Development of caves in the vicinity of Cibolo Creek near New Braunfels progressed in response to two distinct tectonic episodes. First, uplift of the San Marcos Arch subjected Lower Cretaceous rocks to subaerial exposure and heightened topographic relief that promoted deep circulation of groundwater through fractures produced during the tectonism. Primary porosity in the Glen Rose Formation became enhanced as cavities were solutionally enlarged; but openings remained poorly integrated. During the Late Cretaceous, the region was episodically covered by shallow seas that deposited calcareous, siliciclastic, and marly sediments above the Glen Rose Formation. The second disturbance, Balcones faulting during the Miocene, heightened local relief in the immediate area and thus steepened hydraulic gradients. Groundwater moving along the gentle southeastern dip enlarged pre-existing, northwest-trending fractures as well as many of those produced by the faulting. As a result, major caves of the area, such as the Natural Bridge Caverns System, consist of well-integrated conduits of large cross-section that are angulate in plan-view and correspond to flowpaths favorably aligned along open fractures.
During the Quaternary, streams draining the Edwards Plateau incised rapidly in the vicinity of the Balcones Escarpment. Levels of groundwater declined in response to cutting of valleys, and levels of passages in caves developed in a descending sequence. Passages were enlarged to canyon-like cross-section by subsequent vadose streams, and later abandoned as water circulated at greater depths. Major collapse of ceilings in some places blocked active conduits and promoted development of routes of diversion, in places along upper levels.
In general, groundwater flowed to the southeast. Northwest-trending fractures account for a large percentage of the overall orientation of cave passages. However, many northeast-trending fractures and some fractures of conjugate sets provided crossovers in the paths of flow among adjacent, master fractures and thereby account for a significant number of passages as well.
Figure 1. New Braunfels area, central Texas.
Figure 2. Natural Bridge Caverns. Profile shows horizons of cavern development and zones of collapse. Lineations indicated in plan view, many of which correspond to mapped fractures. From Kastning (1978, 1983).
Figure 3. Correlation of horizons of passages within Natural Bridge Caverns System, Bracken Bat Cave, and Double Decker Cave. Topographic elevations at the entrances to the caves and at the nearest point on Cibolo Creek are indicated. From Kastning (1983).
Figure 4. Plan and profile of the South Caverns, Natural Bridge Caverns System, showing faults. Plus sign (+) indicates fault has provided permeable zone for speleogenesis. Negative sign (-) indicates fault has prevented flow of groundwater across it. N indicates faults have not affected flow of groundwater during speleogenesis. Throws are indicated in meters. From Kastning (1983).
Figure 5. Rose-diagram indicating orientations of mapped photolineaments (from both high- and low-altitude imagery) and orientations of mapped passage segments in Natural Bridge Caverns, Bracken Bat Cave, and Double Decker Cave. From Kastning (1981, 1984).
Several large caves occur approximately 20 km west of New Braunfels in west-central Comal County, in an area bounded by Cibolo Creek and Bat Cave Fault (Figure 1). The origin of Natural Bridge Caverns and other nearby caves is intimately tied to the depositional, structural, and erosional evolution of the San Marcos Arch and Balcones Fault Zone. Karstic drainage and conduit development are strongly aligned along favorable lithostratigraphic horizons and zones of fracture. The caves exhibit phenomena indicative of episodes of marine deposition, tectonism, surficial downwasting, fluctuations in groundwater levels and flow rates, and speleothem deposition.
DESCRIPTION OF AREA
The cavern area is generally underlain by Lower and Upper Cretaceous rocks (George and others, 1952; Newcomb, 1971; Barnes, 1974). The lowest cavernous unit is the upper member of the Glen Rose Formation, a thin-bedded sequence of fine-grained dolomite and marl of variable hardness (Stricklin and others, 1971). Beds are alternatingly resistant and recessive, and weather to form stairstep topography. The uppermost beds are massive and dolomitic. Up to 17 m of the upper member of the Glen Rose are exposed along valleys of incised streams.
The Glen Rose is overlain by the Bull Creek Limestone and Bee Cave Marl Members of the Walnut Formation. The Bull Creek is a thin- to medium-bedded, hard, well-sorted biomicrite, intramicrite, and fossiliferous intrasparite, and the Bee Cave is a clayey, fossiliferous micrite and fossiliferous marl (Moore, 1961, 1964). The Walnut is approximately 14 m thick in this area.
The Kainer Formation of the Edwards Group overlies the Walnut and consists of aphanitic to coarsegrained, massive- to thin-bedded, hard, brittle, resistant limestone. It is locally dolomitic and bored, contains abundant chert, and in its basal part is nodular, honeycombed, and contains zones of solutional and collapsed breccia.
The cavern area overlies the faulted outcrop of the Edwards Group, bounded on the northwest by the Tom Creek Fault and on the southeast by the Comal Springs Fault (George and others, 1952; Kastning, 1978). The outcrop of the Edwards Group within the Balcones Fault Zone is 20 km wide in this vicinity, and recharge enters the unconfined zone of the Edwards Aquifer by direct infiltration. The bed of Cibolo Creek between Boerne and the western boundary of the outcrop of the Edwards is part of the zone of recharge for the Edwards Aquifer. Water is lost to the Glen Rose Formation along the bed of Cibolo Creek, and is transmitted into the Edwards Limestone across the Hidden Valley and Bear Creek Faults, where the Glen Rose and Edwards are in juxtaposition (George and others, 1952; Kastning, 1978). Comal Springs, within the New Braunfels city limits, is the largest exsurgence of groundwater from the Edwards Aquifer (Guyton and Associates, 1979).
The cavern area lies along the southwestern flank of the San Marcos Arch, which was raised along its northwest-trending axis during the close of the Early Cretaceous (Rose, 1972; Woodruff and Abbott, 1979; Kastning, 1983). The deformation that produced the arch created several sets of fractures. Those of a northwestern orientation lie parallel to the axis of the arch and are accompanied by two sets of conjugate fractures. These sets resulted from extensional stresses during uplift. Later deformation during Balcones faulting produced another set of extensional fractures trending to the northeast and parallel to the individual faults (Kastning, 1981, 1984).
Subaerial exposure of Lower Cretaceous rocks, brought about by uplift of the San Marcos Arch and incision of streams draining the Edwards Plateau in response to Balcones faulting during the Miocene were crucial in the evolution of circulation of kartstic groundwater and development of caves. The chronology of these events and the influence of geologic structure on speleogenesis are described below.
DISTRIBUTION OF CAVES AND PREVIOUS INVESTIGATIONS
There are several large caves in this area, including the Natural Bridge Caverns System, Bracken Bat, Double Decker, Ebert, Dinosaur, Bear Creek, Beal Ranch, Rompel, Zuercher No. 1, and Brehmer Caves (White, 1948; Reddell, 1964a,b; Beck, 1968). Knox (1975) inventoried points of recharge, including caves and enlarged fractures, in the upper basin of Cibolo Creek in Kendall, Comal and Bexar Counties. She correlated these with available hydrologic data for the basin.
Abbott (1973, 1975, 1977a,b) investigated the hydrogeology of the Edwards Aquifer in the vicinity of New Braunfels and attempted to relate development of caves to the diagenetic evolution of the Edwards Limestone and to the hydrochemistry of the aquifer. The results from this analysis were later incorporated, with an interpretation of the evolution of the drainage-basin, into a plausible hypothesis for the development of cavernous porosity in the area of the San Marcos Platform (Woodruff and Abbott, 1979).
Geologic studies undertaken in recent years in the Natural Bridge Caverns System have focused on the influence of lithology and structure on the development of the cave (Kastning, 1978, 1980, 1981, 1984; Knox, 1981). Renewed surveying of the cave by Orion Knox and others has produced one of the most detailed maps of a major Texas cave (Kastning, 1978, plate 2).
DEVELOPMENT OF CAVES
Description of Natural Bridge Caverns System
The Natural Bridge Caverns System, in the west-central part of the Bat Cave 7.5-minute U.S.G.S. Quadrangle, consists of three separate caves: the North Caverns, the South Caverns, and the Jaremy Room (Figure 2). These are remnants of a once larger cave now segmented by collapse and development of dolines.
This is suggested by the orientations and horizontal and vertical alignments of passage segments. Most known passages are shown on the map, but several side passages at the north end of the caverns remain to be surveyed. Total surveyed length of the system is presently 3354 m. The North Caverns, containing the commercially developed passages, is presently the eleventh longest cave in Texas. The end-to-end extent of Natural Bridge Caverns System is 1160 m. Detailed descriptions of the system, history of its exploration, and bibliography are given by Knox (1962) and Reddell (1964a,b).
Chambers of the Natural Bridge Caverns System are as large as any known from caves of the eastern Edwards Plateau. Speleothems within the cave are among the largest and most spectacular in Texas. They include massive stalagmites, soda-straw stalactites up to 4.3 m long and "fried-egg" stalagmites (Beck, 1978).
The only natural entrance to the system is in a large collapsed doline directly behind the present visitor center. A natural span of limestone bridges the doline, providing the namesake for the cave. The doline, entrance, and side passages to the southwest have been known for some time.
Pluto's Anteroom was discovered in 1960, and all passages to the north were explored during the period of 1960 to 1963. The North Caverns, from the entrance to the Hall of the Mountain Kings, was developed for tourism and opened to the public in 1964 (Heidemann, 1979). An artificial tunnel was excavated between the surface and the Hall of the Mountain Kings in order to provide an exit for tours. It has intersected a small, unnamed cave midway along its length.
The management of Natural Bridge Caverns suspected that a continuation of the large, main passage of the North Caverns might exist to the south, beyond the collapsed doline. This was confirmed by exploratory drilling in the late 1960s, when five boreholes were driven to allow entry into the new chambers. One hole intersected the Jaremy Room and another pierced the South Caverns and a short, overlying passage. Views of the cave in plan and profile (Figure 2) clearly suggest that all segments of the system were once interconnected.
The plan-view of Natural Bridge Caverns shows that most passages consist of linear segments oriented along several preferred directions. An overall north-south trend persists, particularly for the passages of larger cross-section. Several east-trending side passages join the main cave.
The profile clearly indicates that levels of passages have developed along several favorable horizons. Vertical position of levels in the cave is correlative with stratigraphic position (Figure 3). The dip is subhorizontal (less than one half of a degree to the southeast) in this vicinity. Large segments of the master conduit on the lowest level are wholly within the Upper Glen Rose Formation, including the South Caverns, Purgatory Creek, Grendel's Canyon, and the Limbo and Lake Passages. Above this level is another significant horizon of passages that includes the South Fault Passage, Chapel Hall, Jan's Long Crawl, Emerald Lake Passage, and some shorter segments between the Inferno Room and Dome Pit. This level is also within the Upper Glen Rose, but nearly at its contact with the overlying Bull Creek Limestone Member of the Walnut Formation.
Passages of shorter extent are found on two or more levels in the Walnut Formation. These are represented by (1) the isolated passage overlying the South Caverns, (2) St. Mary's Hall, (3) the Coon Rooms, and (4) the passage from the Moors to the Inferno Room and Dome Pit.
Natural Bridge Caverns presently contains two watercourses, both of very short extent within the cave. Purgatory Creek enters on the eastern wall of the main passage at one of the lowest points in the cave (64 m below the entrance). It curves once across the passage and leaves through the eastern wall only 18 m south of where it enters. These openings are alluviated and exploration is not possible. A second watercourse, River Styx, enters and leaves the northernmost section of the cave twice in its southward flow. River Styx may very well be an upstream segment of Purgatory Creek, but this remains to be confirmed by techniques of water-tracing. A well has been driven into the cave at River Styx, from which water is pumped to the surface for use at the visitor center.
Purgatory Creek is frequently dry, but during wet periods it increases in flow. Storms of high intensity cause sufficient flow in Purgatory Creek to inundate lower levels of the cave. A storm on May 11-12, 1972, caused devastating floods on the Guadalupe River and nearby Blieders and Dry Comal Creeks (Colwick and others, 1973; Baker, 1975, 1977). Three days after the storm, water in the cave rose approximately 18 m above the bed of Purgatory Creek. After the crest of the flood, the level of water began dropping at about 20 cm per day (Knox, 1981). The floor of the Castle of the White Giants had been under approximately 2 m of water at this time. This is the largest flood recorded in the cave since its discovery.
Water enters the cave by infiltration even during long, dry periods. Seepage is sufficient to promote continual deposition of speleotheins in most passages. The Chandelier formation, a large drapery-stalactite in the Castle of the White Giants, normally receives enough flow to allow water to fall from it in a continuous trickle rather than in drops. Moisture and humidity are noticeably high throughout the cave. Glass doors have been installed in the entrance and exit tunnels to maintain moisture levels and sustain the growth of speleothems.
Bracken Bat Cave is 960 m to the southwest of Natural Bridge Caverns (Figure 1). The entrance is in a doline 30 m in diameter and 10 m deep. The cave extends N 18o W for 130 m as a single large passage, averaging 20 m in width and 10 m in height. Severed by collapse of the doline at the entrance, the cave may extend to the northwest, but to date no such passage has been discovered.
Bracken Bat Cave has developed on one level. Stratigraphically, the main passage is in the Walnut Formation, just below its contact with the Kainer Formation of the Edwards Group (Figure 3). The slope of the entrance passes through the lower part of the Kainer Formation. Bat Cave Fault, named after the cave, is approximately 200 m to the southeast.
Double Decker Cave is 3.93 km north-northeast of Natural Bridge Caverns in the northwest part of the Bat Cave Quadrangle (Figure 1). The cave has 208 m of surveyed passage and attains a maximum depth of 31.7 m below the entrance. The cave has developed on two levels and along two major directions. The Upper Level extends due east for 60 m from the entrance, and the Lower Level extends N 30 o W for 45 m from the entrance before turning to the southwest for 20 m. Although Double Decker Cave is relatively short, its passages have large cross-sections. The pit at the entrance is in the Kainer Formation, the Upper Level has developed primarily in the Walnut Formation, and the Lower Level is wholly within the Glen Rose Formation (Figure 3). Floors of passages are generally covered by mud, breakdown, or flowstone.
Many caves of the New Braunfels area, including some of the largest ones, have developed in the upper member of the Glen Rose Formation despite a much greater areal exposure of the Edwards Group than that of the Glen Rose. Dissolution and speleogenesis have occurred within the alternating sequence of calcareous, marly, and dolomitic beds of the Glen Rose, but relatively little development has taken place in the overlying, crystalline, Edwards Limestone. Beck (1968) suggested that the caves had developed preferentially within calcareous beds in the Glen Rose and that large caves, such as the Natural Bridge Caverns System and Bracken Bat Cave, have developed parallel to the high-velocity flow of groundwater caused by a convergence of flowpaths in the vicinity of these caves. Features mapped and interpreted by Kastning (1983), and recent hypotheses on the evolution of the cavernous Edwards Aquifer, suggest a new explanation for the origin of the Natural Bridge Caverns System and nearby caves, as described below.
The concordance of passages with stratigraphic horizons in the Natural Bridge Caverns System, Bracken Bat Cave, and Double Decker Cave has been well documented and suggests strong lithologic control (Kastning, 1978, 1983; Knox, 1981). Correlation of stratigraphic horizons within which major passages of the caves have developed is shown in Figure 3. Extensive passages of Bracken Bat Cave, the Natural Bridge Caverns System, and Double Decker Cave have developed in the Walnut Formation. In Natural Bridge Caverns and Double Decker Cave, passages in the Walnut extend upward in places into the basal beds of the Edwards Limestone. This is the result of collapse of ceilings and upward stoping of passages (see below). These two caves also have large passages developed within the upper member of the Glen Rose Formation. The largest of the passages in the New Braunfels area is the one forming the lowest levels of the Natural Bridge Caverns System along its entire extent. It lies in the interval from 21 to 33 m below the contact between the Walnut and Glen Rose Formations (Figure 3).
Development of large passages in the Glen Rose and Walnut Formations may be attributed to enhancement of porosity following Early Cretaceous deposition, when the San Marcos Arch was elevated, and more than 30 m of the uppermost beds of the Edwards Group were removed by subaerial erosion (Rose, 1972). The accompanying increase in circulation of meteoric groundwater within the lower Edwards Group and underlying Walnut and Glen Rose Formations enlarged available pores.
Evidence for development of cavernous porosity predating development of integrated caves can be seen in Natural Bridge Caverns. Cavities in the upper member of the Glen Rose were enlarged during the early stages of circulation of groundwater and were subsequently filled with calcareous clay and marl. Laminations are easily seen in these deposits. Vugs lined with crystals of calcite occur in the Walnut Formation.
Some beds of the upper member of the Glen Rose Formation are highly porous in comparison with others above and below, and are readily seen in walls of passages on most levels of the caves. Much of this porosity has resulted from leaching of burrows, and by dissolution of evaporates and fossils (moldic porosity). Undoubtedly, circulation of groundwater during the mid-Cretaceous enhanced much of this porosity, but most enlarged cavities were presumably small and poorly integrated in comparison to those that developed later.
The greatest effect of enhanced porosity from the mid-Cretaceous occurred during development of caves following uplift of the Edwards Plateau, Balcones faulting, and regional dissection. Groundwater was transmitted to porous zones through joints created by the faulting and along favorable bedding-plane partings. Zones of primary and enhanced porosity were subsequently integrated into a network of conduits. As a result, passages of caves of this area typically occupy stratigraphic horizons that originally had a high effective porosity.
Some stratigraphic control has also resulted from the mineralogic content of the host-rocks. Dolomitic beds are generally poor formers of caves in the presence of strata of greater calcitic content (Rauch and White, 1970; Kastning, 1975). This is true of the caves of the New Braunfels area, where passages commonly occupy calcareous beds rather than dolomitic beds (Beck, 1968). However, the massive, dolomitic unit comprising the uppermost member of the Glen Rose does contain some passages. Moreover, selectivity of calcareous beds over dolomitic beds is present only among strata of uniform porosity and permeability, because these have been the dominant, litbologic determinants for the vertical positioning of passages.
Each of the three caves studied in detail exhibits linearity in plan- and profile-view, suggesting strong structural control (Figure 2). Detailed mapping of the caves and their geologic structure has shown that fractures such as bedding-plane partings, joints, and faults have guided the orientation and morphology of passages.
A few small passages have developed along bedding-plane partings with relatively little control by joints. Examples include meandering segments of the upper two levels in Natural Bridge Caverns. The widths of these passages commonly exceed their height, and some are characterized by anastomosing crawlways averaging 0.3 to 0.4 m in height.
Horizontal permeability along bedding-plane partings is evident in the caves where movement of groundwater can be observed. Purgatory Creek and River Styx in Natural Bridge Caverns occupy phreatic tubes oriented along such partings, and some speleothems have formed where seepage enters the cave along partings. Some seepage entering along the walls of the cave does so along the top surfaces of shaly beds where water has been perched.
Joints have exerted the greatest structural control on the morphology of passages. Joints are readily identified in ceilings of passages in all three caves. Most linear segments of passages in the Natural Bridge Caverns System have been verified as joints (Figure 2).
Joints measured in the Natural Bridge Caverns System have two major trends: N 20-30 o W and N 40-50 o E (Figure 2). The northeastern joint-set is parallel to the Bat Cave Fault (Figure 1) and is a product of Balcones faulting. The northwestern set is not compatible with tensional stresses operating during normal (step) faulting. However, the strike of this set is nearly parallel to the axis of the San Marcos Arch, and joints of this orientation appear to be related to uplift along the San Marcos Platform during the Cretaceous. The northwestern joint-set accounts for 35 % of the total length of passages in the Natural Bridge Caverns System, whereas the northeastern set represents 15 % of this length. The remaining 50 % of the length has been controlled by shorter joints of other orientations, by faults, or by bedding-plane partings.
Trends of most joints observed in Bracken Bat Cave lie in the intervals N 10-30 o W and N 0-10 o E. Joints of the former set trend parallel to the primary northwestern joint-set of the Natural Bridge Caverns System. Joints of the northeastern set, although present, account for little length of passages in this cave. The lower level of Double Decker Cave is mostly oriented along the dominant, northwestern set of joints. The upper level, however, is oriented along joints trending between N 80o E and S 70o E. The trend of this set is subparallel to the Zuercher Ranch Fault (strike of N 85o E) located just 275 m north of the cave.
*Control by Faults
High-angle faults occur in the South Caverns of the Natural Bridge Caverns System and in Double Decker Cave. Faults in the Natural Bridge Caverns System are parallel to the Bat Cave Fault, and their throws generally measure less than a meter. A single fault in Double Decker Cave strikes parallel to the Zuercher Ranch Fault, and its throw is approximately 0.5 m.
Faults can enhance or inhibit the circulation of groundwater in carbonate aquifers, depending on (1) the type of faulting: normal versus thrust, (2) the type of stresses operating during faulting: tensional versus compressional, (3) the lithologies in juxtaposition across faulted strata, (4) the density of joints related to faults, and (5) the degree of brecciation or shatter, and the thickness of such a zone (Kastning, 1977). Faults may also be of neutral influence. Thus, no general statement on the influence of faults on the flow of groundwater or on speleogenesis is universally valid. Oddly enough, examples of positive, neutral, and negative effects of faults on development of passages occur in the Natural Bridge Caverns System and Double Decker Cave. The southernmost 20o m of the South Caverns alone have been affected in each of these ways (Figure 4).
Box Canyon in the South Caverns has developed along a normal fault with a throw of 0.6 m. The fault is clearly visible in the wall at the southwestern end of this segment of passage where the cave turns to the southeast. Joints of narrow spacing, visible in the ceiling of the Box Canyon, suggest that tensional stress during faulting created a zone of enhanced permeability that became integrated with joints of the prominent northwestern set. The master conduit then developed along this integrated flowpath. The fact that the width of Box Canyon is greater than that of other segments of passages connecting with it may be explained by the high density of joints associated with this fault.
In the Fault Room to the south, two visible faults transect the cave but have no expression in the morphology of the passage (Figure 4). Apparently, the small number of associated joints precluded enhancement of permeability, yet the movement of groundwater across the fault remained uninterrupted because displacements were small (0.3 to 0.6 m) and lithologies on both sides of the faults remained similar.
The main conduit of the Natural Bridge Caverns System ends abruptly at the southern terminus of the South Caverns. This termination is coincident with the Bat Cave Fault, which has a throw of approximately 70 m in this vicinity (Figure 4). Here, beds of the Kainer Formation are in juxtaposition with the upper member of the Glen Rose Formation. The lithologic change across the fault prevented the flow of groundwater across it and to the south. Presumably, flow at this location became diverted to the northeast, in conjunction with the regional movement of groundwater to points of discharge at major springs along the Balcones Escarpment. The Bat Cave Fault is the southeastern boundary of the compartment in which groundwater of the area of the Natural Bridge Caverns System moves. There is no visible conduit leading from the Fault Room, apparently because it lies below the floor of breakdown and sediment.
Flowstone is being deposited where the fault transacts Double Decker Cave at the drop between the Upper and Lower Levels. Here, water from the surface infiltrates into the bedrock along the fault-plane and enters the cave as seepage.
*Correlation of Caves with Surficial Fractures
Fractures were mapped from remotely sensed imagery of the New Braunfels area in order to ascertain how well joints and faults can be interpreted from such imagery, and to determine whether this method is helpful in analyzing karstic landforms and speleogenesis. Available imagery included: (1) low-altitude, black-and-white (panchromatic) prints, (2) a controlled photomosaic, (3) high-altitude, color-infrared transparencies, and (4) black-and-white (Band 5) LANDSAT (ERIS) imagery.
Lineaments are visible where fractured rocks crop out or where they are near enough to the surface to transmit structural patterns through the overburden. Lineaments commonly reflect (1) changes in substrate, regolith, and soil in the vicinity of fractures, (2) patterns in vegetation due to availability of moisture in fractures, (3) differences in moisture within surficial material on opposite sides of a fracture, and (4) changes in the albedo of the surface owing to retention of moisture in fractures. Although most fractures can be discerned on the ground, some are too extensive or subtle to be recognized easily in this way. However, these may be readily detected and mapped from aerial, suborbital, or orbital platforms.
A comparison of the map of fractures with known faults on geologic maps shows that many lineaments parallel the Hidden Valley, Zaccaria Ranch, Zuercher Ranch, Bat Cave, and Hueco Springs Faults of the Balcones Fault Zone. A second prominent set of lineaments trending N 10-40o W represents fractures associated with the San Marcos Arch. Numerous shorter lineaments on the map are orthogonal to fractures of the two prominent sets, and may be regarded as conjugate fractures associated with the Balcones Fault Zone and San Marcos Arch.
Orientations of mapped photolineaments and of linear segments of passages in caves may be compared from rose-diagrams (Figure 5). It is readily seen that the orientations of passages have been largely guided by fractures.
Modification by Collapse
Each of the three caves discussed above exhibits significant modification through the collapse of ceilings of the original conduits. Nearly all of the large chambers of the Natural Bridge Caverns System are floored by mounds of debris from collapse (Figure 2). Zones of collapse are evident in the vicinity of the Entrance Pit and along the Upper Level of Double Decker Cave. The slope within the doline at the entrance to Bracken Bat Cave is floored with debris from collapse, and a mound of breakdown occurs 80 m into the cave.
Presumably, buoyant forces provided by phreatic water were removed as the caves drained, allowing collapse of ceilings previously weakened by joints and solutional enlargement of conduits. Continued collapse promoted upward stoping of passages until mechanical stresses were stabilized in accordance with the present configurations of passages. Stoping occurred along the entire length of the master conduit in the lower levels of the Natural Bridge Caverns System (Figure 2), but large rooms of collapse, termed "breakout-domes," are localized in areas of intense fracturing. Commonly, such rooms have formed where two or more prominent fractures intersect (Figure 2), suggesting that such intersections greatly weaken beds of the ceiling. The Hall of the Mountain Kings is a "textbook example" of a breakout-dome, similar in form to those found in Mammoth Cave in Kentucky (White and White, 1969) and Wyandotte Cave in Indiana (Malott, 1951, p. 33). Stoping in Natural Bridge Caverns has transgressed beds of the Walnut Formation and progressed 6 m into the overlying Kainer Formation. Ceilings are remarkably flat and elliptical, and walls are flared in a mechanically stable configuration.
The most massive collapse and stoping occurred in the vicinity of the entrance to Natural Bridge Caverns. Here, stoping progressed from the Glen Rose Formation, through the overlying Walnut and Kainer Formations, and intersected the surface, forming the doline at the entrance. The Natural Bridge, namesake of the cave, is a remnant that has not yet collapsed. The northern end of the South Caverns, Jaremy Room, and Pluto's Anteroom all lie on the flanks of the large mound of debris remaining from collapse (Figure 2). The doline at the entrance to Bracken Bat Cave has formed in an identical manner.
The volume of debris in mounds within the breakout-domes is generally less than that of the material that has collapsed. This suggests that much of the debris has been removed by dissolution. Waters may have gradually drained from the caves over a considerable period of time so that, as buoyant forces were removed, there was still sufficient circulation of water through the lower parts of the chambers to remove material from collapse. Alternatively, vertical fluctuations of water levels may have been sufficient over a long time to allow collapse and concomitant dissolutional removal of debris.
Collapse has greatly modified the flow of vadose water through conduits of Natural Bridge Caverns. River Styx and Purgatory Creek have been diverted away from the master conduit in areas of collapse, and emerge only in the lowest points of the cave where collapse has been minimal. Water rising from these streams during severe flooding is retained behind dams created by the mounds of debris.
Speleothems in the Natural Bridge Caverns System range in size from massive stalagmites and mounds of flowstone to slender, sodastraw-stalactites. Sodastraws in the Fault Room average 0.5 to 2 m in length and one specimen, suspended from the high ceiling of this room, is over 4.3 m in length, one of the longest known in the world. Sodastraws form comparatively rapidly (as much as a few millimeters in length per year), and these formations are relatively recent features associated with slow, vertical seepage along joints. The massive formations of the Castle of the White Giants, on the other hand, are substantially older and still actively forming.
As with other speleothems in caves of the Edwards Plateau many of those that are no longer active show evidence of redissolution. Redissolved stalagmites are common at all levels of the caves, but are more numerous in the Hall of the Mountain Kings in Natural Bridge Caverns. Some stalagmites have been reshaped into streamlined forms by flowing water. The large stalagmites in the Castle of the White Giants have solutional hollows created by redissolution.
Redissolved formations are found as high in the cave as the top of the Walnut Formation in the Hall of the Mountain Kings. Although redissolution may in part be from periodic flooding in lower levels, it is likely that much of this activity is caused by condensation-solution attributable to movement of humid air through the caves. This phenomenon has recently been recognized in the caves of the Guadalupe Mountains of New Mexico and far west Texas.
The developmental history of the three caves proceeded according to the following chronological sequence:
1. The San Marcos Platform was raised episodically along its northwestern axis throughout the post Albian Cretaceous, producing a system of fractures parallel to its axis. Subaerial exposure of the platform in the late Albian resulted in removal of over 30 m of uppermost beds of the Edwards Group, steepening of hydraulic gradients, and circulation of groundwater through fractures. Flow of groundwater was primarily to the south, and original pores were subsequently enlarged into poorly integrated cavities (Woodruff and Abbott, 1979).
2. During the remainder of the Cretaceous, the region was alternately covered by shallow, marine, shelfal waters and exposed to subaerial conditions. Indications of the latter include (1) absence of the Del Rio Clay just west of New Braunfels (where the Buda Limestone rests unconformably on the Georgetown Formation), (2) absence of the lower members of the Austin Chalk on the San Marcos Arch, and (3) presence of Upper Cretaceous rocks (e.g., Austin Chalk) as fill within dolines in units older than those of the Eagle Ford Group. Although further accentuation of the San Marcos Arch in the Late Cretaceous continued to produce fractures, speleogenesis in the area of the Natural Bridge Caverns System was substantially curtailed under conditions of reduced porosity and in the absence of steep hydraulic gradients in the evolving aquifer.
3. The region was lifted above sealevel near the close of the Cretaceous, and rocks attained a gentle (less than 0.50) southeastward dip. However, an absence of nearby points of discharge precluded development of sizeable solutional conduits.
4. Regional (Balcones) faulting during the Miocene significantly accelerated the geomorphic evolution of drainage on the surface and in the subsurface. Northeast-trending fractures produced during faulting were overprinted on the pre-existing northwest-oriented set of fractures. The Balcones Escarpment established a low baselevel to the southeast, toward which surficial and subsurficial geomorphic processes began to operate.
5. Flow of groundwater increased in response to steepened hydraulic gradients created by drainage to springs at baselevel along the Balcones Escarpment. Some faults, such as the Bat Cave Fault, imposed impermeable boundaries on the aquifer, diverting groundwater along the strike of the faults. Flow along fractures and bedding-plane partings within the phreatic zone became integrated into master conduits, producing some of the larger passages of caves. These include the main passage along the lower level of the Natural Bridge Caverns System, the large passage of Bracken Bat Cave, and the Upper and Lower Levels of Double Decker Cave. Some conduits followed beds of favorable primary and solutionally enhanced porosity, and became oriented down the dip, in the direction of greatest hydraulic gradient. However, other passages cut across previous solutional cavities. Most conduits were strongly guided by fractures of the two primary sets. Formed by tensional stresses during uplift of the San Marcos Arch and faulting along the Balcones Fault Zone, these fractures were sufficiently open to easily accommodate phreatic flow. In places, dip-oriented conduits at different levels were integrated through vertical chimneys along fractures, allowing flow to move continuously through a single solutional channel from one stratigraphic horizon to the next. This situation is exemplified by Double Decker Cave where flow was communicated from one level to the other along a major fault.
6. As conduits enlarged, their widths became too great to support overlying beds and collapse of ceilings began. This may have been initiated as water first began to drain from the caves, removing buoyant support of the ceilings.
7. The potentiometric surface in the aquifer began to drop as surficial streams became entrenched along the escarpment and as a consequence of the capture of streams (Woodruff and Abbott, 1979). The caves began to drain as the dropping potentiometric surface intersected the conduits. Mounds of debris in the caves were partially removed by vadose waters still flowing in the conduits. Areas of collapse became mechanically stabilized.
8. Speleothem deposition began.
9. Increased precipitation and decreased evaporation during pluvial climates of the Pleistocene resulted in high discharges in the existing caves. Collapse and alluviation within the caves prevented efficient throughflow of groundwater. As a consequence, caves underwent periodic flooding, which locally promoted development of routes of bypass for floodwater around mounds of debris previously created by collapse. Smaller, middle- and upper-level passages in the Natural Bridge Caverns System developed as routes of bypass in response to blockages. Pre-existing conduits were also enlarged by floodwaters.
10. Vadose flow in the caves diminished to previous levels with the advent of warmer climates. This flow apparently represents underflow derived from recharge along Cibolo Creek. Under conditions of baseflow, discharge is carried through inaccessible conduits beneath the explored levels of Natural Bridge Caverns.
Abbott, P. L., 1973, The Edwards Limestone in the Balcones Fault Zone, south-central Texas: The University of Texas at Austin, Ph.D. dissertation, 122 p.
Abbott, P. L., 1975, On the hydrology of the Edwards Limestone, south-central Texas: Journal of Hydrology, v. 24, p. 251-269.
Abbott, P. L., 1977a, Effect of Balcones faults on groundwater movement, south central Texas: Texas Journal of Science, v. 29, p. 5-14.
Abbott, P. L., 1977b, On the state of saturation of groundwater with respect to dissolved carbonates, Edwards artesian aquifer, south-central Texas: Texas Journal of Science, v. 29, p. 159-167.
Baker, V. R., 1975, Flood hazards along the Balcones escarpment in central Texas: Alternative approaches to their recognition, mapping, and management: The University of Texas at Austin, Bureau of Economic Geology Geological Circular 75-5, 22 p.
Baker, V. R., 1977, Stream-channel response to floods, with examples from central Texas: Geological Society of America Bulletin, v. 88, p. 1057-1071.
Barnes, V. E., 1974, San Antonio sheet: The University of Texas at Austin, Bureau of Economic Geology, Geologic Atlas of Texas, scale 1:250,000.
Beck, B. F., 1968, Speleogenesis in Comal County, Texas: Rice University, Master's thesis, 44 p.
Beck, B. F., 1978, Color differentiation in "fried egg" stalagmites: Journal of Sedimentary Petrology, v. 48, p. 821-824.
Colwick, A. B., McGill, H. N., and Erichsen, F. P., 1973, Severe floods at New Braunfels, Texas: American Society of Agricultural Engineers, Paper 73-206, 8 p.
George, W. O., Breeding, S. D., and Hastings, W. W., 1952, Geology and ground-water resources of Comal County Texas, with sections on surfacewater runoff and chemical character of the water: U.S. Geological Survey Water-Supply Paper 1138, 126 p.
Guyton, W. F. and Associates, 1979, Geohydrology of Comal, San Marcos, and Hueco Springs: Texas Department of Water Resources Report 234, 85 p.
Heidemann, C. W., 1979, Natural Bridge Caverns (abstract): National Speleological Society Bulletin, v. 41, p. 118-119.
Kastning, E. H., 1975, Cavern development in the Helderberg Plateau, east-central New York: New York Cave Survey Bulletin 1, 194 p.
Kastning, E. H., 1977, Faults as positive and negative influences on ground-water flow and conduit enlargement, in Dilamarter, R. R. and Csallany, S. C., eds., Hydrologic Problems in Karst Regions: Bowling Green, Western Kentucky University, p. 193-201.
Kastning, E. H., 1978, Caves and karst hydrogeology of the southeastern Edwards Plateau, central Texas: National Speleological Society Guidebook Series 19A, 46 p.
Kastning, E. H., 1980, Structural, lithologic, and topographic controls on the origin of Natural Bridge Caverns, Comal County, Texas (abstract): National Speleological Society Bulletin, v. 42, p. 32.
Kastning, E. H., 1981, Tectonism, fractures, and speleogenesis in the Edwards Plateau, central Texas, U.S.A., in Beck, B. F., ed., Proceedings of the Eighth International Congress of Speleology, Bowling Green, Kentucky, July 18 to 24, 1981: National Speleological Society, Huntsville, Alabama, v. 2, p. 692-695.
Kastning, E. H., 1983, Geomorphology and hydrogeology of the Edwards Plateau karst, central Texas: The University of Texas at Austin, Ph.D. dissertation, 656 p.
Kastning, E. H., 1984, Hydrogeomorphic evolution of karsted plateaus in response to regional tectonism, in LaFleur, R. G., ed., Groundwater as a geomorphic agent: Proceedings of the Thirteenth Annual Geomorphology Symposium, Troy, New York: London, George Allen and Unwin, p. 351-382.
Knox, J., 1975, Solution features of upper Cibolo Creek basin (abstract): Geological Society of America Abstracts with Programs, v. 7, p. 180.
Knox, J., 1981, Natural Bridge Caverns: Texas Caver, v. 26, p. 84-87.
Knox, O., 1962, Natural Bridge Caverns claims title: Texas Caver, v. 7, p. 107, 109-113.
Malott, C. A., 1951, Wyandotte Cavern: National Speleological Society Bulletin, v. 13, p. 30-35.
Moore, C. H., 1961, Stratigraphy of the Walnut Formation, south-central Texas: Texas Journal of Science, v. 13, p. 17-40.
Moore, C. H., 1964, Stratigraphy of the Fredericksburg Division, south-central Texas; The University of Texas at Austin, Bureau of Economic Geology Report of Investigations 52, 48 p.
Newcomb, J. H., 1971, Geology of Bat Cave quadrangle, Comal and Bexar Counties, Texas: The University of Texas at Austin, Master's thesis, 104 p.
Rauch, H. W. and White, W. B., 1970, Lithologic controls on the development of solution porosity in carbonate aquifers: Water Resources Research, v. 6, p. 1175-1192.
Reddell, J. R., ed., 1964a, The caves of Comal County: Texas Speleological Survey, v. 2, no. 2, 60 p.
Reddell, J. R., ed., 1964b, A guide to the caves of Texas: National Speleological Society Guidebook Series 5, 63 p.
Rose, P. R., 1972, Edwards Group, surface and subsurface, central Texas: The University of Texas at Austin, Bureau of Economic Geology Report of Investigations 74, 198 p.
Stricklin, F. L., Jr., Smith, C. I., and Lozo, F. E., 1971, Stratigraphy of Lower Cretaceous Trinity deposits of central Texas: The University of Texas at Austin, Bureau of Economic Geology Report of Investigations 71, 63 p.
White, E. L., and White, W. B., 1969, Processes of cavern breakdown: National Speleological Society Bulletin, v. 31, p. 83-96.
White, P. J., 1948, Caves of central Texas: National Speleological Society Bulletin, v. 10, p. 46-64.
Woodruff, C. M., Jr., and Abbott, P. L., 1979, Drainage-basin evolution and aquifer development in a karstic limestone terrain, south-central Texas, U.S.A.: Earth Surface Processes, v. 4, p. 319-334.
Abbott, Patrick L. and Woodruff, C.M., Jr., eds., 1986. The Balcones Escarpment, Central Texas: Geological Society f America. P. 91 -100