Citation
Journal of cave and karst studies

Material Information

Title:
Journal of cave and karst studies
Series Title:
Journal of Cave & Karst Studies
Alternate Title:
Continues NSS bulletin (OCLC: 2087737)
Creator:
National Speleological Society
Publisher:
National Speleological Society
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
Geology ( local )
Genre:
serial ( sobekcm )
Location:
United States

Notes

Restriction:
Open Access - Permission by Publisher
Original Version:
Vol. 66, no. 2 (2004)

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
K26-02038 ( USFLDC DOI )
k26.2038 ( USFLDC Handle )
6744 ( karstportal - original NodeID )
0146-9517 ( ISSN )

USFLDC Membership

Aggregations:
Karst Information Portal

Postcard Information

Format:
Serial

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

46• Journal of Cave and Karst Studies August 2004 Darlene M. Anthony and Darryl E. Granger ALate Tertiary origin for multilevel caves along the western escarpment of the Cumbe rland Plateau, Tennessee and Kentucky, established by cosmogenic 26Al and 10Be. Journal of Cave and Karst Studies v. 66, no. 2, p. 46-55. In the southeastern United States, karst features are developed on two topographic surfaces of regional extent known locally as the Cumberland Plateau and Highland Rim (Fig. 1). The most extensive and elevated of the two is the Cumberland Plateau, a rugged upland (550–610 m ASL) bounded on the east by the Valley and Ridge Province and on the west by the solutional surface (275–350 m ASL) of the Eastern Highland Rim. Nearly 180 million years of differential lowering between the sandstone-capped Cumberland Plateau and the limestone surface of the Highland Rim has formed a highlydissected, eastward-retreating escarpment along the western margin of the Cumberland Plateau. The lithologic change from sandstone to limestone along the western escarpment provides an optimum hydrogeologic setting for cave development. Crawford (1984) was the first to describe the “plateau-margin” model of cave development (Fig. 2). In this model, surface streams undersaturated with respect to calcite cross the sandstone caprock of the Cumberland Plateau, sink at the contact between sandstone and limestone, and form cave passage in the vadose zone leading to the local water table. Cave streams emerge as springs along the base of the escarpment or valley wall. Morphometric characteristics of plateau-margin caves include small passage dimensions (in terms of surveyed length and cross-sectional area) and a vertically developed profile (Fig. 3). Of thousands of caves explored along the western escarpment of the Cumberland Plateau, a few do not fit the physical ALATE TERTIARYORIGIN FOR MULTILEVELCAVES ALONGTHE WESTERN ESCARPMENTOFTHE CUMBERLAND PLATEAU, TENNESSEE AND KENTUCKY, ESTABLISHED BYCOSMOGENIC 26ALAND 10BEDARLENEM. ANTHONYANDDARRYLE. GRANGER Department of Earth and Atmospheric Sciences, Purdue University, West Lafayette, IN 47907 anthondm@purdue.edu; dgranger@purdue.edu Cosmogenic burial dating of quartzose cave sediments deposited in multilevel caves beneath the western margin of the Cumberland Plateau dates ~5.7 Ma of cave development in step with episodic incision of the Upper Cumberland River. These particular cave systems are characterized by hydrologically abandoned, low-gradient passages concentrated at common levels above the modern water table. Previous studies recognized morphometric differences between the majority of smaller, hydrologically active “plateau-margin” caves and large, abandoned “fossil” or “Cumberland-style” caves. This study links the origin of multilevel caves on the western margin to a prolonged period of Late Tertiary water table stability, and the development of levels to distinct episodes of Plio-Pleistocene river incision. In this study, clastic sediments in multilevel cave passages are dated using cosmogenic 26Al and 10Be, and are shown to correspond with 1) deposition of upland (“Lafayette-type”) gravels between ~3.5 Ma and ~5 Ma; 2) initial incision of the Cumberland River into the Highland Rim after ~3.5 Ma; 3) development of the Parker strath between ~3.5 Ma and ~2 Ma; 4) incision of the Parker strath at ~2 Ma; 5) shorter cycles of incision after ~1.3 Ma associated with terraces above the modern flood plain; and 6) regional aggradation at ~0.8 Ma. Burial ages of cave sediments record more than five million years of incision history within the unglaciated Appalachian Plateaus and constrain the developmental history of multilevel caves associated with the Upper Cumberland River. Figure 1. The study area in Kentucky and Tennessee (A) with Mammoth Cave (MC) on the Green River, KY. Aportion of UpperCumberland Riverbasin (B) drains the study area, and includes twelve caves on the western margin of the Cumberland Plateau. CC-Cumberland Caverns; BNBone Cave; FH-Foxhole Cave; BS-Blue Spring Cave; SNSkagnasty; LD-Lott Dean (Mountain’s Eye System); ZZarathustra’s Cave; X-Xanadu Cave; BU-Buffalo Cave; WR-Wolf RiverCave; SV-Sloan’s Valley Cave; GS-Great Saltpetre Cave.

PAGE 2

Journal of Cave and Karst Studies August 2004 • 47 ANTHONYANDGRANGER characteristics of the plateau-margin model, although they clearly developed in the same hydrogeologic setting. These were named “fossil caves” (Mann 1982) and later, “Cumberland-style” caves (Sasowsky 1992). Physical attributes of “fossil caves” included large, hydrologically abandoned passages of phreatic origin (Fig. 4). Recharge from the plateau combined with backflooding from surface discharge springs was speculated to produce high hydrostatic pressure in phreatic conduits, which led to the development of large passages under pipe-full conditions (Mann 1982). In a later study, large caves on the western margin were named “Cumberland-style” by geographic association with the highly dissected western margin of the Cumberland Plateau (Sasowsky & White 1994). Characteristic features were similar to those of the “fossil caves,” including abandoned trunk passages concentrated at one or more levels above the modern river level. However, this model linked passage morphology with a different type of speleogenesis. In the Cumberland-style model, large trunk passages were observed to generally follow topographic contours parallel to a surface valley containing a master stream. Subsurface diversion of the master stream is an important constraint for this model, and large caves are hypothesized to be the result of this diversion (Sasowsky et al 1995). The Cumberland-style model attributes large, horizontal passages to high discharge. Both the modified plateau-margin model and the Cumberland-style model require that large, low-gradient horizontal passages form under high discharge conditions. Following either model, abandoned trunk passages could have formed at any time during the past, given the right hydrologic conditions. An alternative hypothesis is that large, multilevel caves on the western Cumberland Plateau escarpment developed synchronously during long periods of river stability. These long periods of time provide an opportunity for modest discharge to dissolve exceptionally large trunk passages. Figure 2. Schematic plateau-margin model of cave development (afterCrawford, 1984). Surface streams originating on sandstone bedrock of the Cumberland Plateau flow down the escarpment and sink at the contact between sandstone and limestone. Sinking streams form cave passages in the vadose zone leading to the local watertable, and emerge as springs along the base of the valley wall. Figure 3. Passages in plateau-margin caves are typically narrow, vertical canyons leading down to the modern watertable. Figure 4. The large, hydrologically abandoned Ten Acre Room in Cumberland Caverns, TN is a passage of phreatic origin above the modern watertable. These passages are referred to as “fossil” caves or“Cumberland-style” caves in the literature. (Photo Bob Biddix.)

PAGE 3

48• Journal of Cave and Karst Studies August 2004 ALATETERTIARYORIGINFORMULTILEVELCAVES Because solution kinetics ultimately control the enlargement rate of conduits, the maximum diameter of a phreatic tube will depend on the length of time the passage is filled with undersaturated water (White 1977; Palmer 1991). Under the baselevel stability model, large multilevel caves on the western margin are related to each other temporally because they all drain to a water table controlled by the elevation of the Cumberland River. To test this hypothesis, we examined cave morphology and sediment structures, and dated sediments in twelve large multilevel caves on the western margin using cosmogenic 26Al and 10Be. BURIALDATINGUSINGCOSMOGENICNUCLIDES. The ability of accelerator mass spectrometry (AMS) to measure small amounts of radionuclides has led to a new way of dating cave sediments up to five million years old (Granger & Muzikar 2001; Muzikar et al 2003). This dating method involves cosmogenic nuclides produced in rocks near the ground surface by cosmic rays (Lal & Peters 1967). The cosmogenic radionuclides aluminum-26 (26Al) and beryllium-10 (10Be) are produced in quartz crystals by reactions with secondary cosmic ray neutrons, which change silicon atoms to26Al and oxygen atoms to 10Be in an approximate 6:1 ratio. Together, these two radionuclides can be used to date when a quartz crystal was carried into a cave. Quartz sediments originating on the Cumberland Plateau caprock are first exposed to cosmic rays, accumulate 26Al and10Be, and are transported underground as part of the bedload of cave streams in the study area. Once underground, the quartz is shielded from further exposure to cosmic radiation by tens of meters of rock. After burial, concentrations of accumulated26Al and 10Be diminish over time due to radioactive decay, with26Al decaying roughly twice as fast as 10Be. The present-day ratio of remaining cosmogenic nuclides yields a burial age for the sediment. DATAANALYSISANDUNCERTAINTIES. Burial ages are determined by iterative solution of equations for measured and inherited concentrations of nuclides (after Granger et al 1997). Accumulation of cosmogenic nuclides for the simple case of a steadily eroding outcrop is described by Equation (1), where the preburial 26Al/10Be ratio ( N26/ N10)0changes with erosion rate as follows: assumed constant for the region and were calculated as P10= 5.22 at g–1a–1and P26= 35.4 at g–1a–1for a latitude of 36 and an elevation of 0.5 km (Stone 2000, modified for a 10Be meanlife of 1.93 m.y.). After shielding from nuclide production by burial underground, the cosmogenic radionuclide production stops, and26Al and 10Be decays according to: N26= ( N26)0e– t / 26(2) and N10= ( N10)0e– t / 10where t is burial time. Because 26Al decays faster than 10Be, the ratio N26/ N10decreases exponentially over time according to: (1) where P26and P10are the production rates of 26Al and 10Be, is the penetration length for neutrons ( 60 cm in rock of density 2.6 g cm-3), 26= 1.02 0.04 m.y. is the radioactive26Al meanlife, and 10= 1.93 0.09 m.y. is the radioactive 10Be meanlife. Local cosmogenic nuclide production rates were where N26and N10are the concentrations of 26Al and 10Be measured by AMS. Equations 1–3 solve for converging solutions of E ( N26/ N10)0, and t after a few iterations (Granger et al 1997). Burial age is reported with two uncertainties; the first is one standard error of analytical uncertainty. The second includes systematic uncertainties in radioactive decay rates, P26/ P10, and production rates, which are added in quadrature and shown as total uncertainties in parentheses. Analytical uncertainties are used when comparing burial ages with each other. Total uncertainties are used when comparing burial ages with other dating methods. METHODSSAMPLESITES. Twelve caves in the Upper Cumberland River basin (Fig. 1) were selected for this study based on: 1) one or more abandoned levels of large cross-sectional area connected by narrow canyons; 2) extensive horizontal development; 3) in-place channel deposits (Fig. 5) with no sediment remobilization from upper levels or surface. [Five caves previously identified as “Cumberland-style” included Xanadu Cave, Zarathustra’s Cave, Mountain’s Eye (Lott Dean), Bone Cave, and Cumberland Caverns (Sasowsky 1992).] Some caves are fragments beneath plateau outliers, with no connection to the modern water table due to escarpment retreat and loss of recharge area. Others have remained connected to their recharge area, and have an active base level conduit today. Extensive horizontal cave passages were grouped by similar heights above the modern river level. The assumption was made that the modern river longitudinal profile was not different from the paleoprofile; therefore caves would develop at similar heights (White & White 1983). Passages were correlated with fluvial (3)

PAGE 4

Journal of Cave and Karst Studies August 2004 • 49 ANTHONYANDGRANGER surface features in the Upper Cumberland River valley, that included the Eastern Highland Rim, the Parker strath, and terraces in the Upper Cumberland River basin (Table 1). Target materials for 26Al and 10Be isotopic measurements were rounded quartz pebbles (Figure 6) and sand weathered from the Rockcastle Conglomerate caprock and deposited in (now) hydrologically abandoned cave passages. Approximately 500 grams of quartz pebbles or one kilogram of cross-bedded sand were collected at each sampling site. COSMOGENICNUCLIDECHEMISTRY. Quartz from each sample site (~120 g) was purified by chemical dissolution (Kohl & Nishiizumi 1992), dissolved in HF and HNO3, and spiked with ~0.7 mg 9Be in a carrier solution. Fluorides were driven out with H2SO4. Aluminum and beryllium were separated and purified by ion chromatography, selectively precipitated as hydroxides, and oxidized at 1100C. AMS measurements of 10Be/9Be and 26Al/27Al isotope ratios were made at the Purdue Rare Isotope Measurement Laboratory (PRIME Lab) and the Lawrence Livermore National Laboratory, CA. RESULTSANDINTERPRETATIONCosmogenic burial dating of sediments shows that caves on the western escarpment were an active part of the regional hydrology in the Late Miocene and throughout the Pliocene (Table 1). The oldest sediments in the study area are found in caves beneath plateau outliers and heavily dissected margins of the Cumberland Plateau, and have no active base level today. Progressively younger burial ages are found in passages at elevations that maintain the modern river profile along two major Cumberland River tributaries. Awidespread, regional aggradation signal occurs in the lowermost levels of multilevel caves across the entire basin. Each of these events is discussed in detail below. 1. Abandonment of Bone Cave at 5.68 1.09 (1.21) Ma. Bone Cave is located beneath Bone Cave Mountain, an elongate spur almost completely separated from the western margin of the Cumberland Plateau. Bone Cave has no physical connection with the modern water table. Stream-deposited quartz pebbles from the main passage of Bone Cave (Muster Ground) yield a burial age of 5.68 1.09 (1.21) Ma, with a large uncertainty due to the very small amount of remaining cosmogenic 26Al. Asmall, discontinuous phreatic level beneath the Muster Ground indicates that incision to a lower level was underway when the cave stream was cut off from its recharge source. The burial age shows that the Muster Ground in Bone Cave carried sediments at a water table nearly 90 m above the modern river level during the Late Miocene, and was abandoned at ~5.7 Ma. Aloss of recharge by surface stream piracy may have caused passage abandonment. 2. Aggradation and abandonment of Cumberland Caverns at 3.52 0.42 (0.49) Ma. Cumberland Caverns lies beneath Cardwell Mountain, a remnant outlier of the Cumberland Plateau separated by a distance of 2.4 km from the retreating edge of the western escarpment. Passages in Cumberland Caverns have no physical connection with the Figure 5. Graded sediments and cut-and-fill structures in the MusterGround of Bone Cave, TN indicate open-channel flow. Waterbottle forscale. Figure 6. Quartz pebbles weathered from the plateau caprock are easily identified in cave sediments and collected forcosmogenic nuclide measurements.

PAGE 5

50• Journal of Cave and Karst Studies August 2004 ALATETERTIARYORIGINFORMULTILEVELCAVES modern water table. Observations of vadose/phreatic transitions in the cave suggest major phreatic development ended with the abandonment of the main passage (Ten Acre Room) via a smaller phreatic passage (Dish Pan Alley). Aggradation after the development of Dish Pan Alley filled Dish Pan Alley to the top of the Volcano Room with over 10 m of sediment, which has been partially removed by minor stream activity. Samples from the top and bottom of the sand fill yielded an average age of 3.52 0.42 (0.49) Ma (weighted by inverse variance), which is interpreted as the time of separation of Cardwell Mountain from the Cumberland Plateau (see Barr, 1961 for discussion) and loss of recharge area for phreatic development of Cumberland Caverns. The Ten Acre Room is inferred to be older than ~3.5 Ma. 3. Abandonment of caves along the Caney ForkCalfkillerRivers. Cave passages concentrated between 40 m and 55 m above the modern river level of the Caney Fork and Calfkiller River (Fig. 1) contain graded stream deposits of quartz pebbles, sandstone gravel, and sand. Burial ages for cave sediments are oldest in caves closest to the Cumberland River, and become progressively younger upstream (Fig. 7). Foxhole Cave (B-survey) was abandoned at 1.97 0.10 (0.17) Ma; Blue Spring Cave (Ship’s Prow) at 1.66 0.23 (0.28) Ma; and Skagnasty Cave (A-survey) at 0.89 0.21 (0.22) Ma. These data suggest that caves on the Caney Fork-Calfkiller River were abandoned in sequence as a knickpoint, or waterfall, migrated upstream. Knickpoint migration would have been initiated by incision of the Cumberland River prior to ~2 million years ago. 4. Abandonment of caves along the Obey River. Wolf River and the East Fork-Obey River (East Fork) are branches of the Obey River, a major tributary of the Cumberland River (Fig. 1). Cave passages concentrated between 40 m and 55 m above the modern river level contain graded deposits of quartz pebbles, sandstone gravel, and cross-bedded sands. Burial ages of sediments in these passages show that Wolf River Cave (Upper Borehole) was abandoned at 2.15 0.47 (0.52) Ma; Buffalo Cave (Saltpetre Passage) at 1.45 0.42 (0.45) Ma; Xanadu Cave (Cumberland Avenue) at 1.64 0.46 (0.48) Ma; and Zarathustra’s Cave (Heaven) at 1.80 0.31 (0.36). There are no significant differences in ages between caves on the East Fork-Obey River, which is not surprising due to the caves’close proximity to each other. Data from the Obey River watershed are consistent with migration of a knickpoint initiated by incision of the Cumberland River prior to ~2 Ma (Fig. 7). Synchronous abandonment of Blue Spring Cave (Ship’s Prow) and Xanadu Cave (Cumberland Avenue) suggests the same incision episode on the Cumberland River is responsible for initiating knickpoints on the tributaries. Table 1. Cosmogenic nuclide concentrations, burial ages, and correlated surface features from Cumberland Plateau caves.Cave and passageElevation above Surface Sample [26Al] [10Be] [26Al]/[10Be]burial ageanamemodern rivers (m)featuretype(106at/g)(106at/g)(Ma) Bone91Highland Rimpebbles0.017 0.0120.017 0.0120.46 0.325.68 1.09 (1.21) (Muster Ground) Cumberland66Highland Rimsand0.158 0.0420.158 0.0421.39 0.373.52 0.42 (0.49) (Volcano Room) Foxhole43Parker strathpebbles0.308 0.0220.308 0.0222.53 0.231.97 0.10 (0.17) (B-survey) Blue Spring49Parker strathpebbles0.380 0.0380.380 0.0383.07 0.381.66 0.23 (0.28) (Ship’s Prow) Skagnasty45Parker strathpebbles0.334 0.0260.334 0.0264.61 0.680.89 0.21 (0.22) (A-survey) Wolf River43Parker strathpebbles0.189 0.0770.189 0.0772.46 0.622.15 0.47 (0.52) (Upper Borehole) Buffalo48Parker strathsand1.127 0.2641.127 0.2643.26 0.771.45 0.42 (0.45) (Main Saltpetre) Xanadu54Parker strathsand1.036 0.1341.036 0.1343.66 0.481.23 0.24 (0.27) (Steven’s Ave.) Xanadu52Parker strathpebbles0.208 0.0260.208 0.0263.13 0.761.64 0.46 (0.48) (Cumberland Ave.) Zarathustra’s40Parker strathsand1.278 0.2281.278 0.2282.65 0.481.80 0.31 (0.36) (Heaven)bXanadu42first terracesand0.763 0.1490.763 0.1494.46 0.880.85 0.37 (0.38) (Sand Hills) Sloan’s Valley48first terracesand1.218 0.2021.218 0.2024.32 0.730.89 0.31 (0.33) (Appalachian Trail)cGreat Saltpetre31first terracesand1.227 0.0621.227 0.0624.31 0.660.95 0.29 (0.31) (Dressing Room) Zarathustra’s28first terracepebbles0.580 0.0530.580 0.0534.47 0.420.86 0.17 (0.19) (Elephant Walk) Zarathustra’s13lower terracespebbles0.899 0.1010.899 0.1014.50 0.520.83 0.21 (0.22) (B-survey) Lott Dean0modern riverpebbles1.416 0.1071.416 0.1076.60 0.540.02 0.13 (0.13) (upstream sump)aUncertainties represent one standard error measurement uncertainty. Systematic uncertainties in production rates (20%), product ion rate ratio (Stone, 2000) and radioactive decay constants are added in quadrature and shown as total uncertainty in parentheses.bHighest passage of three levels in this system.cPassage developed less than 1 km from mainstem Cumberland River.

PAGE 6

Journal of Cave and Karst Studies August 2004 • 51 ANTHONYANDGRANGER 5. Regional aggradation of lowerlevels at ~0.85 Ma. Sediments collected in levels beneath those discussed above indicate widespread aggradation (Fig. 8). Passages that record this event include Xanadu Cave (Sand Hills) at 0.85 0.37 (0.38) Ma; Zarathustra’s Cave (B-survey) at 0.83 0.21 (0.22) Ma; Zarathustra’s Cave (Elephant Walkway) at 0.86 0.17 (0.19); Sloan’s Valley Cave (Appalachian Trail) at 0.89 0.31 (0.33) Ma; and Great Saltpetre Cave (Dressing Room) at 0.95 0.29 (0.31) Ma. These data suggest a widespread regional aggradation event filled one or more of the lower cave levels, overprinting sediment deposited during passage development. 6. Measurement of sediment in active base level passages. The Lott Dean section of the Mountain’s Eye System is an active base level conduit for subsurface drainage of the East Fork-Obey River. Lott Dean is a modern analog for abandonment in progress; a small phreatic tube beneath the floor of the main conduit (Fig. 9) carries the base flow component of the karst aquifer, with the main conduit carrying overflow from storm events (see Hess & White 1989 for discussion of base flow in karst aquifers). Measurements of cosmogenic nuclides from quartz pebbles collected in the overflow conduit yield an age of 0.02 0.13 (0.13) Ma, indistinguishable from a zero burial age found in pebbles on the surface. This confirms that base level conduits carry sediment from the surface, an important assumption in the interpretation of cave sediment burial age. DISCUSSIONIn general, horizontal cave passages in this region form by active solution at a stable water table, and multilevel caves form due to episodic lowering of the local water table in response to changes in the regional base level (White & White 1970; Palmer 1987, 1991). The shape and configuration of multilevel caves on the western margin of the Cumberland Plateau reflect this type of episodic water table lowering, and suggest a common history linked to the changing position of Figure 7. Schematic diagram showing knickpoint migration on Caney Fork-CalfkillerRiverand Obey River-Wolf River. Incision pulses originating on the Cumberland River at t1>2 Ma migrated up the Caney Fork and Obey-Wolf River, lowering the local watertable and abandoning Foxhole Cave and Wolf RiverCave at t2 2 Ma. At t3 1.6 Ma the pulse had migrated up the CalfkillerRiverand East Fork-Obey River, abandoning Blue Spring Cave and the caves in the East Fork. Skagnasty Cave was abandoned at t4 0.9 Ma. Figure 8. Aregional aggradation signal at ~0.8 Ma is found throughout the study area, including this passage in Xanadu Cave, TN (Sand Hills Passage). (Photo Sean Roberts.) Figure 9. The Lott Dean passage of the Mountain’s Eye System, TN is a modern analog forabandonment in progress. Phreatic development beneath the main conduit transmits base flow forthe karst aquifer, and the main conduit carries overflow during storm events. (Photo Brian A. Smith.)

PAGE 7

52• Journal of Cave and Karst Studies August 2004 ALATETERTIARYORIGINFORMULTILEVELCAVES the Cumberland River and its tributaries. Dating sediments in different cave levels can help to firmly establish this history, and constrain the time needed to form large passages. THEASSOCIATIONOFMULTILEVELCAVESWITHLANDSCAPE EVOLUTIONThese caves can be related to features long recognized on the surface. Rivers produce wide straths and alluvial terraces during periods of base level stability, with entrenchment indicative of sudden change in the rate of incision (Fenneman 1938). Widespread fluvial gravels (“Lafayette-type”) scattered across the surface of the Eastern Highland Rim (Potter 1955) are evidence of a wandering, low-gradient Cumberland River prior to initial incision into its present valley (Fenneman 1938; Thornbury 1965). Following incision, a period of stability resulted in development of a wide valley called the Parker strath 65 m beneath the Highland Rim (Butts 1904; Wilson 1948). Discontinuous terraces at 10–15 m intervals beneath the Parker strath represent shorter episodes of incision (McFarlan 1943; Miotke & Palmer 1972). However, determining the exact timing of episodic incision was difficult in the past due to a combination of unsuitable dating methods and poorly preserved surface materials. The development of large cave passages along the western margin may be correlated with periods of base level stability, and their abandonment with incision of the Cumberland River. Large passages in Bone Cave and Cumberland Caverns were moving sediment at least three to five million years ago at a water table controlled by the Cumberland River as it flowed on top of the Eastern Highland Rim. These data constrain initial incision of the Cumberland River into the Highland Rim to a time after ~3.5 Ma. Cave passages along the Caney ForkCalfkiller River and the Obey River were fully developed in cross-sectional area when abandoned by knickpoint migration initiated by the Cumberland River at least two million years ago. These passages formed simultaneously with the Parker strath during a period of base level stability. We suggest these passages developed over a period of ~1.5 m.y. between initial incision into the Highland Rim and incision of the Parker strath, with limited areas of recharge from the Cumberland Plateau. MODERNPHREATICPASSAGESONTHEWESTERNMARGINTwo large, active, base level conduits that drain 172 km and 260 km of the Cumberland Plateau cannot be explained by a long period of base level stability. Presently, the Mountain’s Eye System (Lott Dean) and Blue Spring Cave (Fig. 10), drain the largest recharge areas of the Cumberland Plateau. Climate over the past two million years has changed rapidly and repeatedly as ice sheets grew and receded in North America. Although the Cumberland River was south of the farthest ice extent, it has nonetheless alternately aggraded and incised, raising and lowering the local water table along the Cumberland Plateau margin. Large base level caves that form today thus require large discharges, because the Cumberland River has not maintained a stable position over the past two million years. In contrast, a relatively stable climate in the Late Tertiary resulted in long-term river stability, so large caves could develop from small recharge areas over millions of years. COMPARISONWITHOTHERWORKThe modified plateau-margin model. Our interpretation of speleogenesis differs from that of Mann (1982). The modified plateau-margin model included high hydrostatic pressure in the conduit. We observed in-situ fluvial deposits with cutand-fill features, cross-stratification, and imbricated sediments in several of the named “fossil caves,” which indicate open channel conditions during deposition of the sediment. Sediments in Bone Cave also display several cycles of graded sediments ranging in size from subrounded pebbles 1–2 cm in diameter to flood clays, which indicated periodic flooding of the conduit. We do not think these caves operated under continuous pipe-full conditions. The Cumberland-style model. Our interpretations differ from those of Sasowsky and White (1994) and Sasowsky et al (1995), who relied on paleomagnetic dating of sediments in “Cumberland-style” caves. Paleomagnetic dating of clastic sediments in cave passages involves the construction of a local Figure 10. The Second RiverCrossing in Blue Spring Cave, TN. This large phreatic passage at the modern riverlevel drains roughly 260 km of the Cumberland Plateau. (Photo Bernard Szukalski.)

PAGE 8

Journal of Cave and Karst Studies August 2004 • 53 ANTHONYANDGRANGER magnetostratigraphic column based on the orientation of magnetic grains in fine sediments, and subsequent comparison with the global paleomagnetic record. In the absence of absolute dating means, sediments in caves are analyzed to establish normal or reversed magnetic sequences, the latter implying a minimum of 0.78 Ma in age (Cande & Kent 1995). Paleomagnetic dating of sediments in Xanadu Cave’s Cumberland Avenue (Fig. 11) revealed one reversed-over-normal polarity transition moving stratigraphically upwards, which was interpreted as the younger end of the Jaramillo event at 0.91 Ma (Sasowsky et al 1995). A“missed” reversal in sediments deposited in lower levels of Xanadu Cave would have placed this transition at the younger end of the Olduvai event (1.66 Ma) but was not considered likely by the authors, as sediments in lower levels were of normal polarity. We report a burial age of 1.64 0.46 (0.48) Ma from cosmogenic nuclides in sediments from the same location in Cumberland Avenue, which places the reversed-over-normal sequence at the younger end of the Olduvai Event (1.66 Ma). Where then is the signal from the Jaramillo Event? Lowerlevel passages in Xanadu Cave (Sand Hills) and Zarathustra’s Cave (B-survey) contain sediment with measured normal polarity (Sasowsky et al. 1995). We report burial ages of 0.85 0.37 (0.38) Ma and 0.83 0.21 (0.22) Ma for these same sediments in Xanadu Cave and Zarathustra’s Cave. Based on the paleomagnetic data, we suggest that these sediments are actually younger than 0.78 Ma, which agrees with our data to within measurement uncertainties. This younger sediment fill has likely masked the Jaramillo event in the lower levels of the caves. (Future researchers may want to look for pockets of inplace sediments at the very top of the Sand Hills passage in Xanadu Cave.) If the reversed sequence in Xanadu Cave were actually 0.91 Ma, this would imply that three major cave levels developed within 50 m of elevation above the modern river level over the past 910,000 years (Sasowsky et al 1995). Thus, the cave passages must have formed rapidly, requiring high discharge. According to the Cumberland-style model, the discharge of the East Fork-Obey River (roughly 4.5 m s-1at its point of inflow 10 km upstream from Xanadu Cave) was diverted through both Zarathustra’s Cave (Heaven) and Xanadu Cave (Cumberland Avenue) (Sasowsky 1992). Independent evidence from scallops, however, demonstrated these passages carried low discharge. Scallops in Xanadu Cave (Cumberland Avenue) averaging 25 cm in diameter were used to calculate a paleodischarge of ~0.6 m s-1using Curl’s equations for cylindrical passages (Curl 1974). Scallops in Zarathustra’s Cave (Heaven) averaging 20 cm in diameter were used to calculate a paleodischarge of 0.3 m s-1. These discharges are an order of magnitude smaller than that of the East Fork-Obey River, but are within limits of recharge gathered from small drainage areas of side tributaries such as Lint’s Cove (5.4 km) and Pratt Branch (7.4 km). Comparison with Mammoth Cave, KY. Strong correlation between burial ages of sediments in Mammoth Cave (Fig. 1) and multilevel caves on the western margin of the Cumberland Plateau indicates synchronous incision of both the Green River and Cumberland River. Mammoth Cave shares many similarities with large multilevel caves along the western margin of the Cumberland Plateau, including a location within the unglaciated Ohio River basin, similar lithology and climatic history, and a history of cave development reaching well into the Pliocene. Burial ages of cave sediments at Mammoth Cave reveal a common thread between large caves throughout the Kentucky-Tennessee region and firmly link the speleogenesis of multilevel cave systems to the history of regional river incision. Burial dating of sediments using cosmogenic nuclides in the Mammoth Cave System (Table 2) records nearly four million years of water table position along the Green River (Granger et al 2001). In the Mammoth Cave study, level Aof Miotke and Palmer (1972) is older than 3.62 0.50 (0.52) Ma; both levels Aand B were aggraded at 2.61 0.16 (0.27) Ma. Excavation of sediments in levels Aand B occurred around 2 Ma, when the Green River incised and paused for nearly onehalf million years to form level C. Renewed incision of the Green River occurred to level D at 1.55 0.12 (0.18) Ma, abandoning level C and marking the end of well-developed levels (Palmer 1989). Incision at 1.45 0.12 (0.14) Ma and aggradation at 0.85 0.13 (0.16) Ma followed the abandonment of level D. Sediment fill in level D was re-excavated by incision to the modern river level. [Note: burial ages for Mammoth Cave sediments are recalculated in this paper using an AMS standard made by the U.S. National Institute of Standards and Technology (NIST) that yields a 10Be meanlife 14% lower than that previously accepted, and thus are slightly older than those reported in Granger et al 2001.] Dissolution kinetics and the age of Cumberland Avenue. Burial ages of cave sediments indicate that large passages such Figure 11. Sediments in Xanadu Cave, TN (Cumberland Avenue) contain a magnetically reversed-over-normal sequence, and were dated at ~1.6 Ma using cosmogenic nuclides. The burial age identifies the reversal as the youngerend of the Olduvai Event. (Photo Dave Bunnell.)

PAGE 9

54• Journal of Cave and Karst Studies August 2004 ALATETERTIARYORIGINFORMULTILEVELCAVES as Cumberland Avenue in Xanadu Cave, with a typical diameter of 20 m and a length of 1 km, formed over roughly 1.5 million years. Cleaveland Avenue (level C of Mammoth Cave) is less than 10 m in diameter over a length of 1.5 km and formed over a somewhat shorter interval of 0.5 million years. To first order, this suggests a long-term passage enlargement rate of roughly 0.01 mm/yr. Theoretical maximum enlargement rates calculated from dissolution kinetics are roughly 0.2–1 mm/yr (Palmer 1991, 2000; Dreybrodt & Gabrovšek 2000), which are over an order of magnitude faster than our data suggest. However, these theoretical maximum rates are calculated for highly undersaturated water. Both Palmer (2000) and Dreybrodt and Gabrovšek (2000) caution that natural waters often enter conduits with significant calcium in solution, and thus natural rates of cave enlargement may be 1–2 orders of magnitude less than the theoretical values. Our data indicate this to be the case. CONCLUSIONSLarge, multilevel caves on the western margin of the Cumberland Plateau (including some previously named as “fossil” or “Cumberland-style” caves) formed during a stable, Late Tertiary climate. The development and abandonment of horizontal passages at concentrated elevations above the modern river level is attributed to distinct episodes of stability and accelerated Plio-Pleistocene incision of the Cumberland River and its tributaries, for which there is good geomorphic and geologic evidence to suggest that river incision occurred as knickpoint migration. Achronology for the development of multilevel caves on the western margin may now be written to include: € Uppermost levels of cave passages formed prior to ~5.7 and ~3.5 Ma, when the Cumberland River and its tributaries flowed across the Eastern Highland Rim. € Asecond level of cave passages formed between ~3.5 and ~2 Ma during a major stillstand of the Cumberland River. Incision of the Cumberland River abandoned the second level beginning at ~2 Ma. € Athird level of cave passages formed between ~2 Ma and ~1.5 Ma during a brief stillstand of the Cumberland River. Incision of the Cumberland River abandoned the third level beginning at ~1.5 Ma. € Afourth level of cave passages formed after ~1.5 Ma; regional aggradation at ~0.8 Ma filled the fourth level and into the third group of cave passages. € Incision to the modern river level removed much of the ~0.8 Ma sediment fill. ACKNOWLEDGMENTSWe thank the countless number of cavers who participated in exploration and survey of multilevel caves on the western margin, especially those who continue the task today. The authors are deeply indebted to Phil Bodanza, Nick Crawford, Bill Deane, Ralph Ewers, Chris Groves, Sid Jones, Chris Kerr, Brad Neff, Art Palmer, Ira Sasowsky, Jeff Sims, Bill Walter, and Will White for spirited discussions concerning the origin and development of these particular caves. Access and permission to collect samples was granted by private landowners and the State of Tennessee Department of Environment and Conservation. Financial support for this work was obtained from the National Science Foundation (0092459-EAR); the National Speleological Society (Ralph Stone Research Award); the Geological Society of America; Purdue Research Foundation; and Sigma Xi, the Scientific Research Society. REFERENCESBarr, T.C., 1961. Caves of Tennessee. Tenn. Div. Geol. Bulletin 64, Nashville, 567 p. Butts, C., 1904. Description of the Kittanning quadrangle. USGS Folio 115, p. 2–3. Cande, S.C. & Kent, D.V., 1995. Revised calibration of the geomagnetic polarity timescale for the Late Cretaceous and Cenozoic. Journal of Geophysical Research Bulletin, v. 100, no. 4, p. 6093–6095. Crawford, N.C., 1984. Karst landform development along the Cumberland Plateau Escarpment of TN, in LeFleur, R.G., (ed.), Groundwater as a geomorphic agent. Boston, Allen and Unwin, Inc, p. 294–338. Curl, R.L., 1974. Deducing flow velocity in cave conduits from scallops. NSS Bulletin, v. 36, p. 1–5. Dreybrodt, W. & Gabrovšek, F., 2000. Dynamics of the evolution of single karst conduits, in Klimchouk, A.B., Ford, D.C., Palmer, A.N., & Dreybrodt, W. (eds.), Speleogenesis: Evolution of Karst Aquifers. Huntsville, AL, National Speleological Society, p. 184–193. Fenneman, N.M., 1938. Physiography of the Eastern United States. New York, McGraw-Hill, Inc., 714 p. Granger, D.E. & Muzikar, P.F., 2001. Dating sediment burial with in-situ produced cosmogenic nuclides: theory, techniques, and limitations. Earth and Planetary Science Letters, v. 188, no. 1–2, p. 269–281. Table 2. Cave levels, burial ages, and correlated surface features from the Mammoth Cave System, Kentucky (after Granger et al 2001)LevelElevation aboveTypical morphometric characteristicsAssociated surface featuresBurial ageaGreen River(m) A80+Large passages once filled with sedimentDeposition of “Lafayette-type” gravels3.62 0.50 (0.52) B50-80Very large passages (>100 m) once filled with sedimentBroad straths with thick (6-10 m) gravel2.15 0.24 (0.25) C47Large passages (~30 m) with little sedimentStrath in Green River valley1.55 0.12 (0.18) D30Small passages (~10 m) with little sedimentStrath in Green River valley1.45 0.12 (0.14) lower<30Small passages with undefined levelsAlluvial sediment in Green River0.85 0.13 (0.16)aBurial ages inferred from simultaneous solution of equations; uncertainties represent one standard error measurement uncertaint y, with systematic uncertainties added in quadrature and shown in parentheses.

PAGE 10

Journal of Cave and Karst Studies August 2004 • 55 ANTHONYANDGRANGER Granger, D.E., Kirchner, J. & Finkel, R, 1997. Quaternary downcutting rate of the New River, Virginia, measured from differential decay of cosmogenic 26Al and 10Be in cave-deposited alluvium. Geology, v. 25, no. 2, p. 107–110. Granger, D.E., Fabel, D. & Palmer, A.N. (2001). Plio-Pleistocene incision of the Green River, KYfrom radioactive decay of cosmogenic 26Al and 10Be in Mammoth Cave sediments. GSABulletin, v. 113, no. 7, p. 825–836. Hess, J.W. & White, W.B., 1989. Water budget and physical hydrology, in White, W.B. & White, E.L. (eds.), Karst Hydrology: Concepts from the Mammoth Cave Area. New York, Van Nostrand Reinhold, 346 p. Kohl, C.P. & Nishiizumi, K., 1992. Chemical isolation of quartz for measurement of in-situ produced cosmogenic nuclides. Geochimica et Cosmochimica Acta, v. 56:, p. 3583–3587. Lal, D. & Peters, B., 1967. Cosmic ray produced radioactivity on the Earth, in Flugge, S., (ed.), Handbuch der Physik. Berlin, Springer-Verlag, p. 551–612. Mann, R.A., 1982. Cave development along selected areas of the Western Cumberland Plateau Escarpment. Memphis State University, M.S. thesis. McFarlan, A.C., 1943. Geology of Kentucky. Lexington, KY, University of Kentucky Press, 531 p. Miotke, F. & Palmer, A.N., 1972. Genetic relationship between caves and landforms in the Mammoth Cave National Park area. Wurtzburg, Germany, Bohler-Verlag Press, 69 p. Muzikar, P., Elmore, D. & Granger, D.E., 2003. Accelerator mass spectrometry in geologic research. GSABulletin, v. 115, no. 6, p. 643–654. Palmer, A.N., 1987. Cave levels and their interpretation. NSS Bulletin, v. 49, p. 50–66. Palmer, A.N., 1989. Geomorphic history of the Mammoth Cave System, in White, W.B. & White, E.L. (eds.), Karst hydrology; concepts from the Mammoth Cave area. New York, Van Nostrand Reinhold, p. 317–327. Palmer, A.N., 1991. Origin and morphology of limestone caves. GSABulletin, v. 103, p. 1–21. Palmer, A.N., 2000. Digital modeling of individual solution conduits, in Klimchouk, A.B., Ford, D.C., Palmer, A.N., & Dreybrodt, W. (eds.), Speleogenesis: Evolution of Karst Aquifers. Huntsville, AL, National Speleological Society, p. 367–377. Potter, P.E., 1955. The petrology and origin of the Lafayette gravel part 2. Geomorphic history. Journal of Geology, v. 63, p. 115–132. Sasowsky, I.D., 1992. Evolution of the Appalachian Highlands: East Fork Obey River, Fentress County, TN. The Pennsylvania State University, Ph.D. thesis. Sasowsky, I.D. & White, W.B., 1994. The role of stress release fracturing in the development of cavernous porosity in carbonate aquifers. AGU Water Resources Research, v. 30, no. 12, p. 3523–3530. Sasowsky, I.D., White, W.B. & Schmidt, V., 1995. Determination of streamincision rate in the Appalachian plateaus by using cave-sediment magnetostratigraphy. Geology, v. 23, no. 5, p. 415–418. Stone J.O., 2000. Air pressure and cosmogenic isotope production. Journal of Geophysical Researchv. 105, no. 23, p. 753–759. Thornbury, W.D., 1965. Regional geomorphology of the United States. New York, John Wiley and Sons, Inc., 609 p. White, W.B., 1977. Role of solution kinetics in the development of karst aquifers in Tolson, J.S. & Doyle, F.L. (eds.), Karst Hydrogeology. Huntsville, University of Alabama Press, p. 503–517. White, W.B. & White, E.L,. 1970. Channel hydraulics of free-surface streams in caves. Caves and Karst, v. 12, p. 41–48. White, W.B. & White, E.L., 1983. Karst landforms and drainage basin evolution in the Obey River Basin, north-central Tennessee. Journal of Hydrology, v. 61, p. 69–82. Wilson, C.W., 1948. The geology of Nashville, Tennessee. Tennessee Division of Geology Bulletin, v. 53, 172 p.



PAGE 1

National Speleological Society 2813 Cave Avenue Huntsville, Alabama 35810-4431 New from Elsevier/Academic Press AEG/LB/FR-26164-7/04Elsevier, DM78849, Order Fulfillment 11830 Westline Industrial Dr., St. Louis, MO, 63146 Tel 800.545.2522 € Fax 800.535.9935 Online: http://books.elsevier.com Order from your favorite bookseller or direct from Elsevier: August 2004 € Hardcover € 696 pp. ISBN: 0-12-198651-9 € $99.95ENCYCLOPEDIA of CAVESƒan exciting and significant contribution to the field of caves and karstƒa credit to the fieldƒŽ„ Malcolm Field, Editor, Journal of Cave and Karst StudiesThis monumental and attractively presented reference book belongs on the bookshelf of every person with an interest in caves... an invaluable reference for students, teachers, scientists...Ž „ Larry Master, Chief Zoologist, NatureServe Culver and White have brought together a truly international team of world-class experts...For cavers and professionals [the Encyclopedia] will serve as the most comprehensive state of the art reference in the multidisciplinary field of subterranean sciences.Ž„ Peter Trontelj, University of Ljubljana, Slovenia Edited by:David C. CulverAmerican University, Washington D.C., U.S.A.William B. WhiteThe Pennsylvania State University, University Park, U.S.A.ENCYCLOPEDIA of CAVES Comprehensive Articles € Full Color € For Researchers, Students and Explorersbooks.elsevier.com/caves € Presents a cross-section of contemporary knowledge of caves ranging from biology, geology, and human uses to exploration techniques € Brings together 107 in-depth articles from experts in 15 different countries € Enhanced with hundreds of color photographs, maps, and illustrations € Organized for ease of use with cross references, suggestions for further reading, plus a full glossary and index € Affordably priced for individuals € Highlights many of the great caves of the worldKEY FEATURES:



PAGE 1

70• Journal of Cave and Karst Studies August 2004 BOOKREVIEW John Gunn, Editor with Board of Advisers: Andrew Chamberlain, Emily Davis, Derek Ford, David Gillieson, William Halliday, Elery Hamilton-Smith, Alexander Klimchouk, David Lowe, Arthur Palmer, Trevor Shaw, Boris Sket, Tony Waltham, Paul Williams, and Paul Wood. First Edition (2004). Routledge Taylor and Francis Group. Routledge, New York, NY, 902 p. ISBN 1579583997. $150 (US) and $225 (CAN). Order on-line at http://www.routledgeny.com/books.cfm?isbn=1579583997. Also available on-line at http://www.Amazon.com for $200 (US) with 24-hour shipping. Amazon also had “5 used & new from $167.69” (US) at the time of this writing. (Actually four of the five were listed as new and only one of the five was listed as “like new” but I can’t comprehend why anyone wouldn’t keep this book after purchasing a copy!) Encyclopedia of Caves and Karst Science published in 2004, is a significant addition to the subject of karst in all its various forms and one which the Editor, Board of Advisers, and various authors (who too numerous to mention here) should be proud. This monograph contains over 350 entries (353 according to Mixon 2004, p. 89) and over 500 blackand-white photographs, maps, diagrams, and tables. Fifty-one color photographs are grouped together near the middle of the encyclopedia. All photographs, maps, and diagrams are very clear and relevant to the text, so that clarity is added to material that is difficult to explain, although some diagrams (e.g., Fig. 1 and 2 under the heading “Chemistry of Natural Karst Waters”) may be confusing to non-specialists but not greatly so. According to the Editor’s Introduction, “This is the first encyclopedia on the subject of Caves and Karst Science and provides a unique, comprehensive, and authoritative reference source that can be used both by subject-specialists who wish to obtain information from outside of their immediate area of knowledge and by non-specialists who wish to gain an understanding of the diverse and multi-disciplinary nature of caves and karst science.” This introductory statement by the Editor basically says it all; if you want to learn something new about almost any aspect of caves and karst, you will most likely find it in this monograph. Although not intended as a geographical atlas, the encyclopedia does address scientifically important karst areas at the continent, country, region and/or sitespecific level. In terms of karst science, it addresses “archaeology, biology, chemistry, ecology, geology, geomorphology, history, hydrology, paleontology, physics as well as exploration, survey, photography, literature, and art.” (As with any undertaking of this nature there are always going to be some omissions and errors.) The breadth and scope of the coverage of subcategories of general science in the encyclopedia is a significant accomplishment. As pointed out by Mixon (2004, p. 89), the 202 authors from 36 countries developed exceptionally readable entries which further lends credit to the level of effort by the editor. The articles are relatively short as would be expected for an encyclopedia—one to several pages of two-column 9-point type. Each article ends with a bibliography listed as “works cited” or “further reading” which is probably appropriate for an encyclopedia and for nonscientists, but as a professional scientist, I would have looked ENCYCLOPEDIAOFCAVES AND KARST SCIENCE

PAGE 2

Journal of Cave and Karst Studies August 2004 • 71 BOOKREVIEW for more formal detailed citation/reference list typical of scientific work. However, if a typical scientific citation/reference list had been chosen, then this monograph would probably have increased in length by a factor of 10 or greater. Given the impossibility of such an increase in length and the bibliographic sources listings, I am quite satisfied at being able to find those references that most interest me. An interesting and appropriate aspect of this monograph is the importance placed on exploration and basic science. The study of caves and karst is unique in that cave exploration and science are complementary, which draws individuals from extremely diverse backgrounds together to discuss new findings or new thoughts on older ideas. To integrate exploration and science, the Board of Advisers spent considerable time and energy developing and revising a list of the “world’s important karst areas and most important caves.” Having developed this list, the Board of Advisers then drew up a list of “topical entries considered to be of primary importance to their particular branch of science.” This undertaking has resulted in a good mix of exploration and science, although interested readers will need to do some searching in the “Alphabetic List of Entries,” “Thematic List of Entries,” and/or index (93 pages) to locate all of the items of interest. Many of the exploration and scientific entries may require some extra effort by the reader to fully understand the material presented if the subject entry is not a specialty of the reader. For example, entries such as the “Encantado, Sistema del Rio, Puerto Rico,” “Krubera Cave, Georgia,” “Peak District, England” and other foreign cave and/or karst entries use some geological terms and locality-specific terms that may be unfamiliar to some readers. For the most part, however, the exploration entries are pretty straightforward. The scientific entries are also fairly readable but may be somewhat more difficult for non-specialists. For example, the entry “Dissolution: Carbonate Rocks” and “Dissolution: Evaporite Rocks” necessarily include discussions of the physics and chemistry of dissolution kinetics of carbonates and evaporites. One aspect not readily apparent from the title of this monograph or from the introduction, or flyers announcing its availability, is the inclusion of a significant amount of non-exploration and non-science material. For example, a discussion of “Journals on Caves” with source availability was compiled with a discussion covering two pages. This entry is quite useful for scientists and non-scientists alike because the subject of caves and karst is very diverse, with small publishing groups spread far-and-wide. Asomewhat stranger entry is “Caves in Fiction” which chronicles the history of stories revolving around caves and which falls into the nonscience material. “Art Showing Caves” is a similarly strange entry. So what about omissions and errors? As mentioned above, such was bound to occur in an undertaking of this magnitude. According to Mixon (2004, p. 89), an error occurs in the “America, Central” entry in which some Mexican caves were mislocated on the area map, as well as some confusion over when cave research began in Mexico. There are perhaps more errors of this sort but I suspect not many. Omissions are a minor issue as well. Invariably at any given time, any particular reader will be frustrated that a specific subject of interest to that reader may not have been included in the encyclopedia. Given the immensity of this undertaking and the need to find authoritative authors for each subject entry while keeping the monograph to a “manageable” size, it was necessary that some topics be excluded. For example, I was unable to locate an entry addressing the epikarstic (subcutaneous) zone. I scoured the encyclopedia but the only discussion I could find on the epikarstic zone occurred under the heading “Dolines” and brief mention under the headings “Groundwater in Karst” and “Groundwater in Karst: Conceptual Models.” In all likelihood the epikarstic zone is probably addressed in other parts of this monograph, but it should have had its own entry. Overall I feel that this encyclopedia is a must purchase for anyone with more than a passing interest in caves and karst, including the nonscience entries much of which I found to be interesting reading. It contains a wealth of information that far outweighs its $150 price tag ($225 Canadian) and its relatively insignificant “problem” areas. Students, researchers, cavers, geotechnical consultants, and environmental professionals will all consider this book well worth the purchase price. REFERENCESMixon B. 2004, Review of Encyclopedia of Caves and Karst Science, NSS News, National Speleological Society, p. 89. Book reviewed by Malcolm S. Field, National Center for Environmental Assessment (8623D), Office of Research and Development, U.S. Environmental Protection Agency, 1200 Pennsylvania Ave., NW, Washington, DC 20460 (field.malcolm@epa.gov) May 2004.



PAGE 1

August 2004 Volume 66 Number2 ISSN 1090-6924APublication of the National Speleological SocietyJOURNALOF CAVE AND KARST STUDIES Ten Acre Room, Cumberland Caverns, Tennessee

PAGE 2

EditorMalcolm S. FieldNational Center of Environmental Assessment (8623D) Office of Research and Development U.S. Environmental Protection Agency 1200 Pennsylvania Avenue NW Washington, DC 20460-0001 202-564-3279 Voice 202-565-0079 FAX field.malcolm@epa.govProduction EditorJames A. PisarowiczWind Cave National Park Hot Springs, SD 57747 605-673-5582 pisarowicz@alumni.hamline.eduBOARD OFEDITORS AnthropologyPatty Jo WatsonDepartment of Anthropology Washington University St. Louis, MO 63130 pjwatson@artsci.wustl.eduConservation-Life SciencesJulian J. LewisJ. Lewis & Associates, Biological Consulting 217 West Carter Avenue Clarksville, IN 47129 812-283-6120 lewisbioconsult@aol.comEarth Sciences-Journal IndexIra D. SasowskyDepartment of Geology University of Akron Akron, OH 44325-4101 330-972-5389 ids@uakron.eduExplorationPaul BurgerCave Resources Office 3225 National Parks Highway Carlsbad, NM 88220 (505)785-3106 paul_burger@nps.govPaleontologyGreg McDonaldGeologic Resource Divison National Park Service P.O. Box 25287 Denver, CO 80225 303-969-2821 Greg_McDonald@nps.govSocial SciencesJoseph C. DouglasHistory Department Volunteer State Community College 1480 Nashville Pike Gallatin, TN 37066 615-230-3241 Joe.Douglas@volstate.eduBook ReviewsErnst H. KastningP.O. Box 1048 Radford, VA24141-0048 ehkastni@runet.eduProofreaderDonald G. DavisJOURNALADVISORYBOARDChris GrovesCarol Hill Horton Hobbs IIIDavid Jagnow Julia JamesKathy Lavoie Joyce LundbergDonald MacFarlane William White Journal of Cave and Karst Studies of the National Speleological SocietyVolume 66 Number 2 August 2004CONTENTS Article Application of resistivity and magnetometry geophysical techniques for near-surface investigations in karstic terranes in Ireland P.J. Gibson, P. Lyle, and D.M. George 35 Article Estimating subterranean species richness using intensive sampling and rarefaction curves in a high density cave region in West Virginia Katie Schneider and David C. Culver 39 Article ALate Tertiary origin for multilevel caves along the western escarpment of the Cumberland Plateau, Tennessee and Kentucky, established by cosmogenic 26Al and 10Be Darlene M. Anthony and Darryl E. Granger 46 Long and Deep Caves of the World 56 Bob Gulden Guide to Authors 58 Article Landform differentiation within the Gunung Kidul Kegelkarst, Java, Indonesia Eko Haryono and Mick Day 62 Book Review Encyclopedia of Caves and Karst Science 70The Journal of Cave and Karst Studies (ISSN 1090-6924, CPM Number #40065056) is a multi-disciplinary, refereed journal published three times a year by the National Speleological Society, 2813 Cave Avenue, Huntsville, Alabama 35810-4431 USA; (256) 852-1300; FAX (256) 851-9241, e-mail: nss@caves.org; World Wide Web: http://www.caves.org/pub/journal/. The annual subscription fee, worldwide, by surface mail, is $18 US. Airmail delivery outside the United States of both the NSS News and the Journal of Cave and Karst Studies is available for an additional fee of $40 (total $58); The Journal of Cave and Karst Studies is not available alone by airmail. Back issues and cumulative indices are available from the NSS office. POSTMASTER: send address changes to the Journal of Cave and Karst Studies 2813 Cave Avenue, Huntsville, Alabama 35810-4431 USA. The Journal of Cave and Karst Studies is covered the the following ISIThomson Services: Science Citation Index Expanded, ISI Alerting Services, and Current Contents/Physical, Chemical, and Earth Sciences. Copyright 2004 by the National Speleological Society, Inc. Printed on recycled paper by American Web, 4040 Dahlia Street, Denver, Colorado 80216 USA Front cover: Ten Acre Room, Cumberland Caverns, TN. Photo by Bob Biddix. See Darlene M. Anthony and Darryl E. Granger, p. 46.



PAGE 1

Journal of Cave and Karst Studies August 2004 • 35 P.J. Gibson, P. Lyle, and D.M. George Application of resistivity and magnetometry geophysical techniques for near-surface inv estigations in karstic terranes in Ireland. Journal of Cave and Karst Studies v. 66, no. 2, p. 35-38. Although most karstic regions are characterised by caves, collapse features or passageways, such features often do not have a surface expression, and their presence may go unrecorded. Approximately 35% of Ireland’s land surface is underlain by Mississippian limestone, and karst landforms are known from Counties Roscommon, Fermanagh, Galway and the Burren in County Clare (Figure 1). However, most of the limestone is extensively covered by Quaternary glacial sediments, especially in the Irish midlands. It is believed that widespread karstification occurred in Ireland during the Tertiary, but the character of such karst landscapes is wholly unknown because of this surficial cover (Drew 1997). Geophysical surveying can, in certain circumstances, provide us with the means of locating karst features. Acommonly employed geophysical technique employed in karst terranes is gravity surveying because the density contrast between air and rock is large. This has been employed to a limited extent in Ireland (Hickey & McGrath 2003), but a drawback of this technique is the large number of corrections — latitudinal, elevational, topographical, tidal and drift — that have to be applied to the data before they can be modeled. However, there are other geophysical techniques which can be used in karst terranes, two of which are considered here: magnetometry and resistivity (Gibson et al 1996; El-Behiry & Hanafy 2000). The former technique is used to investigate a paleokarst structure and the latter technique employed to discover an unknown collapse structure and cave in Ireland. The resistivity data were collected and modeled in the field on a laptop computer in less than one hour. The magnetometry study took less than 20 minutes, providing near real-time acquisition of subsurface information which can be acted on while still in the field. APPLICATION OFRESISTIVITYAND MAGNETOMETRY GEOPHYSICALTECHNIQUES FOR NEAR-SURFACE INVESTIGATIONS IN KARSTIC TERRANES IN IRELANDP.J. GIBSON Environmental Geophysics Unit, Department of Geography, National University of Ireland, Maynooth, Co. Kildare, REPUBLICOFIRELANDP. LYLE School of the Built Environment, University of Ulster, Jordanstown, Co. Antrim, NORTHERNIRELANDD.M. GEORGE Environmental Geophysics Unit, Department of Geography, National University of Ireland, Maynooth, Co. Kildare, REPUBLICOFIRELAND Extensive glacial surficial deposits in Ireland prevent the identification of many karst features. Surface magnetic and resistivity geophysical measurements have been used to identify unknown karstic features. Two dimensional resistivity imaging has located an unknown 210-meter-long, 70-meter-wide and 25meter-deep collapse feature in eastern Ireland beneath the surficial sediments. Aresistivity survey over the Cloyne cave system in County Cork has identified the position of an unknown cave. Amagnetic investigation of an infilled paleokarst collapse structure produced a 40 nanoTesla anomaly and illustrates that the technique can be employed in Ireland to locate unknown ones. Figure 1: Location map showing localities mentioned in the text.

PAGE 2

36• Journal of Cave and Karst Studies August 2004 APPLICATIONOFRESISTIVITYANDMAGNETOMETRYGEOPHYSICALTECHNIQUES MAGNETICCASESTUDYAproton precession magnetometer was used to measure the Earth’s total magnetic field which varies with latitude, from about 30,000 nanoTesla (nT) near the equator increasing to around 65,000 nTnear the poles. The theoretical principles regarding such magnetometers can be found in standard geophysical texts (Sharma 1997; Gibson & George 2003). Magnetic susceptibility is a property of a body and is a measure of how easily it can be magnetized. Limestone has an extremely low susceptibility, thus a collapse feature infilled by sediment with a higher susceptibility will be associated with higher magnetic readings. Collapse features are known to exist near Cookstown, northern Ireland, but other unknown ones, which pose a potential risk of collapse, are suspected. Amagnetic study was made of a known one to ascertain if this approach could be adopted in the search for unknown ones. Figure 2 shows a funnel-shaped 15m-deep paleokarst collapse feature. The structure is 8 meters across nearest the surface and is capped by a 1.5 meter thick grainstone which indicates a return to marine conditions after the sub-aerial erosion phase during which the structure formed. The collapse feature is infilled by fine-grained unstratified red-brown sediment which is possibly of aeolian origin. The mass specific susceptibility ( ) and percentage frequency dependent susceptibility ( fd%) of the infill and the limestone were obtained using an MS2 Bartington laboratory magnetic susceptibility system. Aplot of fd% against shows that the limestone is virtually non-magnetic but the infilled sediment has a mass specific susceptibility that is considerably higher (Figure 3a). Amagnetic traverse taken across the feature shows a conspicuous positive 40 nT anomaly (Figure 3b) illustrating that the technique can be successfully employed in such Irish terranes. RESISTIVITYCASESTUDIESElectrical resistivity techniques involve inputting current into the ground via two source electrodes and measuring the potential difference between two sink electrodes — see Gibson & George (2003) for further details. In this study the process was automated using a multi-core cable and 25 electrodes and a two-dimensional apparent resistivity pseudosection was produced. The pseudosection was modeled using RES2DINVprogram which utilizes a least-squares optimization approach in order to determine how the true resistivity varies with depth (Loke & Barker 1995; 1996). In the examples shown here, errors are of the order of 5 per cent. COLLAPSEFEATUREFigure 4a shows the results of a resistivity traverse across a flat football pitch in the town of Maynooth, eastern Ireland (see Figure 1 for location). The limestone in this region is covered by 10m of Quaternary glacial sediments and there are no known karstic features. Other resistivity traverses in this locality have shown that the resistivity of the limestone is typically 500–1000 ohm meters. The acquired data indicate the presence of an unknown collapse feature in the underlying limestone. Bedrock is quite near the surface at the beginning (0–50) and end (170–220) of the traverse and is shown as a red-pink color (Fig. 4a). The central portion of the image is characterized by Figure 2: Paleokarst collapse feature in the Carboniferous limestone in County Tyrone, northern Ireland. Figure 3: (a) Magnetic susceptibility plot of limestone and infilled sediment forpaleokarst collapse feature near Cookstown, County Tyrone, northern Ireland. (b) Results of magnetometertraverse across the same feature.

PAGE 3

Journal of Cave and Karst Studies August 2004 • 37 GIBSON, LYLE, ANDGEORGE resistivity values an order of magnitude less than those expected for the limestone and are shown in blue. These low values are similar to those obtained for glacial sediments in the vicinity, and the observed pattern is interpreted as an unknown infilled collapse feature approximately 70 meters wide and 25–30 meters deep. Asecond traverse was taken at right angles to Figure 4a in order to determine its extent. The results show that in this direction the feature is considerably longer (Fig. 4b). Anumber of such traverses were undertaken and they indicate that the feature is about 25 meters deep, with a 210meter-long axis oriented NW/SE, and a 70 meter shorter axis oriented approximately at right angles to the long axis. CAVESYSTEMOne of the largest subsurface resistivity contrasts is that between solid rock and air such as can occur in a cave system (Morgan et al 1999; Roth et al 1999, 2000). In practice, airfilled caves are typically associated with resistivity values greater than about 15,000 ohm meters, the actual resistivity obtained depending on the size of the caves. Figure 4c shows a 2D resistivity image taken over the Cloyne cave system in Co. Cork, Ireland. The very high resistivity values of over 30,000 ohm meters between 180–210 meters were acquired over a mapped region in which caves are known to exist. However, a similar anomaly associated with high resistivity values can be observed in the 40–70 meter range at a depth of about 20 meters. This area has not been explored and the anomaly is interpreted as an unmapped cave. CONCLUSIONMagnetometry and resistivity are geophysical techniques that can provide useful subsurface information in karst regions. The resistivity of air-filled caves is always significantly higher than the bulk rock and, because limestone is virtually nonmagnetic, even infill with a low magnetic susceptibility will often yield a magnetic contrast. The techniques have been employed in Ireland to show that karst features can be located by such means and to discover an unknown cave and a large unknown collapse feature below the glacial deposits. Figure 4: (a) Resistivity traverse across unknown collapse feature in Maynooth, County Kildare, eastern Ireland. (b) Resistivity traverse across same feature at right angles to (a). (c) Resistivity traverse across a known and unmapped cave.

PAGE 4

38• Journal of Cave and Karst Studies August 2004 APPLICATIONOFRESISTIVITYANDMAGNETOMETRYGEOPHYSICALTECHNIQUES REFERENCES Morgan, F.D., Shi, W., Vichabian, Y., Sogade, J. & Rodi, W., 1999. Resistivity in cave exploration, in Powers, M.H., Cramer, L. & Bell, R.S. (eds.): Proceedings of the Symposium on the Application of Geophysics to Engineering and Environmental Problems. Annual Meeting of the Engineering and Environmental Geophysical Society, Oakland, California, p. 303–308. Roth, M.J.S., Mackey, J.R. & Nyquist, J.E., 1999. Acase study of the use of multi-electrode earth resistivity in thinly mantled karst, in Powers, M.H., Cramer, L. & Bell, R.S. (eds.): Proceedings of the Symposium on the Application of Geophysics to Engineering and Environmental Problems. Annual Meeting of the Engineering and Environmental Geophysical Society, Oakland, California, p. 293–302. Roth, M.J.S., Nyquist, J.E. & Guzas, B., 2000. Locating subsurface voids in karst: a comparison of multi-electrode earth resistivity testing and gravity testing, in Powers, M.H., Ibrahim, A. & Cramer, L. (eds.): Proceedings of the Symposium on the Application of Geophysics to Engineering and Environmental Problems. Annual Meeting of the Engineering and Environmental Geophysical Society, Arlington, Virginia, p. 359–365. Sharma, P.V., 1997. Environmental and Engineering Geophysics. Cambridge University Press, Cambridge. Drew, D., 1997. Landforms and hydrology of the Co. Westmeath lakeland area, in Mitchell, F. & Delaney, C. (eds.): The Quaternary of the Irish Midlands. Irish Association for Quaternary Studies Field Guide No. 21, p. 64–69. El-Behiry, M.G. & Hanafy, F.M., 2000. Geophysical surveys to map the vertical extension of a sinkhole: a comparison study., in Powers, M.H., Ibrahim, A. & Cramer, L. (eds.): Proceedings of the Symposium on the Application of Geophysics to Engineering and Environmental Problems. Annual Meeting of the Engineering and Environmental Geophysical Society, Arlington, Virginia, p. 341–350. Gibson, P.J. & George, D.M., 2003. Environmental Applications of Geophysical Surveying Techniques. Nova Science Publishers Inc., New York. Gibson, P.J., Lyle, P. & George, D.M., 1996. Environmental Applications of Magnetometry Profiling. Journal of Environmental Geology and Water Sciences, v. 27, p. 178–183. Hickey, C. & McGrath, R., 2003. Mapping karst features using microgravity. Geological Survey of Ireland Groundwater Newsletter, v. 42, p. 22–27. Loke, M.H. & Barker, R.D., 1995. Least-squares deconvolution of apparent resistivity pseudosections. Geophysics, v. 60, p. 1682–1690. Loke, M.H. & Barker, R.D., 1996. Rapid least squares inversion of apparent resistivity pseudosections by a quasi-Newton method. Geophysical Prospecting, v. 44, p. 131–152.



PAGE 1

58• Journal of Cave and Karst Studies August 2004 GUIDETOAUTHORS The Journal of Cave and Karst Studies is a multidisciplinary journal devoted to cave and karst research. The Journal is seeking original, unpublished manuscripts concerning the scientific study of caves or other karst features. Authors do not need to be members of the National Speleological Society but preference is given to manuscripts of importance to North American speleology. LANGUAGESThe Journal of Cave and Karst Studies uses Americanstyle English as its standard language and spelling style, with the exception of allowing a second abstract in another language when room allows. In the case of proper names, the Journal tries to accommodate other spellings and punctuation styles. In cases where the editor finds it appropriate to use nonEnglish words outside of proper names (generally where no equivalent English word exist), the Journal italicizes them (i.e., et al .). Authors are encouraged to write for our combined professional and amateur readerships CONTENTEach paper will contain a title with the authors’names and addresses, an abstract, and the text of the paper, including a summary or conclusions section. Acknowledgments and references follow the text. ABSTRACTSAn abstract stating the essential points and results must accompany all articles. An abstract is a summary, not a promise of what topics are covered in the paper. STYLEThe Journal consults The Chicago Manual of Style on most general style issues. REFERENCESIn the text, references to previously published work should be followed by the relevant author’s name and date (and page number, when appropriate). All cited references are alphabetical at the end of the manuscript with senior author’s last name first, followed by date of publication, title, publisher, volume, and page numbers. Geological Society of America format should be used (see http://www.geosociety.org/pubs/geoguid5.htm). Please do not abbreviate periodical titles. Web references are acceptable when deemed appropriate. The references should follow the style of: Author (or publisher), 16 July 2002, Webpage title: Publisher (if a specific author is available), full URL(e.g., http://www.usgs.gov/citguide.html). If there is a specific author given, use their name and list the responsible organization as publisher. Because of the ephemeral nature of websites, please provide the specific date. Citations within the text should read: (Author Year). SUBMISSIONAuthors should submit three copies of their manuscript (include only copies of the illustrations) to the appropriate specialty editor or the senior editor. Manuscripts must be typed, double spaced, and single-sided. Electronic mail submissions are encouraged. Authors will be requested to submit an electronic copy of the text, a photograph, and brief biography upon acceptance of the paper. Extensive supporting data will be placed on the Journal ’s website with a paper copy placed in the NSS archives and library. The data that are used within a paper must be made available. Authors may be required to provide supporting data in a fundamental format, such as ASCII for text data or comma-delimited ASCII for tabular data. DISCUSSIONSCritical discussions of papers previously published in the Journal are welcome. Authors will be given an opportunity to reply. Discussions and replies must be limited to a maximum of 1000 words and discussions will be subject to review before publication. Discussions must be within 6 months after the original article appears. MEASUREMENTSAll measurements will be in Systeme Internationale (metric) except when quoting historical references. Other units will be allowed where necessary if placed in parentheses and following the SI units. FIGURESFigures and lettering must be neat and legible. Figure captions should be on a separate sheet of paper and not within the figure. Figures should be numbered in sequence and referred to in the text by inserting (Fig. x). Most figures will be reduced, hence the lettering should be large. Once the paper has been accepted for publication, the original drawing (with corrections where necessary) must be submitted to the editor. Photographs must be sharp and high contrast. Color will generally only be printed at author’s expense. TABLESSee the “Guidelines for Authors for Producing Tables” on pages 60-61. GUIDE TO AUTHORS

PAGE 2

Journal of Cave and Karst Studies August 2004 • 59 GUIDETOAUTHORS ELECTRONICFILESThe Journal ’s final layout is done using Quark Xpress. Microsoft Word is used in word processing and all figures and photographs should be submitted in either EPS or TIF format. The Journal is printed at 305 dpi. Thus, illustrations that are to be printed at 3.5 inches wide require at least 1068 pixels. COPYRIGHTANDAUTHOR’SRESPONSIBILITIESIt is the author’s responsibility to clear any copyright or acknowledgement matters concerning text, tables, or figures used. Authors should also ensure adequate attention to sensitive or legal issues such as land owner and land manager concerns or policies. PROCESSAll submitted manuscripts are sent out to at least two experts in the field. Reviewed manuscripts are then returned to the author for consideration of the referees’remarks and revision, where appropriate. Revised manuscripts are returned to the appropriate associate editor who then recommends acceptance or rejection. The Senior Editor makes final decisions regarding publication. Upon acceptance, the author should submit all photographs, original drawings, and an electronic copy of the text to the editor. The senior author will be sent one set of PDF proofs for review. Examine the current issue for more information of the format used.

PAGE 3

60• Journal of Cave and Karst Studies August 2004 TABLES TABLECAPTION1. Number tables in the order in which they are cited in the paper. Follow the number with a period and two blank spaces, then the caption. Capitalize only the first letter in the caption, except symbols from chemical elements (e.g., Rn) AND the first letter of formal names and scientific names (except species epithets). Capitalize abbreviations for years before present only whe appropriate (e.g., Ma and ka). End the caption with a period. Italicize all scientific names. Left justify and boldface the entire table caption on one or more lines at the top of the table. 2. Separate the caption from the rest of the table with a thick horizontal line. In the example shown, line thickness is 0.08 em1. TABLEHEADINGS3. Where appropriate place a very thin line underneath a subheading. In the example shown, line thickness is 0.03 em1. 4. Start all column headings just below the thick horizontal lines. Left justify the first column; center all other column headings. Capitalize each initial letter for each heading item unless other capital letters are required (e.g., scientific names or chemical symbols). 5. Abbreviate units of measurement and place them in parentheses on a separate line just below the rest of the heading. Use only Le Systme International d’Units (SI) units of measurement2. Enlarge parentheses as necessary to enclose unit of measure completely (i.e., to account for superscripts and subscripts). 6. Separate the headings from the body of the table with a thin horizontal line. In the example shown, line thickness is 0.05 em1. TABLEBODY7. Start all columns just below the thin horizontal line at the base of the column headings. Left justify the first column and center all the other columns. Do not show units of measurement in the column if they can be abbreviated and placed in parentheses just below the column heading. 8. Align columns of numbers on the decimal or other appropriate marker (e.g., the symbol). Use a zero before the decimal point for values less than one. 9. Align text entries on the left and indent each line after the first and end each sentence with a period. Use only an initial capital for each complete sentence unless other capitals are required. 10. Separate sections of the table with line spaces. Label these sections with a very thin lined heading that is left justified. In the example shown, line thickness is 0.03 em1. Indent subitems one space. 1One em is the width of a capital ‘M’in the current font.2See Nat. Inst. of Stan. and Tech. Publ. 330 and 811 at http://physics.nist.gov/cuu/Units/bibliography.html for correct SI units. GUIDELINES FOR AUTHORS FOR PRODUCINGTABLES FOR THE JOURNALOFCAVE AND KARST STUDIES Table1.Measured222Rnequlibriumactivityandspecicconductivityforselectedsamplingstations. Location Lat.Long.222RnActivityaSpecicConduct.b SampleName .šN /.šW / kBqm 3 Ss 1 Description Wells MunicipalWell39š290000077š200210015 : 21 2 : 74 0.390Principaldrinking-waterwellfor townpopulace. DairyWell39š290250077š2002100 0.380Principalwater-supplywellfor wateringdairycows. FarmWell9š290250077š20022006 : 44 2 : 52 0.448Farmhousedrinking-waterwell. Springs WillowSpring39š290290077š20022009 : 66 4 : 26 0.545Smallseepagespring. FountainRockSpring39š280300077š22000007 : 77 2 : 63 0.520Largeowingspringusedfor shhatchery.c Note: Sampleswerecollectedduringaverywetperiod(1992);dryerconditionswouldlikelyyielddifferentresults.aMeasuredequilibriumactivitydeterminedbyliquidscintillationcounting.bThearithmeticmeanforallmeasuredspecicconductivityvaluesis4 : 85 10 1 Ss 1;nomeasurementseverexceeded7 : 60 10 1 Ss 1.cFountainRockSpringisnolongerusedasashhatchery.Partsofatypicaltable. Highlightednumbersareexplainedbynumbereditemsinthetext. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

PAGE 4

Journal of Cave and Karst Studies August 2004 • 61 TABLES 11. Do not leave blank spaces in the body of the table. These should be marked ‘...’(no data), ‘N.A.’(not applicable) or otherwise as appropriate, and the abbreviations should be marked with a footnote for explanation. 12. Follow the body of the table with a thick horizontal line. In the example shown, line thickness is 0.08 em1. FOOTNOTESYMBOLS€ If several items in a table require footnotes, use relative position in the table to determine the order in which footnotes are assigned. Start at the top of the table, work from left to right, then from top to bottom. € Use lowercase alphabetical characters for footnotes: a, b, c ... z. TABLEFOOTNOTES13. Treat each footnote as a separate paragraph; indent the first line three spaces and end the footnote with a period. Place general information about the table in the first footnote. Precede this entry with ‘Note:’in italics rather than with a symbol. 14. Footnotes should appear in the same order as the symbols were used in the table. Use only an initial capital letter for each sentence in each footnote. ADDITIONALREQUIREMENTS15. Scale SI units using appropriate SI prefixes (e.g., k, , etc.) 16. Always use the mathematical minus sign, ‘–’to indicate subtraction when using mathematical formulae; never substitute an hyphen ‘-’, an en-dash ‘–’, or an em-dash ‘—’for a minus sign ‘–’in mathematical formulae. 17. When reporting data using scientific notation always use the symbol for multiplication, (e.g., 7.60 10–1S s–1). € If a separate section is to be incorporated into the table (e.g., different dates for different sampling events) then separate these sections with a centered and italicized caption within the body of the table. Do not boldface this caption, only capitalize the inital letter of the first word in the caption except as required (e.g., scientific names), and do not end this caption with a period. € Never use vertical lines anywhere in the table. € Never boldface any part of the table other than the caption. € Never use English units of measurement except as allowed (see EXCEPTIONS). € Never italicize units of measure. € Never use nonSI units of measurement except as permissible under specific SI guidelines (e.g., liter). € When reporting data using scientific notation never use the letter ex, ‘x’and never report data using either ‘e’or ‘E to indicate the exponential as would be obtained from a computer program (e.g., 7.60E–1 S s–1). € Never substitute a spreadsheet for a properly constructed table. EXCEPTIONS€ If appropriate, some units of measurement may be used in place of SI units of measurements (e.g., hours may be more appropriate than seconds for long time periods). € In rare instances it may be reasonable to list the correct SI unit of measure followed by its English equivalent enclosed in brackets. For example: (m3s-1) [cfs]; subsequent English numerical values also enclosed in brackets would follow the SI numerical values in the body of the text. € The combination of thick and thin lines may be replaced with a set of uniformly-thick lines. SPECIALEXCEPTION€ If for some reason a proposed data table cannot reasonably match the example shown, then please contact the Editor of the Journal of Cave and Karst Studies for consideration of a special exception. € For those individuals using software or equipment other than MS Word, WordPerfect, or LATEX, (e.g., typewriter) then please contact the Editor of the Journal of Cave and Karst Studies for consideration of a special exception and/or assistance.



PAGE 1

62• Journal of Cave and Karst Studies August 2004 Eko Haryono and Mick Day Landform differentiation within the Gunung Kidul Kegelkarst, Java, Indonesia. Journal of Cave and Karst Studies v. 66, no. 2, p. 62-69. Herbert Lehmann’s (1936) research on the karst of the Gunung Sewu (Thousand Hills) in south-central Java (Figure 1) was the first modern work on humid tropical karst (Sweeting 1981; Jennings 1985) and it made a significant contribution to understanding both the development of the Gunung Sewu landscape itself and tropical karst in general. Subsequent research has revealed that tropical karst morphology varies considerably, particularly as a result of differing geologic environments and hydrologic regimes (Sweeting 1972, 1980; Jennings 1972, 1985; Trudgill 1985; White 1988; Ford & Williams 1989) and the karst of the Gunung Sewu itself demonstrates this differentiation. Lehmann and more recent workers have described the Gunung Sewu landscape as cone(or kegel) karst, characterized by sinusoidal or hemispherical hills (kuppen) interspersed with enclosed star-shaped depressions or interconnected valleys (Lehmann 1936; Flathe & Pfeiffer 1965; Balazs 1968, 1971; Verstappen 1969; Waltham et al 1983). These descriptions of the Gunung Sewu landscape as kegelkarst generalize what is really a variety of different residual hill morphologies, with the conical form not actually being the most characteristic (Flathe & Pfeiffer 1965). The Gunung Sewu karst covers an area of more than 1300 km, and incorporates over 10,000 individual hills (Balazs 1968 estimated 40,000), at densities of about 30/km, whose morphology varies considerably more than previous studies suggest. The diverse forms of residual hills in tropical humid karst are generally considered to be the result of “…structural factors in the broad geomorphological sense.” (Jennings 1985, p. 205). Many individual factors govern carbonate karst development in specific locations, including lithology and structure, which influence the efficacy and the distribution of the dissolution process within the rock mass (Sweeting 1980; Trudgill 1985; White 1988). The objective of this research is to begin to identify the geologic variations and the associated karst landform differentiations within the Gunung Sewu karst, thus refining geomorphological understanding of this classic karst landscape. Because of the paucity of existing data, no specific hypotheses are tested in this initial phase of the research, but it is anticipated that such hypotheses will be developed and tested subsequently. THESTUDYAREAThe research documented herein focuses upon the western two thirds of the broad Gunung Sewu karst area, in the Gunung Kidul Regency of Yogyakarta Special Province, Java, Indonesia between 757' and 812' South latitude (Figure 1). We refer to this study area henceforth as the Gunung Kidul (Southern Hills) karst, reflecting previous and more geographically correct usage (Balazs 1968). The Gunung Kidul area is adjacent to the Indian Ocean on the south central coast of Java (Figure 1). Elevation range is between zero and 400 m above mean sea level, with resurgence springs such as Baron being at sea level and the highest portions centrally located about 25 km from the coastline. LANDFORM DIFFERENTIATION WITHIN THE GUNUNG KIDULKEGELKARST, JAVA, INDONESIAEKOHARYONO Faculty of Geography, Gadjah Mada University, INDONESIAe.haryono@geo.ugm.ac.idMICKDAY Department of Geography, University of Wisconsin-Milwaukee,Milwaukee WI 53201 USAmickday@uwm.edu The Gunung Kidul karst is the western part (65%) of the larger Gunung Sewu (Thousand Hills) karst area, which is generally considered a type example of coneor kegelkarst (Lehmann, 1936). This classification is an over-simplification, however, in that the karst landscape within the Gunung Sewu is considerably differentiated in terms of landform morphology and genesis. In the Gunung Kidul, this differentiation is evident from aerial photographs, which provide basic information about landform patterns, including lineament information. These observations were confirmed by field investigation, which incorporated landform measurement and acquisition of lithological information. These detailed studies distinguish three Gunung Kidul karst subtypes: labyrinth-cone, polygonal, and residual cone karst. The labyrinth-cone subtype occurs in the central Gunung Kidul karst where hard, thick limestones have undergone intensive deformation. Polygonal karst has developed in the western perimeter on hard but thinner limestone beds. The residual cone subtype occurs in the weaker and more porous limestones (wackestones or chalks), despite considerable bed thickness. Figure 1. Location of the Gunung Kidul.

PAGE 2

Journal of Cave and Karst Studies August 2004 • 63 HARYONOANDDAY Physiographically, the Gunung Kidul karst is part of the southern plateau of Java Island (Pannekoek 1948), which extends some 85 km east-west and slopes gently, at approximately a 2% gradient, southward, being marked by a high (25–100 m) cliff along the south coast. Geologically, the study area is dominated by Miocene limestones of the Wonosari Formation, which consists of massive coral reef limestones in the south and bedded chalky limestones in the north (Balazs 1968; van Bemmelen 1970; Waltham et al 1983; Surono et al 1992) (Figure 2). Total thickness exceeds 650 m, and the limestones are underlain by volcanic and clastic rocks (Waltham et al 1983). The coral reef limestone is lithologically highly variable, but dominated by rudstones packstones and framestones Breccias with a clay matrix are not uncommon, biohermal structures are identifiable, and lenses of volcanic ash are interspersed among the carbonates (Waltham et al 1983). The bedded, chalky limestones become more prominent towards the north and northeast, and dominate the Wonosari Plateau (Figure 2). The Wonosari Formation was uplifted during the late Pliocene and/or early Pleistocene and dips gently southward at about a 2% gradient (Balazs 1968; van Bemmelen 1970; Surono et al 1992; Sutoyo 1994). North-south compression associated with tectonic plate convergence produced deformation including intensive northwest-southeast and northeastsouthwest jointing and faulting (Balazs 1968; van Bemmelen 1970; Surono et al 1992; Sutoyo 1994). The structure is most complex along the northern boundary, and the northeastern part was downfaulted, forming the Wonosari Basin, within which karstification is limited. The karst surface within the valleys and depressions is mantled by deeply weathered clays, up to 10 m in thickness, which are remants of volcanic ash intermixed with weathering residue from the limestones (Waltham et al 1983). On the hills, soils are shallow, patchy rendzinas or vertisols, but the karst is intensively cultivated, particularly the red-brown clays in the valleys and depressions, with terracing, irrigation and sophisticated manipulation of available water resources (Uhlig 1980; Urushibara-Yoshino 1993). Karst development in the Gunung Sewu has been influenced by paleoclimatic conditions (Urishibara-Yoshino & Yoshino 1997). Dry valley formation appears to have been associated particularly with the lower sea levels and the cooler, drier conditions of the last glacial stage 18,000 B.P. By contrast, cone karst development was apparently promulgated during subsequent warmer and wetter periods. The prevailing contemporary climate in the Gunung Kidul is strongly influenced by the Northwest and Southeast monsoons, which produce a distinct wet season from October to April and a dry season, which may be extremely arid, between May and September. The annual rainfall is about 2000 mm; records from 14 local rain gauge stations between 1960 and 1997 vary between 1500 mm and 2986 mm annually. An earlier mean annual rainfall, based on 33 years of record, was quoted as 1809 mm (Balazs 1968). Mean annual temperature is about 27 C. Seasonal drought is a serious economic problem, because over 250,000 people live within the Gunung Sewu karst, at a density in excess of 300/km (Uhlig 1980; Waltham et al 1983). METHODOLOGYBroad scale interpretation of the karst landforms was based upon 1:50,000 scale black and white panchromatic aerial photography from September 1993. The aerial photographs were used to produce an uncontrolled photo mosaic, which then was used to identify overall landscape patterns, including visible lineaments, and individual landform morphologies within the study area. On this visual basis, three distinctive broad areas of landform assemblages and patterns were identified and, within each of these, 10, 22 and 29 km sample areas were selected non-randomly, on the basis of photographic quality (absence of clouds) and accessibility, for morphometric analysis and field survey. Valley lineaments were measured from the air photographs, with field verification, and the significance of preferred orientations was tested using one-way analysis of variance. Fieldwork was conducted in November 1999 in order to verify the results of the initial interpretation of the air photographs. Sites were selected non-randomly to correspond with the sample areas within the different landform patterns that had previously been identified on the aerial photographs. The fieldwork involved observation and measurement of individual landforms, together with macroscopic lithological identification and determination of Schmidt Hammer hardness (Day & Goudie 1977; Day 1980, 1982). Rock porosities were determined from thin-section analysis (Curtis 1971). Figure 2. Geology of the Gunung Kidul area (aftervan Bemmelen, 1970).

PAGE 3

64• Journal of Cave and Karst Studies August 2004 LANDFORMDIFFERENTIATIONWITHINTHEGUNUNGKIDULKEGELKARST, JAVA, INDONESIA RESULTSAlthough the Gunung Sewu karst is generally classified as kegelkarst, detailed analysis of the aerial photography and field observation in the Gunung Kidul reveals that there are three distinct landscape subsets, which we refer to as labyrinthcone karst, polygonal karst and residual cone karst. This terminology incorporates a descriptive element (cone) into existing terminology (Ford & Williams 1989). Labyrinth-Cone Karst Labyrinth karst (Figures 3 and 4 [page 66]) “…is a landscape dominated by intersecting solution corridors and solution canyons.” (White 1988, p. 116) or, alternatively, “…aligned or intersecting corridor topographies.” (Ford & Williams 1989, p. 391). Specifics notwithstanding, labyrinth karst development is distinctly linear, incorporating meandering valleys rather than enclosed depressions, and is dominantly controlled by faults or major joints. In the Gunung Kidul area the valley linearity is combined with intervening conical hills, hence we refer to this landscape as labyrinth-cone karst. The labyrinth-cone landscape type is characterized by two series of joint-controlled valleys, which are dry under normal circumstances. In the 29.2 km study area, the dominant trend of the valleys is northwest-southeast, with a secondary trend northeast-southwest (Figure 3). Lineations in the classes 31–40, 41–50, 301–310, 321–330 and 331–340 are statistically significant at the 0.001 level. Valleys extend up to 4.5 km in length, and are typically 50–250 m wide, bordered by steep to moderate slopes on both sides. Most valley thalwegs are thoroughly disordered, with only minimal evidence of descending tributary-trunk sequences. Between the valleys are elongated, interfluvial residual hills, 80–100 m in height, which form long, serrated, ridge-like chains of conical or flattopped hills without intervening closed depressions (Figure 4 [page 66]). Enclosed depressions are uncommon within the labyrinth-cone karst, although some have developed within the valley network, where they tend to be large and elongated (Table 1). The slopes of the residual hills in the labyrinth-cone karst are generally steeper than those in other parts of the Gunung Kidul karst, generally ranging between 60 and 70 degrees. Dry valley sides may be near vertical, although in some localities they grade imperceptibly into the cone slopes. The slope steepness may be attributable to lithological factors. The carbonate lithology of the labyrinth-cone karst comprises floatstones packstones and rudstones which are usually dense, hard limestones. Whereas the mean regional Schmidt Hammer hardness Figure 3. Labyrinth-cone karst: airphotograph (left) and lineaments (right).

PAGE 4

Journal of Cave and Karst Studies August 2004 • 65 HARYONOANDDAY value for the limestones is 40.5 (n=80) for weathered surfaces and 21.2 (n=60) for fresh exposures (Day 1978), the corresponding values for limestones in the labyrinth-cone landscape are 44.3 (n=20) and 24.6 (n=15). Porosity of limestones from the labyrinth-cone karst ranges from 13.0 to 16.6% (n=3), which is quite high for diagenetically mature limestones. More importantly, the bedding is massive, commonly exceeding 5 m. The labyrinth-cone karst is most pronounced in the southern portion of the Gunung Kidul, where the carbonates are most intensively jointed and faulted. This area was subject to the maximum displacement as a result of the compressional stresses associated with the subduction zone of the Australian Plate (Tjia 1966; Dwiyana 1989). Polygonal Karst The most characteristic polygonal karst in Gunung Kidul occurs in the western part of the area (Figures 5 and 6 [page 66]). Polygonal karst is characterized by densely packed or coalesced depressions (cockpits), such that the entire karst landscape, including the residual hills marking the polygonal divides, is consumed by them, and the ratio between closed depression area and karstified area approximates unity (Williams 1971; White 1988). Although owing much to dissolution, the polygonal karst of Gunung Kidul appears to be strongly influenced by fluvial processes and by the general southerly slope of the plateau. Although enclosed depressions dominate the landscape, dismembered meandering valley networks are also present, and these may become activated during intense wet season rains. Whereas surface flow within the depressions is centripetal, flow within the valley segments is dominantly towards the south. Increased discharge from epikarstic springs is of particular importance in generating this channel flow (Haryono, in preparation). Thirty-two springs have been identified to date, 28 close to the margins of the karst, from which they discharge surface runoff in channels, and four in central valleys, where they generate surface runoff that subsequently sinks into valley beds. Polygonal karst is particularly well developed in the western part of the Gunung Kidul karst, where the enclosed depressions in some localities resemble the cockpits of Jamaica and Papua New Guinea (Williams 1971 1972a,b). Elsewhere, the depression slopes are more convex, producing rounded hills, or the sinoid karst of Flathe and Pfeiffer (1965), resembling the “egg-box” terrane described elsewhere (Ford & Williams 1989). Structural control is also evident in the 9.6 km polygonal karst study area (Figure 5), where lineations in the 31–40, 41–50, 51–60 and 311–320 classes are statistically significant at the 0.001 level. Depression slope steepness and morphology are influenced by the spacing of lineaments, and by the relative rates of vertical and lateral corrosion (Tjia 1969). In the field study area in the western Gunung Kidul Table 1. Enclosed depression measurements in labyrinth-cone karst of Gunung Kidul (All measurements in meters). Doline OrderLength RangeMean LengthWidth RangeMean WidthN 0100–23816175–20012442 1150–600445100–20015215 2600–925750175–6253325 31225–16501427250–6504265 Figure 5. Polygonal karst: airphotograph (left) and lineaments (right).

PAGE 5

66• Journal of Cave and Karst Studies August 2004 LANDFORMDIFFERENTIATIONWITHINTHEGUNUNGKIDULKEGELKARST, JAVA, INDONESIA bedding is less massive, commonly on the order of 2 m, with riser heights reflecting those dimensions. There is also a lithological influence, with steeper slopes (mean 31) developed on the harder rudstones and framestones (Schmidt Hammer mean hardness 43.0 weathered, 22.7 fresh, n=15 in both cases) and gentler slopes (mean 15) on softer, impure, marly limestones further north (SH mean hardness 32.6, 19.8, n=10 in both cases). Porosities of the polygonal karst limestones range from 1.1 to 14.0% (n=3). Relative relief ranges considerably, from about 30 m to over 100 m, and enclosed depressions are generally rather smaller than in the labyrinthcone area and less elongated (Table 2). Residual Cone Karst Tower karst consists of residual carbonate hills set in a plain; the residuals may or may not be steep-sided (Ford & Williams 1989). Here we use the term residual cone karst to describe the karst of Gunung Kidul that is characterized by conical isolated hills scattered on a corrosional plain (Figures 7 and 8). Residual cone karst has developed primarily in the northeast of the study area and locally near the south coast where corrosion plains are close to sea level. The main factor governing the development of residual cone karst in the Gunung Kidul karst appears to be lithology. In the 21.5 km study area, bedding generally is not obvious, the limestones being massive, but most of this karst is formed in wackstones which here are relatively soft limestones containing a high percentage of micrite and perhaps best characterized as chalks. In a wet condition, fresh wackstone is easily broken by hand. Mean Schmidt Hammer hardness is 35.0 weathered (n=35), 19.8 fresh (n=20) and porosities are high, ranging from 23.1 to 48.1% (n=3). Closed depressions are not numerous in the residual cone karst, having generally been degraded and coalesced within the larger plain, but such as do occur are broad and shallow, with mean lengths of 1230 m and mean widths of 810 m. Lineations are less conspicuous than in the other landscape types (Figure 7), with only the 31–40 class statistically significant at the 0.001 level in the field study area. Hillslope angles vary from 30 to 40 with a mean height of 90 m (Table 3). DISCUSSIONAlthough the overall karst assemblage in the Gunung Kidul can be described as coneor kegelkarst, a more detailed investigation reveals three subtypes: labyrinth-cone, polygonal, and residual cone karst. These are not randomly distributed throughout the Gunung Kidul area, but are spatially distinct and, on the basis of this preliminary investigation, appear to show a close association with structural and lithological variations in the limestones. Figure 4. Labyrinth-cone karst: ground photograph. Figure 6. Polygonal karst: ground photograph. Figure 8. Residual cone karst: ground photograph.

PAGE 6

Journal of Cave and Karst Studies August 2004 • 67 HARYONOANDDAY The labyrinth-cone subtype occurs in the central part of the Gunung Kidul karst, where hard thick limestones have undergone intensive deformation. This karst sub-type conforms to what Lehmann (1936) termed directed, oriented or gerichteter karst, and it reflects the significance of faulting in the delineation of tropical karst landscapes, as was suggested earlier by Pannekoek (1948). The general north-south alignment mirrors the distribution of depression long axes measured by Quinif and Dupuis (1985). Although the residual hills do not attain the same dimensions or steepness as in the Chinese karst, this landscape resembles Fencong Valley landscape (Lu 1986; Yuan 1991; Smart et al 1986). Polygonal karst develops in the western area on similarly hard but thinner limestone beds. Although the polygonal karst resembles that described elsewhere, many of the residual hills retain their distinctly rounded shape, particularly resembling Table 2. Enclosed depression measurements in polygonal karst of Gunung Kidul (m). Doline OrderLength RangeMean LengthWidth RangeMean WidthN 0100–22516075–22514535 1275–750539200–3753266 2500–1000713200–55045914 3980–14501215350–11507473 Figure 7. Residual cone karst: airphotograph (left) and lineaments (right). Table 3. Enclosed depression measurements in residual cone karst of Gunung Kidul (m). Doline OrderLength RangeMean LengthWidth RangeMean WidthN 0900–2251090775–10258876 1…………0 2…………0 3975–18001230450–10208104

PAGE 7

68• Journal of Cave and Karst Studies August 2004 LANDFORMDIFFERENTIATIONWITHINTHEGUNUNGKIDULKEGELKARST, JAVA, INDONESIA the Chocolate Hills area of Bohol, in the Philippines (Voss 1970). Overall, this sub-type approximates Karst Hill Depression landscape (Lu, 1986; Yuan, 1991; Smart et al 1986). The residual cone subtype occurs in weaker limestones (wackestone) with high porosities but relatively thick beds in the northeast of the study area. The influence of thicker beds on residual cone formation echoes the ideas of Tjia (1969), but the overriding control appears to be the softness and the high porosity of the chalks. Again bearing a striking resemblance to the karst on the periphery of the Chocolate Hills in Bohol, this sub-type resembles a subdued version of the Fenglin Valley landscape of China (Lu 1986; Yuan 1991; Smart et al 1986). It is as yet unclear whether these three different subtypes represent a definite zonation of the overall karst landscape that is related to former surface drainage systems (cp. Smart et al 1986), although this seems quite possible given the existence of obvious valley systems in the contemporary terrain and evidence of previous valley systems (Lehmann 1936; Waltham et al 1983; Quinif & Dupuis 1985). Waltham et al (1983) raised the possibility of the landscape developing by dissection of an anticline, and Quinif and Dupuis (1985) postulated preliminary fluvial development on a Pliocene erosion surface. More recently, Urushibara-Yoshino (1995) and Urushibara-Yoshino and Yoshino (1997) have postulated that the valley systems, and subsequently the cone karst, developed on marine terraces, with the valley systems developed under drier conditions with lower sea levels during the last glacial stage of the Pleistocene, and with karstification more prevalent during wetter interglacial periods. CONCLUSIONSGeneral variation in the landscape morphology, and in the form of individual residual hills, has been observed previously (Tjia 1969; Balazs 1971), although the lithological, structural and hydrologic influences have not been examined in detail before. The evident role of lithology in influencing the karst landscape morphology echoes the results of other studies in tropical karst (e.g. Day 1982; Smart et al 1986), in that the greatest local relief is developed on the limestones with the greatest bed thickness and hardness. The lithological heterogeneity is in contrast to earlier accounts of the carbonate geology, which suggested that homogeneity was the rule (Lehmann 1936), and the landform differentiation also reveals greater geological influence than recognized previously (Waltham et al 1983). Structural variability is also greater than was previously acknowledged, although broadly it is the northeastsouthwest and northwest-southeast lineations that are statistically significant. In addition to lithology and geological structure, regional slope also plays a role in influencing karst landscape development and individual karst landforms. The southerly regional 2% slope controls karst development indirectly through promoting slope-directed runoff, which results in linear depressions or valleys being more numerous than enclosed depressions. This is particularly notable in the southern part of the Gunung Kidul. In this context, Lehman’s (1936) model of karst development progressing from an initial stage dominated by surface runoff and surface valleys to a later stage in which the valleys become increasingly underdrained and dismembered by the development of enclosed depressions seems not inappropriate, particularly given the empirical evidence (Quinif & Dupuis 1985) and the supporting theory put forward since (Smith 1975; White 1988). ACKOWLEDGEMENTSThis article derives from research in progress by E. Haryono on epikarst geomorphology and hydrology. Thanks are due to the P4M Directorate of the Government of Indonesia, who provided funding under the Competitive Research Grant scheme (PHB) XVIII/1. We acknowledge the constructive comments on earlier drafts by Suratman W.S., Sunarto M.S., Pramono H., D. Ford and an anonymous reviewer. Latif and Swarsono assisted with fieldwork and laboratory analysis. REFERENCESBalazs, D., 1968, Karst Regions in Indonesia: Karszt-Es Barlangkutatas, Volume V. Budapest, Globus nyomda, 61 p. Balazs, D., 1971, Intensity of the Tropical Karst Development Based on Cases of Indonesia, Karszt-Es Barlangkutatas, Volume VI. Budapest, Globus nyomda, 67 p. Curtis, B.F., 1971, Measurement of porosity and permeability, in Carver, R.E. (ed.), Procedures in Sedimentary Petrology: New York, Wiley, p. 335–363. Day, M.J., 1978, The morphology of tropical humid karst with particular reference to the Caribbean and Central America [Ph.D. thesis]: Oxford University, 611 p. Day, M.J., 1980, Rock hardness: field assessment and geomorphic importance: The Professional Geographer, v. 32, no. 1, p. 72–81. Day, M.J., 1982, The influence of some material properties on the development of tropical karst terrain: Transactions, British Cave Research Association, v. 9, no. 1, p. 27–37. Day, M.J. & Goudie, A.S., 1977, Field assessment of rock hardness using the Schmidt Test Hammer: British Geomorphological Research Group Technical Bulletin, no. 18, p. 19–29. Dwiyana, R.A., 1989, Geologi dan analisis pengembangan wilayah daerah kemadang dan sekitarnya Kecamatan Tepus, Kabupaten Gunung Kidul [M.Sc. thesis]: Yogyakarta, Universitas Gadjah Mada, 262 p. Flathe, H. & Pfeiffer, D., 1965, Grundzuge der morphologie, Geology und Hydrogeologie im Karstgebiet Gunung Sewu (Java, Indonesien): Geologisches Jahrbuch, v. 83, p. 533–562. Ford, D.C. & Williams, P.W., 1989, Karst Geomorphology and Hydrology: London, Unwin Hyman, 601 p. Jennings, J.N., 1972, The character of tropical humid karst: Zeitschrift fur Geomorphologie, N.F., v. 16, no. 3, p. 336–341. Jennings, J.N., 1985, Karst Geomorphology: Oxford, Basil Blackwell, 293 p. Lehmann, H., 1936, Morphologische Studien auf Java: Geographische Abhandlungen, Series 3, no. 9, p. 1–114. (Also excerpted in Sweeting 1981, p. 320-328) Lu, Y., 1986, Karst in China, landscape, types, rules: Beijing, Geological Publishing House, 172 p.

PAGE 8

Journal of Cave and Karst Studies August 2004 • 69 HARYONOANDDAY Pannekoek, A.J., 1948, Enige Karstterreinen in Indonesie: Tijdschrift Nederlandschaft Aardrijkskundl Genootscher, v. 65, p. 209–213. Quinif, Y. & Dupuis, C., 1985, Un karst en zone intertropicale: le Gunung Sewu a Java: aspects morphologiques et concepts evolutifs: Revue de Geomorphologie Dynamique, v. 34, no. 1, p. 1–16. Smart, P.L., Waltham, A.C., Yang, M. & Zhang, Y., 1986, Karst geomorphology of western Guizhou, China: Transactions, British Cave Research Association, v. 13, no. 3, p. 89–103. Smith, D.I., 1975, The problems of limestone dry valleys — implications of recent work in limestone hydrology, in Peel, R.F., Chisholm, M. and Haggett, P., (eds.), Bristol Essays in Geography: London, Heineman, p. 130–147. Surono, B.T., Sudarno, I. & Wiryosujono, S., 1992, Geology of the SurakartaGiritontro Quadrangles, Java: Bandung, Geological Research and Development Center, Indonesia, scale 1:100,000, 2 sheets. Sutoyo, 1994, Sikuen stratigrafi karbonat Gunung Sewu, in Busono, I., Syarifudin, N. & Alam H. (eds.), Proceedings pertemuan ilmiah tahunan IAGI ke 23: Jakarta, Indonesian Geologists Society, p. 67–76 Sweeting, M.M., 1972, Karst Landforms: London, MacMillan, 362 p. Sweeting, M.M., 1980, Karst and climate: a review: Zeitschrift fur Geomorphologie N.F., Supplementbande 36, p. 203–216. Sweeting, M.M., (ed.), 1981, Karst Geomorphology: Stroudsburg PA, Hutchinson Ross, Benchmark Papers in Geology, v. 59, 427 p. Tjia, H.D., 1966, Structural Analysis of the Pre-Tertiary of the Lukulo Area, Central Java [Ph.D. thesis], Bandung, Institute of Technology, 110 p. Tjia, H.D., 1969, Slope development in tropical karst: Zeitschrift fur Geomorphologie, N.F., v. 13, p. 260–266. Trudgill, S.T., 1985, Limestone Gemorphology: London, Longman, 196 p. Uhlig, H., 1980, Man and tropical karst in Southeast Asia: GeoJournal, v. 4, n. 1, p. 31–44. Urushibara-Yoshino, K., 1993, Human impact on karst soils: Japanese and other examples, in Williams, P.W., (ed.), Karst Terrains: Environmental Changes and Human Impact: Cremlingen, Catena Verlag, Catena Supplement 25, p. 219–233. Urushibara-Yoshino, K., 1995, Environmental change in the karst areas on the island of Java: Journal of the Faculty of Letters, Komazawa University, v. 53, p. 85–97. Urushibara-Yoshino, K. & Yoshino, M., 1997, Palaeoenvironmental change in Java Island and its surrounding areas: Journal of Quaternary Science, v. 12, no. 5, p. 435–442. van Bemmelen, R.W, 1970, The Geology of Indonesia, Volume 1A, General Geology: The Hague, Martinus Nijhoff, 732 p. Verstappen, H.T., 1969, The state of karst research in Indonesia, in Stelcl, O., (ed.), Problems of the Karst Research, Brno, Ceskoslovenska Akademia Sciencias, p. 139–148. Voss, F., 1970, Typische oberflachenformen tropischen kegelkarstes auf den Philippinen. Geographische Zeitschrift, v. 59, p. 214–227. Waltham, A.C., Smart, P.L., Friederich, H., Eavis, A.J. & Atkinson, T.C., 1983, The caves of Gunung Sewu, Java: Cave Science, v. 10, no. 2, p. 55–96. Williams, P.W., 1971, Illustrating morphometric analysis of karst with examples from New Guinea: Zeitschrift fur Geomorphologie, N.F, v. 15, p. 40–61. Williams, P.W., 1972a, Morphometric analysis of polygonal karst in New Guinea: Geological Society of America Bulletin, v. 83, p. 761–796. Williams, P.W., 1972b, The analysis of spatial characteristics of karst terrains, in Chorley, R.J., (ed.), Spatial Analysis in Geomorphology: London, Methuen, p. 135–163. White, W.B., 1988, Geomorphology and Hydrology of Karst Terrains: Oxford, Oxford University Press, 464 p. Yuan, D., 1991, Karst of China: Beijing, Geological Publishing House, 224 p.



PAGE 1

56• Journal of Cave and Karst Studies August 2004 LONGANDDEEPCAVES CAVENAMECOUNTRYSTATEDEPTHLENGTHMETERSMETERS1Voronja Cave (Krubera Cave)GeorgiaAbkhazia1710.0----2Lamprechtsofen Vogelschacht Weg SchachtAustriaSalzburg1632.050000 3Gouffre Mirolda/Lucien Bouclier FranceHaute Savoie1626.013000 4Torca del Cerro del Cuevon (T.33)-Torca de las SaxifragasSpainAsturias1589.04000 5Sarma Georgia Abkhazia1543.06370 6Reseau Jean Bernard FranceHaute Savoie1535.020536 7Cehi 2 “la Vendetta”SloveniaRombonski Podi1533.05061 8Shakta Vjacheslav Pantjukhina GeorgiaAbkhazia1508.05530 9Sistema Cheve (Cuicateco) Mexico Oaxaca1484.026194 10Sistema HuautlaMexicoOaxaca1475.055953 11Sistema del Trave Spain Asturias1441.09167 12Sima de las Puertas de Illaminako Ateeneko Leizea (BU.56)Spain/FranceNavarra/Nafarroa1408.014500 13Sustav Lukina jama Trojama (Manual II) CroatiaMt.Velebit1392.01078 14Evren Gunay Dudeni (Sinkhole) TurkeyAnamur1377.0----15Sniezhnaja-Mezhonnogo (Snezhaya) GeorgiaAbkhazia1370.019000 16Sistema Aranonera (Sima S1-S2)(Tendenera connected)SpainHuesca1349.036468 17Gouffre de la Pierre Saint Martin France/SpainPyrenees-Atlantiques1342.053950 18Siebenhengste-hohgant HohlensystemSwitzerlandBern1340.0145000 19Slovacka jama CroatiaMt.Velebit1320.02519 20Abisso Paolo RoversiItalyToscana1300.04000 21Cosanostraloch-Berger-Platteneck HohlesystemAustriaSalzburg1291.030076 22Cueva CharcoMexicoOaxaca1278.06710 23Gouffre Berger Gouffre de la FromagereFranceIsere1271.031790 24Neide Muruk CavePapua New GuineaNew Britain1258.017000 25Torca dos los RebecosSpainAsturias1255.02228 26Pozo del MadejunoSpainLeon1252.02852 27Crnelsko brezno (Abisso Veliko Sbrego) SloveniaRombonaki Podi 1241.08090 28Vladimir V. Iljukhina System Georgia Abkhazia1240.05890 29Sotano AkematiMexicoPuebla1226.04918 30Kihaje Xontjoa MexicoOaxaca1223.031373 31Schwer-hohlensystem (Batmanhole) Austria Salzburg1219.06101 32Abisso Ulivifer (Olivifer)ItalyToscana1215.010000 33Gorgothakas Greece Crete1208.0----34Dachstein-Mammuthohle AustriaOberosterreich1199.057583 35Complesso del Monte Corchia (Fighiera,Farol.) Italy Toscana1190.052300 36Cukurpinar Dudeni Turkey Anamur1190.03550 37Vandima SloveniaRombonski Podi1182.0 2800 38JubilaumsschachtAustriaSalzburg1173.02380 39Gouffre du Bracas de Thurugne 6 (Reseau de Soudet) France/SpainPyrenees-Atlantiques1170.010340 40Abisso Vive le Donne Italy Lombardia1170.03800 41Anou Ifflis AlgeriaBouira1170.02000 42Sima 56 de Andara(Torca del Cueto de los Senderos)SpainCantabria1169.05620 43Torca Idoubeda SpainAsturias1167.0----44Sistema de las Fuentes de Escuain(Badalona B15-B1) Spain Huesca1151.011450 45Tanne des Pra d'Zeures France Haute1148.03900 46Complesso del Foran del Muss Italy Friuli 1140.0 20000 47Sistema del (Pozu) Xitu (Jitu) Spain Asturias1135.06100 48Sistem Molicka Pec Slovenia DleskovskaPlanto1130.03827 49Abisso SaragatoItalyToscana1125.06000 50Arabikskaja (Kuibyshevskaja/Genrikhova Bezdn)GeorgiaAbkhazia1110.03250 51Kazumura Cave (Lava Tube)U.S.A.Hawaii1101.565500 52 Schneeloch AustriaSalzburg1101.04200 53Sima G.E.S.M.de los Hoyos del PilarSpain Malaga1101.0 3000 54 Gouffre des Partages France Pyrenees-Atlantiques1097.023920 55 Zoou Cave (Dzou)GeorgiaAbkhazia1090.06000 DEEPCAVES OFTHE WORLDCompiled by Bob Gulden

PAGE 2

Journal of Cave and Karst Studies August 2004 • 57 LONGANDDEEPCAVES CAVENAMECOUNTRYSTATELENGTHDEPTHMETERSMETERS1Mammoth Cave SystemU.S.A.Kentucky579364115.5 2Optimisticeskaja (Gypsum) Ukraine Ukrainskaja21400015.0 3 Jewel Cave U.S.A. South Dakota209170192.7 4 Holloch SwitzerlandSchwyz189026940.0 5 Lechuguilla Cave U.S.A. New Mexico180096489.0 6 Wind Cave U.S.A. South Dakota179442202.4 7 Fisher Ridge Cave System U.S.A. Kentucky172747108.6 8 Siebenhengste-hohgant Hohlensystem Switzerland Bern1450001340.0 9 Sistema Ox Bel Ha (Under Water) Mexico Quintana Roo13343933.5 10 Ozernaja (Gypsum) Ukraine Ukrainskaja1220008.0 11 Gua Air Jernih (Clearwater Cave-Black Rock) Malaysia Sarawak109000355.1 12 Reseau Felix Trombe/Henne-Morte France Haute-Garonne105767975.0 13 Toca da Boa Vista Brazil Bahia10200050.0 14 Systeme de Ojo Guarena Spain Burgos100400163.0 15 Sistema Purificacion Mexico Tamaulipas93755953.0 16 Bullita Cave System (Burke's Back Yard) Australia Northern Territory9298523.0 17 Zolushka (Gypsum) Moldova/Ukraine Moldarskaja9020030.0 18 Hirlatzhohle Austria Oberosterreich866061009.0 19 Raucherkarhohle Austria Oberosterreich82686758.0 20 Friars Hole Cave System U.S.A. West Virginia73288191.4 21 Easegill System United Kingdom Yorkshire70500211.0 22 Ogof Draenen United KingdomSouth Wales6612097.8 23 Kazumura Cave (Lava Tube) U.S.A. Hawaii655001101.5 24 Organ (Greenbrier) Cave System U.S.A. West Virginia63569148.1 25 Sistema Nohoch Nah Chich (Under Water) Mexico Quintana Roo61143 71.6 26 Reseau de l'Alpe France Isere Savoie60247655.0 27 Cueva del Valle (Red Del Silencio) Spain Cantabria60223502.0 28 Bol’shaja Oreshnaja (Conglomerate) RussiaRussia58000240.0 29 Barenschacht SwitzerlandBern57800946.0 30 Dachstein-Mammuthohle AustriaOberosterreich575831199.0 31 Botovskaya RussiaLrkutsk57256 6.0 32 Arresteliako ziloa(Souffleur de Larrandaburu) FrancePyrenees57061835.0 33 Kap-Kutan/Promezhutochnaja TurkmenistanTurkistan57000 310.0 34 Cenote Dos Ojos (Under Water) MexicoQuintana Roo56671119.2 35 Schwarzmooskogelhoehlensystem-Kaninchohle AustriaSteiermark560731030.0 36 Sistema Huautla MexicoOaxaca559531475.0 37 Systeme du Granier FranceIsere/Savoie55327564.0 38 Kolkbluser-Monster-Hohlensystem AustriaSalzburg55000711.0 39 Mamo Kananda Papua New GuineaSouthern Highlands54800528.0 40 Gr. Caverna de Palmarito CubaPinar del Rio540000.0 41 Gouffre de la Pierre Saint Martin France/SpainPyrenees-Atlantiques539501342.0 42 Blue Spring Cave (Saltpeter) U.S.A.Tennessee5343171.0 43 Complesso del Monte Corchia (Fighiera,Farol.) ItalyToscana523001190.0 44 Martin Ridge System (Wig.,Jackpot,Martin) U.S.A.Kentucky5188495.7 45 Reseau de la Dent de Crolles FranceIsere50101673.0 46 Lamprechtsofen Vogelschacht Weg Schacht AustriaSalzburg500001632.0 47 Ogof Ffynnon Ddu United KingdomSouth Wales50000308.0 48 Carlsbad Cavern U.S.A.New Mexico49680 315.5 49 Sima del Hayal de Ponata (SI-44,SI-57,SR-7) SpainAlava/Vizcaya49000415.0 50 Sistema Rubicera-Mortero de Astrana SpainCantabria48000546.0 51 Santo Tomas (gran caverna de) CubaPinar del Rio460000.0 52 Crevice Cave U.S.A.Missouri453850.0 53 Grotte de Saint-Marcel FranceArdeche45247233.0 54 Cumberland Caverns (Saltpeter)U.S.A.Tennessee4444461.0 55 Scott Hollow Cave U.S.A.West Virginia43452174.0 LONGCAVES OFTHE WORLDCompiled by Bob Gulden



PAGE 1

Journal of Cave and Karst Studies August 2004 • 39 Katie Schneider and David C. Culver Estimating subterranean species richness using intensive sampling and rarefaction curves in a high density cave region in West Virginia. Journal of Cave and Karst Studies v. 66, no. 2, p. 39-45. Biological sampling and biodiversity mapping have become key components to the understanding of subterranean ecosystems in the face of environmental and anthropogenic threats (Culver et al 2001), and promoting the assessment of the status and vulnerability of cave species facilitates their preservation and protection. Mapping biodiversity is an important step in this endeavor, serving as a tool for education, research, and conservation planning (Culver et al 2001). The information incorporated into maps of species richness in caves can come from a variety of sources, such as inventory or census information or from known occurrence records (Conroy & Noon 1996). The accuracy of these maps and eventual protection of biological diversity therefore hinges on the completeness of these data (Kodric-Brown & Brown 1993, Keating et al 1998). However, there is an inherent bias in relying on occurrence records and compiled lists, in that most of these lists are incomplete and not all caves have been carefully and repeatedly studied, if they have been studied at all. In addition, sites that have been sampled but in which no species were found are typically not displayed on biodiversity maps, making them indistinguishable from unsampled sites (Deharveng 2001). Sampling incompleteness can result in misleading patterns in community structure and species rarity, as Kodric-Brown and Brown’s (1993) study of the effect of different levels of sampling of fish species richness in Australian desert springs shows. This is often compounded by sampling bias towards accessible sites, such as cities and highways (Bojrquez-Tapia et al 1994) and field stations (Pearson & Cassola 1992), as well as bias towards certain taxa (Bojrquez-Tapia et al 1994), that affect the reliability of occurrence data (BojrquezTapia et al 1995). As a consequence of incomplete sampling, not all species may be represented, leading to inaccurate estimates of species richness (Nichols et al 1998), and possible poor decision making in conservation planning and management (Conroy & Noon 1996). Inventories of subterranean fauna may be so inadequate that many species may go extinct before being discovered (e.g., Croatia [USAID 2000]). Thus far, richness estimates for cave faunas have been derived based on extrapolation from a small number of wellstudied caves, which often tend to be the largest and most accessible (Culver et al in press). It is unclear how inaccurate and/or misleading our knowledge of subterranean biodiversity may be. To date, no cave area has been sampled completely, except possibly for the Canary Islands (Izquierdo et al 2001). The Derbyshire region of Britain has had 27% of the 210 caves sampled (Proudlove 2001) and may be the second most completely sampled region. In West Virginia, an area thought to be well-sampled (Culver & Holsinger 1992), less than 10% of the caves have been biologically sampled (Krow & Culver 2001), even though between 1962 and 1973, 152 caves were biologically investigated (Holsinger et al 1976). When variation between sites in species richness is great (as is the case for West Virginia caves), a larger sampling effort is required (Hammond 1994) to estimate total species richness. Sampling effort must be sufficient to minimize sampling bias in order to determine if inventory data are accurate (Hammond 1994). It is therefore critical that the sample of caves be large and unbiased (Krow & Culver 2001). We sampled 65 caves within a high cave density and species-rich area of West Virginia. By sampling a large percentage of caves in an area, it is possible to discover how many species are missed when only a portion of the caves are sampled. We then used these data to examine how many caves need to be sampled to get an accurate estimate of species richness for the study area. Lastly, we predict how many species are indeed present in the study site using rarefaction curves and equations based on species rarity. ESTIMATINGSUBTERRANEAN SPECIES RICHNESS USING INTENSIVE SAMPLINGAND RAREFACTION CURVES IN A HIGH DENSITYCAVE REGION IN WESTVIRGINIAKATIESCHNEIDER1ANDDAVIDC. CULVER Department of Biology, American University, Washington DC 20016 USA Species richness in a group of caves in the 21.25 km corner of the USGS 7 minute Williamsburg quadrangle, West Virginia, was investigated to (1) increase our knowledge of species richness for this area, (2) determine how many caves need to be sampled to achieve an accurate estimate of species richness and (3) estimate how many species are present in this area. Eighteen subterranean invertebrate species were collected from 65 caves within the study area. Seven caves were needed to collect 95% of the species. By sampling only the largest seven caves, 89% of the species were captured. However, the species accumulation curve did not reach an asymptote, and estimations based on species rarity show that half of the species were not collected at all. Therefore, the observed patterns should be interpreted with caution, and more data are needed. 1Now at: Department of Biology, 1204 Biology-Psychology Building, University of Maryland, College Park, MD 20742, USAkatie2@umd.edu

PAGE 2

40• Journal of Cave and Karst Studies August 2004 ESTIMATINGSUBTERRANEANSPECIESRICHNESS STUDYSITEWest Virginia is reported to have 3754 caves (Jones 1997). Of these 195 (5.2%) are reported to have obligate cavedwelling species (Culver, unpublished data). There are 76 known obligate cave species reported from the state (Culver & Sket 2000). As far back as the 1950s, it was acknowledged that Greenbrier County, West Virginia, was rich in cave numbers, possessing some 105 caves — one quarter of all of the caves of West Virginia (Davies 1958). Today, the number of known caves from both locations has increased tenfold, with 1030 known caves from the county (Jones 1997). Greenbrier County is also a national hotspot of cave biodiversity (Culver et al 2000). An area centered around the Buckeye Creek Basin in northeastern Greenbrier County (Fig. 1) was chosen as the study site in part due to its high concentration of caves. The study site was chosen because of its high cave density on a limited number of properties, its proximity to the West Virginia Association of Cave Studies field station, and our good working relationship with the local landowners. We had access to all areas within the study area. The invertebrate fauna of the study area was poorly known, and no systematic survey had been performed prior to this study (Fong & Culver 1994). Only 9 of the 148 caves previously had been biologically sampled, and 10 cave-limited species were reported from this study site (Holsinger et al 1976, Fong & Culver 1994). METHODSCave locations were obtained from files of the West Virginia Speleological Survey (WVSS). The following criteria were used to select caves to be sampled. First, the caves must be located within the study area where 3' of latitude and 2'30" of longitude in the northeast corner of the USGS 7 minute Williamsburg quadrangle was determined to be within the study area and representing an area of 21.25 km. Second, cave enterability was usually assessed by inspection and in some cases from the descriptions in Dasher and Balfour (1994). Caves were located in the field using UTM coordinates, with a map provided by WVSS, and with the help of cavers and local residents. If more than two hours were spent unsuccessfully searching, the cave was classified as “failed to locate.” Located caves were included in the study if they could be safely entered and had a dark zone. Entrances of study caves were then flagged and UTM coordinates recorded to facilitate relocating caves during the sampling period. Sampling took place between June 3, 2002 and July 21, 2002. Once inside a cave, a visual census of organisms on the walls, floor, ceiling, and aquatic areas (if present) was performed for one-person hour, recording all species found. Only potential troglobites and stygobites were collected and these were placed in 70% ethanol. As a general rule, one to five specimens, an adequate number for positive identification to species, were hand-collected from each cave. Terrestrial pitfall traps were constructed of 150 mLplastic jars filled with isopropyl alcohol and covered with 7.5 cm 7.5 cm pieces of 1 cm hardware cloth to exclude cave-crickets, salamanders, and other larger animals. Pitfall traps were baited with limburger cheese and placed in soft mud banks, where mud banks were present. Aquatic traps were constructed of an ordinary kitchen scrubber with a mesh size of approximately 1 cm. The tube of mesh was baited with raw shrimp and tied at both ends. Aquatic traps were placed in slow-running shallow streams or rimstone pools. Traps were placed near areas of high abundance and diversity, as determined by visual sampling. Generally, one terrestrial and one aquatic (if water was present) trap was placed in each cave. If the cave contained various habitats (e.g., rimstone pools and streams, mudbanks and silty shores, etc.), one to three additional traps of each type were placed. This was the case for most caves more than 100 m long. Traps remained in place for two to three days. Prior to Figure 1. Locatormap of the study site. The study site is a ca. 20 km area located north of Lewisburg, in Greenbrier County, West Virginia. Of the 148 caves located in the study site, 73 were enterable. Of these 73 caves, 65 were sampled during this investigation.

PAGE 3

Journal of Cave and Karst Studies August 2004 • 41 SCHNEIDERANDCULVER trap removal, the surrounding area was examined and additional individuals attracted to the bait were collected. Animals from pitfall traps were transferred from isopropyl alcohol to ethanol in the laboratory. Specimens were sorted and identified either by using keys or sending specimens to expert taxonomists. All species remained in ethanol except for the beetles which were transferred to Barber’s fluid, a relaxant used to prevent brittleness and breakage of the specimens (Borror et al 1989). Maps were created using ArcMap GIS (Environmental Systems Research Institute, Redlands, CA, USA) and the UTM data were transformed using Datumpro (Linden Software Ltd, Lincs, United Kingdom). Data were analyzed using Excel (Microsoft Corporation, Redlands, CA, USA), SPSS (SPSS Inc., Chicago, IL, USA), and JMP(SAS Institute, Cary, NC, USA). Rarefaction curves were made by repeatedly sampling all of the collected species at random (Gotelli & Colwell 2001). Rarefaction curves indicate the expected number of species from a collection of random samples and represent what is statistically expected from the accumulation curve (Gotelli & Colwell 2001). With rarefaction curves, differences are no longer attributed to sample size. Rarefaction curves were created using EstimateS software (Colwell, 1997; http://viceroy.eeb.uconn.edu/estimates). Due to incomplete sampling, estimators have been derived to predict the true number of species based on rare species in a sample (Colwell & Coddington 1994). This was done using the equation from Chao (1984), where n is the number of samples. No direct formula for the calculation of the variance is available. We used the algorithm of Csuti et al (1997) to find the minimum number of caves needed to “capture” 95% of the reported troglobites and stygobites. RESULTSThe WVSS (West Virginia Speleological Survey) database showed 148 caves, pits and FROs (for the record only) in the study site. We were able to locate and enter 65 of these caves in the summer of 2002 (Fig. 1). Of the remaining 83 locations, we were unable to locate a physically enterable entrance for 75 of them either because of faulty location data or because the WVSS database contained non-cave karst features. The eight additional enterable caves were located too late to be included in the study (January 2003), but are worth revisiting and sampling in future studies. The average cave length was 165.3 m 64.6 m and varied between 2 m and 3719 m. Most of the caves were short, with 44 of the 65 caves being less than 30 m long. Cave depth averaged 9.8 m 1.4 m with a range of one to 30 m. Twenty-one of 33 caves for which depth data were available were less than 10 m deep. All caves had terrestrial habitats, but only 38 had aquatic habitats. An aquatic habitat was defined as an aquatic area in which a trap could be placed. Overall, six classes, 11 orders, 12 families, 14 genera, and 18 species were collected (Table 1). The two most commonly encountered orders were the Collembola (springtails) and Coleoptera (beetles), followed by the Amphipoda (amphipods), Chordeumatida (millipedes), and Diplura (diplurans). Three rarefaction curves are shown in Figure 2 — one for all caves, one for caves less than 15 m in length, and one for caves greater than 15 m in length. All three curves showed no sign of reaching an asymptote, but the rate of species accumulation for caves greater than 15 m was more than twice that of caves less than 15 m. Two estimates of total species richness are provided in Table 2. The two estimates are 36 and 48, and both are considerably higher (between two and three times) than the observed number of 18. Table 3 shows that seven caves are needed to find 17 of the 18 reported species and suggests which caves need to be sampled in order to collect 95% of the total species collected in the study. If the seven largest caves are used, the result is nearly as good, with 16 of the 18 reported species found in these caves, which implies that 89% of the total species are collected if the largest seven caves are sampled (Table 4). The largest caves themselves are not arranged in order of size but rather in order of their successive contribution of new species, so that in fact that last two caves added to the analysis do not result in the inclusion of any new species. (1) where Sobs is equal to the number of species observed in a sample, L is the number of observed species represented by a single individual (i.e., singletons), and M is the number of observed species represented by two individuals in the sample (i.e., doubletons). The variance on this equation was estimated as (2) Colwell and Coddington (1994) recommend the application of Burnham and Overton’s (1978) jackknife estimators in order to reduce estimation bias in estimating species richness. We calculated this second-order jackknife estimate: (3)

PAGE 4

42• Journal of Cave and Karst Studies August 2004 ESTIMATINGSUBTERRANEANSPECIESRICHNESS DISCUSSIONPrior to this study, knowledge of species richness from caves in this ca. 20 km area was based on sampling of 9 caves (Holsinger et al 1976; Fong & Culver 1994). In most of these collections, techniques other than hand sampling were not used (Holsinger et al 1976). There are striking omissions from the previous faunal list, such as cave snails and flatworms, most likely due to a lack of an adequate census in the area (Fong & Culver 1994). There were 10 species known prior to our study. Our study also had omissions. In spite of extensive hand collections and trapping, no spiders were recovered during our study. As a result of our efforts, the number of caves sampled in this area increased from nine to 65, the number of species recorded from this area increased from 10 to 18 (Table 1). When looking only at the nine caves that were resampled in the current investigation, seven species records were confirmed, and eight new localities were added for species previously reported from the study area. New taxonomic groups were also collected from these nine caves, including planarians, diplurans, collembola, millipedes and mites. As a result of this study, there were 93 new records of species, including eight new species, added to this roughly 20 km area. The eight species new to the study area are all known from West Virginia. Among the most notable species that we collected was the undescribed dipluran, Eumesocampa sp., which has only been collected from one other cave (Steeles Cave, Monroe County, West Virginia). Recent attempts to recollect this species in Steeles Cave have not been successful (L. Ferguson, pers. comm., 2002). The findings of this research showed that subterranean biodiversity for this area had been greatly underestimated. Clearly, by focusing solely on the minimal information known from nine of 148 caves, many species would be unreported and the distributions of others incompletely known. Ideally, homogeneous sampling and intensive sampling are preferred; however, subterranean areas are difficult and expensive to sample and the risk of overcollecting is usually a concern. Therefore, it is necessary to know the minimal sample size needed to get an accurate estimation of species diversity for an area. Using the “simple greedy” algorithm of Csuti et al (1997), we found that only a small number of caves need to be sampled in order to collect all known species in the study area. Table 1. Cave-limited species encountered during the study and theirhabitats.ClassOrderFamilySpeciesReferenceHabitat TurbellariaTricladidaKenkiidae Macrocotyla hoffmasteria(Hyman, 1954)Aquatic MolluscaGastropodaHydrobiidae Fontigens tartareaaHubricht, 1963Aquatic CrustaceaAmphipodaCrangonyctidae Stygobromus emarginatus (Hubricht, 1943)Aquatic CrustaceaAmphipodaCrangonyctidae Stygobromus spinatus (Holsinger, 1967)Aquatic CrustaceaIsopodaAsellidae Caecidotea holsingeri (Steeves, 1963)Aquatic CrustaceaDecapodaCambaridae Cambarus nerterius Hobbs, 1964Aquatic DiplopodaChordeumidaCleidogonidae Pseudotremia sp. nov.aƒ Terrestrial DiplopodaChordeumidaCleidogonidae Pseudotremia sp ƒ Terrestrial InsectaDipluraCampodeidae Eumesocampa sp.aƒ Terrestrial InsectaDipluraCampodeidae Litocampa fieldingaea(Cond, 1949)Terrestrial InsectaCollembolaSminthuridae Arrhopalites clarusaChristiansen, 1966Terrestrial InsectaCollembolaEntomobryidae Pseudosinella gisini Christiansen, 1960Terrestrial InsectaCollembolaEntomobryidae Sinella hoffmaniaWray, 1952Terrestrial InsectaColeopteraCarabidae Pseudanopthalmus grandis Valentine, 1931Terrestrial InsectaColeopteraCarabidae P. higginbothami Valentine, 1932Terrestrial InsectaColeopteraCarabidae P. hypertrichosis Valentine, 1931Terrestrial ArachnidaAcariRhagidiidae Rhagidia variaaZacharda, 1985Terrestrial ArachnidaPseudoscorpionidaChthoniidae Kleptochthonius henroti (Vachon, 1952)Terrestrial Note: Pseudotremia fulgida was previously reported from the study area (Loomis, 1943) but was not collected during the present study.aNot previously recorded from study area. Figure 2. Rarefaction curves fornumberof caves versus numberof species, forall caves ( n = 65), caves less than or equal to 15 m ( n = 32), and caves greaterthan 15 m ( n = 33). Curves generated using EstimateS with the patchiness parameterset to 0.8 as recommended in Gotelli and Colwell (2001).

PAGE 5

Journal of Cave and Karst Studies August 2004 • 43 SCHNEIDERANDCULVER We found that seven caves were sufficient to capture 95% of the known species (Table 3) although a prioi knowledge of which seven caves to sample is lacking. However, using cave length as a surrogate for species richness gives nearly the same results. By examining only the largest seven caves, 89% of the species were collected. This finding has conservation implications, as many of the species (including many of the rare species) could be protected by protecting the largest caves. Izquierdo et al (2001), in a conservation application of Csuti et al .’s greedy algorithm, proposed that in order to maximize the number of species protected in a limited number of sites, conservation decisions should be focused on the cave with the most species, followed by the cave with the most species different from the first cave, and so on. If conservation decisions in the study area were indeed based according to this standard, then conservation priority would be given to the cave with the most species, in this case Buckeye Creek Cave. The next cave of concern would be the cave with the largest number of species different from the first, in this case, Matt’s Black Cave. Here, following the guidelines of Izquierdo et al and their application of the greedy algorithm, it would only take seven caves to protect 95% of the species. With this approach, many of the largest caves (and the species therein) would be protected. Protecting the largest caves does result in the protection of the greatest biodiversity, and the species accumulated in the larger caves represent most of the species found in the smaller caves. Rarefaction curves generated from our data did not reach an asymptote, and the curve rose more steeply for larger caves than for smaller caves (Fig. 2) because species accumulated more quickly in larger caves. When no new taxa are added, an asymptote should, in principle, be reached (Gotelli & Colwell 2001). Due to the failure to reach an asymptote, total troglobitic and stygobitic species richness was estimated using equations provided by Colwell and Coddington (1994). Using Chao’s estimate, S2* total species richness was 36 species, and it was 48 using the second-order jackknife estimate, S4* (Table 2). Colwell and Coddington point out that in practice, the upper bound of the estimate for S4* is approximately twice the observed number, i.e., 36, and the upper estimate for S2* is approximately half the square of the observed number, i.e., 81. This in turn suggests that S4* estimate is unreliable. If we use the S2* value of 36 as the best estimate of the number of species, we have found only half of the species. How did nearly half of the species evade collection? Over 90% of sampled caves in West Virginia have at least one troglobitic species (Culver et al 2004). Here, only 69% of the sampled caves (45 of 65 caves) had at least one troglobite/stygobite collected and approximately one-third of the caves sampled yielded no troglobites or stygobites at all. Repeated visits often are necessary to collect all of the species found in a single cave. In one Italian cave, for example, Fabio Stoch determined that it took six trips to collect all 12 stygobites present (quoted in Culver et al in press). In the nine caves that had been previously examined in our study area, we did not confirm 13 previous species occurrence records. This result could reflect either inadequate sampling or extirpation of these populations. We also did not sample all known caves. An additional 8 caves were found too late to be included in the study, and at least some of the 75 localities in the WVSS database that were reported as having a possible entrance may actually represent cave, at least for the species involved, even if they are not enterable. With an increased sampling size, the detection of rare species increases (Huston 1994). That accumulation curves did Table 2. Estimates of total cave-limited species richness in the study area.ItemEstimate Number of Caves65 Number of Singletons21 Number of Doubletons12 Observed Number of Species18 Chao’s S2* 36.4 1.1 Burnham and Overton’s S4* 47.6 Table 3. Cumulative numbers of cave-limited species based on the “simple greedy” algorithm of Csuti et al (1997) applied to those caves that need to be sampled in orderto collect 95% of total species.CaveNew sp. Buckeye Creek Cave8 Matt’s Black Cave3 Upper Buckeye Creek Cave2 Rapps Cave2 Nellie’s Cave1 Hannah Caverns, Raceway Pit, Sunnyday Pit, Trilium Cave, Seep Cave 2, Short Stuff Cave, Tin Cave, or Wake Robin Cave1 Total Species Collected17 Table 4. Cumulative numbers of cave-limited species based on the “simple greedy” algorithm of Csuti et al (1997) applied to the largest seven caves.CaveNew sp. Buckeye Creek Cave8 Matt’s Black Cave3 Upper Buckeye Creek Cave2 Rapps Cave2 Hannah Caverns1 McFerrin Water (Spur) Cave0 Spencer Cave0 Total Species Collected16

PAGE 6

44• Journal of Cave and Karst Studies August 2004 ESTIMATINGSUBTERRANEANSPECIESRICHNESS not reach an asymptote (Fig. 2) indicates that not all species were discovered. This could indicate heterogeneity within the samples (Culver et al 2004), because caves that have a majority of the troglobites and stygobites are few, whereas caves with few or no troglobites or stygobites are numerous. This could reflect the rarity of cave-limited taxa and differences in observability among species. The need for repeated sampling is evident. We estimated that the true number of species in this area is 36, twice the number of species collected in this study. SUMMARYAlthough our data set is incomplete, it appears that (1) Previous estimates of richness for this 20 km area were quite low and increasing the number of caves sampled from nine to 65 increased the number of species from 10 to 18; (2) Only a small number of the caves need to be sampled in order to collect all of the species observed; and (3) Based on rarefaction curves and mathematical estimations, half of the species from the study area were not collected despite this effort of intensive sampling. This study advances our understanding of cave-limited species, by providing insights into the richness and distribution of stygobites and troglobites, and assessing the efficacy and accuracy of current methods of quantifying subterranean biodiversity. ACKNOWLEDGEMENTSFunding for this project was provided by grants from the Division of Natural Resources of the West Virginia Natural Heritage Program and The Cosmos Club Foundation. Collecting permits were obtained from the West Virginia Division of Natural Resources. We are indebted to H. Hobbs III, J. Hajenga, I. Šereg, R. Payn, C. Belson, M. Kuntner, P. Trontelj, S. Olcos, J. Wykle, M. Osborn, D. Butorac, G. Chepkov, S. Lee, B. Bennet, and C. Stanley for assistance in the field. We thank the late P. Allison, B. Balfour, E. Callison, R. Handley, G. Turner, and W. Jones for generously offering their knowledge of the study site and the caves therein. In addition, we thank WVSS for sharing data. The following specialists kindly provided their expertise in identifying specimens: R. Hershler, R. Muchmore, L. Ferguson, W. Shear, and K. Christiansen. Editorial assistance and insight was provided by D. Fong, K. Kim, and several reviewers, especially S. Taylor, who provided numerous comments and assistance with figures. WVACS provided access to their field station. REFERENCESBojrquez-Tapia, L.A., Balvanera, P., & Cuaron, A.D., 1994, Biological inventories and computer databases: Their role in environmental assessments: Environmental Management, v. 18, p. 775–785. Bojrquez-Tapia, L., Azuara, I., Ezcurra, E., & Flores-Villela, O., 1995, Identifying conservation priorities in Mexico through geographic systems and modeling: Ecological Applications, v. 5, p. 215–231. Borror, D.J., Triplehorn, C.A., & Johnson, N.F., 1989, An introduction to the study of insects: Saunders College Publishing, Philadelphia, 875 p. Burnham, K.P., & Overton, W.S., 1978, Estimation of the size of a closed population when capture probabilities vary among individuals: Biometrika, v. 65, p. 623–633. Chao, A., 1984, Non-parametric estimation of the number of classes in a population: Scandinavian Journal of Statistics, v. 11, p. 265–270. Christiansen, K.A., 1960, The genus Pseuedosinella (Collembola, Entomobryidae) in caves of the United States: Psyche, v. 67, p. 1–24. Christiansen, K.A., 1966, The genus Arrhopalites in the United States and Canada: International Journal of Speleology, v. 2, p. 43–73. Colwell, R.K., 1997, EstimateS: Statistical estimation of species richness and shared species from samples. Version 5. User’s Guide and application published at: http://viceroy.eeb.uconn.edu/estimates. Colwell, R.K., & Coddington, J.A., 1994, Estimating terrestrial biodiversity through extrapolation: Philosophical Transactions of the Royal Society of London B, v. 345, p. 101–118. Cond, B., 1949, Campodids cavernicoles de la rgion des Appalaches: Nots Biospologiques, v. 4, p.125–137. Conroy, M.J., & Noon, B.R., 1996, Mapping of species richness for conservation of biological diversity: Conceptual and methodological issues: Ecological Applications, v. 6, p. 763–773. Csuti, B., Polasky, S., Williams, P.H., Pressey, R.L., Camm, J.D., Kershaw, M., Keister, A.R., Downs, B., Hamilton, R., Huso, M., & Sahr, K., 1997, A comparison of reserve selection algorithms using data on terrestrial vertebrates in Oregon: Biological Conservation, v. 80, p. 83–97. Culver, D.C., & Holsinger, J. R., 1992, How many species of troglobites are there?: Bulletin of the National Speleological Society, v. 54, p. 79–80. Culver, D.C., & Sket, B., 2000, Hotspots of subterranean biodiversity in caves and wells: Journal of Cave and Karst Studies, v. 62, p. 11–17. Culver, D.C., Master, L.L, Christman, M.C., & Hobbs, H. H., 2000, Obligate cave fauna of the 48 contiguous United States: Conservation Biology, v. 14, p. 386–401. Culver, D.C., Deharveng, L., & Gibert, J., 2001, Introduction, in Culver, D.C., Deharveng, L., Gibert, J., & Sasowsky, I.D., (eds), Mapping Subterranean Biodiversity: Karst Waters Institute Special Publication 6, Charles Town, WV, p. 1–2. Culver, D.C., Christman, M.C., Sket, B., & Trontelj, P., 2004, Sampling adequacy in an extreme environment: Species richness patterns in Slovenian caves: Biodiversity and Conservation, v. 13, p. 1209–1229. Dasher, G.R., & Balfour, W.M., (eds), 1994, The Caves and Karst of the Buckeye Creek Basin, Greenbrier County, West Virginia: West Virginia Speleological Society, Maxwelton, WV, 238 p. Davies, W.E., 1958, Caverns of West Virginia with supplement: BiggJohnston-Withrow, Beckley, WV, 402 p. Deharveng, L., 2001, Mapping European endemism: The project “Endemism”, in Culver, D.C., Deharveng, L., Gibert, J., & Sasowsky, I.D., (eds), Mapping Subterranean Biodiversity: Karst Waters Institute Special Publication 6, Charles Town, WV, p. 12–14. Fong, D.W., & Culver, D.C., 1994, Invertebrate fauna, in Dasher, G.R., & Balfour, W.M., (eds), The Caves and Karst of the Buckeye Creek Basin Greenbrier County West Virginia: West Virginia Speleological Survey, Maxwelton, WV, p. 41–42. Gotelli, N.J., & Colwell R.K., 2001, Quantifying biodiversity: Procedures and pitfalls in the measurement and comparison of species richness: Ecology Letters v. 4, p. 379–391. Hammond, P.M., 1994, Practical approaches to the estimation of the extent of biodiversity in speciose groups: Philosophical Transactions of the Royal Society of London B, v. 345, p. 119–136.

PAGE 7

Journal of Cave and Karst Studies August 2004 • 45 SCHNEIDERANDCULVER Hobbs, H.H., Jr., 1964, Anew cave-dwelling crayfish from the Greenbrier drainage system, West Virginia (Decapoda, Astacidae): Proceedings of the Biological Society of Washington, v. 78, p. 189–194. Holsinger, J.R., 1967, Systematics, speciation, and distribution of the subterranean amphipod genus Stygonectes (Gammaridae): Bulletin of the U.S. National Museum, v. 259, p. 1–176. Holsinger, J.R., Baroody, R.A., & Culver, D.C., 1976, The invertebrate cave fauna of West Virginia: West Virginia Speleological Survey, v. 7, p. 82. Hubricht, L., 1943, Studies on the Nearctic fresh-water Amphipoda III. Notes on the freshwater amphipods of the eastern United States, with descriptions of ten new species: American Midland Naturalist, v. 29, p. 683–712. Hubricht, L., 1963, New species of Hydrobiidae: Nautilus, v. 76, p. 138–140. Huston, M.A., 1994, Biological Diversity: The coexistence of species on changing landscapes: Cambridge University Press, Cambridge, 701 p. Hyman, L.H., 1954, North American triclad Turbellaria. XIII. Three new cave planarians: Proceedings U.S. National Museum, v. 103, p. 563–573. Izquierdo, I., Martin, J.L., Zurita, N., & Medina, A.L., 2001, Geo-referenced computer recordings as an instrument for protecting cave-dwelling species of Tenerife (Canary Islands), in Culver, D.C., Deharveng, L., Gibert, J., & Sasowsky, I.D., (eds), Mapping Subterranean Biodiversity: Karst Waters Institute Special Publication 6, Charles Town, WV, p. 45–48. Jones, W.K., 1997, Karst Hydrology Atlas of West Virginia: Karst Waters Institute Special Publication 4, Charles Town, WV, 111 p. Keating, K.A., Quinn, J.F., Ivie, M.A., & Ivie, L.L., 1998, Estimating the effectiveness of further sampling in species inventories: Ecological Applications, v. 8, p. 1239–1249. Kodric-Brown, A., & Brown, J.H., 1993, Incomplete data sets in community ecology and biogeography: Acautionary tale: Ecological Applications, v. 3, p. 736–742. Krow, S., & Culver, D.C., 2001, Gaps in sampling cave fauna: Memoires de Biospeologie (International Journal of Subterranean Biology), v. 28, p. 129–136. Loomis, H.F., 1943, New cave and epigean millipeds of the United States, with notes on established species: Bulletin of the Museum of Comparative Zoology, v. 92, p.373–410. Nichols, J.D., Boulinier, T., Hines, J.E., Pollock, K.H., & Sauer, J.R., 1998, Inference methods for spatial variation in species richness and community composition when not all species are detected: Conservation Biology, v. 12, p. 1390–1398. Pearson, D.L., & Cassola, F., 1992, World-wide species richness patterns of tiger beetles (Coleoptera: Cicindelidae): Indicator taxon for biodiversity and conservation studies: Conservation Biology, v. 6, p. 376–391. Proudlove, G.S., 2001, Subterranean biodiversity in the British Isles, in Culver, D.C., Deharveng, L., Gibert, J., & Sasowsky, I.D., (eds), Mapping Subterranean Biodiversity: Karst Waters Institute Special Publication 6, Charles Town, WV, p. 56–58. Steeves, H.R., III., 1963, Two new troglobitic asellids from West Virginia: American Midland Naturalist, v. 70, p. 462–465. USAID (United States Agency for International Development), 2000, Biodiversity assessment for Croatia: Washington, DC., 89 p. Valentine, J.M., 1931, New cavernicole Carabidae of the subfamily Trechinae Jeannel: Journal of the Elisha Mitchell Scientific Society, v. 46, p. 247–258. Valentine, J.M., 1932, Aclassification of the genus Pseudanophthalmus Jeannel (fam. Carabidae), with descriptions of new species and notes on distribution: Journal of the Elisha Mitchell Scientific Society, v. 48, p. 261–280. Vachon, M., 1952, Apropos d’un pseudoscorpion cavenicole: Notes Biospeologiques, v. 7, p. 105–12. Wray, D.L., 1952, Some new North American Collembola: Bulletin of the Brooklyn Entomological Society, v. 47, p. 95–106. Zacharda, M., 1985, New Rhagidiidae (Acarina: Prostigmata) from caves of the U.S.A.: Vstnik Ceskoslovenske. Spolenosti Zoologicke v. 49, p. 67–80.