of the collapse breccias found in the Lower Ordovician.
Lower Ordoviciankarst and collapsebrecciasassociated with the Sauk-Tippecanoe unconformity have been described extensively in eight areas (Figure 1): (1) the Franklin Mountains, west Texas; (2) the Llano uplift, central Texas; (3) the Nopah Range, eastern California;(4)the St. Lawrence promontory, Newfoundland; (5)the Appalachian Mountains, Virginia; (6)the Arbuckle Mountains, Oklahoma; (7) the upper Mississippi Valley, Wisconsin; and (8) the Ottawa embayment, Ontario. I discuss the characteristics of each location, focusing on observations pertinent to interpreting the origin of the collapse breccias. I could not find refereed articles on collapse breccia or Lower Ordovician karsting in the outcrops of the Ozark Uplift, Missouri.
El Paso and Ellenburger Groups
Franklin Mountains, West Texas
The El Paso Group is exposed along 32 km (20 mi) of outcrop in the Franklin Mountains, El Paso, Texas (Figure 1) (Lucia, 1995)— arguably one of the best exposures of the Lower Ordovician in North America. The El Paso Group is more than 300 m (>1000 ft) thick and is overlain by the Upper Ordovician Montoya Group (Figure 2). The Sauk-Tippecanoe unconformity represents a time gap of 30 m.y. Laterally extensive breccias are found within the upper 45 m (150 ft) of the El Paso Group within the Ranger Peak Formation (Figure 3). These dolomitized breccias extend for several miles and are bounded by unbrecciated limestones. Several pillars of unbrecciated limestone are found within the breccia. Where the Ranger Peak Formation is limestone and unbrecciated, contact with the overlying Upper Ordovician Montoya Group is level and sharp. Where the Ranger Peak Formation is brecciated dolostone, the contact is irregular, and large clasts of El Paso lithology are found embedded in the Montoya sediment. Some breccias contain sandy fill material that is similar in texture to that of thin beds of sandstone found at the unconformity. These observations indicate that the breccias were formed before deposition of the Upper Ordovician Montoya and are therefore properly interpreted as karst breccias.
Figure 3 illustrates that the Ranger Peak Formation is thinner where karsted by an average of 13 m (43 ft) (Lucia, 1995). The thinning is interpreted to be caused by a collapse of small caverns formed within the Ranger Peak Formation. The overlying Upham Member of the Montoya, about 30 m (∼100 ft) thick, does not appear to vary in thickness to compensate for the 13 m (43 ft) of thinning of the Ranger Peak Formation. Given these observations, it is interpreted that karsting at the El Paso unconformity produced a cavern system in the upper part of the Ranger Peak Formation that collapsed after the Montoya time of deposition. Structural sags of this nature have recently been ascribed to strike-slip faulting and flower structures (Davies and Smith, 2006). However, in this well-exposed and extensive outcrop, no tectonic faulting can be found associated with the structural sag. This sag was therefore formed by collapse brecciation of caverns formed in a karst environment in response to gravitational forces and not to tectonic stresses.
Associated with the Ranger Peak Formation karst-ing are vertical breccia pipes that start 300 m (1000 ft) below the Sauk-Tippecanoe unconformity and continue 300 m (1000 ft) above the El Paso into the overlying Silurian Fusselman Formation (Figure 4). The presence of large breccia blocks in the El Paso far below their stratigraphic positions demonstrates the past presence of large caverns. Blocks of the El Paso 30 cm (1 ft) in size can be found as much as 120 m (400 ft) below their stratigraphic positions, suggesting extensive, wide, vertical passageways. Slabs of the El Paso 200 m (700 ft) across can be found 60 m (200 ft) below their Stratigraphic positions and are interpreted to be cavern-roof collapse. These observations establish the presence of a large collapsed-cavern system within the El Paso. The cavern system was formed by dissolution, and brecciation was gravity driven.
Constraints on the time of cavern formation can be inferred by the nature of the cavern fill and the relationship of breccias to stratigraphy, with particular attention to exposure surfaces. Breccia fabrics are generally clast supported, with the volume between clasts filled with laminated and chaotic internal sediment. Only a few examples of breccias with open pore space between clasts have been found. A wide variety of fills are found between the El Paso breccia blocks. The fill is commonly a reddish laminated carbonate, similar to that of sediment found locally at the El Paso unconformity, suggesting that the large caverns were present and partially collapsed at the beginning of the Late Ordovician.
Constraints on timing of cavern collapse can be inferred from the composition of the breccia blocks. Blocks of the Montoya are found in the vertical breccias 3 to 30 m (10–100 ft) below the Montoya, demonstrating that collapse of these large El Paso caverns continued after deposition of the Upper Ordovician. In the Great McKelligon sag, the entire Montoya Group collapsed 60 to 90 m (200–300 ft) into the El Paso (Figure 4). Note that breccia pipes continue up through the Montoya Group, where they contain clasts of Middle Silurian Fusselman Dolostone, demonstrating that collapse continued through the middle Silurian.
Maps of modern caves are generally related to existing fracture patterns, and it can be expected that paleocaverns show some linearity. Some vertical breccias are circular, whereas others have a linearity that could be interpreted as Ordovician fracture control. Collapse-induced faulting is observed in one major collapse feature, illustrating that gravitational forces can be the cause of some faulting. Fractures surrounding collapse breccias have not been studied extensively in this outcrop, but mosaic breccias have been observed in the Montoya Group overlying the El Paso karst.
Several major tectonic faults cut the El Paso Group and extend into Precambrian basement rocks. They are presumed to be of Tertiary age and related to the Basin and Range province structure responsible for the Franklin Mountain exposures. Where these faults cut collapse breccia, the breccia is highly fractured, illustrating that faulting occurred after formation of the collapse breccia. Fault breccia is characterized by being monomictic and granulated, whereas collapse breccia is typically polymictic and composed of discrete rock fragments.
In summary, an extensive cavern system was developed in the El Paso Group before deposition of Upper Ordovician sediments. Caverns ranged in size to as much as 60 m (200 ft) in height and 210 m (700 ft) in width. The cavern system appears to have had two main characteristics: (1) a laterally continuous system in the upper 45 m (150 ft) of the El Paso Group in the Ranger Peak Formation composed of small caverns and crevasses, and (2) a laterally restricted system of large caverns developed mainly 120 to 150 m (400 to 500 ft) below the El Paso top in the McKelligon Canyon Formation.
Some cavern collapse occurred before deposition of the overlying Upper Ordovician Montoya Group, and collapse continued after deposition of the Montoya Group and through the middle Silurian. The collapse resulted in two structural features: (1) collapse of the cavern system in the upper 45 m (150 ft) of the Ranger Peak Formation, producing a 15 m (50 ft) structural sag about 1500 m (∼5000 ft) wide, and (2) breccia pipes about 300 m (∼1000 ft) wide and either circular or elongated, which started in the middle El Paso and extended 300 m (1000 ft) above the El Paso to the middle Silurian Fusselman Dolostone. The Silurian Fusselman Dolostone is known to be capped by significant karsting, and the upper part of the breccia pipes may be influenced by Silurian karsting.
The linearity of some of the vertical breccias may be related to Ordovician fractures. If so, however, the fractures had no significant offset. The only faulting observed that can be linked to these karst and collapse features can be attributed to collapse of the El Paso caverns. Fault-related fractures crosscut the collapse breccias.
The Lower Ordovician Ellenburger Group crops out in the central Texas Llano uplift region (Figure 1). A detailed study of collapse breccia has been made of a quarry that is 1000 m (3300 ft) long and tens of meters high (Loucks, 1999; Loucks et al., 2004). The quarry wall is described in detail, ground-penetrating radar (GPR) is used to image the breccia in 3-D, and cores are used to calibrate the GPR interpretations.
The breccia is composed of tilted large slabs (up to 7 m [23 ft] long), breccia blocks, and cavern fill composed of cobble-size chaotic breccias and fine-laminated carbonate sediment. The similarity of the tilted slabs, chaotic breccia, and laminated infill sediment to features found in modern caverns is used as evidence that this is a collapsed-cavern deposit formed in response to gravity.
Two episodes of cavern formation can be documented from conodonts found in cavern sediment fill and from burial history analysis. The most extensive cavern formation occurred during a 110-m.y. period following Ellenburger deposition— a time gap represented by a major unconformity above the Ellenburger in this region. After burial and uplift, a later cavern formation episode occurred during the Pennsylvanian. Breccias from these caverns contain clasts of brecciated material from the previously lithified cavern breccia.
No discussion exists on the relationship of the breccia geometry to structure. Faults that are observed on the quarry wall and in the GPR data are presumed to be related to collapse of the cavern system.
In summary, this collapse breccia has a history more complex than that of the El Paso caverns. It is difficult to constrain the time and environment of cavern formation because of the 110-m.y. time gap and because the infill sediment cannot be related to overlying units. The caverns are presumed to have formed in the meteoric environment because of the similarity between the breccia fabrics and modern cave breccias and fills. The breccia geometry is not related to structure. In addition, the effect of cavern collapse on overlying units cannot be observed because of the lack of overlying formations in this quarry.
A thick section of the Lower to Middle Ordovician Pogonip Group crops out in the Nopah Range of southeastern California (Figure 1) (Cooper and Keller, 2001). The section is 700 m (2100 ft) thick at its maximum and tapers to zero subcropping against the Upper Ordovician unconformity. It is unconformably overlain by the Eureka Quartzite of the Middle Ordovician (Figure 5). The Pogonip Group is principally dolostone, although some units have significant amounts of terrigenous material.
The breccia is found mostly in bodies that crosscut stratigraphy, but some breccias are stratiform. Most of the breccia is chaotic and polymictic, with some matrix-supported and some frame-supported breccia. Crackle and mosaic fractures are found to be associated with these breccias. Eight vertical breccia bodies, ranging in size from one to many tens of meters wide and 0.5 m (1.5 ft) to more than 100 m (>330 ft) thick, are distributed across about 500 m (∼1650 ft) of section (Figure 6). One breccia pipe is shown to be 180 m (594 ft) deep. The ratio of vertical to lateral dimensions of some pipes is more than 50:1. Mapping of the breccia bodies shows that they are related to a succession of exposure surfaces found in the Pogonip Group. Surprisingly, no karsting is noted at the unconformity.
The nature of the breccia and infilling sediment suggests an Early Ordovician age for cavern formation. The matrix is composed of varying mixtures of silt- and sand-size detrital dolomite, terrigenous clay, quartz silt, and sand. Some of the matrix is sediment stained with red iron oxide similar to overlying terra rossa paleosols.
Breccia clasts tend to resemble the local host rock. Within one unit, the matrixis distinctive orange-stained silt derived from the overlying unit. In the larger breccia pipes, a stratigraphic arrangement of clasts is found. Some pipes cut earlier pipes, forming composite bodies with clear boundaries and distinctly different clast assemblages. These observations show that dissolution and gravity are the principal drivers for the formation of these collapse breccias and that they formed periodically during the Early Ordovician associated with disconformities.
Vertical breccias are distinctly different from fault breccias and not related to faulting. No stratigraphic offset is reported associated with the vertical breccias. However, it is likely that fractures controlled the geometry of the vertical breccias.
In summary, brecciation in the Pogonip Group appears to be concentrated in vertical crevasses throughout the Lower Ordovician, with no karsting concentrated at the unconformity. Instead, breccias are associated with internal disconformities. One breccia is 180 m (594 ft) deep. The observation that breccia blocks are found below their stratigraphic position shows that these were once caverns. The association of breccias with disconformities, the similarity of infilling sediments to the overlying unit, and the lack of brecciation in the Eureka Quartzite suggest that the caverns formed and collapsed during the Early Ordovician. No discussion is presented on the relationship of these vertical breccias to fractures, although logic suggests that their location could be controlled by Early Ordovician fracturing.
St. George Group
Karsting in the St. George Group in northwestern Newfoundland, Canada (Figure 1), was described by Knight et al. (1991). The St. George Group, some 500 m (∼1600 ft) thick, is primarily limestone, with dolostone associated with collapse breccias. The St. George Group is mostly of Early Ordovician age but includes some Middle Ordovician strata. The St. George unconformity, marking the top of the Lower Ordovician, is found at the base of the upper Aguathuna Formation (Figure 7). The time gap is short, estimated at between 1 and 3 m.y. The unconformity is a faulted and karsted surface, with karst breccias found at the surface and as much as 120 m (380 ft) below the unconformity.
Several types of karsting and their relationship to structure and stratigraphy are described. Small fractures and associated stratiform caverns (on a scale of tens of centimeters) filled with dolomite silt and green shale are found below the unconformity to a depth of about 32 m (∼100 ft). Intrastratal caverns, several beds thick and hundreds of meters long, are filled with oligomictic breccia composed of locally derived fragments of burrowed and mottled dolostone and overlain by fractured rock composed of crackle and fitted breccia cemented by spar. Vertical breccias that connect to the unconformity are 1 to 150 m (1 to 500 ft) wide and tens to hundreds of meters deep. They are found along syn-sedimentary fractures and faults. These breccias are filled with polymictic chaotic breccia.
Collapse of one cavern within the upper St. George Group produced an oligomictic breccia and formed a subsidence doline 10 to 70 m (3.3 to 230 ft) deep, about 50 m (∼150 ft) wide, and 200 m (660 ft) long (Figure 8). This collapse feature is filled with middle Aguathuna sediment, indicating collapse before the Lower Ordovician was exposed. Later brecciation of the middle Aguathuna suggests that collapse continued. During formation of the unconformity, dissolution along faults produced vertical breccias and collapse dolines as much as 50 m (150 ft) deep and filled with overlying sediment. Breccia pipes that are circular in plan view cut the unconformity and penetrate the overlying Middle Ordovician Table Point Formation.
Knight et al. (1991) report on breccia clasts that are 50 m (150 ft) below their stratigraphic position. This observation, together with the integrated system of vertical and stratiform caverns, demonstrates the existence of a large cavern system in the upper 100 m (330 ft) of the Lower Ordovician and shows that gravity collapse is the principal control on the formation of the breccia.
The similarity of some of the infill sediment to overlying strata supports the conclusion that this cavern system formed before deposition of overlying Middle Ordovician sediments. In addition, Knight et al. (1991) present evidence that collapse occurred before the Lower Ordovician was exposed, which also suggests the presence of an extensive cavern system in the Lower Ordovician St. George Group before the deposition of the Middle Ordovician Table Point Formation.
Much of the cavern collapse must have occurred before deposition of the Middle Ordovician Table Point Formation, given the observation that dolines are filled with Table Point Formation sediment. However, the presence of breccia pipes extending into the Table Point Formation indicates that some collapse continued into the Middle Ordovician.
Knight et al. (1991) presented compelling evidence that synsedimentary fractures and faults controlled the location of much of the dissolution. These fractures are linked to the Taconic orogeny. In addition, fracture-veined dolomite patterns are found above the stratiform breccias that are similar to those of the crackle breccias and filled fabric (Loucks, 1999), suggesting that these fractures were formed as part of cavern collapse. Unfortunately, the age of these fractures cannot be determined, except that they are younger than the dolomite cement.
In summary, persuasive evidence is presented for the existence of a major cavern system in the upper 120 m (380 ft) of the St. George Group. Vertical fissures linked to the unconformity and the nature of cavern fills attest to the fact that this cavern system existed before deposition of the overlying Middle Ordovician sediments. Indeed, robust evidence suggests that cavern formation began within the St. George Group, producing subsidence dolines. These subsidence dolines were formed by collapse brecciation of Lower Ordovician caverns within the Aguathuna Formation, creating structural sags. Most collapse appears to have occurred before deposition of Middle Ordovician sediments. Important, however, is the description of breccia pipes starting well within the Lower Ordovician and extending up into the overlying Middle Ordovician. The vertical breccias are linked to synsedimentary faults and fractures associated with the Taconic orogeny.
Knox and Beekmantown Groups
Outcrops of the Lower Ordovician Knox Group in Virginia (Figure 1) have been described by Mussman and Read (1986) and by Mussman et al. (1987). The upper Knox Group is as much as 1000 m (3300 ft) thick and is overlain by Middle Ordovician carbonates (Figure 9). In southwestern Virginia, it is composed mostly of dolostone, changing to limestone and dolostone in northern Virginia, where it is called the Beekmantown Group. The unconformity is primarily a disconformity but locally is angular. The time gap is estimated to be 10 m.y.
The unconformity surface is irregular, with relief ranging from 30 m (100 ft) in the southwest to a few meters in the north, with reports of 140 m (460 ft) of relief in the southwest. The paleotopography is thought to be formed by chemical and physical erosion (Mussman and Read, 1986). Local shedding of Early Ordovician-aged debris from paleohighs to fill paleotopographic lows is reported by Mussman and Read (1986).
Breccia bodies are found as sinkholes and sheetlike discordant bodies within the upper 35 m (100 ft) of the Knox and Beekmantown Groups, and as intraforma-tional breccias in the upper 200 to 300 m (660 to 1000 ft) (Figure 9). Sinkholes linked to the unconformity surface are3to35m (10to100 ft) wideand as much as65m (215 ft) deep. They are filled with chaotic breccia that is composed of blocks as large as 2 m (7 ft) in diameter. Discordant breccias, which are interpreted as caverns or collapse dolines, are found 35 m (100 ft) below the Knox unconformity. Subvertical to bedding, they are as much as 12 m (40 ft) long and 2 m (7 ft) high and are filled with massive to laminated sediment, conglomerates, and angular clasts of dolostone. Laminations have scours and dip at various angles. Stratiform breccias, found farther below the unconformity, are a few meters thick and as much as 10 m (33 ft) long. Fitted fabric fractures are found in the host rock lateral to the stratiform breccias.
These breccias are interpreted as collapse breccias. Blocks of the host limestone found on the floor of one outcropping stratiform breccia and the presence of large limestone blocks in one doline fill attest to the presence of a collapsed cavern. The age of cavern formation is based on cavern fills, most of which are non-fossiliferous, and no date of this infill is given. However, some fills are fossiliferous and contain a marine biota of Middle Ordovician age, which is consistent with the conclusion that the caverns were formed before deposition of Middle Ordovician sediments mantling the unconformity. No evidence is presented suggesting that the collapse of the caverns extended above the unconformity. Therefore, it is likely that most of the cavern collapse occurred before deposition of the overlying Middle Ordovician sediments. No relationships between collapse breccias and fractures or structure were reported.
In summary, compelling evidence is reported to conclude that a cavern system developed within the upper 300 m (1000 ft) of the Knox and Beekmantown Groups before deposition of overlying Middle Ordovi-cian sediments in Virginia. Some cavern collapse occurred before deposition of the overlying Middle Ordovician sediments. No evidence is presented that collapse brecciation of the Knox cavern system extended into the overlying strata. It is clear that the top of the Knox Group is erosional, which accounts for some of the local relief. It is possible, however, that some of the paleotopography on the Knox surface is a karst terrain related to dissolution and collapse of the Lower Ordovician cavern system.
Lower Ordovician Arbuckle outcrops are well exposed in the Arbuckle and Wichita Mountains of Oklahoma (Figure 1). The Arbuckle of the Wichita Mountains has been studied in detail by Gao and Land (1991), but these authors do not mention karst breccias. The outcrops of the Arbuckle Mountains are divided into the Arbuckle anticline to the west, the outcrops along Interstate 35, and the Tishomingo anticline to the east (Figure 10). Although the breccias and dolomitization have generally been linked to faulting, Gao et al. (1995) reported the presence of breccias in the Tishomingo anticline that are interpreted to be karst breccias. The breccias are found within thick (100-m [330-ft]-scale) dolostone bodies that occur within the McKenzie Hill, Cool Creek, and Kind-blade Formations (Figure 11). The largest dolostone body is 8 km (12.8 mi) long and is located at the McKenzie Hill-Cool Creek contact.
The breccia is described as composed of “large disturbed and chaotic blocks of dolomitized strata” (Gao et al., 1995, p. 323). Little information as to the size and distribution of the breccia is presented and no evidence as to the age of cavern formation is presented. Gao et al. (1995) interpreted the breccias to be paleokarst because they are found within a dolostone body that generally conforms to stratigraphy. The breccias are found within the Lower Ordovician Arbuckle, and no reports were found in the refereed literature of karsting at the top of the Arbuckle Group in the Arbuckle Mountains. It is possible that the paleokarst described by Gao et al. (1995) is genetically linked to an internal disconformity.
The dolostone body is cut by numerous faults, and the breccias and dolostone have generally been thought to be of tectonic origin. However, the authors conclude that the breccias are not fault related because these breccias are restricted to dolostone and are not associated with faults in the adjacent limestone beds.
Prairie du Chien Formation
Limited outcrop exposures of the Lower Ordovician Prairie du Chien Group (OPdC) exist in southwestern Wisconsin and northeastern Iowa (Figure 1). The OPdC is 30 to 90 m (90 to 300 ft) thick and is unconformably overlain by the Middle Ordovician St. Peter Sandstone (Figure 12). The Middle Ordovician unconformity has more than 60 m (>200 ft) of relief that has been filled by basal units of the St. Peter Sandstone. This relief is thought to be the result of fluvial erosion and karst collapse (Smith et al., 1993).
A detailed study of the Lower Ordovician unconformity was made by Flint (1956), who described two causes for the irregularities found at the OPd∈St. Peter contact: (1) domes produced by draping of flank beds over algal bioherms caused by differential compaction, and (2) solution of the OPdC and compaction of overlying beds caused by overburden weight. These collapse features are more than 30 m (>100 ft) deep and are filled with chert pebbles, sand, and shale.
Flint (1956) concluded that evidence of subsurface solution and compaction deformation of the overlying units under severe pressure of the overlying sandstone is undeniable. From this description, it seems likely that a series of caverns formed mainly near the OPdC unconformity but in places extended as much as 60 m (200 ft) into the OPdC. Most of the cavern collapse occurred during deposition of the conglomerate and shale unit that immediately overlies the OPdC carbonate and before deposition of the main well-sorted sandstone intervals, which is the typical lithology of the St. Peter Sandstone.
Not all dissolution is found to be associated with the Lower Ordovician unconformity. Smith et al. (1993) reported a significant angular discordance between the upper Prairie du Chien Shakopee Formation and the lower Prairie du Chien Oneota Formation, and that dissolution collapse breccias are abundant within the upper 10 m (33 ft) of the Oneota Formation. Breccias are composed of jumbled dolomite and chert clasts in a green sandstone matrix.
No detailed map of the irregular unconformity surface was found. However, Smith et al. (1993) presented a map showing thin areas of the OPdC that are interpreted to represent dissolution and erosion before the Middle Ordovician deposition. The smallest closed contour, however, is 10 km (6.2 mi) across because of sparse data control. No discussion on the relationship of surface irregularities to structure was found.
In summary, the principal karst feature is the large relief on the Middle Ordovician unconformity. In some places, the relief cuts through as much as 60 m (200 ft) of the OPdC or almost the complete section. The contribution of erosion to the formation of this relief is debatable, and it seems likely that some of the paleotopog-raphy iscausedby collapse of caverns that existed in the upper measures of the OPdC and was overlain by a paleosol of shale, sand, and chert pebbles. The caverns are presumed to have been formed by dissolution before deposition of the Middle Ordovician sediments, and most cavern collapse probably occurred before deposition of the St. Peter Sandstone was completed.
I could find no reports that collapse of OPdC caverns affected the overlying Middle Ordovician St. Peter Sandstone or Platteville units. However, the overlying Middle Ordovician carbonates that are the host of zinc deposits in southwestern Wisconsin have a significant karst overprint, as described by Heyl et al. (1959). Shallow, elliptical, structural synclines are described, bounded by fault planes that dip outward. The fracture and breccia pattern described could result from cavern collapse, as described by Loucks (1999). However, outcrops are not sufficient to link these possible collapse structures to collapse breccias in the OPdC.
Dix et al. (1998) described paleokarst in the Lower Ordovician Beekmantown Group in southeastern Ontario, Canada (Figure 1). The Beekmantown Group (Figure 13) is about 130 m (∼400 ft) thick in this region and is overlain by the Middle Ordovician Rockcliff Sandstone. The Beekmantown Group is composed of the upper Carillon, the middle Beauharnois, and the lower Theresa Formations. The Sauk-Tippecanoe contact is reported to be at the top of the Carillon Formation by Dix et al. (1998). However, it is now thought to be at the contact between the Carillon and Beauharnois Formations (G. R. Dix, 2008, personal communication). The time gap was thought by Dix et al. (1998) to be less than that suggested for the St. George outcrops, or about 1 m.y.
Karst is found below a disconformity at the contact between the Carillon and Beauharnois Formations, now thought to be the Sauk-Tippecanoe unconformity. Karst-ing is minimal, being expressed as small caverns and solution-enlarged fractures filled with dolomudstone 1 to 2 m (3 to 6 ft) below the unconformity. Other paleo-caverns found deeper in the Beauharnois Formation are thought to have been formed by burial diagenesis based on relationship with wispy dissolution seams (Dix et al., 1998).
In summary, this location reveals the least amount of karsting described in the literature linked to the Sauk-Tippecanoe unconformity. No relief is described, and paleokarst appears to be restricted to small dissolution features within 1 to 2 m (3 to 6 ft) of the surface. The time gap is described as minor (Dix et al., 1998), which may account for the small amount of karst. The burial argument for the deeper paleocaverns is debatable. The suggestion that laminated and bedded geopetal infilling sediments were transported down faults and into the small paleocaverns is not compelling, nor is the timing between cavern formation and dissolution seams. It would be more compatible with other descriptions if these deeper paleocaverns were Early Ordovician in age.
Summary of Lower Ordovician Outcrop Studies
The most extensive exposures of Lower Ordovician carbonates are in the Franklin Mountains, El Paso, Texas; the Nopah Range, southeastern California; and Newfoundland. All these exposures show extensive cavern systems occurring within the upper hundreds of meters of the Lower Ordovician that were formed before deposition of the sediments overlying the Sauk-Tippecanoe unconformity. All three cavern systems have a strong vertical as wellas a stratiform component. Fracture and/or faulting control to the dissolution and brecciation is expressed in varying ways. Breccia fabrics are similar, that is, chaotic polymictic, monomictic, crackle, chaotic infill, laminated infill, and so on. The breccias all have breccia blocks that have been transported vertically down 10 to more than 100 m (33–>330 ft). The Ellenburger outcrop in central Texas presents a detailed picture of a major cavern-collapse system. This conclusion is primarily based on comparing the breccia fabrics with breccias found in modern cave systems.
The exposures present some interesting differences. No mention of significant karsting at the Lower Or-dovician unconformity is presented in the Pogonip study. Instead, karsting is linked to disconformities within the Pogonip. Most of the collapse in the St. George and Pogonip cavern systems occurred before deposition of the Middle Ordovician. However, collapse of some St. George stratiform caverns produced structural sags within the St. George, and some evidence of breccia pipes that extend into the Middle Ordovician strata is presented. Much of the collapse of the El Paso cavern system occurred after deposition of the overlying strata, resulting in a 15 m (50 ft) structural sag at the El Paso–Montoya contact and major breccia pipes that extend from within the El Paso up 300 m (1000 ft) into the overlying formations.
Although exposures of the Knox and Beekmantown Groups in Virginia and the Prairie du Chien in Wisconsin are not as abundant, studies show that an extensive cavern system was present in the Lower Ordovician before deposition of the overlying Middle Ordovician. Cavern formation occurred at a disconformity within the Prairie du Chien, as well as at the top. Whether cavern collapse affected overlying units is unknown. However, collapse appears to have been completed before deposition of the overlying Middle Ordovician.
The unconformity surface in these two outcrops has significant relief, much more than reported from the other El Paso, Pogonip, and St. George outcrops. Surface relief on the Knox Group ranges from 30 m (100 ft) in southwestern Virginia to a few meters in northern Virginia, and the relief on the Prairie du Chien in southwestern Wisconsin is asmuch as 60 m (200 ft). The relief is interpreted to have formed by erosion, and extensive regional erosion of the Lower Ordovician occurred throughout North America. However, some of the paleotopography may be accounted for by the collapse of a cavern system near the top of the Knox and the Prairie du Chien, forming a karst terrain composed of collapsed dolines and associated sags at the Sauk-Tippecanoe unconformity.
Outcrop descriptions of the Arbuckle Group, Ar-buckle Mountains, are not very extensive but suggest karsting and cavern formation related to a disconfor-mity within the Lower Ordovician. The Beekmantown Group (Ottawa embayment) has little karsting, perhaps related to the short time gap expressed by the Sauk-Tippecanoe unconformity.
A well-illustrated link is found between synsedimen-tary fractures and faults and cavern formation in the St. George outcrop and some suggestion of fracture control in the El Paso outcrop. Vertical breccias in the Pogonip outcrop would seem to have fracture control, but none are reported. The close link between the Lower Ordovician fractures and faults and cavern formation in the St. George and not in the El Paso or Pogonip is probably related to the differencein tectonic activity during the Ordovician. The Taconic orogeny was active in Newfoundland during the Early Ordovician. However, little tectonic activity was found in west Texas and southeastern California during this time.
In conclusion, the outcrops discussed present a picture of an extensive cavern system developed within the Lower Ordovician in response to the formation of the Sauk-Tippecanoe unconformity as well as todiscon-formities within the Lower Ordovician. In Newfoundland, a strong fracture and/or faulting control on cavern formation caused by the tectonic activity associated with the Taconic orogeny is found. In Texas, little fracture control is found because the region was tectonically quiet during this time. Collapse of the cavern system seems to have been confined to the Lower Ordovician in Southeastern California but extended above the Lower Ordovician in Texas and Newfoundland. There, the collapses produced structural sags within and above the Lower Ordovician as well as large breccia pipes rooted in the Lower Ordovician caverns and extending hundreds to thousands of feet into the overlying Ordovician and Silurian sections.
Three subsurface Lower Ordovician karst examples are discussed: the Ellenburger Group of Texas; the Ar-buckle Group of Oklahoma and Kansas; and the Knox Group of Tennessee, Kentucky, and Ohio (Figure 14). As in the outcrop discussions, I focus on characteristics of each location and on observations pertinent to interpreting the origin of the brecciation.
Brecciation in the Lower Ordovician Ellenburger Group in the Permian Basin of west Texas (Figure 14) was extensively described by Kerans (1988, 1989). He reported that brecciation and fracturing are commonly observed in the upper 30 to 90 m (100 to 300 ft) beneath the Ellenburger unconformity. Collapse brecciation is found in places as much as 180 m (600 ft) below the unconformity. Breccias are divided into three breccia facies (Figure 15). The lower facies consists of about 45 m (∼150 ft) of clast-supported chaotic dolostone breccia and minor matrix-supported chaotic breccia. The middle facies consists of about 15 m (∼50 ft) of siliciclastic matrix-supported chaotic breccia containing interbedded graded beds of silty sandstone. Clasts of sandstone and shale fragments are identical in lithology to the overlying Middle Ordovician Simpson Group. The upper facies, ranging from 12 to 30 m (40 to 100 ft) in thickness, is composed of fractured and brecciated dolostone. Fracture and mosaic breccias (fitted and crackle) are characteristic of the upper facies.
Subsurface studies are limited to using core descriptions and wireline logs as data. Descriptions are normally restricted to core slabs 7 to 10 cm (3 to 4 in.) wide, greatly restricting information about the relationship of breccias to stratigraphic information that makes outcrop studies so valuable. Such a limited lateral view makes it difficult to determine how far breccia blocks have been transported. However, it is a logical conclusion that the jumble of chaotic breccias results from the collapse into a cavern of some unknown size (Figure 16). The regionally consistent stratigraphy of karst facies also suggests that the collapse of an extensive cavern system developed in the upper 90 m (300 ft) of the Lower Ordovician. Partly consolidated clasts of Simpson Group sandstone and shale 15 to 30 m (50 to 100 ft) below the unconformity demonstrate that the lower strata of the Simpson Group were deposited and partly lithified before the final roof collapse.
Kerans (1988) did not discuss the relationship to tectonic fractures. However, an interpretive map of an Ellenburger cavern system in the Permian Basin (Canter et al., 1993) suggested fracture control of the collapse breccia. A recent study by Gale and Gomez (2007) concluded that two distinct groups of fractures are present: (1) fractures with irregular shapes, sediment infill, saddle-dolomite cement, and no preferred orientation that are probably related to gravity collapse, and (2) younger subvertical fractures with consistent east–southeast and south–southwest strikes that postdate the saddle-dolomite cement and are probably tectonic in origin.
Some 3-D seismic surveys in the Permian Basin (Loucks, 1999) and in the Fort Worth Basin (Hardage, 1996) show structural sags at the Ellenburger level that extend upward into the overlying strata, suggesting that the collapse of Lower Ordovician caverns affects the structure of overlying units. Loucks (1999) illustrated a seismic interpretation from the Permian Basin, west Texas, showing the Ellenburger collapse affecting strata up to the Devonian Woodford Shale more than 300 m (>1000 ft) above the Ellenburger.
McDonnell et al. (2007) analyzed a 3-D seismic volume from the Fort Worth Basin (Figure 14) and concluded that subcircular sag features found in the seismic data are best related to the collapse of an underlying Ellenburger cavern system. Core descriptions from the Ellenburger show typical clast-supported chaotic breccia, with laminated and chaotic matrix sediment, as well as crackle breccias, with sediment filling the fractures. These sag features are 500 to 1200 m (1600 to 4000 ft) in diameter and extend 760 to 1060 m (2500 to 3500 ft) vertically from the Ellenburger unconformity (Figure 17). Seismic fault patterns and the vertical change in sag width were analyzed and found to be comparable with other collapse features, such as caldera collapse, salt dissolution and withdrawal, and mine subsidence. Typ-ical faulting patterns include inward-dipping reverse faults within the collapse and outward-dipping normal faults on the flanks.
The relationship of these sags to structure has been controversial. Many structural geologists related them to pull-apart basins located along a master strike-slip fault system. However, sandbox models show that pull-apart basins are bounded by steep inward-dipping normal faults and do not have outward-dipping reverse faults (Dooley et al., 2004). Therefore, the sags observed in the Fort Worth Basin are more likely to be the result of gravitational forces instead of tectonic forces.
In summary, the subsurface Ellenburger most commonly contains collapse breccias in the upper hundred meters, which are interpreted to be formed by cavern collapse. Good evidence shows that these caverns formed before deposition of the overlying Middle Ordovician Simpson Group in the Permian Basin. The collapse probably was initiated during cavern formation, and convincing evidence shows that the collapse continued into the Middle Ordovician. Seismic studies show breccia pipes that are rooted in the Ellenburger, extending more than 660 m (2000 ft) into the overlying strata. The relationship of these breccia pipes to tectonic forces is still controversial, but the seismic evidence, together with observations of breccia pipes in the Franklin Mountains, supports a nontectonic origin. It is apparent that much of Texas was underlain by an extensive cavern system during the Early Ordovician and that the collapse of the cavern system affected the structure and sedimentation of overlying strata.
Some of the best exposures of Lower Ordovician breccia are found in the zinc mines of Kentucky and Tennessee (Figure 14). Here, the Knox, as much as 900 m (3000 ft) thick, is composed of dolostone and limestone. The Knox is overlain unconformably by Middle Ordovician carbonates with a time gap of 10 m.y. (Mussman and Read, 1986) (Figure 18). The zinc mines are found in brecciated country rock and fringing host rock (Ohle, 1985). The breccias, which have been generally interpreted as collapse breccias (Kyle, 1976), are found from 60 m (200 ft) above the Knox unconformity to more than 240 m (>800 ft) below the unconformity.
Collapse brecciation into an open cavern was first proposed for the origin of these breccias in 1931 by Ulrich (Ohle, 1985) for the Mascot district of eastern Tennessee. Convincing evidence of downward displacement supports the presence of caverns. Breccia blocks are reported to be 3 to 5 m (10-15 ft) below their stratigraphic positions (Sangster, 1987).
The time of cavern development is arguable, with collapse brecciation postdating cavern development. Kyle (1976) divided the breccias (and, thus, cavern development) in the central Tennessee-southern Kentucky district into early and late (Figure 19). Early breccias have a pale green, finely crystalline dolostone and chert infilling material, similar in lithology to that of the overlying Wells Creek Formation. This breccia extends down to about 150 m (∼500 ft) below the Knox unconformity and does not extend above the unconformity. The similarity of the infilling sediments to the overlying sediments and the presence of altered clast margins, interpreted to represent groundwater oxidation, suggest that these caverns formed before deposition of the Middle Ordovician sediments (Kyle, 1976). These breccias generally are not ore bearing.
Late breccias are composed of a basal part filled with fragments of detrital dolostone, chert, and argillaceous residue and an upper part cemented by saddle dolomite, calcite, sphalerite, fluorite, and barite. In some cases, late breccias envelop and displace early breccias. Most late breccias are developed at several specific stratigraphic levels, are about 10 m (3.3 ft) thick, and are associated with limestone units scattered throughout the dolostone-dominated sequence. Kyle (1976) reported considerable evidence of cavern-roof collapse in the late stratiform breccia bodies. He also reported local breakthrough breccia bodies, referred to here as vertical breccias. These breccias affect most of the Mascot interval and, to some degree, the overlying Middle Ordovician.
Kyle (2006, personal communication) postulates that these caverns were also formed before deposition of the Middle Ordovician sediments but by a different hydro-logic system, which was in part controlled by the early cavern system. The main ore bodies are found in the late breccia in a 60 m (200 ft) interval within the Knox Group.
Sangster (1987) and Montanez (1994, 1997) suggested a post-unconformity dissolution event that created some of the caverns that collapsed to form collapse breccia (commonly referred to as hydrothermal breccia). It is postulated that the ore-bearing solutions alternately formed caverns and deposited sphalerite, dolomite (hydrothermal dolomite), and other minerals after a period of cavern collapse. That is to say that all of the fabrics seen in the Mississippi Valley-type (MVT) deposits were formed by one ore-forming process. Sangster (1987) gave no positive evidence that this has occurred in the Knox Group but pointed out problems with the karst model: no limestone blocks, although limestone beds are present; the lack of cave spele-othems (stalagmites, etc.); and the scarcity of terra rossa cave sediment.
Montanez (1994, 1997) published an extensive study on the burial diagenesis of the Knox Group. She gave an intriguing argument for the origin of intercrystal pore space in late dolostones and for a dissolution phase in the growth of large dolomite crystals. Her conclusion that these late-stage dolostones with inter-crystalline and vuggy porosity were conduits for later fluid movement seems well founded. However, she presented no convincing evidence of late dissolution sufficient to form cavities of sufficient size to collapse and form the collapse breccias observed in the Knox Group. Evidences presented for post-Knox-deposition dissolution are (1) the occurrence of secondary porosity exclusively within the dolostone, (2) the truncation of burial stylolites by solution voids, and (3) the ubiquitous presence of corroded silt flooring solution voids (Montanez, 1994). The dissolution provided conduits for dolomitizing water, so limestone would not be expected to be found, the relationship between stylolites and pore space is always equivocal, and floored sediments say little about the timing of cavity formation, except that the cavity formed before the sediment was deposited.
Despite the evidence of burial dissolution events, it is reasonable to assume that a major cavern system formed in the Knox before the Middle Ordovician and that the collapse started soon after initiation of cavern development. Observations that breccias continue some 60 m (200 ft) into the overlying Middle Ordovician and that breccia clasts in the sediment-filled breccias are cut by late fracturing indicate that cavern collapse continued long after the Middle Ordovician had been deposited. In the deformed Knox strata of eastern Tennessee, laminations of sulfides and clasts of sphalarite are found filling breccias (Sangster, 1987). The age of mineralization is arguable but is thought to be Devonian (Kesler et al., 2004). If so, collapse and deposition of laminated infilling sediment continued after breccia mineralization or at least through the Devonian.
Direct evidence of fracture or fault control for the location of the breccias is scant. Maps of the mines, however, tend to have a rectilinear pattern, suggesting fracture control of the ore bodies and, thus, of the breccias (Mussman et al., 1987). Fractures seem to be organized according to the collapse brecciation model proposed by Ohle (1985) and Loucks (1999) (Figure 20). Chaotic breccia is surrounded by mosaic and crackle breccia vertically and outwardly.
The unconformity surface is reported as highly irregular in Virginia, Tennessee, Kentucky, and Ohio. Furman (1993) reported tower karst 30 m (100 ft) high in Tennessee. The unconformity surface was mapped in detail by Anderson (1991) and Gooding (1992) in Kentucky (Figure 21). Paleotopographic relief can be as much as 90 to 120 m (300–400 ft). According to the map presented by Anderson (1991), closed contours are typically 30 m (100 ft) deep and 0.8 to 4.8 km (0.5–3 mi) across. However, the width of these lows is based on contours subject to well spacing and could be much smaller. The Middle Ordovician Wells Creek Formation is thought to fill in the lows. Gooding (1992), however, reported that the Wells Creek Formation is 3 to 28 m (10–95 ft) thick, which is an insufficient thickness to fill in the 90 to 120 m (300–400 ft) of relief. Perhaps the difference in thickness between the topographic relief and the Wells Creek Formation results from the collapse of an upper Knox cavern system after deposition of the Wells Creek Formation. This idea is supported by a core description that shows brecciated Knox in a green shaly matrix at the base of the Wells Creek Formation, which could represent collapse breccia. According to Harris et al. (1992), lows in the Knox unconformity coincide with vertical breccias in the Knox, suggesting that the lows are dolines or sinkholes linked to the collapse of vertical breccias. Oil production from the overlying Middle Ordovician Black River is concentrated over lows in the Knox unconformity, suggesting that the collapse of Knox caverns induced fractures in the overlying Middle Ordovician.
Smosna et al. (2005) described the Lower Ordovician paleokarst in an oil-productive trend in Ohio near the Beekmantown subcrop. The paleotopographic relief is as much as 20m (6 ft), and oil production is concentrated on the topographic highs. The unconformity is overlain by angular fragments of the underlying Beekmantown, thin beds of green-black shale, and beds of glauconitic sandstone and dolostone. Sand-filled fractures and vugs are found as much as 2 m (7 ft) below the contact. Chaotic breccias that grade into crackle breccias are found within the Beekmantown to a depth of at least 12 m (3.6 ft) below the unconformity. Vugs and caverns, 30 to 90 cm (1–3 ft) high, have been observed in image logs. The interpretation of image logs is a new technology, and it is possible that these are not large vugs but are highly porous patches. The chaotic breccias contain a matrix of green clay and sandy dolomite. This lithol-ogy is similar to that of the overlying sediments of the Wells Creek Formation, indicatingthat the caverns were formed before Middle Ordovician deposition.
In summary, credible evidence exists to show that brecciasin the Lower Ordovician Knox Group owe their origin to the collapse of an extensive cavern system that formed before deposition of the overlying Middle Or-dovician units. Breccias described from Knox subsurface exposures all have a combination of stratiform and vertical breccias. Some vertical breccias contain Middle Ordovician Wells Creek sediment and are genetically linked to the Sauk-Tippecanoe unconformity. Stratiform collapse breccias located within the Knox Group commonly have little internal sediment. The age of these caverns is arguable, but it is thought that they are also genetically related to the unconformity. However, the deeper breccias may be related to cavern development at disconformities within the Knox.
Collapse was initiated soon after or as part of cavern formation and continued through the Middle Ordovician and perhaps the Devonian. The paleotopography found on the Sauk-Tippecanoe surface may represent internal drainage of a karst terrain. Dissolution and collapse of caverns formed near the unconformity may have resulted in dolines and structural sags hundreds of feet deep and thousands of feet across at the unconformity, which may also be expressed by the thickness variations in the Middle Ordovician Wells Creek Formation.
The geometry of the caverns as displayed in maps of mine workings resembles orthogonal fracture patterns. The fractures must be Early Ordovician in age if the caverns are Early Ordovician. Detailed petrographic and geochemical studies have demonstrated other dissolution events that occurred during burial, but these events are post-Ordovician in age and have not been shown to be capable of forming a major cavern system.
Arkoma Basin, Wilburton Field
The Wilburton field, located in the Arkoma Basin of eastern Oklahoma (Figure 14), produces from the Cambrian– Ordovician Arbuckle Group. The Arbuckle Group is more than 600 m (>2000 ft) thick and is overlain by 250 m (800 ft) of the Middle Ordovician Simpson Group. The field is located on a horst block of Pennsylvanian age, with some 300 m (1000 ft) of throw (Mescher et al., 1993). The upper 30 to 75 m (100–250 ft) lacks fracturing and brecciation but has intercrystalline porosity of 4 to 18% and permeabilities of as much as 936 md. Collapse breccia is found between 90 and 150 m (300 and 500 ft) below the Lower Ordovician unconformity and is associated with a major intraformational unconformity within the West Spring Creek Formation (Figure 22).
The types of fractures and breccias are variable. Abundant crackle mosaic breccia and clast- and matrix-supported chaotic collapse breccias with infilling sediment are found in the collapse interval (Bliefnick, 1992). Clasts in chaotic breccias are interpreted from core descriptions to have fallen from their stratigraphic positions. Cavern zones consist of small-scale cavern and collapse zones 1 to 3 m (4–10 ft) thick, with internal sediment fill. The infilling sediment contains detrital carbonate muds with relatively steep dips at variable angles, along with transported detrital quartz sand grains and clasts of carbonate and chert (Mescher et al., 1993). Fractures in this core are numerous. Some fractures are solution, widened from karst dissolution; others are planar and interpreted to be related to later tectonic events.
In summary, descriptions are based on only three cores, which is a limited data set from which to draw any conclusions as to the origin of the brecciated rock. The similarity of the breccia textures to those of the Ellenburger collapse breccia is the basis for concluding that this breccia formed by the collapse of a cavern system that existed within the Arbuckle Group. No attempt to date the cavern-filling sediment has been reported, so the times of cavern formation and collapse are not well constrained. It is curious that no mention is made of suprastratal deformation in the upper interval and that no apparent link to the Lower Or-dovician unconformity is present. Mescher et al. (1993) suggested that cavern formation is related to a major intraformational unconformity located below the upper unbrecciated zone. Bliefnick (1992) presented a schematic diagram (Figure 22) that includes a connection to the Lower Ordovician unconformity. Because this diagram is based on only one core, however, it is basically driven by the Kerans (1988) model. No evidence is presented to support a connection between the brecciated zone and the unconformity as diagrammed in Figure 22.
Anadarko Basin, Oklahoma
Lynch and Al-Shaieb (1991) and Al-Shaieb and Lynch (1993) reported on their studies of cores from the Arbuckle Group in Oklahoma. They surveyed 42 Arbuckle cores from Oklahoma for evidence of karst-ing; of these, only eight cores exhibited such evidence. Five are located in southern Oklahoma on the southern flank of the Anadarko Basin (Figure 14), and three are located on the Central Oklahoma platform near the Kansas state border. All cores are overlain by the Middle Ordovician Simpson Group.
Breccia in the five cores located in southern Oklahoma is found between 125 and 540 m (420 and 1775 ft) below the top of the Arbuckle (Figure 22). The brecciated intervals range in thickness from 3 to 50 m (10–160 ft). Breccia fabrics include crackle breccia and various forms of polymictic chaotic breccia. The breccias and fractures are filled with clayey dolomitic mud-stone. Although no caverns are observed, the 6 m or more (≤20 ft) of chaotic breccia observed in these cores indicates roof collapse into a cavern of similar size. Lynch and Al-Shaieb (1991; also see Al-Shaieb and Lynch, 1993) reported that only Arbuckle conodonts were found in clasts and infilling sediment, suggesting an Arbuckle age for the cavern formation and at least some collapse.
In summary, this study is a good example of what can be learned about the origin of breccias from scattered core material. The fact that only eight of 42 Arbuckle cores exhibit evidence of karst is in contrast to Ellenburger cores, almost all of which have intervals of collapse breccia. The presence of thick intervals of chaotic breccia points to roof collapse into sizable caverns, but the original position of the breccia blocks cannot be determined. The presence of Arbuckle conodonts in the infilling sediment, however, suggests an Ar-buckle age for the cavern formation and for at least some of the cavern collapse. The relationship to fracturing and faulting is not discussed.
Interestingly, the collapse breccias are located within the Arbuckle Group, similar to the karst breccias described from the Arbuckle Mountains. The top of the Arbuckle is not described as karsted. One of the best examples of collapse breccia presented is from 540 m (1775 ft) below the Sauk-Tippecanoe unconformity. It seems unlikely that the cavern formation could be genetically linked to the unconformity. Therefore, as stated by Al-Shaieb and Lynch (1993), the paleokarst is probably caused by multiple episodes of subaerial exposure within the Arbuckle. This agrees with Mescher et al. (1993) that Arbuckle paleokarst in the Arkoma Basin is related to disconformities within the Arbuckle Group and with Gao et al. (1995) that Arbuckle paleokarst observed in the Arbuckle Mountains is within the Ar-buckle Group.
Central Kansas Uplift, Oklahoma and Kansas
The lower Paleozoic section thins from Oklahoma onto the Central Kansas uplift (Figure 14) in response to the pre-Pennsylvanian erosion. On the Central Kansas uplift, Pennsylvanian sediments overlie the Arbuckle. The thickness of the Arbuckle ranges from a few hundred feet to zero, and in some areas the Pennsylvanian lies directly on the Precambrian. From the limited core information available from the Central Kansas uplift, it appears that the upper few feet of the Arbuckle contain solution-enlarged fractures and matrix-supported chaotic breccias. However, insufficient information is present to constrain the age of this breccia or the associated cavern system.
Walters and Price (1948) and, more recently, Cansler and Carr (2001) described the paleotopography of the Arbuckle unconformity in central Kansas. Differential dissolutionhasproducedanirregularsurfacecomposed of dolines that are typically 10 to 20 m (30–60 ft) deep and 300 to 600 m (1000–2000 ft) in diameter (Figure 23). Some dolines are 75 m (250 ft) deep and 1500 m (5000 ft) wide. Estimates of doline diameters are limited by well spacing, and some are probably much narrower than 300 m (1000 ft). The dolines are filled with Pennsylva-nian sediment composed of clay and chert. It is likely that this represents the internal drainage of a karst terrain developed during a long time gap represented by the unconformity.
In northern Oklahoma, the Arbuckle is overlain by the Middle Ordovician Simpson Group. Short cores from the Lower-Middle Ordovician contact in northern Oklahoma have clast-supported chaotic breccias, crackle breccias, and solution-enlarged channels filled with Simpson sand (Al-Shaieb and Lynch, 1993).
In summary, the presence of chaotic breccias in the upper Arbuckle in northern Oklahoma overlain by the Middle Ordovician sandstones suggests that an extensive cavern system was present in the Arbuckle of northern Oklahoma before the Middle Ordovician. On the Central Kansas uplift, the Arbuckle is overlain by Penn-sylvanian formations, and the age of the chaotic breccia found on top of the Arbuckle is poorly constrained. Interestingly, no collapse breccias are described from within the Arbuckle. The highly irregular surface is interpreted to be karst terrain pockmarked by numerous dolines. The Arbuckle Group was extensively eroded, as evidenced by thinning onto the Central Kansas uplift, and some of the paleotopography may be simple erosion.
Summary of Lower Ordovician Subsurface Studies
The most extensive studies on subsurface Lower Or-dovician brecciation are based on core and seismic data from the Ellenburger Group of Texas and on mine and core data from the Knox Group of Tennessee, Kentucky, and Ohio. The 3-D aspect of modern seismic data and of mine workings is key to understanding the history of breccias found in the subsurface. Core studies of the Arbuckle Group in Oklahoma and Kansas are useful for illustrating that similar collapse breccia fabrics are found in the Arbuckle subsurface that may have a similar history to the breccias found in the Texas and Tennessee subsurfaces. Although not discussed here, similar breccias are found in cores from the Black Warrior Basin of Alabama (Wilson et al., 1993).
It is difficult to confirm the presence of paleocaverns from core and seismic data alone using the downward-displacement criteria. Exposures in the Knox zincmines, however, are sufficient to observe that breccia blocks are tens of meters below their stratigraphic positions. The presence of large rotated blocks, found in the Lower Ordovician cores from Texas and Oklahoma, is a strong indication of cavern collapse caused by gravitational forces. Therefore, it is reasonable to conclude that an extensive cavern system was once present in these study areas that extended from the Sauk-Tippecanoe unconformity down to more than 300 m (>1000 ft) into the Lower Ordovician.
The age of the cavern system is also difficult to discern. The similarity of infilling sediment to sediment found in the Middle Ordovician strata immediately above the unconformity argues that a cavern system was present before the Middle Ordovician sedimentation. Simpson sediment appears to be present in some of the Ellenburger collapse breccias. Wells Creek sediment appears to be present in some of the Knox breccias. Cores that cross the unconformity commonly are described as having overlying siliciclastic sediments in solution-enlarged fractures tens of meters below the unconformity, suggesting a karsted surface. Many of the breccias, particularly in the zinc mines in Tennessee, have no distinguishing infilling sediment, so their age remains questionable. Because they seem to be an extension of the breccias having infilling sediment similar to that of the Wells Creek Formation, however, they are probably of the same general age, that is, forming before Middle Ordovician deposition.
Caverns located close to the Sauk-Tippecanoe unconformity are probably linked to the unconformity. In the Arkoma Basin, however, caverns are found lower in the section and are probably related to an internal disconformity within the Arbuckle Group. It may be that the multiple layers of stratiform collapse breccias in the Knox Group can also be explained by a link to as yet undefined internal disconformities.
Cavern collapse has a complicated history. Cavern collapse was probably initiated soon after cavern formation, as observed in modern caves. It is possible that some of the extensive paleotopographic relief on the Knox surface is related to early cavern collapse, producing a karst terrain composed of dolines filled with Wells Creek sediment. In the Knox of Tennessee, vertical brecciation forms local domal structures and associated fracture patterns. Locally, the vertical brecciation produces breccia pipes that extend into the overlying Middle Ordovician.
Cores from the Ellenburger of Texas show extensive collapse brecciation. It is uncommon to find an Ellen-burger core from the Permian Basin that does not contain some collapse breccia. The top of the Ellenburger is commonly difficult to correlate, suggesting an irregular surface. This is expressed in seismic data as a series of structural sagsthat are presumedt of ormby the collapse of the cavern system. They are commonly interpreted as faults. Breccia pipes are observed in seismic data to extend from these structural sags to more than 600 m (>2000 ft) above the Sauk-Tippecanoe unconformity.
The maps of zinc mines in Tennessee commonly show a rectilinear pattern, suggesting fracture control for the collapse brecciation. The fractures must be of Early Or-dovician age, assuming an Early Ordovician age for the cavern system. Brecciation is not related to later tectonic activity. Instead, most fracturing and brecciation seems to be gravity driven and related to collapse brecciation.
Discussion of Outcrop and Subsurface Data
Collapse breccia in the Lower Ordovician carbonates, have been interpreted in many different ways. Establishing the history of collapse brecciation is important to select the best interpretation. In this chapter, I have addressed four basic questions that need to be answered to establish breccia history: (1) Did a cavern exist? (2) What is the age of the cavern relative to the host stratigraphy? (3) What is the age of the collapse? (4) What is the relationship of the cavern to structural elements? The answers to these questions are best sought in outcrop and subsurface studies that have the most complete data sets. They are (1) the Franklin Mountains, (2) the Nopah Range, (3) the St. Lawrence promontory, (4) the zinc mines of Tennessee, and (5) the Texas subsurface (Figure 24).
The observations reported here strongly support the conclusion that the Lower Ordovician breccias were formedby collapse intoan extensive system of caverns. Descriptions of the breccia fabrics include terms typical of collapse breccias, such as chaotic polymictic, mono-mictic, crackle, chaotic infill, and laminated infill. Breccias have been observed from near the Sauk-Tippecanoe unconformity to 240 to 300m(800–1000 ft) below the Sauk-Tippecanoe unconformity and have a strong vertical as well as stratiform geometry.
It is observed that breccia blocks have fallen down stratigraphically, and in some cases, blocks have moved down 10 m (33 ft) to more than 100 m (>330 ft). A crude stratigraphy that has been displaced downward can be seen within a few exposures of vertical breccias. These observations confirm that the breccias have formed by cavern collapse and not by tectonic stresses.
The size of the caverns ranges from smaller than meter size to a scale of hundreds of meters, as deduced by the size of vugs filled with internal sediment, the size of the breccia blocks, and the vertical distance blocks have been transported. In the El Paso Group, cavernsas large as 60m (200 ft) in height and 210m (700 ft) in width have been documented. In the St. George Formation, blocks are found 50 m (150 ft) below their stratigraphic position, suggesting caverns as large as 50 m (150 ft) high. Vertical drops of 5 m (15 ft) have been reported from breccias in the subsurface Knox Group. Although no information has been reported as to the distance breccia blocks have fallen in the Pogonip Group, the scale of brecciation is similar to that found in the El Paso and St. George outcrops, and it is assumed that a large cavern system existed there as well.
Although the age of the caverns is difficult to constrain, all five of the best exposures offer persuasive evidence of cavern formation during the Early Ordovi-cian. The similarity of the internal sediment to overlying sediment is the primary observation for constraining the age of the cavern. In the ElPasoand Pogonipgroups, the presence of internal sediment stained with red iron oxide, similar to terra rossa paleosols found within the Lower Ordovician and at the unconformity, is used to constrain the age to the Early Ordovician. In the St. George Group, collapse occurred before the formation of the unconformity, suggesting cavern formation during the Early Ordovician. Subsurface data from the Knox Group suggests internal sediment sourced from the overlying Middle Ordovician Wells Creek Formation. Core descriptions from the subsurface Ellenburger Group in the Permian Basin suggest infill sediment from the overlying Middle Ordovician Simpson Group. These observations all support an Early Ordovician age for cavern development.
Evidence that some of the cavern formation is genetically related to disconformities within the Lower Ordovician is presented in some studies. In the Pogonip Group, the caverns are genetically associated with a series of disconformities within the Lower Ordovician, and interestingly, not at the unconformity contact. Instead, all 500 m (1650 ft) of breccia are found within the Lower Ordovician. The vertical dimension of the largest breccia is about 100 m (∼330 ft). In the subsurface Arbuckle Group, brecciation is found at a disconformity within the Arbuckle in both the Arkoma and Anadarko basins. Observations from the Arbuckle Mountains also suggest that paleokarsting is constrained to within the Arbuckle Group. One set of caverns within the St. George Group is thought to be genetically associated with a disconformity near the top of the Lower Ordovician. These observations suggest that cavern development linked to disconformities within the Lower Ordovician may be more common than currently thought.
The conclusion from this study is that a large cavern system existed within the upper 300 m (1000 ft) of the Lower Ordovician at the beginning of the Middle Ordovician. This conclusion places constraints on the hy-drochemical or chemical system that formed the cavern system in that it must have been operating during the Early Ordovician and cannot be associated with deep burial processes. Hydrothermal dolomite and sulfide mineralization is after brecciation and cannot be assumed to be genetically linked to the fluids that formed the caverns.
The caverns have a complex history of collapse. Cavern collapse was initiated during the Early Ordovician. In the St. George Group, the unconformity is a faulted and karsted surface with collapse dolines filled with overlying sediment. The collapse of one cavern found within the St. George Group has produced a subsidence doline, which is thought to represent the Early Ordovician cavern collapse. No karsting or collapse is observed at the Pogonip unconformity, suggesting that cavern collapse occurred early in the cavern history. These observations suggest that cavern collapse was initiated during or soon after cavern formation, which is also observed in modern caves.
In many regions, the surface of the unconformity is irregular and has significant topographic relief. The Sauk-Tippecanoe unconformity is clearly a regional erosion surface and can be expected to show some degree of topography. However, it is possible that some of the topography observed on the Sauk-Tippecanoe unconformity can be related to the collapse of caverns developed within a few hundred feet of the unconformity. Note that the unconformity in the well-exposed outcrops in the Franklin Mountains and Nopah Range is flat where no collapse breccia is observed near the unconformity. In the Franklin Mountain outcrop, the unconformity is irregular where the upper 45 m (150 ft) contains collapse breccias. The irregularity is related in part to the cavern collapse after the Late Ordovician.
No dolines have been reported from the El Paso outcrop, suggesting that the irregular nature of the contact is primarily related to the cavern collapse. In the Mississippi Valley outcrop, as much as 75 m (250 ft) of topographic relief is present at the Prairie du Chien unconformity that is related to collapse during deposition of the overlying Middle Ordovician St. Peter Sandstone. The upper surface of the Knox Group in Virginia, Tennessee, Kentucky, and Ohio also has an irregular surface with significant local relief. Similar relief is reported from the Knox Group outcrops in southwestern Virginia. Detailed paleotopographic maps of the Knox surface suggest a closed drainage system similar to a karst terrain. These observations suggest that some of the relief on the Sauk-Tippecanoe surface is related to the cavern collapse and not to simple erosion.
Although the collapse of the caverns most likely started early in the cavern history and continued throughout the unconformity age and into the Middle Ordovi-cian, evidence from the ElPaso and St. George outcrops and from the subsurface Knox and Ellenburger Groups shows that collapse continued to affect large volumes of overlying strata. In the El Paso Group, extensive vertical breccias are found that contain Upper Ordovi-cian and Silurian breccia blocks, attesting to the continual collapse brecciation through 330 m (1000 ft) of overlying sediment. These vertical breccia pipes can be seen on 3-D seismic data from the Permian and Fort Worth basins in Texas. In the subsurface Knox, brec-ciation is observed continuing from the Knox into the overlying Middle Ordovician. Rare sphalerite clasts found in some Knox breccia suggest that the collapse continued through the Devonian.
These observations show a genetic connection between the Lower Ordovician cavern system and the continual collapse into overlying units at a scale of hundreds to thousands of feet. The process by which these breccia pipes are formed has yet to be established and may require post-Early Ordovician dissolution events to account for the space needed to form the collapse breccias. However, these observations indicate that the vertical breccias are rooted in the Lower Ordovician cavern system.
The most striking cases for tectonic control over the location of the Lower Ordovician caverns are in the zinc mines of Tennessee and in the Newfoundland outcrop. Maps of zinc mines in the Knox Group of Tennessee are reported to have patterns similar to orthogonal fracture patterns. These controlling fractures must be Early Ordovician in age because the caverns appear to be Early Ordovician in age. In Newfoundland, a strong fracture and/or faulting control on cavern formation is described. This may be related to the Taconic orogeny that was active during that time. Little evidence of tectonic control for the cavern development in west Texas has been presented. This may be related to the lack of tectonic activity during the Early Ordovician in Texas.
The relationship between structural features formed by cavern collapse (gravitational forces) and structural movements (tectonic forces) is a continuing discussion. The collapse of the Lower Ordovician cavern system has produced structural sags and faults of significant dimensions that are not related to tectonic forces. These collapse synclines have dimensions that can be imaged with seismic data. Strike-slip faulting can also produce structural sags through tectonic extension. In well-exposed outcrops, knowledge of the depositional, dia-genetic, and structural history can be used to distinguish between these two mechanisms. In the subsurface, geologic history is also a valuable tool for sorting out dia-genetic and structural features but is less definitive than outcrop data because of the limited information available from cores, wireline, and seismic data. However, it appears that the fault and/or fracture patterns produced from the collapse (gravity) and the structure (tectonic stress) are sufficiently different to be used to distinguish between them.
Although outcrop exposures of the Knox and Beek-mantown Groups in Virginia, the Prairie du Chien in Wisconsin, and the Beekmantown in Ottawa are not extensive, it is quite clear that a cavern system was present in the upper 100 m (330 ft) before deposition of the overlying Middle Ordovician. The unconformity surface in these two outcrops has significant relief, much more than that reported from the other three outcrops. The surface relief of the Knox Group ranges from 30 m (100 ft) in southwestern Virginia to a few meters in northern Virginia, and the relief on the Prairie du Chieninsouthwestern Wisconsinisas much as 60 m (200 ft). The relief is interpreted to have formed by erosion. Indeed, extensive regional erosion of the Lower Ordovicianis present throughout North America. However, the collapse of a cavern system near the top of the Knox and the Prairie du Chien may have produced collapse dolines and sags, thus accounting for some of the paleotopography.
An extensive cavern system existed in the Lower Or-dovician carbonatesat the time of the Sauk-Tippecanoe unconformity. These caverns collapsed to form collapse breccias. The collapse of this large cavern system began during the Early Ordovician and extended into the Silurian and the Devonian. As a result, the Lower Ordovician cavern system affects the deposition and structure of units hundreds of meters above the Sauk-Tippecanoe unconformity.
The main conclusions from this study are:
Observations reported here strongly support the conclusion that the Lower Ordovician brecciaswere formed by the collapse into an extensive system of caverns located from the surface to at least 300 m (1000 ft) below the Sauk-Tippecanoe unconformity.
The size of the caverns, as deduced by the size of vugs filled with internal sediments, the size of the blocks, and the vertical distance the breccia blocks have been transported, ranges from smaller than a meter to hundreds of meters.
Constraining the age of cavern development is difficult, but persuasive evidence for an Early Or-dovician age of cavern formation is present in this report. However, the age of cavern development is a key question and needs to be further examined.
Some of the cavern formation may be genetically related to disconformities within the Lower Ordovician. Exceptionally good outcrop exposures are needed togeneticallylink caverndevelopment to a specific surface.
The basic conclusion from this study is that a large cavern system existed within the upper 300 m (1000 ft) of the Lower Ordovician at the beginning of the Middle Ordovician.
The caverns have a long and complex history of collapse. The cavern collapse was initiated during the Lower Ordovician. The collapse continued through the formation of the unconformity into the Middle Ordovician. The topography observed on the Sauk-Tippecanoe unconformity is probably a karst terrain related to the dissolution and the collapse of caverns developed within a few hundred feet of the unconformity. The collapse continued through the deposition of overlying strata, forming structural sags and breccia pipes as much as 300 to 600 m (1000–2000 ft) above the Lower Ordovician in some areas and continuing through the Silurian and the Devonian.
A genetic connection is present between the Lower Ordovician cavern system and the continual collapse into overlying units at a scale of hundreds to thousands of feet. The large breccia pipes are rooted in large Lower Ordovician caverns.
Lower Ordovician fracturing and faulting controls cavern development in areas of active tectonic activity. It is likely that the Lower Ordovician fracture patterns control cavern development, even in areas of little tectonic activity.
Structural sags and faults of significant dimension can be formed by the collapse of Lower Ordovician caverns. Distinguishing structural features formed by the cavern collapse from those formed by tectonic stresses is an ongoing problem.
This work was funded by sponsors of the Reservoir Characterization Research Laboratory, an industrial research programat the Bureau of Economic Geology, the Jackson School of Geosciences, the University of Texas at Austin. Discussions with my colleague Bob Loucks were most beneficial. The manuscript was edited by Lana Dieterich and Amanda Masterson. Publication was approved by the Director, Bureau of Economic Geology.
Figures & Tables
Great American Carbonate Bank: The Geology and Economic Resources of the Cambrian—Ordovician Sauk Megasequence of Laurentia
The Great American Carbonate Bank (GACB) comprises the carbonates (and related siliciclastics) of the Sauk megasequence, which were deposited on and around the Laurentian continent during Cambrian through earliest Middle Ordovician, forming one of the largest carbonate-dominated platforms of the Phanerozoic. The Sauk megasequence, which ranges upwards of several thousand meters thick along the Bank's margin, consists of distinctive Lithofacies and fauna that are widely recognized throughout Laurentia. A refined biostratigraphic zonation forms the chronostratigraphic framework for correlating disparate outcrops and subsurface data, providing the basis for interpreting depositional patterns and the evolution of the Bank. GACB hydrocarbon fields have produced 4 BBO and 21 TCFG, mostly from reservoirs near the Sauk-Tippecanoe unconformity. The GACB is also a source of economic minerals and construction material and, locally, serves as either an aquifer or repository for injection of waste material. This Memoir comprises works on biostratigraphy, ichnology, stratigraphy, depositional facies, diagenesis, and petroleum and mineral resources of the GACB. It is dedicated to James Lee Wilson who first conceived of this publication and who worked on many aspects of the GACB during his long and illustrious career.