2009: Proceedings of the 15th International Congress of Speleology, Kerrville, Texas, July 19-26, 2009, Volume 3, Contributed Papers

Citation
2009: Proceedings of the 15th International Congress of Speleology, Kerrville, Texas, July 19-26, 2009, Volume 3, Contributed Papers

Material Information

Title:
2009: Proceedings of the 15th International Congress of Speleology, Kerrville, Texas, July 19-26, 2009, Volume 3, Contributed Papers
Alternate Title:
International Congress of Speleology
Alternate Title:
ICS Proceedings
Creator:
White, William Blaine, 1934---
Publisher:
Greyhound Press
Publication Date:
Language:
English

Subjects

Genre:
Conference Proceeding
serial ( sobekcm )

Notes

General Note:
Note: an uncompressed version is available by selecting the "Additional Digital Versions" URL below. From the Preface: "Five Hundred papers were presented at the Fifteenth International Speleological Congress, Kerrville, Texas, USA on July, 19-26, 2009 by speleologists from all over the world. These volumes contain the written record for those papers. Authors who chose to do so were invited to prepare full papers of up to six pages. Authors who preferred a more limited text contributed abstracts of their papers for the Proceedings. The papers fall into two categories: those that were incorporated into the 13 symposia; 300 papers and those that were contributed to topical sessions; 200 papers. Written accounts appear for both oral presentations and papers that were presented as posters. The papers are arranged alphabetically by first author in the sequence Plenary lectures, Symposia papers, and Contributed papers. Both abstracts and papers received comprehensive technical reviews by the Science Committee. The authors had the opportunity to revise their papers in response to reviewer's comments. It is hoped that the review process has improved the clarity of the papers so that information transfer is enhanced."
Restriction:
Open Access - Permission by Publisher
General Note:
See Extended description for more information.

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:
K26-00112 ( USFLDC DOI )
k26.112 ( USFLDC Handle )
8372 ( karstportal - original NodeID )
9781879961357 ( ISBN )

USFLDC Membership

Aggregations:
Added automatically
Karst Information Portal

Postcard Information

Format:
serial

Downloads

This item has the following downloads:


Full Text

PAGE 3

Cover design: Beth Fratesi Layout and design: Greyhound Press Editor: William B. White15th International Congress of SpeleologyKerrville, Texas United States of America July 19, 2009 Volume 3, Contributed PapersProceedings

PAGE 4

PROcee CEE Di I NGs S 15t T H INte TE RNati ATI ONal AL CONGRess ESS OF Spele PELE Ol L OGYKerrville, Texas United States of America July 19, 2009Produced by the organizing Committee of the 15th Internatioal Congress of Speleology Published by the International Union of Speleology 2009 National Speleological Society, Inc. Individual authors retain their copyrights as indicated in the text. All rights reserved. No part of this work may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any data storage or retrieval system without the express written permission of the copyright owner. All drawings and maps are used with permission of the artists. Unauthorized use is strictly prohibited. Printed in the United States of America. Library of Congress Control Number 2009930608 ISBN 978-1-879961-35-7VOlu LU Me E 3 CONt T Ri I Bute UTE D Pape APE Rs S

PAGE 5

Proceedings of the 15th International Congress of Speleology Errata and Omissions The Proceedings of the 15th International Congress of Speleology contain either abstracts or full papers of the 500 contributions presented at the Congress. The three volumes of the Proceedings total 2130 pages. The pathway to this mass of material was as follows: Prospective authors submitted an initial abstract to the ICS Science Committee. These abstracts were reviewed by the Committee to ascertain that the subject matter was appropriate for the Congress. The abstracts were then returned to the authors with suggestions and an invitation to prepare a full paper limited to six printed pages. Few papers were rejected, but some were withdrawn so that of 540 initial submissions, 500 were presented at the Congress. The draft papers were sent to the Science Committee who distributed them for review after which they were returned to the authors for such adjustments as the reviewers deemed necessary The final papers were received by the Science Committee for formal acceptance and were forwarded to the editor. The edited papers were then transmitted to Production Manager for page layout and preparation for the printer. All of this movement of abstracts and manuscripts was done electronically. In the process of transmittals, various reviews, and editorial handling, a few errors and omissions were created. The lists that follow contain the additions and corrections that have been brought to our attention. We have limited the corrections to matters of fact; small errors in spelling, punctuation, and formatting are not addressed. We apologize to the authors whose papers were mishandled in some manner. The Editorial Team Errata Volume 1, Page 541 Cave Sediments Related to Cretaceous Tertiary Paleokarst Developed in Eogenetic Carbonate Rocks: Examples from SW Slovenia and NW Croatia by Bojan Otoni The abstract was truncated in printing with only the first few lines appearing in the Proceedings. The full abstract follows. In the SW Slovenia and NW Croatia a regional paleokarstic surface separates the passive margin shallow marine carbonate successions of different Cretaceous formations from the Upper Cretaceous to Eocene palustrine and shallow marine limestones of the synorogenic carbonate platform. Thus, the paleokarst corresponds to an uplifted peripheral foreland bulge, when diagenetically immature eugenetic carbonates were subaerially exposed and karstified

PAGE 6

Among the subsurface paleok a rstic features vadose and phreatic forms are recognized. For the epikarst, pedogenic features and enlarged root related channels are characteristic. Vado se channels, shafts and pits penetrate up to a few tens of meters bellow the paleokarstic surface, where they may merge with originally horizontally oriented phreatic cavities. The latter comprise characteristics of caves forming in fresh/brackish water le nses. T he phreatic cavities were found in different positions regarding to the paleokarstic surface, the lowest one being some 75 meters below it. Usually only one distinct paleocave level occurs per location, although indistinct levels of spongy porosity and/or irregularly dispersed cavities of different sizes have been noticed locally. The cavities had been subsequently partly reshaped and entirely filled with detrital sediments and flowstones in the upper part of the phreatic, epiphreatic and vadose zones. The internal cave sediments and flowstones may also occur as clasts in deposits (mostly breccias) that fill subsurface paleokarstic cavities and cover the paleokarstic surface. In general, the variety of cave infilling deposits and the amount of surface derived material decrease with the distance from the paleokarstic surface. Below 1318O values of cavity deposits usually exhibit good correlation with trend sign ificant for meteoric diagenesis. Relatively small p hreatic cavities of the lowermost part of the paleokarstic profile s are commonly geopet ally infilled with laminated mudstone derived from incomplete dissolution of the hostrock overlain by coarse grained blocky calcite of meteoric or mixing meteoric/ marine origin. S omewhat larger phreatic caves located shallower below the paleokarstic surface usually exhibit more complicated stratigraphy. Although the lower parts of the caves are still mainly infilled with reddish stained micritic carbonate sediment different types of flowstone, especially calcite rafts become more prominent higher in the cave profiles. Gradually in the upper parts of the cav es, sediments derived from the paleokarstic surface prevail over autochthonous deposits. Especially channels of the epikarst zone are almost entirely infilled with pedogenically modified material derived directly from the paleokarstic surface. Regardless of their origin, cave deposits had been often intensively modified by pedogenic processes while they were exposed to the paleokarstic surface by denudation. Just prior to marine transgression over the paleokarstic surface some cavities or their parts had been infilled by marine derived microturbidites. It will be shown that especially deposits related to denuded phreatic caves may be of great importance for the study of speleogenetic, geomorphologic and hydrogeologic evolution of a specific karst region. Volume 2, page 650 Medical and Governmental Considerations of CO2 and O2 in Volcanic Caves by William R. Halliday The final sentence of the first paragraph on page 652 contains incorrect wording. The sentence should read: The issue resurfaced when U.S. Geological Survey and National Park Service personnel applied OSHA standards to volunteers in volcanic caves with nontoxic levels of O2 and CO2. Volume 2, page 662 Unusual Rheogenic Caves of the 1919 Postal Rift Lava Flow, Kilauea Caldera, Hawaii by William R. Halliday The first paragraph on page 664 contains several errors and misstatements. The corrected paragraph should read:

PAGE 7

Noxious gas (probably HCl) was encountered only in one tiny cave on the edge of Halemaumau Crater. Presumed sulfate fumes were encountered in numerous caves but were found to be essentially non toxic. Eye irritation rarely was encountered ( Halliday, 2000b). Two types of CO2 monitors previously untested in volcanic caves were required for the last five field trips. They were found to be useless in hyperthermal caves and no sig nificant elevation of CO2 was identi fied in normothermic examples (Halliday, 2007). In no cave was significantly elevated CO2 identified by changes in normal bre athing (Halliday, this volume ). Volume 2, Page 785 Symposium #11, Speleogenesis in Regional Geological Evolution and Its Role in Karst Hydrogeology and Geomorphology was arranged by Alexander Klimchouk and Arthur N. Palmer (no t by John Mylroie and Angel Gins as listed on the title page of the symposium in the Proceedings). Volume 2, Page 1033 Uranium Mapping in Speleothems: Occurrence of Diagenesis, Det rital Contamination and Geochemical Consequences The correct authors for this paper are: Richard Maire, Guillaume Deves, Ann -Sophie Perroux, Bassam Ghaleb, Benjamin Lans, Thomas Bacquart, Cyril Plaisir, Yves Quinif and Richard Ortega. The names of Bassam Ghaleb and Yves Quinif were omitted in the Proceedings Volume. Volume 3, Page 1307 Species Limits, Phylogenetics, and Conservation of Neoleptoneta Spiders in Texas Caves by Joel Ledford, Pierre Paquin, and Charles Griswold James Cokendolpher, Museum of Texas, Texas Tech University, Lubbock, Texas was also a co author for this paper. Omissions The Fossil Bears of Southeast Alaska by Timothy H. Heaton and Frederick Grady was inadvertently omitted in the final stages of page layout. The reviewed and edited paper follows:

PAGE 8

THE FOSSIL BEARS OF SOUTHEAST ALASKA TIMOTHY H. HEATON1, FREDERICK GRADY2 1Department of Earth Sciences, University of South Dakota, Vermillion, SD, 57069, USA 2Department of Paleobiology, Smithsonian Institution, Washington, DC, 20560, USA Southeast Alaska is home to brown bears ( Ursus arctos) and black bears ( U. americanus ) with an unusual distribution. Both species inhabit the mainland while on ly black bears inhabit the i slands south of Frederick Sound and o nly brown bears inhabit the islands north of Frederick Sound. Brown bears of the northern islands belong to a distinct lineage and are genetically more similar to polar bears than their mainland counterparts. Bears are among the most common fossils found in caves in the regi on, and they indicate that both species made greater use of caves as dens when the climate was colder. B ut no bear fossils are known from the Last Glacial Maximum (LGM), even at On Your Knees Cave where foxes and marine mammals have been recovered across m ost of this interval. This begs the question of whether bears survived the LGM on coastal refugia or recolonized the islands after the ice retreated. No evidence has been found to settle the question for black bears. Black bears are far more common than b r own bears in On Your Knees Cave for the period before the LGM but they were slower than brown bears in expanding their range across the islands after the ice melted The evidence for survival in a local refugium is much stronger for brown bears. While the y are less common before the LGM, they had a greater distribution than black bears immediately following the LGM, including some of the outermost islands of the archipelago. The lack of brown bear fossils from mainland sites during early postglacial times may indicate that the mainland was not the source of this population The distinct genetic character of modern island brown bears also suggests that they did not derive from the mainland Two fossil brown bears from caves of Prince of Wales Island have had successful DNA extractions and match the distinct lineage that now lives only on the northern islands of Southeast Alaska. A refugium for brown bears may have been offshore on the continental shelf which was exposed during the LGM but was flooded by rising sea level in the early postglacial period. 1. Introduction Our research in southeast Alaska began in 1991 after several bear skeletons were found in El Capitan Cave on Prince of Wales Island by a caving expedition (HEATON and G RADY, 1992, 1993). El Capi tan Cave is Alaskas largest known cave and has passages that flood during storms, but the fossils were found in a quiet upper passage near the surface. One skeleton was complete and undisturbed, suggesting that the bears were denning in the cave, so caver s called this passage the Hibernaculum. It was apparent that the bears accessed the cave by an entrance that had become sealed with soil and logs, and we were able to reopen this entrance to conduct an excavation of the site. Soon cavers discovered skeleto ns in other caves of the region with similar dimensions, namely horizontal passages 1.5 2.5 meters in diameter. Several natural trap caves with bear fossils were also discovered. Although our research has expanded to include a variety of mammals, birds, an d fishes (H EATON and G RADY, 2003), bears have remained a major focus, and our fossil discoveries have contributed to solving the question of whether animals survived the Ice Age in Southeast Alaska. Most islands of Southeast Alaska are home to bears, but currently there is no more than one species per island. Black bears ( U rsus americanus ) inhabit Prince of Wales Island and most other islands south of Frederick Sound, while brown bears ( Ursus arctos) inhabit the islands north of Frederick Sound, namel y Adm iralty, Baranof, and Chicha gof (ABC) islands. Both species inhabit the nearby mainland (MACDONALD and COOK, 2007). Prior to the discovery of a fossil record KLEIN (1965) proposed that this island distribution resulted from a postglacial colonization histo ry: brown bears arriving from the north and black bears from the south. This hypothesis was based on the prevailing assumption that no land animals survived the Last Glacial Maximum (LGM, 24,00013,000 radiocarbon years B.P.) in Southeast Alaska because of complete ice cover. Although the islands of Southeast Alaska exhibit a nested mammalian fauna suggestive of recent colonization (CONROY et al., 2000), fossil and genetic studies of bears have revealed a much more complex history in the region. The complete skeleton from El Capitan Cave, as well as portions of several others, were of black bears, distinguished from the living bears on the island only by their large size. Their size seemed especially significant since they appeared to be females based on the lack of bacula and the gracile structure of their skulls. Even more significant was the discovery of even larger bear remains that we identified as brown bear. Finding that Prince of

PAGE 9

Wales Island had been home to additional species in early postglacial t ime conflicted with the simple postglacial colonization model held by K LEIN (1965) and other biologists. In addition to brown bears we also discovered fossil remains of Arctic fox ( Alopex lagopus ), red fox (Vulpes vulpes ), and caribou ( Ranifer tarandus ) t hat no longer inhabit the island. Rather than lacking a fauna at the end of the Ice Age, Prince of Wales Island simply had a different fauna that was adapted to the colder and less forested habitat Following this initial discovery we set out to expand ou r dataset both geographically and chronologically by searching for caves with fossil deposits on different islands and the mainland, in diverse habitats, and of greater antiquity. During the 1990s fossil sites were brought to our attention by cavers explor ing the region, often working with the support of Tongass National Forest and guided by forest agendas. After 2000 we began coordinating searches for caves specifically to fill in gaps in our dataset. In spite of limits imposed by limestone distribution an d the difficulty of finding sites over 12,000 years old, a long history for both brown and black bears has emerged. During this same period geneticists began DNA studies on living bear populations in Southeast Alaska that complemented our work (H EATON et a l. 1996), and we have worked in conjunction with ancient DNA researchers to trace bear lineages back in time. What has emerged is a greatly expanded, but not entirely complete, picture of bear history in Southeast Alaska. 2. Postglacial History The postg lacial record of bears in Southeast Alaska is spectacular. Following the discovery of black and brown bears in El Capitan Cave (130 m elevation), additional brown bear skeletons were found in two high elevation caves (over 500 m) on northern Prince of Wale s Island: two juveniles in a natural trap called Blowing in the Wind Cave, and parts of 12 individuals in a horizontal tube called Bumper Cave, including skeletons of what appeared to be a mother and her two cubs (Table 1). By contrast, lower elevation cav es (below 200 m) on the island, such as Kushtaka and On Your Knees caves (den sites) and Tlacatzinacantli Cave (a natural trap) contained only black bears from the postglacial interval (Table 2). This apparent partitioning of den sites by the two species m ust be kept in mind when considering other parts of Southeast Alaska where samples from diverse elevations are not available. This does not mean that brown bears were restricted to high elevations because their isotopic signature indicates a stronger marin e diet than black bears (H EATON 1995; H EATON and G RADY, 2003). Table 1. List of radiocarbon dated brown bear ( Ursus arctos ) fossils from caves of Southeast Alaska in order of age. Laboratory # Age (years B.P. ) 13 C Site Island Sample AA15224 7,205 65 17.9 Bumper Cave POW Dentary AA56996 9,590 95 20.5 Deer Bone Cave Coronation Radius AA 07794 9,760 75 18.0 El Capitan Cave POW Humerus AA 10451 9,995 95 18.5 Blowing in the Wind Cave POW Ribs AA-5 2223 10,700 100 17.1 Enigma Cave Dall Humerus AA15225 10,970 85 19.5 Bumper Cave POW Molar AA15223 11,225 110 16.8 Bumper Cave POW Humerus AA 52221 11,600 100 14.6 Enigma Cave Dall Dentary AA44450 11,630 120 18.2 Colander Cave Coronat ion Humerus AA15222 11,640 80 17.8 Bumper Cave POW Rib AA 15226 11,715 120 16.0 Enigma Cave Dall Humerus AA 32122 11,910 140 18.1 El Capitan Cave POW Rib2 AA52222 11,930 120 14.6 Enigma Cave Dall Skull AA10445 12,295 120 18.3 El Capi tan Cave POW Pelvis AA 33783 26,820 700 16.3 On Your Knees Cave POW Astragalus AA 52219 29,040 600 16.3 On Your Knees Cave POW Rib AA52220 29,590 980 17.7 On Your Knees Cave POW M2/ AA33792 31,700 1900 16.2 On Your Knees Cave POW Molar AA52218 31,900 1,300 19.6 On Your Knees Cave POW Claw AA 52207 33,300 1,500 17.0 On Your Knees Cave POW Phalanx 1 AA15227 35,365 800 15.9 On Your Knees Cave POW Femur AA52215 38,800 3,000 10.0 On Your Knees Cave POW Phalanx 2

PAGE 10

AA 33791 39, 400 3100 17.1 On Your Knees Cave POW Tooth AA 52216 34,000 + 17.4 On Your Knees Cave POW M/1 AA52201 40,900 + 16.8 On Your Knees Cave POW P4/ AA52217 41,100 + 15.4 On Your Knees Cave POW Vertebra Table 2. List of radiocarbon dated black bear ( Ursus americanus) fossils from caves of Southeast Alaska in order of age. Laboratory # Age (years B.P. ) 13 C Site Island Sample CAMS 27263 2,790 60 23.2 Kushtaka Cave POW Artifact AA57000 3,425 50 12.5 Lawyers Cave Mainland Humerus CAMS 31068 3,960 50 20.7 On Your Knees Cave POW Dentary AA 36637 4,847 58 21.2 Hole 52 Cave Mainland Skull SR 5265 6,290 50 Lawyers Cave Mainland Phalanx AA10447 6,415 130 22.1 El Capitan Cave POW Skull CAMS 24967 8,630 60 21.4 Kushtaka Cave POW Rib AA 18451R 9,330 155 23.9 Kushtaka Cave POW Femur AA 32118 10,020 110 22.1 Tlacatzinacantli Cave POW Femur AA36641 10,080 120 21.6 Hole 52 Cave Mainland Phalanx AA 33780 10,090 160 21.2 On Your Knees Cave POW Phalanx CAMS 42381 10,300 50 20.7 On Your Knees Cave POW Artifact AA 36636 10,350 100 18.9 Hole 52 Cave Mainland Skull AA36640 10,420 110 21.6 Hole 52 Cave Mainland Skull AA 07793 10,745 75 21.1 El Capitan Cave POW Humerus AA 32120 10,860 120 21.8 Tlacatzinacantli Cave POW Skull AA 32117 10,870 120 21.8 Tlacatzinacantli Cave POW Ulna AA36638 10,930 140 19.8 H ole 52 Cave Mainland Skull AA 32119 10,970 120 22.4 Tlacatzinacantli Cave POW Fragment AA 33202 11,460 130 19.9 Hole 52 Cave Mainland Canine AA 10446 11,540 110 20.0 El Capitan Cave POW Skull AA10448 11,565 115 18.7 El Capitan Cave POW Sku ll AA 21569 28,695 360 20.7 On Your Knees Cave POW Calcaneum AA 21570 29,820 400 20.8 On Your Knees Cave POW Vertebra AA 33781 36,770 2300 18.6 On Your Knees Cave POW Femur AA33194 38,400 3000 18.4 On Your Knees Cave POW Humerus AA 33198 39,000 3100 19.5 On Your Knees Cave POW Rib AA 16831 41,600 1500 20.7 On Your Knees Cave POW Tibia AA 36653 25,000 + 22.0 On Your Knees Cave POW Premolar AA36655 27,000 + 18.2 On Your Knees Cave POW Baculum AA 33196 38,500 + 19.4 On Your Knee s Cave POW Scapula AA 52206 38,500 + 20.8 On Your Knees Cave POW Metapodial AA 52204 39,100 + 20.2 On Your Knees Cave POW Canine AA33200 39,400 + 19.3 On Your Knees Cave POW Canine AA 33195 40,100 + 18.4 On Your Knees Cave POW Humerus AA 33199 40 ,200 + 19.9 On Your Knees Cave POW Canine AA 44448 41,000 + 21.7 On Your Knees Cave POW Molar SR 5110 43,050 + On Your Knees Cave POW Vertebra SR 5111 44,940 + On Your Knees Cave POW Skull

PAGE 11

Several postglacial deposits have also been found on the mainland near the town of Wrangell and on two of the outermost islands of the Archipelago: Coronation and Dall Islands (HEATON and GRADY, 2003) Today only black bears inhabit Dall Island while no bears inhabit Coronation Island (MACDONALD and COOK, 2007) Three early postglacial cave deposits have turned up six individuals, all of which match brown bear (Table 1). Deer Bone C ave is a den cave while Colander Cave is a natural trap, and Enigma Cave is larger and more complex with bear skeletons both in horiz ontal den passages and at the bottom of pits. All these caves are at 200 m elevation or lower. By contrast, two postglacial cave deposit s on the mainland, a den site called Lawyers Cave and a complex cave with horizontal passages and pits called Hole 52, c ontain only black bear remains (Table 2). Brown bears may have denned at higher elevation, but no such sites are known. The remarkable conclusion from these sites is that the two bear species had nearly the opposite distribution in the early postglacial period than they do today. Currently both species inhabit the mainland while only black bears inhabit the southern islands of Southeast Alaska. Shortly a fter the Ice Age only brown bears inhabited the outer islands, both species occupied the large Prince of Wales Island, and only black bears are documented from the mainland. Discovering the postglacial history of bears in the northern islands of Southeast Alaska, where only brown bears live today, has been hampered by a paucity of limestone and a lack of any fossil discovery. Since brown bears thrived in the southern islands in early postglacial times, there is no reason to doubt their presence farther north. Whether black bears ever colonized the northern islands remains a mystery. To the south of Alaska a p attern similar to Prince of Wales Island has been documented by Canadian investigators Haida Gwaii (Queen Charlotte Islands) and Vancouver Island are currently home only to black bears. Fossil black bears have been found dating back to 10,000 years B.P. o n Haida Gwaii (R AMSEY et al. 2004; F EDJE et al. 2004) and from about 9,800 to 12,000 years B.P. on Vancouver Island (N AGORSEN et al. 1995; N AGORSEN and K EDDIE, 2000). Brown bears from Haida Gwaii have been found dating from 10,000 to 14,500 years B.P., showing that they once were widespread on coastal islands. Another remarkable pattern visible in Tables 1 and 2 is the sheer number of early postglacial bears. With the exception of the sealed hibernaculum of El Capitan Cave, all of these sites remain open for potential denning today. Yet far more specimens of both black and brown bears date between 9,000 and 12,000 years B.P. than date to the 9,000 years since then. Most of these remains were exposed on the cave floors (not fully buried) so were not selected for dating based on their potential antiquity. Either bears were more numerous in early postglacial times or they were denning in caves much more regularly. The fact that natural trap caves (at least a third of the sites) show this same pattern suggests a high bear population. None of the other species we have studied show this distinct chronological pattern. Perhaps the early successional stages of forest development following the melting of the glaciers provided a high density of berries and other edible foods preferred by bears for the herbivorous part of their diet. Since climax forest s are lacking in such foods, modern bears are attracted to forest clear cuts, shorelines, and other disturbed areas where such plants grow. 3. Ice Age History The single site in Southeast Alaska that has produced an extensive Ice Age record (prior to 13,000 radiocarbon years B.P.) is On Your Knees Cave. It is a small cave on the northern tip of Prince of Wales Island discovered during a logging survey and had only a few bones initially exposed. The significance of the site was only recognized when a partial brown bear femur was radiocarbon dated to 35,365 years B.P. (Table 1). Excavation began in 1996 and continued until 2004. An extensive record of mammals, birds, and f ish was discovered covering at least the last 45,000 years (H EATON and G RADY, 2003) plus an extensive archaeological record including the oldest human remains from Alaska or Canada (D IXON et al. 1997). Devils Canopy Cave on Prince of Wales Island is the only other site where we obtained an Ice Age radiocarbon date (on marmot), but extensive excavation produced only a few rodent and insectivore remain s. Our extensive efforts to find an Ice Age site on the outer islands of Southeast Alaska have so far been unsuccessful. For a single site, On Your Knees Cave provides a superb record of animals during the LGM and the preceding interstadial. As can be seen in Tables 1 and 2 many bone dates are beyond the radiocarbon limit, but uranium dates on speleothem fragments date back to 185,800 2,800 years B.P. (D ORALE et al. 2003). Both black and brown bears were present and probably used the cave as a den from at least 41,000 years B.P. until the approach of the LGM (Tables 1 and 2). We have not dated enough samples to be certain exactly when their use of the cave ceased, but no bear remains have been dated to the glacial maximum itself. A sample of 25 ringed seal ( Phoca hispida) specimens were radiocarbon dated from 24,150 490 to 13,690 240 years B.P., which is the very interval that the

PAGE 12

bears (and caribou) are missing. Arctic and red foxes, other marine mammals, and sea birds also date to the LGM so the cave was available and used as a den (by foxes) during that interval. One ringed seal humerus has bite marks that match bear canines, but it could be a polar bear ( Ursus maritimus) kill that was scavenged by foxes. Black bear fossil s outnumber brown bear fossil s in On Your Knees Cave by a ratio of about 10:1. This is not evident in Tables 1 and 2 because we selected specimens of both species for dating. This difference could represent a greater abundance of black bears or a partitioning of den sites by elevation like we see during the postglacial period. Other elements of the fauna suggest that conditions during the interstadial were similar to the early postglacial interval before a climax forest was established. 4. Genetics T ALBOT and S HIELDS (1996) found that brown bears of Admiralty, Baranof, and Chicha gof (ABC) islands (Southeast Alaskan islands north of Frederick Sound) a re distinct from all other populations based on mitochondrial DNA and are more closely related to polar bears than to their mainland counterparts Using nuclear microsatellite variations P AETKAU et al. (1998) confirmed this result for female s but detected some exchange of males with the local mainland population. L EONARD et al. (2000) discovered a fossil from Yukon Territory match ing the ABC bears and dating to 36,500 1,150 years B.P. so this clade had a wider distribution before the LGM Nevertheless, the current restricted range of this clade suggests that the islands of Southeast Alaska acted as a refugium for this population during the glacial maximum (HEATON et al., 1996) Further support for this hypothesis comes from early postglacial fossils of Prince of Wales Island and Haida Gwaii. After several failed attempts at extracting ancient DNA, B ARNES et al. (2002) reported that a brown bear fossil from Blowing in the Wind Cave ( AA10451 on Table 1) belong s to the ABC clade. Further work by S arah B ray (personal communication) also linked a bear from Bumper Cave ( AA16553 on Table 1) and ones from Haida Gwaii to the ABC clade. STONE and COOK (2000) found that black bears from the southern islands of Southeast Alaska belong to a mitochondri al lineage that is also found on the islands and coastal mainland of British Columbia and down the coast to northern California. Several other mammal species have distinct coastal lineages with a similar range, but it remains unclear whether the source of these lineages was south of Cordilleran glaciers or on coastal refugia, possibly in Southeast Alaska (COOK et al., 2001, 2006). 5. Conclusions The absence of a fossil record of bears from the LGM leaves open the question of whether they survived the glaci al expansion in Southeast Alaska on coastal refugia or recolonized afterward. Cave faunas document that both brown and black bears were present during the preceding interstadial and reappeared in great numbers soon after the ice melted. Genetic evidence fo r a distinct coastal lineage, where refugial isolation is the simplest explanation, is strong for brown bears but more equivocal for black bears. Both bears are refugial species in the sense that they were adversely affected by glaciation and struggled to survive under unfavorable climatic conditions. By contrast, other carnivores such as ringed seals, Arctic foxes, and likely polar bears flourished and expanded their ranges during the LGM. The extent to which the Arctic and refugium faunas competed with one another is unknown, but their interactions could have been a factor in the temporary loss of black and brown bears from On Your Knees Cave. What we learn from postglacial bears is that the species were able to move about freely and colonize territory th at was favorable for them, rather than being restricted by barriers and competition. Solving the full puzzle of bear history in Southeast Alaska will require finding additional faunas of similar antiquity to On Your Knees Cave, as a single site cannot document the movements of species. During the LGM the expanding glaciers pushed mammal populations westward, while falling sea level opened up new habitat to the west and changed the configuration of the coastal corridor. The possibility that populations of be ars and other mammals found suitable refugia to survive the LGM in Southeast Alaska is very possible. References BARNES, I., P. MATHEUS, B. SHAPIRO, D. JENSEN, and A. COOPER (2002) Dynamics of Pleistocene population extinctions in Beringian brown bears. S cience 295:2267 2270. CONROY, C. J., J. R. DEMBOSKI, and J. A. COOK (1999) Mammalian biogeography of the Al exander Archipelago of Alaska: a north temperate nested fauna. Journal of Biogeography 26:343 352.

PAGE 13

COOK, J. A., A. L. BIDLACK, C. J. CONROY, J. R. DEMBOSKI, M. A. FLEMING, A. M. RUNCK, K. D. STONE, and S. O. MACDONALD (2001) A phylogeographic perspective on endemism in the Alexander Archipelago of the North Pacific. Biological Conservation 97:215 227. COOK, J. A., N. G. DAWSON, and S. O. MACDONALD (200 6) Conservation of highly fragmented systems: The north temperate Alexander Archipelago. Biological Conservation 133:1 15. DIXON, E. J., T. H. HEATON, T. E. FIFIELD, T. D. HAMILTON, D. E. PUTNAM, and F. GRADY (1997) Late Quaternary regional geoarchaeology of Southeast Alaska karst: A progress report. Geoarchaeology 12(6):689 712. DORALE, J. A., T. H. HEATON, and R. L. EDWARDS (2003) U Th dating of fossil associated cave calcites from southeastern Alaska. Geological Society of America Abstracts with Programs vol. 35, no. 6, p. 334. FEDJE, D. W., Q. MACKIE, E. J. DIXON, and T. H. HEATON (2004) Late Wisconsin environments and archaeological visibility on the northern Northwest Coast. In Entering America: Northeast Asia and Beringia before the Last Glacial Maxi mum D. B. Madsen (ed.), University of Utah Press, p. 97 138 HEATON, T. H. 13C values from vertebrate remains of the Alexander Archipelago, southeast Alaska. Current Research in the Pleistocene, vol. 12, pp. 95 97. HEATON, T. H., and F. GRADY (1992) Preliminary report on the fossil bears of El Capitan Cave, Prince of Wales Island, Alaska. Current Research in the Pleistocene 9:97 99. HEATON, T. H., and F. GRADY (1993) Fossil grizzly bears ( Ursus arctos ) from Prince of Wales Island, Alaska, offer new insights into animal dispersal, interspecific competition, and age of deglaciation. Current Research in the Pleistocene 10:98 100. HEATON, T. H., and F. GRADY (2003) The Late Wisconsin vertebrate history of Prince of Wales Island, Southe ast Alaska. I n Ice Age Cave Faunas of North America, B. W. Schubert, J. I. Mead, and R. W. Graham (eds.), Indiana University Press, p. 17 53. HEATON, T. H., S. L. TALBOT, and G. F. SHIELDS (1996) An Ice Age Refugium for Large Mammals in the Alexander Archipelago, Southeastern Alaska. Quaternary Research, 46(2):186 192. KLEIN, D. R (1965) Postglacial distribution patterns of mammals in the southern coastal regions of Alaska. Arctic 18:7 20. LEONARD, J. A., R. K. WAYNE, and A. COOPER (2000) Population genetics of Ice Age brown bears. Proceedings of the National Academy of Sciences 97(4):1651 1654. MACDONALD, S. O., and J. A. COOK (2007) Mammals and amphibians of Southeast Alaska. Museum of Sou thwestern Biology Special Publication 8:1 191. NAGORSEN, D. W. and G. KEDDIE (2000) Late Pleistocene Mountain Goats ( Oreamnos Americanus ) From Vancouver Island: biogeographic Implications Journal of Mammalogy 81(3): 666 675. NAGORSEN, D. W. G. KEDDIE, a nd R. J. HEBDA (1995) E arly Holocene black bears, Ursus americanus from Vancouver Island Canadian Field Nat uralist 109(1):11 18. PAETKAU, D., G. F. SHIELDS, and C. STROBECK (1998) Gene flow between insular, coastal and interior populations of brown bears in Alaska. Molecular Ecology 7:1283 1292. RAMSEY, C. L., P. A. GRIFFITHS, D. W. FEDJE, R. J. WIGEN, and Q. MACKIE (2004) Preliminary investigation of a late Wisconsinan fauna from K1 cave, Queen Charlotte Islands (Haida Gwaii), Canada. Quaternary Research 62(1):105 109. STONE. K. D., and J. A. COOK (2000) Phylogeography of black bears (Ursus americanus) of the Pacific Northwest. Canadian Journal of Zoology 78:1218 1223. TALBOT, S. L., and G. F. SHIELDS (1996) Phylogeography of brown bears ( Ursus arctos ) o f Alaska and paraphyly within the Ursidae. Molecular Phylogenetics and Evolution 5:477 494.

PAGE 14

CONTRIBUTED PAPERS IN THE BIOLOGICAL SCIENCES

PAGE 16

15th International Congress of Speleology Biological Sciences 1277 2009 ICS Proceedings HOt T Ca A Ve E Record ECORD in IN Me E Xico ICOSAL A A GUILAR, A A DA R R UIZ, JUAN MORALEs S MALACARAL L aboratorio de AA carologa, DDepartamento de Biologa CC omparada. Facultad de CC iencias, UU niversidad NN acional AA utnoma de Mxico, DDelegacin CC oyoacn, CC.P. 04510 DD istrito Federal. Mxico, saulaguilar@hotmail.com Abstract e ecology of bat shelters has been widely reviewed. Many bats require shelter to protect themselves from predators. For others, the shelter requirements are determined by physiological demands of the adults and the young, by social considerations, or by morphological aspects. Selection of habitat is one of the main factors in bat survival. e body temperature and the metabolic rate of the bats are dependent on the ambient temperature except when it is modied by gregarious social behavior. ere are some species that have developed a microclimatic selection very close to their shelter and show physiological adaptations and behavior unique to them. e most important microclimatic factors that intervene in the selection of a bat shelter are temperature and relative humidity. Some authors have found that the selection of a fresh shelter during active periods help facilitate bat digestion, pregnancy, growth, and development of the young. Nevertheless, there are studies recording shelters with very high temperatures and humidity (heat caves) that are selected by the bats. Previous studies have speculated that Mexico has this kind of shelters, but none have been identied. is kind of hyperthermal caves are know as hot caves, hot grottos, or hot caverns. eir most distinctive feature is that biological conditions modify the interior climate, enhanced by the speleomorphologic accident of a single entrance access. Animal populations are established, radically changing the physical conditions of the cave and, thus, modifying the exanten ecology. is activity makes for a unique population composition, density, and dynamics. No other faunal community is known to support a similar biomass with these characteristics in a conned space. e populations of bats vary depending on the type of cave and the zone where they are located.RCORDS DE CUEV V AS CALIENTES EN MX X ICOResumen La ecologa de los refugios de los murcilagos ha sido ampliamente revisada por algunos autores. Muchos murcilagos requieren de un refugio para protegerse de los depredadores y para algunos otros los requerimientos del refugio estn determinados por demandas siolgicas de los adultos o jvenes, por consideraciones sociales, o por cuestiones morfolgicas. La seleccin del hbitat es uno de los factores fundamentales en la sobrevivencia de los murcilagos. La temperatura corporal y la tasa metablica de los quirpteros son dependientes de la temperatura de su entorno excepto cuando son modicadas por el comportamiento social gregario. Existen algunas especies que han desarrollado una seleccin microclimtica muy estrecha hacia sus refugios y presentan adaptaciones siolgicas y conductuales muy particulares hacia los mismos. Los factores microclimticos ms importantes que intervienen en la seleccin del refugio para los murcilagos son la temperatura y la humedad relativa. Algunos autores han encontrado que la seleccin de un refugio fresco, durante periodos activos permite realizar mejor la digestin, gestacin, crecimiento y desarrollo de las cras. Sin embargo, existen estudios en donde se registran refugios con temperaturas y

PAGE 17

Biological Sciences 1278 2009 ICS P roceedings 15th International Congress of Speleology humedades muy altas (cuevas de calor) y que de la misma forma son escogidos por los murcilagos. En los pases en donde se han realizado estos estudios, se cree que Mxico posee refugios de este tipo, pero no se tiene conocimiento de alguno de ellos. Por lo que es de gran importancia el detectar este tipo de refugios y determinar las caractersticas que impera en estos nichos. A este tipo de cuevas trmicamente anmalas se les conoce como cuevas de calor, grutas de calor, o cavernas de calor; lo ms distintivo de estas cuevas o grutas es que el clima interno es modicado por condiciones biolgicas, ya que al aprovechar un accidente espeleomorfolgico de un slo acceso, se establecen poblaciones de animales que cambian radicalmente las condiciones fsicas de la cueva, modicando as, la ecologa existente, tornndola exclusiva por su composicin, densidad y dinmica poblacional. No se conoce ninguna otra comunidad faunstica que mantenga, en un sitio cerrado, una biomasa semejante y con caractersticas tan propias. Esas poblaciones de animales varan de acuerdo al tipo de cueva y a la zona.

PAGE 18

15th International Congress of Speleology Biological Sciences 1279 2009 ICS Proceedings EPIKARST MICROBIAL ASSEMBLAGES IN PADUREA CRAIULUI MOUNTAINS NW ROMANIATRAIAN RAIAN BRAD RAD ADRIANA ADRIANA BICA ICA IOANA IOANA MELEG ELEG OANA OANA MOLDO OLDO VAN ANI I nstitute of Speleology E E mil RR acoita, CC linicilor 5, 400006 CC luj-N N apoca, RRomania Abstract Due to absence of light below ground, and thus the consequent lack of photosynthesizing organisms, microorganisms, and dead organic materials transported from the surface, provide the basis of any groundwater food web. We aim to relate the abundance of microbial assemblages to the presence and density of groundwater fauna in the epikarst of Padurea Craiului Mountains (NW Romania). For that purpose, we have determined the density of aerobic heterotrophic bacteria, coliform microorganisms, and various physiological microbial groups (i.e. Fe-reducing bacteria, ammonifying bacteria, denitrifying bacteria) in samples of water dripping in three caves. Our analyses have revealed that the sampled groundwater is not contaminated by surface microora, i.e., it is free of coliform microorganisms (EE n terobacteriaceae). Although the estimation of the abundance of aerobic heterotrophic bacteria showed low numbers of viable cells (0.4.9 CFU/mL), other physiological tests revealed relatively larger numbers of Fe-reducers (1012 cells/mL), ammonifying bacteria (425 cells/mL) and denitrifying bacteria (345 cells/mL). e relatively large number of microorganisms involved in the nitrogen cycle can be related to ammonication and denitrication processes occurring in the aboveground soils through which percolating water passes. In general, the abundance of microorganisms was larger in locations where there was a larger population density of groundwater fauna and conversely at locations with lower density of groundwater fauna, fewer microorganisms were detected.

PAGE 19

Biological Sciences 1280 2009 ICS P roceedings 15th International Congress of Speleology GUANOPHILE ECOSYSTEM OF AN URBAN CAV V E POLLUTED WITH RAW SEWAGEGREGGREG BRIC RIC KD Department of GG eology and GG eophysics, UU niversity of Minnesota, Minneapolis, MN N 55455 UU SA A Abstract Schieks Cave is a natural maze cave in the St Peter Sandstone. e cave is located 23 m below the streets of downtown Minneapolis, Minnesota, USA, and extends under one city block. When rst discovered in 1904 the cave was barren. Sewer connections from the buildings above were constructed down through the cave and for many years sewage has been leaking into the cave, creating a large pool of stagnant sewage known as the Black Sea. In the 1970s and 1980s enormous numbers of cockroaches inhabited the cave, blackening the walls. By the 1990s there was a switch to an oligochaete/dipteran assemblage. Fungi grow upon the human feces and fungus gnats in turn live upon the fungi. e gnats are preyed upon by webbuilding spiders. Earthworms, entering the cave by stormwater overows, are found in large numbers along the edges of sewage pools. No bats or rats have ever been reported in the cave. Schieks Cave seems to t the model of a hothouse fauna as described from other polluted caves.

PAGE 20

15th International Congress of Speleology Biological Sciences 1281 2009 ICS Proceedings DESCRIPTION OF THE BIOTA ON A REPRESENTATIV V E SAMPLE OF THE SYSTEMS OF KARST GUANENTINA REGION AND V V LEZ PROV V INCE SANTANDER, COLOMBIADD IEGO C C As S ALLAs S PAb B N1, MARIO A A NDRs S MURCIA L L p P EZ2, YANEt T H MUOZSAb B A3 1Estudiante de Biologa, Laboratorio de Mamferos, Instituto de Ciencias Naturales, Universidad Nacional de Colombia, Bogot D.C., Colombia, Asociacin Espeleolgica Colombiana (ESPELEOCOL), casallas45@hotmail.com2Bilogo, Fondo de Biocomercio CC olombia, Bogot DD .C C., CC olombia, mariomurcia30@gmail.com3Profesora AA sociada, II nstituto de CC iencias NN aturales, UU niversidad NN acional de CC olombia, sede Bogot DD .C C., AA.A A. 7495, Bogot D D .C C., CC olombia, Vicepresidenta de la AA sociacin EEspeleolgica CC olombiana (EESPELEOCOL ELEOCOL), ydmunozs@unal.edu.co Abstract In the oriental mountain of the Colombian Andes is found the department of Santander, it is even an karstic area with diversity of formations rocky and great quantity of cavernous systems without exploring; cavities that were formed make millions of years for given way for the action of currents of underground rivers on these. e parental lithology is calcareous rocks. e speleothems in these caves are abundant in their great majority and of a lot of variety. Between 2007 and 2008 were conducted surveys and biological inventories in 32 of these caves in Santander, by identifying specimens collected manually and with some type Pitfall traps. We found approximately 70 families of arthropods, with the crickets, beetles, cockroaches, spiders and Opiliones more common to the naked eye. In addition to that reported 14 species of mammals, of which 13 are Bats, mainly of the genus CC ar ollia, Sturnira and AA rtibeus. ere is evidence of the presence of other groups Cordata, however, were rare. With the collected specimens were calculate the species richness for each cave, and relative abundances in those sampling methods allowed it, according to its topography and ease of sampling. Combine greater wealth and abundance of specimens to the caves with high humidity and greater quantity of guano. In addition to this, it was found that species diversity decreases with the presence of high ow of tourism to the interior of the system. In analyzing the physical conditions and environmental visited the caves and their relation to the composition and richness of the taxa found, it was shown that many of the caves have a similar composition and assembly of fauna, with abundance of troglophile species, few of type and trogloxenous and very rare type troglobies, which implies that the processes of colonization and adaptation of fauna species in tropical cave environments, is given more by processes of colonization of new niches available in the area, which the colonization of these environments, driven by large changes in climate that force species to carry out vertical migration towards more favorable and stable environments in order to prevent its demise (as in the non-tropical caves). is is explained by the large number of groups within the caves that are equal or similar to those present on the exterior of the caves, as evidenced by the abundance of this troglophiles groups and the few troglobite groups.Descripci ESCRIPCI n N de DE la LA biota BIOTA de DE una UNA muestra MUESTRA representati REPRESENTATI Va A de DE los LOS sistemas SISTEMAS c C rsticos RSTICOS de DE regi REGI n N Guanentina UANENTINA y Y Pro RO Vincia INCIA de DE Vlez VLEZ Santander ANTANDER Colombia OLOMBIAResumen En la cordillera oriental de los andes Colombianos se encuentra el departamento de Santander, se trata de una zona crstica con diversidad de formaciones rocosas y gran cantidad de sistemas cavernosos aun sin explorar; cavidades que se formaron hace millones de aos por oradacin dada la accin de corrientes de ros subterrneos sobre estas. El material parental rocoso principal que las compone esta conformado por

PAGE 21

Biological Sciences 1282 2009 ICS P roceedings 15th International Congress of Speleology areniscas y rocas calizas masivas. Los espeleotemas presentes en estas son abundantes en su gran mayora y de mucha variedad. Entre el 2007 y 2008 se realizaron reconocimientos e inventarios biolgicos en 32 cuevas de estas en Santander, mediante la identicacin de especimenes colectados manualmente y en algunas con trampas tipo Pitfall. Se encontraron aproximadamente 70 familias de artrpodos, siendo los grillos, escarabajos, cucarachas, araas y opiliones los ms comunes a simple vista. Adems de esto se reportan 14 especies de Mamferos, de los cuales 13 son Murcilagos, principalmente de los gneros CC arollia, Sturnira y AA rtibeus. Se evidencia la presencia de otros grupos de cordados, sin embargo fueron poco frecuentes. Con los especimenes colectados se calculo la riqueza especca para cada cueva, y las abundancias relativas en aquellas que los mtodos de muestreo lo permitieron, de acuerdo a su topografa y facilidad de muestreo. Se asocian las mayores riquezas y abundancias de especimenes a las cavernas con alta hmedas y con mayor cantidad de guano. Adems de esto, se hall que la diversidad de especies disminuye con la presencia de alto ujo de turismo al interior del sistema. Al analizar las condiciones fsicas y ambientales de las cuevas visitadas y relacionarlas con la composicin y riqueza de los taxones encontrados, se evidencio que muchas de las cuevas poseen una composicin y ensamblaje similar de fauna, con abundancia de especies trogolo, pocas de tipo trogloxeno y muy escasas las de tipo troglobio; lo que hace suponer que los procesos de colonizacin y de adaptacin de las especies de fauna en los trpicos a los ambientes cavernosos, est dada mas por procesos de colonizacin de nuevos nichos disponibles en el medio, que por la colonizacin de estos ambientes, impulsado por grandes cambios en el clima que fuerzan a las especies a realizar emigraciones verticales hacia ambientes ms favorables y estables, a n de evitar su desaparicin (como en las cuevas no tropicales). Esto es explicado por la gran cantidad de grupos dentro de las cavernas que son iguales o similares a los presentes en el exterior de las cavernas, evidenciado esto por la abundancia de grupos troglolos y los pocos grupos troglobios.

PAGE 22

15th International Congress of Speleology Biological Sciences 1283 2009 ICS Proceedings PRELIMINARY REPORT ON THE CAV V E DIPLURA OF COLORADO HEX X APODA: DIPLURA: CAMPODEIDAELL YNN NN M. FERGU ERGU SON OND Department of Biological & EE nironmental SciencesL Longwood UU niversity, Farmville, VA A 23909 UU SA A ; fergusonlm@longwood.edu Two-pronged bristletails, or campodeid diplurans, have been collected from thirteen caves and a gold mine in the state of Colorado. All but two of the caves are inhabited by a dierent species of campodeid. e species found in the mine may be represented by one of the cave inhabiting forms, or it could be a unique species as well. e taxa represented are: E E u mesocampa lutzi Tricampa rileyi two new undescribed species of Metriocampa, and eight or nine new undescribed species of HH ap locampa. EE umesocampa lutzi was originally described from an epigean habitat in Colorado. Tricampa rileyi is a wide ranging endogean species which is also known from Colorado. One of the species of Metriocampa is unique to Colorado, but the other species has also been found in caves in Missouri and Illinois. e potentially nine species of HH aplocampa from Colorado represent 35 percent of all HH ap locampa species known from caves (total number of identied species from caves: 26). e other cavernicolous HH ap locampa species are known from the northwestern United States and Vancouver Island, British Colombia, Canada.1. IntroductionWhen I reported on the cave diplura of the United States (FERGUSON, 1981) at the Eighth International Congress of Speleology held in Bowling Green, Kentucky, to my knowledge no cavernicolous campodeid diplurans had been collected from the state of Colorado. However, DAVIS (1971) had reported the possible sightings in 1970 of a campodeid dipluran in Knoxs Nasty Pit and FixinTo-Die Cave. Both of these caves are near Groaning Cave in Gareld County. HUBBARD (1998; SHEAR & HUBBARD, 1998) reported on his faunal discoveries in seven Colorado caves in 1996, which included four collections of campodeid diplurans. e remaining cavernicolous dipluran collections have been made by David Steinmann and Cyndi Mosch between 1999 and 2007.2. Materials and MethodsSamples examined are from the Denver Museum of Nature & Science collection and the David A. Hubbard, Jr. collection. Whole specimens were mounted ventral side up on glass microscope slides in Hoyers or Andr II media along with any loose cerci and antennae, and oen with two removed hind legs. e slide-mounted specimens were examined using an Olympus compound microscope Figure 1: Distribution of cavernicolous diplurans in Colorado.

PAGE 23

Biological Sciences 1284 2009 ICS P roceedings 15th International Congress of Speleology equipped with Normarski interference diraction contrast, an ocular micrometer, a drawing tube, an Ikegami analog camera, and a Sony videographic printer. Cave location data was from PARRIS (1973), RHINEHART (2001), and D.B. Steinmann (pers. comm.).3. Results and DiscussionMy examination of the 16 collections from 13 caves and one gold mine has revealed 13 distinct taxa (Fig. 1 and Table 1). All but two of the caves are inhabited by a dierent species of campodeid. e species found in the mine may be represented by one of the cave inhabiting forms, or it could be a unique species as well, and is treated as such in this report. e taxa include EE u mesocampa lutzi SILVESTRI 1933, an endogean species which is known from north central Colorado in Larimer, Boulder, and Gilpin counties at altitudes around 3000 meters (COND & GEERAERT, 1962). It has now been discovered in Marble Cave in Fremont County. Tricampa rileyi (SILVESTRI, 1933) has been found in Squeak Cave No. 1 in Gareld County and in Sulphur Cave in Routt County. It is a wide ranging endogean species which has been collected at Cottonwood Pass, Colorado. At an altitude of 3615 meters, this is one of the highest locations known for campodeid diplurans. Two species of Metriocampa are represented in the examined collections. e species from Hourglass Cave in Eagle County is unique to Colorado, but the other species from Manitou Cave in El Paso County has also been found in caves in Missouri and Illinois. e remaining species belong to the genus HH ap locampa SILVESTRI 1933 (Fig. 1 and Table 1). e potentially nine new species of HH ap locampa from Colorado represent 35 percent of all HH ap locampa species known from caves (total number of identied species from caves 26). e other cavernicolous HH ap locampa species are known from lava tubes and a few limestone caves primarily in the northwestern United States (FERGUSON, 1992) and Vancouver Island, British Colombia, Canada.4. AcknowledgmentsI thank the collectors, David A. Hubbard Jr, David B. Steinmann, Cyndi Mosch, and Dr. Frank T. Krell, Curator of Entomology, Denver Museum of Nature & Science, for the loan of the material under his care (Loan # OL-200820). I acknowledge the U.S. Department of Agriculture Forest Service for issuing a Special Use Permit to David Steinmann and Cyndi Mosch for the purpose of collecting and preserving cave faunas from the White River National Forest (SUP# BLA1, expiration date 12/31/2003). I also thank Longwood University for the continuing use of equipment and support for my research following my retirement from teaching.5. ReferencesCOND, B. and P. GEERAERT. (1962) Campodids endogs du centre des Etats-Unis. AA r chives de Z Z oologie EE xprimentale et GG nrale, 101,73. DAVIS, D.G. (1971) e Groaning Cave Complex. NN S S N N ews (National Speleological Society), 29(6):68. FERGUSON, L.M. (1981) Cave Diplura of the United States. Pp 11 (vol. 1), in B.F. Beck (ed.), P roceedings of the EE ighth II nternational CC ongress of Speleology (Bowling Green, Kentucky, 1981). Table 1: Locations and taxa of cavernicolous diplurans in Colorado.

PAGE 24

15th International Congress of Speleology Biological Sciences 1285 2009 ICS Proceedings FERGUSON, L.M. (1992) Diplura of Lava Tube Caves. Pp 281, in G.T. REA (ed.), Proceedings of the 6th I I n ternational Symposium on Vulcanospeleology (Hilo, Hawaii, August, 1991). HUBBARD, D.A. JR (1998) Rocky Mountain Caving at the 1996 Convention A Biological Perspective. R R ocky Mountain CC aving, Autumn, 1998. PARRIS, L.E. (1973) CC a ves of CC olorado. Pruett Publishing Company, Boulder, Colorado, 247 pp. RHINEHART, R.J. (2001) CC o lorado CC aves. Westclie Publishers, Inc., Englewood, Colorado, 136 pp. SHEAR, W.A. and D.A. HUBBARD, JR (1998) Cave millipedes of the United States. IV. A new genus and species from a high altitude cave in Colorado (Diplopoda, Chordeumatida, Tingupidae). Myriapodologica, 5(8):85. SILVESTRI, F. (1933) uarto contributo alla conoscenza dei Campodeidae (ysanura) del Nord America. B ollettino del LL aboratorio di ZZ oologia GG eneral e A A graria della RR Scuola Superiore dA A gricoltura in Portici, 27:156.

PAGE 25

Biological Sciences 1286 2009 ICS P roceedings 15th International Congress of Speleology DEV V ELOPMENT AND APPLICATION OF A DATABASE FOR THE SUBTERRANEAN AMPHIPOD CRUSTACEAN GENERA STYGOBROMUS AND BACTRURUSJOHN R R H H OLs S INGER, JUst ST IN SHAf F ER, G G RACE SCHULt T ED Department of Biological Sciences, OOld DDominion UU niversity, NN orfolk, Virginia 23529, UU SA A A species/locality database for all species of the subterranean amphipod crustacean genera Stygobromus and Bactrurus has been developed in MS Access. Both of these genera are exclusively stygomorphic and recorded from a variety of subterranean groundwater habitats. Currently there are 3,055 entries in the database: 2,817 for Stygobromus; 238 for Bactrurus. Species of Stygobromus are numerous and widespread in North America, and a few occur in the Palearctic region, whereas species of Bactrurus occur only in North America east of the Great Plains. At present the database contains information on approximately 273 species of Stygobromus in North America: 129 described, 22 in manuscript and 122 provisionally recognized. Five other species in the genus are recorded from the Palearctic region. e much smaller genus Bactrurus contains only eight described species. e basic collection data for each species include: species name, specic locality, specic collection number; collection date; brief description of habitat (e.g., cave stream, pool; spring; seep); collection date, number of specimens (broken down into sex and level of maturity where possible); name of collector; deposition of type(s) where applicable; identier; nal remarks. Each dierent collection of a given species is treated as a separate entry. Locality and habitat data are as specic as possible, but vary depending on information supplied by the collector. e second part of the project is development of a Geographic Information System (GIS) for the databases utilizing ESRI ArcGIS 9 soware. is process involves creating location features and linking these data points to the existing MS database, which in turn is used to generate a GIS database. e rst step of the GIS creation and data conversion process is to develop spatial data for each cave, spring, seep, well or other habitat/location in which specimens were found. e preferred source of location data are coordinates (latitude/longitude) that accompany the entries stored in the database. Other sources will be used where coordinates are not available. In addition to generation of distribution maps for publications and presentations, the GIS database will be used to plot species distributions against various backgrounds, including karst terrains, drainage basins, physiographic regions, and to distinguish hot spots of taxonomic diversity for further study, analysis and conservation. We also hope to utilize the capabilities of GIS for further map analyses that will include identication and/or determination of dispersal corridors, potential locations for taxa previously overlooked, and localities threatened by land use change. It may also be useful in some instances for assistance in explaining taxonomic relationships between species and species groups. 1. Introductione need for a workable database for species in large genera of subterranean organisms is obvious. Access to pertinent electronic data on these species is necessary for facilitation of both biogeographic research and conservation eorts. To meet these important needs, we have developed a species/locality database in MS Access for all species of the exclusively subterranean crustacean amphipod genera Stygobromus and Bactrurus (family Crangonyctidae). To maximize the utility of these data, we have linked the basic species data to GIS, which allows the rapid construction of species distribution maps on varying backgrounds, such as karst terrains, drainage basins, etc. 2. Discussione amphipod crustacean genera Stygobromus and Bactrurus are exclusively stygomorphic (i.e., lacking eyes and integumentary pigment) and to date all but a few species of Stygobromus have been found in subterranean groundwater habitats in North America. Species of Stygobromus are numerous and widespread in North America, and a few occur in the Palearctic region (see HOLSINGER, 1967, 1978, in press; WANG and HOLSINGER, 2001; SIDOROV et al., 2008), whereas species of Bactrurus as presently known occur only in North America east of the Great Plains (see KOENEMANN and HOLSINGER, 2001). Currently there are approximately 3,060 entries

PAGE 26

15th International Congress of Speleology Biological Sciences 1287 2009 ICS Proceedings in the database: 2,820 for Stygobromus and 240 for Bactrurus. At present the database contains information on approximately 205 species of Stygobromus in North America: 129 described, 22 in manuscript and at least 54 provisionally recognized. Five other species in the genus are recorded from the Palearctic region. e much smaller genus Bactrurus contains only eight described species. e basic collection data for each species include: species name, specic locality, specic collection number; collection date; brief description of habitat (e.g., cave stream, pool; spring; seep; well); collection date, number of specimens (broken down into sex and level of maturity where possible); name of collector; deposition of type(s) where applicable; identier; nal remarks. Each dierent collection of a given species is treated as a separate entry. Locality and habitat data are as specic as possible, but vary depending on information supplied by the collector. e second part of this project is the development of a Geographic Information System (GIS) database utilizing ESRI ArcGIS 9 soware. is process involves creating location features in the GIS and linking these data points to the existing MS database. e rst step of the GIS creation and data conversion process is to develop spatial data for each cave, spring, seep, well or other habitat/location in which specimens were found. e preferred source of location data are coordinates (latitude/longitude) that accompany the entries stored in the database. Other sources will be used where coordinates are not available, with lesser accuracies noted in the locality database. e use of unique locality codes will allow the expansion of the GIS to include any number of related species databases, increasing its usefulness. We also plan to utilize the capabilities of GIS for further map analyses. In addition to generating a variety of maps for relaying specic information from the database (Figs. 1, 2), the GIS database will be used to plot species distributions against various backgrounds, including karst terrains, drainage basins, physiographic regions (Fig. 3). Such data allows a variety of analyses to be conducted, including the identication of hot spots of taxonomic diversity for further study, analysis and conservation, determination of species dispersal corridors, potential locations for taxa previously overlooked, and localities threatened by land use change (VENI, 2003). By comparing species locations Figure 1: Number of entries to the Stygobromus database per state or proince, proiding a relative indication of the number of localities and/or amount of data gathered in each.

PAGE 27

Biological Sciences 1288 2009 ICS P roceedings 15th International Congress of Speleology to data layers such as karst geology, other surrounding geology, and hydrology, both surface and belowground, predictions of new localities for a given species may be generated. e same data, along with historical geology, might provide insight into the dispersal and speciation within the genus throughout its range. e hydrology and karst layers provide specic information on possible routes through which organisms of this genus may disperse, while data on surrounding geology can indicate boundaries to this expansion. e GIS may also be useful in some instances for assistance in explaining taxonomic relationships between species and species groups. Locations of members of species groups will be compared to surrounding, non-grouped species. Combined with cladistic information this may allow the inclusion of these ungrouped species into existing species groups based on geography and similar characters. Lastly, we hope to promote the use of this database as a tool for the conservation of the genera Stygobromus and Bactrurus, along with associated habitats and species in other genera. Locality data might be analyzed against land-use, parcel ownership and development data provided by local and regional organizations to determine which species and locations are most threatened. With these data, long-range plans can be developed to protect these habitats. Acknowledgements We thank James C. Currens and William R. Elliott for assistance with cave locations in Kentucky and Missouri, respectively, and the Cave Conservancy of the Virginias (CCV) for generous grant support of this project. ReferencesHOLSINGER, J.R. (1967) Systematics, speciation, and distribution of the subterranean amphipod genus Stygonectes (Gammaridae). Bu lletin of the UU nited States NN ational Museum 259, 176 pp. HOLSINGER, J.R. (1978) Systematics of the subterranean amphipod genus Stygobromus (Crangonyctidae), Part II: Species of the eastern United States. Sm ithsonian CC ontributions to ZZ oology, 266, 144 pp. HOLSINGER, J.R. (In press) ree new species of the subterranean amphipod crustacean genus Stygobromus (Crangonyctidae) from the District of Columbia, Maryland and Virginia. Virginia Museum of N N atural HH istory Memoir. Figure 2: Number of species of Stygobromus per state or proince.

PAGE 28

15th International Congress of Speleology Biological Sciences 1289 2009 ICS Proceedings KOENEMANN, STEFAN and J.R. HOLSINGER (2001) Systematics of the North American subterranean amphipod genus Bactrurus (Crangonyctidae). Beaufortia, 51(1), 10. SIDOROV, DMITRY, J.R. HOLSINGER, and V.V. TAKHTEEV (2008) Biogeographic signicance of the recent discovery of a new species of Stygobromus (Amphipoda, Crangonyctidae) from groundwaters in Eastern Siberia near Lake Baikal. 19th International Symposium of Subterranean Biology Abstracts, p88. VENI, G (2003) GIS applications in managing karst groundwater and biological resources. Sinkhole and Engineering/environmental Impacts of karst, 466. WANG, DAQING and J.R. HOLSINGER (2001) Systematics of the subterranean amphipod genus Stygobromus (Crangonyctidae) in western North America, with emphasis on the hubbsi group. A A m phipacica 3(2), 39. Figure 3: Locations of the amphipod genus Stygobromus in Virginia, West Virginia and the District of Columbia. Locations include caves, seeps, springs, wells and other groundwater habitats. Darkly shaded areas represent karst terrains in western Virginia and eastern West Virginia. Lightly shaded areas represent scattered karst-like terrains. (Geological data source: USGS).

PAGE 29

Biological Sciences 1290 2009 ICS P roceedings 15th International Congress of Speleology THE CAV V E FAUNA OF TEX X ASJEAN EAN K. KRE RE JCA CA1, JA A ME E S R R REDDELL REDDELL2, GEORGE GEORGE VENI ENI3 1ZZ ara EE nironmental LLCLLC 118 W. GG oforth RR d., Buda, TX 78610 UU SA A2Texas Memorial Museum, 2400 Trinity St., AA ustin, TX 78705 UU SA A3NN ational CC ave and Karst RResearch II nstitute, 1400 CC ommerce DDrive, CC arlsbad, NN M 88220 UU SA A Abstract Cave biology studies in Texas started in 1895 with the emergence of a remarkable blind salamander and crustaceans from an artesian well in San Marcos. e range of topics studied in Texas caves is varied, and the review of these topics gives insight into one of the most diverse regions in the world for cave biology; as well as how the changing modern landscape aects those species. Topics include taxonomy, ecology, conservation research and biogeography. Regarding taxonomy, the distribution of collecting eorts demonstrates the activity of biospeleologists and cavers, and recent discoveries in well-collected areas (e.g. a blind weevil from Austin, new beetle localities in San Antonio) demonstrate the high potential for new species yet to be found. e number of troglobitic species described per year began to accelerate in the 1960s and is at an all time high in the 1990s and 2000s, with a record of 78 cave species described in one year. A new era of cave related research is in progress, driven by an active community of cave biologists and the rapid urbanization of karst that triggered the federal listing of 16 terrestrial troglobites and 6 aquatic species. Ecology studies include whole ecosystem inventories initiated at caves with endangered species; these show declines in keystone species associated with urbanization. Nutrient ow research indicates dierences in isotopic composition at urban and rural caves. In order to determine and defend the constituent elements aecting species with conservation needs, researchers have called upon biological topics, such as the foraging of cave crickets outside of a cave, and hydrogeological topics, such as the delineation of drainage basins of vadose zone caves. Behavioral studies include spring dwelling salamanders and troglobitic beetles. Regional inventories, species specic inventories, and population studies of rare, threatened and listed species are focused on the Balcones Fault Zone region. Texas is divided into karst regions based on geomorphology, hydrogeology, stratigraphy, structure, and cave density and types. e attributes of each of these karst regions aects the available subterranean habitats for karst species, therefore is a framework we use for examining the distribution of Texas cavernicoles. e distribution patterns for selected species demonstrate both collecting eort and geologic controls. Recently, researchers have tested biogeographic hypotheses using molecular phylogenetic techniques for aquatic and terrestrial cave species.

PAGE 30

15th International Congress of Speleology Biological Sciences 1291 2009 ICS Proceedings WHITENOSE SYNDROME IN HIBERNATING BATS: ARE THESE A FFECTED BATS THE NEX X T CANARY IN THE MINE?THO HO MA A S H H KUNZ UNZ1, DA DAVID ID S. BLEHER LEHER T2, PAUL AUL M. CR CR YAN AN3, JERE ERE MY T.H H COLE COLE MAN AN4, ALAN ALAN HIC HIC KS5, AND AND MERLIN ERLIN D D TU U TTLE LE6 1CC enter for EEcology and CC onservation Biology, Boston UU niversity, Boston, MA A 02215, UU SA A2NN ational Wildlife HH ealth CC enter, UU .S. GG eological Survey, Madison, WI I 53711 UU SA A3Fort CC ollins Science CC enter, UU .S. GG eological Survey, Fort CC ollins 80526 COCO, UU SA A 4UU .S. Fish and Wildlife Service, NN ew York Field OO ce,C C ortland, NN Y 13045, UU SA A5NN ew York State DDepartment of EE nironmental CC onservation, A A lbany, NN Y 12233, UU SA A6Bat CC onservation II nternational, 500 NN orth CC apital of Texas HH ighway,A A ustin, TX 78746, UU SA A Abstract White-Nose Syndrome (WNS) is one the most devastating conditions ever reported for bats in North Americawith losses exceeding 90% in some hibernacula in New York State during the winter of 2007. WNS was rst recorded photographically in Howes Cave, New York, on February 16, 2006. By January 2007, WNS had been documented in nearby Schoharie Caverns, and later that winter in four other hibernacula in New York State, all located within a 15-km radius of a point in Schoharie County, Albany, New York. By March 2008, bats with symptoms of WNS had been observed in hibernacula from at least four states (New York, Vermont, Massachusetts, and Connecticut). Current evidence indicates that several hundred thousand bats have died from conditions associated with WNS in this relatively small area. To date, fatalities associated with WNS have been reported for six of the nine species of bats in the northeastern United States. Field observations have shown that bats aected by WNS are characterized by some or all of the following: (1) a white fungus that grows on the nose, ears, and wing membranes; (2) severely depleted white and brown fat reserves by mid-winter; (3) reduced capacity to arouse from deep torpor; (4) an apparent lack of immune response during hibernation; (5) ulcerated, necrotic and scarred wing membranes; and (6) atypical behavior of bats emerging prematurely from hibernacula in mid-winter. Laboratory studies of bats aected with WNS have isolated a previously undescribed psychrophilic fungus, closely related to GG eo myces spp. that grows optimally at temperatures characteristic of hibernacula. Histological evidence has shown that this fungus sometimes penetrates the dermis, especially in areas associated with sebaceous glands and hair follicles. Genetically identical isolates of this fungus have been collected from aected bats located in widely dispersed hibernacula in the northeastern United States. Preliminary data suggest that concentrations of chlorinated hydrocarbons, pyrethroids, and heavy metals are not markedly elevated in bats thus far examined, nor have known bacterial or viral pathogens been identied. Narrowing the eld of potential causative agents will require an understanding of whether the fungus associated with WNS is pathogenic. Both eld and laboratory investigations are underway to determine the geographic distribution of this fungus, and to assess how it might be directly or indirectly contributing to the demise of hibernating bats in the northeastern United States.

PAGE 31

Biological Sciences 1292 2009 ICS P roceedings 15th International Congress of Speleology Effects FFECTS of OF incubation INCUBATION conditions CONDITIONS on ON uantification UANTIFICATION of OF chemoheterotrophic CHEMOHETEROTROPHIC bacteria BACTERIA from FROM ca CA Ves ES KAt T HLEEN H H L L AvV OIE1, D D IANA E E N N ORt T HUp P2, JEss SS ICA R R SNIDER2, and Nw NW Am M Ak K A A A Nw NW AGb B OLOGU2 1101 HH udson HH all, State UU niversity of NN ew York CC ollege at Plattsburgh, Plattsburgh, NN Y 12901 UU SA A lavoiekh@plattsburgh.edu2 MSC C03 2020, UU niversity of NN ew Mexico, AA lbuquerque, NN M 87131-0001 UU SA A dnorthup@unm.edu Isolation of microorganisms by culture techniques grows less than 1% of the microorganisms in an environmental sample using standard cultivation techniques. Despite this inherent limitation, culturebased studies do have value as they shed light on the metabolic properties and physiology of the microorganisms. A common practice in cave microbiology studies is a period of in-cave incubation on the premise that it contributes to higher growth of more diverse microorganisms than returning sa mples to the surface  before plating. We performed a comparative study of surface vs. in-cave incubation conditions for chemoheterotrophic bacteria from multiple soil and water samples from Great Onyx C ave in Mammoth Cave National Park (MACA), Ky, USA, and Carlsbad  Cavern, Carlsbad Caverns National Park, NM, USA. Samples were diluted and  plated in triplicate in the caves on R2A medium. One set of plates  was le in the cave for incubation for 48 hours, and a duplicate  set was brought to the surface for incubation at cave temperatures. e  same samples were also diluted and plated on the surface. At MACA, one set  was incubated on the surface under cave conditions, and a duplicate  set brought to the caves for incubation. All samples were counted aer 48 hours. MACA plates were le at r oom temperature on the surface and  counted again aer another 48 hours. ere were no dierences in numbers of microbes recovered under any incubation conditions, except for an increase in numbers when the MACA plates were incubated for an additional 48 hours at room temperature. Diversity from the MACA samples as evidenced by appearance of colonies was also identical, except for a few temperature sensitive, pigment-producing colonies. We conclude that cave incubation for isolation and quantication of che moheterotrophic bacteria oers no advantages compared to  removing samples to the surface for cultivation. Because these studies were carried out on samples from o-trail areas of show caves, we do continue to recommend a period of cave incubation for maximum isolation of more specialized or unique cave bacteria. Future studies will repeat the study with sampling in wild caves with areas of low visitation by humans and will explore whether this relationship holds true for chemolithotrophic cave microorganisms.1. IntroductionPlate count studies have traditionally been used in microbial ecology until the advent of culture free techniques, particularly genetic analysis. Many studies of microbes in caves rely on plate count methods that greatly underestimate the total population and tell nothing about the activity of the microbes in that environment. Isolation of microorganisms by culture techniques grows less than 1% of the microorganisms in an environmental sample using standard cultivation techniques (Amann et al. 1995). Plate counts do show many of the same species isolated from caves compared to surface environments, albeit in lower densities from caves. Comparisons of cave silts with rich agricultural soils were made by Gounot (1967) at the CNRS in France. Cave soils tended to have fewer bacteria, yielding several million per gram, while the agricultural soils yielded several hundreds of millions per gram. Most other studies support her general conclusions. Studies of the biomass and activity of microbes in limestone caves in Kentucky were conducted by Feldhake (1986). He concentrated his studies on actual measurements of microbial metabolic rates in 12 sites in four caves, with comparisons to overlying forest soils. Except for a site rich in cricket guano, Feldhake found that organic matter content, microbial activity and biomass were much lower in the cave than in forest soil. He also found signicant variations among sample sites within the cave, and methodological problems when samples were removed from the cave, transported to the lab, and assays made more than 24 hours aer collection. One notable exception to studies that show low numbers and activity of microbes in caves has been work done by Rusterholtz and Mallory (1994). ey compared the microbial activity, density, and diversity of two aquatic

PAGE 32

15th International Congress of Speleology Biological Sciences 1293 2009 ICS Proceedings sediment sites in Mammoth Cave. e study includes counts of cells in the sediment, staining to determine metabolic activity of soil microbes, plate counts using both high and low nutrient media, followed by extensive physiological testing of isolates from the plate counts. ey recovered between 11% of the total cell count on culture medium. e recovery rate for most surface soils is typically 0.4.7%. ey also detected active metabolism in 53% of the population, despite very low nutrient levels of total organic carbon per liter of water. e diversity of populations was extremely high, with 42% of the isolated species similar to surface organisms, with the remainder unidentied. ere were no dominant species, and the type of growth medium used strongly inuenced the types isolated. Despite the inherent limitations, culture-based studies do have value as they shed light on the metabolic properties and physiology of the microorganisms. A common practice in cave microbiology studies is a period of in-cave incubation, as recommended by Rusterholtz and Mallory (1994), on the premise that it contributes to higher growth of more diverse microorganisms than returning samples to the s urface  before plating. We performed a study of the eects of surface vs. in-cave incubation conditions for cultivation of chemoheterotrophic bacteria from multiple soil and water samples from Great Onyx Cave in Mammoth Cave National P ark, Kentucky, USA, and Carlsbad  Cavern, Carlsbad Caverns National Park, New Mexico, USA2. Methods and MaterialsSamples. Water and Soil samples were collected aseptically from the following locations in Great Onyx Cave in Mammoth Cave National Park, Kentucky, and Carlsbad Cavern in Carlsbad National Park, New Mexico. Great Onyx is a small cave (approximately 2 km of passage in Flint Ridge) not connected to the larger Flint-Mammoth System. A small lantern light tour enters the cave and proceeds down Edwards Avenue. e tour takes place, at most 2 times a day. Carlsbad Cavern, located in the Guadalupe Mountains in southeastern New Mexico, is approximately 316 m deep and almost 50 km in length and formed through sulfuric acid driven speleogenesis. Established as a national park in 1930, Carlsbad receives approximately 300,000 visitors per year. Located beneath the Chihuahuan Desert, Carlsbad Cavern is a more oligotrophic habitat than Mammoth Cave. Great Onyx Cave: Two soil samples and two water samples were collected aseptically from Great Onyx Cave from areas with little to no human impact. Soils collected were mud from Cox Avenue and a sandy soil sample from a small side passage o Edwards Avenue. Water samples were collected from dripping water at the base of Edwards Dome and from a small cryptic stream in Bubbly Pit. Carlsbad Cavern: Two soil samples and two water samples were collected aseptically from Le Hand Tunnel, which branches from the Lunch Room in Carlsbad and has one visitor tour of a maximum of 12 people per day. Soils collected were sand from the Beach area of Le Hand Tunnel and dried mud from Le Hand Tunnel. Water samples were collected from two pools, one from a moonmilk area and one from Iron Pool, which has previously been reported to contain iron-oxidizing bacteria. is area receives moderate human impact. Procedures. All soil samples were thoroughly mixed, and a 1 cc subsample mixed with 9 mL of sterile water. A 1 mL aliquot of each water sample was diluted in sterile water. A 1:10 dilution series in sterile water was made of each sample, and 1 mL spread plated in triplicate on R2A Agar (Difco). Samples were diluted and plated in the cave, and one set of triplicate plated samples was incubated in situ in the cave and a duplicate set of plates were removed from the cave, transported on ice to a surface facility, and incubated under cave conditions (temperature and humidity in the dark) on the surface. e same soil and water samples were again diluted and plated on the surface, with one set of plates incubated on the surface under cave conditions, and a duplicate set of plates brought to the cave site for incubation. Plates were counted aer 48 hours of incubation. All plates from Great Onyx were then incubated at room temperature on the surface for another 48 hours and recounted. One additional experiment was done only with the Great Onyx samples. e original soil and water samples were held under refrigeration for 48 hours, and new dilutions made. One set was incubated in the cave and the second under cave conditions on the surface. e plates were counted aer 48 hours. e Carlsbad Cavern results were tested with a KruskalWallis ANOVA for the combined inoculation location and incubation temperature to determine if there is a signicant dierence in the data. For the temperature of incubation and site of inoculation, we performed sign tests, since the data of both data sets were non-normal. 3. ResultsAs shown in Table 1 there were no dierences in numbers of microbes recovered aer 48 hours under any incubation conditions.

PAGE 33

Biological Sciences 1294 2009 ICS P roceedings 15th International Congress of Speleology e Great Onyx plates were all held at room temperature for an additional 48 hours (Table 2), with no change in the numbers of colonies, except for water samples from Edwards Dome, which grew between 0 CFU, but still not signicantly dierent than zero. 4. DiscussionR2A agar was selected because is a low nutrient microbial growth medium intended for recovery of bacteria from potable water samples,\ which are low in nutrients as we would expect in the cave environment. In combination with low temperatures and longer incubation times, R2A agar will improve recovery of stressed bacteria from samples (Reasoner and Geldreich, 1985; Koch, 1997). e amount of growth is inversely proportional to the temperature, and counts are signicantly higher from water samples using a nutrient rich medium such as PCA (Plate Count Agar). Overall, Carlsbad Cavern samples were one to two orders of magnitude lower than Great Onyx Cave soil samples, but the CFU were higher for Carlsbad water samples. Numbers of CFU per cm3 of soil were overall low. ere were no dierences in numbers of microbes recovered under any incubation conditions aer 48 hours regardless of where the dilution series were made or where samples were subsequently incubated (P= 0.082 for the overall ANOVA, 0.2668 for the sign test of incubation temperature, and 0.6875 for the test of incubation site for the Carlsbad Sample Cave dilution: Cave incubation Cave dilution: Surface incubation Surface dilution: Cave incubation Surface dilution: Surface incubation GOCA Mud 8.4 x 10411.0 x 10411.2 x 10411.3 x 104GOCA Sandy soil3.0 x 1043.0 x 1043.1 x 1043.0 x 104GOCA Water Edwards Dome No Growth* No Growth* No Growth* No Growth* GOCA Water Bubbly Pit0-40* 0-40* 0-40* 0-40* CC-LHT 38 Moonmilk soil 7.7 x 103NA 1.3 x 1034.0 x 103CACC-LHT 40 Beach sandy soil No Growth* NA 2.2 x 1033.0 x 103CC-LHT 41 Soil, red clay-silt 1.0 x 102NA 8.3 x 1031.0 x 105CC-LHT 36 Water Moonmilk area 1.3 x 102NA 1 x 1025.7 x 102CC-LHT 39 Water from Iron Pool 4.0 x 102NA 6.0 x 1025.3 x 102* Not statistically signicant. Sample Cave dilution: Cave incubation Cave dilution: Surface incubation Surface dilution: Cave incubation Surface dilution: Surface incubation GOCA Mud 1.96 x 1062.2 x 1067.7x 1067.1 x 106GOCA Sandy soil1.69x 1061.7 x 1067.4 x 1067.7 x 106GOCA Water Edwards Dome 5.3 x 1034.9 x 1036 x 102No Growth* GOCA Water Bubbly Pit 0-100* 0-100* 2 x 1030-100** Not statistically signicant. Table 2. Total number of CFU aer 48 hours of incubation in the cave or under cave conditions on the surface, followed by an additional 48 hours of incubation of all samples at room temperature on the surface. GOCA Great Onyx Cave, MACA. Table 1. Total number of CFU aer 48 hours of incubation. GOCA Great Onyx Cave; CC-LHT Le Hand Tunnel of Carlsbad Cavern.

PAGE 34

15th International Congress of Speleology Biological Sciences 1295 2009 ICS Proceedings Cavern data). Diversity as determined by a visual inspection of plates showed dierences by site in the Great Onyx samples. e Mud sample from Cox had very low diversity, with only two colony types predominating. e sandy soil from Edwards had greater diversity in colonial appearance, and a few colonies of what was probably a temperaturesensitive variety that produced colored colonies under surface incubation conditions, but not in the cave. Subsequent reincubation of the Great Onyx plates at Room Temperature resulted in a large increase in the number of CFU recovered from all of the samples. Culturable CFU from soil samples increased by 10-100 fold. Water samples were much higher for the Edwards Dome sample, and higher, but not statistically signicant, for the Bubbly Pit sample. Contamination and overgrowth of plates by fungi became a problem aer the 48 hrs at room temperature. Storage of Great Onyx samples in the refrigerator for 48 hours and plating a new dilution series had no impact on numbers (results not shown) aer 48 hours of incubation in the cave and on the surface compared to the samples plated immediately aer collection. Many microbes identied from deep caves are identical to surface forms, opportunistic and active only under favorable growth conditions (for example, Dickson and Kirk 1976, James 1994). Most are non-resident chemoheterotrophs transported into caves by water, air, sediment and animals. However, these enrichment-based and cultural studies have focused on typical heterotrophic microbes known from surface studies; and such techniques have been shown to grow less than 1% of microbes present in an environment (Amann et al. 1995). Culture-independent, molecular phylogenetic techniques have since shown that many novel organisms can be found in caves (Angert et al. 1998; Vlasceanu et al. 2000).5. ConclusionsWe conclude that a period of in-cave incubation for isolation and quantication of chemoheterotrophic bacteria oe rs no advantages compared to  removing samples to the surface for cultivation from samples from moderate to low human impact areas of caves. Refrigeration of samples for up to 48 hours prior to dilution had no eect on numbers of CFU recovered from Great Onyx Cave. Because these studies were carried out on samples from o-trail areas of show caves using only one type of medium, we continue to recommend a period of in-cave incubation for maximum isolation of more specialized or unique cave bacteria. Future studies will repeat the study with additional types of media that have been shown to successfully grow cave bacteria; sampling in wild caves with areas of low visitation by humans; and will explore whether isolation of chemolithotrophic cave microorganisms is aected by incave incubation.AcknowledgementsWe thank the Science and Resource Management personnel at Mammoth Cave National Park and Carlsbad Cavern National Park for permission to conduct these studies, especially Kurt Helf, Bob Ward, and Rick Olson at MACA and Dale Pate and Paul Burger at CCNP. is work was done at MACA as part of RSM 106 and CCNP CAVE00035. We appreciate assistance with the eld work from Sandra DeCoste, Elizabeth Lavoie, Steve Jadzack, and Louise Hose. Mark Jensen from Bowling Green State University provided the culture media used at Mammoth Cave. We gratefully acknowledge funding provided by the T&E, Inc., for the Carlsbad portion of the study and from the Research Foundation of SUNY. We thank the Cave Research Foundation for the use of their eld facilities at both locations.ReferencesAmann MANN R.I., W. Ludwig UDWIG and K.-H. Schleifer C HLEIFER (1995) Phylogenetic identication and in situ detection of individual microbial cells without cultivation. M icrobiology RReviews 59, 143. ANGERT, E.R., D.E. NORTHUP, A.-L. REYSENBACH, A.S. PEEK, B.M. GOEBEL, and N.R. PACE (1998) Molecular phylogenetic analysis of a bacterial community in Sulphur River, Parker Cave, Kentucky. AA m erican Mineralogist 83, 1583. DICKSON, G.W. and P.W. KIRK JR (1976) Distribution of heterotrophic microorganisms in relation to detritivores in Virginia caves (with supplemental bibliography on cave mycology and microbiology). In: e distributional history of the biota of the S outhern AA ppalachians. Part IV. Algae and fungi. B.C. Parker and M.K. Roane (Eds.), University Press of Virginia, Charlotteville, VA, pp 205. GOUNOT, A.M. (1967) La microore des limons argileaux souterains: son activit productrice dans la biocnose cavernicole. AA n nales de Splologie 22, 23. JAMES, J.M. (1994) Microorganisms in Australian caves and their inuence on speleogenesis. In:

PAGE 35

Biological Sciences 1296 2009 ICS P roceedings 15th International Congress of Speleology Breakthroughs in karst geomicrobiology and redox geochemistry: abstracts and eld-trip guide for the symposium held February 16 through 19, 1994 C C o lorado Springs, CC olorado. Spec Pub 1. Sasowsky I.D. and M.V. Palmer (Eds.), Karst Waters Institute, Inc., Charles Town, W.V.. pp 31. KOCH, A.L. (1997) Microbial physiology and ecology of slow growth. Microbiology and Molecular Biology R R eviews 61, 305. REASONER, D.J. and E.E. GELDREICH (1985) A new medium for the enumeration and subculture of bacteria from potable water. AA p plied and E E nironmental Microbiology 49(1), 1-7. RUSTERHOLTZ, K.J. and L.M. Mallory (1994) Density, activity, and diversity of bacteria indigenous to a karstic aquifer. M icrobial EEcology 28, 79. VLASCEANU, L., S.M. SARBU, A.S. ENGEL, and B.K. KINKLE (2000) Acidic cave-wall biolms located in the Frasassi Gorge, Italy. GG eo microbiology Journal 17, 125.

PAGE 36

15th International Congress of Speleology Biological Sciences 1297 2009 ICS Proceedings In N Vertebrate ERTEBRATE colonization COLONIZATION and AND deposition DEPOSITION rates RATES of OF guano GUANO in IN a A man MAN made MADE bat BAT ca CA Ve E the THE Chiroptorium HIROPTORIUM Te E Xas AS USAKAt T HLEEN H H L L AvV OIE1 AND D D IANA E E N N ORt T HUp P2 1101 HH udson HH all, State UU niversity of NN ew York CC ollege at Plattsburgh, Plattsburgh, NN Y 12901 UU SA A lavoiekh@plattsburgh.edu 2MSC C03 2020, UU niversity of NN ew Mexico, AA lbuquerque, NN M 87131-0001, dnorthup@unm.edu A man-made bat cave, the Chiroptorium, was built on Selah: the Bamberger Ranch Preserve near Johnson City, Texas, USA, by Margaret and J. David Bamberger. e Chiroptorium was built in 1998 on the principle of if we build it, they will come. It took a few years, but the Chiroptorium was colonized by Tadarida brasiliensis bats in summer 2003. We began monitoring the bat guano in the winter of 2004-2005 to see when the full community of guano invertebrates (excluding mites) would develop in what amounts to a primary succession in a virgin environment. In Year 1, average guano depths in both domes was about 5.5 cm, and the invertebrates very limited in number and diversity, with none of the characteristic beetles. A pseudoscorpion, probably HH e sperochernes mirabilis was common in the guano and on walls from the beginning of our study. Spiders Spermophora senoculata Duges and Tidarren sisyphoides Walckenaer, some with egg cases, were found on the walls. By the winter of 2005, guano depth had roughly doubled to about 10.5 cm, and a diverse community of invertebrates was described, including the dermestid beetles Metoponium sp. and DD ermestes sp. In the third winter (2006), all structure of the guano deposits had been reduced to dust, probably by the action of a large population of beetles and aided by cattle. Several spiders, OO ecobius annulipes Lucas, were found on the walls. e Bambergers built it, and the bats and the invertebrates did come, and very quickly.1. Introductione Bambergers, J. David and Margaret, own Selah the Bamberger Ranch Preserve, in Johnson City, Texas, USA. Avid naturalists and conservationists, they wanted to add a bat colony to the ranch property and began planning in consultation with Bat Conservation International to build a man-made bat roost they described as the Chiroptorium: Chiroptfor bats and -torium from auditorium, a place to gather. e Chiorptorium was designed to house one million bats in two domes, a large outer dome 12.2 m D x 6.1 m H (40 D x 20 H) and a second, smaller inner dome that is 6.1 m D x 6 m H (20 D x 18 H), with a connecting tunnel, an entrance tunnel, and an observation area behind glass. Figure 1 shows the rebar stage of construction of the Chiroptorium with the entrance tunnel to the right and the small tunnel leading to the observation level at the top of the picture. e rebar was then sprayed with gunite, a liquid concrete. To maintain more bat-friendly temperatures, the structure was covered with tar and earth (BATS, 1997). e structure was completed in 1998, but did not attract signicant numbers of bats until the summer of 2003. Gary McCracken conrmed a maternity colony of Mexican freetailed bats (Tadarida brasiliensis) in 2003. Bat biologist omas Kunz estimated a population of 27,000 free-tailed Figure 1: e Chiroptorium under construction. Note person on top of the large outer dome for scale. Photo by James Smith and Brian Keeley.

PAGE 37

Biological Sciences 1298 2009 ICS P roceedings 15th International Congress of Speleology bats in summer 2006 (T. Kunz, personal communication). e plan is to ultimately use the guano as fertilizer at the Ranch. (For a panorama of the large outer dome, go to http://chriswjohnson.blogspot.com/2007/12/bambergerranch-chiroptorium-main.html) Our interest in the Chiroptorium is that is a virgin environment, never before colonized by animals and invertebrates. We wanted to follow the rate of guano deposition and the process of primary succession to determine how long it would take for a full guano invertebrate community to develop, especially Dermestid beetles, and to inventory the invertebrates. We did not include mites in our study, which can represent 95% of the invertebrates in guano. omas Kunz from Boston University and his students are following the changes in bat populations over the same period of time.2. Material and MethodsWe began our study in January of 2005, well aer the majority of bats had le for the winter. We set up an octagonal grid in the large dome, and a hexagonal grid in the second smaller dome. Both domes had wall to wall transects crossing through one central point and an inner and an outer ring 1/3 and 2/3 of the way from the center point to the walls. e points where the rings intersected the transects resulted in n=8 and n=6 sampling sites for the rings in the large and small domes, respectively. At each sampling site we measured guano depth using a meter stick. We did a visual census of invertebrates at each point for 10 cm around the intersection and below the surface. Each dome also contained a large bat box that was particularly heavily colonized by the bats. We measured the piles and made observations of invertebrates. We did a visual survey of the walls of the domes and tunnels, sampling as we encountered something of interest. Year 1 was in January of 2005, Year 2 in November 2005, and Year 3 in January of 2007. Samples of bat guano were taken for a molecular, cultureindependent analysis to detect changes in the guano from year to year (January 2005: CH050106-1 (Dome 2, big pile), 2 (Dome 1, big pile) and 3 (Dome 1, central room); November 2005: CH051112-1 (Dome 2, fresh guano), 2 (Dome 2, older guano), 3 (Dome 1, bat ring), 4 (Dome 1, big pile), 5 (Dome 1, outer ring); CH070115-3 (fresh guano), 4 (fresh guano). DNA was extracted and puried using the MoBio PowerSoil DNA Isolation Kit using the manufacturers protocol (MoBio, Carlsbad, CA). Extracted DNA was amplied with universal bacterial primers 46 forward (5GCYTAAYACATGCAAGTCG-3) and 1409 reverse (5-GTGACGGGCRGTGTGTRCAA-3)(Vesbach, personal communication). Amplicons were cleaned and puried using the Qiagen PCR cleanup kit (Qiagen, Germantown, Maryland) and were cloned using the TOPO TA Cloning kit (Invitrogen, Carlsbad, CA) and sent to Washington University Genome Sequencing Facility for sequencing with primers M13F and M13R. Once received, sequences were edited and contiged with Sequencher 4.8. (Gene Codes, Ann Arbor, Michigan). To check the orientation of our sequences and to convert from antisense to sense, OrientationChecker (www. cardi.ac.uk/biosi/research/bioso/) was used. Sequences were classied using the Ribosomal Database Program Classier soware (http://rdp.cme.msu.edu/). Denaturing Gradient Gel Electrophoresis (DGGE) was carried out on guano samples. e 16S rRNA gene was amplied with universal primers 338 forward with GC clamp(5CGC3 GCCGCGC4 GCGC3GTC3GCCGC5GC3TCCTACG3 AGGCAGCAG-3; and 907 reverse (5-CCGTCAATTC CT3 RAGT3-3). DGGE was conducted using a DGGE2001 System (C.B.S.*Scientic Company, Inc.). PCR products were run on 6% (w/v) acrylamide gels with a denaturing gradient of 30%.3. ResultsFigure 2 shows the changes in guano depths from Year 1 to Year 2 at the sampling sites in each ring. Figure 2A is from Figure 2: Changes in guano depth om Year 1 to Year 2 by sampling ring in A, the large dome, and B, the small dome.

PAGE 38

15th International Congress of Speleology Biological Sciences 1299 2009 ICS Proceedings the large dome, and Figure 2B is from the small dome. Note the dierence in scale. e outer ring in the small dome included the large guano pile under the bat box at site 3. e large guano pile under the bat boxes increased in extent from Year 1 to Year 2 by: Large Dome. Year 1: 31 cm H x 160 cm L x 101.6 cm W to Year 2: 435 cm H x 167.6 L x 127 cm W Small Dome. Year 1: 64 cm H x 182.9 cm L x 152.4 cm W to Year 2: 530 cm D x 323 cm L x 198.1 cm W. Invertebrate numbers from a representative transect from wall to wall in the Large Dome are shown in Table 1 for Year 1 and Year 2. A DGGE analysis of all nine guano samples (Fig. 3) reveals much similarity among samples within a given sampling time, but major dierences between sampling times. Samples from January 2005 and November 2005 showed many bands, indicating the presence of many species, without dominance by any one band/species. e samples from November 2005 appear to be more diverse than those from January 2005, possibly due to sampling earlier in the winter. e samples from January 2007 are dominated by one major band/species and lack diversity. e samples from January 2007 are dominated by one major band/species and did not have as much diversity as formerly seen. e one dominant sequence is LL act obacillus lactis, in the Order LL act obacillales. Two samples form the November 2005 collection from Dome 1 were sequenced and closest relatives were identied. e two samples showed overlap within the Firmicutes phylum, in particular within the Staphylococcus and Sporosarcina genera. e bat ring in Dome 1 contained A A l phaproteobacteria and GG ammaproteobacteria, which were not found in the big pile sample from Dome 1.4. DiscussionIn our Year 1, the winter of 2004, average guano depths in both domes was about 5.5 cm, and the invertebrates Sample location Year 1 Year 2 Outer 0-1(dead lacewing)4 dermestids Inner 0-2Nothing1 ea, 1 dermestid Center1 ant Nothing Inner 5-2Nothing2 dermestids Outer 5-1Nothing1 ant, 1 pseudoscorpion Total1 9Table 1: Inertebrate numbers om a representative transect om wall to wall in the Large Dome. Figure 3: DGGE analysis of all nine guano samples: January 2005, Noember 2005 and January 2007. Each band represents a separate species. Sample Proteo Bartonella ProteoE E n terobacteriaceae FirmicutesE E n terococcus WeissellaA A t opostipesL L actobacillales Firmicutes Staphylcoccus Sporosarcina Bacillaceae 1 Bacilliaceae 2 Bacillales Firmicutes Unclass Bacilli Dome 1 Bat Ring 051112-3 1 2 3 + 1 + 0 + 14 + 4 + 1 + 15 + 5 1 Dome 1 Big Pile 051112-4 0 0 0 + 0 + 24 + 03 + 2 + 0 + 3 + 01Table 2: Taxonomic groups found in clone libraries om two of the guano samples examined with DGGE. Closest relatives by genus or family are given under each phyla (in bold) and the number of clones within each of these classications is given by sample.

PAGE 39

Biological Sciences 1300 2009 ICS P roceedings 15th International Congress of Speleology very limited in number and diversity, with none of the characteristic dermestid beetles. A pseudoscorpion, probably H H e sperochernes mirabilis, was common in the guano and on walls from the beginning of our study. Spiders Spermophora senoculata Duges and Tidarren sisyphoides Walckenaer, some with egg cases, were found on the walls. Mud dauber wasps and a phoebe nest were already well-established in the entrance tunnel by 2005. By Year 2, guano depth had roughly doubled to about 10.5 cm, and a diverse community of invertebrates was described, including the beetles Metoponium sp. and DD ermestes sp. (Fig. 4). In Year 3 (2006), we were surprised to nd all structure of the guano deposits and of individual droppings had been reduced to dust, probably by the action of a large population of beetles (Kunz, personal communication) and aided by cattle. Several spiders, OO ecobius annulipes Lucas, were found on the walls. Sampling was complicated by a major ice storm that made staying in the Hill Country unfeasible. Major changes in the guano bacterial community composition occurred over the three sampling periods, with the last sampling period showing a change from a diverse community, with relatively even abundances, to a community dominated by one organism. e reworking of the guano by dermestid beetles and cows may have contributed to the major change in the bacterial makeup of the guano in 2007. A possible complication of our study is the spreading of about 100 pounds of guano from Bracken Bat Cave prior to 2002 (J. Bamberger, personal communication). e guano was placed in the Chiroptorium in the hopes that the smell would attack bats to the structure. We doubt that this action had any signicant eects on the succession, since any invertebrates transferred with the guano would have had to go through several lean years without inputs of guano, and that a guano community had not fully developed by the time we began our study, despite an abundance of raw material already in place. However, we are conducting a comparison of the Chirpotorium invertebrate community with Bracken Bat Cave, and we plan to investigate other nearby guano deposits.5. ConclusionsA robust guano invertebrate community had developed within three years of reliable colonization of the Chiroptorium by bats. e Bambergers built it, and the bats and the invertebrates did come, and very quickly. e study is still on-going, and comparison of the invertebrate guano community in the Chiroptorium with other bat guano caves will provide additional information.AcknowledgementsWe thank the Bambergers for their interest in our project and all things natural, and for building the Chiorptorium. Selah Educational Programs Manager Colleen Gardner is our regular contact, and has been a gracious host. We thank James Smith and Brian Keeley for Chiroptorium construction pictures provided by the Bambergers. Tom Kunz gave us insights into both bat and Dermestid beetle behavior. Thanks to Jim Kennedy and BCI for providing guano from Bracken and Dr. Sandy Brantley for identification of the spiders. Jerry Lewis made a presumptive identification of the pseudoscorpion. Monica Moya and Amaka Nwagbologu were very helpful with sequence editing and Dr. Armand Dichosa and Jenny Hathaway helpfully provided the DGGE analysis of the guano samples. We thank Jim Lavoie and Kenneth Ingham for field assistance and photography. Funding was provided through Kenneth Ingham Consulting, L.L.C. and the Research Foundation of SUNY. Figure 4: Dermestid beetle on the wall of the Chiroptorium (Photo by Kenneth Ingham).

PAGE 40

15th International Congress of Speleology Biological Sciences 1301 2009 ICS Proceedings ReferencesBATS (How to Build a Cave: A Bold Experiment in Articial Habitat). Winter 1997 http://www. batcon.org/batsmag/batindex.html FERREIRA, R.L., and R.P. MARTINS (1999) Trophic structure and natural history of bat guano invertebrate communities, with special reference to Brazilian caves. T ropical ZZ oology 12: 231.

PAGE 41

Biological Sciences 1302 2009 ICS P roceedings 15th International Congress of Speleology Effects FFECTS of OF missing MISSING legs LEGS on ON distribution DISTRIBUTION and AND Jumping UMPING beha BEHA Vior IOR in IN the THE ca CA Ve E cricket CRICKET Ha A DENo O EcusCUS subt SUBT Erra RRA NEus USKAt T HLEEN H H L L AvV OIE1, MOHAmm MM ED C C HANDOO1, SUGANt T HI THIRUNAvV Uk K ARAs S U1, Uts UTS AvV PANDEy Y1,, E E LIZAb B Et T H L L AvV OIE1, AND KURt T H H ELf F2.1State UU niversity of NN ew York CC ollege at Plattsburgh, Plattsburgh, NN Y 12901 (lavoiekh@plattsburgh.edu)2CC umberland-Piedmont NN etwork, NN ational Park Service, Mammoth CC ave, KY 42259 (kurt_helf@nps.go) e cave cricket (HH ad enoecus subterraneus) is a keystone species in maintaining biological diversity in cave communities in Mammoth Cave National Park through inputs of xed organic material in the form of guano, eggs, and carcasses. Crickets must leave the cave to forage on nights when conditions are favorable. Leaving the refuge of the cave puts them at considerable risk for predation. Invertebrates are preyed upon by a wide range of animals and have developed defenses, including autotomy, or voluntary loss of a limb. We hypothesize that missing legs are a sign of predation pressure on the crickets, and may be dierent in dierent environments. We also hypothesize that missing legs reduce the ability of an adult cricket to escape predation. We used a visual census to record the frequency of missing legs by gender among adult cave crickets at eight dierent entrance locations in Mammoth Cave National Park. We expected males to be missing legs more frequently than females because they must leave the refuge of the cave to forage more frequently than females, but males and females were missing legs in equal frequency. Cave crickets were missing hind legs with greater frequency than other limbs, probably because crickets attempt to jump away from predators, making the larger hind limb closest to the predator. e frequency of crickets with missing limbs varied among cave entrances from a low of 6.6% to a high of nearly 40%: numbers are variable from year to year. In Frozen Niagara, which consistently had a high proportion of crickets missing legs, the percentage was highest in crickets roosting closer to the entrance (30.8%) than deeper (18.7%) into the cave. We collected crickets naturally missing various legs and combinations of missing legs, and measured the distances jumped. Crickets missing one hind leg jumped as far as crickets with both hind legs. Missing one or two of the front and middle legs had no impact on jumping behavior. Crickets missing both hind legs were completely unable to jump. Crickets missing hind legs had lower amounts of food in their crops than crickets with all their legs. We conclude that missing legs has limited negative eects on crickets being able to leave the cave to forage, except in the extreme case of crickets missing both hind legs.1. IntroductionHH adenoecus subterraneus cave crickets are habitual trogloxenes, and are oen abundant around cave entrances (Lavoie AVOIE et al., 2007). HH s ubterraneus is a keystone species, leaving the cave to forage on nights when conditions of temperature and humidity are favorable, returning to the cave for a daytime refuge (Lavoie A VOIE et al. 2007) e crickets maintain communities of specialized egg predators (Poulson OULSON 2002), a diverse guano community under roosts (Poulson OULSON 1992; Poulson OULSON et. al, 1995), and provide dispersed inputs of energy for troglobites. Cave crickets are vulnerable to a wide range of predators. Inside the cave, HH s ubterraneus are preyed upon by spiders, particularly Meta americana, and cave salamanders, EE u rycea lucifuga. Outside the cave crickets are frequently eaten by Peromyscus lucopus, white footed deer mice. Viele I ELE and AND Studier TUDIER (1990) showed that deer mice preferentially overlap feeding ranges around a cave entrance in Mammoth Cave National Park. Helf E LF (2003) found more than four times the number of P. leucopus within 50 m of a cave entrance compared to a control area without a cave entrance. e three sets of paired legs of insects, including crickets, are attached to the three thoracic segments e attachment point at the coxa is very weak, and legs can be easily lost. In many insects, voluntary loss of legs (autotomy) is a strategy to reduce predation (Bateman A TEMAN and AND Fleming LEMING 2005) Cave crickets primary response to a threat is to jump. Insects walk using a characteristic pattern that is described as the alternating triangle gait. e front and rear legs on one side of the body move together with the middle leg on the other side of the body, creating a supportive triangle. e next step switches sides, maintaining the triangle. e movement is highly stereotyped, and controlled by neurons.

PAGE 42

15th International Congress of Speleology Biological Sciences 1303 2009 ICS Proceedings In some insects, the need for coordination decreases at faster speeds. Insects typically adapt rapidly to the loss of one or more legs. Field crickets (GG r yllus bimaculatus) with one missing hind limb had a reduced escape speed and decreased ability to jump (Fleming LEMING and AND Bateman ATEMAN 2007) ey were also slower, moved shorter distances, and used more energy than those with no hind limbs missing (Fleming L EMING and AND Bateman ATEMAN 2007). e same pattern was found with house crickets (AA c heta domestica), and the reduced escape abilities increased their predation risks by lizards and mice (Bateman ATEMAN and AND Fleming LEMING 2006). We hypothesized that male cave crickets would be missing limbs more frequently than females since males must leave the cave to forage more frequently than females (Studier T UDIER et al., 1986). We expected the hind leg to be missing the most oen since it would be in closet proximity to a threat as the cricket jumps away and it is the largest leg, making it a better target. Variability in the proportion of crickets missing legs might be higher in caves with more human impact (visitors) or in caves with sink vs. source populations (Helf ELF et al., 1995). We expected to nd that missing one or both hind legs would reduce a crickets ability to jump in terms of distance per jump and total distance jumped, and that crickets missing legs would not be as eective at foraging as evidenced by lower amounts of food in the crop than crickets with all legs.2. Methods and Materials CC ensus for missing legs. e study was done in eight dierent sites (entrances and caves) within Mammoth Cave National Park, KY, USA. Cave crickets were visually censused from the entrance until population density dropped o, which ranged from 20 m up to 150 m into the cave. We did a visual census of adult crickets, noting gender and which leg or legs were missing on which side of the body. e study is limited to adult crickets with a Hind Femur Length of 20 mm (Studier TUDIER et al., 1986). Adults have highly sclerotized (brown) legs and ovipositors and do not molt so they cannot regenerate a lost leg. Leg loss was analyzed for dierence by gender and limb location (front, middle, and hind limb) by Univariate Analysis of Variance (ANOVA) and a Post-hoc Bonferroni test, respectively. Jumping experiment. We used an area of the tourist trail in Frozen Niagara Cave 30 m in from the entrance that is wide, relatively at, and covered in concrete, with electricity nearby. An electronic balance recording to 4 decimal places was used. Crickets were individually selected and captured by hand so that we would have a good mixture of gender, crickets with all legs, and crickets missing dierent combinations of legs. A total of 44 crickets were tested. Each cricket was placed into a tared plastic 35mm lm container and weighted to four decimal places. We recorded their Hind Femur Length, gender, weight, and which leg(s), if any, were missing. Crickets were placed in the center of the open area and released. As crickets jumped, a coin was placed to record the starting and stopping points for each jump or distance walked. Crickets not compliant with jumping were given a gentle poke. Aer the cricket stopped jumping or le the path area, the distance between each jump was recorded. To make the results more comparable, only the rst ve jumps made by a cricket were used to determine total distance jumped and average distance per jump. e mass of food in the crop was estimated using a regression of Hind Femur Length (HFL) vs. Crop Free Live Weight from dissected crickets (Studier T UDIER et al., 1986). e predictive equation allows us to estimate the amount of material in the crops with a condence of 90% without having to kill the cricket. For crickets missing both hind legs, we used an average HFL of 22 mm for the males and 23 mm for the females. A typical hind leg weighed 0.0382g and we adjusted the total weight by that amount for each missing hind leg to allow us to use the regression.3. Results e number of crickets at each location is shown in Table 1 by gender, with total number of adult crickets observed, the total not missing legs, and the total missing legs. Which leg was missing (Hind, Middle, Front), and the percentage of adults missing legs in the total number censused is also shown. Some individual crickets were missing more than one leg, but the proportion was very low and ranged from 1% to 2.7% in 2008. Multiples legs lost are included in the total reported in Table 1. A Post-hoc Bonferroni test for position of missing leg, using multiple comparisons showed that the Hind leg was missing signicantly more than the other legs. e dependent variable was the number of crickets missing leg at each position. Figure 1 compares the proportion of the population missing legs in May 2007 compared to May 2008, for four caves and entrances in Mammoth Cave National Park. As we censused in 2008, the students noted that there was a higher proportion of broken crickets closer to the entrance than deeper into the cave at the Frozen Niagara entrance. e proportion of crickets missing legs in Frozen Niagara was greater in the rst 10-30 m into the cave (Males 31.9%,

PAGE 43

Biological Sciences 1304 2009 ICS P roceedings 15th International Congress of Speleology Figure 1. e percent of the censused population of adult crickets missing legs in 2007 and 2008 om four entrance and caves in Mammoth Cave National Park. Location (cave or entrance) GenderTotal Total Intact Total Missing Legs Missing Hind Leg Missing Middle Leg Missing Front Leg Percent missing legs(all) New DiscoveryM1139122 171 422.6 F12610422 172 3 Carmichael M59545 50 06.6 F87834 22 0 Frozen NiagaraM16814622 181 217.0 F21818434 301 2 Floyd Collins Crystal Cave M25196 60 023.6 F45367 60 1 Sophys AvenueM30255 11 320 F12102 20 0 Violet City M12111 10 07.4 F17161 10 0 Great Onyx CaveM45423 30 08.2 F47434 41 1 Little Beauty CaveM34277 70 2 38.9 F412714 142 3 TOTALS 1077918159159112117.3%Table 1. Number of crickets at each location and their characteristics.

PAGE 44

15th International Congress of Speleology Biological Sciences 1305 2009 ICS Proceedings Females 30.4%) compared to crickets deeper into the cave 60-80 m (Males 14.6%, Females 20.4%). Jumping experiments. e data for male and female crickets with all legs, missing one hind leg, and missing both hind legs is shown in Table 2 by total distance jumped in the rst ve jumps, for average distance jumped in the rst ve jumps, and in Figure 2 for the calculated crop weight. e data are graphed by individual. 4. Discussion ere is no dierence in the proportion of males and females missing legs (ANOVA F=0.817). Hind legs were autotomized signicantly more oen than other legs (F=34.069, p>0.001). e proportion of crickets missing legs in 2007 varied from a low of 6.6% in Carmichael to a high of 38.9% in Little Beauty Cave (Table 1). e proportion of crickets missing legs from year to year showed some variation (Fig. 1). Sites vary by amount of human impact and source vs. sink populations, but we have found no consistent pattern to explain the variability. Missing one hind leg had no eect of the total distance or the average jump length of a cricket. Crickets missing both hind femurs and were completely unable to jump. We have some indications (Fig. 2) that crickets missing one or both hind legs are less eective at foraging compared to crickets with both hind legs. Crickets missing hind legs were more likely to be found closer to an entrance, suggesting they did not want to move as far as intact crickets with all their legs, or had to leave the cave more frequently to forage. Crickets missing legs may not be as eective at foraging as suggested by a lower amount of food in the crops compared to intact crickets. A variable is that we do not know is when the cricket actually lost the leg. Future studies include repeating the census of crickets missing legs from all of the study sites. We will collect and weigh crickets will all legs and crickets with missing hind legs from near and far populations. e eect of missing legs on cricket endurance (see Fleming L EMING and AND Bates ATES 2007) will be evaluated. We have also noted anecdotally that crickets lose their legs more in winter when we handle them, which is another aspect to study.5. ConclusionsCave crickets (HH adenoecus subterraneus) show limited eects of missing one hind leg on escape behaviors as evidenced by the oen high number of individual missing a leg or legs in many locations. Intact crickets and crickets missing one hind leg showed nearly identical jumping behavior in terms of average jump length and total distance jumped in the rst ve jumps. Crickets Crickets by gender and number of missing hind legs n Average jump distance from rst ve jumps Total distance jumped in ve jumps Males, None 10 21.7 cm 108.4 cm Females, None12 18.1 cm 83.2 cm Males, One 6 18.0 cm 90.1 cm Females, One7 19.61 cm 98.1 cm Males, Two 2 0 0 Females, Two2 0 0Table 2. Cave crickets by gender and number of missing hind legs: none, one, or two. e average jump length and total distance hopped is om the rst ve jumps. Figure 2. Calculated drop weight in adult crickets om the Frozen Niagara entrance of Mammoth Cave National Park, by gender and missing no hind legs, one hind leg, and two hind legs. Circles highlight crickets missing both hind legs.

PAGE 45

Biological Sciences 1306 2009 ICS P roceedings 15th International Congress of Speleology missing both hind legs were not able to jump at all. Losing one or both hind legs may reduce the foraging eectiveness of crickets.AcknowledgementsWe thank Science and Resource Management personnel at MACA for providing access to the caves, and CRF for use of their facilities. We thank Victoria Johnson for assistance with the eld work.ReferencesBATEMAN, P.W and P.A. FLEMING (2006) Increased susceptibility to predation for automized house crickets (AA cheta domestica). EE thology 112, 670. Delcomyn, F. (1997) Insect walking. EE n cyclopedia of N N euroscience, 2nd edition (CD-ROM), G. Adelman and B. Smith (eds.). Amsterdam: Elsevier. FLEMING, P.A. and P.W. BATEMAN (2007) Just drop it and run: the eect of limb autotomy on running distance and locomotion energetics of eld crickets (GG ryllus bimaculatus). Journal of EE xperimental Biology 210, 1446. HELF, K.L., T.L. POULSON, and K.H. LAVOIE. (1995) Protection and management of the cave cricket (HH adenoecus subterraneus) at Mammoth Cave National Park. Proceedings of the National Cave Management Symposium, Spring Mill, IN, pp 155. LAVOIE, K.H. ,K.L. HELF and T.L. POULSON (2007) e biology and ecology of North American cave crickets. Journal of CC ave and Karst Studies 69(1), 114. POULSON, T.L., K. LAVOIE, and K. HELF. 1995. Longterm eects of weather on the cricket (HH ad enoecus subterraneus, Orthoptera, Rhaphidophoridae) guano communities in Mammoth Cave National Park. A A m erican Midland NN aturalist 134, 226. STUDIER, E.H., LAVOIE, K.H., W.D. WARES II, and J. A.-M. LINN. (1986). Bioenergetics of the cave cricket, HH adenoecus subterraneus. CC omparative Biochemistry and Physiology AA 84, 431. VIELE, D.P. and E.H. STUDIER (1990) Use of a localized food source by Permyscus leucopus, determined with am hexagonal grid: Bu lletin of the NN ational Speleological Society 52(1), 52.

PAGE 46

15th International Congress of Speleology Biological Sciences 1307 2009 ICS Proceedings Species PECIES limits LIMITS phylogenetics PHYLOGENETICS and AND conser CONSER Vation ATION of OF N N EoO LEPto TO NEta TA spiders SPIDERS in IN Te E Xas AS ca CA Ves ESJOEL OEL LED LED FORD ORD1, PIERRE IERRE PA A QUIN UIN2, CHARLE CHARLE S GRI GRI SWOLD OLD3 1UU niversity of CC alifornia, Berkeley, DD ivision of II nsect Biology, 137 Mulford HH all, Berkeley, CACA 94720 UU SA A2SWCACA EE nironmental CC onsultants, 4407 Monterey OOaks Bld., Bldg. 1, Suite 110, AA ustin, TX, 78749 UU SA A3AA rachnology LL ab, EE ntomology DDepartment, CC alifornia AA cademy of Sciences, 55 Music CC oncourse DDrive, S an Francisco, CACA 94118 UU SA A Abstract A principal theme in biogeographic studies of troglobites is understanding the relative roles that vicariance and dispersal have played in shaping their geographic distributions. Classical studies oen show troglobites as singlecave endemics, but recent work suggests that some species have wider ranges than expected and are best explained by dispersal. ese ndings not only have implications for understanding speciation in cave organisms, but are also central to developing accurate taxonomies upon which conservation decisions are ultimately based. Here we present a phylogeny of NN eo leptoneta spiders in Texas based on molecular and morphological data, including specimens from published records and several new cave localities. Preliminary results show that while most species are narrowly distributed, some troglobites (including the endangered NN eo leptoneta microps and NN myopica) occur more broadly than previously believed, and may not even be restricted to caves. ese data suggest that both vicariance and dispersal have played important roles in shaping the distribution of NN eo leptoneta and this highlights the need for a comprehensive revision of the genus. A distribution map for NN eo leptoneta is also provided, indicating areas in need of additional sampling or phylogenetic uncertainty.

PAGE 47

Biological Sciences 1308 2009 ICS P roceedings 15th International Congress of Speleology UNTANGLING THE WEBS IN CALIFORNIAS CAV V ES: THE BIOGEOGRAPHY AND SYSTEMATICS OF THE CAV V E SPIDER GENUS USOFIL L AJOEL OEL LED LED FORD ORD1, CHARLE CHARLE S GRI GRI SWOLD OLD2, RO RO SE E MAR AR Y GILLE GILLE SPIE IE3 1, 3 UU niversity of CC alifornia, Berkeley, DD ivision of II nsect Biology, 137 Mulford HH all, Berkeley, CACA 94720 UU SA A2AA rachnology LL ab, EE ntomology DDepartment, CC alifornia AA cademy of Sciences, 55 Music CC oncourse DDrive, San Francisco, CACA 94118 UU SA A Abstract Despite Californias diverse and endemic biota, little attention has been directed towards understanding the composition and relationships of its caveadapted fauna. Basic taxonomic knowledge is lacking in most cases, which hampers research eorts and prevents eective conservation and management planning. is project aims to discover and describe patterns of diversity in the genus U U so la, a group of poorly known spiders in Western North America. e goal of the project is to develop a functional taxonomy for UU so la that will be used to address broader biological questions, particularly those focused on conservation, biogeography, and cave adaptation in Western North America.

PAGE 48

15th International Congress of Speleology Biological Sciences 1309 2009 ICS Proceedings ZOOGEOGRAPHY AND EV V OLUTION OF THE SUBTERRANEAN ASELLID ISOPODS OF NORTH AMERICAJULIAN ULIAN J. LE LE WI I SL Lewis & AA ssociates, LLCLLC : CC ave, Karst & GG roundwater Biological CC onsulting 17903 State RRoad 60, Borden, ININ 47106 UU SA A Over a century elapsed between the discovery of the rst subterranean asellid isopod in North America, C C aecid otea stygia in 1871, and the realization of the true diversity of the fauna. Initially known only from caves of the large karst areas of the eastern United States, knowledge of subterranean asellids in other areas and habitats was slow to accrue. Since 1973 stygobiont asellids have been recognized in a total of 9 genera in North America: CC aecid otea, LL irceus, LL irceolus, Salmasellus, RRemasellus, CC alasellus, CC olumbasellus and two new genera from Oregon and California currently in press. Of these only C C aecid otea and LL irceus have both epigean and hypogean species, the rest possess only subterranean species. From a zoogeographic standpoint CC a lasellus, CC olumbasellus, Salmasellus and the two new genera are restricted to an area west of the Rocky Mountains in the U.S. and Canada. is assemblage of western North American genera is related to an Asian AA se llus assemblage of which one epigean species extends into Alaska. From an evolutionary viewpoint, although morphological similarities exist between the Asian and North American faunas, the subterranean asellids in western North America represent at least two lineages that appear to be only remotely related to AA se llus. No vestige of eyes nor pigmentation remain in any of the species of CC alasellus, CC olumbasellus, Salmasellus, or the two new genera. is combined with the absence of epigean relatives suggests that this group of asellids invaded subterranean habitats long ago. In contrast, the evolutionary history of some of the subterranean CC aecid otea is clearer, with examples of species with varying degrees of morphological adaptation to groundwaters. e best example is that of the H H o bbsi species group. Among the members of this group are species ranging from eyed and pigmented, some with vestigial eyes and traces of pigmentation, to those that are completely eyeless and unpigmented. Many of the species with intermediate degrees of morphological adapation are those that invaded shallow groundwaters, inhabiting the interstices of unconsolidated glacial sediments. ere are species of CC aecidotea on both sides of the Rocky Mountains, but none in the mountains, suggesting that these isopods have been present in North America since prior to the upli of the range. e creation of the great plains east of the mountains apparently extirpated surface populations of asellids, with only a few subterranean species remaining. Unlike C C aecid otea, in which the majority of species are subterranean, only two stygobiont species of LL ir ceus are known, representing localized groundwater invasion. RR emasellus parvus in Florida is a bizarre example of an adaptive shi, in which the isopods have evolved the ability to swim in the water column, rather than crawling across the substrate as in most other asellids. Similarly, L L ir ceolus smithii possesses adaptations for its existence in the deep groundwaters of the Edwards Aquifer in Texas.1. IntroductionWhen Alpheus Packard collected C C aecidotea stygia from Richardsons Spring in Mammoth Cave, Kentucky, it was the rst subterranean isopod discovered in North America (Packard 1871). Since that time dozens of species of the isopod Family Asellidae have been described among ten genera summarized by Lewis (in press 2009a). With one exception (A A s ellus, represented by the epigean species A A alaskensis ), these genera are endemic to North America. e genus C C aecidotea (Packard 1871) is the dominant asellid group in North America, with over 84 species, of which about 70 are obligate groundwater inhabitants, followed by L L i rceus with 15 species (2 stygobites)(Henry et al. 1986). e other seven genera possess only obligate subterranean species. R R e masellus (monotypic) and L L i rceolus (7 species) occur in the eastern U.S. and are apparent derivatives of C C a ecidotea (Lewis & Bowman 1996). Five genera are restricted in distribution to the far western parts of the U.S. and Canada, of which three are currently described (Lewis 2001, Lewis et al 2003): Salmasellus (2 species), C C a lasellus (2 described, 1 undescribed species) and C C o lumbasellus (monotypic). Two other genera that are referred to here as new genus 1 (from Malheur Cave, Oregon) and new genus 2 (formerly C C a ecidotea sequoia Bowman 1975) (Lewis, in press 2009b).

PAGE 49

Biological Sciences 1310 2009 ICS P roceedings 15th International Congress of Speleology 2. Zoogeography and EvolutionIn analyzing the evolution of the subterranean asellids not surprisingly the fossil record is of little help. e fossil record of various kinds of isopods dates back to the Paleozoic (Wilson 1999), which arms the antiquity of the group. Little work has proceeded with genetic analysis of the asellids in North America yet and given the extreme rarity of some of the genera and species (e.g., CC o lumbasellus is known from one specimen), the prospects for additional knowledge soon is unlikely. us, for the moment we remain dependent on the implications from morphological, zoogeographic and paleogeographic evidence. e holarctic distribution of the asellids (Henry & Magniez 1970) suggests that the group has been in freshwater for a long time, predating the Mesozoic separation of North American and Eurasia. e species of CC aecid otea, and to a lesser extent LL ir ceus, are nearly continuous in distribution from the east coast through the central U.S. e asellids are absent in the Rocky Mountains except for Salmasellus steganothrix (known from an area encompassing Alberta to Montana), with several mostly rare and poorly known species occurring along the Pacic coastal states. Also present in far western North America, CC aecid otea is represented by CC tomalensis and CC occidentalis, although the western taxa are not very similar to those in the east. is attests to the dispersal of this group across temperate North America at some ancient time, prior to the creation of the large swath of western North America in which the asellids are now absent. e paleogeography of North America provides insight into this distribution. During the Cretaceous, marine transgression separated North America into eastern and western landmasses separated by the Western Interior Seaway that inundated what was later to become the Great Plains region (Fig. 1). e marine waters subsequently retreated, leaving much of present-day North America above sea level. e Laramide orogeny occurred creating the Rocky Mountains and with them, the relatively arid great plains region to the east that served to further isolate the CC aecid otea into eastern and western components (Lewis et al. 2006). Concerning the group of asellid genera endemic to western North America, CC a lasellus, CC olumbasaellus and new genus 1 are related to the AA se llus pattern genera found in Asia (Henry & Magniez 1995, Magniez 1996). Considering that the 4 species assigned to these three genera form a geographic cluster in the U.S. Pacic coastal region, it Figure 1. Map of North America as it appeared during the Cretaceous, when much of the Great Plains and southern U.S. were under seas. e Western Interior and Hudson seaways divided North American into three disjunct landmasses. Approximate present day asellid distributions are superimposed (shown as darker stippled areas) onto the Cretaceous landmasses.

PAGE 50

15th International Congress of Speleology Biological Sciences 1311 2009 ICS Proceedings seems likely that they form a sister group to the MesoasellusPhreatoasellus-UU enasellus-NN ipponasellus-Siberasellus assemblage of Henry & Magniez (1995). e close similarity of the male second pleopods and geographic proximity of the species of CC a lasellus, CC olumbasaellus acheron and new genus 1 suggests a common ancestor. Since new genus 1 displays morphological characteristics on both sides of Henry & Magniez (1995) phylogeny, i.e.,mandibular palp reduced (e.g., NN i pponasellus, Sibirasellus) and rst maxilla inner lobe with 4 setae (e.g., AA se llus), one or more of these characters in the new genus 1 asellid clouds the picture with apparent convergence. Other pieces of the puzzle are the genera Salmasellus and new genus 2, which appear closely related as evidenced by synapomorphies listed by Lewis (in press 2009b). Although it was tempting to assign CC aecid otea sequoiae to the genus Salmasellus, among other things the large setae present on the male pleopod 2 endopod in Salmasellus remained unique. It was preferred to erect a new genus rather than broaden the diagnosis of Salmasellus. e two genera certainly have zoogeographic anities with the AA se llus pattern genera, but the morphological resemblance is not as strong as in the other North American genera discussed above. Considering the North American and Asian distribution of the AA sellus pattern genera, freshwater habitats connecting the areas in which these genera now exist around the Pacic Rim must have been present across Beringia. Paleontological evidence suggests that this bridge dates to the mid-Cretaceous (Fiorillo et al 2008; Weishampel et al. 2004). e timing of groundwater invasions in western North American asellids remains obscure. All of the western asellids are either strictly epigean or completely subterranean no morphological intermediates between eyed/pigmented and eyeless/unpigmented species have been discovered. It would seem that the subterranean species invaded groundwaters long ago and their epigean ancestors are long gone. is is not the case in the eastern CC aecid otea, where many examples of evolutionary intermediates are present. Although some species group lineages are completely of subterranean facies, others have species of varying degrees of morphological adaptation to groundwaters. e members of the CC an nula Group are completely eyeless and unpigmented, with bizarre modications of the male second pleopod that have suggested their placement in the genus Pseudobaicalasellus (Henry & Magniez 1970, Lewis 1980). Most of the species of the Stygia Group are also completely stygomorphic, with the exception of CC aecid otea beattyi, which possesses vestiges of pigmentation (Lewis & Bowman 1981). Among the members of the HH o bbsi Group are found species with varying degrees of adaptations to subterranean habitats (Lewis 1982; Lewis et al. 2006). e epigean species C C aecid otea kenki and CC brevicauda inhabit surface streams, springs and cave streams. Both are eyed and pigmented (Fig. 2), although some populations become depigmented in caves. CC aecidotea kendeighi and CC spatulata are inhabitants of saturated interstices of till plains or other unconsolidated deposits in the midwestern U.S. ey have the vermiform appearance typical of such subterranean isopods, but retain vestigial eyes and light magenta pigmentation. Other phreatobites, like CC phreatica or CC lesliei, or many of the stygobites like CC packardi are completely eyeless and lack any pigmentation. Some of the subterranean asellids inhabiting saturated interstices (rather than caves) have penetrated signicant Figure 2. e possession of eyes and pigmentation varies widely in species of the Hobbsi Group in the genus Caecidotea. (A) Caecidotea kenki, epigean inhabitant of streams and springs, eyes and pigment are prominent, (B) Caecidotea kendeighi and (C) Caecidotea spatulata, phreatobites with varying degrees of pigmentation and vestigial eyes; (D) Caecidotea phreatica, eyeless and unpigmented. [Drawing of C. kenki by Carolyn Gast om Bowman 1967 with permission.]

PAGE 51

Biological Sciences 1312 2009 ICS P roceedings 15th International Congress of Speleology distances into areas covered by the Pleistocene glaciers. One of the outstanding questions of subterranean asellid zoogeography is how these species arrived in the glaciated areas. Two schools of thought have been advanced. e rst entails the animals remaining in subglacial refugia like the aquifers in deep unconsolidated deposits in lled stream basins (Koenemann & Holsinger 2001). e other theory is that the crustaceans followed the retreating glaciers northward. Koenemann & Holsingers map indeed shows a correlation with deep deposits and collection sites for the phreatobitic amphipod Bactrurus mucronatus However, recent C C aecidotea collections have been made in areas where sediments are shallowest, like a collection from southeastern Bartholomew County, Indiana. Within a given surface drainage basin, C C a ecidotea beattyi is typically found in a headwater area where sediment deposits are shallowist (Lewis 2008). I am not satised that a good solution to the question of how phreatobitic crustaceans arrived in the glaciated regions has been found yet. As Lewis & Bowman (1981) noted, none of the theories are necessarily mutually exclusive. e genera LL ir ceolus and RR emasellus represent asellids in which adaptive shis to have produced unusual morphological characteristics. Both genera are probably oshoots of CC aecid otea. ey occur in geographic areas inundated by the Cretaceous seas and represent subsequent dispersions from eastern North America into the coastal plain areas. RR emasellus has anatomical modications to accommodate swimming, with legs with elongate setae rows that transform the typical asellid pereopods into oar-like structures (Fig. 3) (Bowman and Sket 1985). Other North American asellids typically crawl across the substrate, but R R emasellus is also capable of swimming weakly in the water column, an incredible behavioral adaptation. e genus LL ir ceolus is dened by a mosaic of similar morphological characteristics of males, particularly the simple rst pereopod, elongate and sparsely setose rst and second pleopods, and relatively simple structures of the second pleopod endopodite. All LL ir ceolus are eyeless and unpigmented, and many of the species are minute in size, with the largest species in the genus reaching a length of 8.8mm (Lewis & Bowman 1996; Lewis 2001). e zoogeography, ecology and morphology of the seven species suggest division of the genus into three subgroups: (1) LL irceolus serrata, a stygobite in the Ozarks of southern Missouri and one of the smallest asellids in North America at a maximum length of 2.5mm. is species has some facets of its anatomy that are unique, including the lack of retinaculae on the male rst pleopod, and male second pleopod endopod exhibiting apparent torsion as well as a prominent apical process. at not-withstanding the species appears to t the diagnosis of LL ir ceolus. e morphology displayed by another Ozark species, CC aecid otea ancyla, is suggestive of an intermediate between CC aecid otea and LL irceolus. (2) Stygobitic LL irceolus in Texas and adjacent border area of Mexico, including LL hardeni, LL. bisetus, LL. pilus, LL. nidulus and LL. cocytus. is group ranges from 2.2 to 8.8 mm in length and includes the smallest asellid (LL hardeni) in North America. All have maxilla 1 inner lobe with 5 setae except L L hardeni, which has only 4 setae. Some of the species exhibit varying degrees of reduction or simplication of pleopods 4 and 5. (3)L L irceolus smithii inhabiting the phreatic groundwater of the Edwards Aquifer feeding the artesian well at San Marcos, Hays County, Texas. is species has perhaps the most morphologically derived mouthparts of a North American asellid. e rst maxilla has inner/outer lobes with 8/10 setae, rather than the 5/13 seen in all C C a ecidotea. Likewise, the mandible has 3-cuspate incisors and 2-cuspate lacinia mobilis, as opposed to 4/4 cuspate in most other North American asellids, including the other species of L L irceolus as well as all C C a ecidotea Furthermore, pleopods 4 and 5 have the endopod and exopod fused into a single eshy ramus (Bowman & Longley 1976). Figure 3. Comparison of pereopods of (A) Remasellus parvus, modied to accommodate swimming; and (B) Lirceolus nidulus, adapted for conentional crawling. [Remasellus drawing by late omas E. Bowman with permission.]

PAGE 52

15th International Congress of Speleology Biological Sciences 1313 2009 ICS Proceedings Since Texas was inundated during the Cretaceous marine transgression, the source of the fauna was probably via postCretaceous dispersal from the Ozarks. LL ir ceolus serratus could be a geographic relict of a group that dispersed into Texas and northern Mexico. Further inference of the eects of the dispersal into Texas is elucidated by examination of the well-known amphipod/isopod assemblage of the artesian well at San Marcos. e seeming paucity of asellid diversity (with one species, LL smithii) as compared to subterranean amphipods (10 species) found in the artesian well is more understandable when the source of the faunas are considered (Holsinger & Longley 1980). Eight of the 10 amphipods were clearly derived from marine ancestors, leaving only the two species of the ancient freshwater crangonyctid genus Stygobromus. us, when compared to the part of the artesian well fauna that was derived from freshwater (rather than marine) ancestors, the asellids measure more comparably. 3. SummaryAsellids have been in North America dating back to at least the Mesozoic. eir distribution has been shaped by many factors including marine transgressions and the upli of the Rocky Mountains. Today the majority of the species occur in the eastern United States, with most of those occurring in subterranean waters. In contrast, generic diversity is greater in western North America.ReferencesBowman, T.E. 1975. ree new troglobitic asellids from western North America (Crustacea: Isopoda: Asellidae). International Journal of Speleology 7: 339. Bowman, T.E. and G. Longley. 1976. Redescription and assignment of the new genus LL ir ceolus of the Texas troglobitic water slater AA se llus smithii (Ulrich)(Crustacea: Isopoda: Asellidae). Proceedings of the Biological Society of Washington 88 (45): 489. Bowman, T.E. and B. Sket. 1985. RR emasaellus, a new genus for the troglobitic swimming Florida asellid isopod, A A se llus Parvus Steeves. Proceedings of the Biological Society of Washington 98 (3): 554. Fiorillo, A.R., D.W. Norton, and P.J. McCarthy. 2008. Beringia from a Cretaceous perspective. 2008 Park Science Symposium in conjunction with the Beringia Days 2008 International Conference. Abstract. Henry, J-P. and G. Magniez. 1970. Contribution a la Systmatique des Asellides (Crustacea Isopoda) Annales de Splologie 25 (2): 335. Henry, J-P. and G. Magniez. 1995. Nouvelles donnes sur les Asellidae pigs dExtrme-Orient (Crustacea, Isopoda, Asellota). Contributions to Zoology 65 (2): 101, SPB Academic Publishing, Amsterdam. Henry, J-P, J.J. Lewis, and G. Magniez. 1986. Isopoda: Asellota: Aselloidea, Gnathostenetroidoidea, Stenetriodea. In Stygofauna Mundi, a faunistic, distributional and ecological synthesis of the world fauna inhabiting subterranean waters (including the marine interstitial). L. Botosaneanu, ed. E.J. Brill, Leiden, e Netherlands, 740 pp. Holsinger, J.R. and G. Longley. 1980. e subterranean amphipod crustacean fauna of an artesian well in Texas.Smithsonian Contributions to Zoology 308: 1. Koenemann, S. and J.R. Holsinger. 2001. Systematics of the North American subterranean amphipod genus Bactrurus (Crangonyctidae). Beaufortia 51 (1): 1. Lewis, J.J. 1980. A comparison of Pseudobaicalasellus and CC aecidoteai, with a description of CC aecid otea bowmani,n. sp. (Crustacea: Isopoda: Asellidae). Proceedings of the Biological Society of Washington 93 (2): 314. Lewis, J.J. 1982. A diagnosis of the Hobbsi Group, with descriptions of CC aecid otea teresae, n. sp., and CC. macropropoda Chase and Blair (Crustacea: Isopoda: Asellidae). Proceedings of the Biological Society of Washington 95 (2): 338. Lewis, J.J. 2001. ree new species of subterranean asellids from western North America, with a synopsis of the species of the region (Crustacea: Isopoda: Asellidae). Texas Memorial Museum, Speleological Monographs 5: 1. Lewis, J.J. 2008. Non-karst groundwater fauna of glaciated Indiana. Final report, Lewis & Associates Biological Consulting, to Indiana Department of Natural Resources, 8 pages.

PAGE 53

Biological Sciences 1314 2009 ICS P roceedings 15th International Congress of Speleology Lewis, J.J. 2009a. Isopoda (Aquatic Sowbugs). In: G. E. Likens, editor, Encyclopedia of Inland Waters. Volume 2: 346. Elsevier, London, UK. Lewis, J.J. 2009b in press. Two new subterranean isopod genera from the western United States (Crustacea: Isopoda: Asellidae). Bulletin of the Socit Internationale de Biospologie. Lewis, J.J and T.E. Bowman. 1981. e subterranean asellids (CC aecid otea) of Illinois (Crustacea: Isopoda: Asellidae).Smithsonian Contributions to Zoology, 335: 1. Lewis, J.J and T.E. Bowman. 1996. e subterranean asellids of Texas (Crustacea: Isopoda: Asellidae). Proceedings of the Biological Society of Washington 109 (3): 482-500. Lewis, J.J., J.W. Martin, and R. Wetzer. (2003). C C o lumbasellus acheron, a new genus and species of subterranean isopod from Washington (Crustacea: Isopoda: Asellidae). Proceedings of the Biological Society of Washington 116 (1): 190. Lewis, J.J., G.O. Graening, D.B. Fenolio, and E.A. Bergey. 2006. CC aecidotea mackini, new species, with a synopsis of thesubterranean asellids of Oklahoma (Crustacea: Isopoda: Asellidae). Proceedings of the Biological Society of Washington 119 (4): 563. Magniez, G.J. 1996. AA se llus aquaticus et se proches parents: un tranger parmi laf aune asellidienne dEurope. Mmoires de Biospologie, 23: 181. Packard, A.S. 1871. On the crustaceans and insects. Pp. 744-761, in Packard, A.S. and F.W. Putnam, editors. e Mammoth Cave and its inhabitants. American Naturalist 5 (12): 739. Weishampel, D.B., P.M. Barrett, R.A. Coria, J. Le Loue, X. Xu, X. Zhao, A. Sahni, E.M.P. Gomani and C.N. Noto. 2004. Dinosaur distribution, in Weishampel, D. B., P. Dodson and H. Osmlska, eds. e D D in osauria (2nd ed.). Berkeley: University of California Press. Pp 517. Wilson, G.D.F. 1999. Some of the deep-sea fauna is ancient. Crustaceana 72 (8): 1019.

PAGE 54

15th International Congress of Speleology Biological Sciences 1315 2009 ICS Proceedings STRUCTURE CHARACTERIZATION OF THE ARTHROPOD FAUNA COMMUNITY IN CAV V E OF THE RASCADERO, SANTANDER, COLOMBIAMARIO A A NDRs S MURCIA L L p P EZ1, D D IEGO C C As S ALLAs S PAb B N2, AND YANEt T H MUOZSAb B A3 1Bilogo, II nstituto AA lexander on HH umboldt, Bogot DD .C C., CC olombia, mariomurcia30@gmail.com2EEstudiante de Biologa, LL aboratorio de Mamferos, II nstituto de CC iencias NN aturales, UU niversidad NN acional de CC olombia, Bogot D D .C C., CC olombia, AA sociacin EEspeleolgica CC olombiana (EESPELEOCOL ELEOCOL), casallas45@hotmail.com3Profesora AA sociada, II nstituto de CC iencias NN aturales, UU niversidad NN acional de CC olombia, sede Bogot DD .C C., AA.A A. 7495, Bogot D D .C C., CC olombia, Vicepresidenta de la AA sociacin EEspeleolgica CC olombiana (EESPELEOCOL ELEOCOL), ydmunozs@unal.edu.co Abstract We have characterized the communities of arthropods of Rascadero Cave in the Guanentina region of the Santander Department, Colombia. Manually collected samples were taken at intervals of 1 hour for each transect using pitfall traps without bait, and three-line transects, with 5 traps each. We identied the specimens collected at the taxonomic family level, nding a wealth of nearly 60 taxa. We assessed richness and rarefaction for the purpose of establishing diversity at the Alpha level for the cavern, and sought to identify patterns in the structure and zonation of the communities of arthropod cave fauna. We evaluated the relative abundance of species found in the pitfall traps. e estimated value of its importance within the communities was analyzed using indices of dominance and equity. Very diuse patterns found in community structures constitute a complex web of nonlinear dimensional trophic relationships. e boundaries of communities only allow us to distinguish three major trophic levels: the so-called guanivores (detritivores and saprophages), who depend almost entirely on guano, the omnivorous generalists, and the predators both small and large, the latter being the dominant forces in the community. Resumen Con el n de caracterizar las comunidades de Artrpodos de la cueva el Rascadero pertenecientes a la regin Guanentina en el departamento de Santander-Colombia, se realizaron muestreos de recoleccin manual en intervalos de tiempo establecidos de 1 hora por cada transecto, y se usaron trampas de cada sin cebo (Pitfall), para tres transectos lineales, con 5 trampas cada uno. Se identicaron los especimenes colectados a nivel taxonmico de Familia en su mayora, encontrndose una riqueza de cerca de 60 taxones. Se hallaron ndices de riqueza especica y de rarefaccin con el n de establecer la diversidad Alpha para la caverna; y con el propsito de aproximarnos a identicar patrones en la estructuracin y zonacin de las comunidades de artropofauna caverncolas. Se evalu la abundancia relativa para las especies halladas en Pitfall, se calcul su valor de importancia dentro de las comunidades analizadas haciendo uso de ndices de Dominancia y de Equidad. Encontrndose patrones muy difusos en la estructura de las comunidades, que se constituyen en una compleja red tridimensional de relaciones trcas no lineales, determinados por las adaptaciones de las especies all presentes, en los cuales los limites de las comunidades solo son claros para distinguir tres grandes niveles trcos: los denominados Guanivoros (Detritivoros y Saprofagos), que son los que dependen casi completamente del recurso guano, los Omnvoros o Generalistas, y los Depredadores pequeos y grandes, siendo estos ltimos fuerzas dominantes en la comunidad.

PAGE 55

Biological Sciences 1316 2009 ICS P roceedings 15th International Congress of Speleology SMALLSCALE SPATIAL DISTRIBUTION OF AQ UATIC FAUNA IN CAV V ES FROM NORTHWESTERN ROMANIA EASTERN EUROPEII OANA NICOLE NICOLE TA A MELEG, OO ANA TEODORA EODORA MOLDOv V AN, SANDA I I Ep P URE, TRAIAN BRAD E E mil RR acoi II nstitute of Speleology, CC linicilor 5, 400006 CC luj-N N apoca, RRomania, ioana.meleg@hasdeu.ubbcluj.ro Abstract Spatial patterns of groundwater biodiversity are poorly known and this type of knowledge is essential for understanding local variation in groundwater assemblages. We have investigated the local diversity patterns of aquatic fauna from percolation water in ve karst caves in the Pdurea Craiului Mountains (Apuseni Mountains, northwestern Romania). e caves are located between the catchment areas of Criul Repede and Criul Negru rivers, in an area of approximately 3 km2, at an average elevation of 450 m a.s.l., and drain water towards the two rivers. Percolating water was sampled continuously between November 2007 and March 2009. Variation of physico-chemical parameters (i.e. pH, temperature, electrical conductivity, Al, Cr, Fe, NO2 -, oxygen consumption CCOMn) and species distribution pattern have been evaluated. Spatial correlations were established between stations inside one cave (with a distance from one site to another that ranged from 3 to 500 m), as well as between caves (with a distance that ranged from 1 to 15 km). e aquatic fauna of the selected caves is heterogeneously distributed and varies largely between stations. e fauna found in percolating water consists of a mixture of epigean and hypogean fauna, and includes eleven taxonomic groups: Nematoda, Oligochaeta, Gastropoda, Hydrachnidia, Amphipoda, Isopoda, Cyclopoida, Harpacticoida, Ostracoda, Collembola, and insect larvae. e biodiversity in the ve caves was dominated by crustaceans with both high relative abundance and species richness, especially harpacticoid copepods. Among crustaceans, four species out of fourteen are endemic for this area, represented by harpacticoid copepods (two taxa) and niphargid amphipods (two taxa). e percolating water was not contaminated and reveals small chemical variations throughout the year. Statistical analysis showed that the pattern of species distribution is inuenced by water chemistry within caves, while between caves no signicant correlation was established. e factors responsible for the observed patterns are further discussed within local and regional context. e present study has yielded new data regarding the groundwater fauna distribution in this karst region. Long-term sampling eorts can reveal an increase in species richness that can be of high interest in developing conservation policies in the future.

PAGE 56

15th International Congress of Speleology Biological Sciences 1317 2009 ICS Proceedings ECOLOGICAL STUDIES IN INTERSTITIAL HABITATS OF ROMANIAN CARPATHIANS EASTERN EUROPEOANAOANA TEODORA EODORA MOLDO OLDO VAN AN E E mil RR acoi II nstitute of Speleology, CC linicilor 5, 400006 CC luj-N N apoca, RRomania, oanamol@hasdeu.ubbcluj.ro Abstract Only the animals that live between sediment grains are considered interstitials although, according to the Latin etymology of the word interstitial, many of the subterranean habitats can be considered as the living space of interstitials. Whether terrestrial or aquatic, and whether living in karstic or non-karstic areas, their habitats are various and encompass the crack and fracture networks of karstic massifs, the alluvia of surface rivers, the sediments of pools in a cave, etc. Environmental variations in such habitats are buered and they represent the ecotone zone between two dierent habitats and faunas. eir study is important for their important biodiversity, and as a mosaic of animals with dierent origins. Our present study is a parallel comparison between two interstitial habitats for groundwater fauna, (1) the water in the alluvia or surface waters, and (2) the water of the caves unsaturated zone. Ecological studies were undertaken during a period of 14 months in two dierent areas of the Apuseni Mountains of Romania (Western Carpathians). Sampling on a monthly basis (alluvia) or continuously (dripping water in caves) together with measurements and analysis of the physico-chemical parameters has revealed the unknown biodiversity of hypogean animals in such habitats in relation to related epigean communities and water quality. Dierences and common features of these two interstitial habitats are discussed. e purpose of such study is to emphasize the high biodiversity of interstitial habitats, their mosaic composition of fauna of dierent origin, their importance in the economy of the subterranean environment and the need for complex biological studies for the conservation and protection of underground fauna.

PAGE 57

Biological Sciences 1318 2009 ICS P roceedings 15th International Congress of Speleology ENERGY FLOW IN THE CAV V ES WITH EMPHASIS IN ITS ARTROPOFAUNA, WITH SPECIAL REFERENCE TO SANTANDER COLOMBIAN CAV V ESYANEt T H MUOZSAb B A1, L L IGIA R R Os S ARIO BENAvV IDEs S SILv V A2, MANUEL A A Nt T ONIO H H Oy Y Os S R R ODRGUEZ3, MARIO A A NDRs S MURCIA L L p P EZ4, D D IEGO C C As S ALLAs S PAb B N5 1Profesora AA sociada, II nstituto de CC iencias NN aturales, UU niversidad NN acional de CC olombia, sede Bogot DD .C C., AA.A A. 7495, Bogot D D .C C., CC olombia, Vicepresidenta de la AA sociacin EEspeleolgica CC olombiana (EESPELEOCOL ELEOCOL), ydmunozs@unal.edu.co2DDepartment of Biological Sciences, e GG eorge Washington UU niversity, 2023 GG Street, NN W, Washington DD .C C. 20052. Miembro Vocal de la AA sociacin EEspeleolgica CC olombiana (EESPELEOCOL ELEOCOL), lbenavides@oeb.harvard.edu3Bilogo, LL aboratorio de Mamferos, II nstituto de CC iencias NN aturales, UU niversidad NN acional de CC olombia, Bogot DD .C C., C C olombia, Miembro Vocal de la AA sociacin EEspeleolgica CC olombiana (EESPELEOCOL ELEOCOL), mahoyosr@unal.edu.co4Bilogo, II nstituto AA lexander on HH umboldt, Bogot DD .C C., CC olombia, mariomurcia30@gmail.com5EEstudiante de Biologa, LL aboratorio de Mamferos, II nstituto de CC iencias NN aturales, UU niversidad NN acional de CC olombia, Bogot D D .C C., CC olombia, AA sociacin EEspeleolgica CC olombiana (EESPELEOCOL ELEOCOL), casallas45@hotmail.com Abstract To establish how the energy ows inside the caves, we have explored 79 Colombian caves taking into account its fauna composition, its location, the landscape around them, the ecological relation between the species and the system and its functional group. To solve it, there is a theoretical proposal about the energy ow in the neotropical underground ecosystems. Considering that there is a highest richness of species in the ecosystem, there will be the highest number of connections in the dierent tropical levels. is is a changeable characteristic depending on the system, its size, its fauna diversity, the aquatic component interaction and the functional groups that it has. e caverns as semi-closed systems have Trogloxenous organisms such as the bats and other bigger vertebrates that include energy in guano form. erefore, the Trogophiles that live inside the guano are the most important energy transformers. e guano, is the main source of exogenous energy that feeds the neotropical caves and its organisms. To do it, it needs to develop assimilation process and take advantage of it, thank to some arthropod the energy could be re-included in the guano. When the organisms that live there died, or when rst and second degree predator energy could be come out of the trophic net of the guano. In any case, the energy could be lost if the cavern suers a washing seasonal, which cause could be the oods. In Santander, Colombia at the rural zone of the town Chipat, there has been found two caves (Cueva del Naranjo y cueva del Cenicero) with particular ecological characteristics that gives indications of special trophic conditions. In both of them, some guano tokens have been analyzed zones with dierent light intensities, evaluating by a bromatological study of the chemical elements in there. e Cueva del Cenicero has a lack of process related to the organic matter deterioration; it is known that it is not a dry cavern but here was a big quantity of dead bats mummied (Phyllostomus discolor), these results show a variation in some levels of the decomposition cycle. Besides, la Cueva del Naranjo is a system because its morphology let the oblique entrance of the sun light during the whole cavity, making a possible dierentiation in the ways of starting of the energetic ow. And also, with the knowledge of the fauna composition indoor it is possible to propose some energy ow diagrams for those systems that explain the peculiarities that contribute to both caverns to have a particular trophic behavior, far away from a conventional ow system.

PAGE 58

15th International Congress of Speleology Biological Sciences 1319 2009 ICS Proceedings Flu LU Jo O de DE energ ENERG a A en EN ca CA Vernas ERNAS con CON nfasis NFASIS en EN su SU Artropofauna RTROPOFAUNA con CON referencia REFERENCIA especial ESPECIAL un UN cue CUE Vas AS de DE Santander ANTANDER Colombia OLOMBIAResumen a al interior de la cueva, se propone posibles esquemas dCon el n de establecer cmo uye la energa dentro de las cuevas; se exploraron 79 cuevas colombianas, teniendo en cuenta la composicin de fauna, la ubicacin de esta en la caverna, el paisaje que rodea la cueva, el grado de delidad de las especies al sistema (relacin ecolgica) y su grupo funcional. A partir de esto se plantea una propuesta terica del ujo de energa de los ecosistemas subterrneos neotropicales, considerando que mientras mayor sea la riqueza de especies en un sistema, mayor va a ser el nmero de conexiones entre los diferentes niveles trcos. Siendo esto variable dependiendo del sistema, el tamao del mismo, la abundancia y diversidad de su fauna, la interaccin de un componente acutico y los grupos funcionales que presente. Las cuevas como sistemas semicerrados, poseen organismos trogloxenos tales como murcilagos y otros grandes vertebrados que incorporan energa en forma de guano, en tanto que los trogllos, que viven dentro del guano son los principales transformadores de energa. El guano, es la principal fuente de energa exgena que nutre la mayora de cuevas neotropicales y sus organismos, este sufre diferentes procesos asimilacin y aprovechamiento, sensacional es por parte de una diversa cantidad de artrpodos, esta energa puede ser reincorporada al guano, al morir los organismos que habitan en l, o en caso que los predadores de primer y segundo orden sea trogllos, puede salir de la red trca del guano. En cualquier caso esta energa se puede perder si la caverna sufre proceso de lavado estacional, a causa de inundaciones. En Colombia, para el departamento de Santander en la zona rural del municipio de Chipat, se encontraron dos cuevas (Cueva del Naranjo y cueva del Cenicero), con patrones ecolgicos particulares, que dan indicios de condiciones trcas especiales. En ellas se analizaron muestras de guano en zonas con diferentes intensidades de luz, evaluando mediante un estudio bromatolgico los elementos qumicos presentes. Se encontr que la Cueva del Cenicero, posee una carencia de procesos relacionados con la degradacin de materia orgnica, Si bien no es una cueva seca, se hallaron en ella una gran cantidad de cuerpos de murcilagos momicados (Phyllostomus discolor), que evidencia una variacin en algunas etapas del ciclo de descomposicin. Por otra parte la cueva del Naranjo es un sistema que debido a su morfologa permite la entrada oblicua de luz solar, a lo largo de la cavidad, generando una posible diferenciacin en las vas por las que inicia el ujo energtico en esta cavidad. Tomando en cuenta la composicin faunstice ujo energtico puntuales para estos dos sistemas, que expliquen las peculiaridades que llevan a estas cuevas a tener un comportamiento trco particular, lejos de un sistema de ujo convencional.

PAGE 59

Biological Sciences 1320 2009 ICS P roceedings 15th International Congress of Speleology COLOMBIAN BATS FROM THE CAV V ES YANEt T H MUOZSAb B A1, MANUEL A A Nt T ONIO H H Oy Y Os S R R ODRGUEZ2, D D IEGO C C As S ALLAs S PAb B N3 1Profesora AA sociada, II nstituto de CC iencias NN aturales, UU niversidad NN acional de CC olombia, sede Bogot DD .C C., AA.A A. 7495, Bogot D D .C C., CC olombia, Vicepresidenta de la AA sociacin EEspeleolgica CC olombiana (EESPELEOCOL ELEOCOL), ydmunozs@unal.edu.co2Bilogo, LL aboratorio de Mamferos, II nstituto de CC iencias NN aturales, UU niversidad NN acional de CC olombia, Bogot DD .C C., C C olombia, Miembro Vocal de la AA sociacin EEspeleolgica CC olombiana (EESPELEOCOL ELEOCOL), mahoyosr@unal.edu.co3EEstudiante de Biologa, LL aboratorio de Mamferos, II nstituto de CC iencias NN aturales, UU niversidad NN acional de CC olombia, Bogot D D .C C., CC olombia, AA sociacin EEspeleolgica CC olombiana (EESPELEOCOL ELEOCOL), casallas45@hotmail.com Abstract Since 1994 to our days, we have had some visits to Antioquia, Boyac and Santander (Colombian departments), nearly 80 caves in which there is a report of 28 dierent bats species of the 178 types that live in Colombia. More than a half of them are Insectivores, followed by the fungivores, nectarivorespollenivores, and hematophagous. Some of these species are LL onchorhina marinkelli Colombian an endemic species and typical of these systems. e cave with the most species richness with 7 species, is the Cueva Caja de Agua located in the Huila department.MurcilagosURCILAGOS colombianos COLOMBIANOS habitantes HABITANTES de DE ca CA Vernas ERNASResumen Desde 1994 hasta la fecha se han visitado en los departamentos de Antioquia, Boyac y Santander (Departamentos de Colombia) cerca de 80 cavernas, en las cuales se reportan 28 especies de murcilagos de las 178 que hay en Colombia. La gran mayora de estas son insectvoras, seguidas por las frugvoras, nectarvoras polinvoras y sanguinvoras. Algunas de estas especies como L L onchorhina marinkelli, especie endmica de Colombia y propias de estos sistemas. La caverna con mayor riqueza de especies es la Cueva Caja de Agua en el departamento del Huila con siete especies.

PAGE 60

15th International Congress of Speleology Biological Sciences 1321 2009 ICS Proceedings BIOSPELEOLOGICAL PROV V INCES IN COLOMBIAYANEt T H MUOZSAb B A1, L L Us S G G UILLERm M O BApt PT Ist ST E2, D D ANILO SALAs S3 1Profesora AA sociada, II nstituto de CC iencias NN aturales, UU niversidad NN acional de CC olombia, sede Bogot DD .C C., AA.A A. 7495, Bogot D D .C C., CC olombia, Vicepresidenta de la AA sociacin EEspeleolgica CC olombiana (EESPELEOCOL ELEOCOL), ydmunozs@unal.edu.co2Profesor AA sociado, Ponticia UU niversidad Javeriana, Bogot DD .C C. CC olombia, Miembro Vocal de EESPELEOCOL ELEOCOL guillermo. baptiste@javeriana.edu.co3CC oordinador de Programas. Fundacin Moiss Bertoni. Paraguay. danilosalas@gmail.com Abstract In 1998, the rst map showing Colombian cavern ecosystems represented 84 reported caves and grottoes. At present, there are about 130 caves located in the following bioespeleological provinces: (1) Cordillera Boyacense-Santander (Boyaca, Cundinamarca, Norte de Santander, Santander), (2) southeastern Antioquia, (3) Central Andes Mountains (Cauca, Huila, uindio, Tolima, Valle del Cauca), (4) border (or piedmont) of the Eastern Andes (Caqueta, Guaviare, Meta, Putumayo), (5) Guianese Shield along the mountains of La Macarena, e Lindoso, Araracuara, and the complex of mountains of Vaups; (6) mountains of the southern Colombian Andes (Nario and west border of Putumayo montains), (7) and the mountains of Choc (8) Serranas Cordoba; (9) Sierra Nevada de Santa Marta (10) Serrania de Perija, (11) La Guajira mountains and (12) North of San Andres Islands, for a total of twelve biospeleological provinces, which have established partnerships with regard to their origin, biota and faunal exchange.Pro RO Vincias INCIAS Bioespeleol IOESPELEOL gicas GICAS en EN Colombia OLOMBIAResumen En 1998 se ubicaron por primera vez en un mapa de ecosistemas en Colombia las cavernas, cuevas y grutas, las cuales fueron cerca de 84 reportes. En la actualidad se registran alrededor de 130, localizados en las siguientes provincias bioespeleolgicas: (1) Cordillera BoyacenseSantandereana (departamentos de Boyac, Cundinamarca, Norte de Santander, Santander); (2) Suroriente Antioqueo; (3) Montaas de los Andes Centrales (departamentos del Cauca, Huila, uindo, Tolima, Valle del Cauca); (4) Piedemonte de la Cordillera Oriental (departamentos del Caquet, Guaviare, Meta, Putumayo); (5) Escudo Guyans junto con las serranas de La Macarena, La Lindosa, Araracuara y el complejo de serranas del departamento del Vaups; (6) Montaas del Sur de los Andes colombianos (departamentos de Nario y pie de monte occidental del Putumayo); (7) Serranas y Cerros del Choc Biogeogrco; (8) Serranas de Crdoba; (9) Sierra Nevada de Santa Marta; (10) Serrana de Perij; (11) Cerros y Serranas de La Guajira y; (12) Regin de San Andrs Islas, para un total de doce provincias bioespeleolgicas, las cuales se han establecido por asociaciones en cuanto a su origen, biota e intercambio faunstico.

PAGE 61

Biological Sciences 1322 2009 ICS P roceedings 15th International Congress of Speleology METHODS FOR COUNTING CAV V E CRICKETSGG RO RO BER ER T MYER ER S III IIIZ Z ara EE nironmental LLCLLC 118 W. GG oforth RR d, Buda, TX 78610 Abstract Cave crickets in the genus CC eut hophilus are a primary source of energy for cave ecosystems that support endangered species in central Texas. Many cave preserve managers monitor cricket populations as an indicator of cave community health using various methods. Crickets are well suited as indicator species because they are more numerous and easy to detect than troglobites. ey are also key to introducing nutrients into cave ecosystems by foraging on the surface at night and roosting in caves during the day. Researchers have found that in central Texas the number of individuals of CC eut hophilus spp. crickets counted in a cave correlate fairly well with the level of human impact at that cave, and the total numbers of individuals of other taxa in a cave strongly correlate with the total number of cave crickets counted in that cave. Two methods, in-cave counts and exit counts, are used to monitor cricket populations; however, little is known about the relative precision or accuracy of either. A study was initiated in 2005 to compare results produced by each method. In-cave counts and exit counts are performed for three days every other month at three caves on Camp Bullis Military Reservation in northern Bexar County, Texas. e in-cave counts are performed during the day and exit counts on those same evenings, allowing comparison of each method on the same day and examination of day-to-day variation. During in-cave counts all crickets detected on the ceilings and walls are counted using timed visual searches. During exit counts crickets are counted as they emerge from cave entrances for approximately two hours following sunset. Crickets are distinguished by age class (nymph, sub-adult, adult) and species (CC uethophilus secretus, CC. sp. B, and occasionally CC cunicularis). Initial data analysis reveals that each method has strengths and weaknesses, which are important to understand when interpreting data that guide management decisions. It is also important to understand the capabilities of monitoring methods when setting monitoring goals and designing monitoring programs, for instance detecting a given degree of population decline within a certain time frame. Since monitoring cave species is dicult and funding for management is oen limited, it is important that cricket monitoring programs be well designed with clear and achievable goals.

PAGE 62

15th International Congress of Speleology Biological Sciences 1323 2009 ICS Proceedings REV V IEW AND ASSESSMENT OF KNOWN CAV V ERNICOLES A ND RARE EPIGEAN BIOLOGY OF KARST AND CAV V ES, GREAT SMOKY MOUNTAINS NATIONAL PARKDANIELDANIEL C C NOL NOL FI IN N ational Park Service, GG reat Smoky Mountains NN ational Park. 107 Park HH eadquarters RR d., GG atlinburg, TNN 37738 UU SA A and Western Kentucky UU niversity, HH oman EE nironmental RResearch II nstitute. Bowling GG reen, KY 42101UU SA A e Great Smoky Mountains National Park (GRSM) encompasses 2100 km2 in the states of North Carolina and Tennessee, United States of America. e park receives over 9.5 million visitors a year, making it the most visited national park in the U.S. Karst areas in GRSM account for less than 1% of the total surface land area in the park, but most karst areas receive highly concentrated visitation. GRSM is the most biologically diverse national park in the National Park System; only 12% of the suggested 100,000 species who live in the park have been identied. Research to date suggests that karst areas in GRSM exhibit signicantly high biological diversity. Six karst areas in GRSM contain 16 known caves with signicant biological resources. Several endemic species have been identied in the caves and karst areas. In addition, park caves provide habitat for federally listed threatened and endangered species. Under the guidance of the Federal Cave Resources Protection Act of 1988, the National Park Service is required to inventory and list signicant caves on federal lands, and to provide management and dissemination of information about caves. e park does not currently have a cave management plan in place in order to eectively manage the delicate cave and karst ecosystems. Although signicant advances have been made in recent years in the inventory and monitoring of the biology of the cave and karst areas in the Great Smoky Mountains National Park, resource managers still do not have a complete picture of the cavernicoles and rare epigean biology of the park. To date, there has been no compilation of the list of known cavernicoles and biota associated with the caves and karst of Great Smoky Mountains National Park. is paper describes important cave and karst areas of concern in Great Smoky Mountains National Park, provides a review of relevant literature, both peer reviewed and National Park Service internal documentation related to cave and karst areas, and begins to identify and compile a list of the cavernicoles and rare epigean biology which have been thus far described in the literature of karst and caves in the Great Smoky Mountains National Park. is document should provide the basis for creation of a structured inventory and monitoring eort to further identify and describe species of concern for managers of the cave and karst resources of the Great Smoky Mountains National Park.1. IntroductionGreat Smoky Mountains National Park (GRSM) encompasses 2,108.76 square kilometers in the states of North Carolina and Tennesse. It receives over 9 million visitors a year, making it the most visited national park in the world. e author recognizes 16 solution caves in six main karst areas. Karst areas within the Great Smoky Mountains exist in the western portion of the Tennessee side of the park. Current management of cave resources is not based on an approved management plan for caves and karst; two dra management plans (GRSM 1979b, GRSM 1989) were developed, but their implementation has been limited due to their lack of accuracy and completion. In addition, GRSM currently recognizes and addresses management of 12 caves. Management decisions in regards to cave and karst resources currently follow general directives and the parks superintendents compendium. Caves and karst of the Great Smoky Mountains National Park have formed in windows of calcium carbonate in the Great Smoky Mountain thrust sheet. e older metamorphic rocks that make up the Great Smoky Mountains were thrust over much younger limestone and dolostone (dolomitic limestone) along the Great Smokies

PAGE 63

Biological Sciences 1324 2009 ICS P roceedings 15th International Congress of Speleology Fault (King ING 1968; Neuman EUMAN 1947; Southworth OUTHWORTH et al., 2000; Wilson I LSON 1935). Erosion of the metamorphic rock exposed the underlying limestone and dolostone, forming the aforementioned windows, which in turn eroded into the at, grassy valleys typical of the Valley and Ridge Provinces of the Appalachians. It is in the lower walls and oors of these valleys where karst topography and their caves have formed (Barr A RR 1968; GRSM, 1989). Ecologically, karst areas of GRSM support diverse biota oen completely dierent than areas of GRSM with noncarbonate basement rock. Although park managers are just beginning to work towards a comprehensive inventory, over 270 organisms, (including rare and endemic species), have been described in association with caves and karst. Caves in the park provide important habitat for several species of bats (including one federally listed species), and are home to at least one endemic species. In addition, species diversity of GRSM epigean (above ground) karst terrain is known to be higher than non-karst areas of comparable size. Characteristics such as these exemplify the importance of continued cave and karst research and management eorts.2. Cave Biology (Biospeleology)Although interest in the study of cave biota in the United States began in the midto late 1800s, the majority of current information has been compiled since the 1950s (Barr ARR 1968; Elliot LLIOT 2000). Cave biota exhibit unique morphology, traits associated with subterranean life such as eye and pigment loss, delicate form, and enhanced extraoptic sensory structures (Culver U LVER et al., 1999; Culver U LVER et al., 2000; Peck E CK 1998; New York University (NYU) Website, 2008). ese adaptations are of great interest to scientists in the study of natural selection, gene ow, and genetic and morphologic changes in species (Culver U LVER et al., 1999; Culver U LVER et al., 2000; NYU Web site, 2008). In addition to morphological adaptations, obligate aquatic cave organisms have long life spans, and are therefore likely to accumulate toxins. Such organisms are highly sensitive to water contamination, and are of interest to scientists as indicators of groundwater quality (Culver U LVER et al., 1999; Culver U LVER et al., 2000; NYU Web site, 2008). Cave inhabitants can be grouped into four major groups: troglobionts (and stygobionts), troglophiles (and stygophiles), trogloxenes (and stygoxenes) and accidental cave inhabitants. Troglobionts (terrestrial) and stygobionts (aquatic) are obligate cave dwellers, species that spend their entire life cycle inside caves, and have (over time) formed adaptations in order to survive life only in caves (Reeves E EVES 2000). Species of animals including sh, salamanders, insects and spiders oen have adaptations such as reduced eyes and/ or no eyesight, loss of pigmentation, elongated antennae, and increased senses of smell and touch (Culver U LVER et al., 1999; Culver ULVER et al., 2000; Peck E CK 1998). ere are over 1300 (425 stygobiont and 928 troglobiont) obligate cave species known (Peck E CK 1998), although it is thought that probably less than half of the obligate subterranean species in the United States have been described (Elliot L LIOT 2000). Obligate cave fauna have the highest reported level of endemism of any taxonomic or ecologic group of organisms in the United States (Culver U LVER et al., 2003). Troglobionts (and stygobionts) are generally of greatest interest to cave biologists and managers, since they are restricted to cave habitats and are sensitive to management practices (Reeves EEVES 2000). Troglophiles and stygophiles are considered to breed and live in cave environments, but are not obligate cave dwellers. An example would be the Cave salamander (EE u rycea lucifuga), which has been found to forage outside of caves. Trogloxenes (terrestrial) or stygoxenes (aquatic) are transient cave dwellers, needing to spend a portion of their life cycle in caves. For example the Indiana bat (Myotis sodalis) uses the constant environment of specic caves to hibernate through the winter. Accidental cave inhabitants cannot survive to reproduce in caves. 3. Cave Biology and Great Smoky Mountains National Park e southeastern United States, including the Appalachian Mountains, exhibits signicant diversity of obligate cave fauna (troglobionts and stygobionts). In addition, Tennessee, in which all of Great Smoky Mountains National Parks caves and karst areas lie, is the fourth most diverse state in regards to genus-level diversity of obligate cave fauna (Peck ECK 1998; Culver ULVER et al., 1999). Great Smoky Mountains National Park is one of the most biologically diverse places in the world. Designated as a International Biosphere Reserve in 1988, the park is home to 10,000 known species, and scientists estimate there are 90,000 species yet to be identied and described (NPS Website, 2008). Although the park lies in one of the most diverse regions and states with regards to obligate cave fauna species richness, park managers are just beginning to work towards a comprehensive assessment of the distribution of obligate cave fauna and rare epigean biology of the cave and karst areas withn the park boundary. As early as the 1930s, park managers expressed an interest in better understanding the cave and karst areas in the park. Early eorts to describe the parks caves were focused more

PAGE 64

15th International Congress of Speleology Biological Sciences 1325 2009 ICS Proceedings on the geology and physical conditions inside the caves rather than the fauna (GRSM, 1936). In 1974, Great Smoky Mountains National Park began studying human impact and restricting access to park caves because of the presence of Indiana bats (Myotis sodalis), a federally protected species (Rabinowitz ABINOWITZ and AND Nottingham OTTINGHAM 1979; GRSM, 1979b). Currently, the only continuous sampling of cave biota is conducted in conjunction with the United States Fish and Wildlife Services Indiana bat hibernacula counts. It was not until the early 1980s when there was a concerted eort by Great Smoky Mountains resource managers to identify and describe species associated with caves and karst. Beginning in 1984, under the direction of Great Smoky Mountains National Park resource managers, Richard L. Wallace completed a series of biological survey reports of the Great Smoky Mountains National Parks caves. Although there existed data on bat fauna that occurred within the parks caves, Wallace noted that knowledge of other cave fauna was incomplete or unknown. Wallaces biological surveys uncovered several new, rare, and endemic species (Wallace ALLACE 1984, 1989, 1990). Based on these ndings, park resource managers began monitoring populations of cavernicoles (e.g., Stygobromus fecundus) in at least one location, Gregorys Cave (Johnson O HNSON 1990). Management of Gregorys Cave was changed in order to provide protection for this endemic species. Without continued interest and direction through a cave management plan, over time, the monitoring eorts have fallen by the wayside with occasional spot checks by visiting scientists. Several independent researchers have surveyed a variety of cave and karst associated biota in the caves of Great Smoky Mountains National Park: amphibians (Dodd O DD and AND Griffey RIFFEY 2001; Dodd ODD 2003), salamanders (Taylor AYLOR and AND Mays AYS 2006), spiders (NPS, 1992), annelids (Reeves EEVES and AND Reynolds EYNOLDS 1999), and general invertebrate cavernicoles (Reeves E EVES 2000). In addition, in 2007, Discover Life in America, a non-prot partner of the national park invited scientists with a specic interest in karst areas to participate in a Karst uest to identify and describe both cave life and associated epigean biology related to the karst areas in the park. Great Smoky Mountains National Parks Cave Management and Management of reatened and Endangered Indiana Bats (GRSM, 1979a) states that Given the unique character of the above-ground environment, there is good reason to believe that a thorough survey and analysis of the underground environment will uncover additional unique and rare biota and minerals. In addition, caves with signicant vertical development, like those of the Smokies, are likely to exhibit higher species diversity than caves with extensive horizontal development (Culver U LVER et al., 2003). Although considerable advances have been made in identifying and monitoring the biota of cave and karst areas within Great Smoky Mountains National Park, park managers still do not have a complete picture of the cavernicoles and rare epigean biology associated with the cave and karst areas in the park. To date, biological data have not been collected, identied, or published on ve of the sixteen caves within the park. 4. Known Cave and Karst Biota of Great Smoky Mountains National ParkAs previously discussed, troglobionts (and stygobionts) are generally of the greatest interest to cave biologists and resource managers. rough the aforementioned inventories of cavernicoles and associated biota of the karst areas in the Great Smoky Mountains National Park, a total of ten troglobiotic organisms have been described in the caves of the park (Table 1). Of these ten organisms, ve have been identied to species, and ve to genus. One of the ten organisms (Stygobromus fecundus) is endemic to one cave (Gregorys Cave). Other organisms include amphipods, isopods, diplurans, millipedes, atworms, and arachnids. Approximately 270 organisms have been otherwise described in association with the caves and karst areas in Great Smoky Mountains National Park. Based on the compilation of available research to date, of these 270, over 200 organisms have been identied to generic level, and 50 to family. Over y of these organisms have been identied as troglophiles, and 40 as trogloxenes. Countless other more commonly occurring organisms use karst habitats nonspecically (DLIA, 2007). Great Smoky Mountains National Park is one of the most biologically diverse ecosystems in the world. Scientists are beginning to comprehend the number of species that call the Smokies home. It is likely that many of the epigean species that inhabit the karst areas in the park are widespread in non-karst areas as well, although researchers have recently begun to inventory the species associated with the habitat surrounding the cave and karst areas (DLIA, 2007). Many species that thrive in karst habitats in the park (Table 2) are rare elsewhere within the park boundary (although they are not considered rare outside of the park). But it is interesting to note, for example, that one species of Asian liverworts only known North American population is restricted to one of the karst areas in the park (Hentschel E NTSCHEL et al., 2008). A wide variety of terrestrial snails inhabit the karst areas in the park, and based on the high level of endemism in karst

PAGE 65

Biological Sciences 1326 2009 ICS P roceedings 15th International Congress of Speleology areas and the Great Smoky Mountains National Park, it is likely once these species have been identied and described, that there will be several endemic species (Culver U LVER et al., 2003). e karst areas in the park exhibit signicant diversity of epigean species, and further research may identify additional rare and endemic epigean species. Although several of the parks known caves have been extensively inventoried, and the cavernicoles well described, it is important to point out that not all of the parks caves Species Common name Record location (cave) Citation Stygobromus fecundus amphipodGRC Wallace 1984, 1989; Mays 2002 Stygobromus sp. (fecundus?) amphipodGRC NPS 1991; NPS 1992; NPS 1993; NPS 1994; NPS 1995; NPS 1996; DLIA 2007 Stygobromus sparsus amphipodGRC, SGC, SPC, RMB, WOBWallace 1984, 1989; Reeves 2000; Mays 2002L L i tocampa sp. dipluran BLC, CC1, RMB, SGC, SPC, WOB Reeves 2000; DLIA 2007 Sphalloplana sp. atwormGRC, RMB Wallace 1984, 1989; NPS 1991; NPS 1992; Reeves 2000; DLIA 2007C C aecid otea incurva isopodGRC, MYC, RMB, WOB Wallace 1984, 1989; NPS 1993; Reeves 2000; Mays 2002; DLIA 2007C C aecid otea sp. isopodMYC Reeves 2000 Scoterpes sp. millipede BLC, CC1, GRC, RMB, RBC, SGC, SPC, WOB Wallace 1984, 1989; Reeves 2000; Mays 2002; DLIA 2007A A p poleptoneta sp. spiderRMB Reeves 2000 Phanetta subterranea spiderBLC, CC1, GRC, RMB, SGCReeves 2000; Mays 2002N N e sticus barrowsi spider CC1, CC2, GRC, RBC, RMB, SGC, SPC, WOB Wallace 1984, 1989; Reeves 2000; Mays 2002; DLIA 2007 Cave location codes: BLC= Bull Cave; CC1 and CC2= Calf Cave 1 and 2; GRC= Gregorys Cave; MYC= Myhr Cave; RMB= Rich Mountain Blowhole; RBC= Rainbow Falls Cave; SGC= Scott Gap Cave; SPC= Saltpeter Cave; WOB= White Oak BlowholeTable 1: Identied and Described Troglobionts and Stygobionts of Great Smoky Mountains National Park. Species Common name Species Common nameA A sp lenium resiliens blackstem spleenwort Frasera/Swertia caroliniensis American columboA A sp lenium ruta-muraria wall-rueH H y banthus concolor Eastern greenvioletC C ar ex eburnean bristle-leaf sedge Hydrophyllum macrophyllumlargeleaf waterlieafC C ar ex jamesii James sedge Mertensia virginica Virginia bluebellsC C e lastrus scandens American bittersweet Pedinophyllum interruptum liverwort*C C o rallorhiza wisteriana early coral root Pellaea atropurpurea purple clireakC C r ataegus calpodendrin pear-thorn Polymnia canadensis whiteower leafcupC C ys topteris bulbifera bladder fern Porella gracillima liverwort*D D ecumaria barbara woodvamp Quercus muhlenbergii Chinkapin oakD D elphinium tricorne dwarf larkspur Smilax hugerii Hugers carrionowerD D entaria multida Muhl toothwort Solidago caesia bluestem goldenrodD D ispo rum lanuginosum yellow fairy bells Spigelia marilandica Indian pinkD D odecatheon maedia American cowslip/shooting star Staphylea trifoliate American bladdernutE E r yngium yuccifolium button eryngo Viburnum rudulum rusty blackhaw  Vitus cinerea graybark grapeTable 2: Vascular and non-vascular* plant species associated with karst areas rare to Great Smoky Mountains National Park.

PAGE 66

15th International Congress of Speleology Biological Sciences 1327 2009 ICS Proceedings biology has been well dened. Much of the research and longer-term monitoring eorts have focused on several specic caves (e.g., Gregorys Cave, S. fecundus and White Oak Blowhole, M. sodalis) and has resulted in a more comprehensive list of species for these caves. Several known caves in the park have not been extensively surveyed, nor has the cave biota been well described. It is also interesting to note that, to date, there has been little to no research concerning the microfauna in the parks caves. In developing a comprehensive cave and karst management strategy, it is important to consider the management of both cave biota and the epigean biology associated with the karst areas.5. Conclusionse relatively recent progress made by the NPS and independent researchers in inventorying the cavernicoles and epigean biology of the caves and karst areas of Great Smoky Mountains National Park has highlighted need for implementing a structured inventory and monitoring program and an approved Cave and Karst Management Plan in order to manage more eectively and eciently the cave and karst resources in the park. ere is no long-term inventory and monitoring program in place for caves and karst areas, although these areas have demonstrated great diversity of species, including rare and endemic species, both troglobiotic and epigean in nature. Because several caves in the park have not been extensively inventoried, it is possible that additional species associated with caves and karst in Great Smoky Mountains National Park have yet to be discovered. In developing and implementing a Cave Management Plan, it would be benecial to park resource managers to incorporate a structured inventory and monitoring program to identify and describe not only the biota, but also the environmental quality and physical conditions of the caves and karst areas in the park. ere is no baseline data for the caves in the park in most cases; establishing a record of current conditions, through a variety of indices (e.g., water quality sampling, photo documentation of current resource conditions, establishing use patterns and carrying capacity of sensitive cave and karst areas, more complete biological inventories in some cases) would provide a basis for future management decisions with regards to caves and karst in the Smokies. Developing a more complete understanding of current conditions, in combination with a long-term monitoring eort and implementation of a Cave Management Plan will allow Great Smoky Mountains National Park to protect the unique cave and karst resources in a consistent and eective manner.6. ReferencesBarr Jr., T.C. 1968. Cave Ecology and the Evolution of Troglolbites. Dobzhansky, T., M.K. Hecht and W.C. Steere (eds). Evolutionary Biology, Volume 2. K. Holland, pp 35 Culver, D.C., H.H. Hobbs III, M.C. Christman, and L.L. Master. 1999. Distribution Map of Caves and Cave Animals in the United States. Journal of Cave and Karst Studies 61(3):139 Culver, D.C., L.L. Master, M.C. Christman, and H.H. Hobbs III. 2000. Obligate Cave Fauna of the 48 Contiguous United States. Conservation Biology 14(2): 386 Culver, D.C., M.C Christman, W.R. Elliot, H.H. Hobbs III, and J.R. Reddell. 2003. North American Obligate Cave Fauna: Regional Patterns. Biodiversity and Conservation 12: 441 Department of the Interior. National Park Service Web site. http://www.nps.gov/ Discover Life in America. 2007. Karst uest Data. Great Smoky Mountains National Park Natural History Files: Caves and Karst Dodd Jr., K.C. and M.L. Griey. 2001. e Cave Associated Amphibians of Great Smoky Mountains National Park: Review and Monitoring. e Journal of the Elisha Mitchell Scientic Society 117(3): 139 Dodd, C.K. 2003. Monitoring Amphibians in Great Smoky Mountains National Park. United States Geological Survey Web site: pubs.usgs.gov Elliot, W.R. 2000. Conservation of the North American Cave and Karst Biota. Dr. H. Wilkens, D.C. Culver and W.F. Humphreys (eds). Ecosystems of the World 30, Subterranean Ecosystems. Elsevier, Amsterdam, pp 665 GRSM. 1936. Caves and Sinks of Cades Cove and Vicinity. Great Smoky Mountains National Park Natural History Files: Caves and Karst GRSM. 1979a. Cave Management and Management of Indiana Bats (GRSM-N-16). Great Smoky Mountains National Park Natural History Files: Caves and Karst

PAGE 67

Biological Sciences 1328 2009 ICS P roceedings 15th International Congress of Speleology GRSM. 1979b. Dra Cave (Indiana bat) Resources Management Plan. Great Smoky Mountains National Park Natural History Files: Caves and Karst GRSM. 1989. Dra Cave Resources Management Plan. Great Smoky Mountains National Park Natural History Files: Caves and Karst Hentschel, J., P. Davidson, and J. Heinrichs. 2008. Porella in North America. Fieldiana: Botany. N.S. (47):183191 Johnson, K. 1990. Report on the Status of Stygobromus Sp. of Gregorys Cave. Great Smoky Mountains National Park Natural History Files: Caves and Karst King, P.B., R.B. Neuman, and J.B. Hadley. 1968. Geology of the Great Smoky Mountains National Park, Tennessee and North Carolina. U.S. Geological Survey Professional Paper 587 Mays, J.D. 2002. A Systematic Approach to Sampling the Arthropod Assemblage of Gregorys Cave. Western Carolin University Masters esis. National Park Service, Great Smoky Mountains National Park Biological Monitoring Report, Gregorys Cave. 1992. Great Smoky Mountains National Park Natural History Files: Caves and Karst Neuman, R.B. 1947. Notes on the Geology of Cades Cove, Great Smoky Mountains National Park, Tennessee. J ournal of the Tennessee AA cademy of Science. Vol. XXII (3) New York University Department of Biology Cave Biology Research Group Web site: http://www.nyu.edu/fas/ cave Peck, S. 1998. A Summary of Diversity and Distribution of the Obligate Cave-Inhabiting Faunas of the United States and Canada. Journal of Cave and Karst Studies 60(1): 18 Rabinowitz, A. and B. Nottingham 1979. Human Visitation and Fall-Winter Cave Usage by Bats in the Great Smoky Mountains National Park. Great Smoky Mountains National Park Natural History Files: Bat Studies Reeves, W.K. 2000. Invertebrate Cavernicoles of the Great Smoky Mountains National Park, USA. e Journal of the Elisha Mitchell Scientic Society 116(4): 334 Reeves, W.K. and J.W. Reynolds. 1999. New Records of Cvae-Dwelling Earthworms in the Southeastern United States, With Notes on eir Ecology. Medrologica 7(10): 65 Southworth, S., P.G. Chirico, and T. Putbrese. 2000. Digital Geologic Map of parts of the Cades Cove and Calderwood uadrangles, Tennessee and North Carolina, Great Smoky Mountains National Park. U.S. Geological Survey Open-File Report 99-175 Taylor, N. and J. Mays. 2006. e Salamanders Eurycea longicauda and Plethodon glutionsus in Gregorys Cave, TN: Monitoring and Observations on Ecology and Natural History. Southeastern Naturalist 5(3): 435 Wallace, R.L. 1984. Biological Survey Report of the Great Smoky Mountains National Park Caves. Great Smoky Mountains National Park Natural History Files: Caves and Karst Wallace, R.L. 1989. Biological Survey Report of the Great Smoky Mountains National Park Caves. Great Smoky Mountains National Park Natural History Files: Caves and Karst Wallace, R.L. 1990. Myhr Cave Fauna. Great Smoky Mountains National Park Natural History Files: Caves and Karst Wilson Jr., C.W. 1935. e Great Smoky rust Fault in the Vicinity of Tuchaleechee, Wear, and Cades Cove, Blount and Sevier Counties, Tennessee. Journal of t he Tennessee AA cademy of Science. Vol. 10(1). pp 57

PAGE 68

15th International Congress of Speleology Biological Sciences 1329 2009 ICS Proceedings From ROM cryophily CRYOPHILY to TO troglomorphy TROGLOMORPHY : morphological MORPHOLOGICAL and AND molecular MOLECULAR e E Vidence IDENCE for FOR the THE e E Volution OLUTION towards TOWARDS troglomorphy TROGLOMORPHY in IN three THREE different DIFFERENT lineages LINEAGES of OF the THE genus GENUS Cicuri ICURI Na A Araneae RANEAE : Dictynidae ICTYNIDAE PIERRE PAq Q UIN1, N N ADINE D D Up P RR2 1Karst and EE ndangered II nertebrate RResearch LL aboratory, SWCACA EE nironmental CC onsultants, 4407 Monterey OOaks Boulevard, Building 1, Suite 110, AA ustin, Texas, 78149, UU .S.A A ppaquin@swca.com2AA merican Museum of NN atural HH istory, DD ivision of II nertebrate ZZ oology, CC entral Park West at 79th Street, NN ew York, NN ew York, 10024 UU SA A ,: nduperre@amnh.org Abstract e spider genus CC icu rina is composed of 120 species found in North America, South East Asia and one species in Europe. e genus includes an impressive radiation of 58 troglobitic species occurring mainly in Central Texas (USA). Following the early classication of the genus proposed by Chamberlin & Ivie (1940), these troglobitic species are all placed in the subgenus C C icu rina (C C icurella). Recent investigations of phylogenetic relationships within the genus support the monophyly of C C icu rina (C C icurella), and the validity of the subgenera CC icu rina (C C icurusta) and CC icurina (C C icurina). e discovery of a new troglobitic CC icu rina from Georgia (USA), and the inclusion of an obscure Mexican species in this broader phylogenetic context, showed that evolution for cave life happened at least three times within the genus: the Texas radiation with 58 species in C C icurina (C C icurella), the new species from Georgia in CC icu rina (C C icurina), and the Mexican species in CC icu rina (C C icurusta), a scenario supported with both morphology and molecules. Anity for cold conditions (cryophily) is proposed as the driving force towards evolution of troglomorphy in CC icu rina, favoring the views of Barr (1967) over Howarth (1986). Examples of cryophilic anities are given for CC icu rina. e success of the Texas radiation CC icu rina (C C icurella) with 58 species when compared to the other lineages where only one troglobitic species is known, may be partly due to the particular nature of the caves in Central Texas where the lineage evolved. Actual data suggests that karst of this area forms a complex network, with actual and past connections. We hypothesize that C C icu rina (C C icurella) evolved from one karst island to another through faulting that developed with geologic age, resulting in extreme narrow endemism of numerous species found in well dened geological clusters of caves, while such connections do not exist where troglobitic CC icu rina (C C icurusta) and CC icurina (C C icurina) are known.

PAGE 69

Biological Sciences 1330 2009 ICS P roceedings 15th International Congress of Speleology CAV V E MYOTIS MYOTIS VEL EL IFE E R IN N CAUTUS ROOST MONITORING AND P ROTECTION ON FORT HOOD ARMY INSTALLATION, TEX X AS, USA.CHARLECHARLE S E E PE E KIN IN S Wildlife Biologist, UU .S. AA rmy, Fort HH ood NN atural RResources Branch, Fort HH ood, Texas, UU SA A 76544 Suitable bat caves are rare on karst landscapes because the microclimate must meet temperature and humidity requirements for optimal metabolic regulation by roosting bats. Maternity colonies oen select caves that allow nursing females to conserve stored energy and neonatal bats to develop and mature. ese caves oen have microroosts (recessed ceilings, domes, solutioned fractures) which trap and retain heat. Because maternity caves are rare on karst landscapes and are crucial for species persistence, they must be studied, monitored, and protected from anthropogenic roost disturbance and microclimate alteration. Fort Hood is an 88,000 ha U.S. Army Installation located in central Texas. Over 250 caves and 620 sinks and rockshelters occur on the installation; many features contain karst-dependent biota, including 17 endemic, invertebrate species. Five caves are known roosts for cave myotis (Myotis velifer incautus), a Federal species of concern; one cave system is unstudied, two caves are abandoned, one cave is an alternate roost, and one cave is a warm-season maternity roost. e primary and secondary entrances to the maternity roost were protected with marginally usable bat gates. To maximize emergence ow and bat safety, the inadequate gates were replaced with cupolastyle bat gates. At the primary and secondary entrances (respectively), gate volume increased 1,200% and 920%, gate surface area increased 880% and 1,050%, and gate angle iron spacing increased to the recommended 14.6 cm. Increased gate volume, surface area, and angle iron spacing provide a safer and more ecient emergence, while ensuring roost protection from anthropogenic disturbance. From 2005 to 2008, an estimated 10,500 bats (2006) and 23,550 bats (2008) [mean = 17,200; SD = 5,300] occupied the maternity cave during the warm-seasons. Yearly population trends exhibited typical maternity cave occupation dynamics, such as gradual warm-season population increases followed by population doublings aer young bats became volant. From 2004 to 2008, cave microclimate parameters (temperature and absolute humidity) followed predictable, cyclical patterns that ranged between 10.6 C, 1.2 g/m3 (winter) and 29 C, 29 g/m3 (summer). Cave microclimate provided ecient maternity conditions, but inecient hibernaculum conditions. Cave microclimate parameters uctuated synchronously with surface parameters, suggesting surface air ow inuence. However, cave conditions were dampened with respect to surface extremes and rapid changes. Despite their availability and suitability, the two abandoned roosts remained unoccupied during the warm-seasons. It is critical to study caves exhibiting bat roost signs to determine if the roost is a maternity site. Research needs include: bat roost delineation on cave maps; population and microclimate monitoring at maternity caves; monitoring and microclimate studies at abandoned roosts; careful consideration of cave gate styles, bat emergence patterns, and cave morphology before maternity roost protection; and regular surface vegetation maintenance at gated cave entrances and abandoned roosts.1. IntroductionBats are the only group of vertebrates that have successfully exploited caves for permanent shelter (KUNZ 1982). Cave use by bats is determined by roost temperature range and stability, which ultimately inuence bat body temperature and metabolism (HOCK 1951). In warm caves, less energy is expended maintaining an optimal body temperature (Burnett and August 1981). Metabolic conservation is especially important for pregnant and lactating bats because energy expenditure costs and burdens are greatest during neo-natal growth and care (KURTA et al. 1987, SPEAKMAN and THOMAS 2003). An important factor inuencing cave microclimate parameters (temperature and humidity) is the interaction between cave morphology,

PAGE 70

15th International Congress of Speleology Biological Sciences 1331 2009 ICS Proceedings and air ow exchange and entrapment (TUTTLE and STEVENSON 1978, PALMER 2007). Air ow exchange between the cave and the surface coupled with cave morphology and micro-roosts (i.e. domes, recessed ceilings, solutioned fractures) oer a range of temperature gradients for roosting bats (TUTTLE 2000). Spatial variation in roost temperature and humidity can be important for bats (HENSHAW 1966, RAESLY and GATES 1987, RICHTER et al. 1993, BRUNET and Medelln E DELLN 2001). Prevention of air ow into or out of caves can change the temperature, thereby making a roost unsuitable to bats (RICHTER et al. 1993, CURRIE 2001). Specic roost microclimate requirements mean that suitable bat caves are severely limited on karst landscapes (TUTTLE and STEVENSON 1978, KUNZ 1982), which results in patchy distribution (HUMPHREY 1975), oen restricting large bat populations to a few caves. For example, endangered gray bats (Myotis grisescens), an obligate cave-dwelling bat, aggregate in less than 5% of available caves (TUTTLE 1979). Anthropogenic roost disturbance and alteration are the greatest threats and causes for decline for most cave-dwelling bats (MOHR 1976, McCRACKEN 1989). Human disturbance at cave roosts were related to an 89% decrease in gray bat populations in Kentucky (RABINOWITZ and TUTTLE 1980). LACKI (2000) suggests that human intrusion likely caused Raneques big-eared bats (CC orynorhinus ranesquii) to abandon their maternity roost. In Arizona, a maternity colony of cave myotis (Myotis velifer) [hereaer velifer] abandoned a roost where increased human recreational use occurred (OSHEA and VAUGHAN 1999). Large populations concentrated in a handful of caves, and bats low reproductive potential (BARCLAY and HARDER 2003), predispose them to rapid population decline should disturbance or destruction befall key roosts. Bat caves can be protected from anthropogenic disturbance with gates, which allow bat emergence yet prevent human entry. However, improperly gated caves can cause bat roost abandonment and population decline (TUTTLE 1977, CURRIE 2000, 2001; PETIT et al. 2006), and likely alter emergence ight behavior, increase the potential for collisions, and increase energy consumption associated with ight (SPANJER and FENTON 2005). Even with bat-friendly gate designs, some species may reject full gates (i.e. gate occupies entire passage/entrance), and improper gate placement can cause roost abandonment (Currie 2000). Alternative gate designs are box-like cupolas constructed over an entrance (TUTTLE and TAYLOR 1998, KRETZMANN 2002), which allow bats to stage in the gate volume before emerging. SPANJER and FENON (2005) recommend the construction of bat-friendly gates at entrances rather than within passages and on at ground rather than an incline. erefore, gate type, species acceptance, cave morphology, and surface topography should receive careful consideration before a cave is gated. Fort Hood is an 88,000 ha U.S. Army Installation located in central Texas. Over 250 caves and 620 sinks and rockshelters occur on the installation; many features contain karstdependent biota, including 17 endemic, invertebrate species. Five caves are known roosts for velifers, a Federal species of concern; one cave system is unstudied, two caves are abandoned, one cave is an alternate roost, and one cave is a warm-season maternity roost. Herein, I present results from 4 years of protection, management, and monitoring at the maternity roost and the two abandoned roosts.2. Study Site and MethodsShell Mountain Bat Cave (SMBC) is a 117 m long, 11 m deep velifer maternity roost with two entrances. Both entrances are used by bats, and both are gated with boxstyle cupola gates. e 5 paneled gates are designed so that bats may safely emerge through horizontal spaces on 4 gate panels. e h panel (gate top) is covered with expanded metal grid that is impassable by bats. ere are 10 known roosts (darkened ceilings with guano accumulation underneath) in the cave, mostly in recessed ceilings and micro-roosts. Observations suggest that vegetation growth around the cupola gates may restrict/prevent emergence ow and increase the potential to ground bats, thus increasing the likelihood of roost abandonment. To reduce these possibilities, I routinely cut herbaceous and woody vegetation to cm high around the cupola gates in a 2 m perimeter (main pit entrance) and a 1 m perimeter (crawlway entrance). Tippit Cave is a 91 m long, 17 m deep abandoned velifer roost with a single, small diameter sinkhole entrance. ere are 6 known roosts in the cave, mostly in recessed ceilings and micro-roosts. Egypt Cave is a 124 m long, 10 m deep abandoned velifer roost with two entrances. ere are two known roosts in the cave, both in recessed ceilings and micro-roosts. Once a year, I cut woody vegetation to ground level in a 7.5 m perimeter around Tippit and Egypt cave entrances. To count emerging bats at SMBC, I used 3 infra-red illuminators and a video camera with 0 lux capabilities (main pit cupola), and a thermal imager (crawlway cupola). Night vision equipment placed at a distance minimizes disturbance and does not alter bat emergence patterns

PAGE 71

Biological Sciences 1332 2009 ICS P roceedings 15th International Congress of Speleology during counts (THOMAS and LaVAL 1988, KUNZ et al. 1996). Once a month from March-October, I counted all bats emerging from the south gate face of the main cupola during a 1-minute period every 5 minutes for at least 65 minutes, which resulted in fourteen 5-minute count intervals. en, I multiplied each 1-minute count by four (accounts for all four gate surfaces) to obtain bats-emergingper-minute every 5 minutes. Concurrent with main cupola counts, I counted all bats emerging from the north gate face of the crawlway cupola during a 1-minute period every 10-minutes for at least 65 minutes, which resulted in eight 5-minute count intervals. For the 14 main cupola and 8 crawlway cupola intervals, I calculated parametric values and 95% condence intervals based on the t-distribution equation. I used dataloggers (3 at SMBC, 1 at Tippit Cave, 1 at Egypt Cave) to record surface and cave microclimate parameters (temperature and absolute humidity) every 1.5 hours for 1 year periods. 3. Results and DiscussionSMBC entrances are gated to protect the velifer maternity colony. At the crawlway entrance, we believed there was negligible contribution to nightly emergence, so a at, gridstyle gate hinged to a receiver set in a concrete footer was constructed. Grid gates are unsuitable for emergence due to the lack of gate volume and the narrow spacing of the angle irons. is gate essentially prevented emergence and likely impacted surface-to-cave air ow exchange because the footer occupied 66% of the entrance. At the main pit entrance, an A-frame tent-style gate was constructed. is gate was marginally useable by bats because: angle iron spacing (i.e. horizontal gate space through which bats y) was smaller than recommended, optimal spacing is 14.61 cm (TUTTLE and TAYLOR 1998); gate base dimensions were slightly larger than the main pit entrance dimensions, thereby limiting lateral ight space needed for cave-to-gate transition; and the tapering, A-frame design denied bats usage of the upper 1/3 and the two side faces of the gate (i.e. only 2/3s of two sides were useable (Table 1). Despite gate inadequacies, approximately 16,000 velifers roosted in SMBC for 9 years (Land 2001). We replaced the inadequate gates with bat-friendly, cupola style gates that greatly increased gate volume, gate surface area, and angle iron spacing (Table 1). Increased volume allows more bats to stage inside the gate, increased gate surface area allows greater exit opportunities, and optimal angle iron spacing allows faster, safer exits. e crawlway cupola currently provides an alternate route for emerging bats that did not exist with the grid gate. Monitoring during the past 3 years suggests that 20% of the population exited from the crawlway (PEKINS 2008). is route likely contributes to gate overcrowding reduction at the main pit cupola and further expedites overall emergence ow. Cupola gate installation positively aected emergence behavior. e tent gate forced bats to cyclically y between the cave and the gate (similar to popcorn popping), which increased the collision potential, severely restricted gate staging and emergence ow, and likely increased energy expenditure. e cupola gates allow bats to exit the cave entrance, then transition to circular staging ights in the gate space, and then rapidly exit from the gate. However, when compared to a cave with no gate where velifers can freely exit (see TWENTE 1955b, HAYWARD 1970, KUNZ 1974), SMBC cupola gates lengthen emergence time and slow emergence ow (Pekins 2008). Overall, SMBC gate replacements provide a safer and more ecient emergence, while ensuring adequate roost protection against anthropogenic disturbance. From yearly population maxima during 2005-2008, an estimated 10,500 bats (2006) to 23,550 bats (2008) [mean = 17,200; SD = 5,300] occupied SMBC (Fig. 1). Others have reported velifer populations as large as 15,000-20,000 bats (TWENTE 1955a, 1955b, HAYWARD 1970, FITCH et al. 1981). Yearly population trend exhibited     Crawlway Entrance    Main Pit Entrance  Gate parameter Grid Style Cupola Style Change (%)Tent StyleCupola StyleChange (%) Volume (m3) 0.383.88921 0.94 12.5 1,230 Surface Area (m2) 0.647.41,056 4.95 16.3 229 Useable Area (m2) 0.647.41,056 1.65 16.3 887 Angle iron space (cm)4.4514.61228 8.99 14.61 63Table 1. Gate style changes and dimension improements at two entrances of Shell Mountain Bat Cave, Fort Hood, Texas. Gates were changed to improe emergence ow and safety for cave myotis (Myotis velifer incautus). Main pit gate was changed in 2004, crawlway gate in 2006. Angle iron space is the horizontal space through which bats exit.

PAGE 72

15th International Congress of Speleology Biological Sciences 1333 2009 ICS Proceedings typical maternal cave dynamics. e pattern followed a predictable trend of early-spring arrival of migratory bats, then late-spring occupation by parturient adults, then birth and ight of young mixed with post-lactating adults (summer peak population), then late-summer local dispersal/colony break-up, and nally autumn bat departure; the cave was unoccupied in winter. Others have observed the same generalized pattern at velifer maternal sites (DUNNIGAN and FITCH 1967, HAYWARD 1970, KUNZ 1974) and long-ngered bat (Myotis capaccinii) maternity sites (Papadatou et al. 2008). A caveat of my estimate is that it reects relative abundance, which is an approximation of the true population at best. THOMAS and LaVAL (1988) suggest that emerging bats can be counted in 1or 5-minute intervals, assuming emergence ow is constant. During all counts, SMBC emergence ow rapidly increased and peaked during the rst 30 minutes, then gradually tapered during the remaining 35 minutes. Despite this pattern, ow was still constant enough to reliably estimate bats-emerging-per-minute every 5 minutes. Velifers oen emerge in distinct group bursts (i.e. adult and juvenile) that may blur together during emergence (KUNZ 1974). is behavior may contribute to count inaccuracy and unreliability if count interval spacing is too far apart (miss the bursts) or if bursts overwhelm an observer (THOMAS and LaVAL 1988). SMBC gates pool the bats into one area (cupola gate space) for short time periods, thereby eliminating the burst eect and further reducing the possibility of count inaccuracy. From 2004-2008, SMBC microclimate parameters followed predictable, cyclical patterns that ranged 10.6 C, 1.2 g/m3 (winter) and 29 C, 29 g/m3 (summer), suggesting the roost was suciently warm and moist for neonatal bats and nursing adults, but too warm for velifer hibernation. Others have reported velifer maternal cave temperatures ranging from 15 C to 35 C (TWENTE 1955b, KUNZ 1973, HAYWARD 1970). During the warm seasons of Figure 1. Monthly population estimates (mean 95% limits) and yearly occupation trends for cave myotis (Myotis velifer incautus) roosting in Shell Mountain Bat Cave, Fort Hood, Texas during 2005-2008. For each year, x-axis begins with March and ends with October. e cave was unoccupied by cave myotis om Noember to February. Closed circles = main pit cupola; open circles = crawlway cupola. Count was not conducted during May 2005. Crawlway counts began in 2006. Y-axis is variable amongst years.

PAGE 73

Biological Sciences 1334 2009 ICS P roceedings 15th International Congress of Speleology 20042008, monthly temperatures at four non-bat caves and 1 alternate velifer roost ranged 12 C to 28 C, suggesting velifers may select SMBC for additional reasons ( e.g. cave morphology or ceiling topography). In general, cave microclimate parameters uctuated synchronously with surface parameters, suggesting surface air ow inuence. However, cave conditions were dampened with respect to surface extremes and rapid changes, and were generally warmer than surface in the winter and cooler than surface in the summer. During the warm seasons of 20042008, Tippit Cave and Egypt Cave temperatures ranged 19 C to 20 C, and 14.5 C to 18 C, respectively. Despite their availability and suitability, these two roosts remained unoccupied by velifers. 4. Conclusions and Research NeedsIt is critical to study caves exhibiting bat roost signs to determine: if the roost is occupied or abandoned, if the roost is a maternity site, what species occupy the site, and threats to the site. Monthly and seasonal population trends, emergence patterns, and microclimate trends should be documented at maternity sites during a minimum 5 year period so patterns can be established. Abandoned bat caves and alternate roost sites should be identied, studied, monitored, and if necessary, restored. Before a maternity site or hibernaculum is gated, careful consideration should be given to cave gate styles, bat emergence patterns, microclimate alteration, and cave morphology. I recommend regular surface vegetation maintenance at gated caves to ensure a safe emergence. To better assess where bats are roosting and to what extent, I recommend that bat roost area delineation be incorporated onto cave maps.AcknowledgmentsM. Warton, J. Fant and J. Reddell for building the cupola gates, J. Kennedy for direction with bat cave management and monitoring, D. Cimprich for preparing the graphs, and anonymous reviewers who greatly improved this manuscript. ReferencesBarclay ARCLAY R.M.R. and L.D. HARDER (2003) Life histories of bats: life in the slow lane. Chap. 5 in Bat E E cology, omas H. Kunz and M. Brock Fenton (Eds.), University of Chicago Press, Chicago, Illinois, USA, pp 209. BURNET, A.K. and R.A. MEDELLIN (2001) e speciesarea relationship in bat assemblages of tropical caves. Journal of Mammalogy 82(4), 1114. BURNET, C.D., and P.V. AUGUST (1981) Time and energy budgets for day roosting in a maternity colony of Myotis lucifugus. Journal of Mammalogy 62, 758. CURRIE, R.R. (2000) An overview of the response of bats to protective eorts. P roceedings of Bat CC onservation and Mining: AA Technical II nteractive Forum. U.S. Department of Interior, Oce of Surface Mining, Alton, Illinois, USA, pp 173. CURRIE, R.R. (2001) Response of the Indiana bat (Myotis sodalis) to gates. Bat RResearch NN ews 42(2), 29. Dunnigan U NNIGAN P.B. and J.H. FITCH (1967) Seasonal movements and population uctuations of the cave bat (Myotis velifer) in south-central Kansas. T ransactions of the Kansas AA cademy of Science 70(2), 210. FITCH, J.H., K.A. SHUMP Jr., and A.U. SHUMP (1981) Myotis velifer. Mammalian Species 149, 1. e American Society of Mammalogists, Lawrence, Kansas, USA. HAYWARD, B.J. (1970) e natural history of the cave bat Myotis velifer. RR es. Sci., Western NN ew Mexico U U niversity 1, 1. HENSHAW, R.E. (1966) Microclimate selection within the cave by bats. NN a tional Speleological Society Bulletin 28(2), 99. HUMPHREY, S.R. (1975) Nursery roosts and community diversity of nearctic bats. Journal of Mammalogy 56(2), 321. HOCK, R.J. (1951) e metabolic rates and body temperatures of bats. Biological Bulletin 101, 289 299. Kretzmann R ETZMANN J.A. (2002) Bat cupola design considerations. P roceedings of Bat GG ate DDesign: AA Technical II nteractive Forum. U.S. Department of Interior, Oce of Surface Mining, Alton, Illinois and Coal Research Center, Southern Illinois University, Carbondale, Illinois, USA, pp 207. KUNZ, T.H. (1973) Population studies of the cave bat (Myotis velifer): reproduction, growth, and development. OO ccasional Papers Museum of NN atural H H istory UU niversity of Kansas 15, 1.

PAGE 74

15th International Congress of Speleology Biological Sciences 1335 2009 ICS Proceedings KUNZ, T.H. (1974) Feeding ecology of a temperate insectivorous bat Myotis velifer. EE cology 55(4), 693. KUNZ, T.H. (1982) Roosting ecology of bats. Chap. 1 in EE cology of Bats, omas H. Kunz (Ed.), Plenum Publishing Corp., New York, New York, USA, pp 1. KUNZ, T.H., D.W. THOMAS, G.C. RICHARDS, C.R. TIDEMANN, E.D. PIERSON, and P.A. RACEY (1996) Observational techniques for bats. Chap. 7 in Measuring and Monitoring Biological DD iversity, Standard Methods for Mammals, WILSON, D. E., J. NICHOLS, R. RUDRIN, R. COLE, and M. FOSTER (Eds.), Smithsonian Institution Press, Washington, D. C., USA, pp 105. KURTA, A., K.A. JOHNSON, and T.H. KUNZ (1987) Oxygen consumption and body temperature of female little brown bats (Myotis lucifugus) under simulated roost conditions. P hysiological ZZ oology 60(4), 386. LACKI, M.J. (2000) Eect of trail users at a maternity roost of Ranesques big-eared bats. J ournal of CC ave and Karst Studies 62(3), 163. LAND, T.A. (2001) Population size and contaminant exposure of bats using caves of Fort Hood Military Base. Unpublished Master of Science thesis, Texas A & M University, College Station, Texas, USA, 103 pp. McCRACKEN, G.F. (1989) Cave conservation: special problems of bats. NN a tional Speleological Society Bulletin 51, 49. MOHR, C. (1976) e protection of threatened cave bats. P roceedings of the 1976 NN ational CC ave Management Symposium, Albuquerque, New Mexico, USA, p. 57-. OSHEA, T.J., and T.A. VAUGHAN (1999) Population changes in bats from central Arizona: 1972 and 1997. Southwestern NN aturalist 44(4), 495. Papadatou A PADATOU E., R.K. BUTLIN, and J.D. Altringham L TRINGHAM (2008) Seasonal roosting habits and population structure of the long-ngered bat (Myotis capaccinii) in Greece. Journal of Mammalogy 89(2), 503. PALMER, A.N. (2007) Cave meteorology and internal weathering. Chap. 12 in CC a ve GG eology, R. A. WATSON (Ed.), Cave Books Publishing, Dayton, Ohio, USA, pp 324. PEKINS, C.E. (2008) Cave myotis (Myotis velifer incautus) population and roost microclimate monitoring at a maternity site on Fort Hood, Texas during 2008. Endangered species monitoring and management at Fort Hood, Texas: 2008 annual report, e Nature Conservancy, Fort Hood Project, Fort Hood, Texas, USA, 66 pp. PETIT, S., A. RPJER, and L. PORS (2006) Surveying bats for conservation: the status of cave-dwelling bats on Curacao from 1993 to 2003. AA n imal CC onservation 9, 201. Rabinowitz A BINOWITZ A., and M.D. TUTTLE (1980) Status of summer colonies of the endangered gray bat in Kentucky. Journal of Wildlife Management 44(4), 955. RAESLY, R.L. and J.E. GATES (1987) Winter habitat selection by north temperate cave bats. AA m erican Midland NN aturalist 118(1), 15. RICHTER, A.S., S.R. HUMPHREY, J.B. COPE, and V. BRACK JR. (1993) Modied cave entrances: thermal eect on body mass and resulting decline of endangered Indiana bats (Myotis sodalis). C C o nservation Biology 7(2), 407. SPEAKMAN, J.R. and D.W. THOMAS (2003) Physiological ecology and energetics of bats. Chap. 10 in Bat EEcology, T.H. KUNZ and M. BROCK FENTON (Eds.), University of Chicago Press, Chicago, Illinois, USA, pp 430. SPANJER, G.R., and M.B. FENTON (2005) Behavioral responses of bats to gates at caves and mines. Wildlife Society Bulletin 33(3), 1101. THOMAS, D.W. and R.K. LaVAL (1988) Survey and census methods. Chap. 5 in EE cological and Behavioral Methods for the Study of Bats, T. H. KUNZ (Ed.), Smithsonian Institution Press, Washington D. C., USA, pp 77. TUTTLE, M.D. (1977) Gating as a means of protecting cave dwelling bats. P roceedings of the 1976 NN ational

PAGE 75

Biological Sciences 1336 2009 ICS P roceedings 15th International Congress of Speleology CC ave Management Symposium, Albuquerque, New Mexico, USA, pp 77. TUTTLE, M.D. and D.E. STEVENSON (1978) Variation in the cave environment and its biological implications. NN a tional CC ave Management Symposium Proceedings, Albuquerque, New Mexico, USA, pp 108. TUTTLE, M.D. (1979) Status, causes of decline, and management of endangered gray bats. Journal of Wildlife Management 43, 1. TUTTLE, M.D., and D.A.R. TAYLOR (1998) Bats and mines. Bat Conservation International, Inc., Austin, Texas, USA, 50 pp. TUTTLE, M.D.. (2000) Where the bats are part III: clis, caves, and rock crevices. Bats 18(1), 6. TWENTE, J.W. Jr (1955a) Aspects of a population study of cavern dwelling bats. Journal of Mammalogy 36(3), 379. TWENTE, J.W. Jr (1955b) Some aspects of habitat selection and other behavior of cavern dwelling bats. E E cology 36(4), 706.

PAGE 76

15th International Congress of Speleology Biological Sciences 1337 2009 ICS Proceedings NEW STUDIES OF SPEoO PLatyrhi ATYRHI Nus US Pou OU Lso SO Ni I PISCES: AMBLYOPSIDAETHO HO MA A S L L POUL OUL SON ON 318 Marlberry CC ircle, Jupiter, FLL 33458-2850 UU SA A tomandliz@bellsouth.net In my comparative studies of the sh family Amblyopsidae (1961, 1963), I predicted what a more cave-adapted Amblyopsid would look like based on changes with increasing time of isolation in caves (decreasing eye size and complexity and decreasing density and complexity of pigmented melanophores). With increasing time of isolation (Typhlichthys subterraneus to AA mblyopsis spelaea to AA mblyopsis rosae) each cave species has strong convergence in troglomorphy with slight changes in relative head size (+), free lateral line exposure and numbers (+), sense organ integrative parts of the brain (+), caudal and pectoral n length (+) and weight per unit length (-). I suggested that these trends are neotenic. In 1967 I received two black and white photos of an apparently pigmentless and eyeless cavesh with the cryptic note Is this what you predicted? Initially, the sh was identied as a 10-12mm Typhlichthys based on its peculiar spatulate head shape, large head, relatively thin body, and unbranched n rays. In fact, it was the 58mm holotype of what its discoverer John Cooper and his colleague Bob Kuehne described as Speoplatyrhinus poulsoni (1974). uantication of the shes brain morphology, free neuromast densities, pigmented melanophores, and gill morphology as an index of metabolic rate indicated more extreme neoteny than in any other Amblyopsid. It is found in only one cave in northwestern Alabama and is protected by the Key Cave National Wildlife Refuge. It is listed as critically endangered. On November 5, 2008, I got to see my namesake in nature and study specimens in an aquarium for several hours while Dante Fenolio took a series of fantastic color photos. e sh were returned aer three hours. I studied three sh (42, 45, and 62mm SL) in a 25 liter aquarium with water temperature 15C and depth of 15cm. ere were two ~ 10 x 10 x 6cm rocks. e sh showed no change in behavior when with 1, 2, or 3 individuals or with a high density of small active amphipods and sedentary isopods. ey caught no prey even when the prey touched a snout and showed no change in behavior when near or contacting another sh. ey swam with short partial pectoral-caudal strokes, instead of a leisurely stroke and glide. ey oen repeatedly bumped gently into the side of the aquarium or nosed under the edge of a rock. ey were easy to pick up by hand and returned to normal swimming rates and ventilation frequency and depth within 1 minutes. ey oen stopped swimming and oated to the surface before swimming back down. ey explored all parts of the water volume and were continuously active.1. IntroductionA comprehensive review of published, unpublished, and new data on the North American freshwater sh family Amblyopsidae is given by NIEMILLER and POULSON (2009) as a chapter in a book on cave shes dedicated to Poulson for his many, long, and continued contributions. Here I consider Speoplatyrhinus poulsoni in even more detail with particular attention to my recent observations of my namesake in Key Cave and in an aquarium just outside the cave. My predictions for the next stage in Amblyopsid troglomorphy (POULSON 1961, 1963) were based on trends seen with increasing evolutionary time in caves for the then known troglobitic species. Estimates of evolutionary time in caves were based on EIGENMANS classic studies of eye histology and my own studies of melanophore pigment cells. My rationale, which has better support today (POULSON 1985 and NIEMILLER and POULSON 2009), was that mutations of simplication and loss are near neutral aer the early stages of isolation in caves. In the rest of this paper, I will describe what we now know about which features of Speoplatyrhinus poulsoni anatomy are shared with and dierent from other troglobitic Amblyopsids. en I will tie the anatomy to my recent observations of sh in an aquarium. Next, I will consider alternative hypotheses to explain the progressive restriction of geographic range of troglobitic Amblyopsids ending

PAGE 77

Biological Sciences 1338 2009 ICS P roceedings 15th International Congress of Speleology in the relict distribution of my namesake. Finally I will consider what size-frequency distributions and scale marks for aging suggest about life history.2. Convergent Troglomorphic Anatomy Based on partial data, the eyes and pigment of Speoplatyrhinus poulsoni (Sp) are the most reduced of the troglobitic Amblyopsid species. Here I compare it to A A m blyopsis rosae (Ar) which has the most reduction of eyes and pigment of the other Amblyopsids. ough we have no data on my namesakes eye histology, the optic lobe volume is more reduced. For a 45 m SL sh the mm3 volume for Sp is 0.38 < 0.96 for Ar. is suggests that the eye is smaller with more parts lost and indistinguishable. It also suggests that the optic tectum visual tracts have not been co-opted by tactile tracts via negative pleiotropy. e density of pigmented melanophores of Sp is 0.6 per .175mm2 on the side of the body compared to 0.7 for Ar. e average melanophore size of .04mm2 x 10-3 is about half that for Ar. And unlike Ar the pigmented melanophores are completely lost in individuals greater than 45mm SL so the negative allometry for pigment cell number must be even lower than the b of -2.4 for Ar. e most important non-visual sensory system in Amblyopsids is the lateral line which is greatly hypertrophied in all the troglobitic species. e surcial ridges (aka stitches) of neuromasts are necessary and sucient to explain avoidance of obstacles and capture of living prey. Each ridge has 7 34 neuromast cups each with a gelatinous cupula. e height above the skin of neuromast cup plus cuopula is 0.8 1.1mm for all species. e ridges are concentrated on the head and most adjacent ridges are oriented at right angles to each other. Among the troglobitic Amblyopsids, Speoplatyrhinus poulsoni has about 25 percent more total neuromasts partly because many of the ridges have more neuromasts than in other species but mainly because Sp projected head area is so large relative to projected body area, i.e., 1.73 compared to 0.67 0.73 for the three other troglobites (NIEMILLER and POULSON 2009: Figure 5). e brain areas for primary lateral line sensory input, the cristae cerebelli and eminentia granularii, and for integration of all sensory information, the forebrain, show relatively the same degree of hypertrophy in Sp as in the other troglobites. Despite an elaborated lateral line system none of the three Speoplatyrhinus observed in an aquarium for about 2 hours caught any of the approximately 40 amphipods and 15 isopods introduced to their aquarium. Surprisingly they showed no increase in swimming speed or changes in search behavior. ey did not even react when they encountered a moving amphipod or when a fast swimming amphipod touched their head. Perhaps they were not hungry. ey certainly seemed to have full stomachs based on the rounded belly. My past lab studies of other Amblyopsid troglobites, that showed striking changes in behavior and successful capture of living prey in the laboratory, always involved sh that had not been fed for several days. And once sated the sh returned to the behavior seen before prey were introduced. Surprisingly, the brain areas that are responsible for input and integration of equilibrium information in Speoplatyrhinus poulsoni are not as hypertrophied as in the other Amblyopsid troglobites (Figure 11 in NEIMILLER and POULSON 2009). e cerebellum is smaller than for any other troglobites and not much bigger than in the troglophile Forbesichthys agassizi.. e semicircular canals and otoliths are larger than in Forbesichthys but smaller than in Typhlichthys and AA mblyopsis. I saw no dierences in swimming behavior which might be correlates of these dierences. My only hypothesis is that the narrow and small frontal area of Sp head, relative to the blunt and broad shape of the other species, may not require as large a cerebellum and semicircular canal system (see rationale for this in POULSON 1963). Sps thin and spathulate head shape can displace much water as it approaches an obstacle and this may explain why all three individuals observed repeatedly nosed into rocks or corners without any obvious avoidance behavior. Unlike all other species, Speoplatyrhinus swam using short partial caudal and pectoral n strokes and so moved very slowly. Despite having as large or larger pectoral n and caudal n area than other species, they did not show leisurely coordinated caudal pectoral strong stroke and glide swimming expected of approximately one body length moved per stroke. I observed some body positions of swimming and exploring Speoplatyrhinus that are correlated with its spathulate snout. e sh oriented their bodies so that the bottom of the head was parallel to and with maximum contact with the substrate. As a result the body angled upwards at about 30 degrees. Another unusual behavior seen in all three individuals was that they oen stopped swimming and oated to the water surface. Once at the surface they oen rested for a few seconds with their back slightly out of the water. en they leisurely swam back to the bottom and continued exploring the aquarium.

PAGE 78

15th International Congress of Speleology Biological Sciences 1339 2009 ICS Proceedings As with all the Amblyopsids, Speoplatyrhinus poulsoni has a moderately developed external olfactory intake tube and olfactory rosette and internal brain olfactory lobe and similar development of the caudal papillae system, which may be unique to the family. Sp has fewer but much larger caudal papillae than Ts (Figure 12 in NIEMILLER and POULSON 2009). We have no clue as to the function of this sensory system.3. Phylogeny Based on Anatomy A major diculty is to separate convergent cave-dependent characters of troglomorphy from cave independent autapomorphies that can be used to determine phylogeny. ARMBUSTER (personal communication) is developing a phylogeny of the Amblyopsidae based on skeletal and other anatomical characters. e following are what I judge to be clear autapomorphies of Speoplatyrhinus poulsoni. 1. e absence of presumed tactile sense organs seen on the head in all other Amblyopsid troglobites (NIEMILLER and POULSON 2009: Figure 8). 2. e presence of short lines of thin rod like (~ 0.5mm) presumed sense organs on the head (thinner than cupulae Figure 1 this paper). 3. e dierence in relative sizes of otoliths with the sagitta smaller and utriculus larger than in any other Amblyopsid (Figure 11 in NIEMILLER and POULSON 2009). 4. e very large caudal papillae (Figure 12 in NIEMILLER and POULSON 2009). 5. Possibly the extremely thin and spathulate snout (Fig. 1). Alternative hypotheses are that it is a spandrel with no adaptive function (sensu LEWONTIN and LEVINS) related to reduction of skeletal structures around the eye orbit or it is a spandrel related to extreme neoteny. Part of my basis for predicting trends in troglomorphy with increasing evolutionary time was the clear neotenic trend in the three Amblyopsid troglobites (POULSON 1961, 1965, 1985, NIEMILLER and POULSON 2009). Clearly Speoplatyrhinus poulsoni is the most neotenic of all Amblyopsids as recognized in the description of the species by COOPER and KUEHNE (1974). 1. e absence of bifurcate n rays. 2. e very large head relative to body size. 3. Absence of a supraopercular papilla opening to the enclosed head lateral line system. 4. e possibly further reduction in calcication of the skeleton suggested by poor resolution radiographs and poor alizarine staining of cleared specimens. 5. e apparently translucent appearance of even large individuals. When histological studies of the eyes are done we may nd that the parts that are lost or simplied are dierent than in other species of Amblyopsids. For example eye muscles, scleral cartilages, and pigmented epithelium are missing in Typhlichthys but vestiges are present in both species of Amblyopsis even though their eyes are more simplied overall (Table 2 in NIEMILLER and POULSON 2009).4. Biogeography Speoplatyrhinus poulsoni shows an extreme relict distribution. It shows the extreme of decreasing geographic range and number of caves occupied among the four species of troglobitic Amblyopisds (Figure 3 in NIEMILLER and POULSON 2009). Even though number of sites and caves is a much better measure of ability of the sh to move underground, total counties with any record gives an index of geographic range. Records of sh exist from 72 counties in 4 states along the eastern US coastal plain for C C h ologaster cornuta 32 counties in eastern Missouri, southern Illinois, Kentucky, and Tennessee for Forbesichthys agassizi 54 counties in seven states for Typhlichthys subterraneus ( central Missouri and northern Arkansas and central Kenttucky and Tennessee and northern Alabama), ten counties in southern Indiana and northern Kentucky for A A m blyopsis spelaea nine counties in southwestern Missouri and adjacent areas of Arkansas and Oklahoma for A A m blyopsis rosae and one county in northwestern Alabama for Speoplatyrhinus poulsoni One hypothesis to explain current distributions of Amblyopsid cave species is that it reects the past distributions of their surface ancestors. e Ozark vs. Figure 1: Photos of 42 mm SL Speoplatyrhinus poulsoni by Dant Fenolio.

PAGE 79

Biological Sciences 1340 2009 ICS P roceedings 15th International Congress of Speleology Interior Lowland Province distribution of the two A A mblyopsis species is one example. In surface teleost sh there are several species groups of darters (EE theostoma) with this distribution. It is also known for salamanders, with both the EE u rycea lucifuga complex and the EE u rycea longicauda complex found both west and east of the Misiisippi. Plethodon dorsalis and Plethodon angusticlavius salamander distributions show the same dichotomy. But I know of no surface sh or salamanders that have geographic distributions matching those of either Forbesichthys agassizi or Typhlichthys subterraneus. And there are no freshwater sh, salamanders, craysh, or shrimp in the range of the Amblyopsidae that show the extreme relict distribution of Speoplatyrhinus poulsoni (there are both salamanders and sh restricted to only one or a few surface springs along the Balcones Escarpment in Texas). A second hypothesis of competitive displacement has been proposed (e.g. WOODS and INGER in NIEMILLER and POULSON 2009) to explain why Typhlichthys has such a wide and central geographic distribution compared to either AA mblyopsis species and why Speoplatyrhinus has such a narrow and peripheral distribution compared to Typhlichthys (e.g. NIEMILLER personal communication). Competition can be of two types where indirect is less eective than direct. Indirect competition involves a dierential eciency of exploiting a food resource that is not defensible. Direct competition involves direct harm during agonistic interactions at a defensible resource. Despite some evidence for density dependent population in one population of AA m blyopsis spelaea, I now argue that neither kind of competition is likely to contribute to the dierences in geographic range just described. Indirect competition is unlikely to result in competitive exclusion among troglobitic Amblyopsids because the species show only slight dierences in lateral line hypertrophy, eciency of nding dierent live foods, and metabolic rates (Tables 4, 6, 8, and 9 NIEMILLER and POULSON 2009). In addition, in the few cases where putative competitor species are syntopic there is no indication that either species is doing more poorly, as judged by gut fullness, or reproduction, judged from sizefrequency distributions. e cases are of Typhlichthys and A A m blyopsis spelaea in Mammoth Cave and Typhlichthys and Speoplatyrhinus in Key Cave. In Mammoth the two species separate along a stream gradient from upstream high food (Typhlichthys) to downstream low food (AA m blyopsis). is is the wrong direction for purported exclusion of AA m blyopsis by Typhlichthys. In Key Cave one large Typhlichthys has been seen for 15 years (growing from ~ 50mm to a measured 82mm SL) along with as many as 10 Speoplatyrhinus on any one trip. Individuals of both species appear healthy with no signs of emaciation. Again these data, especially the ~ 0.5mm per year growth of the large Typhlichthys, are in the wrong direction to support a prediction that Typhlichthys is the better indirect competitor. Direct competition is unlikely to result in competitive exclusion among troglobitic Amblyopsids because the species have little or no agonistic intraspecic interactions and there is no defensible resource in their cave habitats. In dyadic interactions with resident vs. intruder in small aquaria Bechler found much lower frequency, diversity, and intensity of agonistic acts in troglobitic Typhlichthys and AA mblyopsis spelaea compared to the troglophilic Forbesichthys agassizii (Table 7 in NIEMILLER and POULSON 2009). AA m blyopsis rosae showed almost no agonistic behavior. And in a small volume aquarium three Speoplatyrhinus (42, 45, and 62mm SL) showed no evidence that they even recognized each other as a sh when they were introduced to the aquarium sequentially and frequently contacted each other. In a large articial stream used by Bechler the agonistic behavior was almost absent even in the species that showed occasional strong interactions in a small aquarium. And nobody has ever seen agonist behavior in a cave. In caves I can conceive of no defensible food resource and even deep, quiet water hiding places during oods are abundant and not defensible. e only mode of possible competitive exclusion that I can hypothesize is an indirect demographic swamping. Demographic swamping could occur if one species lifespan was shorter and its reproductive rate was much higher than for a competitor even if that competitor was a more ecient feeder with a lower metabolic rate. At present the size-frequency distributions among the Amblyopsid troglobites are similar and so no species could have a reproductive advantage. Of course, in the past there could have been exclusion by demographic swamping if a species newly isolated in caves had a shorter life span and higher reproductive rate than a species that had been isolated at an earlier time. is seems unlikely since the climatic event most likely to have resulted in isolation would be interglacial drying when the organic matter washing into or seeping into caves would be decreasing.5. Metabolic Economies and Behavior When DisturbedMy observations of Speoplatyrhinus in an aquarium show almost continuous activity despite indirect evidenc for a very low metabolic rate. Among the other troglobitic

PAGE 80

15th International Congress of Speleology Biological Sciences 1341 2009 ICS Proceedings Amblyopsids there is a trend for decreasing metabolic rates but increasing percent of the day active and greater distance swum per day (Table 9 NIEMILLER and POULSON 2009). e three Speoplatyrhinus that I observed for ~ 2 hours were continuously active and explored all parts of the aquarium. Given the good correlation of gill surface area with metabolic rate in the other species, I infer that Speoplatyrhinus has about 40 percent lower metabolic rate than AA mblyopsis rosae which has the lowest measured metabolic rate of the other three troglobites (Figure 13 in NIEMILLER and POULSON 2009). e barely visible amplitude and ~1 per two second frequency of its opercular ventilation movements are consistent with a very low metabolic rate. As with other troglobitic Amblyopsids, Speoplatyrhinus was easy to pick up by hand, scarcely struggled when held, and when returned to the aquarium showed no excited swimming and returned within three minutes to imperceptible ventilation. In contrast, the 82mm SL Typhlichthys studied at Key Cave wiggled continuously when picked up and showed clearly faster swimming when returned to the water.6. Life History and Inferred Growth Rates and LongevitiesAll the indirect data suggest an intermediate lifespan and growth rate for Speoplatyrhinus compared to other Amblyopsid troglobites. e putative marks of decreased growth rates on the scales look the same as in all Amblyopsids with the following growth checks and standard lengths from the type series: 33mm = 3 checks, 42mm = 4 checks, 47mm = 5 checks, and 52 mm = 6 checks. ese indicate an age of the 52mm sh as 18 24 years based on the relation of scale checks to actual growth rates determined from mark-recapture data in the other troglobites. In the troglophile Forbesichthys the scale checks match the mark recapture growth rates with the known interruption of growth seasonally and a 2 and rarely 3 year lifespan. To explain the discrepancy between ages estimated by scale growth checks and actual measured growth rates, Bechler has suggested to me that cavesh may not even grow some years and so not even lay down any circuli on scales. is ts with mark recapture data that show that as many as 20 30 percent of recaptured sh and more of larger individuals sh show << 1mm body length growth per year or even ~ 1mm negative growth (BROWN in NIEMILLER and POULSON 2009 for AA m blyopsis rosae and PEARSON personal communication for AA m blyopsis spelaea and Typhlichthys). It does not t with the fact that I have never found stomachs of preserved cavesh empty; >= 95 percent of the frequency and 70 95 percent of the volume of prey are copepods. One thought is that the marks on scales represent occasional large meals rather than times with no or few prey eaten. is is an extension of my RARE lucky large prey captured hypothesis (NIEMILLER and POULSON 2009). e size frequency distributions of the surface and springcave species of Amblyopsids show clear cohorts associated with 2-3 year classes but clear size cohorts are never seen for the cave species (Figure 4 POULSON 1961 from preserved collections). In AA m blyopsis spelaea, the only species with veried branchial incubation, there are incubating females or rst year young 8mm SL found only in about 10 percent of many trips over 40 years to the two caves with the largest populations in streams where all the habitat is accessible for census. In caves with the largest A A m blyopsis rosae populations there are many hiding places among rocks for small sh but the sh 20-30mm are found all the time and sh 12 15mm are found on perhaps a quarter of trips. I have inferred from these data that AA m blyopsis rosae has the most frequent reproduction of any Amblyopsid troglobite. is is associated with the lowest metabolic rate, fastest growth rates, lowest cost for reproduction, and shortest inferred longevity of any cave species (Table 5 in NIEMILLER and POULSON 2009). In contrast Typhlichthys has only one population, Shelta Cave in Alabama, in which we routinely found small 11-21mm SL sh. And, in contrast to the case for CC hologaster cornuta and Forbesichthys agassizi, we never see preserved or live individuals with bellies obviously swollen with eggs even in AA m blyopsis spelaea with known clutch sizes of 50 70 in 60 80 mm SL females. Putting all of this together I hypothesize that, except for AA m blyopsis spelaea, clutch sizes of troglobitic Amblyopsids are exible and small, perhaps as few as 10. Along with iteroparity and long life this would minimize the cost of reproduction and maximize spreading of the risk of reproductive failure. e data on Speoplatyrhinus poulsoni from Key Cave are consistent with a very low level but regular level of reproduction. e type series has individuals of the following mm SL; 32, 33.5, 34, 38.5, 42, 47, 49, and 58.3. e most accurately estimated and measured sizes in nature are for October and November 2008 mm SL of 15, 30, 30, 39, 41, 42, 48, 52, and 62. e 62 mm sh is the largest known and only one sh was estimated to be smaller than 15 mm in the past (12mm). e cave is dicult to census with water depths approaching 10m where the habitat is accessible in only a few isolated passage segments that are dicult to access. And only a few of these areas are large enough to snorkel.

PAGE 81

Biological Sciences 1342 2009 ICS P roceedings 15th International Congress of Speleology Based on observations in the eld Key Cave has a low to moderate food supply for Speoplatyrhinus and the two species of cave craysh. e water is usually crystal clear and there are scattered small rocks on the bottom that could be hiding places for the one cave amphipod and one cave isopod known from the cave. e area of seasonally deposited gray bat guano next to one pool never has obviously higher numbers of sh or craysh than other areas and no visible isopods or amphipods. On any one trip the maximum number of sh seen has been 10 with about twice that many craysh. Baited traps yield occasional craysh but not amphipods or isopods. If we are ever allowed to examine the stomachs and intestines of the type series I expect to nd only copepods.7. ConclusionsSpeoplatyrhinus poulsoni has been isolated in caves longer than any of the other three troglobitic species but has only slightly greater elaboration of the lateral line sensory system. It shows more extreme neoteny than any other species. It has several autapomorphic anatomical features which suggest that it is very distantly related to the other troglobites But, except for its lack of stroke and glide swimming, its other behavioral and physiological traits are similar to and convergent with the other species. ese include diminished reaction to disturbance, low metabolic rate, moderately long longevity, low growth rate, and very low frequency of but regular occurrence of reproduction.AcknowledgementsI especially thank Bernie Kuhajda, Brook Fluker, and Shane Stacy for catching the sh and thanks to Dante Fenolio for his extraordinary photographs. I thank John Cooper for naming the sh in my honor and allowing me to help with its description. And I thank Matt Niemiller for continuing discussion of my ideas. ReferencesCOOPER, J.E., and R.A. KUEHNE (1974) Speoplatyrhinus poulsoni, a new genus and species of subterranean sh from Alabama. CC o peia 1974 486. KUHAJDA, B.R. and R.L. MAYDEN (2001) Status of the federally endangered Alabama cavesh, Speoplatyrhinus poulsoni (Amblyopsidae), in Key Cave and surrounding caves, Alabama. E E n ironmental Biology of Fishes 62, 215. Niemiller, M.L. and T.L. Poulson (in press) Studies of the Amblyopsidae: Past, Present, and Future. Chap. in e Biology of Subterranean Fishes, Eleonora Trajano ,M Bichuette, and B. Kapoor (Eds.), Science Publishers, Eneld. Poulson, T.L. (1960) Cave adaptation in Amblyopsid shes. PhD Dissertation, Department of Zoology, University of Michigan, Ann Arbor, Michigan. U U n iversity Microlms 61-2787 POULSON, T.L. (1963) Cave adaptation in Amblyopsid shes. AA merican Midland NN aturalist 70, 257. POULSON, T.L. (1985) Evolutionary reduction by neutral mutations: Plausibility arguments and data from Amblyopsid shes and Linyphiid spiders. Bulletin of t he NN ational Speleological Society 47, 109.

PAGE 82

15th International Congress of Speleology Biological Sciences 1343 2009 ICS Proceedings FIELD METRICS FOR CAV V E STREAM BIOINTEGRITYTHO HO MA A S L L POUL OUL SON ON 318 Marlberry CC ircle, Jupiter, FLL 33458-2850 I build on the concept of Indices of Biological Integrity (IBI) successfully used for sh and macroinvertebrates in surface streams. Unlike grab samples for assay of microbes and pesticides and water chemistry, biological communities have a memory so one does not have to sample at the time of discharge (smoking gun) or at the front end of a ood hydrograph to hypothesize why a streams IBI is compromised. Also, biological censuses are much less expensive. To develop an IBI for cave streams one needs to use all of the fauna. I suggest: 1. Counting and measuring and assessing health of all individuals of sh, salamanders, and craysh; 2. Random timed and dened areal samples for isopods, amphipods, atworms and stream-bank terrestrial fauna at the generic level; 3. Baited traps with rock and leaf litter refuges to assess copepods, isopods, amphipods, and atworms that can otherwise only be reliably censused by getting frequency and abundance data picking up rocks; 4. Assess microbial biolm by slipperiness on rocks; 5. Estimate sizes and abundances of particulate organic matter; and estimate the texture of stream bottom / bank inorganic substrate (clay, silt, sand, gravel, and rocks). One overall metric of integrity useful across streams and regions is a graph of log abundance vs. species rank (most to least abundant). High dominance of the highest ranked species and great rarity of the lowest rank species suggests that the stream is compromised. I hypothesize dierent community signatures for siltation, organic enrichment, high level acute toxicity, and low level chronic toxicity. Silt homogenizes microhabitat, organic enrichment favors trogloxenes and troglophiles, acute toxicity dierentially aects the fastest growing species, and chronic toxicity harms the top predators by biomagnication and bioaccumulation.1. IntroductionJames Karr and his colleagues pioneered the concept of an Index of Biological Integrity (IBI) for streams and its application to sh in the 1980s. e US Environmental Protection Agency subsequently developed an IBI using macro-invertebrates in the 1990s because they agreed with Karr that the EPAs past reliance on chemical grab samples gave only a brief snapshot of the water chemistry and did not give an integrated picture of impacts of water chemistry, or other stressors, on biological communities. Especially important is that chemical sampling to detect a point source requires being present at the time of release from a discharge pipe (smoking gun) or at the lead edge of a ood hydrograph to detect non-point source contamination. In addition, IBIs are cheaper and communities have a memory in that species identities and relative abundance reect cumulative eects of all stressors over a time period of at least the generation time of the species present.2. A Traditional IBI for Cave Streams May Not be PossibleCaves have very few species so a traditional IBI based on sh and/or macro-invertebrates and including indices of diversity is not practicable. So we should use all the species that can be identied at least to the generic level. Many species of isopods and amphipods cannot be separated in the eld and even bait-trapped individuals would have to be sent to experts to identify making the time involved prohibitive. In addition, with only a few species, the presence or absence of the large species like sh, will have huge eects on a species diversity-based IBI due just to sampling error (WILLIAM PEARSON personal communication). So what is the solution? I suggest that we census all the species of stream and stream bank fauna. It should be obvious that we census when the water is clear and low both for increased reliability and lowered risk of ood. Here I amplify the lower case letters in the abstract. a. Careful measuring of a few captured individuals plus estimation of all sizes of sh, salamanders, and craysh

PAGE 83

Biological Sciences 1344 2009 ICS P roceedings 15th International Congress of Speleology is important because the size-frequency distribution is a sensitive measure of reproductive frequency and success in long-lived cave animals. In cave craysh we also can see the sizes and numbers of developing eggs through the translucent exoskeleton. Measurement of lesions, missing or broken limbs or eroded ns and ns, and visible external parasites is done only on a follow-up trip if the initial community signature suggests acute or chronic toxicity. But any signs of lethargy, unusual behavior, or lack of reaction to disturbance should be noted and lead to careful scrutiny of captured individuals for signs of emaciation or disease. b. At one spot in each downstream to upstream quintile of the stream we census macro-invertebrates at least identied to group both in the water (atworm, snail, isopod, amphipod, shrimp) and on the stream bank (e.g. springtail, mite, spider, pseudoscorpion, daddy-long-legs, carabid beetle, leiodid beetle, millipede, bristletail). Sketch unknown species with a scale indicated. Stream bank fauna and organic matter can become food for aquatic species during oods and can tell us dierent things about adverse impacts. If there are rocks we count the numbers and estimate sizes of associated animals with a sample of 10 rocks. If there are no rocks we do a standard time and area search. At the same locations we estimate the amounts and size distributions of sediment (silt-clay, sand, gravel, rocks, and breakdown) and the relative amounts of coarse particulate organic matter (sizes identiable to leaf, twig, bat guano, and worm castings) and ne particulate organic matter (not identiable). In addition we note the color of the organic matter since black indicates old and scarcely decomposeable material. Note color of any rocks (e.g. black for manganese and orange for iron) and its feeling when rubbed (slippery means a microbial biolm). c. Especially if there are not rocks or mud clods to census for small fauna, I suggest using baited traps. If you are making only one trip then leave traps as you census upstream and retrieve and census them as you return. If you can return in a few days you are more likely to catch things. In either case I suggest a trap of your design that allows water circulation so that the smell of bait can be carried downstream. And I suggest small rocks plus washed leaf litter as a refuge for small species from predation by larger species or individuals. Place the contents in a white bottomed pan and census what you catch. d. In all cases it is important and easy to distinguish among troglobitic, troglophilic, trogloxenic, and accidental species. In general troglobites are apparently white and eyeless and troglophiles are pigmented and have visible eyes on close inspection. Trogloxenes can not normally complete there life cycles in caves and in the extreme include tubicid worms and colonial sewage bacteria.3. Potential Problems with Implementing My Community SignaturesI have had some feedback from folks at a workshop in which my ideas were discussed so here I briey respond to a few of the perceived problems and comment on some misunderstanding. Some of my comments hold for IBIs in general. ey include possible need to develop metrics for dierent karst regions, diculty of identifying animals to species in the eld, the misunderstanding that health of a stream is useful concept, and the perception that dierent potential implementers could not agree on the degree of expertise and training needed to use the methodology. a. e beauty of Karrs sh IBI is that it is easy, with a little training, to identify all species quickly in the eld and release them. But even Karrs sh IBI, that includes scores for generalist species and pollution intolerant species, has had to be modied for geographic regions with very dierent sh faunas. Dierent karst regions also have dierent cave stream faunas but my proposed signatures do not use a scoring system and deal with categories of species (troglobites, troglophiles, trogloxenes, and accidentals) that can be identied in the eld. I also rely heavily on sizefrequency data for these categories of animals to hypothesize the kind of impact. It could be argued that lack of a scoring system is a disadvantage but all systems have negative and positive tradeos. b. Use of categories of cave animals and size-frequency distributions make the inability to separate similar copepod, isopod, and amphipod species in the eld less of an impediment to identifying the category of impact. Further, in caves there are rarely more than two species of amphipods or isopods or atworms that are dicult to distinguish in the eld and they do not seem to have very dierent pollution tolerances. In contrast surface stream oen have dozens of species of oligochaete worms and midge larvae that are impossible to tell apart in the eld, very dicult to distinguish microscopically, and have striking dierences in pollution tolerance. c. I have problems with the concept of a healthy ecosystem or community because health is dicult to dene. An ecosystem that is highly eutrophic, due to excess nutrients, is greatly simplied but it still has fully functional primary and

PAGE 84

15th International Congress of Speleology Biological Sciences 1345 2009 ICS Proceedings secondary production, structured food chains, and nutrient recycling. And all systems change over time with natural variations in weather and climate; the eects of rare but important oods for food renewal in caves is a case in point. A better measure of impact on a community is whether it is resilient in that it returns quickly to the same state aer natural or anthropogenic disturbance. d. In my 40 years of doing biosurveys of cave streams I have found three non-scientists who spent the time needed to get the same data that I got in the same stream at the same time. e issue is interest and dedication to the task rather than initial ability or formal training. Unlike with surface streams, we do not need a huge cadre of volunteers to do the work. And there are increasing numbers of individuals who even make a living doing biosurvey work in caves.4. Background for a Community Signature Approach to Cave Stream Bio-IntegrityI rst suggested the use of community signatures in a paper on e Mammoth Cave Ecosystem (POULSON, 1992) and followed up on the idea in a proposal for Long-term Ecological Monitoring (LTEM) for which I was the lead as a Consulting Ecologist (GS-14) for summers of 1992-1994 at Mammoth Cave National Park. e proposal ranked 1 of 150 proposals, submitted by National Parks and Monuments, for scientic merit and has now been funded for over a decade. e water sampling program includes measurement of Total Organic Carbon, conductivity, temperature, and pH at major springs. And sampling is done every two years synoptically and conditionally with large rain events. Unusual results lead to more extensive sampling, e.g. for pesticides and heavy metals, and sampling in the aquifer progressively upstream to try to nd the source of any unusual change in water chemistry. A biological survey is done where a cave stream is available for in any spring watershed. A central precept of our LTEM proposal was based on a pioneering paper on e Central Kentucky Karst (WHITE et al., 1970 in POULSON, 1992). We argued that karst surface features and knowledge of the processes of cave formation would help us to identify potential threats to the cave system and that these threats should drive management. Based on horizontal diagrams of the Central Kentucky Karst and vertical proles of dierent geological, mineralogical, and biological zones in the Mammoth Cave System (POULSON, 1992), we identied potential impacts on aquatic ecosystems. ese exemplify worldwide problems of groundwater pollution in karst regions. In Karst, there is rapid transfer of water and contaminants from the surface to the subsurface and most contaminant transfer occurs with rainfall events. Extensive areas of recharge encompass a wide variety of land uses and threats. Here are particulars for the Central Kentucky Karst. In the recharge areas starting with sinking streams and continuing under the sinkhole plain, agricultural activities are the most obvious threats. Productive farmlands are located among the sinkholes and sinking creeks. ey are non-point sources with chronic contributions of sediment, pesticides, nutrients, and bacteria to cave streams. Point source spills have occurred and will continue to occur along Interstate 65, the Cumberland Parkway, and the CSX railroad which traverse the watershed. Examples of spills include anhydrous ammonia, ink, diesel fuel, gasoline, heating oil, and paint. ey are a constant reminder of the vulnerability of the karst aquifer. Along the transportation corridors, urban development is expanding. Such growth will increasingly burden aquatic resources downstream. Common problems are sewage and solid waste disposal and leakage from buried storage tanks and pipelines. Emerging problems are industrial wastes. Another class of point source pollution is associated with oil and gas production. ese include spills and leaks of drilling uids, muds, hydrocarbons, and brines. Brines encountered during drilling are high in sodium, chloride, sulfates, and some heavy metals. Illegal release of petroleum production brines has and is occurring. Dams on the Green and Nolin Rivers have caused problems because of ooding and hydraulic damming via the major springs back into base level rivers of the Mammoth Cave System. By altering timing, of holding water back and releasing it, the normal seasonal hydrograph of level, ow, and ow direction have been altered. Assessment of potential threats can be combined with hypothesized community signatures for each class of threat.5. Hypothesized Community Signatures for Dierent Classes of ImpactBased on ecological, toxicological, and physiological principles, I have hypothesized community signatures for negative impacts of ve classes of threats. ese kinds of threats may be predicted a priori from activities and land use patterns on the surface and/or they may be hypothesized from the community signatures in cave streams. Of course interpretation may be complicated if there are multiple kinds of impacts.

PAGE 85

Biological Sciences 1346 2009 ICS P roceedings 15th International Congress of Speleology a. OO rganic enrichment can have a range of eects from positive to extremely negative. Since caves are food-limited, there are always positive and negative tradeos. Inorganic fertilizers may have no eects on stream communities since there are no plants to respond in caves. However inorganic fertilizers will cause algal blooms in surface waters. If these blooms wash into caves that results in large amounts of decomposeable organic matter which can have negative impacts in cave streams. Of course bacterial and fecal input from septic systems, farm feed lots, manure on elds, or failed sewage treatment plants also introduce decomposeable organic matter. In either case I expect increases of secondary productivity that can be inferred from a continuum of progressive eects. I expect the following sequence of impacts with increasing amounts of decomposeable organic matter. First the frequency and degree of biolms will increase with lowest level organic enrichment. Next I expect increased reproduction of the shortest lived troglobitic detritivores, like atworms and isopods that graze on biolms and ne particulate organic matter. is eect can be inferred from size-frequency distributions and abundances. With greater organic enrichment I predict a shi to presence and then dominance of troglophiles. With still further enrichment I predict that even troglophiles are replaced by huge densities of trogloxenes, especially dense mats of red tubicid worms and stringy colonial sewage bacteria. At worst there is so much decomposeable organic matter that the biological oxygen demand of aerobic bacteria results in elimination of oxygen and death of all macroscopic organisms. And to reiterate an earlier point, even the most simplied of these communities demonstrate all aspects of ecosystems function even though our value judgment may be that they are not like what we expect or desire without disruption by human activities. b. Toxins aka poisons have a continuum of eects. Chemicals that are toxic include inorganic compounds and organic compounds. In both cases they are toxic because they are rare in nature and organisms have not evolved ways to detoxify them by breaking them down or sequestering them in parts of their body where they have no eects. As a result the toxins will build up in the body as long as intake and absorption is greater than egestion, excretion, and detoxication. Inorganic toxins include heavy metals like cadmium, mercury, and lead. Organic toxins include hydrocarbons like oil and gasoline and pesticides including chlorinated hydrocarbons (e.g. DDT and chlordane), organophosphates, carbamates, and perhaps pyrethroids. I predict that pulsed (point source) inputs of high concentrations of toxins will aect the shortest lived and fastest growing species at the beginning of the food chain. is is the reason that herbicides most aect weeds and antibiotics most aect bacteria. e species most aected in a cave stream include troglophiles, if present, and the species of troglobites at the beginning of food chains, especially isopods and copepods. I expect that their numbers will be greatly reduced if there is acute toxicity. Larger troglobites with lower growth and metabolic rates may crawl out of the water onto the stream bank. I would expect this for craysh and some salamanders when there is a sudden pulsed input of a toxin. It is counterintuitive, but true, that long-lived large troglobites, like sh and craysh, are not so sensitive to pulsed toxin input. e reason is that they have very low metabolic rates. On the other hand, I predict that chronic (non-point source) inputs of toxins will most aect the longest lived and slowest growing species at the end of the food chain. ere are two reasons. First, the longer an individual lives the more the toxin slowly increases over months and years by the process called bioaccumulation. Second the more steps there are in a food chain the more the toxin quickly increases by a process called biomagnication. A toxin builds up at least 10 fold at each step in a food chain and this can happen every day or week for years. e combination of biomagnication and bioaccumulation puts top predator troglobites, like sh and salamanders, at double jeopardy. Even craysh in caves are partially predaceous, especially on isopods. As a result I predict that the largest and oldest troglobites, especially sh and salamanders, will be the rst to die and so the sizefrequency distribution will be shied toward the smaller sizes. Live large individuals might show signs of weakness or might have lesions or deformities. e older individuals are normally the ones reproducing most, so the frequency of very small individuals might also be reduced. Due to a reduced number and smaller size of predators, the number of prey individuals (copepods, isopods, and amphipods) may increase. c. Siltation will homogenize the stream bottom habitat and so cause a decline in diversity of species and numbers of individual in remaining species. If gravels and rocks are covered by silt then there are few hiding places for isopod and amphipod prey of large species. In addition silt may mix with ne particulate organic matter and cover biolms; both reduce the availability of food to detritivores causing their numbers to decline. An indirect eect of less prey may fewer predators.

PAGE 86

15th International Congress of Speleology Biological Sciences 1347 2009 ICS Proceedings Troglobitic sh and craysh are not necessarily compromised by siltation because they have food other than isopods and amphipods. Fish eat mostly copepods even when other preys are present and craysh eat coarse particulate organic matter or even ne particulate organic matter mixed with silt. Many caves in the United States are in areas where there was massive land clearing for agriculture up until the early 1900s and these caves oen have streams with mud and silt banks and much silt on the bottom. d. HH ydropatterns of cave streams can be changed by dams and/or reservoirs on surface streams either upstream or downstream of the cave. Reservoirs and dams upstream can reduce the amount of particulate organic matter washed into a cave and its seasonal pattern. Many cave stream species synchronize each others reproductive state and reproduce based on subtle seasonal increases in food input. Dams and reservoirs on rivers upstream or downstream from spring resurgences of cave streams may also change cave stream hydropattern by both backooding and hydraulic damming. Hydraulic damming slows or stops the normal recession of water aer a ood and so results in deposition of silt and may stop ne particulate organic matter from reaching the downstream parts of cave streams and rivers. is may also compromise shrimp and copepods, which feed only on biolms and ne particulate organic matter. Without these prey species the food supply of cavesh is reduced. And, as with direct siltation, I predict that interruption of normal hydropatterns may compromise all troglobites and so I predict that only the large individuals of sh and craysh can persist since they have the greatest capacity to forage over great areas and live long times with little or no food. 6. Tests of Community Signature Hypotheses; Inadvertent and Natural Experiments In this nal section I give some examples of tests of my community signature hypotheses. As a baseline I review several multidecade and continuing stream biosurveys in cave streams in the Mammoth Cave system that are local watersheds with near pristine surface watersheds. ese show the kinds of variation possible within and between three streams with only natural variation including the eects of 100-year oods. en I give snapshots of cave streams where the community structure suggests dierent adverse impacts due to human activities in the surface watershed. a. ree pristine Mammoth Cave streams have maintained the same community signatures from 1965. e rst 25 years of study including 3 surveys is reported with actual data in POULSON (1992). All three streams showed a persistent change in species composition and density along an upstream to downstream decrease in substrate heterogeneity and particulate organic matter and with an increase in depth and width. And persistent great dierences in particulate organic matter and occurrence of rock refuges among streams only changed overall abundances and sizefrequency distributions of species (dierential reproduction) rather than the relative abundance of species both for the aquatic and streambank terrestrial communities. A 100-year ood changed the distribution of ries and pools along the stream but not the community of animals. One of the streambank terrestrial communities was virtually lost for a year from 1966 to 1967 as a result of a pollution event but the stream community was not aected. Two sewage treatment lagoons overowed into the headwaters leaving a line of green cyanobacteria (Spirulina) at the water edge along the length of the stream. is alga is known to produce toxins and I argued, as part of a successful suit to have the Job Corps and the lagoons removed from above the cave, that the short-lived terrestrial organisms ate the algae and died. b. With catastrophic spills of toxins there is oen no interpretation necessary to determine cause of complete loss of the cave stream communities. One example was a gasoline spill into Pless Cave in Indiana. It resulted in in immediate loss of a large population of cave craysh. A second example was a breakage of a fertilizer line that released urea into the cave spring system in Meramec Missouri. e resulting huge biological oxygen demand of bacterial decomposers causing a sudden loss of dissolved oxygen killing 100s of cave sh and 1000s of craysh that oated out of the huge spring. c. Hidden River Cave in Horse Cave Kentucky is perhaps the best studied case where pollution caused local loss of a cave stream community. e problem with interpretation is that both heavy metals from an electroplating operation and decomposeable organic matter from a creamery spilled into the cave over several decades. At that time the sequence of decline in the cave community was not documented but the intermittent recovery has been well-studied. e electroplating stopped and so the recovery involves decreasing amounts of decomposeable organic matter with time.as the creamery wastes could be treated by a new regional sewage plant.. As predicted the rst species to reappear as dissolved oxygen started to increase were colonial sewage bacteria and mats of tubicid worms in the shallow waters along shore where maximum oxygen levels were found. e hemoglobin in the

PAGE 87

Biological Sciences 1348 2009 ICS P roceedings 15th International Congress of Speleology red worms binds oxygen at extremely low concentrations. As the residual creamery wastes decomposed oxygen levels increased high densities of cave isopods appeared as sewage bacteria and worms disappeared. About the same time troglophilic craysh appeared. Finally, aer several years troglobitic craysh and amphipods and a few troglobitic cavesh appeared. All of these had recolonized the area from upstream small tributaries that had been refuges from pollution. As of 2002 the main cave stream had still not recovered to the point that only troglobitic species occurred, as they still do in the small upstream refuges. Every time there has been a big ood the incompletely decomposed creamery wastes that had been buried by sediments were remobilized. e rst time this happened the community reverted to colonial sewage bacteria and tubicid worms. Aer several cycles of burial and remobilization of the decomposeable organic matter the community no longer reverts to the simplest community. We do not know ho long it will be before all the buried wastes are remobilized and fully decomposed. d. Water leached particulate organic matter, bat guano and leaves and twigs, should not change cave stream community signatures because the potential food has already lost most of the easily decomposeable organic matter. My observations over the past 50 years support this prediction. Cave streams with maternity colonies of gray bats or massive amounts of leached particulate wood and leaves have only the expected troglobites. ey simply grow faster and reproduce more than in caves with very low amounts of particulate organic matter. e. e loss of a diverse community of troglobites in Shelta Cave in the 1970s may be explained by two alternative hypotheses of human impacts. Due to a cave gate not optimum for a small colony of gray bats it was rst hypothesized that disappearance of the energy from bat guano cased loss of the troglobites including three species of craysh, one shrimp, and one sh. For the reasons summarized in the previous paragraph, I had always considered this hypothesis unlikely. Recently it has been suggested that chlordane used to control termites caused the problem as the city expanded around the cave. is hypothesis ts a community signature I predict with chronic (non-point) input of low concentrations of a toxin. Even in the 1960s, large sh were mostly absent and reproduction by the largest craysh was very low. I expect that the last organisms that were censused in the 1970s, as the community disappeared were small individuals of sh and craysh. f. e historic change in community signature of base level stream communities in Mammoth Cave is consistent with the negative impact of changes in hydropattern and hydraulic damming I predicted above for eects of dams on Green River. In 1906 Lock ands Dam 6 was built just downstream of the spring resurgences from the cave and in the 1950s a large upstream dam was built. e observations of loss of the cave community have been reviewed (POULSON, 1992) so I only summarize the initial recovery here subsequent changes here. Briey the rst animals to reappear when release patterns from the upstream dam were changed to be more seasonally natural were very large cave craysh and then large cavesh. Recently shrimp have been seen again. All of the fauna has recolonized from more upstream parts of cave streams that are not subject o hydraulic damming and still have ne particulate organic matter and refuges for isopod and amphipod prey.Acknowledgments and a RequestStarting in the 1960s Cave Research Foundation colleagues that helped most with stream surveys over many years were Stan Sides and Richard Zopf. My work in Hidden River Cave complements studies by Jerry Lewis. John Cooper and Horton Hobbs III shared invaluable observations and data on Shelta Cave. And, since the 1990s, Bill Pearson, and his students along with Rick Olson, and Kurt Helf have continued and improved my surveys. Over the years many additional friends, colleagues, and students have helped me in the eld them and have shared ideas. You know who you are and I thank you one and all. And I solicit thoughts and comments for any readers of this work in progress.ReferencePOULSON, T.L. (1992) e Mammoth Cave Ecosystem. Chap. in A. Camacho (ed.). e NN atural HH istory of Biospeleology, Museo Nacional de Ciencias Naturales, Madrid, pp 569.

PAGE 88

15th International Congress of Speleology Biological Sciences 1349 2009 ICS Proceedings VV ISUAL OBSERV V ATIONS OF THE MACROSCOPIC LIFE IN PUERTO RICAN CAV V ES FROM 2002 TO 2008RONALDRONALD T. RICHARD RICHARD SE E nironmental Science, Turabo UU niversity and Sociedad EEspeleolgica de Puerto RR ico II nc. (SE E PRI RI) PO O Box 366894, San Juan, PR R 00936-6894 Between 2002 and 2008 biological notes were collected during 95 visits to 43 caves in Puerto Rico. e data is non-specic, presence/absence. No census data were collected. is data set includes only the main island of Puerto Rico and does not include oshore islands such as Isla de Mona. All data were collected in total darkness with the exception of bats, the Puerto Rican boa, and depigmented crustaceans, which were noted even if found in the entrance areas of the cave. e life of the caves was photographed but no traps were set, nor were rocks systematically overturned, and no specimens were collected. Identication is not to species but to categories that can be indentied in the eld. All notes were written down in the cave or shortly aerwards. Most life observed is larger than 1 cm and does not live under water or mud. e goal of this paper is to determine what forms of life are common and what are not common in Puerto Rican caves. Large numbers of depigmented adult animals were observed on only two of the 95 trips. In Cueva de Murcielagos in Guanica hundreds of depigmented crustaceans (amphipods and isopods) were observed in a saline pool of water under a bat roost in the entrance area of the cave. In Cueva Lechuga in Camuy dozens of depigmented millipedes were observed. Both of these animals have been observed by others over a span of years. Most of the depigmented animals observed in Puerto Rican caves appear to be in a phase of their life and are not permanently without pigmentation. Depigmented cockroaches, whip spiders, and millipedes have been observed and in these cases it is a phase of their life. A total of 36 taxonomical groups have been observed in Puerto Rican caves. Life was observed in 91% of the caves visited. In Puerto Rico, caves without life tend to ood completely or be small enough to explore in 15 minutes or less. e data were normalized so that each cave has the same weight. e most common life found in Puerto Rican caves and their frequency are bats (73% of caves visited), cockroaches (47%), whip spiders (45%), crickets (42%), plants deposited by bats (41%), crabs (39%), gnats (38%), spiders (23%), snails (22%), fungi (21%), roots (18%), rats (11%), small guano animals (8.8%), large amphibians (6.8%), sh (6.7%), small amphibians (6.0%), benthic shrimp (5.8%), water-deposited plants (4.6%), swimming shrimp (4.4%), boas (4.3%), scorpions (2.8%), earthworms (2.6%), depigmented millipedes (2.3%), pigmented millipedes (2.3%), ants (1.6%), unidentied human parasites (1.4%), depigmented crustaceans (1.2%), mosquitoes (0.5%), and earwigs (0.2%). ree (Yuyu, Ro Encantado, and Vientos) of the four caves with the highest biodiversity are in the Sistema de Ro Encantado. is is the largest mapped cave in Puerto Rico. e cave has 17 km of continuously traversable passage, some of which is underwater. e land surface above the cave is rugged karst and poor farmland. e area has been abandoned and has low levels of human population. Above the Sistema de Ro Encantado is one of the largest road-less tracks in Puerto Rico. Biodiversity is not the same as biomass.1. Introductione animal life found in caves is oen classied into four categories. Troglobites are animals that live only in the total darkness of caves and have morphological adaptations to live in caves. e most common adaptations are loss of pigmentation and eyes, extended appendages, and a slower metabolism. Troglophiles are animals that live their entire life in caves but are also found outside of caves. Frequently they are nocturnal and live in humid environments. Trogloxenes are animals that visit caves but cannot complete their entire life cycle in the cave. Bats, which must leave the cave to feed, are the best-known example of trogloxenes.

PAGE 89

Biological Sciences 1350 2009 ICS P roceedings 15th International Congress of Speleology Accidentals are animals that are in the cave by accident and do not regularly visit the total darkness of caves. In Puerto Rico, species lists have been published for Cueva Aguas Buenas (Fenton E NTON 1968; Beck ECK et al., 1976) in Aguas Buenas, Cueva El Convento (Nicholas I CHOLAS 1974) in Guayanilla, and the caves in the Gunica dry forest (Conde ONDE Costas OSTAS and AND Gonz ONZ lez LEZ 1990). Invertebrate species lists for several caves have been published by Peck E CK (1974). ere are 13 species of bats on the island of which nine regularly roost in caves (Gannon A NNON et al., 2006). Bat populations in Puerto Rico have been published for two caves. Cueva Cucaracha has 700,000 and Cueva Cuelebrones has 300,000 bats (Gannon A NNON et al., 2005). It is fair to assume that many caves in Puerto Rico have more than 10,000 bats. e largest bat populations are found in hot caves where the body temperature of the bats can raise the air temperature to more than 30 C (Gannon A NNON et al., 2005). Hot caves are formed when tens of thousands of bats roost in a room at higher elevation than a single constricted entrance. Under these conditions bats use less energy and water but they are more vulnerable to predators that wait at the constricted entrance. Fruit-eating bats bring seeds and leaves back to the cave roost (Kunz UNZ and AND Daz AZ 1995). Viable seeds are dropped to the oor of the caves where they sprout and grow in the humid and dark conditions. e seeds are doomed as they will never encounter the sunlight needed for photosynthesis. e seeds grow until they have exhausted the supply of energy in the seed and then they die. ese seeds are an important source of nutrients for other species that live in the cave. Crickets in Puerto Rican caves are from the genus A A m phiacusta. e genus AA m phiacusta is described in Desutter E SUTTER Grandcolas RANDCOLAS and AND Otte TTE (1997). Fungi have been studied in the Sistema de Ro Camuy (Nieves IEVES Rivera IVERA 2003). Ship rats (RR attus rattus) were introduced into Puerto Rico by Spanish settlers in the 16th century. Norwegian rats (RR norvegicus) arrived later. Both species live in the cities but the ship rat is dominant in rural areas such as the El Yunque National Forest (Engeman N GEMAN et al., 2006). Terrestrial coastal areas of Puerto Rico have a variety of crab species but the center of the island has only the Puerto Rican Freshwater Crab (EE pi lobocera sinuations). e reason for this is that all other crabs and shrimp spend part of their time in saltwater. ey may spend most of their lives on land but they must migrate to the ocean to breed. e freshwater crab reproduces in freshwater and does not migrate (Cook O OK et al., 2008). Puerto Rico has freshwater shrimp from the genera Macrobrachium, Xiphocaris, and AA tya (Crowl ROWL et al., 2000). Macrobrachium are benthic and have pincers. In the 1960s a 50 cm long Macrobrachium carcinus was captured in the Tres Pueblos Sinkhole (Gurnee U RNEE and AND Gurnee URNEE 1987), which is part of the Sistema de Ro Camuy. Xiphocaris are usually less than 4 cm long and are depigmented even though they live in surface waters. ey also have the tendency to jump over obstacles such as swi water. AA tya is a swimming shrimp intermediate in size between Macrobrachium and Xiphocaris. All these shrimp need to migrate between freshwater and the estuary and all have suered population losses as the construction of dams has broken their migratory routes. e largest undammed river in Puerto Rico is the Ro Grande de Manat in the north central part of the island. At caves with huge bat populations the Puerto Rican Boa (EE picrates inornatus) waits at the constricted entrance and then captures bats as they nightly exit the cave (Rodrguez ODRGUEZ and AND Reagan EAGAN 1984). e boa is protected by both Puerto Rican and federal law. In a study in El Yunque National Forest the boas averaged 1.4 m in length (Wunderle UNDERLE 2004). e boas in El Yunque have limited access to bats. e cave in Puerto Rico with the best-documented sightings of depigmented animals is Cueva de Murcielagos in Gunica and all the depigmented animals are crustaceans. Peck E CK (1974) saw the depigmented crustaceans and reported that the water was hypersaline. Conde O NDE Costas OSTAS and AND Gonz ONZ lez LEZ (1990) saw Typhlatya monae, Stgiomysis holtuisi, and Metaniphargus bouseldi and the specic conductivity of the water was 7,600 S/cm. which is about 15 percent of the ocean. For centuries the residents of Puerto Rico have used caves as another resource in the struggle to survive. Water, guano, crabs, and shrimp have been extracted from caves. In the 20th century caves were lit with homemade kerosene lanterns, constructed with glass bottles and wicks. ese are called jachos in Puerto Rican Spanish. Gurnee U RNEE and AND Gurnee URNEE (1974) described the air in Cueva Aguas Buenas in Aguas Buenas as smoky.

PAGE 90

15th International Congress of Speleology Biological Sciences 1351 2009 ICS Proceedings e largest mapped cave system in Puerto Rico is the Sistema de Ro Encantado (Courbon O URBON et al., 1989) in the municipios of Floriada, Ciales, and Manat. From upstream to downstream this cave includes Cueva La Escalera, Cueva Ro Encantado, Cueva Juan Nieves, and Cueva Aguas Frias. Cueva Aguas Frias is the resurgence of the Ro Encantado. is system has 17 km of continuously traversable passage, short sections of which are underwater. Non-traversable but linked by dye tracing to the Sistema de Ro Encantado are Cueva Yuyu, Cueva Zumbo, and Cueva Vientos (Morales ORALES 2007). A dry cave in the drainage basin of the Ro Encantado is Cueva Balcones. e dierent entrances of the system are widely separated and separated by sumps. e Ro Encantado is a tributary of the Ro Grande de Manat.2. MethodsFor this study 43 caves were visited 92 times between 2002 and 2008. Cueva Clara de Camuy in Camuy is the section of the Sistema de Ro Camuy that is a show cave. All the other caves were not developed for commercial use. Life in the caves was observed while caving and no trip made was solely to collect biological data. e data is non-specic presence/ absence. No attempt was made to collect census data. No traps were used and no rocks were systematically overturned. Attempts were made to photograph the life but no specimens were collected. Most life observed is larger than 1 cm and does not live under water or in mud. e life was observed by the author or another member of the party. e data was normalized so that each cave has the same weight in the numbers presented. For the purpose of this study each cave entrance in the Sistema de Ro Encantado is considered a separate cave. Living organisms were classied to categories that could be identied in the eld. In most cases the species cannot be identied. ere are two problems with species level identication. First it would require collecting samples and sacricing the animals and second it would require specialists that may not be readily available.3. Resultse results are shown in Table 1. Living organisms were observed in 91% of the caves surveyed. In Puerto Rico, totally dark caves without observed life tend to be small enough that they can be explored in 15 minutes or ood completely. Bats were observed in 73 percent of the caves surveyed. I have photographed bats with orange parasites. Cockroaches or cockroach-like animals were observed in 47 percent of the caves in this survey. is category includes anything called cucaracha in Spanish. is category includes beetles, native species such as the Puerto Rican Cave Cockroach (AA spid uchus cavernicola) and the invasive American cockroach (Periplaneta americana). In Cueva Mucara in Aguas Buenas cockroaches were photographed in the act of copulation. In caves with thousands of cockroaches it is common to observe depigmented animals. For several hours aer shedding their exoskeletons cockroaches are depigmented. Whip spiders were observed in 45 percent of the caves surveyed. Most whip spiders observed in Puerto Rican caves are probably Phrynus longipes as this is the largest and most readily observed species on the island. Whip spiders have been observed eating cockroaches and crickets. Between December and June whip spiders have been observed mating, carrying reddish-brown eggs on their stomach, or yellowish-green newly hatched ospring on their back. e reproductive season of the whip spiders coincides with the driest part of the year. In caves where whip spiders are observed it is common to observe 5 to 10 during the cave trip. Crickets were observed in 42 percent of the caves surveyed. It is common to observe 5 to 10 crickets at a time. In Cueva Aguas Buenas in Aguas Buenas crickets were photographed in the act of copulation. Sprouting plants that were deposited by bats were observed in 41 percent of the caves surveyed. Crabs were observed in 39 percent of the caves surveyed. Crabs can be found either under or over the water surface. A dead crab was observed in a dry cave, Cueva La Mora in Comerio. Crabs have been photographed eating bats and seeds. Gravid females have their ospring in a compartment in their abdomen and if they are near to hatching and the crab is picked up it can rain baby crabs. is has been observed in Cueva Represa in Hatillo. Gnats were observed in 38 percent of the caves surveyed. Gnats are detected in caves solely because they are attracted to the headlamps of the cavers. Usually they are only a few millimeters in size and would pass unobserved if they were further from the lights. In some caves, with large bat populations, the gnats ying into the mouth, ears, and nose become so bothersome that cavers turn o their headlamps and use only hand lights. Spiders or spider webs were observed in 23 percent of the caves surveyed. Observed spiders have been small and did not have hairy tarantula-like bodies. In Cueva Represa in Hatillo a spider was photographed carrying an egg sack.

PAGE 91

Biological Sciences 1352 2009 ICS P roceedings 15th International Congress of Speleology Live snails or snail shells were observed in 22 percent of the caves surveyed. Based on color and shell morphology there are at least three species of snails present in Puerto Rican caves. No slugs have been observed. Fungi were observed in 21 percent of the caves surveyed. e laments of the fungi are usually white but are sometimes black. e most common substrate of the fungi is organic material such as wood or a dead animal. Sometimes the laments were on a soil or rock substrate that did not appear to be organic. No fruiting bodies were observed. Roots extending into total darkness were observed in 18 percent of the caves surveyed. Trees on the surface are extending their roots into the cave to extract water and nutrients. ese roots are then eaten by other residents of the cave. Rats were observed in 11 percent of the caves surveyed. A dead rat was found in Cueva Aguas Buenas in Aguas Buenas and the tail was longer than the body which identies it as a ship rat. Small guano animals were observed in 8.8 percent of the caves surveyed. is non-specic category includes the myriad of small animals seen moving on guano piles. Large amphibians have bodies larger than 10 cm and were observed in 6.8 percent of the caves surveyed. is category is probably mostly Cane Toads (Bufo marinus) and to a lesser extent Bullfrogs (RR ana ca tesbeiana). Both of these species are invasive and not native.

PAGE 92

15th International Congress of Speleology Biological Sciences 1353 2009 ICS Proceedings Fish were observed in 6.7 percent of the caves surveyed. Most were under 5 cm in length and were described by the observer as guppies. Small amphibians were observed in 6.0 percent of the caves surveyed. Small amphibians are less than 5 cm long. e native amphibians in this category include several species from the genus EE l eutherodactylus as well the white-lipped frog (LL eptodactylus labialis). is category would also include several invasive species. Benthic shrimp from the genus Macrobrachium were observed in 5.8 percent of the caves surveyed. ese are typically from 15 to 30 cm long. Most of these shrimp were observed in the Sistema de Ro Encantado which is a tributary of the Ro Grande de Manat. Unlike crabs, shrimp have only been observed only underwater. Plants deposited by water were observed in 4.6 percent of the caves surveyed. Water-deposited plants can be distinguished from bat-deposited plants because they tend to be in alluvial deposits and at lower elevation than batdeposited plants. Swimming shrimp from the genus AA t ya were observed in 4.4 percent of the caves surveyed. e Puerto Rican Boa was observed in 4.3 percent of the caves surveyed. In this data set if a boa was observed in all cases only one was observed. Cueva uintero in Corozal is a horizontal dry river passage with a large bat population. Typical passage diameters are 10 to 20 m with no constrictions. e arrival of the cave explorers disturbed the bats and they ew to other locations in the cave. During this process a boa on the wall about 1.8 m above the oor was observed capturing a bat in room that was about 10 m wide. Scorpions were observed in 2.8 percent of the caves surveyed. In all cases if a scorpion was observed then only was seen. Earthworms were seen in 2.6 percent of the caves surveyed. In Cueva Lechuga in Camuy cockroaches appeared to be feeding on an earthworm. Depigmented millipedes were observed in 2.3 percent of the caves surveyed. e lone cave where this was observed is Cueva Lechuga in Camuy. Another cave explorer had observed the same animals in this cave in 2001 (Miller, written communication, 2008). Pigmented millipedes were observed in 2.3 percent of the caves surveyed. In Cueva Ro Encantado an adult millipede was observed and then one week later an adult was observed in the same location with newly hatched millipedes. One interpretation of the observation is that the adult was guarding a nest. It is impossible to know if the adult was the same even though the location was the same. Ants were observed in 1.6 percent of the caves surveyed. In 1.4 percent of the caves surveyed I received bites from an unknown animal. is unidentied human parasite is the only thing in a cave that has ever bitten me. Depigmented crustaceans were observed in 1.2 percent of the caves surveyed. e only cave where this was observed is Cueva de Murcielagos in Gunica where depigmented amphipods and isopods were seen in an entrance area in the slurry from Lago Guano. e cave has thousands of bats and is the only cave in Puerto Rico where the bat fecal material falls directly into saline water. Visitation and resource extraction from Puerto Rican caves have declined dramatically. In the 1960s and 1970s cave explorers saw many groups of Puerto Ricans visiting Cueva Aguas Buenas in Aguas Buenas (Beck E CK et al., 1976). e author has visited caves in Aguas Buenas 31 times and has never seen another group. ere is no evidence that anyone is entering caves to mine guano or collect crabs or shrimp. Most of the water-supply systems in caves have been abandoned. e author is aware of two operating watersupply systems in caves. Jachos have been observed were they have been le at the entrances to the cave. In most cases the kerosene has evaporated long ago. In 2006 a functional jacho was observed and lit in Cueva Aguas Buenas in Aguas Buenas. e jacho had been made from a 2003 wine bottle from Argentina.4. DiscussionOf the 43 caves surveyed only one, Cueva Lechuga in Camuy, has freshwater and depigmented animals that live in total darkness and have been observed on two visits over seven years. It is a mystery why depigmented crustaceans are routinely observed in the entrance area of Cueva Murcielagos in Gunica. is cave has tens of thousands of bats and is the only place known to the author where bats roost over saline water. e salinity varies signicantly. It has been speculated that the depigmented crustaceans evolved in total darkness and then migrated out to the entrance that has higher nutrients (Conde O NDE Costas OSTAS and AND Gonz ONZ lez LEZ 1990). In the entrance area of the cave being white should make it easier for predators to capture them.

PAGE 93

Biological Sciences 1354 2009 ICS P roceedings 15th International Congress of Speleology For many animals found in Puerto Rican caves we do not have enough knowledge to classify them. In temperate caves it is well documented that crickets leave the cave at night to feed near the entrance. Nicholas I CHOLAS (1974) speculated that crickets in Puerto Rico did the same. ere is no evidence that crickets must leave caves in Puerto Rico to nd food. ere is the only cave where a change in the fauna can be identied. irty years ago the Bigclaw River Shrimp (Macrobrachium carcinus) and the Web-footed coqui (EE leutherodactylus karlschmidtii) were observed in Cueva Aguas Buenas in Aguas Buenas. In 10 visits between 2002 and 2008 neither of these animals were observed. e migration of the shrimp has been cut by the construction of the Carraizo Dam in 1953. e fact that the Bigclaw River Shrimp were still being observed in the 1970s indicates that this shrimp can live for at least 20 years in the wild. e Web-footed Coqui is extinct. In both cases the changes inside the cave were the result of changes outside of the cave. Puerto Ricans are no longer visiting caves or using them for resource extraction. is means that the life in the caves has fewer disturbances. It also means that people have less direct knowledge of the caves of the island. It is dicult to conserve what is unknown.ReferencesBECK, B.F., M. FRAM, and J.R. CARVAJAL-ZAMORA (1976) e Aguas Buenas Caves, Puerto Rico: Geology, Hydrology and Ecology with Special Reference to the Histoplasmosis Fungus. NN a tional Speleological Society Bulletin, 38, 1. CONDE COSTAS, C. and C. GONZALEZ (1990) Las Cuevas y Cavernas en el Bosque Xerotico de Gunica AA cta CC ientca, 4(1-3), 113. COOK, B.D., C.M. PRINGLE, and J.M. HUGHES (2008) Phylogeography of an Island Endemic, the Puerto Rican Freshwater Crab (EE pi lobocera sinuations). Journal of HH eredity, doi:10.1093/ jhered/esm126. COURBON, P., C. CHABERT, P. BOSTED, and K. LINDSLEY (1989) AA t las of the GG reat CC aves of the World: Cave Books, St. Louis, MO, 369 pp. CROWL, T.A., N. BOUWES, M.J. TOWNSEND,A.P. COVICH, and F.N. SCATENA (2000) Estimating the Potential Role of Freshwater Shrimp on an Aquatic Insect Assmeblage in a Tropical Headwater Stream: a Bioenergetics Approach. V erh. II nternat. Verein. LL imnol., 27, 1. DESUTTER-GRANDCOLAS, L. and D. OTTE (1997) Revision of the West Indian Genus AA m phiacusta, Saussure, 1874, with Descriptions of Twenty New Species (Orthoptera: Grylloidea: Phalangopdidae). A A n nales de la Socit EE ntomologique de France, 33(1), 101. ENGEMAN, R., D. WHISSON, J. UINN, F. CANO, P. UIONES, and T.H. WHITE JR. (2006) Monitoring Invasive mammalian Populations Sharing Habitat with the Critically Endangered Puerto Rico Parrot AA maz ona vittata. OO ryx, 40(1), 95, doi:1017/S0030605305001286. FENTON, M.B. (1968) Fauna of Aguas Buenas Caves in National Speleological Society Field Trip to Aguas Buenas Caves, Puerto Rico (GURNEE, J., ed.): Special Report to the Municipality of Aguas Buenas, pp 11. Gannon A NNON M.R. Kurta URTA Allen LLEN Rodrguez ODRGUEZ Dur UR n N and Willig ILLIG M.R. (2005) Bats of Puerto R R ic o: AA n II sland Focus and a CC aribbean Perspective Texas Tech University Press, 239 pp. GURNEE, R.H. and J. GURNEE (1974) DD isc oery at RR o C C amuy. Crown Publisher, New York, 183 p. KUNZ, T.H. and C.A. DAZ (1995) Folivory in FruitEating Bats, With New Evidence from AA rt ibeus jamaicensis (Chiroptera: Phyllostomidae). Biotropica 27 (1) 106. MORALES, J. (2007) Estudio de la Hidrulica y Movimiento de las Aguas en los Sistemas de Cuevas de los Ros Encantado y Vientos. Focus, 6(1), 135. NICHOLAS, B.F. (1974) Biology and Ecology of the El Convento Cave/Spring System (Puerto Rico). I I n ternational Journal of Speleology, 6, 109. NIEVES-RIVERA, A.M. (2003) Mycological Survey of Ro Camuy Caves Park, Puerto Rico. J ournal of CC ave and Karst Studies, 65(1) 23. PECK, S.B. (1974) e Invertebrate Fauna of Tropical American Caves Part II: Puerto Rico An Ecological and Zoogeographical Analysis. Biotropica 6(1)

PAGE 94

15th International Congress of Speleology Biological Sciences 1355 2009 ICS Proceedings 14. RODRGUEZ, G.A. and D.P. Reagan EAGAN (1984) Bat Predation by the Puerto Rican Boa, EE picr ates inornatus. C C opeia, 1984(1), 219. WUNDERLE JR., J.M., J.E. MERCADO, B. PARRESOL, and E. TERRANOVA (2004) Spatial Ecology of Puerto Rican Boas (EE picr ates inornatus) in a Hurricane Impacted Forest. Biotropica 36(4) 555 571.

PAGE 95

Biological Sciences 1356 2009 ICS P roceedings 15th International Congress of Speleology VertebrateVERTEBRATE Species PECIES in IN Underground NDERGROUND Features EATURES of OF Arizona RIZONATHOm M As S R R St T RONG WestL L and RResources, II nc., 4001 EE Paradise Falls DDrive, Tucson, AZAZ 85712 UU SA A tstrong@westlandresources.com Underground features, including caves and abandoned mines, provide important resources for a wide variety of vertebrate species in arid and semi-arid regions. Resources used by these species include shelter, nest sites, water, foraging sites, and relief from extreme environmental conditions. Arizona has a wide variety of vegetative communities resulting from high variability in elevation, precipitation, and average temperatures. ese vegetation types range from Sonoran Desert in the southwest part of the state to Alpine tundra on the San Francisco Peaks. Agency les, literature sources, internet sources, and site visits were used to compile records of species using caves and mines throughout Arizona. e majority of features with records of vertebrates were abandoned mines, but data were also obtained from many limestone caves and a few basalt and gypsum caves. At least 101 vertebrate species have been recorded in these underground features, including 63 mammal species, 15 birds, 17 reptiles, and 6 amphibians. Records of vertebrates were obtained for sites in the Sonoran Desert, Semi-desert Grassland, Madrean Evergreen Woodland, and seven other biotic communities. A common pattern in each of these communities is that a few species are found in many sites, and numerous species are found in relatively few sites. Similarly, many caves or mines support a few vertebrate species, and a few sites have larger numbers of species. Dierences in the species composition among the biotic communities are due to dierences in the pools of species present in each biome and to the dierences in the availability of data for each area. e availability and structural characteristics of the caves and mines in each biome may also contribute to dierences in species use of these features. ese underground features provide important resources that may allow individuals or species to survive in harsh climatic conditions. With the large range in elevations, these conditions can include extremely hot and dry summers and very cold winters. Protection of these resources may be critical for the survival of some species and for the maintenance of regional biodiversity.1. IntroductionCaves are widely known to provide habitat for a variety of vertebrate species that spend all or signicant portions of their life cycles inside the totally dark areas of the caves. It is less well known that caves, and particularly cave entrance areas, can provide an important resource for an even wider variety of species. Particularly in arid regions, caves may provide temporary relief from extreme temperature or low humidity conditions. In addition, they may provide hiding places to escape predators, den sites, nest substrates, or hunting locations for predators. While many individual observations of vertebrate species in caves in Arizona have been reported, there have been few attempts to compile this information in any systematic way. Kingsley I NGSLEY et al. (2001) compiled a list of species known to use caves or abandoned mines in the Sonoran Desert. is list was based primarily on personal observations or knowledge of the four authors and on records listed in Hoffmeister O FFMEISTER (1986), but it did not include a thorough literature review. However, this preliminary list included 67 vertebrate species (4 amphibians, 14 reptiles, 11 birds, and 38 mammals). e primary conclusion of that paper was that caves and mines in the Sonoran Desert are signicant wildlife habitat resources and worthy of protection. More recently, Strong TRONG (2006 and 2007) has documented many more sites and species in research on the Sonoran Desert of Arizona and the Chihuahuan Desert of New Mexico. e state of Arizona covers a wide range in elevations, from barely 30 m above sea level on the Colorado River near Yuma to about 3,850 m at the top of Mt. Humphreys near Flagsta. is range of elevations in turn produces corresponding variations in temperatures and rainfall, which support dierent vegetative communities. e biotic communities present in Arizona include desertscrub, semidesert, and other grasslands, evergreen woodlands and savannas, coniferous forests, and alpine tundra. e southern and western part of Arizona is in the Basin and Range physiographic province, with broad, low valleys separated by mountain ranges. Some of these mountain ranges may be over two thousand meters above the valleys, such that the mountain tops are in Petran Montane

PAGE 96

15th International Congress of Speleology Biological Sciences 1357 2009 ICS Proceedings Coniferous Forest and while Sonoran Desertscrub covers the valleys. Geology of this region is complex, with bedrock exposures in the fault-block mountains separated by wide valleys of alluvial deposits. Sedimentary rocks, including limestone and gypsum are present in isolated patches, and igneous and metamorphic rocks are common in the mountain ranges. e northeastern part of the state is in the Colorado Plateau physiographic province, underlain by relatively at-lying Paleozoic and Mesozoic sediments, with some broad folds and volcanic areas producing higher mountains. is region also has desertscrub, grasslands, evergreen woodlands, and coniferous forests, and it includes the only areas of Alpine Tundra in the state. A Transition Zone lies between the Colorado Plateau and the Basin and Range provinces. is zone also has complex geology and very rugged topography. e wide ranges in topography, climate, and vegetation lead to high levels of species diversity across the state in all classes of vertebrates. Caves or abandoned mines are present in all physiographic areas of Arizona, and many vertebrate species are available to take advantage of the special resource conditions provided by these features.2. MethodsData in this analysis were compiled from a variety of sources. Kingsley INGSLEY et al. (2001) provide a reliable list of species known to utilize caves and mines in the Sonoran Desert of southern Arizona, but they did not document species in any specic sites. Other sources, in particular Hoffmeister O FFMEISTER (1986) and Cockrum O CKRUM and Petryszyn ETRYSZYN (1991), provided valuable information, including site records. Several internet sites have extensive information on vertebrate species and their habitat usage and requirements. In particular, the Biotic Information System of New Mexico (BISON 2008) supported by the New Mexico Department of Game and Fish, the heritage data management system supported by Arizona Game and Fish Department (AGFD 2008), and NatureServe Explorer (Nature A TURE Serve ERVE 2008), supported by natural heritage programs. Other standard literature sources were also searched for relevant information. In Arizona, there are no dense concentrations of caves on public lands, and therefore, no le records comparable to those at the Bureau of Land Management (BLM) or Carlsbad Caverns National Park in New Mexico. Caves are relatively rare in the Sonoran Desert, largely because of the limited exposures of limestone or gypsum in this region, but they are more common in other parts of the state. Areas of basalt are also relatively rare in the Sonoran Desert of Arizona, although there are extensive basalt ows in central and northern Arizona. In contrast, abandoned mines are relatively common throughout Arizona and provide underground resources similar to natural caves. Direct observations of vertebrate species were made in many caves and abandoned mines throughout the state. Observational evidence of vertebrate species can take several forms. Visual observations of living animals are the most reliable evidence of species using underground resources. ese animals got into the cave or mine under their own power, which suggests that they made a conscious choice to use the underground site for some purpose. e presence of nests or den sites is also positive evidence of the mammal or bird that constructed the nest, and it also provides evidence of the reason for the use of the underground feature. Similarly, the presence of feathers and egg shells indicates that birds were nesting in the cave or mine, and the condition of the egg shells can indicate successful hatching. e presence of scat in a cave or mine demonstrates that the animal was alive while in the site, and it also indicates that the animal was in the site for a signicant period of time. Tracks indicate the presence of a live animal in the site, but they provide no evidence for the reason for or the duration of the visit. Skeletal material in a cave or mine is more ambiguous. While it can generally be identied to species, it does not necessarily provide conrmation that the animal entered the site by its own choice, or even whether it was alive when it entered the site. Skeletal material at the bottom of a pit generally indicates that the animal died as a result of the fall. Although it may have entered the cave by choice, it probably did not intend to fall into a pit. Data from all of these sources were entered into spreadsheets, including species, sites, biotic communities, dates of observation (if known), and original observers (if known). Biotic communities are based on Brown R OWN (1982). Each report of a species from a cave or mine was entered as a separate record. For common species, particularly bats, there were oen multiple reports from a single site. Site information was compiled in a separate spreadsheet to avoid duplication of sites, particularly with the large number of unnamed, abandoned mines.3. Resultse literature reports, personal communications, and personal observations provided 582 records of 101 species of vertebrates in the caves and abandoned mines of Arizona. Eighty-two of these species are documented with specic locations, and another 19 species are considered conrmed but without specic locations, based on Kingsley I NGSLEY et

PAGE 97

Biological Sciences 1358 2009 ICS P roceedings 15th International Congress of Speleology al. (2001). ese records are primarily in the Sonoran Desert scrub, Semidesert Grassland, and Madrean Evergreen Woodland, which accounted for 480 records. e remaining records are distributed among seven other biotic communities, as listed in Table 1. e 238 sites with vertebrate records include 156 abandoned mines and 82 natural features, including limestone, gypsum, and basalt caves, rock shelters, and sinkholes.3.1 Taxonomic analysise data collected in this research can be analyzed in a variety of way. e rst approach is to sort the records by taxonomic group. e 101 species recorded include 63 mammals, 15 birds, 17 reptiles, and 6 amphibians. No sh have been documented in the caves of Arizona. e distribution of these groups among the ten biotic communities is shown in Table 1. ese vertebrate classes can be broken down further into orders and families. e 63 mammal species are distributed among six orders and 19 families. ese six orders represent all of the native orders of mammals in Arizona. As expected the order Chiroptera (bats) is well represented, with 20 species in three families documented in the caves and mines, but Rodentia (rodents) is the most diverse group with 22 species and ve families documented. Carnivora (carnivores) is also well represented with 13 species in ve families. Artiodactyla (even-toed ungulates), Lagomorpha (rabbits and hares), and Soricimorpha (shrews and moles) complete the list with four, three, and one species, respectively. e 15 bird species are distributed among ve orders and 11 families, with nine species from the order Passerifornes. e 17 reptile species include three orders and seven families. e dominant group of reptiles is in the family Viperidae, represented by six species of rattlesnakes. ree species of turtles have been reported in these underground features. e six amphibian species include two orders, Caudata and Anura, and ve families. e distribution of classes, orders, and families among the dierent biotic communities is generally similar to the overall distribution. e Sonoran Desertscrub, Semidesert Grassland, and Madrean Evergreen Woodland have the greatest numbers of species, but this diversity may only reect the fact that these communities also have the greatest numbers of sites and total records. e strong correlation between species number and the numbers of sites or records suggests that additional research in the other communities would likely increase the numbers of species observed in those areas.3.2 Species distributionA further analytical approach with these data is to examine the distribution of species among the sites. When the data are sorted by species and sites, it immediately becomes apparent that a few common species are found in many sites across a wide range of biotic communities. Some other species are found in numerous sites but may be limited to only one or two biotic communities. However, the majority of species are sparsely distributed, and many are only reported from a single site. is pattern is illustrated in Figure 1, with bars broken down by taxonomic class. From this gure, it is again obvious that very few species are found in large numbers of sites. e species with the widest distribution is the white-throated woodrat (NN eot oma albigula), which was recorded in 46 sites. Seven of the ten most widely distributed species are bats, which is an indication of the intensity of research and availability of data on bats. Biotic Community Number of Sites Number of Records Number of Species MammalsBirdsReptilesAmphibians Sonoran Desertscrub 4118477481115 3 Semidesert Grassland 3967241842 0 Petran Montane Conifer Forest51410820 0 Mohave Desertscrub 913111010 0 Madrean Evergreen Woodland102229322531 3 Interior Chaparral 1023141400 0 Great Basin Conifer Woodland101710801 1 Great Basin Desertscrub 152410820 0 Plains Grassland 11 1 010 0 Great Basin Grassland 5105 221 0 Summary 238582101631517 6Table 1. Distribution of vertebrate species by class and biotic community.

PAGE 98

15th International Congress of Speleology Biological Sciences 1359 2009 ICS Proceedings 3.3 Site diversityAn alternative way to analyze these data is to examine the number of vertebrate species using each cave or mine. ese results are illustrated in Figure 2 for the summary of all biotic communities. e vast majority of sites (almost 83%) have records of only one or two species, and very few sites have more than ve species. e site with the greatest diversity is a small basalt cave in Sonoran Desertscrub, where 11 species were recorded during one site visit with intensive searching. Although fewer data are available for the individual communities, the same pattern is persistent across all communities. 4. DiscussionWhile the diversity and distribution of vertebrates using underground features is of interest, the reasons why vertebrates seek out these resources is even more interesting from an ecological point of view. e simplest use of a cave or mine would be as a temporary roost site. Several species of birds have been observed roosting in caves, particularly owls using caves as daytime roost sites. Similarly, many species of bats use caves or mine as temporary nighttime roosts between foraging bouts. Caves and abandoned mines provide good sites for nests or dens for a wide variety of birds and mammals. In particular, woodrat (NN eot oma sp.) nests are found in many sites in nearly all biotic communities. Caves and mines are also important as maternity sites for many bat species, and the only reported maternity sites for the endangered lesser longnosed bat (LL eptonycteris yerbabuenae) are in caves and mines of the Sonoran Desert (Cockrum O CKRUM and Petryszyn ETRYSZYN 1991). Many vertebrate species are known to use caves or abandoned mines as hibernation sites. Underground features in arid regions are likely to provide water sources for a variety of animals. Caves and mines may provide points of access to groundwater or small perched aquifers in areas where surface water is rare. In addition, caves or mines may provide relief from extreme environmental conditions. In desert regions with very high summer temperatures and very low humidity or in mountainous areas with severe winter conditions, caves and mines could provide sites with moderate temperatures and higher relative humidity. It seems Figure 1. Species distribution among sites. e horizontal axis represents the number of vertebrate species that are found in the number of underground features shown on the vertical axis. e bars are broken down by taxonomic class. Figure 2. Species diversity within sites. e horizontal axis represents the diversity of species within a site and the vertical axis represents the number of sites with that diversity.

PAGE 99

Biological Sciences 1360 2009 ICS P roceedings 15th International Congress of Speleology likely that many vertebrate species would deliberately select these favorable microclimates, although choice might be very dicult to demonstrate. Some species, particularly ringtails, snakes, and raptors, are apparently using caves as foraging sites. It is likely that some of the skeletal material found in caves or mines was carried there by predators. Other skeletal material may represent an animal that merely fell into a pit entrance of a cave or a mine sha and was unable to escape.5. Conclusions and Recommendationse results of this study clearly demonstrate that underground features of many biotic communities of Arizona are being used regularly by a wide variety of vertebrates. In spite of the reliance on random observations and the lack of extensive, systematic surveys of vertebrates in caves, there are reliable records of over one hundred species using a large number of caves and mines. is level of usage and the documented types of usage by these species demonstrate that these underground sites provide a habitat feature that is an important resource for many species. In arid or semiarid environments with extremes of high temperatures and low relative humidity, these sites could be critical to the survival of many vertebrates, including some threatened or endangered species. With so many species depending on the underground features of Arizona, it is imperative that federal and local land management agencies maintain policies that provide protection for these species and their habitat requirements. Under the National Environmental Policy Act (NEPA), federal agencies are required to analyze potential environmental impacts prior to taking any action. Based on the evidence of vertebrate use of these caves and mines, potential impacts on these wildlife species and their habitat requirements must be considered in any action. In addition, when federal or state agencies are giving permits for recreational caving or for scientic research, they should provide the permitees with information about wildlife species using the caves and any precautions they should take when visiting the caves. When mines are to be closed for public safety, surveys should determine the use or potential for use by vertebrates, primarily bats, and specially-designed gates could provide access for bats and other vertebrates while restricting human access.6. ReferencesArizona RIZONA Game AME and AND Fish ISH Department EPARTMENT (2008) Species abstracts, Heritage Data Management System. Arizona Game and Fish Department, Phoenix, Arizona. BISON-M. (2008) Biotic Information System of New Mexico. New Mexico Game and Fish Department and e Fish & Wildlife Information Exchange (Conservation Management Institute, Va Tech, Blacksburg, Va). Available at internet: http://www. bison-m.org/speciesreports.aspx. Brown R OWN D.E., ed. (1982) Biotic Communities of the American Southwest United States and Mexico. Desert Plants, vol. 4, nos. 1. Cockrum O CKRUM E.L. and Y. Petryszyn E TRYSZYN (1991) e longnosed bat, LL eptonycteris: an endangered species in the Southwest? Occasional Papers. Museum Texas Tech University. 142,1. Hoffmeister O FFMEISTER D.F. (1986) Mammals of Arizona. University of Arizona Press, Tucson. 602 pp. Kingsley I NGSLEY K.J., T.R. Strong T RONG E.L. SMITH, and T.K. SNOW (2001) Caves and Mine Adits as Wildlife Resources in the Sonoran Desert Region. Pp. 138 140 in Proceedings of the 15th National Cave and Karst Management Symposium, October 16, 2001, Tucson, Arizona. Nature A TURE Serve ERVE (2008) NatureServe Explorer, an online encyclopedia of life. Available at internet site: http://www.natureserve.org/explorer/. STRONG, T. R. (2006) Vertebrate Species in Caves of the Chihuahuan Desert A Comparison Between Park Service and BLM Caves. Paper presented at the National Speleological Society Convention, Bellingham, Washington, August 7, 2006. Strong T RONG T.R. (2007) Vertebrate Species in Desert Caves and Mines A Comparison Between the Chihuahuan and Sonoran Deserts. Paper presented at the National Speleological Society Convention, Marengo, Indiana, July 23, 2007.

PAGE 100

15th International Congress of Speleology Biological Sciences 1361 2009 ICS Proceedings CAV V E COMMUNITIES IN MISSOURI A COMPARISON OF NUTRIENT RICH AND NUTRIENT POOR SETTINGSMICHAEL ICHAEL SU U TTON ONC C ave RResearch Foundation, 5544 CC ounty RRoad 204, AA nnapolis, MO O 63620 Projects within Cave Research Foundations Ozark Operation have resulted in detailed aquatic census data for two large caves in dierent sinkhole-plain settings: Devils Icebox Cave, Boone County; and Crevice Cave, Perry County. Baseline biological inventories over the past 15 years have meanwhile resulted in a considerable amount of data on community make-up in caves of the Missouri Ozarks, primarily within the Mark Twain National Forest and Ozark National Scenic Riverways. In stark contrast to the sinkhole plain caves, a majority of Ozark Highland caves are highly oligotrophic as a result of their setting within a covered karst, where sinkholes are usually cryptic, and coarse organic material is almost entirely excluded by a deep covering of residuum. Although there is a great deal of overlap in species composition, stream populations are more diverse, and population densities are on average at least an order of magnitude more in the sinkhole-plain caves compared with Ozark caves. Stream population densities are most readily compared by restricting faunal counts to the most favorable habitat type shallow ries with much loose rock, gravel and/ or bedrock and by focusing on C C aecid otea sp. isopods, the most widespread and common group of aquatic invertebrates. Community makeup in sinkholeplain caves also tends to be skewed towards stygophiles and away from stygobites. Some large stygobitic invertebrates such as Kenkia sp. atworms, on the other hand, appear to be conned to the nutrient-rich sinkhole-plain caves. Although CC aecid otea population densities show wide variations, the ranges of measured population densities do not overlap, and the mean density in Ozark Highland Caves is at least an order of magnitude less than those of the sinkhole plain caves.1. Introductionere is a fairly large amount of data on the species makeup of Missouri caves, although this still represents only a rather scanty sampling. About 1000 caves have some sort of biological information on record approximately 15% of the more than 6,400 caves recorded (Elliott, 2007, 2008; Missouri Speleological Survey database). However, many of those caves have had little research, and there has been very limited quantitative or semiquantitative assessment of population densities. Several Cave Research Foundation (CRF) projects over the past two decades have serendipitously allowed some conclusions to be drawn regarding cave stream population makeup and densities in caves in two dierent physical settings: sinkhole plain caves within the Ozark Border province and caves within the covered karst of the Ozark Highlands province (Fig. 1). e longest running of these projects, which continues today, began in 1990 as a response to a short-term management problem for the Mark Twain National Forest (MTNF). A mineral lease application in an environmentally sensitive, karstic area in southern Missouri required development of a detailed Environmental Impact Statement. As part of this process CRF undertook to perform mapping and baseline biological surveys for a group of 50 or so MTNF and privately owned caves within and near the proposed lease area. e project was in part an extension of earlier work by Gardner A RDNER (1986), who had published fauna lists for a large number of publically owned caves, including some within the lease area. Although Gardner made occasional comments as to species abundance and Figure 1: Location of caves contributing to the stream fauna density data. Physiographic proinces adapted om Nelson, 2005.

PAGE 101

Biological Sciences 1362 2009 ICS P roceedings 15th International Congress of Speleology ecology, no systematic attempt was made to quantify populations. e CRF project was likewise focused on obtaining a quick, once-o assessment of the biological makeup of a large number of caves (Sutton U TTON 1993, 1998). Nevertheless, some quantication was attempted in that macrofauna vertebrates and craysh were individually counted and, somewhat sporadically, counts were attempted of terrestrial and aquatic invertebrates. Aer the initial project ended, MTNF elected to continue funding, and to expand the project to other caves within the Forest. MTNF is the largest cave owner in Missouri, with nearly 600 caves currently recorded, and these are distributed throughout the Ozarks. As the project has evolved and matured, simple timed and/ or areal faunal counts have become routine practice (Sutton, unpublished results). Several other Ozarks Highlands caves were assessed as a result of two short-term projects within Ozark National Scenic Riverways (ONSR), Shannon and Carter Counties. One project was closely tied to the prospecting lease application, which also threatened to compromise groundwater quality within ONSR. Several stream caves were inventoried and semi-permanent stream census plots established in gravel/ rock rie habitat (Sutton U TTON 1999). Another project to assess caves subject to high levels of public use also involved some quantitative stream census, generally by timed counts of suitable habitat (Sutton U TTON 2006). Caves within the Ozarks Highland province are for the most part in a covered karst setting the soluble bedrock surface is mantled with a thick layer of residuum consisting largely of clay and insoluble (chert) rock. Open sinkholes are scarce, and cave entrances are generally gravity spring outlets or fossil spring outlets. Water inltrates the subsurface through the insoluble mantle via cryptic, residuum-lled sinkholes and gravel-covered losing stretches of stream bed (Aley L EY 1978). Coarse organic material is almost entirely ltered out, and routine nutrient input consists only of soluble and microscopic material, together with small amounts of coarser material brought into the cave via the downstream entrances by trogloxenes. Periodically, sinkhole collapse occurs by stopping upward through the insoluble cover (Gillman I LLMAN et al ., 2007), injecting a quantity of forest litter into the system, but although such events are relatively common in Missouri as a whole, they are rare within the watershed of any particular cave stream. e streams typically have a clean-washed appearance with little or no particulate organic matter, except near entrances. In 2002, CRF undertook a project in a very dierent cave setting. e project was to develop census protocols for an endemic stygobite in Devils Icebox Cave, a large stream cave in Rock Bridge Memorial State Park within the Ozark Border province of central Missouri (Boone County). e cave occurs in a sinkhole plain with very open water input. e future of the pink planarian (Kenkia glandulosa) was considered questionable owing to water quality issues exacerbated by urbanization of the caves watershed (Frueh RUEH et al., 2007). Although the focus was on the large, relatively scarce atworm, the approach was to put Kenkia within a broader ecological context by counting all stream fauna. Aer trying various approaches, the most eective strategy turned out to be focusing census eorts on those types of stream habitat most likely to be occupied by the atworm. e stream habitat was extremely patchy, with no atworms and little other fauna observed in pooled sections with ne sediments rather, atworms together with stream fauna in general, were strongly concentrated in the relatively limited sections of shallow rocky ries, provided that rocks in the ries were not cemented in place by calcite deposition. Several such areas were set up as semi-permanent census plots, covering 5 linear meters of stream bed. e protocol was to count all fauna within each plot (Sutton UTTON 2004). In 2007, CRF conducted a biological census in another Ozark Border cave in a sinkhole plain in southeastern Missouri (Perry County) (Sutton U TTON 2008). Crevice is the longest known cave in Missouri (> 40 km), and like Devils Icebox Cave, it is in a mixed agricultural and urban setting, leading to serious water quality issues (Pobst OBST and AND Taylor AYLOR 2007). A practical approach to compare population densities between the two sinkhole plain caves and the Ozark Highland caves was achieved by focusing on the group of CC aecidotea sp. isopods. ere is a suite of 15 species in the Missouri subsurface, 14 of which are stygobites or phreatobites. e exception is CC brevicauda, a stygophile which is predominant in the sinkhole plain caves. Being far less troglomorphic than the other species, CC brevicauda is easily distinguished at a glance, even in young, small specimens. e various troglomorphic species are almost ubiquitous in Missouri Highlands cave streams, and with the exception of several small species of extremely limited distribution, they are of comparable size. Only two species, C C antricola and CC salemensis, are widespread and common (Elliott LLIOTT 2008). 2. MethodsMethods were simple and straightforward. Semi-permanent census plots were established by labeling the downstream

PAGE 102

15th International Congress of Speleology Biological Sciences 1363 2009 ICS Proceedings end of each plot, which were 3-5 m in length along a stretch of rie with a high proportion of rock and/ or gravel. e investigator then moved slowly upstream, turning over each moveable rock, siing through coarse gravel, and visually searching areas of open bedrock or ne sediment, while a note keeper recorded numbers of each species encountered. For the wide stream channels typically found in Devils Icebox Cave, a measuring tape was used as a central dividing line so that two counters could work in parallel. e characteristics of the plots were recorded in terms of percentage of each broadly dened substrate, average width, and average depth of water. For the Ozark Border caves, most semi-permanent census plots were counted multiple times over the period of the respective studies. Counts and the resulting population densities were averaged for each plot, and those means were then averaged to obtain an overall mean population density for rocky rie habitat in the cave. Ozark Highland cave census plots were counted only once. Timed counts, usually of 5 or 10 minutes, were also used these were especially useful in situations where available time was limited, or a suitably extensive continuous stretch of rocky rie habitat was not available. In the Devils Icebox study, and for some of the Crevice Cave counts, the areal counts were also timed, so that data from timed counts could be directly compared with the counts by area. e timed counts in Crevice Cave were at quasirandom locations outside the xed census plots; for the Ozark Highland caves the timed counts were also at quasirandom locations. In two caves multiple random counts were conducted; these counts were averaged for each cave, and the resulting means used to calculate the overall mean. Several caves with large gray bat colonies were excluded from the analysis in order to avoid the potential complicating factor of guano input to the system.3. ResultsSee Table 1, Table 2, Table 3, and Table 4 Devils Icebox Cave Crevice Cave Ozarks Highlands caves C C ottus cf carolinae SB? Typhlichthys subterraneus S B shE E urycea longicauda TP E E urycea longicauda TP E E urycea longicauda TP salamanderE E urycea lucifuga TP E E urycea lucifuga TP E E urycea lucifuga TP salamanderE E urycea spelaea TB salamander Kenkia glandulosa SB Sp halloplana evaginata SB (Sphalloplana hubrichti S B) atwormC C aecidotea brevicauda SP C C aecidotea brevicauda SP isopod (CC aecidotea new sp.) SB C C aecidotea antricola SB C C aecidotea sp. SB isopod Bactrurus brachycaudus P B Bactrurus brachycaudus P B Bactrurus sp. PB amphipod (AA llocrangonyx hubrichti SB) amphipodC C rangonyx forbesi SP C C rangonyx forbesi SP C C rangonyx forbesi SP amphipodG G ammarus troglophilus SP G G ammarus sp. SP amphipod Stygobromus s p. SB amphipodO Orconectes virilis SP? C C ambarus sp. SB craysh Physa sp. SP Physa sp. SP snail Fontigens aldrichi SP Fontigens aldrichi S P snailA A gabus sp. SP? (AA gabus sp. SP?) (AA gabus sp. SP?) diving beetle Table 1. Cave adapted stream fauna lists for the two sinkhole plain caves and a generalized list for the total of Ozark highlands caves. PB phreatobite. SB stygobite, SP stygophile, TP troglobite, TP troglophile. Taxa in parentheses are scarce, and single site endemics for Ozark Highlands caves are omitted. P lot area (m2) no. of counts Caecidotea population densities (animals/ m2) s (animals/person-minute) s 1 21 5 65.4 27.5 15.8 6.6 2 15 2 63.5 -- 22.4 -3 12 6 52.4 19.2 8.5 5.0 4 9.5 2 48.2 -- 9.7 -5 9.5 1 54.7 -- 8.7 -Overall population density = 56.8 Caecidotea/ m2, s = 7.4, cv = 13%; Overall population density = 13.0 Caecidotea/ person-minute, range = 4.2-37.8, s = 5.9, cv = 45% Table 2. summary of Caecidotea sp. counts for Devils Icebox Cave. s = standard deviation, cv = coecient of variation.

PAGE 103

Biological Sciences 1364 2009 ICS P roceedings 15th International Congress of Speleology 3. DiscussionTable 1 compares community makeup in the two sinkhole plain caves and the Ozark Highland Caves. Note that in the latter case, all fauna is not represented in all caves the community structure is generally a good deal simpler than the list of potential inhabitants would imply (Sutton U TTON 1993, 1998; Elliott L LIOTT 2008). Stream community structure is skewed towards stygobites and phreatobites in the sinkhole plain caves compared with the Ozark Highlands caves. A possibly stygobitic sh, the grotto sculpin (CC ottus cf carolinae), occurs in Crevice Cave but not in Devils Icebox. e sh is slightly troglomorphic, but not nearly as much as the widespread southern cavesh (Typhlichthys subterraneus) of the Ozark Highlands (Burr U RR et al., 2001). Troglophilic salamanders of two closely related species, EE u rycea lucifuga and EE longicauda, occur commonly in both Ozark Border and Ozark Highland caves, but a troglobitic salamander, EE spe laea, is conned to the Ozark Highlands. Stygobitic isopods (CC aecid otea sp.), common throughout the Ozark Highlands, are partly replaced by the stygophile CC brevicauda in Crevice Cave (32% of > 2000 observations), and almost entirely in Devils Icebox (>99.9%). A new species of stygobitic isopod was discovered in Devils Icebox in the course of the study, but is represented by only a single specimen and a total of two observed individuals. An interesting pattern emerged in Crevice Cave, in that the stygobite strongly predominates in the most upstream reaches of the various streams examined, but is gradually replaced by the stygophile downstream (Sutton UTTON 2008). e main exceptions to stygobitic paucity in the sinkhole plain caves are the large (up to 3.5 cm) triclad atworms. ese two species together with a third, Kenkia lewisi, occur only in Devils Icebox and the Perry County sinkhole plain caves; in the Ozark Highlands, a much smaller species of Sphalloplana occurs very sparsely in widely scattered locations. Phreatobitic amphipods P lot area (m2) no. of counts Caecidotea population densities (animals/ m2) s 1 1.1 5 45.8 23.8 2 1.75 2 16.2 -- 3 2.4 3 23.1 12.5 4 3 3 22.2 15.3 5 3 3 30.1 13.5 Overall population density = 27.5 Caecidotea/ m2, s = 11.6, cv = 42% Random timed counts: n = 28, mean = 7.9 Caecidotea/ person-minute, range = 0.2-30.4, s = 9.2, cv = 116% Table 3. summary of Caecidotea sp. counts for Crevice Cave. Table 4. Ozark Highlands caves included in population density analysis. Figures in parentheses are accession numbers correlating caves to location and other data in the Missouri Speleological Survey database. County Cave Barry Chimney Rock Cave (BRY018) Barry Mushroom Rock Cave (BRY048) Barry Twin Cave (BRY053) Carter Blue Spring Cave (CTR022) Carter Camp Yarn Cave (CTR003) Carter Panther Spring Cave (CTR037) Christian Rattlesnake Cave (CHR013) Douglas Still Spring Cave (DGL036) Howell Crocker Cave (HWL037) Laclede Pittman Cave (LAC041) Oregon Bliss Camp Cave (ORE040) Oregon Blowing Spring Cave (ORE111) Oregon Boze Mill Cave (ORE038) Oregon Chaney Cave (ORE082) Oregon Kelly Hollow Cave (ORE007) Oregon Falling Spring Cave (ORE022) Oregon Pipe Spring Cave (ORE036) Oregon Posy Spring Cave (ORE112) Oregon River Level Cave (ORE055) Oregon Statue Cave (ORE092) Oregon Walters Cave (ORE020) Phelps White Pine Cave (PLP094) Pulaski Onyx Cave (PUL027) Pulaski Peninsula (PUL190) Shannon Bay Branch Arch Cave (SHN150) Shannon Bounds Branch Cave (SHN222) Shannon Bunker Hill Cave (SHN014) Shannon Jam Up Cave (SHN020) Shannon Powder Mill Creek Cave (SHN021) Taney Gilbert Cave (TNY120) Washington Chimney Cave (WSH015) Washington Little Scott Cave (WSH004) Washington Susan Cave (WSH002) Overall population density: n = 23 caves, mean = 2.6 Caecidotea/ m2, range = 0-11.4, s = 3.8, cv = 146% Timed counts: n = 12 caves, mean = 0.6 Caecidotea/ person-minute, range = 0-2.3, s = 0.8, cv = 133%

PAGE 104

15th International Congress of Speleology Biological Sciences 1365 2009 ICS Proceedings various species of Bactrurus are relatively common in both settings. By focusing the population density analysis on CC aecid otea sp. isopods, the problem of comparing species of widely dierent size is obviated. For example, the tiny troglophilic snail Fontigens aldrichi is numerically the dominant stream animal in Crevice Cave, but is absent from Devils Icebox, and occurs only sporadically throughout the Ozarks. Since Fontigens is a good deal smaller than CC aecid otea, including the snail in the analysis would require a problematic correction for biomass, for which data is lacking. Large animals (sh, craysh, salamanders) rarely turn up in census plots or random counts, but if there is signicant predation pressure on isopod populations from these animals, it could signicantly reduce isopod densities, and may be one reason for the high variability discussed next. Ideally, one would like to compare biomass densities, but this would be a dicult undertaking data on the masses of Ozark cave stream animals is lacking, and it is not obvious how one would adequately assess densities for large, rare animals. Comparing population densities has a number of obvious pitfalls. e habitat is extremely patchy, and even by excluding all habitats except supercially similar rock/ gravel ries, species densities vary greatly. ere is also considerable temporal variation, as shown by repeated counts of the same plots in the sinkhole plain caves, where the coecients of variation for isopod density (cv = s/mean x 100%) for multiple counts of the same plot are typically in the range of 50% (36%-69%). Overall coecients of variation for areal counts in the sinkhole plain caves are 13% for Devils Icebox Cave and 42% for Crevice Cave. By comparison, Ozark Highland caves are much more variable, with counts of zero animals being fairly common, and a cv for the entire set of areal observations of 146%. (Zero density does not imply an absence of isopods from the cave, just from a particular plot). e wide variation potentially makes it dicult to draw conclusions regarding the relative abundance of isopods in the dierent types of cave. However, the means population density in Ozark Highland caves diers by more than an order of magnitude from those of the sinkhole plain caves, and even with such a wide variability, the ranges of areal isopod densities in the three cases do not overlap: Oza rks Highlands caves: 0-11.4 isopods/m2Crevice Cave: 16.2-45.8 isopods/m2 Devils Icebox Cave: 48.2-65.4 isopods/m2e Devils Icebox study showed that the relative abundance of CC aecidotea between plots was similar whether plotted by area or by eort. However, Tables 2 and 3 show that the coecient of variation is considerably higher in the sinkhole plain caves for timed versus areal counts. e dierences between the two settings are therefore less clear when population densities by unit search eort are compared, but the range for the Ozark Highland caves still dier greatly (Tables 2-4). It is therefore very likely that isopod abundance is genuinely less in Ozark Highland caves. It is also at least somewhat probable that Devils Icebox Cave densities are signicantly higher than those of Crevice Cave. e relative scarcity of cave stream fauna in the Ozark Highlands is not an unexpected result, given the hydrological dierences between the two settings open sinkhole input in the one case and a combination of diuse and highly ltered discrete input in the other. In both sinkhole plain caves, and in strong contrast to the cleanwashed streams of typical Ozark Highland caves, coarse and ne particulate organic matter are conspicuous components of the stream bed material. Moreover, all of the Ozark Highland caves examined are in relatively undeveloped oak/hickory forest in areas with low human population density. e sinkhole plain caves are both in urban-edge situations with intensive agricultural activity within the caves watersheds. Although the Ozark Highland have been subject to considerable post-settlement alteration, including a period of widespread deforestation, it is safe to assume that the sinkhole plain cave streams are more severely aected by post-settlement activity. Unfortunately, there are few or no sinkhole plain caves in the Ozark Border province which have not been subject to the eects of agriculture and/ or urbanization. It is therefore impossible to deduce how much of the dierence in population densities is due to human alteration of the environment and how much to the geohydrological dierences between the caves. Within the Ozark Highlands, stream caves with major gray bat (Myotis grisescens) colonies form a special case, in that the resulting guano might render the streams a good deal less oligotrophic. However, the small amount of data on the isopod densities in bat caves collected during these projects falls well within the range of non-bat cave streams (Sutton UTTON 1993, 1998). In one case where a large colony roosts almost entirely over the water (Bat Cave, Ozark County), the population of twilight zone crustaceans does seem subjectively high, but the eect does not extend into the dark zone, which is upstream from the bat roost (unpublished data). In Turner Spring Cave (Oregon County), and again rather subjectively, the population density of isopods is comparable upstream and downstream

PAGE 105

Biological Sciences 1366 2009 ICS P roceedings 15th International Congress of Speleology from the main bat roost. Possibly, bat guano falling into a cave stream is leached of nutrients rather rapidly, but a good deal more data on bat cave stream communities is required. In conclusion, the analysis here presented attempts to make a start at putting hard numbers on Missouri cave stream population densities, an aspect of Missouri cave stream ecology which has been largely lacking, and demonstrates numerical dierences in the densities of two very dierent cave settings. Acknowledgments Funding for the various projects were provided by Mark Twain National Forest, Ozark National Scenic Riverways, Missouri Department of Natural Resources, and Missouri Department of Conservation. e projects here reported would have been impossible without copious amounts of eld-work time volunteered by CRF members. While the list of helpers is much too long to reproduce here, I must single out the two most prolic and longest-term workers, Sue Hagan and Scott House.ReferencesBURR, B.M., G. ADAMS, J. KREJCA, R. PAUL, and M. WARREN Jr (2001) Troglomorphic sculpins of the CC ottus carolinae species group in Perry County, Missouri: distribution, external morphology, and conservation status. EE n ironmental Biology of Fishes 62, 279. ELLIOTT, W. R. (2007) Zoogeography and Biodiversity of Missouri Caves and Karst. J ournal of CC ave and Karst Studies 69, 135-162. Elliott, W.R. (2008) Cave Life Database. Missouri Department of Conservation. Gardner, J.E. (1986) II n ertebrate fauna om Missouri CC aves and Springs. Missouri Department of Conservation, Natural History Series no. 3, Jeerson City, MO, 72 pp. FRUEH, T., R. CAMPBELL, and R.N. LERCH (2007) Successful watershed planning by stakeholders, Boone County, Missouri. Proceedings of the 18th N N a tional CC ave and Karst Management Symposium, St. Louis, pp 144 GILLMAN, J., J. PALMER, G. YOUNG, and J. PREWETT (2007) Soil-cover Karst Collapse: A Geological Hazard in Missouri, Proceedings of t he 18th NN ational CC ave and Karst Management Symposium, St. Louis, pp 175. Nelson, P W. (2005) T errestrial NN atural CC ommunities of Missouri revised edition. Missouri Department of Natural Resources, Jeerson City, 550 pp, p 4 Pobst, B. and S.L. TAYLOR (2007) Water uality Monitoring in the Perry County Karst, Missouri, P roceedings of the 18th NN ational CC ave and Karst Management Symposium, St. Louis, pp 68. Sutton, M.R. (1993) Caves and cave wildlife in a mineral prospecting area, Oregon and Shannon Counties, Missouri. Missouri Speleology, 33, pp 1. Sutton, M.R. (1998) Baseline Mapping and Biological Inventory of Caves on the Mark Twain National Forest Doniphan-Eleven Point District, Missouri: Phase 2. Report to Mark Twain National Forest, 105 pp. Sutton, M.R. (1999) Inventory of the aquatic and ood zone macrofauna communities of certain caves within Ozark National Scenic Riverways, Carter County, Missouri. Report to Ozark National Scenic Riverways, 20 pp. Sutton, M.R. (2004) e Pink Planarians of Devils Icebox Cave census protocols. Report to Missouri Department of Natural Resources and Missouri Department of Conservation, 34 pp. Sutton, M.R. (2006) Baseline biological inventory to manage public use caves at Ozark Riverways, 20052006. Report to Ozark National Scenic Riverways, 20 pp. Sutton, M.R. (2008) Crevice Cave Biology a Cave Research Foundation Study, 2006-2008. Report to Missouri Department of Conservation, 42 pp.

PAGE 106

15th International Congress of Speleology Biological Sciences 1367 2009 ICS Proceedings SPATIAL AND TEMPORAL DISTRIBUTION OF TERRESTRIAL MACROINV V ERTEBRATES IN LEHMAN CAV V ES, A TOURIST CAV V E IN GREAT BASIN NATIONAL PARK, NEV V ADA, USASTE E VEN EN J. TA A YLOR LOR1, JEAN EAN K. KRE RE JCA CA2, MICHAEL ICHAEL E E SLA LA Y3 1DD ivision of Biodiversity and EEcological EE ntomology, II llinois NN atural HH istory Survey, 1816 S. OOak St., C C h ampaign, ILIL 61820 UU SA A 2ZZ ara EE nironmental, LLCLLC 118 W. GG oforth RR d., Buda, TX 78610 UU SA A 3OO zark HH ighlands OO ce, e NN ature CC onservancy, 675 NN LLollar LL n., Fayetteville, ARAR 72701 UU SA A mslay@tnc.org Abstract Monthly macroinvertebrate counts and measures of physical and environmental parameters were used to study the biology and distribution of cave species in Lehman Caves (White Pine County, Nevada), which averages more than 30,000 visitors per year. Counts at paired bait stations (near-trail and far-from-trail) demonstrated that diversity and abundance drop o with increasing distance into the cave. is pattern is correlated with changes in environmental parameters (2 cm soil temperature, air temperature, humidity, lower available nutrients), but also with human visitation levels which decrease with increasing distance into the cave. Analysis of near-trail vs. far-from-trail bait stations showed no consistent dierences in abundance and diversity. Measures of physical and environmental parameters in 30 study plots showed an expected distribution of troglophiles nearer to the entrance and troglobites far from the entrance. We were not able to demonstrate that tourist trails in Lehman Caves are aecting the diversity or abundance of the fauna, although two Park-endemic troglobites were more abundant at low impact sites and far from trails. Microcreagris grandis (Pseudoscorpionida: Chelonethida: Neobisiidae), presently classied as a troglobite, was found much more commonly proximal to the entrance than in more remote areas of the cave, suggesting possible reclassication as a troglophile. Our results point to the importance of maintaining healthy, low impact areas near entrances of commercial caves.

PAGE 107

Biological Sciences 1368 2009 ICS P roceedings 15th International Congress of Speleology

PAGE 108

CONTRIBUTED PAPERS IN C ONSERV V ATION AND MANAGEMENT

PAGE 110

15th International Congress of Speleology Conservation and Management 1371 2009 ICS Proceedings DOCUMENTATION OF CAV V ES AND KARST IN WESTERN AUSTRALIARR Oss SS A A NDERs S ON WA A Speleological GG roup (WA A SG G ) PO O Box 443, CC loerdale 6985, Perth, WA A. AA ustralia, rossjay@iinet.net.au Abstract In Western Australia, speleologists work in partnership with land managers in relation to the conservation, management and protection of karst systems. is paper will discuss two projects that have involved collaboration and partnerships between land managers and speleologists. In 2006, the WA Speleological Group (WASG) received funding to undertake research on Christmas Island, o the WA northern coast. en, in 2007, the WASG participated in a project with the State Government, undertaking local research in karst north of Perth. Both of these projects involved improving land manager awareness of karst systems and important biological values subterranean fauna and habitats. e Christmas Island Expedition, involved a team of about 10 speleologists undertaking research on the subterranean fauna of Christmas Island. Christmas Island is located o the north coast of WA, and although considered the legal responsibility of WA, it is managed by Parks Australia North. e team were on the island for a month, visiting a number of caves on a regular basis, documenting fauna and collecting specimens that were then examined by the WA Museum. e team visited and relocated all known karst features and also discovered and described several new karst features. e HH idden Treasures Project studied the biodiversity of the Northern Agricultural Region (NAR). Caves and their associated biodiversity were one type of natural asset being considered. e 2007 Subterranean Biodiversity Inventory was a key component of the HH idd en Treasures Project. e project aimed to collect information to allow the highest priority conservation areas in the NAR to be identied and to assist in the development of a Biodiversity Conservation Plan. e study was funded by the Northern Agricultural Catchment Council (NACC) and delivered by the Department of Environment & Conservation in association with the Australian Speleological Federation (ASF). Fauna sampling within caves focussed on invertebrate populations and revealed a diverse assemblage of organisms. Several teams of speleologists assisted in eld trips over the year.

PAGE 111

Conservation and Management 1372 2009 ICS P roceedings 15th International Congress of Speleology Preliminary RELIMINARY studies STUDIES on ON soil SOIL erosion EROSION intensity INTENSITY grading GRADING in IN southwest SOUTHWEST karst KARST area AREA China HINACC AO JIANHUA1 JIANG Z Z HONGCHENG1 YANG D D Es S HENG2 PEI JIANGUO1 YANG H H UI1 L L UO WEIq Q UN1 1Karst II nstitute of GG eology, CAGCAGS, Key LL aboratory of Karst DD ynamics, MLR LR 541004, GG uilin, GG uangxi, CC hina2Pear RR iver Water RResources CC ommission, MWR R 510611, GG uangzhou, GG uangdong, CC hina e concentration of insoluble matter in carbonate rocks is very low in southwest China. Consequently, soil formation is slow and soil layers are usually thin on karst slope. e results from remote sensing and monitoring sites show that soil loss intensity is mostly weak and light. In fact, the area of rock desertication in karst regions, is continuously increasing, with the soil eroding and the ecological situation becoming worse. e traditional standard for soil erosion gradation is confronted with challenge. Risk valuation and intensity classication of soil erosion in karst region should be redened. Depending on the factors of limestone formation and carbonate rock dissolution, related data and images have been collected. With the help of ArcView3.2, the rate of limestone soil formation can be produced. By also considering a soil loss tolerance, a new standard for soil erosion gradation suitable to karst region can be put forward. e duration of feeble, light, middle, strong, very strong, and acute soil loss gradation, is <30, 30~100, 100~200, 200~500, 500~1000, and >1000t/(km2a), respectively. On the basis of features of soil loss processes and the karst environment, four suggestions for soil protection in karst region are brought forward: (1) depending on the karst hydrological structure, the key part to prevent the soil loss from surface to ground, is the sinkholes and dolines, and eective measurement for soil protection should be taken; (2) on karst hill, in top part, the carbonate rock is largely exposed with little soil cover, the way to close hillside to facilitate aorestation can be selected; the middle slope part, the rock desertication is middle gradation, to build the wall and prevent the soil loss, and planting the economic plants to increase the land productivity, should be suitable measurement; in the foothill part, thick soil layer and little rock, the farmer land construction, modern agriculture activity and food production are taken; (3) in the karst region with thin soil layer, biogenic fence technology is advocated, shrub species are recommended, particularly shrubs belonging to the C4 plant with high photosynthesis capacity and bean family with nitrogen xation. Moreover, they have ourishing roots to bear drought and poor habitat; (4) limestone soil is characterized by high calcium content and alkaline. e high calcium level activates many nutritional functions. e eective soil improvement can enhance agricultural productivity.1. Introduction In the factors that aect ecosystem formation and evolution, the climatic and hydrological elements are long-term and slow driving forces shaping the ecosystem, while geological and geomorphological factors are the carrier and material foundation of ecosystem existence and development (LIU Yanhua et al., 2001). Most of the parent rock of soil formation are hard and thick carbonate rock in the southwest of China, and the concentration of acid insolvable matter in carbonate rock is very low, so the speed of soil formation is very slow (YUAN Daoxian et al., 1988). e depth of soil layer is usually only 10 to a few dozens centimeters, and the soil drains away to become rock desertication aer a few heavy rains without covering vegetation. Aer evaluating the harm of soil erosion, the new standard for soil erosion gradation that is suitable to karst regions should be put forward. According to the scientic data collected during scientic surveys on soil erosion and ecological safety conducted by the Ministry of Water Resources, the Chinese Academy of Sciences, and the Academy of Chinese Technology in 2006, and mechanism studies on karst dynamic systems, the preliminary standard for soil erosion gradation suitable to the southwest karst region has been put forward.2. Process and Impacting-Factor of Soil FormationWang ANG et al.s (1999) research data on five limestone soil profiles in Guizhou and Hunan provinces, showed the geochemistry process of carbonate rock being eroded to form limestone soil following two stages: carbonate rock dissolution and acid insolvable matter transform to soil.2.1 Content of acid insolvable matter in carbonate rocks

PAGE 112

15th International Congress of Speleology Conservation and Management 1373 2009 ICS Proceedings ere are much data showing that carbonate rock dissolution is impacted by rock and mineral types and structure and chemical component. For example, in the Guilin karst area, the carbonate rocks are the pure and thick layer limestone rock in the Devonian through Carboniferous Periods. e results of carbonate rock dissolution experiment on samples from 103 dierent strata indicate that in the pure carbonate rocks, dissolution rate has positive correlation with the CaO content in the rocks and negative with MgO content; in the impure carbonate rocks, the physical destruction increased due to the addition of acid insolvable matter, but the rate of chemical dissolution reduced (Weng et al., 1987). Nie (1994) collected 124 carbonate rock samples from Cambrian, Ordovician, Devonian, Permian, and Triassic beds and the results of carbonate rock dissolution experiments show that in the pure carbonate rock, the relative dissolutional rate decreased as Mg content increasing, from 0.96 for pure limestone to 0.50 for pure dolomite; in the impure carbonate rock, the relative dissolutional rate decreased as acid insolvable matter content increasing, from 0.96 to 0.52 for limestone with dierent acid insolvable matter content, and from 0.50 to 0.26 for that of dolomite.2.2 Condition of hydrodynamic and hydrochemistryCarbonate rock dissolution is a typical interaction between water and rock. ere is a diusedboundary level (DBL) (Liu et al., 2000a) between solid carbonate rock and water interface. As the rate of the ow increasing, the DBL becomes thin, and the carbonate rock dissolution rate increases. Liu et al. (2000b) conducted a test to quantify the allogenic water erosion capability at Yaoshan, Guilin. e dissolutional rate of pure limestone of Rongxian group in Devonian was 1,000,500 mm/ka, 10-15 times higher than that of limestone water. Meanwhile, it was proven that the dissolution rate of pure limestone in ow-water at the speed of 20 cm/s and 60 cm/s was two and six times that of non-ow water, respectively.2.3 Temperature and precipitation Temperature and rainfall are the most important factors on the carbonate rock dissolution because the amount of rainfall inuences the hydrological and runo conditions and the temperature inuences bioactivities and exchange rate between water and CO2. Most of the prevailing mathematic models were made in the past. Liu (2000b) established linear correlation between carbonate dissolution rate and runo:D DR=0.0544(P-E)-0.0215 r=0.98 where P is rainfall(mm), E is evaporation(mm). Pulina(1974)established the relationship between dissolution rates, temperature, and rainfall utilizing a large set of data. e relationship is that the inuence of the rainfall change on dissolution rate will be weak at lower temperatures, but the dissolution rate would increase rapidly with the increase of rainfall at higher temperatures (1620oC).2.4 Biological activitiesIn the earth surface system, the biosphere links up the entire system by exchanging and transferring matter and energy with adjacent spheres (Golubic et al., 1978; Zhang, 1992). Organisms are one of the most active geological agent in the Earth Surface System (Yin, et al., 1994). Organisms could stimulate carbonate rock dissolution, mainly through the metabolic activities of vegetation and microorganisms to produce high concentration of CO2 and erosive secretion. e results of the simulated test by Cao et al. (2004) show that dierent vegetation covers lead to carbonate rock dissolution dierence. e dissolution rate under the system of tree-soil-rock with ourishing root is 3.84 times higher than that of soil-rock and 2.36 times higher than that of herb-soil-rock. Jennings (1985) found that soil-vegetation cover enhanced the carbonate rock dissolution when he collected data and established the relationship between the dissolution rate and the average annual runo from dierent countries. 3. Estimation of Soil Formation Rate 3.1 MethodsCarbonate rock soil formation closely correlates with carbonate rock dissolution and its content of acid insolvable matter. erefore, a mathematical regression equation could be established depending on the limestone dissolution rate of typical sites. e collection impact-factors on limestone soil formation including lithology, temperature, precipitation, net primary productivity of vegetation, and soil respiration. We calculated the spatial distribution of carbonate rock dissolution rate with the regression equation by comprehensive inuence from the dierence of carbonate rock lithology and the hydrological conditions. And then, we estimated the soil formation rate of the carbonate rock by adding content of acid insolvable matter and dissolution rate together.

PAGE 113

Conservation and Management 1374 2009 ICS P roceedings 15th International Congress of Speleology 3.2 Data collection and processing e data were collected from the region of southwest China karst, including east of Yunnan, Guizhou Province, and Guangxi Province. e soil area is 553 thousand km2, the limestone area is 2555 thousand km2. Lithological type, combination and distribution: Carbonate rock types and distribution are shown on the geological map along with detailed stratum descriptions (Fig.1). e gure shows that continuous limestone and limestone-dolomite are mainly distributed in the east and southeast of Yunnan, the southwest of Guizhou and Guangxi; dolomite are mainly in the northeast of Guizhou, meanwhile, the dolomite usually interbedded with limestone and clastic rocks; impure limestone in the northeast of Yunnan. Content of acid insolvable matter: e most data on the composition of acid insolvable matter in carbonate rocks are from Institute of Geology, the Chinese Academy of Sciences(Karst Research Group from the Chinese Academy of Sciences, 1979). e carbonate rock in the east of Yunnan are Permian and Triassic limestones eir composition data for acid insolvable matter came from Wang et al. (2003). e carbonate rock in the south of Guizhou Province and Guangxi Province karst region are pure, blocky, thick limestone from the Carboniferous and the Devonian Periods. Data on their acid insolvable matter composition came from Weng et al. (1987). Temperature and precipitation and hydrological condition: Temperature and precipitation data are from the average temperature and precipitation grid image of 19932000, provided by the Chinese Academy of Sciences. In the research region, there are 3 dierent hydrological zones from west to east: Yunnan-Guizhou Plateau, Guangxi Basin and the slopes transition zone between them. (1) Yunnan-Guizhou Plateau: Gradient ratio of surface and underground rivers is small, with slow water ow, thick DBL. erefore, the intensity of carbonate rock dissolution decreases. (2) e slopes transition zone: Gradient ratio of surface and underground rivers is large, the altitude has a big dierence, water cycle is fast in both vertical and horizontal directions, and with thin DBL. Karst processes are enhanced. Figure 1: Geologic map of research area showing carbonate rock outcrop.

PAGE 114

15th International Congress of Speleology Conservation and Management 1375 2009 ICS Proceedings (3) Guangxi Basin: the basic condition for karst processes is characterized with low elevation, gentle water ow gradient, and long time of waterrock interaction. As with high precipitation, the water ow is fast with thin DBL. e karst process is also accelerated. In the research region, limestone, dolomite, and clastic rocks are interbedded in the middle and the northeast of Guizhou Province; impure limestone and clastic rocks are interbedded in northwest Guizhou Province and in northeast Yunnan Province. Allogenic water from clastic regions with low-pH strongly accelerates carbonate rock dissolution. Biological activities: e biological activities on carbonate RegionSiteP/mmT/oCNpp/gCm-2a-1Sr/gCm-2a-1Ldr /mmka-1East of Yunnan Mengzi105818.1 316.5 1179.2 41.1 Mile88416.5 244.8 1365.8 30.0 ujing103014.6 190.0 1352.4 45.0 Luoping149215.1 202.8 1498.1 65.0 Huyuan133215.0 190.4 1446.2 50.0 Lunan93215.3 219.6 1344.1 32.0 Tonghai88015.9 196.3 1339.6 30.0 Guizhou Libo119417.9 1116.2 309.4 65.5 Qinglong140714.5 268.3 1445.8 55.0 Anlong126016.1 203.7 1502.5 60.0 Ziyun120316.0 194.5 1487.5 50.0 Guanling134215.7 208.4 1496.6 50.0 Zhengfeng129016.6 229.6 1527.4 70.0 Luodian118218.9 222.6 1605.7 55.0 Guiyang111214.9 248.2 1398.1 44.5 Wuchuan109416.0 245.0 1438.1 36.1 Guangxi Huanjiang133819.1 313.3 911.5 59.1 Yongfu170219.0 498.1 1759.9 60.0 Yizhou132120. 2 304.4 1733.9 55.0 Liucheng137220.0 421.5 1733.5 55.0 Liuzhou136220.6 398.8 1745.5 56.0 Tiane137719.9 226.9 1721.4 50.0 Duan149620.9 264.7 1814.5 80.0 Longlin115818.6 297.1 1577.6 70.0 Mashan154221.4 344.5 1842.1 80.0 Shanglin155921.0 365.7 1845.8 65.0 Luzhai148119.7 474.7 1747.8 54.0 Xincheng137920.8 283.9 1778.9 60.0 Gongcheng145219.4 470.5 1718.6 60.0 Rongan162219.2 455.6 1755.3 70.0 Rongshui143818.8 422.5 1671.9 70.0 Nandan135519.4 211.9 1708.9 50.0 Guilin170318.9 383.0 1753.3 89.7Note: the data of the table is from Jiang et al. (2000); Zhu et al. (2008); Fang et al. (1993 Table 1: Limestone dissolution rate and relative temperature, precipitation, net primary production of vegetation and soil respiration rate of typical sites in southwest karst region, China.

PAGE 115

Conservation and Management 1376 2009 ICS P roceedings 15th International Congress of Speleology rock dissolution are mainly through production of high CO2 concentration in soils and secretion of erosive organic matters, such as organic acid, enzyme and chelate compound (Cao et al., 2000a; 2000b; Liu, et al., 2001). e inuence also can work via biophysical and biochemical processes (Cao et al., 1999). Soil respiration (Sr) is a good index for ecosystem metabolic activity including plant roots respiration, soil microorganism respiration and oxidation, and decomposition of soil organic matter (Pan, 2001). e net primary production of vegetation (Npp) is the amount of organic matter through photosynthesis by producers, minus organic matter consumed by plant metabolism. What remains can be used for plant growth and reproduction, and by other animals and microorganisms. e Npp of vegetation is the important index maintaining and promoting the progress and evolution of ecosystem (Sun et al., 1993). e images of soil respiration and the Npp of vegetation were provided by Institution of Geography Science and Resource, the Chinese Academy of Sciences (Company of Zhongkeyongsheng, Beijing). Data collection from typical sites: e data of limestone dissolution rate (Ldr) were collected from the research area and the relative data including the precipitation(P), temperature(T), net primary productivity(Npp) of vegetation, and soil respiration(Sr) were obtained (Table 1).4. Soil Tolerant Loss and Erosion Intensity Gradatione soil tolerant loss was calculated by the process as follows.4.1 Producing the image of limestone dissolutionAccording to the data from Table 1, the regression equation was calculated to a condence level of 95%, with the help of EXCEL2003.D DR=0.038P+0.071NNpp-2.758T-0.029Sr-48.002 (r=0.66) Where DDR is dissolution rate (mm/ka); P is annual average precipitation (mm); T is annual average temperature(oC); Sr is soil respiration (gC/m2a), NNpp is net primary production of vegetation(gC/m2.a). With the help of ArcView3.2, the image of limestone dissolution rate can be produced.4.2 Image of soil formation ratee soil formation rate was inuenced by carbonate rock type and the content of acid insolvable matter. e dissolution rate of limestone is about two times higher than that of dolomite. But, the dolomite readily weathers Figure 2: Limestone soil formation rate in research area.

PAGE 116

15th International Congress of Speleology Conservation and Management 1377 2009 ICS Proceedings physically to produce dolomite sands, which can become the mechanical component of soil. e rate of dolomite physical weathering is 4-5 times higher than that of limestone. erefore, under the same condition of acid insolvable matter and others, the rate of soil formation in dolomite terrain is 2-2.5 times higher than that in limestone areas. Comprehensively considering the inuence of stratication of dolomite and limestone and allogenic water from clastic rock area, the rate of soil formation in the middle-northeast Guizhou, 2.5 coecient was added. And then with the help of ArcView3.2, the image of the soil formation rate was produced with the grid image of acid insolvable matter folded to the grid image of limestone dissolution (Fig.2). Figure 2 shows in northeast Guizhou, where the conditions of soil formation are relatively superior, the maximum soil formation rate is 40~120t/(km2a), while it is just 4~20 t/(km2a) in more than 70% of the other study areas. e average rate of 30~40 t/(km2a) is dened as the soil tolerant loss. Li et al. (2006) estimated the soil tolerant loss in karst region of Guizhou as 6.84~103.46 t/(km2a).4.3 Soil erosion intensity gradation According to SL 190-1996 Standard of Soil Ersion Gradation (SSEG), the soil tolerant loss of dierent types of erosional areas are 200 (northeast blackland region and north soil-rock region) to 1000(northwest plateau) t/(km2a). Based on this standard, the lower limits of light degree are dierence for various soil erosional regions, but the upper limit all are 2500 t/(km2a). is standard could not be adopted in southwest karst area. us, a new standard suitable to a karst region is needed. Depending the above results, the duration of feeble, light, middle, strong, very strong and acute soil loss gradation is <30, 30~100, 100~200, 200~500, 500~1000, and >1000t/(km2a), respectively. e gradation is compared with past standards (Table 2).5. Evaluation on Soil Erosion and Suggestione rain water goes to underground directly through the sinkholes and dolines, on account of the hydrogeologic structure with upper-under ground double layers. e typical soil erosional process starts with soil falling to underground rivers then emerges back in a surface river. While in the non-karst region, the surface river system developed. e typical soil erosion is that the soil is eroded down the hillside surface with water ow, then goes into the surface river directly. 5.1 New requirement of soil erosion hazard evaluation Beyond soil fertility declined, soil layer thinning, and river channel blockage, soil erosion in the southwest karst region of China causes rocky desertication formation. Consequently, evaluation of soil erosion hazards in karst region should be worked out in two stages: 1) on the premise of soil-covered karst, evaluated with the soil erosion coecient; 2) when gave soil loss has occurred and rock is exposed (shi-ga-la land), evaluated with the ratio of the rock desertication area.5.2 SuggestionOn the basis of the particulars of soil erosion processes and mechanisms in karst regions, we put forward countermeasures for soil conservation. Four suggestions for soil protection in karst region are brought forward: (1) Depending on the karst hydrological structure, the key part to prevent the soil loss from surface to ground is at sinkholes and dolines, and eective measurement for soil protection should be taken. (2) On karst hill tops, where the carbonate rock is largely exposed with little soil cover, closing hillsides to facilitate aorestation can be helpful; the middle slope part, the rock desertication is middle gradation, to build the wall and prevent soil loss, and planting economic plants to increase the land productivity, should be suitable measurement; in the foothill part, thick soil layer and little rock, farm land construction, modern agriculture activities, and food production are conducted. Gradation feeble light middlestrongvery strong acute SSEG<200,500,1000200,500,1000~25002500~50005000~80008000~15000>15000 Cai(1989) <68 68~100 100~200200~500 Wei(1996) <50 50~100 100~200200~500500~1000 Wan, et al(2003)<46 46~230 230~460460~700700~1300 authors <30 30~100 100~200200~500500~1000Table 2: Preliminary standard of soil erosion intensity in southwest karst region, PR China, t/(km2a).

PAGE 117

Conservation and Management 1378 2009 ICS P roceedings 15th International Congress of Speleology (3) In the karst region with thin soil layer, biogenic fence technology is advocated, shrub species are recommended, particularly some shrubs belonging to the C4 plant with high photosynthesis capacity and bean family with nitrogen xation. Moreover, they have ourishing roots to bear the drought and poor habitat. (4) Limestone soil is characterized by high calcium content and alkaline. e high calcium level activates nutritional productivity. e eective soil improvement can enhance agricultural yield.Acknowledgmentse authors thank Zhu Minjie to help preparing the paper. is study were supported by the projects Rehabilitation and reconstruction on degenerated eco-system in Chongqing karst region with series No. 2006BAC01A16, Scientic Survey on soil loss and ecological safety in southwest karst region with series No. 2005SBKK05, and Key Laboratory Cultivation on Guangxi Karst Dynamics Laboratory with series No. GuiKeNeng0842008ReferencesChai Zhongxin (1989) Soil erosion in karst area of Guangxi autonomous region. M ountain RResearch 7 (4), 255. Cao Jianhua, Yuan Daoxian (1999) Relationship between water-holding of carbonate rock and saxicolos algae, lichen, and moss and its ecological signication. G G eo chemistry28 (3), 248. Cao Jianhua, Yuan Daoxian, Pan Genxing et al. (2001a) Preliminary study on biological action in karst dynamics system. EE art h Science Frontiers8 (1), 203. Cao Jianhua, Yuan Daoxian, Pan Genxing (2001b) Simulation experiment on dissolution of calcite by citric and its karst importance. C C a rsologica Sinica 20 (1), 14. Cao Jianhua, Yuan Daoxian, Pan Genxing et al. (2004) Carbon cycle of karst dynamics system inuenced by soil carbon transfer covered with various vegetation. E E art h and EE nironment32 (1), 90. Fang Jingfu, Lin Junshu, Li Juzhang et al (1993) Relation of solution to environment in karst areaA case study of Hong-shui River Basin. AA ct a GG eologica Sinca 48 (2),122. Golubic S (1978) Carbonate Dissolution. In e Biology of C C y anobacteria. N.G. Carr and B.A.Whitton (eds.), Oxford Blackwell Scientic Publications Ltd, pp 107. Jennings J.N (1985) K arst GG eomorphology. Oxford Basil Blackwell Ltd., London, 238 pp. Jiang Zhongcheng, Jiang Xiaozhen, Lei Mingtang (2000) Estimation of atmospheric CO2 sink of karst areas in china based on GIS and limestone tablet loss data. C C ar sologica Sinca19 (1), 1. Karst Research Group from the Chinese Academy of Sciences (1979) K arst Study of CHINACHINA. Science Press, Beijing, 336 pp. Li Yangbing, Wang Shijie, Wei Chaofu et al (2006) e spatial distribution of soil loss tolerance in carbonate area in Guizhou province, EE art h and E E nironment34 (4),36. Liu Yanhua and Li Xiubin (2001) F ragile EEco-enironment and Sustainable DDevelopment. e Commercial Press, Beijing, 471 pp. Liu Zaihua (2000a) Field experimental research on the corrosion kinetics of limestone and dolomite in allogenic waterCase from Yaoshan Mt., Guilin. C C ar sologica Sinca19 (1), 1. Liu Zaihua (2000b) Contribution of carbonate rock weathering to the atmospheric CO2 sink. C C arsologica Sinca19 (4), 293. Nie Yueping (1994) Karst development Caracteristics under the Lithological control of carbonate rocksA case study in south-central Guizhou. CC ar sologica Sinca 13 (1), 31. Pan Genxing (2000) P edology in the EE arth Surface System. Geological Publishing House, Beijing, 144 pp. Pulina Marian (1974) Preliminary studies on denudation in SW Spitsbergen. B iulletin de lA A cademie Polonaise de Sciences, Serie de Sciences de la Terre22, 2. Sun Ruyong, Li Bo, Zhu Geyang et al (1993) GG e neral E Ecology. High Education Publishing House, Beijing, 324 pp.

PAGE 118

15th International Congress of Speleology Conservation and Management 1379 2009 ICS Proceedings Wan Jun, Cai Yunlong, Lu Yunge et al (2003) Soil erosion risk assessment in karst area--Case Study in Guanling County of Guizhou Province. R Research of Soil and Water C C onservation 10 (3), 148 0150153. Wang Shijie, Jie Hongjun, Ouyang Zhiyuan et al ( 1999) Preliminary study on carbonate rock weathering and soil formation. Science in C C hina (S eries DD ) 29 (5),441. Wang Yu, Li Yan, Tan Jizhong et al (2003) Storage R R u le of Karst Water in Fault Basins Yunnan Science and Technology Publishing House, Kunming, 140 pp. Weng Jintao (1987), K arst and CC arbonate RRocks in GG uilin. Chongqing Publishing House, Chongqing, 180 pp. Wei Qifan (1996) Soil Erosion in karst region of south China and its control. RR esearch of Soil and Water CC onservation 3 (4), 72. Ying Hongfu, Xie Shuchen, and Zhou Xiugao (1994) New advance prospect on microbiogenal mineralization. E E art h Science Frontiers1 (3), 148. Yuan Daoxian and Cai Guihong (1988) e S cience of Karst EE nironment Chongqing Publishing House, Chongqing, 328pp. Zhang Yun (1992) New outlook on earth, A A dvance in earth science 7 (1), 57. Zhun Mingqiu, Cao Jianhua, and Guo Fang (2007) Analysis on the carbon amounts originated by the weathering of carbonate rocks and the inuence of soils on the carbon turnover process in karst areas. C C ar sologica Sinca 26 (3), 202.

PAGE 119

Conservation and Management 1380 2009 ICS P roceedings 15th International Congress of Speleology Research ESEARCH Priorities RIORITIES for FOR Underground NDERGROUND Ecosystems C OSYSTEMS in IN Colombia OLOMBIABRIGItt TT E L L Us S G G UILLERm M O BApt PT Ist ST E1, YANEt T H MUOZSAb B A2, D D ANILO SALAs S3, MANUEL A A Nt T O NIO H H Oy Y Os S R R ODRGUEZ4 1Profesor AA sociado, Ponticia UU niversidad Javeriana, Bogot DD .C C. CC olombia, Miembro Vocal de EESPELEOCOL ELEOCOL CC olumbia, guillermo.baptiste@javeriana.edu.co2Profesora AA sociada, II nstituto de CC iencias NN aturales, UU niversidad NN acional de CC olombia, sede Bogot DD .C C., AA.A A. 7495, Bogot D D .C C., CC olombia, Vicepresidenta de la AA sociacin EEspeleolgica CC olombiana (EESPELEOCOL ELEOCOL), ydmunozs@unal.edu.co3CC oordinador de Programas. Fundacin Moiss Bertoni. Paraguay. danilosalas@gmail.com4Bilogo, LL aboratorio de Mamferos, II nstituto de CC iencias NN aturales, UU niversidad NN acional de CC olombia, Bogot DD .C C., C C olombia, Miembro Vocal de la AA sociacin EEspeleolgica CC olombiana (EESPELEOCOL ELEOCOL), mahoyosr@unal.edu.co Abstract Cave research in Colombia has been developed mostly as an occasional activity since the beginnings of the XX century, and throughout the Andean cordillera. Just during the last three decades there has been some development of systematic eorts about the subject, and recently they have received the benet from the Convention on Biological Diversity (CBD) principles of conservation and sustainable use, applied to underground ecosystems. e road has not been easy, although plenty of pleasant moments. What have we accomplished? First at all, 130 reported cave systems, with just 6% of them with topography, 10% with some geology, and less than a third with some fauna and ora information. Despite the available data, they have not been used for conservation or sustainable use purposes, which should be the goals for the next decades. We suggest a regional planning approach, but beforehand, a specic policy development for underground ecosystems, something missing in the Colombian environmental law. ere are unsolved questions about the property of cave systems (public or private), and rules for its management, that must be adopted by the Ministry of Environment, Housing and Territorial Development, as well as the regional environmental authorities in Colombia and other stakeholders. We are proposing to complete a National Strategy for the Conservation of Colombian Caves and other Underground Ecosystems within the frame of the National Biodiversity Policy initiative that is being developed at the ministerial level, together with a Biodiversity National Action Plan 2009 in perspective.

PAGE 120

15th International Congress of Speleology Conservation and Management 1381 2009 ICS Proceedings A CAV V E SURV V EY FOR RESEARCH AND TOURIST CAV V E MANAGEMENTJULIA ULIA M. JA A ME E S1, DA DAVID ID J. MAR AR TIN IN2, GREGOR GREGOR Y M. TUNNOC UNNOC K3, and ALAN ALAN T. WARILD ARILD4 1Sydney Speleological Society and AA ustralian Speleological Federation, School of CC hemistry, F11, U U n iversity of Sydney, NN SW 2006, AA ustralia2Sydney Speleological Society 52 William James DDr, Mt Kembla, NN SW 2526, AA ustralia 312 LL andsdowne St, EE astwood, NN SW 2122, AA ustralia4Sydney UU niversity Speleological Society and AA ustralian Speleological Federation 41 NN orthwood St, NN ewtown, 2042, AA ustralia e Jenolan Caves Survey Project Group is preparing a State of the Art survey of the Jenolan Cave System, New South Wales (NSW), Australia. e survey has a number of potential uses but this paper only discusses those pertinent to speleological research and tourist cave management. e survey is a total station traverse through the main tourist routes. Less accurate survey instruments were used elsewhere. All entrances were linked using a surface theodolite network that was tied in to the Australian Map grid using dierential GPS. e surface network is critical to tying the entrances of the cave system together, proving the accuracy of main tourist loops and xing underground data to the map grid allowing GPS use on the surface. Radio location equipment was used to check selected less accurate traverses. e Walls cave-surveying program developed by the Texas Speleological Society was used for reduction of the cave survey data. Walls was chosen because of its features: simple text le input and Scalable Vector Graphics (SVG) output that can be imported directly into Adobe Illustrator. Other useful features are XYZ coordinates output, SVG round tripping and SMAPS Exchange File. Computer draing was performed using Adobe Illustrator as it uses brush and symbol libraries to provide a consistent style. Details such as geology, mineralogy, infrastructure, lighting and text are placed on separate layers. ese layers are maintained when exported to Adobe Portable Document Format (PDF). roughout the system, permanent-survey stations have been placed as aids for passage sketching, research and management use. A private contractor has carried out an aerial survey with a resolution of ve meters of the general area. From this a 3D map of the area has been produced with the caves below the surface terrain. e cave coordinates have been used to nd the points on the surface above underground features using a GPS. Uses of the survey are illustrated. e 3D map has been used to show the relationship between the cupolas in an area of the cave that has been formed by rising thermal waters. e surface location of these features and their depth below the surface has been obtained from the survey plan and GPS studies. GPS studies also allow the depth to be calculated for drip sites that have been sampled and analyzed for their water chemistry. e survey plan, GPS and radio location were used to establish the route that the fossil jaw bone of a Diprotodon, a species of Australian mega fauna extinct for some 30,000 years, entered the cave. For general management of the caves the survey plan is the most useful. e Adobe Illustrator les have additional layers that may be devoted to any particular task such as speleothem cleaning or infrastructure changes such as re-lighting a cave. e plan and the contour map of the surface have been used to assess the impact of a car park on the cave that it partially covers. However, for showing what tourists will encounter on a traditional cave tour or cave adventure tour, a developed long section of data and sketches is best.1. IntroductionIn 1986 the Jenolan Scientic Advisory Committee established that a complete high-grade survey of the Jenolan Tourist Cave System (New South Wales, Australia) was required. e survey was needed to allow scientic studies, notably of the geology and hydrology, to be adequately interpreted (James et al., 1988). e Jenolan Caves Survey Project Group undertook to prepare a State of the Art survey of the Jenolan Cave System. In addition, the committee felt that the survey would be a useful aid to cave

PAGE 121

Conservation and Management 1382 2009 ICS P roceedings 15th International Congress of Speleology management and development decisions. e proposal to survey the tourist caves and connected wild caves was endorsed and supported (without nancial commitment) by the Jenolan Caves Reserve Trust. e surveying project commenced in 1987. e Jenolan Tourist Cave System has historically been considered to comprise a number of almost separate caves although they are the sections of one cave system. e names of these caves are used here when discussing the survey as they divide the cave system into manageable sections for surveying, management and publication. is paper focuses on the detailed survey of the tourist cave section which comprises approximately one third of the 20+ km system. e tourist sections within the Jenolan Cave System are complex: parts are a 3D maze, in other areas boulder collapse dominates and there are many entrances. e eld work for the initial surveying exercise was undertaken between 1987 and 1990. Although the early traverse data was of the highest standard, the cave map when draed was little more than a record of the work reecting a wide range of sketchers abilities and hence could only be used to answer minor management questions. e traverse data was never published in an accessible format for use by research workers. e project was re-initiated in 2005. As survey data reduction and computer draing soware had advanced considerably the project was recast to take advantage of a mixture of free and commercially available soware in order to produce a state of the art cave survey that could be used for research and tourist cave management. 2. Instruments and Methods 2.1 Underground eld measurementse 1987/90 survey through the tourist routes is a total station traverse. A non-magnetic technique was necessary because of interference from cave furniture and power supplies. A Sokkisha total station with an SDR2 electronic eld book, data logger/computer was used. is instrument has infra-red electronic distance measurement (EDM) and a resolution of 1 mm for distance and one second for angles. In terms of accuracy, the standard deviation is +/(5 mm + 3 ppm) for distance measurement and +/4 seconds for angles. e Sokkisha computer calculates rectangular station coordinates as each leg is surveyed. ese were recorded in the data loggers memory (subsequently down loaded to a PC) and on booking sheets by a recorder. In some areas less accurate survey instruments were used. ese are listed in Bonwick et al. 1988 with their method of use and calibration. e Sokkisha was also used to take the large number of measurements required for the 1:200 contour maps of the large chambers. 2. 2 Surface eld measurementsAll entrances to the system were located using a surface theodolite network. is allowed the accuracy of main tourist cave loops to be proven and once converted to map grid coordinates allowing GPS use on the surface. An important dierence of the surface survey from the underground survey was the use of triangulation (and trilateration measuring distances). In contrast the underground survey was almost entirely traversing, i.e. progression from station to station one leg at a time. Because of the relatively long survey legs involved in the surface survey (up to 850 m), it was desirable to establish a network of stations where the connecting legs form a series of braced gures. To illustrate: a rectangular gure where the diagonals have also been measured is a mathematically stronger gure, than one where only the sides have been measured. Surface survey stations are typically marked with a drill hole and wing; a surveyors peg or a New South Wales state survey disk. 2.3 Survey station recording In 2005, the initial ve digit system for station numbering designed to prevent duplication of survey station numbers was replaced with a six digit system that contained two letters that allowed the data to be identied to a specic cave. e survey stations also needed to be recovered. At times, this required considerable eort as the original markers had disappeared. Passage junction stations were marked with 30 mm stainless steel disks stamped with the six digit survey station numbers and xed with stainless steel screws and epoxy. All survey stations have been photographed and a digital photograph library compiled to assist in relocating stations for passage sketching, research and management. Station locations have also been recorded by sketches giving the distance from the nearest infrastructure such as handrail posts and stairs.2.4 Reduction of eld data e early data had been reduced and traverses plotted using both mainframe computers and PCs involving a complex series of steps and programming (Bonwick et al. 1988). By 2005, survey reduction programs had advanced and become more numerous and many were freely available. For this project several such programs were considered, resulting in the choice of Walls, a cave-surveying program developed by the Texas Speleological Society (www.utexas.edu/tmm/sponsored_sites/tss/Walls/). It features simple text-le input and Scalable Vector Graphics (SVG) output that can be imported directly into Adobe Illustrator. Other useful features are XYZ coordinates

PAGE 122

15th International Congress of Speleology Conservation and Management 1383 2009 ICS Proceedings output and SVG round tripping. Walls allowed traverses to be produced both as proles and plans. e program also closes traverse loops. However, particular care has had to be taken where a loop contains both high and lower accuracy data. e loop is closed on the basis that any adjustment is restricted to the lower accuracy legs. is is justied as closure errors in loops surveyed entirely with the Sokkisha total station are typically less than 50 mm. e high grade total station survey eectively constrains the coordinates of many junctions in the complex network resulting in the closure problem being reduced to fairly small partial loops. e survey was originally based on an arbitrary coordinate system with magnetic alignment. In 1998 Australian Map coordinates were established for two points in the Jenolan area using dierential GPS (part of a topographical mapping exercise commissioned by the Jenolan Caves Reserve Trust). e surface network was extended to pick up these two points thus allowing the surface and underground surveys to be tied to map grid coordinates, currently the Australian Geodetic Datum 1966 (AGD66) coordinates. However, today the datum for mapping is the Geodetic Datum of Australia 1994 (GDA94). e project has plans to convert to GDA94.2.5 SketchingIn the 1987/90 survey the sketchers accompanied the surveying teams. e Sokkisha total station provided instant XYZ coordinates which allowed the sketcher to plot to scale. Despite this when the archived sketches were examined most were found to be lacking. In 2005, the decision was made to re-sketch the entire tourist cave routes in a separate exercise. e sketchers worked from a 1:200 scale traverse line. e widest section of the passage or chambers was used as the limits of the cave walls with the conventional methods of representing overhangs and ledges. Ideally a sketcher would have had one or more assistants and a laser distance measurer. ey drew their rendering of a plan in the cave and at each marked station drew a cross section of the passage. is sketch was draed, then printed to scale and taken back into the cave for checking and the addition of further detail. When resketching commenced the following was observed: e total station survey had too many long survey legs to adequately sketch from (survey legs had initially been maximized). Inll survey legs were required. ese were a lower survey grade but in the view of the sketchers, the shorter the better. e tourist cave paths, stairs and viewing platforms if surveyed in were of immense value to the sketcher. Inll stations, platform and stair locations were located using triangulation methods. Hand held laser EDM units, steel tapes and handyman laser levels facilitated this work. During the period of re-sketching many of the hand rails in the caves were replaced by stainless steel and provided they were used with caution, magnetic bearings could be taken and plotted to assist the sketchers Not all cavers have the ability to produce a suitable sketch: that is one that can be rst interpreted by a drasman, checked, and used by others. To produce developed long sections, the proles of the caves were sketched along the traverse lines. 2.6 Computer draing e initial plan was hand drawn on Mylar lm in black ink at a scale of 1:200. It was limited by the quality of the sketches available to the drasman. Adobe Illustrator allowed the current survey to be draed in color. First dras of the sketches could be prepared by a number of drasmen using Illustrator brush, swatch and symbol libraries to provide a consistent style. ese libraries were based on the UIS survey symbols as published by the UIS surveying commission (www.carto.net/neumann/caving/ cave-symbols/). A number of modications were made to the set with special symbols added when required. e master plan of the tourist caves is kept by one person to ensure that the draing of the entire map is consistent. e Illustrator layer and sub-layer construction has been used extensively to separate the component caves on the master plan and their associated details such as geology, mineralogy, infrastructure, lighting and text. One level of these layers is maintained when exported to Adobe Portable Document Format (PDF). When additional work is carried out in the cave the most recent version of the master plan is used to produce maps for us in the cave and then the information obtained is up dated on the master plan. 2.7 Preparation of 3D Modelse surface and caves require dierent treatment. e 3D cave models were constructed using custom soware to generate VRML les based on the cave survey and le right up and down data. DEM data (altitude points on a regular grid) was used to produce a surface to put into the VRML le along with the cave 3D data. VRML allows a map, air photo, Google earth photo or other graphic le to be pasted onto surface. Although aligning the photo to the surface

PAGE 123

Conservation and Management 1384 2009 ICS P roceedings 15th International Congress of Speleology can be problematic. is VRML le is then processed into Acrobat Professional Extended which gives a product that can be viewed on any computer with Acrobat Reader. 2.8 Preparation of high resolution 3D cave models A 3D cave model generated from le right up and down data gives a general idea of cave passages, but it oversimplies the cave. To get a better result more cross section data is required. e number used for this project was 12 points. ese were taken using a pole that could be leveled with a board at its top marked with 12 white lines radiating at 30 degree increments from its centre. e length of the pole was adjustable to allow the marked card to be as near as possible to the center of the passage, but still reachable. e white lines were used as guides to angle the laser distance measurer attached the centre of the board. e survey legs need to be short and depending on the situation. Where the total station survey leg was along a at surface for example on the tourist paths a tape was stretched between existing stations and the 12 point data taken at regular intervals or where there was a signicant change in passage shape or size. e distance along the traverse line was noted and the data given the same bearings as the traverse line. e alternative is to run a completely new survey line with short legs down the centre of the passage tying its ends into the total station survey stations. e 12 point survey data is input into an Excel le that outputs data in a simple 3D format that can be imported into AC3D. Once in AC3D the results can be edited to tidy up poorly dened corners and a smoothing process run to round o sharp edges. Finally the AC3D model is output as VRML and integrated with the rest of the 3D model prepared as described above.2.9 Data organization and archivingAll original data, line drawings and photographs are digitized and led either with respect to cave and/or station number. e computer reduced material and map dras are backed up on two external hard disks which are kept in separate locations. All maps will be presented to the Jenolan Caves Reserve Trust in so copy so that they can be used for management purposes. It is proposed to publish the cave and surface maps as a special addition of Helictite, the Australasian Journal of Cave Research. All maps and data will be available to researchers.3. Results and Discussion 3.1 e plan e plan of the tourist Caves ts on a sheet 5 m x 1.5 m at a scale of 1.200. Individual cave plans and cross sections have been printed at 1:200 and have been already used for Figure 1: e Plan of the Jenolan Tourist Caves.

PAGE 124

15th International Congress of Speleology Conservation and Management 1385 2009 ICS Proceedings a number projects, for example, a Spelean History (Whitby et al. 2009) and a Spelean education project (Kennedy et al., 2009) (Fig. 1). e detailed survey master plan is the most useful tool for management of the caves and recording research. e Adobe Illustrator les have layers that may be devoted to any task such as speleothem cleaning, infrastructure changes, or relighting a cave.3.2 Plan beneath the contoured surface A management problem oen encountered at Jenolan is whether infrastructure that has been placed in the area in the past is likely to impact on the caves. Figure 2 shows how the plan and the contour map were used to assess the relationship between a car park and the tourist cave below. 3.3 XX YZ coordinates and GPSFor the rst time, questions such as how deep sites within the caves are below the precipitous Jenolan terrain can be rapidly answered. e xyz coordinates converted to GPS coordinates have given depth data for drip sites that have been sampled and analyzed for their water chemistry. e same method was used when the jaw bone of the rhinocerossized diprotondon optatum was found in the caves. e species became extinct some 30,000 years ago (Long et al., 2002). e jaw bone was in a boulder collapse in a section of the tourist caves inaccessible to such a large animal through known passages. On the surface above the jaw bone site is a pit trap cave containing an extensive boulder collapse the bottom of which is a few metres above the jaw bone. To illustrate for tourists what they will encounter on a traditional cave or adventure tour, a developed long section is best. Figure 3 shows the developed long section of the adventure cave tour that is marketed as the Plug Hole Adventure Cave. Examination of the plan shown in Figure 1 will show that even to those used to interpreting such complex plans (the cave has several levels above each other) it is dicult, but for the inexperienced it is impossible. 3.4 3D-Modelse 1987/90 survey data was suciently complete for it to be used for three dimensional computer graphics model (Fig. 4) generated using le, right up and down data taken at the time of surveying. It has been used on a number of scientic posters to show the location of sampling sites within the cave and is on display in the interpretation centre at Jenolan Caves. e model illustrates geological features of the system such as the down cutting of the Jenolan River to form the northern caves. Initially it oversimplied the shape of the Devils Coach House and the Grand Arch so it was improved by the use of 12 point cross section data taken at 10 m intervals along their center lines. e representation of the two chambers is shown without smoothing so that the 12 point net can be observed. e area (Fig. 5A) in the southern part of the southern caves (Fig. 1) was attributed to the solution of the limestone by rising aggressive water resulting in a series of domed chambers known as cupolas (Armstrong, 2004). Figure 5B illustrates how poorly le, right, up and down data taken with long survey legs represents such a complex area of the cave. It is inadequate for depicting the shape of cupolas and Figure 2: Cave plan in silhouette beneath contoured surface. Figure 3: A developed long section through Elder Cave.

PAGE 125

Conservation and Management 1386 2009 ICS P roceedings 15th International Congress of Speleology the way they are interconnected. Figure 5C shows the rst attempt to represent part of this series of domes by the 3D high resolution method. e methods have described the computing involved in placing the 3D cave map below the Jenolan Surface. A monochrome gure does not adequately demonstrate the power of this method of representing the cave system and its surface. In addition, the advantages of ability to rotate and enlarge are lost in any passive mode of presentation. It is proposed that the model will be made available on a webpage so that it can be viewed in Acrobat Reader enabling all of the above features to be used.4. Conclusions e accuracy of the 1987 surface survey and underground traverse allowed it to be used in the 2005 redressing of the project. A systematic approach to the organization of data and additional surveying has resulted in high quality maps of the caves being produced. e availability of freeware, shareware and commercial computer programs has enabled the end product to be produced both as hard and so copy. e project has achieved one of its major goals to produce the survey in an accessible form so that it can be used by management and research workers. Finally the cave survey which commenced as obtaining, recording and presenting eld data as a cave map has turned into a continuing research project in itself.Acknowledgmentse Jenolan Caves Reserve Trust for nancial support for soware and consumables and making available an oce in the JCRT research facility that is dedicated to the production of the survey. e Jenolan Cave guides and numerous cavers both national and international that have worked on various aspects of the survey project. Both R.Q. Bridge and C.D. Dunne for operating theodolites and reducing data. e School of Civil and Mining Engineering, University of Sydney for supplying surveying equipment and computer facilities.ReferencesARMSTRONG R.A.L. (2004) e troubles with cupolas. A A cta CC arsologica. 33 9. BONWICK M.H., R.Q. BRIDGE, C.D. DUNNE, J.M. JAMES, D.J. MARTIN and G.M. TUNNOCK (1988) e Jenolan Cave System Surveying Project Part 2 techniques and computing. Proceedings of t he 17th Biennial CC onference of the AA ustralian Speleological Federation, Cairns, pp 3. JAMES J.M., D.J. MARTIN and LK.TUNNOCK (1988). e Jenolan Cave System Surveying Project Part 1 history, organisation and assessment. Proceedings of t he 17th Biennial CC onference of the AA ustralian Figure 4: A modied 3D of the Jenolan Tourist caves which also shows the route of the Jenolan River. Figure 5: A, the plan of the southernmost caves in the Jenolan Tourist Caves (Fig. 1). B, the same section as A of the original 3D model using le right up and down. C, the high resolution 3D model of the same area.

PAGE 126

15th International Congress of Speleology Conservation and Management 1387 2009 ICS Proceedings Speleological Federation, Cairns, pp 25. KENNEDY, S., C.M. BARNES, and J.M. JAMES (2009) A project for the Duke of Edinburghs silver award. ese proceedings. LONG J., M. ARCHER, T. FLANNERY and S. HAND (2002) Prehistoric mammals of AA ustralia and NN ew G G uinea: one hundred million years of evolution. University of NSW Press limited Sydney, Australia. WHITBY, J.L., K.A. BELLAMY, and J.M. JAMES (2009) A spelean history revealed when naming features for a cave survey. ese proceedings.

PAGE 127

Conservation and Management 1388 2009 ICS P roceedings 15th International Congress of Speleology CAV V E AND KARST RESOURCE INV V ENTORY AND MONITORING ON THE TONGASS NATIONAL FOREST, SOUTHEAST ALASKA, USAJOHANNA OHANNA L L KO O VARI ARI K Karst RResource Specialist/ GG eologist, Tongass NN ational Forest 1312 Federal Way, orne Bay AA K 99919, UU SA A jkoarik@fs.fed.us e Tongass National Forest is the largest forest in the National Forest System in the United States, encompassing over 6.9 million hectares covering the islands of the Alexander Archipelago and the narrow band of mainland from Dixon Entrance to Icy Bay. e Tongass contains 85% of the total karst in southeast Alaska, approximately 400,000 hectares primarily on Chicagof, Prince of Wales, and surrounding smaller islands. In this century, inventory and monitoring of our karst resources is a new priority. Most recently in 2008 an amendment to the Tongass Land Management Plan updated the karst standards and guidelines developed and put into place in 1997. Expeditions to map and inventory karst features were revived in 2008. Projects to evaluate the baseline characteristics of spring ow in karst watersheds were completed in the last 4 years. A project utilizing these baseline data is being developed for quantitatively evaluating the eectiveness of our karst standards and guidelines through karst watershed monitoring on Kosciusko Island. A massive eort to digitize and organize cave and karst inventory data is currently underway, with cave locations being recorded with high resolution GPS for integration into a comprehensive GIS database with resource inventories. ese data will be used to organize and guide eorts to conduct a biological inventory of caves on the Tongass in the future. As budgets grow tighter every scal year, cooperation with local groups such as the Glacier Grotto of the National Speleological Society and national groups such as Geocorps America of the Geologic Society of America have helped to support our cave and karst programs.1. IntroductionOn the Tongass National Forest, the protection of cave and karst resources came about shortly aer the Federal Cave Resource Protection Act of 1988. e Tongass Cave Project (TCP) began inventorying and exploring caves in southeast Alaska in the early 1980s, and basic karst resource inventories began in the late 1980s and early 1990s. e initiation of karst resource protection on the Tongass together with the mapping projects led by the TCP spurred interest in karst and cave resources in southeast Alaska. El Capitan Pit, the deepest limestone pit in the United States, was mapped at 182.4 meters, and El Capitan Cave, the longest cave in Alaska was mapped at approximately 3.2 kilometers (Lewis 1997). Initial work began on inventorying the biological resources in Tongass caves on Prince of Wales and surrounding islands in the early 1990s. Researchers found that mammal species and birds used caves as critical roosting and hibernating habitat, including ve species of bats (Baichtal and Swanston 1996). Invertebrate collections from over 300 caves and resurgence sites yielded at least ve troglobitic and forty troglophilic invertebrate species, three of those newly discovered (Carlson 1994 and 1996). e caves were found to be rich in paleontological and cultural resources, including the nding of the oldest bones in North America in On Your Knees Cave (Carlson 1993, Dixon et al. 1997). As a result, the Forest Service included the Karst and Cave Resource Signicance Assessment of 1993 as part of the Tongass Land Management Plan (TLMP) revision process. e 1997 TLMP incorporated a karst management strategy in accordance with Aley et al.s recommendations in 1993, including integration of continuing research as well as inventory and monitoring priorities in southeast Alaskas karst. is karst management strategy was implemented on two timber projects in highly developed karst areas (Lab Bay and Heceta Sawy Salvage) with mixed results. While both areas were inventoried for karst features prior to harvest, monitoring of the harvest contractors work on the ground was thought to have not been conducted properly due to the, limited experience of resource managers (Baichtal 1997). In addition, no physical monitoring of the karst watershed occurred in order to determine actual eects on the water quality and quantity at karst springs due to timber harvest. Inventories of cave and karst features continued through the TCP and Forest Service expeditions. Cave resource inventories were not documented at the time of cave survey, but short reports were lled out on each

PAGE 128

15th International Congress of Speleology Conservation and Management 1389 2009 ICS Proceedings cave including observations, if any, on cave resources such as biological, archaeological, and paleontological. Since Carlsons studies in 1993, however, no further professional work inventorying the potentially diverse populations of cave fauna in southeast Alaska occurred. In the 21st Century, the Tongass again began to revise the forest management plan, and resource managers proposed changes to the karst management strategy. In 2002, a panel was contracted by the Forest Service to assess the implementation of the karst standards and guidelines established in the 1997 TLMP and to analyze proposed changes. e Karst Review Panel found that generally the implementation of Karst Standards and Guidelines had ensured a high level of protection for karst resources; however they recommended a higher level of training for karst specialists and identied some revisions to the proposed changes (Griths et al. 2002). Finally, in 2008 the Amendment to the 1997 Tongass Land Management Plan was published, including some changes recommended by the karst review panel for cave and karst resource inventory and monitoring. Today, the current inventory and monitoring objective for cave and karst resources on the Tongass is stated as, e signicant cave and karst ecosystems should be maintained and protected Forest-wide, and natural karst processes should continue while the productivity of the karst landscape should be maintained (TLMP 2008). e current inventory and monitoring question for cave and karst resources on the Tongass is, Are the biological, mineralogical, cultural, paleontological components, and recreational values of the karst and caves maintained? (TLMP 08). e budget decline across the Forest Service reached the Tongass in approximately 2004. Since that year, budgets for inventory and monitoring have all but dried up, and nding funding for projects such as these has required creative thinking and partnerships.2. Settinge Tongass National Forest is the largest forest in the National Forest System in the United States, encompassing over 6.9 million hectares covering the islands of the Alexander Archipelago and the narrow band of mainland from Dixon Entrance to Icy Bay (Fig. 1). e Tongass contains 85% of the total karst in southeast Alaska, approximately 400,000 hectares primarily on Chicagof, Prince of Wales, and surrounding smaller islands (Fig. 2). e largest area of karst development exists on Prince of Wales and surrounding islands, which contain approximately 1,813 km2 of karst (Baichtal 2006). ese karst areas are concentrated on the north end of the island and surrounding smaller islands, where over 600 caves have been mapped. Karst formed to some extent on Prince of Wales prior to the Wisconsin glacial advance 21,000 to 14,000 years ago. is period of glaciation caused scouring, passage collapse, and sediment ll in karst systems, as well as leaving thick glacial till deposits and razing epikarst development at lower elevations. e intense development of karst on the Tongass National Forest is controlled by several factors including the high percentage of calcium carbonate (CaCO3) in the limestone of southeast Alaska averaged at 97.65 percent (Maas et al. 1992). In addition, faults and fractures resulting from the northward movements of the Alexander Terrane are dominated by northwesterly trending strike-slip faults and second order intersecting north-trending strike-slip faults, which dene karst conduit formation (Gehrels and Berg 1992, Aley et al. 1993, Baichtal and Swanston 1996).3. Methods 3.1 Inventory In order to maintain and protect cave and karst resources forest wide, they rst must be inventoried. Inventory methods for cave and karst features on the Tongass National Forest are quite progressive, with the exception of biological surveys. Areas with karst development are now assessed using a whole ecosystem approach an area of carbonate Figure 1:e Tongass National Forest is located in southeast Alaska, USA (USDA Forest Service).

PAGE 129

Conservation and Management 1390 2009 ICS P roceedings 15th International Congress of Speleology bedrock is assessed as a complete karst system, as opposed to individual karst features. Aerial photograph, geologic, and topographic map interpretation is utilized prior to eld reconnaissance. Surface streams and where they seem to disappear are noted as well as obvious disturbances in the forest, which could suggest karst features. Geologists locate karst features and caves out on the ground, mark them with GPS points, and integrate them into the master GIS database of features. Caves are generally noted and added to a list to be explored further during a cave expedition in the summer season. Cave resource inventories are conducted with surveys. Generally karst inventory work is associated with timber sales due to budget constraints. Karst features located are oen found within proposed timber harvest units, and part of the GIS database process is associating buers with highly sensitive or vulnerable karst features and caves representing a high degree of openness, or ability to transport sediment or contaminants to the subsurface watershed. If the majority of the proposed timber harvest occurs in areas of karst development, oen dye tracing will be conducted to fully understand the eects of harvest on the karst watersheds. In addition, large dye trace projects were conducted on Prince of Wales and Tuxecan Islands in conjunction with road project evaluations (Prussian and Baichtal 2007). e results of the traces help resource managers understand the scope of karst watersheds on the Tongass, and how timber harvest and road building impact these areas. No current methods exist for biological monitoring of cave fauna on any level.3.2 MonitoringIn order to maintain the natural processes of the karst systems on the Tongass, rst resource managers must understand how those natural processes work. Monitoring methods on the Tongass lag signicantly behind inventory eorts. Currently, the only karst and cave resource monitoring accomplished is qualitative in nature and in conjunction with the actual timber harvest itself. Geologists go out on the ground in current harvest units and assess the harvest contractors compliance with the karst standard and guidelines for that particular sale. is involves evaluating whether or not buers placed around karst features were successfully applied in practice and analyzing new temporary roads built and culvert placement. In addition, the Tongass boasts the only show cave in the state of Alaska, El Capitan Cave. El Capitan Cave also happens to the longest mapped cave in Alaska at approximately 3.2 kilometers. is cave is open to visitors from May through August each year, and averages around 500 visitors during that time. e cave is gated to prohibit access unless on a guided tour. Other than the 372 wooden stairs to the entrance and the bat gate, the cave is undeveloped. Photographic visitor use impact monitoring was attempted in the early 1990s, however no follow up work on this project has occurred. Other caves on the Tongass also see recreational visits from guided adventure groups and locals. One cave displays a basic register in the entrance, however none of the other caves are monitored for visitor use and impacts. One of the diculties with monitoring visitor and land use impacts on Tongass caves is the lack of base level data to compare changes. 4. Results and DiscussionWhile inventory methods on the Tongass are producing a comprehensive catalog of cave and karst features, monitoring methods are in the very beginning stages of development. In addition, the management of cave and karst inventory data needs to be reorganized and updated for the 21st century into a useable tool for resource managers. While the new inventory and monitoring objective and question begin to address the issues, a more ecient statement would be useful. Resource geologists and karst resource specialists were not consulted in the development of the current monitoring question. A new question is proposed which would allow for the monitoring of key indicators of karst system health. In a Forest wide inventory and monitoring workshop in 2008, geologists proposed the monitoring question, Are we maintaining the functionality of the Figure 2: Areas of carbonate bedrock on the Tongass National Forest (USDA Forest Service).

PAGE 130

15th International Congress of Speleology Conservation and Management 1391 2009 ICS Proceedings karst ecosystem in order to not diminish the hydrological, biological, minerological, cultural, paleontological, and recreational resources? (Kovarik 2008). Water quality in a karst system aects all the resources within it. rough monitoring certain gauges within the systems such as changes in water quality, impacts to all resources can be assessed. To date, no quantitative monitoring of changes in water quality, quantity, sediment, or turbidity below harvested units in karst areas has occurred during or post harvest. A forest service hydrologist monitored water quality in 2004 and 2005 at four karst springs in an attempt to understand general water quality trends of the karst systems. In 2006 a Masters thesis analyzed storm response at two karst springs in order to look at hydrologic function and storage in unharvested karst watersheds on the Tongass (Kovarik 2007). ese projects are the rst step at establishing base line conditions on the Tongass. A project is in development for Koscisuko Island, where a large timber sale is proposed. Both old growth and young growth (areas harvested in the 1940s and 50s) units are being assessed for harvest. Watershed analyzation is in progress right now in order to place instruments at karst springs below units that will be harvested in the future. Data gathered from these sites will address the ecacy of the current karst standards and guidelines and the feasibility of thinning in young growth units as a restoration method for harvest impacted karst watersheds. Biological inventories are the next logical step for the cave and karst inventory program on the Tongass. However, before a comprehensive biological survey can begin, the current cave inventory data needs reorganized and updated. e current cave and karst resource inventory is in a master GIS database, however the majority of cave locations are lacking data such as the name of the cave. While hard copies of all cave maps are led, survey data has either been lost or only exists as the original survey sheets taken into the cave. e TCP wrote short reports on each cave found, surveyed or unsurveyed. ese reports contain vital resource inventory data. Currently the only copies of these are in paper format. All cave data needs scanned to .pdf and saved, and all cave surveys need entered into a cave data management program. Cave locations lacking names need to be relocated in the eld, and eld checked for the metal tags labeling each cave with its inventory number. is project will require many hours of eld time attempting to relocate caves based on fairly vague location data. During the summer of 2008, the Tongass received funding through working with the Geological Society of America Geocorps Interns for two cave technician positions to begin the relocation and digitizing. In addition, through contracting with the Glacier Grotto of the National Speleological Society, two TCP project members will be assisting with the relocation of caves found in the early days of the project. During cave relocation, additional information is gathered utilizing cave and karst feature inventory sheets developed by the Hong Megui Cave Exploration Society. Data gathered on this sheet includes photos of the cave entrance and general location of the cave, sketches of the entrance location, and precise location details. Once the locations are linked with the cave names, and survey data has been entered electronically, the master GIS database will include line plots for each cave. ese line plots will be a sort of underground watershed map to which resource managers can link all pertinent resource information, including entrance photos, archived reports, inventories linked to cave survey stations. is database will be a tool that will greatly diminish the chances of loss of corporate data from the Tongass cave and karst program, and will provide a strong foundation for future resource inventory information such as comprehensive biological surveys.ReferencesAley, T.; C. Aley, W.R. Elliot, P.W. Huntoon, 1993. Karst and cave resource signicance assessment Ketchikan area, Tongass National Forest, Alaska. Ozark Underground Laboratory, Protem, Missouri. 79 pp. Baichtal, J.F. 2008. P ersonnel CC ommunication. Baichtal, J.F. and D.N. Swanston. 1996. Karst Landscapes and Associated Resources: A Resource Assessment. General Technical Report, PNW-GTR-383. Portland, Oregon: U.S. Department of Agriculture, Forest Service, Pacic Northwest Research Station. 13 pp. Baichtal, J. 1997. Application of a Karst Management Strategy: Two Case Studies from e Tongass National Forest, Southeastern Alaska, e Challenges of Implementation. Proceedings of the 1997 Karst and Cave Management Symposium. Bellingham,Washington. AA m erican CC ave AA ssociation. 4. Carlson, K.R. 1994. Inventory and Assessment of Ecological Relationships between Cavernicolous (Cave associated) Invertebrate Species and eir

PAGE 131

Conservation and Management 1392 2009 ICS P roceedings 15th International Congress of Speleology Interactions in Representative Karst Ecosystems on Carbonate Terrain in the Ketchikan Area National Forest: Part I. Dall Island. Karst Biosciences. Christiansburg, Virginia. 60 pp. Carlson, K.R. 1996. Inventory and Assessment of Ecological Relationships between Cavernicolous (Cave associated) Invertebrate Species and eir Interactions in Representative Karst Ecosystems on Carbonate Terrain in the Ketchikan Area National Forest: Part II. Coronation Island. Karst Biosciences. Christiansburg,Virginia. 93 pp. Carlson, R. 1993. Overview of Archaeological Resources associated with caves and rockshelters in southern southeast Alaska. AA l aska AA nthropological A A ssociation.20th annual meeting, 12 pp. Dixon, E.J., T.J. Heaton, T.E. Field, T.D. Hamilton, D.E. Putnam, and F. Grady, 1997. Late uaternary Regional Geoarchaeology of Southeast Alaska Karst: A Progress Report. GG eoar chaeology. 12: 689. Freitas, T.A., C.M. Soja, and A.I. Antoshkina, 1998. Coeval development of Silurian Stromatolite reefs in Alaska and the Ural Mountains: Implications for paleogeography of the Alexander Terrane: Comment and Reply. GG eo logy.382. Gehrels, G. and H. Berg, 1992. Geologic Map of Southeast Alaska. U.S. Department of the Interior, U.S. Geological Survey. Kovarik, J.L. 2007. Storm response and water balance of two temperate rainforest karst watersheds, southeast Alaska. Masters esis, Western Kentucky University, Bowling Green, Kentucky. Kovarik, J.L. 2008. Cave and Karst Inventory and Monitoring. Presentation to the Tongass National Forest Inventory and Monitoring Workshop, Southeast Alaska Discovery Center, Ketchikan, Alaska. Lewis, S.W. 1997. e Tongass Cave Project: An Advocate for Karst Ecosystem Protection in Southeast Alaska. Proceedings of the 1997 National Cave Management Symposium. Mass, K.M., J.C. Still, and P.E. Bittenbender, 1992. Mineral Investigations in the Ketchikan Mining District, Alaska, 1991: Prince of Wales Island and Vincinity. U.S. Bureau of Mines. OFR 81. Prussian, K. and J. Baichtal, 2007. Watershed Delineation on Prince of Wales Island, Southeast Alaska. US Forest Service, Tongass National Forest, Alaska. Soja, C.M. 1991. Origin of Silurian Reefs in the Alexander Terrance of Southeastern Alaska. Palaios. 6: 111 125. Soja, C.M. and A.I. Antoshkina, 1997. Coeval Development of Silurian stromatolite reefs in Alaska and the Ural Mountains: Implications for paleogeography of the Alexander Terrane. GG eo logy. 25: 539.

PAGE 132

15th International Congress of Speleology Conservation and Management 1393 2009 ICS Proceedings KARST CAV V E SYSTEM TYPES AND THEIR PROTECTION IN CHINAYAORU AORU LU LU1 and KE E QIANG IANG HE HE2 1School of CC ivil EE ngineering Tongji UU niversity NN o.1239,Siping RRoad ,Shanghai ,C C hina2Qingdao Technological UU niversity, Qingdao,Shandong,P.R R CC hina In China, there is widely developed karst cave systems, (1) particularly the carbonate rocks have formed a lot of special and wonderful cave systems as the important resources. On the other hand, the karst cave systems can harm local peoples living and constructions. In ancient Chinese saying for the carbonate rocks regions are no mountain without cave and no cave without wonderful. Some problems related to the karst cave systems are discussed in following:1. Basic Features of Karst Cave SystemsBased on dierent principles, karst cave systems will be separated. For examples, by the lithological features, the cave systems developed in the carbonate rock, sulfate rock, and halide (Lu and Zhang, 2007) are classied, but in the carbonate rocks, usually the two main sub-types must be distinguish, (Lu et al., 2006).1.1. Sub-surface karst river stream: e surface river water sinks into underground cave and passage way through a certain distance, and then the substream is owing out to the surface to form the surface river again. e related formula is expressed in the following: Qd m 1f tft,f tpi,p ti,IOiq q QQn 1p (1) where Qoi = exit ow quantity of the sub-surface karst river stream in i time, m3/s; QI, i-t = entrance ow quantity of the sub-surface karst river stream in i-t time, m3/s; qp, i-tp = the ow quantity of p branch of the sub-surface karst river stream in i-tp time, m3/s; qf, i-tf = the ow quantity through corroded ssure into the sub-surface stream in i-tf time, m3/ s; Qd = the condensation water between i-t to i time into the sub-surface stream, m3/s. 1.2. Karst underground river systems: e karst subterranean river systems are mainly collecting the percolating raining water through sinkholes; corroded ssures to subsurface. e related formula is expressed in the following. Qd Oi m 1f tfi.f n 1p tpi.p n 1s tsi.s OiQqqqQ (2) qs.i-tswhen prior i-ts time, the ow quantity directly sinking into ground by depression (m3/s), other marks are ditto formula (Lu, 1986). Referenced to the formations of cave systems and related environments, the cave systems in China will be classied into twelve models (Lu, 1999) (Fig. 1). All the developments of cave models are controlled and inuenced by local structures (Fig. 2).2. Protection and Exploitation of Cave SystemsConsidering both karst cave system resource conditions and hazard factors for economic constructions and mankinds living, the main purposes for studying cave systems are reasonably exploitation the dierent resources related to cave systems while providing protections, and avoiding or reducing the hazards to harm the constructions and people lives. Some important problems are discussed in following.2.1 Exploitation of resources related to cave systemsMany kinds of liquid, gaseous, and solid mineral resources have been formed in cave systems, but karst water resources are the most important ones. In Southwest China, about 3358 cave systems have been investigated for their water resources (Table. 1). e comparison related to karst water systems between Southern China and Northern China will be listed in Table 2. e karst water recourse in cave systems will be exploiting through many ways, such as pumping karst water from vertical passages, guiding water by articial tunnel to connect the underground river, or constructing a dam in a cave system to form an underground reservoir (Lu, 2003; Milanovic, 1981). But, the important problem is that the karst collapses and other hazards have oen been promoted. erefore, to protecting cave systems is the rst duty before exploiting water and other resources in cave systems.

PAGE 133

Conservation and Management 1394 2009 ICS P roceedings 15th International Congress of Speleology Figure 1: Main models related to formation enironments of karst cave systems: 1.sea eroded coastal cave model; 2reef cave model; 3hill and shallow valley cave model; 4lacustrine cave model; 5--tower peakvalley cave model; 6--high mountain deep gorge cave model; 7slope mountain land cave model; 8ridge mountain cave model; 9high-cold plateau cave model; 10moist-hot plateau cave system;11glacial cave model; 12thermal cave model. Figure 2: Analyses of cave systems controlled by structures: 1cave passages; 2large valley; 3--depression and sinkhole; 4ssures and vertical cave passage and sinkhole; 5entrance of surface river into ground; 6exit of karst subterranean stream; 7fault; 8anticline axis; 9syncline axis; 10carbonate rock; 11sandstone and shale; 12sandstone and shale with less carbonate rock; Athe cave development controlled by structural ssures; Bunderground river controlled by elongated anticline; Cthe karst subterranean river and spring controlled by fault; Dvertical karst cave and passages controlled by ssures; E subterranean river controlled by syncline; Fcross tension ssures controlled the development of subterranean river.

PAGE 134

15th International Congress of Speleology Conservation and Management 1395 2009 ICS Proceedings Table 1. Some conditions of water resources in cave systems in Southwest China (5) Content Regions Ground river Number Flowing quantity in dry season108m3/a Yunnan 148 39.02 Guizhou 1130 71.35 Sichuan 895 63.96 Chongqing 201 28.68 Guangxi 435 191 Hunan 338 17.65 Hubei 211 14.85 Total 3358 426.69 Table 2 CC omparison of Karst water systems features between Southern CC hina and NN orthern CC hina Contents Region Southern China Northern China System feature main karst ground river systemsKarst spring systems Passage feature main karst cave and larger corroded passages net main corroded ssures net Catchment area Km250>3000 500>4000 Ground water owing speed m/d 100>11400 1>50 Ground run-o modulus L/Km2.s 8>35 2>10 Karst water age a main 1>10 main 3>20 2.2. Prevention and reduction of harms resulted by the compound ows from cave system.In karst cave systems, there are usually existing compound ows, which are mixed with liquid (water). Gas (air, vapor, CO2 and other gas masses) and solid (e.g., clay, sand and gravel) (Milanovic, 2000). For the uncompressed liquid and solid materials: (3) x, y, zexpress the component of owing speed in x, y, z axis. (intensity)as constant then : (4) So that: (5) But for the compressed gas mass: (6) (7) (8) (9) In a subterranean cave system, gas mass compressed together with liquid and solid materials in the closed passages under high pressures results in the speed potential r. When subsurface engineering destroys their balance, the compressed three phrases mass will suddenly explode up to form the three phrases ows with very high starting speed Uf to harm the underground engineering.. In China, some tunnels with their lengths of 12-17 km had been utilized as warming systems and undergone a series of treatments for the hazards. On the other hand, the mechanical stability of the big cave systems also are the important problem either for the subsurface constructions or for the surface buildings. e Karst cave systems commonly have speleothems (Fig.3),

PAGE 135

Conservation and Management 1396 2009 ICS P roceedings 15th International Congress of Speleology but big underground halls are also developed. A big cave near the railway tunnel is enhanced by a lot of ferroconcrete columns (Fig. 4). 3. Development of Tourism Resources Related to Cave SystemsIn China, about four hundred cave systems are already opening for tourism and some are designated as a National Geopark or World Geopark. At present, only the Furong Dong Cave System is designated as a World Heritage Site. Multiple purposes for geopark include: 1. protecting natural resources; 2. decreasing hazards; 3. studying geosciences; 4. improving environmental quality; 5. promoting local development; 6. gaining nancial benets. e sustainable development of geoparks must include the six qualifying characteristics, which are potential benets to the elds of science, evolution, culture, ecology, architecture, and economics. Harmonious relationships for geopark are also very important, which include: 1. resources and hazards; 2.development and protection; 3. water supply and environment; 4. arbor forest and shrub cover; 5. energy and impacts; 6. communication and reception; 7. building style and landscape; 8. articial actions and natural processes. Otherwise, the speleothems and fractural depositions in cave systems may provide a series of messages for studying the environments evolution.4. Evaluation of Cave Systemse about mentioned factors related to sustainable development of cave systems are expressed in Fig. 5. QTT = f(G, E, W, A, B, Hg, HC. HB) (10) QTTTotal quality of geo-environment; Gfactors of trip resources; EEco-hydrology; Wfactor of water environment; Aair quality; Bbiological resources; HGfactor of geo-hazards; HCfactor of climatic hazards; HBfactor of biological hazards. By a comprehensive evaluation related to karst water activity, eco-hydrology, speleothems depositions, karst collapse and destruction of cave passages, fractural depositions in cave, and vegetation in surface, it is possible to separate cave systems into best ecological, good ecological, deteriorated ecological, and bad ecological conditions --the four types of cave systems.5. Conclusionsere are numerous cave systems of a wide variety in China, which are closely related to peoples life and economic construction. However, at present, very few careful surveys and studies have been done to reveal the karst development rules. A large number of cave systems have developed into tourism resources, while more are eectively exploited into karst water resources. Karst cave systems are precious resources, and we must insist on the principle to exploit while protecting and protect while exploiting. Special Figure 3: A lot of special speleothems in Lianhua Dong Cave, Guilin. Figure 4: A big cave to be treated by ferroconcrete columns. Figure 5: e evaluation of enironmental quality related to geopark and/or cave systems must consider the chemical, physical, mechanical, and biological compound processes.

PAGE 136

15th International Congress of Speleology Conservation and Management 1397 2009 ICS Proceedings attention must be paid to the hazards provoked by cave systems during construction activities. erefore, we must avoid geo-hazards provoked by cave systems during engineering constructions as well as prevent and control disasters caused by caves.ReferencesLu Yaoru, 1986. Process of karst caves development and three phrases ow. Proceedings of the 9th International Congress of Speleology. Bacelona. Spain, Vol.I 273. Lu Yaoru, 1986, KARST IN CHINA.Landscapes. Types. Rules. Beijing: Geological Publication. Lu Yaoru, 1999. Research on the evolutions of karst hydrogeological environments and their engineering impacts (Chinese, with English abstract). Beijing: Science Press. Lu Yaoru, 2003. Geo-ecology and sustainable developmentdeveloping ways for karst regions in Southwest and adjacent karst regions of China (with English contents ). Nanjing: Hohai University Press. Lu Yaoru, 2007. Karst water resources and geo-ecology in typical regions of China. Environmental Geology, 51:695. Lu Yaoru, and Zhang Fenge. 2007. Sulphate rock karst and carbonate-sulphate rocks compound karst Developmental mechanism and engineering impacts. (Chinese with English abstract and English tables and gures illustrations )Beijing: High Education Press. Lu Yaoru, Zhang Fenge, Liu Changli, Tong Guobang and Zhang yun., 2006. Ground water systems and ecohydrological features in the main karst regions of China. ACTA GEOLOGICA CHINA,80(5): pp 743. Milanovic,P. 1981. Karst hydrology. Water Resources Publication, Little, Colorado, USA. Milanovic,P. 2000. Geological engineering in karst monograph. Belgrade: Zebra Publication Ltd.

PAGE 137

Conservation and Management 1398 2009 ICS P roceedings 15th International Congress of Speleology

PAGE 138

CONTRIBUTED PAPERS IN THE EARTH SCIENCES

PAGE 140

15th International Congress of Speleology Earth Sciences 1401 2009 ICS Proceedings KARST CAV V E FEATURES OF MONGOLIAEE A AVIR IR MED ED PHD D .I I nstitute of GG eography of the Mongolian AA cademy of Sciences Abstract In the research paper are reected cave geneses, morphology and cave evolution features of Mongolia. e karstic caves are distributed in Mongolia accordingly to the distribution of carbonate rocks. e sharp climate, plateau height and precipitation changes have had an inuence on the origin and evolution of the karst caves. ere are three large regions where karst caves are widespread. In these regions the biggest multi-phase caves are formed. Caves in Mongolia have developed through the following 5 stages: 1. Split stage or rst stage. e surface water penetrates down a ssure, erodes rocks and causes the initiation of caves; 2. Canal stage. Given further penetration of water from the surface, the cave is enlarged and the volume of water owing through it increases, causing karst processes to be more active; 3. e stage to originate underground water ow. Underground canals get wider and the erosion activities increase; 4. Water stage to originate the karst. Underground canals get wider permanently, the canal water corrodes the bottom part and penetrates downwards, developing new passages below and leaving dry upper-level passages ; 5. Failure and dry stage. In this stage, the cave roof fails and some parts of the cave become open. Karst caves of Mongolia are famous for their rich archeology. In 1995, the scientists of Mongolia, Russia, and America made excavation in the TSagaan agui and proved that ancient people lived there before 750,000 years ago. Many hundreds of archeological nds were made, including stone tools. In 1997, excavations were made in the Chehen agui and 317 ancient stone tools were found.

PAGE 141

Earth Sciences 1402 2009 ICS P roceedings 15th International Congress of Speleology THERMAL SEDIMENTATION IN CAV V ES: A NOTEGIOGIO VANNI ANNI BADINO ADINOL L a Venta EEsplorazioni GG eograche and DD ip. Fisica GG enerale UU ni-TO O, Via Pietro GG iuria 1, II -10125, Torino, II taly Abstract e concept of temperature is meaningful only for a system in the condition of thermodynamical equilibrium. Its application to a system quite far from it, like the external air, is usually purely conventional. For instance, the local temperature is dened as the temperature of a thermometer exposed to air but shielded from radiation, at an height of 2 m above ground covered by grass, on a at surface, far from buildings and so on. e idea of cave temperature is usually considered more meaningful, because underground uids and radiation are essentially at the equilibrium condition. Nevertheless, if measures are pushed strongly ahead, signicant gaps from equilibrium appear. e main unbalance is surely the thermal sedimentation, i.e. the trapping of relatively warm air above relatively cold one. is can introduce signicant biases in the denition of local cave temperature, which then depends on the height above the oor. is can aect the estimations of underground lapse rates, it introduces biases in the estimation of human impact on very sensitive show caves, it can result in condensation processes in case of forced mixing of air sheets, it can explain probably Martels degree rule and can nally aect local thermal insulation, then explaining something about the location of bats and crystals. It is important to emphasize that thermal sedimentation cannot be a stable structure in an insulated system (it would violate the Second Principle), and it has to be fed by some external forcing. e paper shows that the local vertical temperature gradient in conduits is usually around 30 mK/m in alpine caves, but it can increase to 100 mK/m in special cases and to 0.4 K/m in very dry caves. In the Rio Martino laboratory we have shown a seasonal disappearance of thermal sedimentation, which corresponds to expectations. ermal sedimentation measurements are nevertheless quite complex to be taken. As regards this matter, the new generation of thermocameras is going to allow a signicant step ahead.

PAGE 142

15th International Congress of Speleology Earth Sciences 1403 2009 ICS Proceedings THE SOUND OF NATURAL CAV V ESGIOGIO VANNI ANNI BADINO ADINO1,2, RO RO BER ER TO O CHIGNOLA CHIGNOLA3,4 1LL a Venta EEsplorazioni GG eograche2DD ip. Fisica GG enerale UU ni-TO O, Via Pietro GG iuria 1, II -10125, Torino, II taly3DD ip. Scienza e Tecnologia, UU ni-VRR Strada LLe GG razie 15 CCV1, II -37134 Verona, II taly4UU nione Speleologica Veronese Complex karstic underground systems, i.e. deep and long caves with more than one entrance, are known to be characterized by intense underground airows. During the last decade, the physics of underground winds has started to be unravelled and this natural phenomenon has been shown to depend upon the interaction between the external and the internal atmospheres. We have used sonic anemometry to measure and record wind speed at various entrances of Italian caves with dierent known depths and lengths. Signals were sampled at 1 Hz and analyzed using classical method from signal processing theory. Spectral analysis of the overall signals showed a 1/fn decay with an exponent ranging between 0.7 and 2.5 for dierent caves and within dierent frequency ranges, the exponent being very close to unity for the air owing in the greatest underground systems analyzed. Spectrograms were obtained using the method of the sliding windows and the results subjected to rigorous statistical analysis. Signicant peaks were isolated from spectrograms at various frequencies for all caves, demonstrating that each cave can play sounds if properly excited by the external atmosphere. 1. Introduction Most cavers know from experience that big underground systems are characterized by intense air circulation. Indeed, this natural phenomenon is empirically exploited by cavers to guide their explorations of new caves or of yet unexplored branches of known karstic systems. During the last decade, however, the physics of underground winds have started to be unraveled and experimental data and theoretical models in this eld are rapidly accumulating. e air circulation in caves is basically driven by two independent forces: one is proportional to the density dierence between the internal and external atmospheres and the other to uctuations in the pressure of external atmosphere (Lewis, 1990). Both components uctuate in time and this means that a large cave system never attains equilibrium. Indeed, the underground atmosphere is unstable and receives energy from the outside (Badino, 1995). It is therefore expected that underground winds measured at caves entrances should contain information on atmosphere dynamics, both of the external and the internal ones, as well as on the geometry of the underground system itself (Plummer, 1969;, Lismonde, 2002). Sonic anemometers equipped with data loggers are now available and measure wind speeds and directions with great precision. We therefore exploited sonic anemometry to measure wind velocity in dierent underground systems and provide rst analyses of recorded signals using conventional methods borrowed from signal processing theory. Here we present some result of our rst data acquisition campaign (2006), with 2.5 million wind velocity measurements in ve dierent caves.2. Recording of SignalsMeasurement was accomplished with one or two sonic anemometers Gill Windsonic (accuracy 2% at 12 m/s, resolution 0.01 m/s), usually in the middle point of smallest conduit section, in order to increase accuracy and reduce noise. Acquisition frequency rate was set to 1 Hz. Signals were sampled for a total time interval of two days and nine hours. Measurements required continuous human control in order to avoid interferences (or thes) because the anemometers were almost always in the middle of wellvisited cave entrances. For this reason the acquisition could not last more than two or three days.4. Wind Speed Sampled at the Entrances of V V arious Cavese importance of choosing very large caves and with strong air ows became apparent quickly; the large inertia of air movements and the high wind velocities might in fact reduce noise created by local uctuations and increase the relative accuracy in the spectral analysis. e rst test was made in the Corchia underground system

PAGE 143

Earth Sciences 1404 2009 ICS P roceedings 15th International Congress of Speleology (near Lucca), the largest in Italy, which is characterized by very strong winds in its many entrances. e Corchia cave system is a conduit net extended inside the mountain over 1.2 km of vertical extent and 3 square kilometers of surface area; 53 km of conduits are surveyed, but the cave probably has a development 2 or 3 times larger. Our rst choice was the Eolo entrance, because it is the most classical and has the strongest air discharge. At this entrance, the airow, apparently quite continuous, is in fact highly modulated and shows a complex structure which is very likely to depend upon the dynamics of internal oscillators. e Piaggia Bella complex (near Cuneo) has 42 km of known conduits over 1 km of vertical relief and 10 square kilometres of surface area (so, it is much atter than Corchia). It has several entrances, and we chose to record wind speed at the Mastrelle entrance. e Grotta Gigante, near Trieste, is a vast cave essentially characterized by a huge internal room (approximately 600,000 m3) connected to the external karstic plateau by three entrances; regular oscillations of its internal wind have already been observed and discussed (Cigna, 1968). 5. Signal Analysise signals shown in Figs. 1-3 were analyzed using standard methods borrowed from signal processing theory. Preprocessing of non-stationary signals is usually recommended and the most common way to do it is to subtract the trend from raw data. Fig.4 shows the periodogram of the detrended signals. e periodgrams show a region where the power spectral density decreases with increasing frequencies. is trend follows the power law S ( f ) 1/ f and this equation has been nonlinearly tted to data within the frequency intervals to provide estimates of the exponent .6. Statistical Analysis of Spectrograms and Peaks Identication We used the local Gaussian white noise assumption as the null hypothesis. In essence, we assumed that within a narrow time interval the noise associated to the signal is Gaussian white noise. White noise provides at power spectra and its amplitude uctuations are exponentially distributed. e probability density distribution of the amplitudes in the spectrum of Gaussian white noise is given by: p ( Sk) 1 Sk e SkSk (1) where Sk is the mean power spectral density. us, it is straightforward to calculate the probability PSk of each peak in power spectra for the null hypothesis: PSk p ( Sk) SkdSk (2) A peak was labeled as signicant if its probability turned out to be below the threshold value PSk=0.005 (in other words, we have <0.5% probability that the peak is due to a random uctuation). e rationale for considering this value stems from the observed relative uctuation (i.e. the CV) of estimated power spectra that, on average, was 15 %. If this Figure 1: Data for Corchia System. Figure 2: Data for Piaggia Bella Complex. Figure 3: Data for Grotta Giganta.

PAGE 144

15th International Congress of Speleology Earth Sciences 1405 2009 ICS Proceedings uctuation is associated to a given peak with PSk=0.005, then its probability, at worse, would still be <0.01 and the peak would nonetheless be considered as highly signicant. Figure 5 shows the period of the signicant peaks isolated from all signals and plotted as a function of time. Periodic winds can be observed for all caves and in some instances such as for Grotta Gigante a great number of low period waves appear to concentrate within a given time interval. us, the caves play sounds and each entrance seems to be characterized by a dierent sound print which likely depends on the local shape of internal paths. 7. ConclusionsWe have started to investigate the sound of natural caves with the hope of highlighting possible correlations between the acoustic properties and the structure of underground systems. In principle, one might think to carry out a sort of remote acoustic survey of a cave. However, internal winds, that at rst sight appear to ow very regularly, show unexpected complex dynamics that pose important problems of physical interpretation. Overall power spectra of sampled signals decay with increasing frequencies following a power law. Depending on the exponent of the power law dierent scenarios can be drawn: in the 1/f regimen (pink or icker noise) the energy decreases linearly with frequency, then each octave contains the same energy. Pink noise is intermediate between the 1/f0 noise (white noise), in which the system has no memory of previous states and energy is equally distributed among frequencies, and the 1/f2 noise (brownian noise), in which the system has no memory of previous increments. We show that the overall power spectra of large underground systems tend to behave as pink noise, with the exception of the Grotta Gigante cave. Pink noise is rather ubiquitous and shows up in very dierent natural systems such as: the variation of time intervals between human heart contractions, the postural sway of a person standing on a platform, the electrical noise originated by ions dynamics in brain neurons, and many others. Pink noise has been interpreted by means of the superposition of dynamic processes that relax exponentially with dierent characteristic times (Milotti, 2002). is is very interesting because from the infra-acoustic point of view, complex caves can be considered as systems of Figure 4: e trends have been estimated by the moing average method on the number of samples. e trends were then subtracted om the signals. Figure 5: Exponent of the power law decay of spectral amplitude with increasing equencies as estimated by nonlinear tting of the spectra within the indicated equency ranges (see also Fig.1-3).

PAGE 145

Earth Sciences 1406 2009 ICS P roceedings 15th International Congress of Speleology many dierent damped and coupled oscillators, and this is probably the origin of the pink noise. Indeed, the Grotta Gigante is not a complex cave since it is essentially formed by a huge chamber connected to the karstic plateau by three entrances and thus it cannot be considered as a system of many coupled oscillators. is might explain why noise in this cave appears to be dierent (i.e. brown vs. pink) with respect to that measured in the vaster and more complex caves. Modulation of wind speed appears to vary in time and this is surely correlated to the dynamics of the external atmosphere (e.g. windy days), which is able to trigger oscillations. Our rst approach (measurements carried out at a single entrance, sampling at 1 Hz) was probably not appropriate. In fact, the data show a lot of noise in the high frequency range and important daily variations. is suggests an appropriate modication of the experimental approach. In future data acquisition campaigns sampling will be switched to 0.1 Hz and carried out on much longer acquisition times. Also, the use of two anemometers in the same cave, in tandem located at dierent entrances with simultaneous acquisition of outside meteorological data, is highly warranted.Acknowledgementis work has been partially supported by the Associazione Gruppi Speleologici Piemontesi.ReferencesBadino G., 1995; Fisica del Clima Sotterraneo, Memorie dellIst. Italiano di Speleologia, 7, serie II, Bologna. Cigna A.; An Analytic Study of Air Circulation In Caves Int. J. Spelol. 3, 1968. Lewis W., 1992; Atmospheric Pressure Changes And Cave Airow NSS Bulletin 53. Lismonde B., 2002; Arologie des systmes karstiques, CDS Isre, 361 pp. Milotti, E., 2002; A pedagogical review of 1/f noise, Arxiv preprint physics/0204033, 2002 arxiv.org Plummer, W.T., 1969, Infrasonic resonances in natural underground cavities, Journal of the Acoustical Soc. of America, v. 46 (5), pt. 1, pp 1074

PAGE 146

15th International Congress of Speleology Earth Sciences 1407 2009 ICS Proceedings THE CUEV V A DE LOS CRISTALES MICROMETEOROLOGYGIOGIO VANNI ANNI BADINO ADINOD D ip. Fisica GG enerale UU ni-TO O, Via Pietro GG iuria 1, II -10125, Torino, II taly Cueva de los Cristales, Naica, is one of the most interesting caves ever explored. During the Proyecto Naica, led by SpeleoResearch & Film, we undertook a complex series of measurements with the goal of understanding its current physical state. is was mostly carried out by the Department of General Physics of the University of Turin. Its natural state is at a depth of 170 m immersed in 54C mineralized water. Now it is lled with air, partially surrounded by ventilated galleries at 35 C. All kinds of micrometeorological processes are happening and unfortunately we are rather unprepared to follow the details because it is what in physics is called a transitional state. It is, in other words, experiencing a fall towards a new state of equilibrium that we still have not determined, but that we hope to be able to inuence in some way in the future. In the meantime, all the environmental parameters vary far more than expected and in an irreversible way, so the techniques usually used to study caves arent applicable here. e climate of a normal cave is in fact substantially static, with minimal oscillations, whether daily or seasonal, around a point of equilibrium. ey are oscillations related to the shape of the cave, but which also partly determine it, because they are able to start air currents and condensation processes which, over millennia, can signicantly alter the rock. But we are still speaking about systems near equilibrium and which are therefore relatively easy to study. e climate of Cristales is evolving in an irreversible way, as well as quickly. Our measurements have shown various phenomena. e cave continues cooling by approximately half a degree per year, because it loses heat by conduction towards the nearby mine galleries to the North-West, as well as by irradiation along the access corridor. We have also noticed that in the upper areas the air is stably warmer and more humid than the lower zones and those close to the exit. An unexpected nd has been that, while the temperature is very stable, even if in slow decline, the humidity shows strong variations, on both the short and seasonal time scales. is is probably due to meteoric water inltrations along the fractures created by the mining activity. Finally, there is an air current of about 50 L/s which starts when the access door is opened.1. Introductione Cueva de los Cristales Gigantes, or simply Cueva de los Cristales, is located in Mexico, near the Naica village in the state of Chihuahua, at an altitude of 1100 m asl, is one of the most interesting caves until now explored. It was discovered during the excavation of a tunnel in 2000. Since then it is closed with a steel door (not airtight) and more recently by a transparent veranda which protect the visitors from the exposure to the hostile atmosphere. In fact, at the moment, around 2 people per year are permitted to visit the cave during weekends, but without opening the last, transparent door. During the Proyecto Naica, led by SpeleoResearch&Film, we undertook a complex series of measurements with the goal of understanding its current physical state. is was mostly carried out by the Department of General Physics of the University of Turin.2. General e climate of a usual cave is in fact substantially static, with minimal oscillations, whether daily or seasonal, around a point of equilibrium. ey are oscillations related to the shape of the cave and its contact with the external environment, but which also partly determine it, because these uctuations around the equilibrium are able to create micro-meteora (transient processes like air currents and condensation) which, over millennia, can signicantly alter the rock. But we are still speaking about systems near the equilibrium and which are therefore relatively easy to study. e main characteristic of the Cristales micro-climate is

PAGE 147

Earth Sciences 1408 2009 ICS P roceedings 15th International Congress of Speleology that it is not stationary or quasi-stationary; it is evolving in an irreversible way, as well as quickly. e Cristales natural state is at a depth of 170 m immersed in 54 C highly mineralized water. Now it is lled with air, partially surrounded by ventilated galleries at 35 C. All kinds of micrometeorological processes are happening and unfortunately we are rather unprepared to follow the details because it is what in physics is called a transitional state. It is then experiencing a fall towards a new state of equilibrium that we still have not determined, but that we hope to be able to inuence in some way in the future. In the meantime, all the environmental parameters vary far more than expected and in an irreversible way, so the techniques usually used to study caves arent applicable here.3. Cristales Environmental Conditionse equilibrium temperature of cave in the Sierra Naica is given by the average temperature of inltrating waters (Badino, 2005). In Europe the yearly average temperatures of intrating waters and air are quite similar, but here they are not because precipitations are concentrated during summer. Corrections due to the Latitude (-4) and Altitude (+300 m) eects give a Sierra Naica average temperature 1 C higher than El Paso (data from worldclimate.com). Rainfall average temperature is generally 0.5 C less than the air, then we can assume that in Naica inltration is roughly at the same temperature of El Paso. We obtain then an average yearly temperature of 17.5 C, but an average rain temperature of 21 C if weighted with precipitation intensities. Figure 1 shows precipitations and average monthly temperatures. In the fall to the aquifer surface altitude (Cueva de las Espadas) the water temperature increases by about 0.75 C. We can then assume that the equilibrium temperature of a cave at 1250 m asl is 22 C. is has to be considered the asymptotic temperature of external heat source which drives thermal exchange processes in contact with the Naica caves. Some temperature measurements inside Cristales were rst taken in October 2002 (Testo 910, 0.01 C resolution), giving 47.1 C at the oor and 47.4 C at 2 metres with a Testo 910. We repeated the same measurements with the same instrument in January 2006, obtaining 45.5 C. e humidity, taken in June 2006 with an Assman psicrometre, was between 92 amd 94 %, then monitoring of relative humidity was possible. We have done this from May 2006 until now with a set of six Testo 175-H. Humidex is an index of the perceived temperature, which depends on temperature and humidity (Masterton et al., 1979). It is in degree Humidex; 35: uncomfortable; 40: strong and general discomfort, danger; 46-53: highly dangerous; >54: imminent heat shock, death (Fig. 2). Cristales Humidex is around 90-95, twice the lethal, which gives an idea of the technical problems faced during exploration. Our measurements inside Cristales have shown various phenomena, due to its non-equilibrium state and its Figure 1: Rainfall intensity and temperature at Naica. Figure 2: Perceived temperature.

PAGE 148

15th International Congress of Speleology Earth Sciences 1409 2009 ICS Proceedings probable thermal connection with larger structures in the north-eastern branch. Main phenomena are: Strong thermal sedimentation; General temperature decrease; Important air draught with open doors; Strong relative humidity variations. ermal Sedimentation In April 2006 we have begun a set of measures with an acquisition system Sigma 3000 (20 PT100, in four groups of four sensors inside the cave) of Lombard&Marozzini, which has given non-continuous data until August 2007 with a relative accuracy <4mK. e thermal sedimentation in Cristales is extremely strong and shows diurnal variation. In fact, the highest part of the cave is quite stable, at a temperature around 45.5 C, highly humid; also the cold trap at the bottom of the cave is quite stable at 44.6 C (Fig. 4). e intermediate region shows important variations. It looks like a hot bubble that lls the highest part of the cave (and surrounding regions) has a diurnal variation and lls or empties the cave, leaving the highest and lowest parts undisturbed, but lling the intermediate area with hot air, including the access corridor (Fig. 3). Figure 3: Diurnal temperature variations near the Cristales entrance. Figure 4: Temperature variations and values in dierent stations.

PAGE 149

Earth Sciences 1410 2009 ICS P roceedings 15th International Congress of Speleology It shows a strong, regular daily eect, not depending on direct intervention on the doors. e maximum is at 00.00 UTC (16.00 LT), with oen a sudden, very short drop. Aer that, the temperature falls down in few hours. ere is a secondary, less regular maximum around 0910 UTC (01-02 LT). It seems due to temperature daily variations, maybe with some trapping eects, or pressure variations inside the mine due to regular actions on the ventilation. ere is a vertical sedimentation with an average gradient of 1.4/10 C/m at the end of 2006, reduced to 0.9/10 C/m at the end of 2007; there was also a N-S gradient of 0.6/25 C/m (Fig. 4). e downward hot bubble dri causes condensation on crystals surface in the highest part of cave.5. Temperature variationsIn the period 2002-2008 the cave has cooled by approximately half a degree per year, because it was losing heat by conduction towards the nearby mine galleries to the North-West, as well as by irradiation along the access corridor (Fig. 5). e estimated heat losses (the two of around 100 W each), as compared with the temperature decrease allow to calculate in 69 J/K the total thermal capacity of the Cristales thermal system, roughly 30-50 times larger than the thermal capacity of known crystals. At the end of 2007 the mine conduits surrounding the cave have been closed to airow and their temperature quickly increased to approximately 40 C. is fact, accompanied probably with a careful management of internal door, which is now always kept closed, has almost stopped the temperature decrease of Cristales. e temperature at the top (North-East) has become stable around 45.5 C, whereas the temperature near the door, where there was a strong heat loss, has increased by 0.7 C to 45.2 C in January 2009 (Fig. 6). Condensation onto crystals in the upper parts of the cave is now virtually stopped. Fig. 5 shows the general temperature trend since the beginning of operations, in the region South, near the entrance. Geochemical data suggest that the Cristales temperature when drained by waters was around 56 C. An unexpected nding has been that, while the temperature is very stable, even if in slow decline, the humidity shows strong variations, on both the short and seasonal time scales. is is probably due to meteoric water inltrations along the Naica Fault and to movements of the hot-humid air bubble, which is in contact with Cristales (Figs. 7, 8, 9) 6. Air draughtWith the door nearly closed it is possible to perceive an air current coming from the cave. In May 2007 we have measured the air ow (50 l/s) and the moving pressure with a Testo instrument for dierential pressure Figure 5: T-eld in Cristales and surroundings. Figure 6: Temperatures in Cristales stations. Figure 7: Temperature vs Time in Cristales Center-South.

PAGE 150

15th International Congress of Speleology Earth Sciences 1411 2009 ICS Proceedings measurements. e Testo 6349 was outside the main door, connected with the cave by a tube (Fig. 10). As expected, the pressure changes a lot due to infra-acoustic eects inside the mine, nevertheless it has an average value around 10 Pa, which corresponds to an air column 10-15 m high with a temperature dierence around 10 C. Also mines explosions have large eects on the cave. We have seen two types of behaviour: the overpressure ows out from cave aer 6 s from the explosions, the overpressure coming to the cave from the mine galleries and bounces on the door. In the two cases, the overpressure or its bouncing is able to open the ajar door. e cave general structure as deduced by micro-meteorological data is shown in Figure 11.7. Conclusionsese measurements have therefore shown us that the part of the cave which we know is only a fragment of a much vaster structure, which, depending on environmental conditions, introduces or extracts air from the environment we call Cristales, including the surrounding cave like Ojo de la Reina. Moreover, the air currents show that the cave is connected to the mine through another passageway, probably fractures. Basically, these measurements show that the cave continues. e heat loss, which was able to damage the crystal surfaces with vapor condensation,is probably nished, but a systematic monitoring of environmental parameters inside the cave is necessary to preserve it and, if necessary, to drive the internal atmosphere to a new state compatible with cave maintenance and study.ReferencesBadino G., 2005; Underground Drainage Systems and Geothermal Flux; Acta Carsologica, 34/2, 1, 277 316 Masterton, J.M., and F.A. Richardson, 1979: Humidex, A Method of uantifying Human Discomfort Due to Excessive Heat and Humidity, CLI 1-79, Environment Canada, 45 pp. Garofalo P.S., D. Gunter, P. Forti, S.-E. Lauritzen, and S. Constantin 2007; e uids of the giant selenite Figure 8: Temperature and Relative Humidity variations in Cristales. Figure 9: Relative Humidity in the Top Station, 3 sensors. Figure 10: Dierential Pressure on the main door aer a sudden closure. Figure 11: Probable shape of Cristales and connected structures.

PAGE 151

Earth Sciences 1412 2009 ICS P roceedings 15th International Congress of Speleology crystals of Naica (Chiuhahua, Mexico), ECROFIXIX 17 Garofalo P.S., P. Forti, S.-E. Lauritzen, S. Constantin 2007; e uids that generated the giant selenite crystals of Naica (Chihuahua, Mexico), Proc.Congresso FIST Rimini Settembre 2007

PAGE 152

15th International Congress of Speleology Earth Sciences 1413 2009 ICS Proceedings Decline ECLINE of OF ca CA Ve E ice ICE a A case CASE study STUDY from FROM the THE Austrian USTRIAN Alps LPS Europe UROPE based BASED on ON 416 years YEARS of OF obser OBSER Vation ATIONMICHAEL BEHm M1, VERONIk K A D D Itt TT Es S2, RO RO BER ER T GREILINGER GREILINGER2, H H ELGA H H ARtm TM ANN 2, L L Uk K As S PLAN3, D D IEt T ER SULZb B ACHER2 1Vienna UU niversity of Technology, II nstitute of GG eodesy and GG eophysics, GG usshausstrasse 27-29, AA -1040 Wien, AA ustria, mbehm@ mail.tuwien.ac.at2Speleological Society of Vienna and LLower AA ustria, OObere DDonaustrasse 97/1/61, AA -1020 Wien, AA ustria, oce@cave.at3NN atural HH istory Museum Vienna, DDepartment for karst and caves; Museumsplatz 1/10, AA -1060 Wien, AA ustria, lukas.plan@ nhm-wien.ac.at Due to their geographic position and high elevation, the Austrian Alps in Europe host a large number of ice caves. About 14,000 caves are currently known in this region, and some 900 of them feature permanent ice llings. Our study object is the Geldloch cave in Eastern Austria, which is a part of a 27 km long and 662 m deep cave system. e main entrance of Geldloch opens at 1446 m a.s.l. e cave shows dynamic air ow due to openings at dierent elevations. Ice is mainly found in a chamber (ice dome, 50 x 20 m) about 100 m behind the entrance and in the passage leading into it. e cave was visited rst in 1592, and a report including detailed drawings was published. e visitors were impressed by the ice formations, including a large frozen lake. Expeditions in 1747, 1808, 1855, 1897, and 1900 also observed the ice formations and the ice lake which oen set an end to their exploration. First photographs of the ice were taken in 1902. In the second half of the 20th century, regular cave exploring trips allowed to make accurate notes on the ice decrease. e comparison reveals that in general the ice is declining. e most recent visit in October 2008 showed that the cave is completely free of ice aer summer nowadays. We present the dierent data sources (drawings, photographs, and descriptions) and quantitatively estimate the ice loss based on the height changes of the ice lake.1 Cave Characteristics and Regional Settingse GG eldloch (Fig. 1) is part of the 27 km long and 662 m deep tscher CC ave System that is located in Lower Austria (Europe). Mount tscher reaches 1893 m and is a rather isolated high alpine mountain in a pre alpine environment. e tscher CC ave System (Hartmann & Hartmann, 1985; 1990; 2000) comprises two major caves: GG e ldloch and Taubenloch, which both open on the southern hillside at 1446 and 1505 m a.s.l. respectively. A connection between those two caves was found as recently as 1994 (Hartmann & Hartmann, 2000). e tscher is part of the Northern Calcareous Alps and consists of Upper Triassic, well karstied limeand dolostones of the Dachstein Formation. e alpine climate is chilly with heavy snow falls in winter. e average temperature at the meteorological station of the nearby village of Lackenhof at 805 m asl is 6.0 C (Land Niedersterreich, 2009). Assuming a gradient of 0.35/100 m, the temperature at the altitude of 1500 m extrapolates to 3.5 C. Average annual precipitation at Lackenhof is 1814 mm. Figure 1: Overview map of the Geldloch as part of the tscher Cave System. Map by E. Herrmann, Survey: Speleological Society of Vienna and Lower Austria.

PAGE 153

Earth Sciences 1414 2009 ICS P roceedings 15th International Congress of Speleology 2. Description of the Ice Formationse GG eldloch has two entrances: the main entrance, a large portal 10 m wide and 5 m high at 1450 m asl and a very small upper opening at 1663 m a.s.l. called Sisyphusloch. e latter was discovered and articially enlarged in 1988 (Hartmann & Hartmann, 1990). Within the cave, six sections can be distinguished, of which especially the 700 m long main gallery and its connection to the Sisyphusloch are of interest for our study. Ice formations are restricted to the beginning of this section. eir growth and decline is controlled by the air current between the main entrance and upper opening: the known Sisyphusloch and many others that are not passable by humans. e main entrance leads to a 10 m wide gallery which is partly covered with snow in winter and spring. Approximately 100 m behind and 30 m below the entrance the II ce LL ake is situated (Fig. 1). e II c e LL ake is historically described as an accumulation of water in summer, mostly frozen in winter (e.g. Levenstein & Riedl, 1879). e biggest reported extent is 25 to 35 m while it can cover the whole width of the gallery which is about 10 m. Behind the II c e LL ake, early visitors reached the II c e Wall, a frozen waterfall that was gone by 1992 (Greilinger, 1992). It was ca. 12 m wide and reached an impressive height of up to 10 m over two cascades. On top two ice stalagmites grew having several meters in diameter. e II c e DDome, a chamber measuring 50 m x 20 m, was reached by climbing the II c e Wall. Early descriptions mentioned the II c e DDome to be covered with massive ice formations. Behind the Ice Dome, the cave divides into two major parts. To the North, a series of vertical drops leads down to a sump 444 m below the entrance. To the West a large horizontal gallery terminates aer some 600 meters in a boulder chock. In the second half of this section, a series of chimneys leads to the upper entrance Sisyphusloch 217 m above the main entrance. Several other chimneys with air dra have not been explored yet. 3. History of Exploration and Remarks on Ice FormationIn the past GG eldloch was considered to host one of the most magnicent subterranean ice formations (Haselbach, 1876, p. 204) which is one of the reasons why its rst exploration dates back as early as 06-09-1592. is fact also explains the importance of the GG e ldloch in the history of the scientic exploration of European caves (Shaw 1992, p 255f). A review of the exploration history is given in Hartmann & Hartmann (1992). During all expeditions, notes on the ice formations were taken, but not in a systematic way. e ice oen was a major obstacle. e reports were sometimes sparse and some were lost. Nevertheless, it is known that, at the time of the rst expedition (1592), the II c e LL ake was frozen and the Ice Pyramids had a circumference of ve meters (Fugger, 1891; Soukup, 1887; Haselbach, 1876; Hartmann & Hartmann, 1985). In 1747 an expedition was ordered by Emperor Franz I. of Austria. It was led by A. J. Hacker and J. A. Nagel. e group failed to cross the II c e LL ake but produced detailed sketches of the entrance part which show the ice formations in the entrance part and at the II c e LL ake (Fig. 2). e II ce LL ake was reported to have completely vanished in 1847 (Haselbach, 1876). Relling obviously occurred relatively quickly. e lake itself had dierent manifestations throughout the centuries (unfrozen, totally or partially frozen surface, dierent ice thicknesses). In September 1855 another map was drawn by Becker and Figure 2: Basic map of Geldloch om Nagel, 1747.

PAGE 154

15th International Congress of Speleology Earth Sciences 1415 2009 ICS Proceedings Hrtler (Becker, 1859), including the ice formations, but the map covers only the entrance parts. It is simple but gives an impression of the ice formations at that time. In 1897 the maps were extended by Crammer & Sieger (1899) who also investigated temperatures and ice appearance (Fig. 3). e entire horizontal part of the cave was fully explored by the beginning of the 20th century and mapped from 1900 to 1902. is was also when rst photographs of the cave were taken (Anonymous, 1953). e vertical section was explored and mapped in 1923. Aer that, research became more extensive with several expeditions (e.g. 1953, 1974). Major degeneration of the ice was reported before 1960 (Pfau, 1960). Up to then, the reports are frequent and diverse. Greilinger (1992) sums them up, revealing a massive decline at the II ce LL ake since the beginning of the 20th century. In 1992, 400 years aer the rst exploration, only little of the ice in the GG e ldloch was le. In 1988 the introduction of the battery driven drilling machine led to the discovery of the Sisyphusloch. In 1994, the link between GG e ldloch and Taubenloch was discovered. e most recent visit in October 2008 showed the GG e ldloch being completely free of ice. Figure 4 shows a summary of historic data available on the ice formation of the II c e L L ake.4. Historic Explanation of Periodic Ice Formation in the GeldlochAs opposed to the nearby Taubenloch, the GG eldloch hosted permanent ice. Until 1747, it was believed that the ice of the II ce LL ake existed in summer only, melting down completely in winter. Various explanations were given on this phenomenon. Most of them involved a cooling of the rocks of the tscher in winter which conserves the ice during the hot summer months. Water on the otherwise frozen lake was explained as melting water originally coming from the snow at the entrance. Crammer & Sieger (1899) proved these theories wrong aer doing extensive meteorological research. is led to the modern theories on air circulation in caves and ice caves in particular (Crammer, 1899). 5. Discussion and ConclusionsAlthough there are many observations dating back to 1592, only few of them provide accurate data on the thicknesses of ice at the II c e LL ake. Further diculties arise from ice level changes due to seasonal variations, sometimes contradicting descriptions and the fact that the observations are based on infrequent and irregular visits. Despite these limitations, the data summarized in Figure 4 allow for the following interpretation: e II c e LL ake had its largest extent around the mid of the 19th century and has been declining since then. is coincides with the progress of glaciation in the Alps. In general, the glaciers advanced until around 1850 and have been retreating since then. e mid of the 19th century is also the end of the little ice age in the Northern Hemisphere which followed the Medieval Warm Period. e II ce LL ake seems to have completely disappeared in the mid of the 20th century. e constant decline of ice in the adjacent ice dome Figure 3: Section of Geldloch om Crammer & Sieger, 1899. a: Entrance, d-e: Ice Lake; e-f: Ice Wall; I: Ice Dome. Figure 4: Level (aboe ground) of the Ice Lake in the years 1747 until 2008.

PAGE 155

Earth Sciences 1416 2009 ICS P roceedings 15th International Congress of Speleology supports this interpretation. Few but reliable observations indicate that the II c e LL ake can completely disappear and reform within a few years. ReferencesBecker, M.A. (Ed.)(1859) Der tscher und sein Gebiet, aus eigener Beobachtung und bisher unbentzten uellen geschp. Erster eil. Zwettel. Crammer, H. and R. Sieger, (1899) Untersuchungen in den tscherhhlen. Globus, Braunschweig, Vol: LXXV: 313. Crammer, H. (1899) Eishhlenund Windrhren Studien. K.K. Geographische Gesellscha, Wien, I: 19. Fugger, E. (1891) Eishhlen und Windrhren, I. XXIV Jahresbericht der K.K. Ober-Realschule in Salzburg. Salzburg. Greilinger, R. (1992) Eisuntersuchungen im Geldund Taubenloch. HH hlenkundliche Mitteilungen des L L andesvereins fr HH hlenkunde in Wien und NN, 48: 219. Haselbach, K. (1876) Oetscherfahrten. Blatt des Vereins fr L L and eskunde NN iedersterreich 1876: 201. Hartmann, H. and W. Hartmann, (1985) Die Hhlen Niedersterreichs, Band 3. Landesverein fr Hhlenkunde in Wien und N, Wien, 432 pp. Hartmann, H. and W. Hartmann, (1990) Die Hhlen Niedersterreichs, Band 4. Landesverein fr Hhlenkunde in Wien und N, Wien, 624 pp. Hartmann, H. and W. Hartmann, (1992) Geschichte der tscherhhlen in vier Jahrhunderten. H H hlenkundliche Mitteilungen des LL andesvereins fr H H hlenkunde in Wien und NN, 48: 190. Hartmann, H. and W. Hartmann, (2000) Die Hhlen Niedersterreichs, Band 5. Landesverein fr Hhlenkunde in Wien und N, Wien, 616 pp. Land Niedersterreich (2009) Wasserstandsnachrichten und Hochwasserprognosen Niedersterreich. www.noel. gv.at/Externeseiten/wasserstand/wiskiwebpublic/ maps_N_0.htm. checked 2, 2009. Levenstein, J. and V. Riedl (1879) Das Geldloch im Oetscher. AA lpen-Z Z eitung 1879/9: 2. Nagel, J. (1747) Beschreibung des auf allerhchsten Befehl Ihro Maytt: das Rm Kaysers und Knigs Francisci I untersuchten tscher-Berges. Von dem tscher Berg. sterr. National Library, HS 7920. Pfau, H. (1960) Das Geldloch im tscher. HH hlenkundliche Mitteilungen des LL andesvereins fr HH hlenkunde in Wien und NN, 7: 72. Shaw, T. (1992) History of Cave Science. e Exploration and Study of Limestone Caves, to 1900. Sydney Speleological Society, Sydney, 2nd Edition. Soukup, R. (1887) Vom tscher und seinen Hhlen. O O esterreichische Touristen-Z Z eitung 1880: 183. Anonym (1953) Die Expedition 1953 zur Erforschung des Geldlochs im tscher (Niedersterreich). DD ie H H hle, 3: 40.

PAGE 156

15th International Congress of Speleology Earth Sciences 1417 2009 ICS Proceedings EFFECTS OF SURFACE MORPHOLOGIES ON FLOW B EHAV V IOR IN KARST CONDUITSAARONAARON J. BIRD IRD1, GREGOR GREGOR Y S. SPRINGER RINGER2, RACHEL RACHEL F. BO O SCH CH3, RANE RANE L L CURL CURL4 1OOakland UU niversity, 2200 Squirrel RR d., RR M 363 HH annah HH all, RRochester, MI I 48309 UU SA A bird2@oakland.edu, 2OOhio UU niversity, DDepartment of GG eological Sciences, 316 CC lippinger LL aboratories, AA thens, OHOH 45701 UU SA A springeg@ohio.edu31269 Woodeld St., LL ake OOrion, MI I 48362 UU SA A rachel.bosch@gmail.com4UU niversity of Michigan, DDepartment of CC hemical EE ngineering, AA nn AA rbor, MI I 48109-2136 UU SA A ranecurl@umich.edu e 1974 work from Blumberg and Curl on scallops in karst conduits is revisited using threedimensional, time-dependent computational uid dynamics simulations. e original work presented the results from theoretical analysis and experimentally observed scallop formation as an expression including Reynolds number, friction factors, and scallop size. is expression is oen used to determine karst conduit ow velocities based on observed scallop sizes in caves. In the current work, ow behavior is simulated using the CFD code, STARCCM+, for vadose and phreatic cave passages containing scalloped walls in order to create comparisons with and extend the results of the original work. Results of simulations are used to demonstrate potential impacts of momentum-dominated speleogenesis.1. IntroductionLong-term development of karst groundwater ow conduits is governed by discharge, the hydraulic gradient, rock properties (e.g., solubility and porosity), and chemical aggressiveness of the water (Palmer, 1991). Each makes a signicant contribution in dening the end-product groundwater system. Porosity, i.e. percent of void volume, of the medium is what allows the karst development to occur. e pores are commonly characterized by their shapes and volumes. eir connectivity denes rock permeability (White, 2002). Mechanically-derived groundwater systems with ow path apertures of <1 cm have been referred to as fracture systems while voids having diameters of cm have been referred to as conduits. Conduit ow is governed by viscous drag on walls, oors, ceilings, and ow-path obstructions, including pathway bends. ese are oen described by hydraulicow and geometry based variables. ey aect travel time through the groundwater system, and subsequently, inuence the formation of the conduit itself. Furthermore they have a signicant impact on larger-scale descriptors of groundwater systems, namely determination of tracer breakthrough curve tailing behavior, and estimation of total system discharge. A large amount of research has been conducted to determine the eects of various hydraulic-ow and geometry factors, including Field et al. (1997), Hauns et al. (2001), Peterson et al. (2006), Geyer et al. (2007), and others. Much of this work describes the application of numerical approaches to predict the impact of conduit surface features on tracer breakthrough curves. A smaller group of researchers, e.g., Hauns et al. (2001), have generated detailed simulations of critical portions of large conduit systems, such as potholes, meanders, and pools. Work in this area is important for determining the friction factors that may be present due to conduit morphologies and conduit surface morphologies, including scallops, which are concavities eroded into conduit surfaces by ow associated with near-wall detachment and reattachment patterns (White et al. 2005). In this regard, Blumberg and Curl (1974) developed a friction factor expression for ow in conduits with fully-developed scalloped walls, dependent only on ow Reynolds number. e current paper revisits the work of Blumberg and Curl (1974) with 3D ow simulations using STARCCM+V3 (CD-adapco, New York, NY) with the goal of determining a level of validation with the original ndings and to examine ne-scale phenomena that could not be observed or measured in the original experiments.2. Previous Work Blumberg and Curl (1974) conducted a theoretical and experimental investigation to obtain an expression for how a mean size of scalloping depends upon conduit ow velocity, by applying dimensional analysis and the universal law of the wall for turbulent ow over rough surfaces. Such an approach was critical for proceeding past the recirculating dilemma caused by coupling dissolution with generation of turbulence inducing surface features. eir work involved the calculation, measurement, and dimensional determination of key descriptors of the uid behavior. ese included channel Reynolds number, friction velocity and

PAGE 157

Earth Sciences 1418 2009 ICS P roceedings 15th International Congress of Speleology friction-velocity Reynolds number, mean channel velocity, scallop (or ute) period length, a Sauter mean characteristic length, and a roughness function. eir resulting expression provides a friction factor, as follows: (1) where is the Reynolds number based on conduit diameter, d, and on mean ow velocity, and where is a roughness function for pseudo-smooth pipe ow. e above equation can be further simplied to predict paleoow conditions based on scallop size (Curl, 1974a, 1974b). In this equation is a Reynolds number for mean scallop size as follows: (2) where is measured as the maximum length of the ith scallop determined by (3) and is density, is mean ow velocity, and is dynamic viscosity. Previous to the 1974 work, Curl (1966) had observed a unique relationship between Reynolds number and scallop formation, independent of kinematic viscosity, therefore results obtained from the experimental observations in the 1974 work can be extended to other conditions beyond those described here.3. eoretical BackgroundDetails of uid ow systems can be predicted numerically in nite-volume formulations where matrix solutions from multi-diagonal matrices are solved based on discretizations of the Navier-Stokes equations with mass, momentum, and energy being continuous and conserved. More extensive details of CFD are described by Ferziger and Peric (2002), Anderson (1995), Pope (2000), and Patankar (1980). Regarding CFD in speleology, Jeannin (2001) and Houns et al. (2001), as well as others, have presented the governing equations in terms of karst-conduit ow applications. In this work, the CFD package, STARCCM+V3 (CD-adapco, New York, NY) was used. is CFD tool was selected for its ability (1) to create closed, clean surfaces of dirty CAD representations, (2) to create polyhedral control volumes that promote orthogonality, (3) to study transition turbulent ows, and (4) to calculate near-wall modeling parameters needed for turbulence models. For viscous ows, two near-wall layers are commonly described and need to be accurately measured (experiment) or calculated (simulation). ey are the viscous sublayer and the turbulent boundary layer located just above. In the past it had been commonplace to populate a near-wall region with large numbers of cells in order to accurately calculate viscous shear-dominated ow in this region. However, such an approach is very costly in time and computer memory resources. To overcome these limitations, the use of a blended wall function, originally based on law-ofthe-wall approaches, that requires many fewer cells in the boundary and near-wall layers is applied here. is approach is described in CD-adapco (2007) and receives detailed attention in Popovac et al. (2007). For making useful comparisons with results from Blumberg and Curl (1974), the near-wall velocity (which also can give friction-velocity Reynolds number, etc.) and the scallop-length (peak to peak) velocity are calculated in the simulation. e latter is the result of advective ow calculation in a turbulent regime, while the former is determined from the following equation used in STARCCM+V3 (CD-adapco, 2007): (4) where energy generation, is initially by default 0.42, and energy dissipation E = 9.0, is a measure of the distance to the wall in the sublayer, represents the intersection of the viscous sublayer with the turbulent region just above, and b is a further expansion of the discretization. 4. Computational Grid3D surfaces of the gypsum scallop elds generated in the work by Blumberg and Curl (1974) (Figure 1) were scanned by the University of Michigan 3D Lab (Ann Arbor, MI) and stored in a stereolithography le (STL). Scans were made at a resolution of 1e-3 m. e resulting STLs were imported into STARCCM+V3 where a computational region was added above the scallop eld to create a uid ow region. An inside surface wrapping procedure was applied to extract the scallop eld and ow region of interest. A closed surface with surface cell size of 1e-3 m resulted. Based on surface, a volume mesh was built from polyhedral cell control volumes to promote orthogonality and was grown at a rate of 17% from the surface to the maximum cell size of 0.01 m. Figure 1 shows the result of the surface wrap of the scanned scallop eld for the 1 m/s Blumberg-Curl experiments.

PAGE 158

15th International Congress of Speleology Earth Sciences 1419 2009 ICS Proceedings 5. ResultsSimulations were obtained for individual scallops, a scallop eld, fully submerged potholes, and free surface potholes. Comparisons with Blumberg and Curl experiments were possible for the two former cases, while qualitative comparison with Hauns et al. (2001) is made for the two latter cases. Since Reynolds number provides a common comparison, CFD calculation results for 1 m/s BlumbergCurl scallop elds are presented in Figure 2. The location of measurements in the 1 m/s Blumberg and Curl (1974) scallop field were at 53.4 cm downstream from the leading edge of the gypsum block. Of note, the velocity used in Blumberg (1970) and adopted by Blumberg and Curl (1974) is which is the maximum measurable near-wall velocity and which is not a traditional near-wall friction velocity. Using this value of velocity, Blumberg and Curl reported near-wall of 2120 and 2320 for two different roughness functions, of 8.8 and 8.9. CFD simulation in this work slightly under predicts with a value of 1405 for a corresponding = 9.0. For the location above the scallop where was determined, CFD agrees well with Blumberg and Curl. Their reported values were 20500 and 21000, while CFD gives 22300. This shows a reasonable qualitative validation of the results, and allows further expansion of applications of the current CFD models. While it is usually thought that conduit meanders, intersections, and sharp bends have the most impact on the velocity of the water ow, CFD results here show that as scallop eld population increases for small numbers of scallops, the drag increases and subsequently uid velocity decreases. Drag force coecient for the Blumberg and Curl scallop eld is 0.0182, which contains approximately 360 scallops. Drag coecients calculated for elds containing 1, 3, 5, and 7 scallops were 0.002409, 0.002577, 0.002742, and 0.002778, respectively. Decrease in ow velocities, as drag increases with increase in scallop number, are visible in Figure 3a,b. Figure 4 shows centerline ow velocities for the 1m/s Blumberg-Curl scallop eld. Figure 1: Looking downstream over the 3D scanned and discretized surface scallop elds om Blumberg and Curl, 1974. Upstream is at the bottom of the image and downstream at the top. Figure 2: Velocity vectors om centerline section slice of calculated ow over scallop eld. Upstream is top-right side of image. Scallop where Blumberg and Curl made experimental measurements is just to the le and below of center. Figure 3: From le (a), for single scallop centerline ow velocities at heights aboe scallop basin of 0.0001, 0.05, 0.1, 0.15, 0.2, 0.225, 0.325, 0.675, and 0.775 m; and right (b) 7 scallops. Simulations with 3 and 5 scallops further demonstrate this trend.

PAGE 159

Earth Sciences 1420 2009 ICS P roceedings 15th International Congress of Speleology Overall mean ow velocities are reduced when features are present on conduit walls. Simulation results for 1, 3, 5, and 7 scallops show a 3% mean-ow velocity reduction (Fig. 3). For a fully developed scallop eld, signicant ow eects are present throughout the ow eld due to turbulence generated from the scallops. However direct impact of a scallop eld in fully developed turbulent conditions is not as signicant as are lightly populated scallops. is is due to the acceleration of the water at the leading edge of a scallop (Fig. 5) as it is squeezed between the inertially dominant channel ow and the slower moving eddy in the scallop. Detachment of the ow in the scallop occurs where the acceleration is highest, which is at the leading edge. Reattachment occurs at a location approximately 2/3 scallop length downstream from the leading (Fig. 5). Experimental observations suggest this to be the location where scallop propagation, i.e. signicant dissolution, occurs in the scallop. is location corresponds with a condition of locally dominant momentum being perpendicular to the aected medium, i.e. the wall of the conduit. Another location where locally dominant momentum is perpendicular to the conduit surface is the oor of a pourover or waterfall. Field investigation by Brucker et al. (1972), also suggests vertical enlargement of shas, i.e. where momentum is perpendicular to the aected medium, is a dominant event in karst conduit evolution. Figure 6 shows simulation results and turbulent kinetic energy (TKE) plot for a 1-meter high pour over. Together with the falling mass of the water and the already high momentum, a great deal of force is available to act on the oor and walls of a conduit.6. ConclusionsScallops are a relatively common geometric and hydrodynamic feature of karst conduits. e work of Blumberg and Curl (1974) led to further understanding of the formation of scallops as well as an expression to relate scallop size to ow Reynolds number, which allows estimation of paleoow discharge. e current work reported in this paper has qualitatively validated the original experimental results using CFD. is allows further extension of the CFD simulations to additional conditions, including singular scallops and their eect on ow conditions, as well as other conduit features such as pourovers. In the 1974 work it was observed that reattachment occurred at approximately 2/3 scallop-length downstream from the leading edge of the scallop. is is the location where new scallop development is most signicant. e location of reattachment in the scallop is seen at the Figure 4: Simulation results om the Blumberg-Curl 1 m/s scallop eld for centerline velocities at heights aboe scallop basin of 0.005, 0.01, 0.015, 0.02, 0.025, 0.03 m. Figure 5: Vector plot of ow in and aboe scallop. Vector size indicates magnitude. Figure 6: Vector plot of ee surface ow in a pouroer (a), and (b) plot of turbulent kinetic energy in the pouroer.

PAGE 160

15th International Congress of Speleology Earth Sciences 1421 2009 ICS Proceedings same location in the CFD results. It is of not that this is the point where locally dominant momentum is perpendicular, or at a high angle, to the aected medium, i.e. the wall of the conduit. is condition of locally dominant momentum being perpendicular to the surface is present not only at the reattachment point in scallops but also at the base of pourovers and waterfalls.ReferencesANDERSON, J.D. (1995) CC omputational Fluid DD ynamics e Basics With AA pplications. McGraw-Hill, Inc. BLUMBERG, P.N., and R.L. CURL (1974) Experimental and theoretical studies of dissolution roughness. Journal of Fluid Mechanics 65, 735. BRUCKER, R.W., J.W. HESS, and W.B. WHITE (1972) Role of Vertical Shas in the Movement of Groundwater in Carbonate Aquifers. GG r ound Water 10(6), pp 5. CD-ADAPCO (2007) USER GUIDE, STARCCM+ Version 3.02.003, pp 1272-1282. CURL, R.L. (1966) Scallops and Flutes. CC a ve RResearch G G roup of GG reat Britain Transactions 7, pp 121. CURL, R.L. (1974a) Deducing Flow Velocity in Cave Conduits from Scallops. e NN SS Bulletin 36(2), pp 1. CURL, R.L. (1974b) Errata: R.L. Curl, (1974a). e NN SS Bulletin 36(3), p 22. FERZIGER, J.H., and M. PERIC (2002) CC o mputational Methods for Fluid DD ynamics. Springer-Verlag, New York. FIELD, M.S. and S.G. NASH (1997) Risk Assessment Methodology for Karst Aquifers: (1) Estimating Karst Conduit-Flow Parameters. EE n ironmental Monitoring AA ssessment 47, pp 1. GEYER, T. and S. BIRK, T. LICHA, R. LIEDL, M. SAUTER (2007) Multitracer Test Approach to Characterize Reactive Transport in Karst Aquifers. G G r oundwater 45, pp 36-45. HAUNS, M. and P.-Y. JEANNIN, O. ATTEIA (2001) Dispersion, retardation and scale eect in tracer breakthrough curves in karst conduits. Journal of H H yd rology 241, pp 177. PALMER, A.N. (1991) Origin and morphology of limestone caves, GG eo logical Society of AA merica Bulletin 103, pp 1. PATANKAR, S.V. (1980) NN u merical HH eat Transfer and Fluid Flow. Hemisphere Publishing Corporation, New York. PETERSON, E.W. and C.M. WICKS (2006) Assessing the importance of conduit geometry and physical parameters in karst systems using the storm water management model (SWMM). J ournal of HH ydrology 329, pp 294. POPE, S.B. (2000) Turbulent Flows. University Press, Cambridge. POPOVAC, M. and K. HANJALIC (2007) Compound Wall Treatment for RANS Computation of Complex Turbulent Flows and Heat Transfer. Flow, T urbulence, and CC ombustion 78, pp 177-202. WHITE, W.B. (2002) Karst hydrology: recent developments and open questions. EE n gineering G G eology 65, pp 85-105. WHITE, W.B. and E.L. WHITE (2005) Groundwater ux distribution between matrix, fractures, and conduits: constraints on modeling. Spe leogenesis and EE volution of Karst AAquifers 3, 6 pgs.

PAGE 161

Earth Sciences 1422 2009 ICS P roceedings 15th International Congress of Speleology PRELIMINARY WATER Q UALITY AND BOUNDARY REDEFINITION OF THE SCOTT HOLLOW DRAINAGE BASIN, APPALACHIAN PLATEAU, USAMELI ELI SA A R R BI I SHO HO P1, IRA IRA D D SA A SO O WSKY1, WILLIA ILLIA M K. JONE ONE S2 1DDept. of GG eology and EE nironmental Science; UU niversity of AA kron; AA kron, OHOH 44325-4101 UU SA A ; melisa_bishop979@ sbcglobal.net; ids@uakron.edu2EE nironmental DData PO O Box 356; Warm Springs, VA A 24484 UU SA A ; wkj30@hotmail.com Scott Hollow Cave is a major (>43 km) vadose conduit system developed within a mature karst upland in Greenbrier and Monroe Counties, West Virginia, USA. Previous mapping and dye tracing revealed that the cave drains an estimated 48.2 km2, with a major conduit transmitting water northward to discharge at the Greenbrier River. e majority of conduits are developed in the lower units of the Greenbrier Group (Mississippian), especially the Hillsdale and Sinks Grove Limestones. e present study sought to rene the boundaries of the overall basin, identify sub-basins within the system, and to evaluate specic water quality parameters. A series of qualitative dye-traces were undertaken to identify hydrologic connections between the surface and subsurface and to help better dene the drainage basin boundary. Multiple dyes, including Sodium Fluorescein, Rhodamine WT, and Eosin, were used, along with charcoal detectors. Within and outside the cave 14 locations were identied for water quality sampling. Coliform counts, and nitrate were repeatedly measured. e incised Second Creek must limit the eastern margin of the basin. Even so, no injections near this boundary have owed in that direction; they have moved westward toward the cave. Extreme complexity of ow, apparently caused by structural features, is found. For example, three dierent dyes were placed in close injection points in a single sinkhole, the entrance sink of the main cave entrance, and were recovered in four dierent places. In the southern part of the area the basin appears to wrap around the nose of the plunging Sinks Grove Anticline. Time-averaged nitrate values vary from 4 to 47 mg/L at the in-cave locations fed by known surface catchments. 1. IntroductionKarst areas are recognized for their susceptibility to groundwater contamination. e current research project sought to characterize and explain the variability of water quality entering Scott Hollow Cave, an extensive dendritic cave network. e cave is developed in the Big Levels region Figure 1: Partial Scott Hollow Cave map showing the in-cave sample locations.

PAGE 162

15th International Congress of Speleology Earth Sciences 1423 2009 ICS Proceedings of southeastern West Virginia, which is a rolling limestone upland surface that hosts many of the long caves of North America (Fig. 1, inset). is karst plain is a rural area dominated by agriculture, mostly cattle grazing and feed crops (Figure 2D). e animal waste from pastures has the potential to seep into the soil or be carried underground by sinking streams and impact the aquifers below. Water samples were collected at 10 locations within the cave. 2. Location and Overview of Study Areae Scott Hollow drainage basin is located in a moderately deformed portion of the Appalachian Plateau physiographic province. e stratigraphy consists of Mississippian interbedded limestones (Fig. 2B) and shales of the Figure 2: A. Map of the hypothesized drainage basin (unpublished map by Michael Dore 2009) illustrating dye traces conducted. Geologic contacts and structural features om Ogden (1976) B. Photo taken in the North South passage within Scott Hollow Cave (scale about 3 meters) C. Inset of where the Scott Hollow dye trace was conducted D. Photo illustrating the local surface topography.

PAGE 163

Earth Sciences 1424 2009 ICS P roceedings 15th International Congress of Speleology Greenbrier Group (Reger and Price, 1926). Many of the extensive cave systems present in Monroe County occur within the Hillsdale and Sinks Grove Limestones, including Scott Hollow Cave (Fig. 1). e Scott Hollow drainage basin is a mature karst plain that drains an estimated 48.2 km2 surrounding the small town of Sinks Grove (Jones, 1997). Overall drainage is towards the north, where the deeply incised Greenbrier River serves as local base level (Fig. 2A). Earlier work (Jones 1997) showed that the cave drains to Glorias Spring on the south bank of this river. e two dominant folds within the basin are the Sinks Grove anticline and the Maple Grove syncline (Fig. 2A), both trending to the NE (Ogden, 1976). Scott Hollow Cave developed in the west limb of the Sinks Grove anticline with its main passage, Mystic River, running parallel to strike (Davis, 1999). Many of the eastern tributaries of Scott Hollow Cave ow down dip before reaching Mystic River. Joints, faults and fractures also inuence conduit orientation and morphology (Ogden, 1976). Demrovsky (2003) discovered that nitrate levels in cave streams peaked when large quantities of snow melt saturated the overlying soil recharging the aquifer. is increase is attributed to the fact that nitrates are stored in soils during dryer periods and are released during wetter periods, indicating seasonal inuences on concentration levels (Pasquarell & Boyer, 1996).3. Methods 3.1 Dye tracingA series of qualitative dye traces were conducted using a method based on that of Mull et al. (1988). is information is being used to 1) dene the boundaries of the Scott Hollow drainage basin, 2) establish hydrologic connections between the surface and sub-surface, and 3) delineate groundwater ow paths. ree uorescent dyes were used for this study; Sodium Fluorescein (CI Acid Yellow 73), Rhodamine WT (CI Acid Red 388), and Eosin (CI Acid Red 87). Background levels of uorescence were checked with activated charcoal. All traps were switched with new ones every few weeks and sent to the Environmental Data lab in Warm Springs, Virginia to be analyzed using a spectrouorometer. e traps were continually replaced until dye was detected. First Dye Trace Scott Hollow: e rst dye trace was conducted in the closed valley entrance sinkhole of Scott Hollow Cave (Fig. 2B). Approximately 1 liter of each dye was injected in 3 dierent locations on March 22, 2008. Eosin was used at a stream coming from a wooded area at the east end of the entrance valley, Sodium Fluorescein was placed where the stream enters Scott Hollow Pond, and Rhodamine WT was used in the stream that exits the pond and sinks underground. Detectors were placed at the following 8 locations throughout the cave: Popcorn Alley, Root Canal, Mastodon Ave., Chris Trunk, upstream Mystic River across from Johns Flowstone, downstream Mystic River across from Craigs Creek, NS Sump, and Middle Earth. Two detectors were placed at Motel 8 and Main St., the upper Scoop City section of Scott Hollow Cave, in the event that the dye was to travel in that direction. Second Dye Trace Eastern Boundary: e second dye trace was conducted in the southeastern section of the study area. 0.7 liters of Rhodamine WT was poured into a sinking stream exiting the pond at Anderson Sink in the eastern section of the drainage basin boundary. 2 liters of Eosin was used in a small sinking stream near the entrance of Rolling Rock Cave, and the same amount of Sodium Fluorescein was used at Dores Sink on May 18, 2008, which are also located in the eastern portion of the basin. Traps were placed in the upstream section of the cave. ese locations include Mystic River across from Craigs Creek, Mystic River across from Johns Flowstone, Iceland, Land of Lost Footing, Greenwater Junction in Little Mystic River, and at the Upstream Sump. ree traps were placed on the surface along Second Creek at Hokes Mill, Nickels Mill, and at the mouth of Second Creek in the event that groundwater ow went in the opposite direction.3.2 Water quality samplingWater samples were collected on a monthly basis from 10 locations within the cave and 4 surface locations. ree of the surface locations are known insurgences into the cave system and the last location is the only known discharge location, Glorias Spring, along the Greenbrier River. Each month water samples for nitrate analysis were collected, and analyzed using a Dionex DX-120 Ion Chromatograph. A coliform bacteria water analysis was performed alternate months from samples collected on the surface and within the cave. e Coliscan Easygel method (Micrology Laboratories) was used. e analytical mixture was placed in a small incubator at 35C (95F) for 24 hours. All pink and purple colonies were counted and reported as total coliform per mL of water used. 4. Results and Discussion 4.1 First dye trace Scott Hollow

PAGE 164

15th International Congress of Speleology Earth Sciences 1425 2009 ICS Proceedings e rst set of detectors monitoring background levels contained no dye. e second [9 days] and third [18 days] set of detectors tested positive for dye. e Rhodamine WT used in the sinking stream exiting the pond was detected at Mastodon Ave. and was visibly detectable within a few hours of injection in this stream. Rhodamine WT was also detected in Mystic River across from Craigs Creek, the NS Sump, and was also detected in the Scoop City section of Scott Hollow Cave at Popcorn Alley. e Sodium Fluorescein used at the stream entering the pond was detected in Mystic River across from Craigs Creek, and NS Sump. Rhodamine WT and Sodium Fluorescein were also detected at Popcorn Alley where there is a stream coming from the upper Scoop City section of the cave (Fig. 2A). e detection of both Sodium Fluorescein and Rhodamine WT in the upper Scoop City section and the lower section of Scott Hollow Cave at Mastodon Ave. and at the NS Sump indicates that water from the same sinkhole drains into two dierent parts of the cave. Eosin was not detected on any of the traps. is was probably because a high concentration of Rhodamine WT may have masked a relatively lower concentration of any Eosin present in the sample. is was taken into consideration for planning the second dye trace.4.2 Second dye Trace eastern boundary Dye was detected on the rst [12 days], second [27 days] and third [41 days] set of traps. Rhodamine WT was detected at the Upstream Sump, Land of Lost Footing, and Iceland, which are all located in the upstream section of the cave. Dye was also detected at down stream Mystic River across from Johns Flowstone and across from Craigs Creek. Greenwater Junction was the only location that Rhodamine WT was not detected. e Eosin and Sodium Fluorescein were detected at Greenwater Junction, Land of Lost Footing, Iceland and downstream Mystic River across from Johns Flowstone and Craigs Creek. No dye was detected at any of the surface locations along Second Creek (Fig. 2A). e Eosin used at Rolling Rock Cave and Sodium Fluorescein used at the Dores Sink were not detected at the Upstream Sump and were both detected at Greenwater Junction. All three dyes eventually made their way down Mystic River and were detected across from Johns Flowstone and Craigs Creek. It appears that groundwater could be owing around the southern nose of the Sinks Grove Anticline from the Anderson Sink, eventually making its way to the Upstream Sump at Mystic River. Groundwater ow from Rolling Rock Cave and the Dores Sink seem to be taking a similar path, both ending up at Greenwater Junction in the Little Mystic River and eventually joining Mystic River. is Eastern boundary dye trace suggests that Scott Hollow Cave may possibly be much larger than originally speculated and the potential for the discovery of new cave passages is very probable.4.3 Nitrate analysisNitrate in the groundwater from March through December varies between 0 mg/L and 47 mg/L (Fig. 3). It appears that subsurface water sample locations containing the greatest amount of nitrates originate from areas with a high intensity of cattle. Locations with lower concentrations are in sub-basins that are mostly wooded or have fewer cattle. It also appears that seasonality aects the amount of nitrates present in cave streams. Samples collected from March through May all have values that dont substantially vary. is can be attributed to steady amounts of rainfall common in the spring. Beginning in June, nitrate values start to rise and then gradually decrease in August (late summer) through October (fall) (Fig. 3). is time of year precipitation was rather low, allowing nitrates to become trapped in the overlying soils and can account for the decrease in nitrates in the groundwater. Sample locations Craigs Creek and Middle Earth values were 10 to 15 mg/L higher than all the other sample locations. Based on current cave survey data, aerial photos and topographic maps of this area it was possible to determine the headwaters for these 2 areas. e land use for these locations is primarily for cattle grazing and other live stock, indicating that runo from these pastures is entering the cave system. e cattle are rotated to dierent parts of the farm throughout the year and could attribute to these higher values at these specic locations. In November and December, nitrate values increased substantially due to an increase in rainfall and snow melt occurring aer a month with little to no precipitation, ushing nitrates from the surface into the aquifer.4.4 Coliform Analysisere are gaps in the total coliform data for this study due to analytical problems. Interpretation is based on 4 months of data; March, May, July, and December 2008 (Fig. 4). Coliform colony totals range from about 550 (per 3 mL of the water sample) to 0. Coliform levels in March and May are between 0 and 300 per 3 mL of sample and continue to decrease in July in correlation to low precipitation rates. Samples collected in December had the greatest amount of coliform present which was a result of increased precipitation, which ushes the microbes into the cave via swallets and sinking streams.

PAGE 165

Earth Sciences 1426 2009 ICS P roceedings 15th International Congress of Speleology 5. ConclusionsSubterranean drainage in the Scott Hollow basin is complex as a result of the local structure and stratigraphy. e rst traces examined ow on the scale of a single sinkhole. ey show that minor variations in position of an insurgence can result in owpaths that diverge on the scale of 100s of meters. is is likely caused by the presence of multiple permeable bedding planes near the base of the Hillsdale Limestone. On the scale of the mapped cave, ow from sinkhole inputs is mainly downdip along bedding planes, with collected ow traversing northward along strike to resurgence at the Greenbrier River. e second set of traces examined portions of the drainage which go beyond the mapped cave. ese results indicate that a) e basin may extend beyond the original estimated 48.2 km2, b) It is likely that the cave has a much more extensive footprint than is Figure 3: Graph of nitrate concentrations om March December 2008 and total precipitation (in). Precipitation data proided courtesy of Doug Boyer (Lewisburg Airport station). Triangles indicate sampling days. Figure 4: Graph of the total coliform data for March, May, July, and December 2008.

PAGE 166

15th International Congress of Speleology Earth Sciences 1427 2009 ICS Proceedings currently mapped, and c) Flow appears to wrap around the plunging southern end of the Sinks Grove Anticline, going southward and then northward towards the Greenbrier River. e nitrate and total coliform data collected within this drainage basin indicate that agricultural practices can have a direct aect on groundwater quality. e uctuations in nitrate concentrations and total coliform values also illustrate how seasonality and precipitation rates have an eect on the groundwater quality of this area. During spring precipitation rates dont vary much. Nitrate concentrations during this time ranged from 0-25 mg/L, but each individual location didnt vary much due to the fact that nitrates were readily washed into the aquifer with every precipitation event. Late summer through fall, precipitation rates began to slowly decrease to almost no rainfall in October. October nitrate values range from 0 15 mg/L. November and December had signicant precipitation events following a long dry period which allowed high quantities of nitrates to be ushed into the groundwater, yielding higher concentrations ranging from .2 37 mg/L.AcknowledgementsWe would like to thank Mike Dore and Mike McFall for their input and assistance with eld work; Pat Dore for her hospitality; Tom uick for lab and technical assistance. Our appreciation is expressed to landowners Mike Christie, Mel McDowell, Lewis McCormick and Brad Smith for their cooperation. e Department of Geology and Environmental Science, e Cleveland Grotto and Doug Boyer provided additional support. ReferencesDavis, S., 1999, Aquifer development in folded and fractured limestone e Scott Hollow Drainage Basin, Monroe County, West Virginia: M.S. esis, University of Akron, 210 pp. Demrovsky, N., 2003, e eects of well drilling on karst aquifers in Monroe and Greenbrier Counties, West Virginia: M.S. esis, University of Akron, 198 pp. Jones, W.K., 1997, Karst hydrology atlas of West Virginia: Special Publication 4, Karst Waters Institute, 111 pp. Mull, D.S., T.D. Lieberman, J.L. Smoot, and L.H. Woosley Jr, 1988, Application of dye-tracing techniques for determining solute-transport characteristics of ground water in karst terranes: EPA 904/6-88001, Ground-Water Protection Branch, Region IV, Atlanta, GA., 103 pp. Ogden, A.E., 1976, e hydrogeology of the central Monroe County karst, West Virginia: Dissertation, West Virginia University, 262 pp. Pfa, J.D., 1993, Determination of inorganic anions by ion chromatography, Environmental Monitoring Systems Laboratory Oce of Research and Development, USEPA, Cincinnati, Ohio, 27 pp. Pasquarell, G.C. and D.G. Boyer, 1996, Herbicides in karst groundwater in southeast West Virginia: Journal of Environmental uality, vol 25, pp 755. Reger, D.B. and P.H. Price, 1926, County Reports: Mercer, Monroe, and Summers Counties, West Virginia Geological Survey, 963 pp. Shank, D., 2002, Hydrologic and structural controls on the evolution of a karst aquifer, Windy Mouth Cave, West Virginia: M.S. esis, University of Akron, 121 pp.

PAGE 167

Earth Sciences 1428 2009 ICS P roceedings 15th International Congress of Speleology FLOW CHARACTERIZATION FROM EPIKARST AND SHALLOW BEDROCK SPRINGS ACROSS A RANGE OF HYDROLOGIC CONDITIONS, SAV V OY EX X PERIMENTAL WATERSHED, OZARK REGION, U.S.VAN AN BRAHANA RAHANA1, RAL RAL PH H DA DAVI I S1, PHILLI HILLI P D D HA HA YS2, KENNE ENNE TH H F. STEELE EELE1 1DDepartment of GG eosciences, UU niversity of AA rkansas, Fayetteville, AA rkansas 727012UU .S. GG eological Survey, 113 OO zark HH all, Fayetteville, AA rkansas 72701 Abstract is study describes the results of a series of spring-discharge measurements, continuous monitoring of eld parameters, and tracer studies that were conducted at the Savoy Experimental Watershed (SEW), and it oers explanations for the observed variations in springow for varying hydrologic, temporal, and extreme stress conditions. is specic study is part of a longer-term assessment of hydrologic budget components and speleogenesis at SEW, a long-term, well-instrumented karst research site in the Ozark region of northwestern Arkansas, midcontinent United States. Discharge measurements for Pond, Tree, Woodpecker, Memory, Dribble, and Red Dawg Springs (epikarst) were conducted twice a day, during the early morning (7:00 t.b. during July and August; 8:00 t.b. during December because of late sunrise), and the early aernoon, at about 2:00 .b. Data collection periods were chosen reecting the growing season (21 days in July and August, 2005 and 2006) and the dormant season (5 days in December 2005). Volumetric measurements were made with 500 to 2000 mL graduated glass cylinders, accurate to +/2 mL; time measurements were made with a stopwatch readable to 0.01 seconds. e mean of 10 repetitions per spring (volume/time) for each measurement provided an accurate, reproducible assessment of discharge. Results indicate that (1) diurnal uctuation of these epikarst springs is dominant during the growing season and minimal during the dormant season when deciduous trees have lost their leaves and transpiration is minimal. is loss of water from the shallow ground-water system is interpreted to be evapotranspiration from the ground-water system, and ranges from 5 to 25 mL/s over the course of a diurnal cycle for each spring; (2) extreme low-ow conditions are manifest dierently in dierent spring basins, and these provide an understanding of the ow mechanisms that may be active in karst settings; (3) the wide range of hydrogeologic response in epikarst springs to identical stresses indicates that our models of these systems are likely grossly oversimplied. Continuous monitoring and multiple tracer studies at Langle and Copperhead Springs (shallow bedrock) over long time periods and varying hydrologic conditions illustrate the varying hydrogeologic responses to diering ground-water levels in rapid-ow karst aquifers, and provide strong evidence that predictive numerical modeling in such settings at grid scales less than several kilometers is risky and misleading. Application of models to test hypotheses does seem apprpriate, but only within the range of hydrologic conditions that have been observed and used to develop the conceptual model.

PAGE 168

15th International Congress of Speleology Earth Sciences 1429 2009 ICS Proceedings Fluorescence LUORESCENCE Characterization HARACTERIZATION of OF Karst ARST Auifers UIFERSTERRI BROw W N, SID JONEs S L L ARRy Y MCKAy YD Department of EE arth and Planetary Sciences e UU niversity of Tennessee, Knoxville TNN 37996 UU SA A e uorescent properties of natural dissolved organic matter (DOM) were investigated as indicators of organic carbon ux and pollutant/pathogen transport through groundwater basins. e study focused on the character and potential monitoring value of naturally-occurring uorescence in karst aquifers typical of the Southern Appalachian region. Well and spring samples were collected during severe drought conditions, and provide a reference for DOM behavior at base ow. e evidence suggests that DOM uorescence changes predictably on multiple temporal scales. In carbonate settings, DOM composition changes seasonally across the entire region, and more suddenly in response to local recharge events. Groundwater was dominated by fulvic acid-like in winter months, reecting fresh inputs of leaf litter and woody plant material from leaf fall. Refractory, humic acid-type uorophores were only detected aer signicant recharge events in a minority of the sources. Fluorescent DOM exhibited a compositional shi towards protein-like compounds during warmer weather, possibly due to increased soil microbial production, root exudates, and soil moisture decits. e uorescent properties of low-risk wells varied less than those of all springs. e uorescent properties of some high-risk springs were observed to change over very short time intervals (hours), while others exhibited little uctuation regardless of time frame. With more work, uorescent DOM characterization could provide a useful tool for watershed monitoring and modeling, with benets for the design of low-concentration dye tracer tests, and renement of sampling strategies for organic pollutants and waterborne pathogens. 1. IntroductionSampling groundwater for background uorescence is recommended as to test the potential for analytical interference between tracer dyes and ambient uorophores, both naturally-occurring and anthropogenic (Aley, 1999; Smart and Karunaratne, 2002). In practice, lowlevel uorescent signals are subtracted from dye spectra without much consideration of the information being lost (Alexander, 2005). Natural uorophores are ubiquitous in aquatic systems, consisting of dissolved organic matter (DOM) derived from decomposed plant litter and the exudates of roots and soil microbes (Chen et al., 2003; Zech et al., 1997). Humic acids (HA) and fulvic acids (FA), plus the tryptophan-like (P1) and tyrosine-like (P2) proteins make up the bulk of this complex mixture. e uorescent properties of DOM are attributed to reactive, oxygen-rich functional groups (carboxylic acids, phenols, amines) that help drive essential soil processes and biogeochemical cycles (Kramer et al., 1990; Mobed et. al, 1996; McDowell, 2003). DOM is a key player in the processes of metal complexation, immobilization of microbial pathogens, and the subsurface retardation and transport of radionucleides (Scott et al., 1998; Jardine et al., 1990), and though its specic role in the global carbon cycle is not fully understood, DOM is considered a major link in the transfer of carbon and nutrients between terrestrial and aquatic ecosystems (Smart et al., 1976; Jae et. al., 2008). Fluorescent properties are expressed in terms of excitation and emission wavelengths (ex/ em) and intensity (I) of the spectral peak. Natural uorescence is most commonly observed in the long ultraviolet and blue parts of the spectrum, with ex in the 200 nm range, and em from 350 nm (Newsom et al., 2001). Exceptions include algal products and chlorophyll which are detected by extending the scan range up to 650 nm, near the analytical limits of most instruments. Fluorescent DOM has been distinguished in lake water samples from Antarctica to Colorado (McKnight et al., 2001), the Amazon River basin and Gulf of Mexico (Coble, 1996), peat catchments in northern England (Baker and Spencer, 2004) and caves in Hungary (Tatar et al., 2002). e position and intensity of spectral signals can be correlated with relative age, function, composition and

PAGE 169

Earth Sciences 1430 2009 ICS P roceedings 15th International Congress of Speleology origin of uorophore solutions (Senesi, 1991; Mobed et al., 1996; Frimmel, 1990; Chen et al, 2003; Kalbitz et al., 1999). Correlations between DOM and aquatic environments have been reported despite dierences in analytical techniques and instrumentation, suggesting the need for wider applications of uorescence characterization in monitoring karst watersheds. e objectives of this study were to examine the temporal and spatial variability of natural uorescence in karst aquifers, and to assess the viability of uorescence characterization as an environmental tracer and indicator parameter for assessment and compliance monitoring. e research was conducted in partnership with the Tennessee Department of Environment and Conservation (TDEC) Division of Water Supply under the auspices of the Source Water Protection Program.2. Study Area We sampled 23 groundwater-based community water supply sources in the Upper Tennessee River Basin (UTRB; Figure 1). e Tennessee portion of the watershed encompasses an area of about 2 million hectares between Johnson County (36o 33 N latitude) and Hamilton County, Tenn. (35o 12 N latitude). Altitudes in the northern part of the study area exceed 1200 m near Mountain City, TN and vary around 200 m in the south near Chattanooga. e climate is temperate and humid, with average annual temperatures from 9o C to 17o C. Mean annual precipitation (1935) ranges from 988 mm in the north to 1244 mm in the south (Johnson, 2002). e sampling program was conducted from fall 2006 through fall 2007 during a period of severe drought. e study area encompassed Paleozoic carbonate aquifer settings in the Valley and Ridge physiographic province. Bedrock generally dips 40o to 50o to the southeast, and is heavily folded and fractured by multiple thrust sheets aligned sub-parallel to the northeastward trend of the Appalachian orogen (Lloyd and Lyke, 1995). Subsurface drainage networks and direct surface water-groundwater connections are common, and most oen manifested by karst topography. Karst aquifers are mantled by clay-rich soils (Order, Ultisols) ranging from very thin to more than 30 m thick. Figure 1: Locations of community water supply sources sampled for uorescence characterization, Upper Tennessee River Basin, USA.

PAGE 170

15th International Congress of Speleology Earth Sciences 1431 2009 ICS Proceedings Spring and well waters are generally of the calciummagnesium-bicarbonate type, with relatively high hardness measured as total dissolved solids (TDS) or Specic Conductivity (SpC; Lloyd & Lyke, 1995; Krawczyk and Ford, 2006). A summary of physical and geochemical eld parameters measured during this study is presented in Table 1. Recent surveys by both the USGS and UT detected EE coli in up to half of all water supply wells and the majority of springs, along with low concentrations of enteric viruses (Johnson, 2002; Johnson et al., 2005). Despite their inherent vulnerability to surface-derived contamination, karst aquifers are the primary and most productive sources of private and community water supplies in the region.3. MethodologyRaw water samples were collected directly from community water supply wells and springs on a quarterly basis, in addition to one week-long automated sampler study. Field parameters (pH, electrical conductivity, and temperature) were measured prior to sample collection with a multiparameter meter (YSI Model 63). Replicate raw water samples were collected in pre-cleaned 40 ml glass vials for use in analyses, as eld blanks, and for comparison of sample degradation and holding times. Samples were ltered in the eld using 0.2 um glass ber syringe-mounted lters and immediately placed on ice, protected from light, and transported to the UT Hydrogeology Program laboratory in Knoxville. Samples were temporary stored in a dedicated refrigerator at 4oC. One replicate sample from each source was acidied to just above pH = 2.0 for total organic carbon analysis (TOC; Shimadzu) via EPA Method 5310. Ultra-violet (UV) absorbance (Beckman Coulter 640B) was measured at 254 nm and 270 nm per Standard Methods 5910B and others (Kalbitz et al, 1999). Fluorescence analysis was performed on a luminescence spectrophotometer (Perkin Elmer LS55) with a xenon lamp. Instrument settings are listed in Table 2. Samples were analyzed in both single scan and in 3-D mode to illuminate spectral details over a range of excitation wavelengths at xed wavelength intervals ().4. Resultsroughout the latter half of the study period, most sources exhibited the spectra of at least two distinct uorophores, usually FA in combination with one or more proteins or HA. A snapshot of groundwater DOM under low base ow conditions was obtained by plotting excitation and emission wavelengths at maximum uorescent intensity (single scan spectra) for August and September 2007 (Fig. 2). e data were grouped into ve distinct components. Protein-like uorophores, P1 and P2, plotted at the lowest ex/em, reecting their low molecular weight.and intense uorescence. FA and HA compounds plotted near the center of the graph at higher and lower excitation wavelengths, respectively. A h component detected only in later summer samples (ex/ em= 275/600) was thought to be related to algal products (AL), since some spring-based supplies were experiencing nuisance algae problems at the time. Other possibilities include the introduction of anthropogenic material or vegetative signs of drought stress. e seasonal dynamics of DOM are evident from progressive changes in ex/em, presented as scatter plots in Figure 3. e well data is generally more tightly clustered than the spring data, which is probably a function of the wells having smaller immediate recharge areas and less mixing. Humic acid only appeared in response to storm events in springs, and occurred in only one high-risk well. Fulvic acid dominated all groundwater DOM in winter, SPRINGSpH Temperature ( O C) Conductivity (S/cm/s2) Absorbance TOC (mg/L) Max 8.919.3 567 0.361.75 Min 7.111.0 112 <0.0040.31 Avg 8.015.1 253 0.070.94 WELLS Max 8.619.7 573 0.163.36 Min 7.211.3 187 <0.0040.37 Avg 7.915.6 332 0.021.11Table 1: Summary of 2006 Geochemical Data.scan range: 200 nm 620 nm : 15 nm ex slit: 4 nm 5 nm (dependent on maximum intensity) e m slit: 3 nm 4 nm (dependent on maximum intensity) sc an speed: 1000 nm/min emission increment: 10 nm number of scans: 1 15 Table 2: Instrument Settings for Perkin-Elmer LS55B Spectrophotometer.

PAGE 171

Earth Sciences 1432 2009 ICS P roceedings 15th International Congress of Speleology while warmer weather samples showed a distinct blue shi towards proteins. is pattern was mirrored in the absorbance data, a surrogate for dissolved organic carbon (not shown), which dropped over time as biodegradation progressed, temperatures increased and recharge events became increasingly scarce. High FA uorescent intensity occurred in winter and exhibited the most variability, apparently driven by the inux of organics from freshly fallen leaf litter during winter storms. High I values in summer were mainly related to protein uorescence, with a low co-ecient of variation. e most intense uorescence was sustained in forested catchments as opposed to those sources surrounded by grasslands. Natural uorescence at one high-risk spring within a protected forested watershed greatly quenched the uorescence of low concentrations (10 g/ml) of uranine dye. e short-term dynamics of groundwater DOM were measured at this spring (Clear Creek spring) from June 1 through June 6, 2007. ree small rain events (22.9, 0.76, and 55.9 mm at 24, 48, and 72 hours, respectively) occurred during the monitoring period and were recorded at the nearby water ltration plant. Fluorescent intensity and emission wavelength are plotted as a time-series in Figure 4. Diurnal uctuations were not apparent at this resolution, yet distinct peaks resulted from storm ows. Fluorescent intensity peaked at the spring within 10 to 20 hours of the storms, indicating rapid recharge through fractures and conduits most likely enhanced by the soil moisture decit. e DOM signature was that of FA, except for a broad (600 nm), highly uorescent pulse that appeared in response to the third and largest rain event, suggesting a contribution of algae or autochthonous turbidity (Pronk et al., 2005). 5. Conclusionsis regional survey of the uorescence properties of karst groundwaters revealed predictable, seasonal patterns in DOM evolution that could be exploited for water supply protection and management. While the annual trend was similar across the study area, local variations indicate the potential value of using natural uorescence to trace pathways in specic groundwater settings. While annual to millennial changes in climate, vegetation and land use aect DOM source material and supply, short term uctuations occur in response to recharge events and diurnal cycles. In well-characterized systems, the introduction of organic anthropogenic contaminants such as sewage, pesticides and detergents could be viewed as incidental uorescent tracers in and of themselves. Few studies have observed the entire spectral range of natural uorophores in groundwater. Fulvic acids were found to be the dominant uorophore in these karst sources, consistent with the ndings of others (Baker and Genty, 1999; van Beynen et al., 2001; Tatar et al, 2002; Lapworth et al., 2008), however, the presence of intense protein uorescence in summer months was unexpected. In these particular wells and springs, all closely regulated drinking water supply sources, the proteins appear to be associated with increased soil microbial activity (McKnight et al., 2001) rather than nonpoint source impacts such as wastewater euent, storm water disposal and agricultural runo (Baker and Lamont-Black, 2001; Baker and Spencer, 2004). e humic acid fraction of uorescent DOM Figure 2: Typical DOM Components in East Tennessee karst aquifers (Late Summer 2007). Data points represent single synchronous scan uorescence maxima labeled accordingly: P1 = Protein-like (Tyrosine); P2 = Protein-like (Tryptophan); FA = Fulic acid-like; HA = Humic acidlike; AL = Algal by-products. Figure 3: DOM Dynamics as Seasonal Fluctuations in Fluorescence Maxima. Figure 4: Short-term DOM Dynamics at Clear Creek Spring, June 2007.

PAGE 172

15th International Congress of Speleology Earth Sciences 1433 2009 ICS Proceedings occurred in groundwater in association with signicant recharge events, and in keeping with its hydrophobic nature, appeared to remain bound by sorption and other mechanisms to clays and biolms in the epikarst (Perrin et al., 2003; Kramer et al., 1990; McKnight et al., 1997). e nal segment of this study proved that some sources can be highly sensitive to very small recharge events, especially at base ow conditions. e lag time between baseline and elevated uorescent intensity is expected to be unique to each source, depending on environmental and site-specic factors such as season, vegetation, ambient moisture conditions and position in the watershed. ese short-term uctuations in uorescent intensity and DOM content have implications for the timing of low-concentration dye traces and the application of uorescent biomarkers. Data analysis is incomplete and the nal presentation of this paper will include correlations with geochemical parameters such as turbidity and temperature. e results suggest that with more work, uorescence characterization at various locations throughout a watershed could be used as a natural tracer of suspected nonpoint sources and discrete inputs of pollutants and other organic matter.ReferencesAlexander LEXANDER Scott COTT C. ) Spectral deconvolution and quantication of natural organic material and uorescent tracer dyes. Proceedings of t he Tenth Multidisciplinary CC onference, San Antonio, Texas., B. Beck (Ed.), ASCE Geotechnical Special Publication 144, 441. Aley L EY Thomas HOMAS ) Groundwater Tracing Handbook, Ozark Underground Laboratory, 1572 Aley Lane, Protem, MO 65733. Baker AKER A. and D. GENTY (1999) Fluorescence wavelength and intensity variations of cave waters. J ournal of HH ydrology 217, 19. Baker A KER A., and J. Lamont AMONT Black LACK ) Fluorescence of dissolved organic matter as a natural tracer of ground water. GG r ound Water 39, 5, 745. BAKER, A., and R. SPENCER (2004) Characterization of dissolved organic matter from source to sea using uorescence and absorbance spectroscopy. Science of t he Total EE nironment 333, 217. Chen H EN W., P. Westerhoff ESTERHOFF J. Leenheer EENHEER and K Brooksh ROOKSH 2003) Fluorescence excitationemission matrix regional integration to quantify spectra for dissolved organic matter. E E nironment, Scence and Technology 37 (24), 5701. COBLE, P.G. (1996) Characterization of marine and terrestrial DOM in seawater using excitationemission matrix spectroscopy. M arine CC hemistry 51 (4), 325. Frimmel R IMMEL F.H. (1990) Characterization of organic acids in freshwater: Current status and limitations. O O rganic AA cids in AAquatic EEcosystems, E.M. Perdue and E.T. Gjessing (Eds.), John Wiley & Sons Ltd: Chichester. JAFFE, R., D. McKNIGHT, N. MAIE, R. CORY, W.H. McDOWELL, and J.L. CAMPBELL (2008) Spatial and temporal variations in DOM composition in ecosystems: e importance of longterm monitoring of optical properties. Journal of G G eo physical RResearch 113, 1. JARDINE, P.M., G.V. WILSON, J.M. McCARTHY, R.J. LUXMOORE, D.L. TAYLOR, and L.W. ZELAZNY (1990) Hydrogeochemical processes controlling the transport of dissolved organic carbon through a forested hillslope. J ournal of CC ontaminant H H ydrology 6, 3. JOHNSON, G.C. (2002) Water quality of springs in the Valley and Ridge Physiographic Province in the Upper Tennessee River Basin, USGS Water Resources Investigations Report 02-4180. Johnson O HNSON T.B. (2005) Detection of enteric viruses in East Tennessee public ground water systems, MS Degree esis, University of Tennessee, Dept. of Earth & Planetary Sciences, Knoxville, Tenn. Kalbitz A LBITZ K., W. Geyer EYER and S. Geyer EYER ) Spectroscopic properties of dissolved humic substances a reection of land use history in a fen area. Biogeochemistry 47, 219. Kass A SS W. (1998) Tracing Techniques in Geohydrology, A.A. Balkema, Rotterdam, Netherlands. Kramer R AMER J.R., P. Brassard RASSARD P. Collins OLLINS T.A. Clair LAIR and P. Takats AKATS Variability of

PAGE 173

Earth Sciences 1434 2009 ICS P roceedings 15th International Congress of Speleology organic acids in watersheds. OOrganic AA cids in AAquatic E Ecosystems, E.M. Perdue and E.T. Gjessing, (Eds.), John Wiley & Sons Ltd: Chichester. Lapworth A PWORTH D.J., D.C. Gooddy OODDY A.S. Butcher UTCHER and B.L. Morris ORRIS ) Tracing groundwater ow and sources of organic carbon in sandstone aquifers using uorescence properties of dissolved organic carbon (DOC), AA p plied GG eochemistry, in press. Lloyd L OYD O.B. and W.L. Lyke YKE (1995) Ground Water Atlas of the United States, Segment 10 Illinois, Indiana, Kentucky, Ohio, Tennessee. Hydrologic Investigations Atlas 730-K, U.S. Geological Survey, Reston Virginia. McDOWELL, W.H. (2003) Dissolved organic matter in soils future directions and unanswered questions. G G eod erma 113, 179. Mc C Knight N IGHT D.M., E.W. Boyer OYER P.K. Westerfhoff ESTERFHOFF P.T. Doran ORAN T. Kulbe ULBE and D.T. Andersen NDERSEN (2001) Spectrouorometric characterization of dissolved organic matter for indication of precursor organic material and aromaticity. L L i mnology and O O ceanography 46 (1), 3848. Mc C Knight N IGHT D.M., R. Harnish ARNISH R.L. Wershaw ERSHAW J.S. Baron ARON and S. Schiff CHIFF (1997) Chemical characteristics of particulate, colloidal, and dissolved organic material in Loch Vale watershed, Rocky Mountains, National Park. Biogeochemistry 36, 99. Mobed O BED J., S. Hemmingsen EMMINGSEN J. Autry UTRY and L. Mc C Gown O WN 1996) Fluorescence characterization of IHSS humic substances: total luminescence spectra with absorbance correction, EE n ironment Science & Technology 30, 10. Newsom E WSOM M., A. Baker AKER and S. Mounsey OUNSEY (2001) e potential role of freshwater luminescence measurements in exploring runo pathways in upland catchments, H H ydrological Processes, 15, 989. Perrin E RRIN ,J., PY Jeannin EANNIN and F. Zwahlen WAHLEN (2003) Epikarst storage in a karst aquifer: a conceptual model based on isotopic data, Milandre test site, Switzerand. Journal of HH ydrology 279, 106. PRONK, M., N. GOLDSCHEIDER, and J. ZOPFI (2005) Dynamics and interaction of organic carbon, turbidity and bacteria in a karst aquifer system. H H yd rogeology Journal 14, 473. Senesi E NESI N., T.M. Miano IANO M.R. Provenzano ROVENZANO and G. Brunetti RUNETTI (1991) Characterization, dierentiation, and classication of humic substances by uorescence spectroscopy. Soil Science 152, 259. SCOTT, D.T., MCKNIGHT, D.M., BLUNT-HARRIS, E.L., KOLESAR, S.E. and D.R. LOVLEY (1998) uinine Moieties act as electron acceptors in the reduction of humic substances by humics-reducing microorganisms. EE n ironment, Science & Technology 32, 2984-2989. Smart M ART C.C. and k K .c C Karunaratne ARUNARATNE ) Characterization of uorescence background in dye tracing, EE nironmental GG eology 42, 492. SMART, P.L., B.L. FINLAYSON, W.D. RYLANDS, and C.M. BALL (1976) e relation of uorescence to dissolved organic carbon in surface waters. Water R R esearch 16, 805. Tatar A TAR E., E. Csintalan SINTALAN V.G. Mihucz IHUCZ K. Tompa OMPA L. Poppl OPPL and G. Zaray ARAY (2002) Determination of fulvic acids in water samples of Hungarian caverns, Microchemical Journal 73, 11. VAN BEYNEN, P.E., H.P. SCHWARCZ, D.C. FORD, and G.T. TIMMINS (1991) Organic substances in cave drip waters: studies from Marengo Cave, Indiana. CC anad ian Journal of EE arth Sciences 39, 279. Zech E CH W., N. Senesi ENESI G. Guggenberger UGGENBERGER K. Kaiser AISER J. Lehmann EHMANN T.M. Miano IANO A. Miltner ILTNER and G. Schroth CHROTH ) Factors controlling humication and mineralization of soil organic matter in the tropics. GG eod erma 79, 117.

PAGE 174

15th International Congress of Speleology Earth Sciences 1435 2009 ICS Proceedings ATRAZINE CONTAMINATION AND SUSPENDED SEDIMENT TRANSPORT WITHIN LOGSDON RIV V ER, MAMMOTH CAV V E, KENTUCKY, U.S.A.JULIE E E SCHENCk K BROw W N1, St T Ep P HEN T. KENw W ORt T Hy Y2 1146 Shooting Star Trail, GG urley, AA labama 35748 UU SA A2101 Bridgewater LL ane, OOak RR idge, Tennessee 37830 UU SA A Understanding the potential for karst aquifer contamination by sediment-sorbed pesticides is important for cave conservation eorts in agricultural landscapes. Flow rate, water quality parameters and suspended sediment concentrations were measured in Logsdon River, a ~10 km karst conduit within the Turnhole Spring Groundwater Basin of Mammoth Cave National Park, Kentucky, U.S.A., to determine characteristics of storm-period transport of sediment-sorbed atrazine through a conduit-ow karst aquifer. Analysis of two independent precipitation events occurring in the Spring of 2008 from May 2-4 and May 27 demonstrated the rapid response of the Logsdon River to precipitation events with detections of atrazine increasing during the initial turbidity peak and decline in specic conductance, indicating that the atrazine arrives with the initial ush of surface waters that enters the conduit. Distinct peaks of atrazine did not coincide with ne grained (silt and clay-sized) sediment peaks and concentrations of atrazine remained elevated on the falling limb of the hydrograph as turbidity declined. In addition, no systematic relation between ltered and unltered samples was evident. ere was also exceedingly weak correlation between the concentration of atrazine and suspended sediment, suggesting that if atrazine is sorbed to ne sediment particles, this sorption involves only the fractions ner than 0.22 m. 1. Introductionis research was conducted to evaluate whether stormperiod transport of atrazine through conduit-ow karst aquifers depends on the magnitude and characteristics (particularly grain size) of surface-derived ne sediment inputs. In addition, we evaluated the use of turbidity, a commonly measured water quality parameter, to provide an easily measured proxy of the characteristics of the suspended sediment load and an indication of probable atrazine contamination during the pesticide application season. e Turnhole Spring Groundwater Basin constitutes the largest karst aquifer within Mammoth Cave National Park, Kentucky, U.S.A (Fig. 1). Land use is a mixture of forest (53%), agriculture (43%) and developed use (4%) (Meiman, 2006). It drains 245 km2 that discharge into the Green River, including the Cave City and Pakota Creek subbasins. e Cave City basin is drained by Logsdon River, a vadose stream with a total drainage area of 25 km2 and a base ow rate of 100 liters/second. e Logsdon River parallels the Chester escarpment along the southern edge of the Mammoth Cave Plateau and is traversable for ~10 km before ending in a sump where it intersects the Hawkins River tributary inside the Mammoth Cave National Park boundary. Anderson (2002) collected suspended sediment and water samples from the 145 m drilled well site that intersects the Hawkins River in Mammoth Cave and determined that atrazine was primarily associated with suspended sediments, thus indicating that adsorption to sediments can be a major mechanism for atrazine transport in karst conduit aquifers. Aer a one year study of hydrochemical changes within the Logsdon River, Raeisi et al., (2007) attributed initial minimums in specic conductance during full pipe conditions to the early movement of storm water through the conduit, while the second minimum was interpreted as the return of storm water that was temporarily stored in the aquifer adjacent to the conduit. In contrast, changes in specic conductance during partially-full pipe conditions were mainly controlled by external recharge conditions, such as the behavior of sinking streams (Raeisi et al., 2007).2. MethodsTo document patterns of atrazine transport in relation to suspended sediment ux in Logsdon River, continuous water quality monitoring and water sample collection were conducted for two storm events during the Spring 2008 pesticide application season from May 2-4 and May 2729. e primary source of data collected was from water quality instrumentation installed in an active ow channel of Logsdon River approximately 100 m upstream from

PAGE 175

Earth Sciences 1436 2009 ICS P roceedings 15th International Congress of Speleology Pete Strange Falls. A Sontek Argonaut acoustic Doppler velocity proler (aDcp) mounted on the river bed measured ow depth and velocity for estimation of ow rates. Water temperature, pH, specic conductance (spC) and turbidity were measured at 10 minute intervals with a Hydrolab MS5 multiparameter water quality sonde. Supplementary information on sediment concentrations and particle size was obtained from a Sequoia Scientic Inc. LISST 25-X laser diraction sediment sensor. All instruments operated on battery power and were mounted to the cave passage to prevent movement during high ow conditions. Data collected from rain gages located near the Cave City subbasin was used to estimate the timing and magnitude of precipitation inputs to the groundwater basin. Water quality data collected since the deployment of instrumentation in August 2005 were analyzed to establish baseline values and typical ood pulse responses for temperature, pH, specic conductance and turbidity. Water samples for analysis of atrazine and suspended sediment concentration were collected by a Teledyne ISCO sampler that was programmed to collect samples at 40 minute intervals during ow events. Sample collection was triggered when turbidity readings exceeded a threshold value of100 NTU to capture data from larger runo producing precipitation events. Additional samples for atrazine and total suspended solids were collected from a surface tributary sink point by a Teledyne ISCO sampler and from a pump below the 145 m well sha that intersects the Logsdon River near the conuence of the Hawkins River. Analysis of atrazine levels within a 50 ml subsample was performed by the enzyme linked immunosorbent assay (ELISA) method. Water samples obtained from the Pete Strange Falls instrument site were ltered in the lab with a 0.22 m syringe lter. 3. Results and DiscussionHydrologic data for the two sampled events showed a rapid response typical of conduit ow karst aquifers (Figs. 2, 3). While the peak stage and ow rate were higher for the May 2-4 precipitation event, the hydrologic response was much quicker for the May 27 precipitation event due to the higher intensity of precipitation near the Cave City subbasin. Sudden drops in spC indicate the ushing of storm water through the conduit, followed by peaks and dips in spC that reect subsequent pulses of storm water Figure 1: Map of study area (Based on NLCD 2001 data).

PAGE 176

15th International Congress of Speleology Earth Sciences 1437 2009 ICS Proceedings passing the instrument station. An overall correlation of spC reductions with increases in turbidity early on the ow hydrograph (before and immediately following the peak) suggests the arrival of one or more discrete inputs of relatively dilute, sediment laden water from the surface, followed by a secondary pulse of low spC, turbid water later on the falling limb (Figs. 4, 5). is pattern of spC and turbidity responses suggests the initial input of water arrives from land within or near the park boundary, while the secondary pulse arrives from areas further away from the observation point that has a lower proportion of forested land. An alternative explanation of the secondary water quality perturbation that occurs later on the falling limb is that it reects the return of storm water temporarily stored in the aquifer adjacent to the conduit (Raeisi et al., 2007). Atrazine values ranging from 12.4 ppb to >50 ppb at a surface tributary of the Logsdon River reect the mobilization of atrazine from agricultural areas during the early May event (Fig. 6). Detections of atrazine during the initial turbidity increase at both in-cave monitoring sites (Logsdon River Wells and Pete Strange Falls) indicate the atrazine arrived before the secondary pulse of turbid water passed the sampling point (Figs. 7, 8). e increase of atrazine concentrations to near steady values between 2 and 3 ppb was correlated with the secondary pulse of ne sediment laden water for both precipitation events. However, distinct peaks in atrazine levels did not occur during periods of peak turbidity and continued to stay elevated on the falling limb of the hydrographs as turbidity declined. ese observations are similar to results from Anderson (2002) who reported peak levels of atrazine aer the turbidity peak in Hawkins River. e timing of atrazine transport relative to ow and other water quality variations in Logsdon River may reect relatively low atrazine availability in surface recharge areas closer to the monitoring point. In addition, no systematic relation between atrazine concentration in ltered and unltered samples was evident, suggesting that if atrazine is sorbed to ne sediment particles this sorption involves only the fractions ner than 0.22 m. Nor was atrazine positively correlated with measured concentrations of suspended sediment or the grain size Figure 2: Precipitation, stage response and ow rate in Logsdon River and surface tributary: May 2, 2008. Figure 3: Precipitation, stage response and ow rate in Logsdon River: May 27, 2008. Figure 4: Stage, specic conductance and turbidity in Logsdon River: May 2, 2008. Figure 5: Stage, specic conductance and turbidity in Logsdon River: May 27, 2008.

PAGE 177

Earth Sciences 1438 2009 ICS P roceedings 15th International Congress of Speleology distributions of samples for the May 27 precipitation event (Fig. 8). erefore, increased atrazine levels do not correlate simply with ne sediment concentration, but may be related to the geochemical composition of materials in suspension or to other factors aecting sorption such as preferential association with particular mineral fractions or an increase in ne organic material, both of which were not evaluated in the present study. Additional work is required to establish whether such geochemical factors are important controls on pesticide transport in the Cave City basin or other conduit-ow karst aquifers in agricultural landscapes.4. ConclusionsObserved patterns of suspended sediment and atrazine concentrations were related to the transport of surface runoff through the karst aquifer and to patterns of agricultural land use within the Cave City subbasin. This study determined that storm-period transport of atrazine through a conduit-flow aquifer was associated with an initial peak of surface derived fine sediment inputs, but a commonly measured water quality parameter (turbidity) was not correlated to the concentration of atrazine and could not provide an indication of possible atrazine contamination during the pesticide application season. Water quality monitoring within Mammoth Cave National Park clearly cites non-point source runo from agricultural practices as the major cause of contamination in the Turnhole Spring groundwater basin (Meiman, 2006). Atrazine levels at a surface tributary of Logsdon River greatly exceeded the U.S. Environmental Protection Agencys (USEPA) maximum contaminant level (MCL) of 3.0 ppb which was detected in 81 percent of these samples. Only one of the in-cave samples from either event exceeded the MCL, but seventeen of the thirty-three samples from the early May event and twenty of the twenty-four samples from the late May event did exceed the USEPA aquatic life criterion of 1.8 g / L. ese ndings support the need for on-going monitoring and mitigation of contamination within karst aquifers to protect cave fauna, particularly within areas that receive surface precipitation inputs from agricultural lands.Acknowledgementse authors want to thank Ted Baker, Leigh Anne Bledsoe, Andrea Croskey, Dr. C. Warren Campbell, Jim Currens, Graydon Dill, Brian Ham, Melissa Hendrickson, Johanna Kovarik, Jodi Lindsey, Dan Nol, Randy Paylor, Narcisa Pricope, Dr. Chris Smart, and Mark Tracy for their assistance with data collection at the Pete Strange Falls Instrument Site in Logsdon River over the course of several years. Jeremy Goldsmith, Chrissie Hollon, Erin Lynch and James Otoo provided assistance with data collection from the eld for the early May 2008 sampling event. Dr. Chris Groves and Joe Meiman provided insightful commentary about Logsdon River and the Turnhole Groundwater Basin. Figure 6: Atrazine levels detected at surface tributary of Logsdon River: May 3, 2008. Figure 7: Atrazine levels, specic conductance and turbidity in Logsdon River: May 3, 2008. Figure 8: Atrazine levels, specic conductance, turbidity and suspended sediment concentration in Logsdon River: May 27, 2008.

PAGE 178

15th International Congress of Speleology Earth Sciences 1439 2009 ICS Proceedings ReferencesAnderson, M. S. (2002), Transport of the Herbicide Atrazine on Suspended Sediments during a Spring Storm Event in Mammoth Cave, Kentucky. M.S. esis, Western Kentucky University, Bowling Green, KY. 62 pp. Meiman, J. (2006), Mammoth Cave National Park: Water Resources Management Plan. National Park Service, U.S. Department of the Interior. 256 pp. Raeisi, E., C. Groves, and J. Meiman, (2007), Eects of partial and full pipe ow on hydrochemographs of Logsdon River, Mammoth Cave Kentucky, USA. Journal of Hydrology, 337: 1, 1.

PAGE 179

Earth Sciences 1440 2009 ICS P roceedings 15th International Congress of Speleology TIANKENGS IN THE KARST OF CHINACHENCHEN WEIHAI EIHAI ZHU ZHU XUE UE WEN ENI I nstitute of Karst GG eology, CC hinese AA cademy of GG eological Sciences, GG uilin, CC hina, chinacave@163.com China has the most extensive and diversied karst terrains in the world and most of them are rich in caves and dolines. e cone karst (fengcong) and tower karst (fenglin) developed in the humid climate in southern China form the most distinctive karst landscapes. Tiankengs are giant dolines that are a feature in some areas of the cone karst. In recent years, more than y tiankengs have been discovered in the cone karst in southern China, notably in the provinces of Chongqing, Guangxi, Sichuan, and Guizhou. Current research indicates that tiankengs develop in specic environments of geomorphology, geology, and hydrogeology, and are, therefore, distinguished from normal karst dolines.1. IntroductionIn carbonate rock terrains, one kind of negative karst landform has not previously been recorded because it is relatively rare and occurs only in more remote regions. It is the great or giant doline, with steep walls and several hundred meters in depth and diameter; it is a collapse doline distinguished by its very large size. is landform was rst observed by geologists in China in the early 1980s, in the Xingwen karst in Sichuan, in an area that was being developed for tourism. In the past 20 years, many more of these great karst dolines have been discovered in southern China. Some, including Xiaozhai, Dashiwei, Qingkou, and Haolong, have been found by vigorous tourism development that has been searching for spectacular features in the more remote karst areas. Others have been found during explorations by the China Caves Project, and this has led to considerable research to identify the special features of the giant dolines that distinguish them from more common, smaller, dolines. is kind of giant doline is distinguished from normal dolines by its size, and also by major dierences in basic characteristics, geomorphic evolution, and hydrogeologic conditions. e understanding of its geomorphology and its importance matured through the 1990s. In October 2001, it was proposed that this giant doline could be distinguished from normal dolines, and a new term, tiankeng, was proposed for the karst literature (Zhu, 2001). ere are two types of tiankeng collapse tiankengs and erosion tiankengs. e former developed by dynamic underground water ow, while the latter were formed by allogenic surface drainage that fed into an underground river in the karst. Collapse tiankengs are much more widespread and more numerous than the erosional forms.2. Distribution of Tiankengs in ChinaCurrent records show that the tiankengs in the karst of China are mainly in the south of the country, especially in the cone karst terrain, although some lie outside this core zone. e distinctive cone karst of southern China covers an area of about 150,000 km2, mainly in northern and western Guangxi, southern Guizhou, around the Yangtze Gorges, and across southeastern Chongqing, northern Guizhou, southeastern Sichuan, western Hunan, and western Hubei, as well as in southeastern Yunnan. e important tiankengs discovered so far in China are those described below, and these are located on Figure 1.3. e Major Tiankengs in ChinaTiankengs are best described as collapse dolines that are more than about 100 m wide and deep, and this is recommended as the internationally accepted denition of a tiankeng. ere are, however, many more features that are between 50 m and 100 m deep and wide, and these are already widely known as tiankengs within China; they have been referred to as small tiankengs (Zhu, 2001). ese include three small tiankengs in the Mengzi basin, Figure 1: Locations of tiankengs in southern China.

PAGE 180

15th International Congress of Speleology Earth Sciences 1441 2009 ICS Proceedings within the karst of Yunnan. ough many of these smaller features are very signicant karst landforms, they are all omitted from Table 1 and the descriptions that follow. Except for the Xiaoyanwan and Dayanwan tiankengs in the Xingwen stone forest tourism area in Sichuan Province, which have been known for many years, the important discoveries of tiankengs have occurred since 1994. In that year, the largest tiankeng, Xiaozhai, was discovered near the Yangtze Gorges during the search for a new exploration site for British cavers in the China Caves Project. In 2001, a group of 26 tiankengs was discovered in the Leye karst in Guangxi during investigations for karst tourism resources and the search for another venue for cavers of the China Caves Project. e discoveries of this special karst feature generated interest in scientic research, which was pursued in subsequent years. Almost at the same time, Qingkou Tiankeng was discovered by the senior author in the Wulong karst, and was later explored by the Hongmeigui Cave Club. Around its vertical walls there are several hanging waterfalls and these converge on the oor and ow into a large cave passage. Qingkou Tiankeng was the rst erosional tainkeng to be recognized. e 49 known tiankengs are listed in Table 1. Tiankeng Length x Width m Area m2Depth max m Depth min mV V o lume Mm3Xinlong karst, Fengjie, Chongqing Xiaozhai 625 x 535 274,000662 511119.3 Xiaokeng 330 x 180 45,000286137 12.0 Chongtianyan 300 x 160 41,500168 1037.0 Luokuangyan 135 x 100 10,800101 1001.1 Wujiazhai 200 x 150 24,000103 472.5 Daqinkeng 200 x 1508,000137 661.1 Qingshui, Yunyang, Chongqing Longgang 350 x 170 53,000350 2509.2 Houping, Wulong, Chongqing Qingkou 250 x 220 40,700295195 9.2 Niubizi 380 x 100 27,000195 1003.5 Taipingmiao 180 x 180 26,400420 3009.9 Daluodang 240 x 220 32,400372 28210.4 Shiwangdong170 x 15025,900252 1725.1 Sanqiao, Wulong, Chongqing Zhongshiyuan 565 x 555 278,200214 7534.8 Xiashiyuan 990 x 545 352,100373 5031.5 Qinlong 520 x 200194,000276 19531.7 Shenying 300 x 260 51,200285 1909.7 Xingwen karst, Sichuan Xiaoyanwan 625 x 475 200,000248 17836.0 Dayanwan 680 x 280164,000110 4015.0 Dashiwei group, Leye, Guangxi Baidong 220 x 160 22,000312 2635.8 Chadong 400 x 350 80,50025 16513.3 Chuandong 370 x 27073,000312 17511.7 Dacao 300 x 140 30,00092 561.3 Dalong 240 x 200 35,000125 953.3 Dashiwei 600 x 420 167,000613 51175.0D D atuo 530 x 380 149,000290 26332.7D D engjiatuo 370 x 240 128,200278 22226.2 Diaojing 290 x 280 86,300170145 12.6 Gaicao 440 x 95 24,700120 902.2 Huangjing 320 x 170 51,700161 1406.3 Jiameng 90 x 80 8,800271 2111.6 Ladong 202 x 12521,600215 1462.8 Laowuji 300 x 275 75,600171 1108.3 Longtao 210 x 175 14,400115 951.4L L uoj ia 140 x 100 10,200128 710.7 Lanjiawan 150 x 115 10,700130 670.6 Shenmu 370 x 340 70,900234 18613.2

PAGE 181

Earth Sciences 1442 2009 ICS P roceedings 15th International Congress of Speleology 3.1 Tiankengs of XXingwen, Sichuanese two tiankengs are situated in the stone forest tourism area of Xingwen, in Sichuan Province. In September 1992, a British cave expedition team explored the underground river and cave systems to reveal the hydrogeological characteristics of the karst (Waltham and Willis, 1993; Waltham et al, 1993; Zhu et al, 1995). More than 30 km of passages were surveyed in 89 caves; the two longest caves Tianquan Dong (8100 m) and Zhucaojing (8800 m) both have passages opening directly into the sides of the Xiaoyanwan tiankeng. Xiaoyanwan Tiankeng is roughly circular in plan, 625 m from east to west, and 475 m across; its vertical walls are 60130 m round the entire perimeter. e maximum elevation on the rim is 870 m, and the lowest point in the tiankeng oor is at 622 m, giving a maximum depth of 248 m, and the total volume is 40M m3. Dayanwan lies 400 m to the west of Xiaoyanwan, and is 680 m long east to west, 280 m across, 110 m deep, and 15M m3 in volume. e Dayanwan and Xiaoyanwan tiankengs appear to be older than many of the other tiankengs in China, but cannot yet be dated; clearly Dayanwan formed earlier than Xiaoyanwan since it is so degraded.3.2 XXiaozhai Tiankeng, Fengjie, ChongqingLocated near Xinlong town in Fengjie county, the Xiaozhai tiankeng is in the karst on the right bank of the Jiupan River, a tributary to the Yangtze River. Developed in gently dipping Lower Triassic limestone, the area is a typical cone karst at elevations of 1300-2000 m. Xiaozhai Tiankeng may rank as the largest tiankeng in the world, with an entrance diameter of 537 to 626 m, a depth of 662 m, and a volume of 119.35M m3 (Fig. 2). In prole, it has a double nested structure; the upper bowl is 320 m in depth, and the lower sha is a rectangle 342 m in depth and 257-268 m across; the sloping ledge between these two parts is formed at the level of a muddy limestone. Across the oor of the tiankeng, a cave river has a maximum discharge of 174 m3/sec. Within the karst around Xiaozhai tiankeng there are six other tiankengs of medium-size and 101 m deep; these are Chongtianyan, Luokuangyan, Wujiazhai, Houzishi, Xiaokeng, and Daqinkeng (Table 1). ey all appear to have developed at an earlier stage than Xiaozhai Tiankeng, and all of them are of the collapse type. 3.3 Dashiwei tiankeng group, Leye, Guangxie large group of tiankengs that includes Dashiwei lies in western Leye county, close to Tongle town (Zhu et al, 2003b). ere are 26 tiankengs discovered so far. e largest is Dashiwei Tiankeng, which is pear-shaped in plan, 600 m long from east to west, 420 m wide from north to south, 1580 m around the perimeter, and 613 m at its maximum depth. It is surrounded by vertical clis (Fig. 3) and its oor is covered by a ramp of collapse debris more than 100 Shizilu 130 x 70 7,00012090 0.6 Shuijia 245 x 135 23,700167 1112.6 Xiangdang 310 x 230 45,000146 8012.6 Bama karst, GuangxiH H ao long 800 x 600 320,000509185 110.0 Jiaole 750 x 400 220,000325 28367.0 Dongdang, Shuitang and Yijiehe groups, Guizhou Detian 200 x 130 20,000145 1152.5 Dachang 550 x 180 80,000320160 10.0 Bajiao 280 x 160 24,300195 1504.0 Tongtian 210 x 130 20,000370 3607.2 Xiaoshui 180 x 130 13,000230 2102.8 Dacaokou 920 x 240 140,000220 16025.0 Xiaocaokou 300 x 120 22,000180 1203.3 Bandong 190 x 100 15,000240 2252.0Table 1: List of tianken. Figure 2: Aerial view of Xiaozhai Tiankeng, Fengjie.

PAGE 182

15th International Congress of Speleology Earth Sciences 1443 2009 ICS Proceedings m high that slopes steeply down from east to west and is covered by ourishing secondary forest. e cave river is accessible at the lowest point of the tiankeng oor (Fig. 3), under its western wall; it emerges from a pile of collapsed rocks, and 6000 m of passage has been mapped downstream, as far as a point where the river drops into deep, narrow, and inaccessible ssures. All the tiankengs in the Dashiwei group are of the collapse type, though Huangjing Tiankeng has been modied by allogenic water since its collapse development, and is intermediate to an erosional type of tiankeng (see below). Of the 26 tiankengs in the Leye karst, Dashiwei is classied as a very large tiankeng. Chuandong, Datuo, and Dengjiatuo are all large, though only Chuandong Tiankeng has vertical walls on its entire perimeter, while the other two have degraded to leave debris slopes round about half their perimeters. ese and 17 other tiankengs of normal size are listed in Table 1.3.4 Tiankengs of Wulong, ChongqingAround Wulong, the terrain along the banks of the Wu Jiang, a tributary of the Yangtze River, is a typical fengcong cone karst that is deeply dissected by a river valley. e major local relief creates a vadose zone up to 1000 m deep in the succession of carbonate rocks of Cambrian, Permian, and Triassic age. Several collapse tiankengs have been recognized in the karst of Wulong county, in the natural bridges tourism area, and close by it. e most important are Zhongshiyuan, Xiashiyuan, Qinlong, and Shenying Tiankengs. Xiashiyuan is the largest of the tiankengs and is 1 km long. Both the Qinlong and Shenying Tiankengs (Fig.4) developed in the dry valley that originated as a major cave passage over 2 km long and is now spanned by three natural bridges. Qingkou Tiankeng is the nest example of an erosional tiankeng yet discovered in China. It has been formed by a concentrated ow of allogenic surface water draining into karst with a deep vadose zone. e tiankengs of Qingkou, Niubizi, Daluodang, and Taipingmiao were developed at the sandstone-limestone boundary within a broad valley. Of these, Qingkou Tiankeng is the largest and most mature; it is 250 m wide and 295 m deep (Table 1). is type of tiankeng diers from the collapse form because it has been formed by erosion from the surface into the limestone vadose zone by an allogenic surface stream. It is the less common type of tiankeng. Caves have developed at dierent levels in the Qingkou tiankeng karst, and a cave system about 10 km long discharges southward to the Mawandong rising of the Muzong River.4. Nomenclature and Denition of Tiankengs 4.1 e denition of a tiankengBased on the research and conclusions outlined above, it is proposed that the denition of a tiankeng is a large, steep-walled, pit-like, negative, karst landform that opened from beneath towards the surface, with both its depth and diameter more than 100 m, developed in a great thickness of Figure 3: Aerial view of Dashiwei tiankeng, Leye. Figure 4: Shenying Tiankeng, between the natural bridges in Wulong.

PAGE 183

Earth Sciences 1444 2009 ICS P roceedings 15th International Congress of Speleology continuous soluble rocks within a deep vadose zone of the aquifer and connecting with an active cave river at its foot. A tiankeng is characterized by its tourism values of rarity, grandeur, and spectacular magnicence, and also by its special ecological environment. ese are all characteristics that dierentiate between tiankengs and normal karst dolines. ere are three size groups Very large tiankengs, more than 500-600 m in diameter and depth, are very rare worldwide, and there are only three known so far in China Xiaozhai, Dashiwei and Haolong. Large tiankengs, 300-500 m in diameter and depth, are also few around the world. Normal tiankengs, 100-300 m in diameter and depth, are more numerous and more widely distributed.4.2 Two types of tiankengsCurrent research suggests that tiankengs in carbonate rocks may be divided into two types collapse tiankengs and erosional tiankengs. e former are much more widespread and numerous than the latter (Zhu et al, 2003a, 2003b).Collapse tiankengs have been formed in soluble rocks where massive amounts of rock material have been dissolved and eroded away at depth by a powerful and dynamic underground drainage system, notably through a large cave river passage. Under specic geologic and hydrogeologic conditions, a cave chamber evolved as its roof failed gradually while the fallen rock debris was carried away by water; eventually, the chamber roof opened out to the ground surface. Erosional tiankengs is developed in the vadose zone of soluble rocks by allogenic water that dissolves and erodes the rock in its vertical descent from the surface. ey are rare because of their special environmental conditions. 4.3 Development environments of tiankengsCollapse and erosional tiankengs have some features in common, but also have some important dierences in their development conditions. ere are ve essential environmental conditions. 1) A great and continuous thickness of soluble rocks. 2) A deep vadose zone in the karst. 3) A favorable geologic structure. 4) A highly active hydrodynamic system with underground conduit ow in cave rivers. 5) Favorable climate and hydrogeology. ese environmental factors are all optimally developed in the karst of southern China, especially in the cone karst areas with high relief which include the most important tiankeng sites in the world.4.4 e geologic age of tiankengsTiankengs provide data for research of karstication intensities and rates, and oer a new way to study time-scales of karst geologic processes. It is suggested that the tiankengs in China are among the youngest negative landforms of the karst. Ongoing research is relating the high intensities and rates of karstication to the relatively young geologic ages of the tiankengs. Tectonic upli of the karst region of southern China dates only from the Himalayan orogeny in the early uaternary. is caused the deep incision by surface trunk rivers (including the Yangtze, the Wu, and the Hongshsui, each with their deep gorge sections), and the consequent decline of the karst water table (and increase in vadose zone thickness) caused development of the underground rivers and development of the tiankengs. is suggests that the tiankengs in China formed mainly in the late Pleistocene, within the last 128,000 years. 5. Conclusionse doline is one of the most familiar and distinctive landforms in karst terrains. ere are several doline types that form in soluble rock terrains (mainly of carbonate rock and gypsum), including the solution, collapse, and subsidence dolines. Collapse dolines are karst bedrock collapses (and caprock dolines involve collapse of insoluble rock that covers a buried karst), but the sizes of all dolines do not match the size of a tiankeng. Dolines in carbonate rocks generally have diameters and depths up to about 100 m. Very large, vertical-walled, karst depressions dier from all the main doline types, though some have been called just large dolines or large collapse dolines. ese are distinguished by not only their size but also by their development mechanisms and conditions, so it is proposed that they should be separated from dolines and referred to as tiankengs.5.1 Distinctive features of a tiankengA very large doline is now called a tiankeng for certain basic reasons: A tiankeng develops in special environmental conditions that integrate aspects of geology, geomorphology, and hydrogeology, but a normal doline develops in a much wider range of karst environments. Consequently, dolines and collapse dolines are widespread while tiankengs are very few within the world's karst. A tiankeng diers greatly from dolines and collapse dolines in its development and erosional mechanisms. A collapse doline forms by dissolution and suosion in normal geologic and

PAGE 184

15th International Congress of Speleology Earth Sciences 1445 2009 ICS Proceedings hydrodynamic environments. Collapse tiankengs have developed through an unusual hydrodynamic combination of erosion, dissolution, and collapse, where three evolutionary stages may be distinguished, from a cave river passage, to a large cave chamber, to a tiankeng open to the surface. A tiankeng is very much larger than a normal collapse doline. It is more than 100,000 m3 in volume, 1000-100,000 m2 in area at both the surface level and its oor, and more than 100 m in depth. A tiankeng diers from normal dolines in its development processes. Tiankengs have close relationships with the regional development of cave river systems, and their distribution, evolution, age, and development rates are not comparable to those of normal karst dolines.5.2 e importance of the scientic study of tiankengse study of tiankengs has important scientic implications for karst hydrogeology and geomorphology, and for studies of karst processes and neotectonics. Where tiankengs develop within a regional karst drainage system, they indicate the presence of a powerful karst hydrodynamic system, and they relate to the input and output balance of material and energy in a strong conduit ow or cave river in the karst aquifer. is can supply valuable data towards research into the basic characters of an aquifer and into the evolution and variance of conduit ow in drainage areas where tiankengs occur. e evidence to date suggests that collapse tiankengs develop anywhere they can, and always destroy any surface karst landform, including depressions, dolines, wind-gaps, blind valleys, cones, hills, towers, and any other positive landforms. It appears that tiankengs are among the younger karst landforms, and a negative feature of 10 to 100 million cubic meters volume developed in a short geologic time gives us a new concept of intensity, mode, and rate of karstication. It appears that the tiankengs of southern China have largely or entirely developed within the later Pleistocene, and the oldest tiankeng appears to be no more than 128,000 years old. Finally, local neotectonic upli appears to have contributed to the critical environment for tiankeng development. Conversely, tiankeng research can supply important information for the study of the rates and characteristics of neotectonic movement.AcknowledgementsFirst of all, to our colleagues, Prof. Zhu Dehao, senior engineer Li Yusheng and Tan Kaiou, as well as Zhang Yuanhai, Huang Baojian, Zhang Ren, Han Daoshan, Cai Wutian, Zhu Jinbo, Deng Yadong for their contribution to the tiankeng investigation and research, tiankeng data analysis and map compilation. Also to Prof. Yang Mingde of Guizhou Normal University for supplying tiankeng data in Guizhou, Andy Eavis for his great contribution to our tiankeng exploration as the international coordinator of the China Caves Project, Tony Waltham for his research cooperation on tiankengs inside and outside China, and the many cavers from Britain, Ireland, USA, Australia, Italy, France, and Switzerland who have enriched the paper by supplying their cave surveys and data. is text was translated from Chinese by Zhang Yuanhai and Zhu Xuewen, and the grammar was edited by Tony Waltham.ReferencesWaltham, A C and Willis, R G, 1993. Xingwen: CC hina C C aves Project 1989-1992. British Cave Research Association: Bridgewater, 48pp. Waltham, T, Brook, D and Bottrell, S, 1993. e caves and karst of Xingwen, China. CC a ve Science, 20(3), 75-86. Zhu Xuewen. 2001. Chinas karst tiankeng and its value for science and tourism. S cience & Technology RReview, 10(160), 60-63. Zhu Xuewen, Zhu Dehao and Chen Weihai, 2003a. A brief study on karst tiankeng. CC ar sologica Sinica, 22(1), 51-65. Zhu Xuewen, Huang Baojian, Zhu Dehao and Chen Weihai, 2003b. DD ashiwei Tiankeng GG roup, LLeye, G G uangxi: discoeries, exploration, denition and research. Guangxi Scientic & Technical Publishers (and China International Publishes: Beijing), 184pp. Zhu Xuewen, Zhang Ren, Zhang Yuanhai and Han Daoshan, 1995. Karst and caves in Xingwen stone forest area, Sichuan. CC ar sologica Sinica, 14(suppl.), 28-48.

PAGE 185

Earth Sciences 1446 2009 ICS P roceedings 15th International Congress of Speleology Delineating ELINEATING Spring PRING Flow LOW Systems YSTEMS in IN the T HE Te E Xas AS Hill ILL Country OUNTRY USAAA LI H H C C HOw W DHURy Y1, C C HAD N N ORRIs S2 1Texas Water DDevelopment Board, 1700 NN orth CC ongress AA venue, AA ustin, Texas 78711, UU SA A2Texas Parks and Wildlife DDepartment, 4200 Smith School RRoad, AA ustin, Texas78744, UU SA A To improve our understanding of the spring ow systems in the Edwards and Trinity aquifers of the Texas Hill Country, we have examined more than 250 spring waters for their chemical compositions, estimated ows, and water temperature. Springs in the area arise along zones of weaknesses in the aquifer materials, such as, bedding planes, faults, karst features, and river bottoms due to gravity drainage or artesian pressure. Continued ow of recharge water under-saturated with respect to carbonate along these zones causes preferential dissolution of host carbonates creating cavities and facilitating rapid transport of groundwater discharging through springs. ese springs are important to the area as they provide clean drinking water, maintain streamow through baseow discharges, nourish ecological habitats, and supply water for the survival of the rare, endemic, and common species. A good understanding of the origin of the spring waters can provide valuable information regarding their owpaths and residence times and for evaluating vulnerability of these springs and the biota they support to drought and pumping. We observed that most of the springs are mainly composed of calcium-magnesium-bicarbonate waters characteristic of shallow groundwater recharge. Some of the spring waters have sodium and chloride ions that are well correlated, and a few of the spring waters have chloride/bromide ratios in excess of 400 suggesting minor halite dissolution, possibly from halite contained in evaporites of the Upper Glen Rose Limestone or upward migration of deeper saline water. Sodium and chloride enriched spring waters preferentially occur to the north (south of the Llano Upli) and northeast along the Balcones Fault Zone. About 80 percent of the springs have estimated ows that range from 270 to 2,700 cubic liters per day. Most of the spring water temperatures are less than 25 degrees centigrade similar to the adjoining groundwater in the aquifers. We have also analyzed about 24 of these springs for detailed stable and radiogenic isotopes and chemical compositions. We observed that isotopic and chemical compositions of the spring waters vary with dierences in elevations. Spring waters at higher elevations commonly have lower total dissolved solids and lighter carbon, oxygen, and deuterium isotopes suggesting shorter ow paths and residence times between recharge and discharge. At lower elevations, heavier carbon, oxygen, and deuterium isotopes in the spring waters suggest relatively longer ow paths and residence times. Strontium isotopes suggest that most of the spring waters are derived from various degrees of mixing of the shallow groundwater and rainfall. Nearly all of the springs have tritium as high as three tritium units and percent modern carbon content ranging from 70 to 110 percent, supporting a recent origin of these waters. Our investigation suggests that most of the spring waters are modern in age and result from relatively rapid ow through the aquifer. e modern age of the waters suggest that the spring ows are highly vulnerable to short term climatic variations and pumping, such as the response of Jacobs Well Springs during the drought of 2000 and 2008.1. IntroductionSprings are natural discharge points in a groundwater ow system. Groundwater from unconned aquifers may naturally drain under gravity to spring orices through bedding planes, faults, karst features, or river beds. Under artesian pressure, deeper groundwater may move vertically upwards to the land surface and discharge through springs. Duration and intensity of spring discharge may serve as clues to our understanding of shallow groundwater recharge, their role in maintaining streamow, and hydraulic characteristics of the aquifer. In the Texas Hill Country, USA, springs are of signicant importance as they are used for potable water

PAGE 186

15th International Congress of Speleology Earth Sciences 1447 2009 ICS Proceedings supply sources, in shaping surface drainage and maintenance of streamow, and sustaining ecological habitats for various rare, endemic, and common biological species. Many of the larger springs in the study area, including Comal, San Marcos, and Hueco Springs have been extensively studied for their source waters, ow paths, and recharge characteristics (Guyton and Associates, 1979; Ogden and others, 1986; McKinney and Sharp, 1994; Brune, 2002; and Johnson and Schindel, 2008). erefore, while we examined geochemical data for about 250 spring waters from the Hill Country area including the larger springs, we concentrated our investigation on the 24 smaller springs that were not studied earlier. In this study our objectives were: (1) determine the origin and ages of the spring waters, (2) better understand ow characteristics to these springs, and (3) determine vulnerability of these springs to drought or pumping from the shallow aquifers.2. MethodsWe examined about 250 spring water compositions from the Texas Water Development Boards groundwater database for this investigation. In addition, we selected 24 springs based on the spring ow amount and ease of accessibility for collection of chemical and isotopic compositions. We collected 24 spring water samples for deuterium (2H) and oxygen-18 (18O), carbon-13 ( 13C), carbon-14 (14CDIC), tritium (3H), and strontium (87Sr/86Sr) isotopes. All groundwater samples were analyzed by ion chromatographymass spectrometry (ICP-MS) for chemical parameters at the Energy Laboratories in Wyoming. Isotopic analyses of 18O and 2H were carried out at the Coastal Sciences laboratory in Austin, Texas. Isotopes of 18O were analyzed on a VG Micromass SIRA Series II mass spectrometer using the carbon dioxide (CO2) equilibration method (Epstein and Mayeda, 1953). Isotopes of 14CDIC were analyzed at the Beta Analytic Inc. using an accelerator-mass-spectrometer (AMS) and the results were presented as percent modern carbon (pmC). Isotopes of 13C were measured with reference to the PDB standard. Isotopes of 3H were analyzed at the Tritium Laboratory of the Rosenstiel School of Marine and Atmospheric Sciences, University of Miami. Isotopic analysis of 3H was performed by chromium or zinc reduction to hydrogen (H2) gas on a Micromass 602D mass spectrometer. 87Sr/86Sr was analyzed by ermal Ionization Mass Spectrometer (TIMS) at the isotope laboratory of Massachusetts Institute of Technology in Boston.3. Hydrogeology of the SpringsSprings are fed by gravity or they may appear under pressure through natural openings (dissolved limestone along impermeable bedding contacts, caverns, and faults) in a conning unit. e Edwards and associated limestones form the most prolic springs in Texas (Brune, 2002). Certain sections of the vugular limestones that form the aquifers are interconnected to facilitate rapid inltration and replenishment of recharge. e Edwards Group of rocks are characterized by a steep drop in land surface elevation allowing groundwater to drain under gravity along its slope through permeable parts of the aquifers and discharges along impermeable bedding contacts of the Upper Glen Rose Limestone. Springs also preferentially occur along many of the river beds as they occupy the lowest elevations in the land surface (Fig. 1). In addition, locations of many of the headsprings may move by tens of meters with changes in the duration and intensity of rainfall. Many of the springs are perennial sources of water to the creeks and rivers. For example, a well developed trend is observed between spring discharges at the Fessenden Springs and ow in the adjacent Johnson Creek located in the Upper Guadalupe watershed in Kerr County. When recharge is not available, such as during the droughts that aect parts of central Texas, water levels in the springs may decline, with some springs eventually drying up temporarily. For example, ow in the Jacobs Well Springs trickled down to nothing during the drought of 2000 and 2008, probably due to an absence of recharge that was further exacerbated by increased pumping in the shallow aquifer. 4. ResultsChemical and isotopic characteristics of the spring waters can help delineate the origin of the spring waters and their potential sources. We report chemical and isotopic characteristics of the spring waters in the following sections.4.1 Chemical compositione spring waters are dominantly composed of calciummagnesium-bicarbonate waters except for about six spring waters that have higher sulfate and chloride. Springs in the north and northeast of the study area have higher concentrations of sodium and chloride. A plot of sodium versus chloride to determine the source(s) of sodium and chloride indicates that they are well correlated (r2 = 0.74) with most of the samples plotting on or close to the 1:1 line (Fig. 2). However, one sample plots above the 1:1 line showing excess sodium compared to chloride concentrations. Both sodium and chloride concentrations of these spring waters are much higher than the rain water. Most of the spring waters have low chloride/ bromide (<400 molar) with only a few springs having values greater than 400. ese values of chloride/bromide

PAGE 187

Earth Sciences 1448 2009 ICS P roceedings 15th International Congress of Speleology ratios are much lower than groundwater from the adjacent aquifers where many samples have chloride/bromide ratios in excess of 10,000 (Chowdhury, 2008). A plot of land surface elevation versus total dissolved solids of spring waters indicate that salinity progressively increases in the springs from the higher to the lower elevations (Fig. 3)4.2. Isotopic compositionSpring waters have 18O values that range from -3 to -6.1 SMOW similar to groundwater (Fig. 4). Springs in Bandera and Real counties have much lighter 18O compositions than the springs further to the east. A plot of 18O versus 2H values show that most of the spring waters fall along the Local Meteoric Water Line that represents average composition of local precipitation (Fig. 4). 13C isotopes of the spring waters range from -3 to -12.8 PDB with the lighter isotopes occurring at higher elevations in the west. Heavier 13C isotopes in the spring waters are more common at lower elevations (Fig. 5). Most of the spring waters have 14C isotope values containing more than 80 percent modern carbon with the exception Figure 1: Map showing land surface elevation in the study area and spring locations.

PAGE 188

15th International Congress of Speleology Earth Sciences 1449 2009 ICS Proceedings of one spring water which has as low as 40 percent modern carbon. Nearly all of the spring waters that we studied have tritium concentrations that range from 2 to 3 tritium units. Only three spring waters have concentrations less than 1.5 tritium units. We note that most of the spring waters have 87Sr/86Sr values of 0.7076 to 0.7079 with a few samples having values of 0.7080 to 0.7086, and one spring located in the north near the Llano Upli has an elevated 87Sr/86Sr value of 0.7098.5. DiscussionMost of the spring waters have a dominant calciummagnesium-bicarbonate composition with low total dissolved solids. is chemical composition is characteristic of waters derived from modern recharge from the shallow subsurface parts of a limestone aquifer. Higher concentrations of sodium and chloride in springs located in the northern-northeastern parts of the study area suggest some involvement of halite dissolution. is observation is further supported by (1) good correlation between sodium and chloride (r 2 = 0.74), (2) chloride/bromide ratios in excess of 400, and (3) low concentration of sodium and chloride in the rain water (Oetting, 1995; Musgrove and Banner, 2004). Higher concentrations of sodium and chloride in the springwater could potentially develop due to evaporation and accumulation of these elements in the soil zone with subsequent discharges under recharge events, but the spring waters do not show any signicant eect of evaporation as observed from their depleted 18O isotopes. One sample with high sodium may suggest that the excess sodium is derived from dissolution of feldspars contained in the aquifer materials, supported from its proximity to the Figure 2: Plot of Na versus Cl of the spring waters. Note that most samples fall on the 1:1 line. Figure 3: Plot of land surface elevation versus total dissoled solids. Note total dissoled solids in the spring waters increases at lower elevations. Figure 4: Plot of 18O versus 2H for the spring waters and groundwater. Note most samples plot along the Local Meteoric Water Line (part of the groundwater data om Jones and others, 1997 and Lambert and others, 2000). Figure 5: Plot of land surface elevation versus 13C. Note 13C values in the spring waters gets heavier at lower elevations.

PAGE 189

Earth Sciences 1450 2009 ICS P roceedings 15th International Congress of Speleology Llano Upli. A subset of the spring water samples, with detailed analyses of chemical and isotopic data, shows moderate correlation (r2 = 0.61) between land surface elevation and total dissolved solids (Fig. 3). is relationship may suggest that springs at higher elevations discharge rapidly due to a steep drop in elevations, have shorter ow paths, and thus retain fresh water compositions. Some of the groundwater that still recharges further from the springs, moves deeper into the aquifer acquiring more dissolved solids through waterrock interaction and nds its way as spring discharges along the contacts of the Edwards Group and the tighter Upper Glen Rose Limestone outcrop. Changes in 13C in the spring waters at dierent elevations are also observed. For example, lighter 13C isotopes preferentially occur in the spring waters at higher elevations and heavier 13C values at lower elevations (Fig. 5). Under open system conditions of the shallow groundwater, 13CTDIC approach 13C values of soil CO2 with values of -18 to -22 PDB. Considering a equilibrium fractionation factor of about ~ +8 PDB between soil CO2 and HCO3, 13CTDIC of the recharge water would range from -10 to -14 PDB as observed in the spring waters at elevations of 450 to 600 meters (Figure 5). On the contrary, some spring waters at elevations between 150 to 400 meters have 13C values of -4 to -8 PDB suggesting carbon contribution through carbonate dissolution from the deeper parts of the aquifer under close system condition. Isotope values of 87Sr/86Sr in the range of 0.7076 to 0.7079, with a few samples having values of 0.7080 to 0.7086, suggest that the spring waters largely retain 87Sr/86Sr values of local marine limestones negating involvement of any deeper uids. Variations observed in 87Sr/86Sr are probably functions of various degrees of mixing of the rainwater and the shallow groundwater. One spring water to the northern part of the study area shows 87Sr/86Sr values of about 0.710, suggesting ow from silicate-bearing rocks in the Llano Upli aquifers. Other lines of evidence, such as plot of 18O isotopes along the Local Meteoric Water Line suggest that the spring waters are largely derived from local precipitation with no signicant eects of evaporation. However, three springs, with 18O values lighter than -5.5 SMOW fall to the le of the Local Meteoric Water Line, suggesting that they are not directly derived from local precipitation and were modied by subsurface processes during recharge. Most of the spring waters are also shied to the le of the Global Meteoric Water Line (Fig. 4) whereas most of the groundwater plots to the right of the Global Meteoric Water Line indicating eects of evaporation. Given these dierences between adjoining groundwater and the spring waters, it can be argued that the spring waters were largely inltrated into the subsurface through fracture openings, and thus escaping evaporation. Nearly all of the spring waters contain 70 percent or more modern carbon indicating that the waters are relatively modern. Tritium concentrations of as much as about 3 tritium units in several of the springs further suggest that the spring waters are derived from modern recharge. e above lines of evidence suggest that the springs are rapidly recharged and replenished by modern recharge and are therefore, highly vulnerable due to a reduction in rainfall during drought and/or excessive pumping of the shallow aquifer. Reduction in ow at the springs will threaten protection of the rare, endemic, and common species found in the springs as well as reduce instream ows to the receiving basin.6. ConclusionsA modern, shallow subsurface origin for most of the spring waters is supported by their chemical and isotopic compositions as well as their compositional dierences with land surface elevations. A moderate correlation between total dissolved solids and land surface elevation suggest that the length of the ow paths and groundwater residence time control spring water composition. For example, shorter ow paths between recharge and discharge at higher elevations result in fresher spring waters and longer ow paths for springs that recharge further from the springs and discharge at lower elevations result in greater water-rock interaction retaining higher dissolved solids. Similarly, lighter 13C values in the spring waters at higher elevations and their heavier 13C values at lower elevations support origin of the dissolved carbonates, and in turn, the host water that contains them, from shallow and deeper in the subsurface, respectively. Isotope values of 87Sr/86Sr further suggest that the spring waters retain characteristic isotopic signatures of local marine limestone aquifer and their variations are probably caused by various degrees of mixing of the rain water and shallow groundwater. Nearly all of the spring waters contain 70 percent or more modern carbon indicating that the waters are relatively modern. Tritium concentrations of as much as about three tritium units further support this. Given that the springs are rapidly recharged and replenished by modern recharge, they are highly vulnerable due to a reduction in rainfall during drought and/or excessive pumping of the shallow aquifer.

PAGE 190

15th International Congress of Speleology Earth Sciences 1451 2009 ICS Proceedings AcknowledgmentsWe thank Mr. Gary Franklin for collecting the spring water samples and Ms. Janie Hopkins for coordinating the sampling plan. We also thank Drs. Ian Jones, Shirley Wade, and Robert Mace, Ms. Cindy Ridgeway, and Mr. Bill Mullican for their reviews and comments. ReferencesBRUNE, G. (2002) Springs of Texas, Volume I, Texas A&M University Press, College Station, 566 pp. CHOWDHURY, A.H. (2008) Hydrogeochemistry, groundwater recharge, and groundwater ages in the aquifers of south-central Texas: Gulf Coast Association of Geological Societies Transactions, 58, 177. EPSTEIN, S.E. and T.K. MAYEDA, (1953) Variations of the 18O/16O ratios in natural waters, Geochimica et Cosmochimica Acta, 4, 213. Guyton U YTON W.F. and associates (1979) Geohydrology of Comal, San Marcos, and Hueco Springs, Texas Department of Water Resources Report 234, 85 pp. JOHNSON, S.B. and G.M. SCHINDEL, (2008) Source water determination for San Marcos Springs, San Marcos, Texas: Gulf Coast Association of Geological Societies Transactions, 58, 469. JONES, S.A., W.L. ROGER, and J.F. BUSBY, (1997) Chemical evolution and estimated ow velocity of water in the Trinity aquifer, south-central Texas, U.S. Geological Survey Water-Resources Investigations Report 97-4078, 22pp. Mc C Kinney I NNEY D.C. and J.M. Sharp H ARP (1995) Springow augmentation of Comal Springs and San Marcos Springs, Texas, Phase I-Feasibility Study, Center for Research in Water Resources Technical Report 247, 314 pp. LAMBERT, R.B., K.C.,p GRIMMl and R.W. LEE, (2000) Hydrogeology, hydrologic budget, and water chemistry of the Medina Lake area, Texas: U.S. Geological Survey Water-Resources Investigation Report 00-4148, 54 pp. MUSGROVE, M., and J.L. BANNER, (2004) Controls on the spatial and temporal variability of vadose dripwater geochemistry: Edwards Aquifer, central Texas: Geochimica et Cosmochimica Acta, 68, 1007. OETTING, G.C. (1995) Evolution of fresh and saline groundwater in the Edwards Aquifer: Geochemical and Strontium isotopic evidence for regional uid mixing and uid-rock interaction, M.A. thesis, University of Texas at Austin, 203pp. OGDEN, A.E., A.Q. RAY, and S.R. ROTHERMEL, (1986) Hydrogeochemistry of the Comal, Hueco, and San Marcos springs, Edwards Aquifer, Texas, Chapter 13 in e Balcones EEscarpment, CC entral Texas, Patrick Abbott and Chuck Woodru Jr. (Eds.), Geological Society of America, 115.

PAGE 191

Earth Sciences 1452 2009 ICS P roceedings 15th International Congress of Speleology KARST MICROCLIMATE MONITORING IN THE NORTHERN ALPS, AUSTRIA: INITIAL RESULTSAA ARON C C URt T ISC C ambridge UU niversity CC aving CC lub, aaron.curtis@cantab.net An ongoing micrometeorological program was initiated in the summer of 2007 by the Cambridge Austrian Cave Science Expedition (CASCE) with the intention of characterizing the heat ux across the entrances of two caves on the Loser plateau in the Totes Gebirge mountains, Rundreishhle and Steinbruckenhhle. e relative importance of the diusive, convective (airow), and latent (condensation and evaporation) components of that ux were examined. ree weeks of intensive monitoring conducted in 2007 resulted in 105,500 temperature data points in Rundreishhle and 64,000 in Steinbruckenhhle, as well as surface meteorological data. e long penetration distance observed for the diurnal temperature cycle into the caves implies an entrance heat ux several orders of magnitude greater than could be explained by diusive processes alone, suggesting dominance of convective and / or latent processes. 1.1 L/m of condensate was estimated in the entrance series of Steinbruckenhhle, potentially transferring 34 kJ/day to cave walls.1. Aimse Loser plateau karst microclimate study seeks to characterize the heat ux between the caves investigated, Rundreisehhle and Steinbrckenhhle, and the external air mass in terms of the relative magnitude of its component processes. Viewing the cave entrance as a plane, the heat ux QEnt across that surface is expected to be the sum of the ux of three types of heat transfer across that plane. ese are the advective sensible heat ux QS, the advective latent heat ux QL, and the diusive heat ux QC through limestone, such that QEnt=QC+QS+QL (1) e relative importance of these components is assessed here using data from simultaneous datalogged measurements of cave and surface atmospheric variables, as well as visual observation by a team of cavers. To begin with, the importance of the diusive ux QC is investigated. Because QC cannot be measured directly, I tested a theory that assumes all heat transfer in caves is via Badinos (2004) penetration length model by observing the distance from entrances at which diurnal temperature cycles are no longer visible. is zone of diurnal temperature variation is similar to the the heterothermic zone of Luetscher and Jeannin (2004) for a daily rather than annual periodicity. Evidence that QC alone is a very small component of heat transfer through the caves leads to a discussion of airow through the cave and the chimney eect (Michie, 1997), representing QS. Likely latent heat ux (QL) is addressed through wet and dry bulb measurements of atmospheric water vapor content and visual observation of condensation. Existing theories are critically evaluated in the light of our data, to develop a conceptual model of heat transfer with empirical limits based on our observations. 2. Cyclical VV ariations and Penetration LengthsHeat transfer across the cave entrance is investigated initially on the diurnal scale in this study. If the cave is modeled as rock, we would expect a sinusoidal temperature cycle that decreases in amplitude with distance into the cave. Longer period cyclicities (such as seasonal or even glacial cycles) would be expected to cause the same amplitude signal further into the cave. For each period length, there is a distance at which the amplitude approaches zero. is is referred to as the penetration length, xcycl, dened as (2) where a is thermal diusivity (m2s-1) of the cave air or limestone, and is the period of the forcing cycle (Badino, 2004). Taking only diusive heat transfer through cave walls into account, we can use values for a obtained for limestone through laboratory analysis, around 1.0x10-6 m2s-1. Using this value, the annual cycle (8760 hours) should disappear only 3.4 m into the cave, and glacial-cycle scale inuences should be extend 339 m into the cave. Taking into account impurities such as water in the limestone, thermal diusivity can be a factor of 10 higher, implying that the annual cycle would extend a more reasonable 10 m into the cave.

PAGE 192

15th International Congress of Speleology Earth Sciences 1453 2009 ICS Proceedings Measurements of air temperature at distances from cave entrances have found values of penetration depth much higher than those predicted by these values of limestone thermal diusivity. For example, Forbes (1998) detected diurnal variation in both temperature and humidity 75 m into a cave. e present study found diurnal temperature variations 15 m into the cave. ese cannot be accounted for by the thermal diusivities of any of the components of the cave system. erefore, other processes must be having a signicant eect. 3. Diusive Heat Transfer (QC)To investigate penetration length in Rundreisehhle, thermistors (labeled rA3, rA4, rB1, rB2, rB4, rC1, rC2, rC3), were placed roughly every three meters along the cave (their locations were surveyed to BCRA grade 5). Next, spectral analysis was carried out to determine whether the external diurnal temperature forcing cycle was observable at each distance into the cave, and if so its amplitude and wave energy (Fig. 1). If the (Badino, 2004) diusive model of penetration length holds true for Rundreisehhle, then, using the standard thermal diusivity for limestone of 1.0x10-6 m2s-1, we would expect the amplitude of the diurnal cycle to approach zero at 0.08 m. Instead, all six in-cave thermistors on the plot demonstrated convincingly diurnal peaks in their spectral analyses. Referring to Figure 1, we see that the peaks occur at slightly longer periods of 1549 and 1570 minutes, or 25.8 and 26.1 hours. Noise is probably responsible for the deviation of these peaks from 24h. e surface weather station demonstrated peak wave energy at 23.75 h. Because Rundreisehhle is a simple 25 m tube of a cave with two entrances, and the thermistors extended the entire length of the cave, the diurnal cycle had a penetration length of at least 12.5 m. is is more than an order of magnitude further than would be predicted by conduction through limestone. erefore, Qc is only responsible for a very small portion of the total heat transfer into the cave. However, the peaks do attenuate towards the center of the cave, as shown in Figure 2, eectively an east-west transect of Rundreisehhle where the energy of the diurnal peaks from the spectral analyses has been plotted against distance along the cave. We can investigate the nature of the relationship between diurnal amplitude and distance into the cave by tting regressions to our data. A parabolic regression ts best, and explains 98.4% of the data (adjusted R2), well within the 95% condence intervals displayed on the graph. is weakening of the diurnal cycles wave energy with distance into the cave is a pattern that is not exclusive to a Qc-driven system; it would be observed in a QS or QL dominated system as well. Before discussing these possibilities, we can consider the penetration length results for one entrance of Steinbrckenhhle, an extensive cave system with nine entrances. e Rundreisehhle measurements provided us with a lower bound for penetration length of 12.5 m, but no upper bound on that value because the cave was too short for the wave energy of the diurnal cycle to approach zero at any point. Because Steinbrckenhhle is a larger cave, however, we were able to place thermistors as far as 180 m from any entrance. Table 1 shows the results of the spectral analysis, which raises our lower bound for penetration length to 15 m, and provides an upper bound of 35 m. Figure 1: Fourier transform of Rundreisehhle temperature timeseries. Figure 2: Amplitude of energy peak for diurnal wavelength om Fourier transform for each sensor, with distance into the cave. A parabolic regression, with the equation amplitude=6273934.8m into cave+ 39.26m into cave2 ts well. Entrances are at 0 and 25m.

PAGE 193

Earth Sciences 1454 2009 ICS P roceedings 15th International Congress of Speleology 4. Airow (QS)Because Rundreisehhle is nearly straight, we can test the inuence of external, cave-parallel wind using weather station and cave thermistor data. If external wind controls air movement in the cave (rather than chimney eect), we would expect to see a temperature increase when a strong wind blows in a direction parallel to the cave, as this would increase QS, the advective ux of heat from the entrance of warm outside air into the cave. Rundreisehhle runs almost exactly east-west. A Pearson correlation matrix of the eight Rundreisehhle thermistors with wind speed, north-south wind component, and east-west wind component showed that the temperatures are not controlled by surface airow, suggesting either a lack of signicant cave air movement or a dominance of the chimney eect. Correlations with p=0.05 for the wind variables occur for thermistor rA4, rB2, rB4, rC2 and rC3, and rA3 has one at p=0.08. rA4 and rA3 are the surface and entrance thermistors. It is interesting to note that rA3, located in the sheltered entrance depression but outside the cave entrance, shows a fairly strong relationship with north-south wind but not east-west wind. ermistor rA4, on the surface but below the dwarf pine canopy, correlates with wind blowing in both directions. e remaining wind-correlated thermistors, rB2, rB4, rC2, and rC3, are correlated with wind but not cave-parallel wind. External wind is almost certainly not a factor in Steinbrckenhhle, where two transects of sensors were placed. Although this cave would be classied as a Type V (multiple entrances at dierent levels, chimney eect likely) cave according to the scheme of (Michie, 1997), for our purposes we can think of the specic sections we are monitoring as simpler caves. e E entrance transect can be approximated as a type III cave (single entrance, descending passage), and the CSB area can be viewed as a type IV (single entrance, ascending passage). In this case, we are merely considering the eect of entrances 204E and 204C on the nearby cave microclimate.5. Condensation (QL)Extensive measurements of condensation have been carried out in the caves of the Crimea and the Caucasus, reviewed by (Dublyansky and Dublyansky, 1998). In that paper, condensation was related to discharge of streams fed by karst aquifers, and several proposed formulas for predicting underground condensation were put forth. Corrosion in association with condensation is an agent of speleogenesis and has been discussed by Dreybrodt et al. (2005). Here, we are particularly interested in the latent heat transfer (QL) associated with cave condensation. (Michie, 1997) stresses that condensation always results in a net transfer of heat from the cave air to the cave walls. However, because cave air is eventually replaced, the cave as a whole is a net recipient of heat from condensation. To observe condensation, expedition members were instructed to look for morphologies that suggest condensation corrosion (Dreybrodt et al., 2005). According to (Jameson, 2005), these include drop dents, rill trails, and splash patches. However, despite 30 pairs of observant caver eyes inspecting the rock for these features, none were reported. Direct, visual surveys of condensation itself proved more eective. Making use of the frequent trips to various parts of Steinbrckenhhle, I asked expedition members to keep an eye out for any walls covered partially or completely in water droplets, and aerwards surveyed the cavers informally. Responses highlighted three main areas in which striking examples of the phenomenon were commonly observed. ese three areas are shown in Figure 3. All three were near entrances, and conveniently, these happened to be in close proximity to our temperature transects: 204E entrance, CSB passage, and Crowning Glory. One caver described the elds of droplets as really pretty, while another remarked that droplets were quite extensive and occurred generally on `underside surfaces. ere was a general feeling that condensation on this scale is not typical of the UK caves in which expedition members did most of their caving, perhaps a conrmation of Dublyansky and Dublyanskys (1998) prediction that this altitude and latitude provides better conditions for condensation than those of Britain. To an observer, the condensation droplets appear similar to the reective dots described by Rowling (2001) (Fig. 4). Rowling held that such droplets are not caused sA3 & sA4 sA1 & sA2 sB1 & sB2 sB3 & sB4 sC1 & sC2 sC3 & sC4 Distance from 204e entrance 4 15 35 75 150 180 Spectral analysis peak period 23.8 hours 24.0 hoursNone None None None Peak amplitude 795.8 719.8 None None None NoneTable 1: Spectral analysis of thermistors in Steinbrckenhhle.

PAGE 194

15th International Congress of Speleology Earth Sciences 1455 2009 ICS Proceedings solely by thermodynamic phase change but that bacteria colonies of the genus AA ct inomyce encourage nucleation of condensation droplets and attract water using hydrophilic bers. is is in accordance with the increasing realization that bacteria are important agents in caves and that many geophysical processes that occur underground depend on microbiology, as evidenced in Barton (2006) and in a special speleological issue of Geomicrobiology Journal (v18, 2001). However, the condensation droplets that we observed do not show the features that Rowling considers indicative of the biological origin of these droplets. Rowling expects the droplets may exhibit a gold, yellow, or brown color, which Moore and Sullivan (1997) attributed to the presence of the pigment beta carotene in association with actinomycetes. Additionally, bacterial droplets should be small: 0.1 mm to 2 mm. Because our droplets are clear and range from 4 mm to 8 mm, it is unlikely that this bacterial mode of formation is involved. Two trips into Stienbrckenhhle were conducted to photograph condensation. By measuring the pixel size of droplets and scaling with the use of a measuring tape included in the photos, we found that the vast majority of droplets were between 4.6 and 7.7 mm in diameter, and were spaced at one drop every 2 to 4 cm2. With an average passage perimeter of 4 m, this amounts to about 1.1 L per one meter of passage. So, in the 170 m of passage we were observing, there was roughly 187 L of condensate present as droplets on the walls. While it is possible to make these limited guesses as to the abundance of condensed water, the actual net rate of condensation rather than the amount visible at any one time is required in order to understand the thermal eects of the condensation. ere are indications that the condensation process was active during our study period, and most of this condensation may have been formed each day. Condensation occurred on the Hobo datalogger in CSB passage during the period of investigation (Fig. 5). Surveys of cavers suggest that there is a diurnal cyclicity to the formation of this condensation. e majority of reports of extensive condensation occurred in the evening; cavers noticed condensation while leaving rather than entering the cave, and one caver remarked that there was more in the evening/aernoon. Condensation in the 204E entrance passage, Germkndels Revenge, was observed exclusively in the evening. According to the expedition logbook, few trips were underground during the period between 2 am and 10 am, however, so we cannot be certain that droplets were not present then, although it seems likely due to the absence of condensation from 11 am to around 2 pm is is not entirely in accordance with the predictions put forth by Dublyansky and Dublyansky (1998) with the Figure 3: CSB passage, with condensation droplets. Figure 4: Condensation in Crowning Glory. Streaks detailed in inset are interpreted in the text as rivulets down which condensate slowly ows (although this is not visible on an observers timescale.) Figure 5: Condensation droplets in CSB Passage. Moisture on the logger (inset) demonstrates that the droplets formed during the study period.

PAGE 195

Earth Sciences 1456 2009 ICS P roceedings 15th International Congress of Speleology microclimatic method for condensation estimation. ey imagine a maximum at 10 am to 4 pm and a minimum at 10 pm to 2 am, corresponding with maxima of surface temperature and humidity. However, by observing condensation droplets, we are not observing the rate of condensation, but rather the cumulative volume since the beginning of the period t. Maximum visible condensation is likely to occur at the end of the period t, which is likely to be the aernoon through early evening. If droplets were not present in the morning, this implies that not only are condensation processes more active during the evening, but crucially that the droplets were removed from the walls by another process during the night or morning. Evaporation is one possibility. de Freitas and Schmekal (2006) produced a conceptual model of the vapor ux between cave air and walls as a continuum, cycling sinusoidally between condensation and evaporation. Besides evaporation, it is also possible that the droplets are removed by mechanical means. Gravity is the most likely culprit. e majority are attached to the underside of rock by adhesion and surface tension, and appear to be stationary, without owing or dripping, to casual observation. However, the droplets could be moving slowly, coalescing and owing down the side of the rock. Dreybrodt et al. (2005) described such ow from the rock surface down to the cave oor. Close observation of Figure 4, a photograph taken of the roof of Crowning Glory Passage, suggests this is such a case. One can clearly make out vertically aligned stripes in the condensation pattern, presumably representing long timescale rivulets. In addition to ow down the wall through these rivulets, it has been suggested that the limestone in our area is porous enough that the droplets could ow directly into the rock and enter the pore spaces, potentially another signicant pathway for mechanical removal of the condensation water (Charles Self, personal communication, 2008). Assuming that the droplets are removed nightly by a non-evaporative process, we can estimate the daily latent transfer of heat to the cave walls, using waters enthalpy of condensation, which is roughly -2.5 kJ g-1. at the temperatures relevant here (near 0 degrees C) 228.64g day-1-2.5kJ g-1 = -555.6 kJ day-1 (3) If this much condensation is occurring each day, it may be a major control on the cave wall temperature. However, this is only a rst order approximation at best, and it is very likely that a signicant percentage of the condensation droplets remain in the cave overnight, reducing the daily heat transfer into the cave. To determine whether the droplets are removed by owing, dripping or by evaporation, time-lapse photography could be employed. 6. Conclusions and Foundation for Continuing Research is study provided a rst look at the likely relative magnitudes of component processes in diurnal heat transfer at Rundreisehhle and Stienbrckenhhle. e wave energy of the diurnal cycle approaches zero at a distance greater than 15 m but less than 34 m from entrances, more than an order of magnitude further than would be expected if diusive heat transfer through cave walls (QC) alone was responsible. e abundant condensation observed in Stienbrckenhhle suggests that condensation on cave walls may be a greater factor. Disentangling the importance of airow through caves is more complicated; at Rundreisehhle, cave temperature correlated with surface windspeed, but not specically in a cave-parallel direction. Future investigations in these caves should conduct anemometry at multiple entrances to measure the chimney eect.Acknowledgmentse 31 members of the Cambridge Austrian Cave Science Expedition 2008 each contributed in an important way to the collection of the data for this project. I would particularly like to acknowledge Djuke Veldhuis for research coordination and sponsorship eorts, Richard Mundy for assistance running uid dynamic simulations, and John Billings, Edvin Deadman, Oliver Stevens, Duncan Collis, and Andreas Forsberg for help with data collection and transportation of equipment. Assistance with electronic and hardware construction was provided by Wookey, Chris Hopkins, and Adrian Hayes, and Oliver Madge. Approval and nancial support from the British Cave Research Association (BCRA), the Royal Geographical Society with IBG (RGS), and the Cambridge Expeditions Committee were vital to the success of this investigation. Equipment and consumables were supplied by our corporate sponsors: Tunnocks, Peli, Princeton Tec, Silva, Hilti, Mickies Place, Mornake, and Whitworths. Cooperation of those in Austria was also vital. I would like especially to acknowledge Hilde Wilpernig and the Gasthof Staudnwirt, Robert Seebacher of the Verein fur Hlenkunde in Obersteier and the workers of the Loser Panoramastrasse.ReferencesBadino, G. (2004). Cave Temperatures and Global Climatic Change. International Journal of Speleology, 33(1/4): 103.

PAGE 196

15th International Congress of Speleology Earth Sciences 1457 2009 ICS Proceedings Barton, H. A. (2006). Introduction to cave microbiology: a review for the non-specialist. Journal of Cave and Karst Studies 68(2): 43. de Freitas, C. and A. Schmekal (2006). Studies of condensation/evaporation processes in the Glowworm Cave, New Zealand. International Journal of Speleology 35(2):75. Dreybrodt, W., F. Gabrovek, and M. Perne. (2005). Condensation corrosion: a theoretical approach. Acta Carsologica 34(2): 317. Dublyansky, V.N. and Y.V. Dublyansky (1998). e problem of condensation in karst studies. Journal of Cave and Karst Studies 60(1): 3. Forbes, J. (1998). Air temperature and relative humidity study: Torgac Cave, New Mexico. Journal of Cave and Karst Studies 60(1): 27. Jameson, R.A. (2005). Modication of scallops by condensation corrosion in Snedgars Cave, West Virginia, USA. 2005 Salt Lake City Annual Meeting 207-10. Abstract retrieved from http://gsa. confex.com/gsa/2005AM/nalprogram/abstract_ 97210.htm Luetscher, M. and P.Y. Jeannin (2004). Temperature distribution in karst systems: the role of air and water uxes. Terra Nova 16(6): 344. Michie, N. (1997). An Investigation of the Climate, Carbon Dioxide and Dust in Jenolan Caves, N.S.W. PhD esis, Macquarie University Earth Sciences.. Moore, G.W. and G.N. Sullivan (1997). Speleology: Caves and the Cave Environment, Cave Books. Rowling, J. (2001). Cave entrances reective dots. retrieved from http://www.speleonics.com.au/jills/ byzone/byzoneReDots.html on September 10, 2007

PAGE 197

Earth Sciences 1458 2009 ICS P roceedings 15th International Congress of Speleology GEOMORPHOLOGICAL EV V OLUTION AND DIGITAL MAPPING OF THE KSIROMERO REGION, WESTERN GREECEMIL IL JANA ANA GOLU GOLU BO O VIC IC DELIGIANNI DELIGIANNI1, KO O SMA A S PA AVLO LO POULO OULO S1, GEORGE GEORGE VENI ENI2, I I SSAA AA K PARCHARIDI ARCHARIDI S1 1HH arokopio UU niversity, Faculty of GG eography, 70 EE l. Venizelou Str., 17671 AA thens, GG reece, miljanamakis@yahoo.com, kpavlop@hua.gr, parchar@hua.gr2NN ational CC ave and Karst RResearch II nstitute, 1400 CC ommerce DDrive, CC arlsbad, NN ew Mexico 88220, UU SA A gveni@nckri.org Ksiromero is a region in the northeast section of the Prefecture of Aitoloakarnania, in western Greece. e region has a total areal extent of 106.76 km2 and ranges in elevation from 178 m to 1,314 m above sea level. e karst of Ksiromero is primarily developed on beds of 10-200 m thick Triassic carbonate breccia conglomerates which occupy 70% of the region; 200-300 m thick Upper Triassic to Early Jurassic limestones and dolomites occupy 20%, uaternary deposits 17%, and 150 m thick Triassic gypsum 3%. e region is normally dry, but streams ow aer strong rains. e Ksiromero region contains a dense network of dolines, and more than 75 water reservoirs, which include depressions excavated in terra rossa. is geomorphic study is based on eld research, laboratory analysis of water and rock samples, and work with Able Soware R2V, ARCGIS 9.2, and ERDAS 9.1 geographic information systems soware. We created an ArcGIS karst feature and geodatabase and developed a mechanism for symbolically identifying the ephemeral, temporal, or inferred features common for karst in this region. All 1:5000 topographic maps of the region were digitized at 2-m contour intervals to allow detailed visualization and analysis. Arc Info 9.2, with 3DAnalyst Tools and Spatial Analyst, were used to create .TIN, grid and 3D shapeles and high resolution digital elevation models. We also created a digital geological map of the study area and examined Landsat satellite images using principal components analysis, where we could recognize, compare and measure karst landforms from the digital karst geomorphological map model with the satellite imagery. Karst in the carbonate breccia conglomerates has dramatic features, including richly developed elds of dolines, uvalas, and poljes. e main karst drainage basin has three smaller peripheral basins. In some cases, it was dicult to categorize karst landform types by eld research and the digital topographic model. From our parallel study of the digital karst geomorphological map with the satellite imagery, we conclude that the geologic structure of the carbonate breccia conglomerates does not provide the information needed to remotely classify small karst forms. Additionally, the gypsum-cemented clasts create relatively high uniform permeability throughout that outcrop, compared to most karst terrains, which minimizes runo and development of additional karst features. Poljes of tectonic origin were easily recognized. e analysis yielded information to better evaluate eld results for other possible poljes in the region. We identied 14 poljes; however, six previously identied poljes are now recognized as part of a larger compound basin. Residual hills are predominantly intercalated limestone and gypsum, likely the result of lesser solubility than tectonic factors. e system of solution dolines and terra rossa cisterns are adequate water reservoirs, but it is necessary to protect their water quality from pesticides used in the watershed. e karst geomorphological map with the geodatabase will be used to guide future land use in this region.1. Introductionirty-three percent of Greece is karst (PapadopoulouVrynioti, 2004), yet Greece has no specialized governmental or private karst research association. Consequently, knowledge is scattered, and hydrogeological karst investigations are oen not based on a clear understanding of karst geomorphology. With this project, we try to provide a model for the country that integrates karst evolution with the study of the organization and origin of karst landforms, their inuence on human activities, and the impact of those activities on karst. Ksiromero is a 106.76 km2 region in the northeast section of the Prefecture of Aitoloakarnania, in western Greece. It is geographically bounded by the mountainous area of Akarnanika Ori to the west (Psili Kory, the highest peak

PAGE 198

15th International Congress of Speleology Earth Sciences 1459 2009 ICS Proceedings is 1,157 m above sea level), the basin of Lake Ambrakia to the east, Amvrakikos Gulf to the north, and the southern watersheds whose rivers ow to the Ionian Sea through the Department of Akarnania. e karst of Ksiromero is richly developed with dierent types of dolines, poljes, and uvalas. Since ancient times, it has been sparsely populated, and agriculture is limited due to small water supplies. Animal husbandry, mainly sheep-herding, is the populations primary livelihood; farming of cultivable karst surfaces is secondary. In this paper we present some of the puzzling questions of our karst geomorphological research of Ksiromero, and answers we developed through the creation of a karst geomorphological map and geodatabase, with insights from Landsat satellite images. Karst geomorphological research that focuses on the resolution of land use problems, combined with an understanding of the natural history and development and functions of karst features is necessary to sustainably live with karst (Veni, 1999).2. Regional Geological Setting e Ksiromero region is within the Ionian geotectonic zone, part of the large External Hellenides Platform that extends west from mainland Greece and appears on the western edge of Peloponnesus and in some of the Dodecanese Islands. Primary tectonic features of the Ionian Zone are long thrust faults that trend east-west and northeast-southwest, and long reversed and normal faults that trend northwestsoutheast. e faults congurations have created the large tectonic basin of Ksiromero and other basins in western Greece. Geologic mapping of the region is provided by the Institute of Geology and Mineral Exploration (1986, 1987). Figure 1 illustrates these units and karst geomorphic and hydrologic features. During Permo-Triassic time, the Ionian Zone was a shallow, restricted, marine basin which accumulated over 150 m of evaporites (Karakitsios, 1992), of which gypsum is notably exposed. eir episodic deposition may have allowed the development of a paleointrastratal karst, as described in other regions by Bosk et al. (1989), although no direct evidence has been found. Triassic Tryphos Formation carbonate breccia conglomerates were deposited over the evaporites, followed by up to 200 m of dolomite and as much as 300 m of the Pantocrator Limestone into the Early Jurassic. At that time, the shallow ethos Sea covered a continental platform which extended throughout nearly all of western Greece. e carbonate breccia conglomerates are epigenetic, formed during the Triassic from major tectonic activity, diapiric deformation, and dissolution of underling evaporites. ese conditions continued with small modications up to the end of the Jurassic. From the Pliocene to the uaternary, more recent gypsum deformation occurred at the surface due to underlying diapiric movement along prominent faults (Underhill, 1988). e Tryphos carbonate breccia conglomerates are the main karstied rocks of the region. ey range from 10-200 m thick and cover 70% of the area. Petrographic analyses, the geological mapping, and studies of hand samples identify them as microcrystalline to sparry calcite limestone and dolomite breccia and pebbles, up to 30 cm in diameter. In some places, the carbonate clasts have become marble. Traces of pyrite and quartzite are present. e clasts occur in a cohesive to slightly cohesive gypsum matrix that weathers into a terra rossa matrix.3. Local Karst Hydrologye word Ksiromero is Greek for dry place. From a geographic perspective, the region internally drains water to where it is not exploitable for use. Hydrogeologically, it is a system of closed karst watersheds whose recharge Figure 1: Geomorphic and hydrogeologic map of Ksiromero, Greece (geology based on Institute of Geology and Mineral Exploration, 1986, 1987).

PAGE 199

Earth Sciences 1460 2009 ICS P roceedings 15th International Congress of Speleology characteristics are poorly dened and where the downgradient destination of its groundwater is unknown. While recharge occurs through the higher elevation limestones and dolomites, most of it is transmitted though the Tryphos breccia conglomerate and the underlying gypsum. e highly soluble gypsum matrix of the conglomerate produces a relatively uniform highpermeability surface that minimizes surface runo. While most karst features of the region are developed in the Tryphos, their density may have been greater if the matrix had been carbonate. Additionally, the few open caves and conduits in the region likely result from the high production of residual terra rossa soils from the gypsum. e soils accumulate on all surfaces, especially in dolines, and runo is insucient to transport them through the karst to create more open caves. In some lower-central elevation areas, groundwater occurs in two horizons, separated by a thin bed of clay and marl. One is 4-12 m below the surface and the other between 15-30 m. ey serve as small, local water supplies where they pool in some dolines. Deep wells in the region are not known to yield usable quantities of water because they either do not reach the water table, and/or possibly because most groundwater is restricted to conduits which are unlikely to be intersected by drilling. Generally, water is supplied to Ksiromero from neighboring regions and stored in more than 75 reservoirs. Some are natural solution dolines fed by groundwater, and others are natural or excavated depressions in terra rossa. To prevent loss of water into the ground, the bottoms of some of the terra rossa reservoirs are cemented. Chemical analysis of the imported water shows it is usually within drinking water standards, although due to vulnerability to contamination, it is more suitable for irrigation. 4. Karst Geomorphology and Geographic Information System (GIS) AnalysisTo better evaluate the morphologic and hydrologic characteristics of the region, we used ArcGIS Info 9.2 to create a detailed geodatabase of karst landforms, and geographic and hydrologic features that represent and/or inuence karst development. We digitized surface water ows, which exist only following periodic rain. Topographic contours were digitized at 2-m intervals above mean sea level. We used Spatial Analyst to create a digital elevation model (DEM) to calculate surface area, ow direction, ow accumulation, ow sinks, stream networks, stream links, and watershed boundaries. We created a database with all existing karst forms and their sub-groups in dierent layers. We identied 14 poljes in the region; six were previously identied but we recognized them as part of a larger compound basin. rough watershed delineation, we identied one large compound karst basin, which includes nine smaller polje watersheds, Doline Elevation above mean sea level (m) Depth (m) 3-D surface area (m2)Type of doline (per ar, 2001) Amuistra323.8 352 28.21,080,500shallow, near-fault, point recharge Stadopigado135.4 240 104.6481,400funnel-like, near-fault, point recharge Palaiofrazata243.5 284 40.5265,900funnel-like, near-fault, point recharge Mpoimo337.9 376 29.1262,600funnel-like, fault doline, point rechargeTable 1: Morphometric characteristics of major dolines in Ksiromero. Polje Elevation above mean sea level (m) Watershed 3-D surface area (m2) Polje 3-D surface area (m2) Type of polje Residual hills (Hum) Fraksias178.8 20020,518,2001,397,900 tectonic, evolved from dolines and uvalas 1 Geladomandra235.2 240 5,627,200860,700 tectonic, border 2 Kserolagia210.3 246 10,160,600854,600 border, evolved from dolines and uvalas 6 Spilia246.4 28810,673,500278,500 tectonic, structural, point recharge, ponor 0 Stinadia200 23027,489,3005,259,800 radial, tectonic, structural, point recharge 3Table 2: Morphometric characteristics of polje watersheds in Ksiromero.

PAGE 200

15th International Congress of Speleology Earth Sciences 1461 2009 ICS Proceedings and ve smaller separate watersheds for poljes. We found 17 residual hills in the poljes, comprised predominantly of limestone intercalated with gypsum and likely the result of lesser solubility than tectonic factors. We identied a total of 278 dolines with funnel, shallow, and a few collapse morphologies. We also identied 12 uvalas. Ponors are rare but exist at the contact of the limestone with the breccia conglomerates and follow faults. Using ars (2001) classication scheme, Ksiromero primarily has broken, broken collapse, near-fault, and fault dolines. Our database included dierent layers for each type of karst features, and using the DEM, we applied Spatial Analyst and 3D Analyst to calculate surface areas and slopes. Some results are summarized in Tables 1 and 2.5. Satellite imagingWe adapted Landsat satellite imaging of Ksiromero, using ERDAS Imagine 9.1 soware, for principal components analysis (PCA). is linear transformation technique in image processing reduces data redundancy between spectral bands (Sabins, 1996). Using PCA for this study, standardized principle component transformations were conducted using the six bands of the Landsat image (excluded the thermal band to recognize karst morphologies. We determined: PC1 = 92.7%, PC2 = 4.2%, PC3 = 2.4%, PC4 = 0.4%, PC5 = 0.25% and PC6 = 0.05%. Figure 2 is an example of PCA-examined imagery, specically PC2. In combination with PCA, we used ArcMap 9.2 to compare, evaluate, and further reduce image noise and better recognize karst features. With that technique, we evaluated a combination of stretched symbology values while changing color schemes to best accentuate the features. Both PCA and color scheme methods produced the same results. e rst principle component looks like a panchromatic image and corresponds to the brightness, providing information about topography and albedo. is principal component contains most of the information needed for recognizing large karst features (the largest poljes) on the breccia conglomerates. PCA and color scheme analysis also identied parts of the largest poljes watershed where limestone pavement and smaller karst features like dolines or uvalas are otherwise dicult to impossible to remotely discern.6. ConclusionKsiromero is a region of carbonate karst, with morphologic and hydrologic characteristics complicated by underlying evaporites and preferential dissolution of the gypsum matrix in the carbonate breccia conglomerate. Its features cannot be adequately understood based on carbonate karst conditions alone. e size of a dolines watershed is oen roughly proportional to the size of the karst conduit at its base. In contrast, the more uniform high permeability surface of the breccia conglomerate in Ksiromero recharges water more diusely than a well-cemented limestone, resulting in less runo and more sediment deposition in dolines to eectively reduce conduit size. However, the high, though diuse, permeability suggests that groundwater fed by such recharge, as well as through open conduits, is highly vulnerability to contamination from the drained surface. e system of solution dolines and terra rossa reservoirs are adequate for the areas modest water demand, but water quality must be protected from pesticides and other agricultural contaminants. e GIS-based map and geodatabase from this study can be used to assess vulnerability, plan protective measures of individual sites of outstanding vulnerability, test future regional groundwater recharge models, and develop supplemental water supply and conservation plans. Figure 2: Hydrologic and geomorphic Landsat image of Ksiromero examined for second level principle components analysis (original imagery was evaluated in color).

PAGE 201

Earth Sciences 1462 2009 ICS P roceedings 15th International Congress of Speleology ReferencesBosk, P., D.C. Ford, J. Glazek, and I. Horacek (Eds.) (1989) Paleokarsta systematic and regional view, Academia Prague, Czechoslovakia, 725 pp. ar J. (2001) Structural bases for shaping of dolines. AA ct a C C arsologica, 30(2), 239. Institute of Geology and Mineral Exploration (1986) Geological map of Greece: Astrakos Sheet. Institute of Geology and Mineral Exploration, 1 sheet. Institute of Geology and Mineral Exploration (1987) Geological map of Greece: Amphilochia Sheet. Institute of Geology and Mineral Exploration, 1 sheet. Karakitsos V. (1992) Ouverture et Inversion Tectonique du Bassin Ionien (Epire, Grece). AA n nales GG eologues des pays HH elleniques, 35, 185. Papadopoulou-Vrynioti, K. (2004) e role of epikarst in the morphogenesis of the karstic forms in Greece and specially of the karstic hollow forms. AA ct a C C arsologica, 33(1), 219. Sabins, F.F. (1996) RR emote Sensing: Principles and I I nterpretation, W. H. Freeman and Company, New York, 3rd ed., 494 pp. U nderhill, JJ .R. (1988) Triassic evaporates and Pliouaternary diapirism in western Greece, Journal of t he GG eological Society, 145(2), 269-282. Veni, G. (1999) A geomorphological strategy for conducting environmental impact assessments in karst areas. GG eo morphology, 31, 151-180.

PAGE 202

15th International Congress of Speleology Earth Sciences 1463 2009 ICS Proceedings MORPHOLOGICAL RELATIONSHIPS BETWEEN ERRATIC BOULDERS AND ASSOCIATED BEDROCK LIMESTONE FINS OR ROCK COMETS, MADRE DE DIOS ARCHIPELAGO, CHILEJOEL OEL DE DE SPAIN AIN1, RICHARD RICHARD MAIRE AIRE2, STE E PHAN HAN JALLIE ALLIE T31. SEq Q UOIA AND KINGs S C C ANy Y ON N N At T IONAL PARks KS 47050 G G ENERALs S H H IGHw W Ay Y THREE R R Iv V ERs S CA CA 93271, U U SA A 2. L L Ab B ORAt T OIRE ADE ADE SDym DYM Et T U U NIv V ERs S It T E BORDEAUx X 3, CNR CNR S, MAIs S ON DEs S SUDs S 33607 PEss SS AC, FRANCE 3 L L Ab B O RAt T ORIE ED ED YTE E M, U U NIv V ERs S It T E DE SAvV OIE, CNR CNR S, C C Amp MP Us S s S CIENt T If F Iq Q UE, 73376 L L E BOURGEt T DU L L AC, FRANCE Rock comets are apparently unique karst landforms composed of glacial erratic boulders and bedrock limestone ns from the Madre de Dios Archipelago, Chile. Fins form due to dierential erosion under conditions of prolic rains and strong frequent winds. We measured 50 erratics at two sites as well as their proximal bedrock n, n orientation and grooves adjacent to the windward side of the erratic boulders and analyzed this data with standard statistics. Comet orientation varied 29 at Suplador Pass while on Tarlton Island comet orientation varied 52 due to the wider distribution of the comets across a ridgeline. Erratic width was a good predictor of n width and groove width. But, these were the only strong relationships demonstrated. Erratic height is not a good predictor or n height or length and erratic average dimensions are also not a good predictor of n length. Erratic height also does not predict groove depth. Overall comets seem to develop based upon more specic and localized conditions such as the overall shape of the erratic, location relative to other comets, bedrock features, or vegetation and specic location along the ridge line or pass area all of which eect wind ow and rain fall.1. Introductione Madre de Dios Archipelago of Chile at 52 south latitude hosts the karst-rich Tarlton Limestone of Pennsylvanian to Permian age. is rock has been heavily altered and includes many folds, faults, thrust faults, localized marble, and sills and dikes of volcanic rock. Outcrops of Duchess of York Sandstone also occur across the karst (Forsythe and Mpodozis, 1983). is region of Chile is extremely wet with annual rainfall between 6 and 10 m and as few as three days per year with no rainfall. e region also routinely has very strong winds gales are common and local ports are oen closed due to the high winds. e islands and adjacent ords are entirely glaciated, although the specics of glacial activity and timing are unknown (Jalliet et al., 2008). International cave and karst researchers have travelled to the area on four major expeditions over the past 10 years. Based at the open-air limestone mine on Guarello Island, the caver teams have explored many caves and pits as well as the amazing surface karst of the islands. One of the most interesting discoveries in 2000 were rock comets small erratic boulders of volcanic rock or sandstone behind which a bedrock limestone tail or n had developed (Figs. 1 and 2) originally described in Jalliet and Hoblea (2000). In addition a groove of varying size, but generally a few centimeters deep, is eroded into the bedrock at the windward margin of each erratic. Comets occur in groups only at the most windy sites in the archipelago on at benches at the tops of passes and along narrow ridges and where erratics have also been deposited during glacial retreat (Fig. 2). A handful of comet sites have been recorded and most have only a few comets. e sites used in this study on Tarlton and Madre de Dios Islands contain the densest concentrations of comets known. Figure 1: Comets on Tarlton Island looking along the axis of the bedrock n. Joel Despain, Centre Terre.

PAGE 203

Earth Sciences 1464 2009 ICS P roceedings 15th International Congress of Speleology Previous work has shown that erosion rates of the Tarlton limestone are very high with a possible down cutting rate of 16 mm per century (Centre Terre, 2007). Winds on the islands are associated with frequent periods of heavy rain and the rain does literally move sideways across the passes and ridges during storms. Observations and photos of the comets during rainstorms revealed that the limestone ns at times remain dry while adjacent areas are exposed to rain and subsequent erosion (Fig. 3). 2. MethodsTo understand the relationship between the erratic boulders and their adjacent bedrock ns, we measured their respective sizes. Eighteen comets were measured along a ridge on the northwest corner of Tarlton Island and 32 comets at a single site were measured at Suplador Pass on Madre de Dios Island, for a total of 50. For each the maximum width, height, and thickness of each erratic, the length, and maximum width and height of each limestone n, and the depth and width of the bedrock groove on the windward side of the erratic were measured using a berglass measuring tape to an accuracy of 0.5 cm except for groove depths, which were measured to the nearest millimeter. In addition, the azimuth orientation of each n was recorded. Some comets at both sites were not appropriate for this study such as those formed by two erratics or those that merged into each other and, as such, were not included. Standard statistical analysis was applied to these data. 3. Results and DiscussionComets are oriented dierently at the two sites reecting a variation in the direction of the prevailing winds. On Tarlton Island comet azimuths vary from 245 to 297 with a range of 52 and a mean orientation of 273. At Suplador Pass, comet orientations vary from 305 to 334 for a range of 29 and a mean of 324.5. e comets on Tarlton Island are across a larger area compared to Suplador and are within several at exposures of limestone that are proximal but at dierent elevations up a ridgeline. With varied locations comes an increasing likelihood of localized variations in the wind direction and variation in comet orientation. Erratics vary in width from 11 to 105 cm while the bedrock ns vary from 7.5 93 cm. Forty ve of 50 erratics are wider than their adjacent bedrock n. Width variations between ns and erratics range from 0.5 19 cm with a mean of 4.5 cm. A linear polynomial regression of erratic width versus n width produces a t and r2 of 0.844 demonstrating a strong relationship between the width of erratic boulders and bedrock ns (Fig. 4). Erratics range in height from 11.5 to 105 cm. and bedrock ns vary in height from 4 to 42 cm. All of the erratics were taller than their proximal ns. e average dierence Figure 2: Comets at Suplador Pass on Madre de Dios Island that weremeasured for this study. Joel Despain, Centre Terre. Figure 3: Rock comet on Tarlton Island aer a light rain with strong winds and with a prominent dry area along a small limestone n in the lee of the erratic. Joel Despain, Centre Terre. Figure 4: Regression between the width of erratic boulders and their adjacent bedrock n. e t is an r2 of 0.844.

PAGE 204

15th International Congress of Speleology Earth Sciences 1465 2009 ICS Proceedings between erratic and comet height is 11.3 cm. with a range of 3 to 70 cm. Linear polynomial regressions between the height of the erratics and the ns produce a resulting r2 of 0.382 showing that erratic height is a poor predictor of comet height. Unlike the other dimensions, the length of the comet does not have a direct corollary in the size of the associated erratic. We compared erratic height and width and the average of erratic height and width (mean dimension) to n length. Both erratic width and height are poor predictors of comet length. e regression of erratic width versus n length produces an r2 of 0.322 and for erratic width the t is an r2 of 0.187. e mean of the two is no better with a t of 0.25 (Fig. 5). e grooves parallel to the front of the erratic boulders are variable. Fiy-eight percent of grooves are 0.5 6 cm wider than their adjacent erratic, but 25.5% are shorter and a few comets lack the feature entirely. A linear polynomial regression with erratic width versus groove width produces an r2 of 0.74 showing a relationship between the widths. Groove depths range from 0 7 cm with a mean of 1.37 cm. A regression between erratic height and groove depth produces a poor t with an r2 of 0.255. While comets themselves are uncommon, related landforms are throughout the islands. Dierential erosion along volcanic dikes and sandstone units at varying orientations and angles and vertical erosion below erratic boulders in areas of low wind have been widely noted (Centre Terre, 2007). Forty meters from the comets at Suplador Pass, volcanic rock 0.1 m thick and 0.7 m in diameter shielded a bedrock arm in its lee from erosion. e bedrock limestone arm is 2 m in length and projects upward and into the prevailing winds (Fig. 6). While the width of the erratics has a clear eect on the width of the limestone n and groove for each comet, only general predictions on other bedrock parameters can be made from an examination of the erratic. Small variations in the shape of the erratic and the adjacent ground surface, the proximity to adjacent comets, the shape and slope of the bedrock surface, and likely other factors that eect wind velocity and direction wield signicant and localized inuence on the size and shape of the bedrock ns, making these features highly variable.Acknowledgmentse authors acknowledge the contributions of the members of the Ultima Patagonia Expeditions 2000, 2006, and 2008ReferencesFORSYTHE, R., and C. MPODOZIS (1983) Geologia del Basamento pre-Jurasico Superior en El Archipielago Madre de Dios, Magallanes, Chile. Servicio Nacional de Geologia y Mineria, Chile, Boletin 39, 63 pp. JAILLET, STEPHANE and RICHARD MAIRE, FRANK BREHIER, JOEL DESPAIN, BENJAMEN LANS, LAURENT MOREL, JEAN-FRANCOIS PERNETTE, ESTELLE PLOYON, and BERNARD TOURTE (2008) Englacement, eustatisme et reajustements karstiques de la bordure sud de larchipel de Madre de Dios. Karstologia 51, 1. Figure 5: Mean erratic boulder size versus bedrock n length with a poor r2 of 0.25. Figure 6: Bedrock arm at Suplador Pass. Joel Despain, Centre Terre.

PAGE 205

Earth Sciences 1466 2009 ICS P roceedings 15th International Congress of Speleology JALLIET, STEPHANE and FABIEN HOBLEA (2000) Une Morphologie Originale Liee au Vent: Les Fusees ou Cretes Eoliennes de Lapiaz de Lile Madre de Dios: AA ctes de la 10 RRencontre d OOctobre, Paris, pp 73. Centre Terre (2007) RR ap port dexpedition UU ltima Patagonia 2006 et projet UU ltima Patagonia 2008. Centre Terre, Toulouse, France, 40 pp.

PAGE 206

15th International Congress of Speleology Earth Sciences 1467 2009 ICS Proceedings EX X TREME INCREASE OF CO2 IN BELGIAN CAV V ESCC Am M ILLE E E K1 and JEAN GODI GODI SSAR AR T2 1UU niversit de LL ige, Belgium, camille.ek@ulg.ac.be2UU nion belge de Splologie, Belgium, jean.godissart@skynet.be We began making CO2 measurements in Belgian caves in 1966. Analyses were conducted in cave halls, galleries, ssures and shas. ey were collected during various seasons and at dierent levels above the oors of the conduits, and in the absence or presence of other human beings. Our rst results were published in 1968. From that time on, we have carried out studies in Poland, uebec, China and other countries. However, we have focused most of our work in Belgium. We have discovered over these forty years of study, a strong increase in the observed values of CO2 in Belgian caves. For example, a few of our observations follow. Trou Joney (Comblain-au-Pont, province of Liege) is a small and shallow cave. We measured at the central point of the main gallery, 1870 ppm CO2 in July 1966, and 13,800 ppm at the same location in July 2007. At the bottom of the sha of the Comblain-au Pont cave, we measured 600 ppm in July 1966 and found 1500 ppm in July 2008. In La Merveilleuse cave (Dinant, province of Namur), we measured 800 ppm at the central point of the Big Hall in August 1990, rising to 1700 ppm in August 2008. In the main gallery of the Fontaine de Rivire cave (Hamoir, province of Liege), we found 5000 ppm in August 1972, increasing to 12,000 ppm in 1991. We conclude that the CO2 content of the atmospheres of many caves in Belgium (at least) has become higher e increase is very variable, but omnipresent. Are our measurements signicant? We believe that we have validated our instruments and our methods. e increase is probably not a result of local industrial activities. e CO2 curves of Mauna Loa Observatory (Hawaii) and Mace Head (Ireland) both show an increase of atmospheric carbon dioxide during the last half century. However, the upsurge of CO2 observed in the caves is proportionately much greater than the increase in those well-known surface measurements. ere is a very complex interrelationship between temperature, vegetation and biomass activity, and CO2 in the soil and underground. e increase of any one of these three parameters can induce changes in the two others, and hence in the partial pressure of CO2 in cave air. Ination du CO2 dans les grottes de Belgique. Nos premires mesures de dioxyde de carbone dans les grottes de Belgique datent de 1966 et les premiers rsultats furent publis en 1968. Nous avons mesur le CO2 dans les salles, les puits, les galeries et les ssures, en direntes saisons. Au l de ces quelque quarante annes, nous avons observ dans toutes les grottes une forte augmentation des concentrations de lair en dioxyde de carbone au cours du temps. Ainsi, par exemple, dans le fond du Trou Joney, une petite grotte peu profonde situe Comblain-auPont (province de Lige), nous avons mesur 1870 ppm de CO2 en juillet 1966 et 13800 ppm au mme endroit en juillet 2007. Dans la grotte La Merveilleuse Dinant (province de Namur), la teneur en CO2 dans la grande salle tait de 800 ppm en aot 1990 et en aot 2008 elle tait passe 1700 ppm. Dans la g rande galerie de la grotte de Fontaine de Rivre Hamoir  (province de Lige), nous avions trouv 5000 ppm en aot 1972 et nous en avons mesur 9400 en octobre 2008. Les teneurs en CO2, dans les grottes belges en tout cas, sont donc en forte hausse. Cette augmentation est trs variable mais elle est trs gnrale et, notre avis, elle nest pas inuence par lactivit industrielle locale. Certes, les courbes de lobservatoire de Mauna Loa et de Mace Head montrent une augmentation du CO2 dans latmosphre au cours du dernier demi-sicle, mais, dans les grottes tudies, nous sommes en prsence dun phnomne beaucoup plus important.

PAGE 207

Earth Sciences 1468 2009 ICS P roceedings 15th International Congress of Speleology 1. IntroductionOur CO2 measurements in cave air began in 1966. ey were conducted in galleries, chambers, shas, ssures; we analysed the air near the entrances and in remote parts, near the ceiling and near the oor. e rst publication of our results appeared in the NN a tional Speleological Society Bulletin (Delecour et al., 1968). It showed that there is much more carbon dioxide in cave air than in the open air, much more in ssures than in galleries or chambers, and more in remote parts than near the entrances. Other results were published the same year in the AA n nales de Splologie, showing that CO2 in caves mainly comes from the surface soil through ssures (Ek et al., 1968). In 1985, we described the seasonal rhythm of CO2 partial pressure with a summer maximum and a winter minimum, and the slow decline of carbon dioxide from the soil to lower and lower levels in the caves (Ek & Gewelt, 1985). Six Belgian caves have been selected here to display our results. All of them are located in the Paleozoic limestones south of Lige (Belgium), close to the 50th parallel North (Fig. 1). All caves exist in an oceanic temperate climate, with a mean temperature of about 10C and an average annual rainfall of 800 mm. ey are in a covered karst, under grasslands and woods, at elevations ranging between 80 and 240 m. e thickness of the roof between cave and surface ranges from 15 to 90 m. We have conducted analyses in several other countries (i.e., Canada, Poland, China), but those are not dealt with in this paper. Some of our measurements done between 1966 and 1990 were repeated in several caves in similar conditions between 2000 and 2009. is has allowed us to discover that carbon dioxide in these cave atmospheres has strongly increased in the recent decades, considerably more so than in the outer atmosphere.2. Instruments and Methodse rst measurements of CO2 were made in 1966 by C. Ek and his colleagues with an electrolytic eld device (Ek et al., 1968). e analysis was based upon the titration of the carbon dioxide in a known volume of air absorbed in a 0.1 N NaCl solution. e time necessary to neutralize the absorbed carbon dioxide is measured. e CO2 content of the sample is computed from the current (mA), the time (seconds), and the air sample volume (mL). e apparatus, packed in a wooden case, weighed 15 kg, and was thus relatively heavy; however it worked in caves for several years (Fig. 2). From 1981 onward, we used the gas pump Precision Detector (by Gastec) which is lighter and faster but working with single use detector tubes. In these tubes, carbon Les relations entre les paramtres climatiques, biologiques (tels la respiration de la biomasse) et le dioxyde de carbone dans les sols sont complexes et les variations de chacun de ces paramtres peuvent inuencer les deux autres et par l la pression partielle du CO2 dans les grottes. Figure 1: Location of the caves studied in Belgium. C: Comblain-au-Pont. D : Dinant. E : Esneux. H: Hamoir. R : Remouchamps. Figure 2: Electrolytic eld device for measuring atmospheric CO2, by Koepf (Ek et al., 1968).

PAGE 208

15th International Congress of Speleology Earth Sciences 1469 2009 ICS Proceedings dioxide reacts with hydrazine or, for higher concentrations, with potassium hydroxide; in both cases, the CO2 concentration is given by a direct reading on the graduated scale of the tube. e minimal precision guaranteed by the manufacturer is +/-25% but the eective precision is much better and the reproducibility is about +/-10%. In some cases, for example in conned places, the operator used a carbon dioxide absorbing mask in order to avoid human CO2 exhalation (Fig. 3). Current surveys (since 2008) are conducted using a X-am 7000, a portable gas measuring and monitoring instrument by Draeger equipped with an IR (infrared radiation) probe ranging from 300 ppm up to 50,000 ppm CO2 which has been the most suitable device for our purpose (Fig. 4). is equipment, weighing 1.6 kg (one tenth that of our rst device), is also able to record data from remote places thanks to its pump sampling function, thus avoiding human contamination. Before each working day, the device was calibrated in the open air against the known concentration of CO2. Whenever we changed measuring instruments we checked several times the concordance of the techniques.3. ResultsSix caves have been selected here to display our results (Fig. 1). All of them are located in the Paleozoic limestones south of Lige (Belgium). e carbon dioxide content of the air of all the caves displays seasonal uctuations, with a summer maximum and a winter minimum, as exemplied in Figure 5. Over the course of the 40 years covered by this study, the values of CO2, and particularly the summer maximums, have been on a noticeable increase. Here are a few examples of these rises:3.1. Fontaine de Rivre Cave in Hamoiris cave lies in Devonian limestones, in the province of Lige, some 40 km south of the city of Lige. Its entrance opens at 135 m a.s.l. It is about 1100 m long. e rock above the cave is about 85 m thick. It is a phreatic maze connected to a wide gallery ending in a large chamber (20,000 m3) with a lake (Godissart, 1994). Between 1972 and 2008, the summer concentration of CO2 increased from 5000 to 9000 ppm (Fig. 6.1).3.2. Trou JJ oney in Comblain-au-PontTrou Joney is a small cave, 30 km south of Lige, in the same province. It lies in Carboniferous limestone, at an elevation of 185 m. Its length is only 60 m and the rock above the cave is about 15 m thick. e single gallery is choked at the bottom with gravel and ne sediments. We carried out a carbon dioxide survey in July 1966, and a monthly survey all along 1978 (Ek, 1979). At the bottom of the cavity, the CO2 rose in summer from 1715 to 13,800 ppm between the rst and the last date (Fig. 6.2).3.3. Sainte-Anne Cave and Brialmont Caves in Esneuxese two caves are located 9 km south of Lige, in Devonian limestones. e entrance of Sainte-Anne is at an altitude of 85 m a.s.l. whereas Brialmont, which is situated above the rst one, opens at 180 m a.s.l. ey are probably Figure 3: Gastec pump detector giving a direct reading of CO2 concentration (Ek & Gewelt, 1985). Figure 4: X-am 7000 by Draeger, inared portable device. Figure 5: Seasonal variation in the CO2 concentration in the atmosphere of Comblain-au-Pont Cave.

PAGE 209

Earth Sciences 1470 2009 ICS P roceedings 15th International Congress of Speleology Figure 6: Carbon dioxide in cave air. 6.1. Rivre Cave, Hamoir. Sketch map and location of the principal point of CO2 measurements. 6.2. Trou Joney, Comblain-au-Pont. Long prole and comparison of CO2 values in summer 1978 and 2008 and in winter 1978 and 2009. 6.3. Brialmont Cave, Esneux. Long prole. Location of comparative CO2 measurements. Topographical survey by Naveau, S.S.D., 1971. 6.4. Ste-Anne Cave, Esneux. Long prole. CO2 concentrations in January 1983 and January 2008.

PAGE 210

15th International Congress of Speleology Earth Sciences 1471 2009 ICS Proceedings parts of the same system, but to date the connection has not been made by cavers. Both caves consist of subhorizontal galleries. Brialmont is two-tiered. Beneath it, Sainte-Anne is four-tiered. e galleries are spacious and connected by shas and ssures. e total development of SainteAnne is about 1500 m and an underground river ows in its two lower levels. Above it, Brialmont, with 180 m of development, is dry (Figures 6.3 and 6.4). It is in this cave system in 1984 that we discovered the slow diusion of carbon dioxide from the soil to the successive lower levels of the caves. We also established the strong annual rhythm of CO2 concentration (Ek & Gewelt, 1985). Returning to these caves a quarter of a century later, we discovered a strong rise in carbon dioxide. In August 2006, the galleries of Brialmont showed an increase of much more than 30% (Fig. 6.3). In January 2008, Sainte-Anne displayed values approximately double those of 1982 (Fig. 6.4). In these caves, although the measurements of 1982 were carried out each month, unfortunately we have only single values in 2006 and 2008.3.4. La Merveilleuse Cave in DinantLa Merveilleuse lies in the Carboniferous limestone, 75 km south of Brussels as the crow ies. It opens at about 150 m a.s.l. and its development measures about 750 m. We analyzed the CO2 there in 1990, 2003 and 2008. e values measured in the halls and galleries in August of each of these three years are displayed in gure 7.1. It is clear that between our rst and last measurements, the CO2 concentration in the cave air has approximately doubled, except at the aperture of two shas, where carbon dioxide variations are probably related to a deep underground stream (Fig. 7.1). In August 2003, we carried out a detailed survey of a small conned chamber: the Temple de Diane (Fig. 7.2 and 7.3). We measured 4000 ppm CO2 near the ceiling and 5100 near the oor. In August 2008, these gures had increased to 15,400 and 18,000 ppm respectively!4. ConclusionsWe observed in the caves of Belgium a huge rise in pCO2. is generally has been an increase of around 100% in forty Figure 7: La Merveilleuse Cave, Dinant. Carbon dioxide in air. 7.1. La Merveilleuse. Long prole. CO2 measurements in August 1990, 2003 and 2008. 7.2. Le Temple de Diane. Long prole. Comparison of some CO2 measurements in Aug. 2003 and Aug. 2008. 7.3. Le Temple de Diane. Long prole. Detailed CO2 survey in August 2008.

PAGE 211

Earth Sciences 1472 2009 ICS P roceedings 15th International Congress of Speleology years, and in some places even more. is increase is variable, as shown in Figure 8, but all of the caves visited showed some increase. Of course the pCO2 curve of Mauna Loa (Machta, 1972), from 1960 to 2008 displays a slight increase, but this is much less signicant: rising from 314 ppm in 1960 to 387 ppm in 2008, about 22% in a half-century. From 1960 until now, the rate of rise at Mauna Loa has increased from 0.9 ppm per year in the nineteen sixties to 2.0 ppm per year during the beginning of the third millennium. Closer to Belgium, Mace Head (Ireland) shows the same slope. e worldwide rise of atmospheric CO2, although sharing the same trend, is much too small to explain the inated values observed in the caves. Carbon dioxide in most caves mainly originates from the soil. is idea is supported by investigations by Bourges et al. (2001), Calmels et al. (2005) and Baldini and colleagues (2006). It is the emission of CO2 by the soil biomass which is the presumed cause of the increase observed in the caves. A slight rise of temperature has been observed in Belgium, which can favor increased activity by the vegetation. e year of 2006 was the warmest since the beginning of meteorological surveys (1833), and 2007 was warmer than 2006. On the other hand, the general rise of atmospheric pCO2 induces a rise of the rate of photosynthesis which leads, in turn, to an increase of root respiration and of microbial activity in the soil (Koerner et al., 2005). is could certainly enhance the carbon dioxide concentration in soil air, hence in the caves.AcknowledgementsWe are grateful to Dr. J.M. Bouquegneau, Dean of the Sciences Faculty of the University of Lige, who provided us, the two X-am 7000 infrared CO2-meters that we operated in the recent years, as well as with experienced advice, Ph. Labarbe helped with the illustrations: graphs and photographs. Hearty thanks to Myriam Herman, Peter Blackie and Amand Trine for their precious help in the improvement of the English text. ReferencesALDINI, J., L. BALDINI, F. McDERMOTT, and N. CLIPSON (2006) Carbon dioxide sources, sinks, and spatial variability in shallow temperate zone c aves  : Evidence from Ballynamintra Cave, Ireland. Journal of CC ave and Karst Studies, v. 68(1)  : 4. BOURGES, F., A. MANGIN, and D. DHULST (2001) Le gaz carbonique dans la dynamique de l atmosphre des cavits karstiques  : lexemple de laven dOrgnac (Ardche). CC o mptes rendus de lA A cadmie des Sciences de Paris, Sciences de la Terre, 333  :685. CALMELS, D., J. GAILLARDET, and L. FRANCOIS (2005) Modeling the impact of vegetation on carbonate weathering rates. GG eo physical RResearch A A bstracts, 8, 09690. SRef-ID  : 1607-8692/gra/ EGU06-A-09690. Figure 8: Changes in CO2 concentrations in some Belgian caves between 1966 and 2008. Below, Mauna Loa curve of CO2 evolution.

PAGE 212

15th International Congress of Speleology Earth Sciences 1473 2009 ICS Proceedings DELECOUR, F., C. EK, and F. WEISSEN (1968) An electrolytic eld device for the titration of CO2 in air. NN ational Speleological Society Bulletin, 30  : 131. EK, C., F. DELECOUR, and F. WEISSEN, (1968) Teneur en CO2 de lair de quelques grottes belges. Technique employe et premiers rsultats. AA n nales de Splologie, 23  : 243. EK, C. (1979) Les phnomnes karstiques dans les roches palozoques de la Belgique. Processus actuels, problmes actuels. AA n nales de la Socit gologique de Belgique, 102  : 13. EK, C. and M. GEWELT, (1985) Carbon dioxide in cave atmospheres. New results in Belgium and comparison with some other countries. EE art h Surface Processes and LL andforms, 10  : 173. GODISSART, J. (1994) Le cycle annuel des tempratures et du CO2 dans la grotte de Fontaine de Rivre Hamoir (Belgique). CC o mptes rendus du CC olloque international de Karstologie de 1992 LL uxembourg. Service gologique du Luxembourg  : 179. KOERNER, C., R. ANSSHOFF, O. BIGNUCOLO, S. HAETTENSCHWILER, S. G. KEEL, S. PELAEZ-RIEDL, S. PEPIN, R.T.W. SIEGWOLF, and G. ZOTZ (2005) Carbon Flux and Growth in Mature Deciduous Forest Trees Exposed to Elevated CO2. Science, 309  :1360. MACHTA, J. (1972) Mauna Loa and global trends in air quality. Bulletin of the AA merican Meteorological Society, 53  :402.

PAGE 213

Earth Sciences 1474 2009 ICS P roceedings 15th International Congress of Speleology OV V ERV V IEW OF THE NONKARST CAV V ES IN HUNGARY Istv ISTV N Es ES Zt T ERHs S1, G G EORGE SZENt T Es S 2 1Kztrsasg-t 157, 8045 II sztimr, HH ungary2 AA lte Frankfurter Str. 22 b, 61118 Bad Vilbel, GG ermany, szentesg@aol.com Hungary is a country in central Europe with a land area of 93,030 km2. e greater part of its territory consists of a plain and a low hilly landscape. A moderately high mountain range, with a highest point of 1014 m comprises a third of the country. ese mountains are mainly composed of limestone, other sedimentary rocks, and volcanic formations. In Hungary, 2800 karst caves and 892 non-karst caves have been recorded. e majority of the non-karst caves developed in basalt, andesite, rhyolite, and their tus, as well as in sandstone, conglomerate and a variety of metamorphic rocks. A signicant part of the karst limestone caves are large and well decorated. e length of 30 of these caves varies between 1 and 25 km. However the non-karst caves are shorter the longest being the 428 m long Csrg Hole they represent various types of cave development. e syngenetic caves in volcanic rocks include gas bubbles, steam explosion caves, fumarole caves, and crystal chambers. Typical syngenetic caves formed at the same time as the deposition of calc-tufas. e caves that originated postgenetically formed by four major eects: mass movement, erosion (physical weathering), fragmentation (physical weathering), and chemical weathering. ese four main processes have resulted in the development of about 25 dierent types of non-karst caves. e 892 non-karst caves are located in 20 regions. e number of caves varies greatly from region to region. ere is one area with nearly 300 caves identied, while in other regions only have one or two caves documented. Generally the Hungarian non-karst caves are poor in secondary mineral deposits and speleothems, although silicate speleothems and other mineral formations occur in some caves. Ice remains until the summer in some smaller caves. In the caves 198 species of fauna and 18 species of ora some very rare have been found and classied. In some non-karst caves, signicant archaeological and palaeontological remains have been discovered. 1. IntroductionIn Hungary, 892 natural non-karst caves have been listed. Organized research began in 1983 with the launch of the Volcanspeleological Collective. eir comprehensive activity is still ongoing. e present study summarizes the development, the number and location, the mineral occurrences, the climatological and biological observations i n the non-karst caves in Hungary (ESZTERH S 2005, ESZTERH S-SZENTES 2004, OZORAY 1952). 2. Geographical Sketch of HungaryHungary is situated in the center of Europe in what is known as the Carpathian Basin (BULLA 1964, JUH SZ 1987). e total surface area of Hungary is 93,000 km. Its largest river, the Danube, runs through the country. e main geographical units of Hungary: are the Great Hungarian Plain, the Little Alfld, the North Hungarian Mountains, the Transdanubian Mountains, the Transdanubian Hills and Island Mountains and the preAlps. e GG reat HH ungarian Plain lies at an average of 100 m a.s.l. and comprises of lowland which covers nearly the half of the country. Its surface is made up of alluvial at and low ridges covered with sand dunes. Structurally, the lowland represents a cauldron subsidence, which has been lled up with loose sediments. e Variscan crystalline basement and the Mesozoic sequence is overlain by 1,000,500 m thick Pannonian and uaternary sediments such as clay, clayey marl and sand. e LL i ttle AA lfld is a plain in the north-western part of the country. Its landscape has been moulded as a result of the inuence of the Danube River. e sunken fault-blocks of the basement are covered by Miocene limestone and 15002500 m thick Pannonian limnic sediment, above which has been deposited a 20 m thick sandy-pebbly sediment of the primal Danube River. e NN orth HH ungarian Mountains are a part of the inner range of the Carpathian Mountains. e main ridge is the highest summit in Hungary, the 1,014 m high Mount Kkes. Small basins and tectonic-erosion valleys form a strongly dissected landscape. e geological history can be traced back to the Carboniferous Age. In the Triassic thick limestone layers were deposited in the Bkk and Aggteleki Mountains.vvolcanic activity occurred, which has created the Brzsny, the Cserht, the Mtra and the

PAGE 214

15th International Congress of Speleology Earth Sciences 1475 2009 ICS Proceedings Tokaji Mountains. e basalt plateaux and cones of the Medves-Ajncski Mountains were formed at the end of the Pliocene. At the end of the Tertiary the mountains underwent a peneplanation, resulting in further upli. e Transdanubian Mountains are a 200 km long, 40 km wide SW-NE trending fault-block range mainly composed of Mesozoic sequences. e range includes the following Visegrdi, Budai, Pilis, Gerecse, Vrtes, Velencei and Bakony Mountains, which are divided by tectonic valleys and wider basins. In the Mesozoic formations the Velencei Mountains are composed of Carboniferous granite, the Visegrdi Mountains of Miocene andesite whilst a part of the Bakony Mountains are covered by Pliocene basalt ow. e T ransdanubian HH ills and II sland Mountains lie west of the Danube River. e crystalline basement and the Mesozoic sequence is overlain by Tertiary sediment several thousand metres thick. e most signicant are the Pannonian formation and the extensive loess cover. From this hilly landscape rise two distinctive mountains, the Mecsek and the Villnyi Mountains. Above the granite basement they are composed of Permian sandstone, Triassic and Jurassic limestone. e easternmost foothills of the Eastern Alps, the Soproni, Kszegi and the Vas Mountains form the pre-A A l ps on the western border of Hungary. ey are mainly composed of metamorphic sequences (Fig 1).3. e Number and Location of the Non-Karst CavesBefore beginning the discussion of the non-karst caves we should mention briey the karst caves of Hungary. Approximately 2800 karst caves are known in the karstic limestone ranges of the Bkk, Aggteleki, Bakony and the Budai Mountains. Some of them are extensive and spacious cave systems. In the Aggteleki Mountains the longest cave, the 25 km long B aradla-D Domica System, is to be found. is runs under the border into Slovakia. In the Bkk Mountains the deepest cave in Hungary is located, the 253 m deep I I s tvn-lpai CC ave. 4. Rsearch History of the Non-Karst Cavese rst written reference to a non-karst cave, the Likask near the village of Lovasberny, dates from 1295. e cave has developed in quartzite and it was mentioned in a charter as a border reference point. urz Gyrgy, Palatine of Hungary had visited the Pokol-lik near the village of Kapolcs in 1610. From the middle of the XVIIth century more and more descriptions and data appeared in the archives concerning non-karst caves. In 1869 the geologist Jzsef Szab led the rst scientic expedition to the Mtra Mountains in order to explore the Csrg Hole. It was many years late in the1930s, that various researchers again investigated non-karst caves. In the 1950s and 1960s explorations revived further. Organized research began in 1983 with the launch of the Volcanspeleological Collective. eir comprehensive activity is still ongoing. e organization, led by Istvn Eszterhs, consists of a nucleus of 15 persons, who are occasionally joined by several more cavers. ey have listed many new caves and in the last 25 years they have dug and discovered nearly 1000 m of new cave passages in 43 caves. ey have studied the development of the non-karstic caves, and have identied several new types of cave development, identied as the consequence caves, holes formed by alkaline solution and the fumarole cavities. ey have discovered, and described, some speleothems previously unknown in Hungary, such as silica stalactites and isingerit discs. ey have solved the problem of the reasons for ice development in low elevation basalt caves. ey have classied 198 species of animals and 18 species of fungi (some of them are unusual) which are to be found in the caves. e results have been summarized in seven separate volumes and in 160 articles mainly in Hungarian, but occasionally in German or in English (E SZTERH S 2003). In 2004 the digital list of the nonkarst caves in Hungary, was completed. is can be viewed on the website http://geogr.elte.hu/nonkarstic using either Netscape or Mozilla browser. e list is updated every year (E SZTERH S SZENTES 2004). 5. Genotypes of the Non-Karst Cavese development of the non-karst caves is a complex process in both space and in time. Several inuences, of varying importance, prevail, related to the formation of these caves. e most important of these inuences are described below (E SZTERH S 1993, OZORAY 1952, SZENTES 1971). e syngenetc cavities formed concurrently with the processes Figure 1: e main geographical units and the mountains in Hungary.

PAGE 215

Earth Sciences 1476 2009 ICS P roceedings 15th International Congress of Speleology of the rock formation. In Hungary only a few syngenetic cavities are known, because the youngest lava ow of the country is over one and a half million years old. us the cavities near to the surface, for instance the lava tubes have fallen victim to denudation. In some lava rock, steam explosion cavities can be found, for example in Explosion Cave near the village of Gdrs. In syngenetic caves in the magmatic rocks, gas bubbles occurred (Gyula Cave in the Mtra Mountains). Crystal caves, which have been opened by mine workings (e.g. Andsite Cavity of Rzsa Mine in the Brzsny Mountains) also formed. Additionally, in magmatic rocks, holes formed by alkaline solution can be found, as well as fumarole cavities and hollows (Kmori Fox Hole in the Brzsny Mountains). At the same time as the deposition of the calc-tufas, characteristic syngenetic caves formed (Anna Cave in the town of Miskolc). Most of the non-karstic caves in Hungary have postgenetic origin. ese caves were formed aer the development of their surrounding rock. e formation of these caves can be divided into four major categories namely mass moement, erosion ( physical weathering), agmentation ( physical weathering) and chemical weathering, representing about 25 dierent types of the cave development. e caves which originated through mass moement were formed as a result of the shiing of the rock mass. is type of tectonic cave falls into one of four groups, which are dependent on the rock structure and the tectonic features of the region, namely: caves which have developed perpendicular to the rim of the outcrop such as the Fissure Cave of Mount Ttika, caves which have formed parallel to the rim of the outcropping rock formation such as Pokol Hole near the village of Kapolcs, the combined ssure cave such as the Tzifs Cave in the Kszegi Mountains which developed along several vertical faults and caves which have formed along the bedding plane such the Lena Cave near the village of Bozsok. AA t ectonic caves which were formed by the equalization of tension as the rock mass moved down the slope. In the talus deposits of the basalt cones, typical extensional atectonic pseudocave such as the Vadln-lik in the Mount Kovcsi are to be found. e other type of the atectonic cave that such as the Csrg Hole in the Mtra Mountains. e development of this cave can be traced back to the continuous sliding of the rhyodacite tu and the consequent aggradation. As a natural cavity breaks down, realignment of the hollows forms a high level collapsed labyrinth (e.g. Basalt Cave near the village of Pula). e consequence caves are a particular example of caves which originated through collapse. e collapse of ceilings in articial cavities may form apparently natural holes in the higher elevations of a system (e.g. a Szilvs-ki Cave near the town of Salgtarjn). Rock dri has resulted in the formation of smaller cavities, resulting in the formation of the leaning pseudocaves (e.g.Asztag-ki Hole in the Mtra Mountains) and the talus caves (e.g. Csrg-pataki Pseudocave near the village of Mtraszentimre). e erosion (physical weathering) has created three dierent types of caves in Hungary. Channel erosion has formed the Szarvas-ki-klyuk in Tarnalelesz village. e Grgeteges Rock Shelter near the village of Domoszl was formed by the lateral erosion. Near the town of Btonyterenye the Macska Rock Shelter was shaped by evorsion in rhyodacite tu below a waterfall. Deation has not formed caves in Hungary, but the inuence of deation can be seen in several rock shelters (e.g. the Pesk Cave near the village of Tarnalelesz). A further cave developing process caused by physical weathering is agmentation, which has come as a result of the inuence of temperature and moisture variation. Examples of such caves are the cavities between basalt columns (e.g. Sziklakonyha near the village of Somlvsrhely) and woolsack caves which have developed in granite (e.g. Zsivny Cave near the village of Pkozd). Chemical weathering has resulted in the formation of dierent types of solution caves. Acidic solution forms cavities in the rock, which has a high lime content, such as in the Kurta-vlgyi Cave near village of Velem which has formed in calcareous phyllite. In silicate rocks, alkaline solution is able to dissolve cavities. In siliceous conglomerate Mkus Bcsi Cave in the Mtra Mountains is an example of this type of dissolved cave development. Alkaline solution has been partly responsible for the formation of the Aranyhz Hole a small cave in geysirit in the Tihanyi Peninsula. H H ydratation, hydrolysis, oxidation and biological weathering are not decisive in the development of non-karst caves in Hungary. ese eects have only minor secondary inuences in cave formation (e.g. Iker-k Cave near the village of Pkozd and the Lyukas-k-vlgyi Cave near the village of Ivd).6. e Most Important Non-Karst Caves Below we try to give a short introduction to some nonkarst cave in Hungary. ese caves are either the longest or are of particular genetic, mineralogical or archaeological importance. All of the non-karst caves can be found on the website http://geogr.elte.hu/nonkarstic.C C s rg HH ole, the longest non-karstic cave in Hungary in the Mtra Mountains. is cave has recently been extended to 428 m long and 29.6 m deep. e development of the

PAGE 216

15th International Congress of Speleology Earth Sciences 1477 2009 ICS Proceedings cave can be traced back to the continuous sliding of the rhyodacite tu boulders on a 20 slope and the consequent aggradation.. As a result of this, the cave is a complicated labyrinth, consisting of several chambers, passages of varying length and shas (Fig.2., Fig. 3). At the deepest level of the cave a perennial spring emerges and in the Bat Chamber an intermittent lake can sometimes be found. In the cave 18 species of animal have been identied. In the Lost Passage large colonies of Lesser Horseshoe bats roost throughout the winter. e Pulai Basalt C C ave opens in the Bakony Mountains. e cave is a 151 m long and 21 m deep break down cave. e 25-30 m thick basalt is underlain by soluble limestone. Along cracks in the basalt, seeping water has dissolved holes in the limestone, into which the basalt layer has broken. e cave is accessible through a narrow sha, which leads to a bigger chamber. From this chamber various small passages and shas open in dierent directions. e cave wall shows nicely the dierent basalt layers, which are the witness to the several thousand years of volcanic activity. In the cave a rare silicate mineral can be found, the disc-shaped isingerit. e 51 m long Pokol H H o le is also to be found in the basalt of the Bakony Mountains. e basalt is overlain by a loose sandstone layer and its rim has broken away. e basalt blocks have not slid down on the slope, but leant against the bedrock, forming a leaning pseudocave. In the sandstone, thick basalt blocks dam up the seeping water, which emerges in the cave as a spring. is spring feeds a small lake. e size of the lake varies, because aer the water reaches a certain level a siphon system drains it. e 32 m long NN a gy Srkny II ce CC ave can be found in the basalt of the Mount Szent Gyrgy in the Bakony Mountains. is cave is unique in that the annual average temperature is +10 C e entrance is 270 m a.s.l. e cave consist of a narrow labyrinth, which has developed in the debris of the collapsed basalt columns. Ice (ice coating, icicles) can be Figure 2: Survey of the Csrg Hole, the longest non-karst cave in Hungary (Mtra Mts.).

PAGE 217

Earth Sciences 1478 2009 ICS P roceedings 15th International Congress of Speleology observed only in summer. During this time the air ows over the stone slope outwards from the cave. Evaporation on the extensive surface of the moist basalt blocks causes so much heat extraction, that the surroundings are cooled to below the freezing point. e GG yula CC ave is a gas bubble cave in the andesite of the Matra Mountains. Erosion by the nearby stream opened the cave, which originally had an entrance about 7 m in diameter. Now only a quarter of the sphere shaped hole exists, the rest having been eroded away with the lower part being lled in by gravel. e excavations in the gravel found potsherds from the Bronze Age and the Middle Ages. is suggests that, the cave was a transient shelter for people in those times. e 24 m long F ul-hegyi Big CC ave has developed in hydroquartzitic rhyolite tu in the southern part of the Tokaji Mountains. e Miocene rhyolite tu was percolated by the alkaline, silicic water of the hot springs, which silicied the tu. In a later period, when the reaction of the water exceeded the pH 9, which has precipitated earlier, dissolved. is solution formed several cavities in the Mount Ful region. e excavations of the inll of the Big Cave revealed many potsherd and tools from the Neolithic times to the Middle Age. e Szils-ki CC ave opens in a basalt inselberg in the Medves-Ajncski Mountains. e 68 m long and 15 m deep cave is a so called consequence cave. In the early 20th Century, below the basalt, a 3 m thick coal seam was mined out. In May 1917 the mine collapsed. As a consequence of this collapse the 80 m thick basalt layer downfaulted about 1 m and a 300 m long 20 m deep crevice developed. e basalt blocks formed part of the ceiling of this ssure system and 30 known consequence caves were created. e largest of these is the Szilvs-ki Cave (Fig. 4) e Betyr CC ave in the Cserht Mountains is 87 m long, and is the longest sandstone cave in Hungary. e cave was formed in calcareous sandstone as a result of dierent types of cave development. It is has formed along a fault line. Along the fault, the seeping water has dissolved the limy cementing material and the sand has peeled o from the loose rock surface and accumulated in the lower level of the distended ssure. is iterative process has resulted in a gradually widening ssure cave lled with loose sand. Excavations in this sand have revealed a large number of Pleistocene mammal bones and potsherds dating from the Middle Ages to modern times. e 50 m long AA r any CC ave opens in the Tokaji Mountains in rhyolite tu. e way in which this partly collapsed horizontal cave has developed is complex. Originally it was formed along a ssure by an underground stream. e natural passage was widened in the 17th century for an ore prospecting mining tunnel. Later, the rst part of the Figure 3:.A very narrow passage in the Csrg Hole (Mtra Mts.). Figure 4: Large collapsed basalt boulders in the Szilski Cave (Medves-Ajncski Mts.).

PAGE 218

15th International Congress of Speleology Earth Sciences 1479 2009 ICS Proceedings tunnel was collapsed and was subsequently dug out again in 1993. In the cave there is a slow running knee deep stream. e roof is nicely decorated with a large number of 20 cm long silicate stalactites. e M kus Bcsi CC ave can be found in the Kszr Valley in the Mtra Mountains. is 14 m long, tubular shaped cave was formed in siliceous conglomerate by the alkaline solution. e alkaline solutions originated from the geysers, which were active above the cavities long ago. 7. Mineralogical, Climatological, Biological and Archaeological ObservationsGenerally, the non-karst caves in Hungary are poor in mineral content, but in some caves various mineral fo rmations occur. (ESZTERH S 2003). Calcite speleothems can be found in those caves, which are covered or surrounded by calcareous rock formation (e.g. Remete Cave near the village of Zalasznt, Explosion Cave near the village of Gdrs). More common are the siliceous speleothems. Large quartz, amethyst and celestine crystals are to be found in those caves, which were opened by deep ore mine operations. e roof of the Arany Cave near the village of Tllya is nicely decorated with a large number of 20 cm long silicate stalactites. ermal water has resulted in albite, anorthite and kaolinite covering the walls in the Fels Cave near the town of Srospatak and in the Baglyas Hole in the town of Salgtarjn. Siliceous pisolites are fairly common (e.g. Galrias Cave near the village of Hromhuta and the Smirgli Cave near the village of Regc). However, the large number of isingerite disks on the wall of the Basalt Cave near the village of Pula are a rarity. e climatic conditions in the relatively small non-karst caves in Hungary do not show signicant deviation from the surface climate. e exceptions are those caves, which have developed in porous or detritic rock formations. Here the evaporation on the large rock surface causes such signicant a heat extraction, that the surroundings are cooled below freezing point. Five small ice caves are known in Hungary. Each of these caves have developed in detritic volcanic rocks which open at a low elevations (270 m) (E SZTERH S 1999-a). One hundred ninety-eight species of animals and 18 species of fungi which were classied, which are to be found in the non-karst caves. e fauna and the ora do not show signicant dierences from those found in karstcaves, although variations in proportions found can be observed. For instance the proportions of the penicillin ora and the lepidoptera fauna are higher than in the karst-caves. In almost every caves springtails (CC o llembola) occur, as do rove beetles (Staphylinidae), humpbacked ies (Phoridae), spiders (AA raneidae) and butteries and moths (LL epidoptera). Furthermore some unique species were identied, such as for instant the NN i phargus hungaricus, the H H o loscotolemon jaquet, and the O Orobainosoma hungaricum (ESZTERH S 2003). e palaeontological ndings in the non-karst caves are poorer than in the karst-caves, due to the lack of the lime to preserve them. e most ancient nds were rodents (C C ric etus cricetus major, OOchotona pusilla) from the Riss-Wrm interglacial period from the Seybold Cave near the town of Kszeg. e most spectacular nd was a cave bear (UU r sus spelaeus) skeleton, from the Wrm period in the Betyr Cave near the shrine of Szentkt. AA r chaeological remains occurred in 20 non-karst caves. e oldest ndings, Neolithic potsherds and lithic tools (int obsidian), were excavated from the Nagy Cave near the village of Legyesbnye. e nds from the metal ages, from the ancient times and from the Middle Ages occur in many caves all over the country. Many artefacts and documentation prove that, the nonkarst caves and articial cavities were used for dwellings, resting-places, sanctuaies as economic or military objects (E SZTERH S 1999-b, 2003).ReferencesBULLA B. (1964): Magyarorszg termszeti fldrajza Tanknyvkiad, Budapest pp 1. E SZTERH S I. (1993): Genotypes of caves in volcanic rocks in Hungary Conference on the karst and research activities of educational and research institutions in Hungary, Jsvaf pp 81. E SZTERH S I. (1994): Lychnis a Vulknszpeleolgiai Kollektva kiadvnya, Kapolcs pp 1. E SZTERH S I. (1999-a): Eishhlen des gemssigten Grtels in Basalt Proceedings of the VIIth International Symposium for Pseudokarst, Arad-Moneasa p. 5-13 & Jahresbericht der Hhlenforschergruppe Rhein-Main (Jg. 20.), Frankfurt a. M. pp 107. E SZTERH S I. (1999-b): Legenden ber vulkanische Hhlen in Ungarn Slovensk kras XXXVII. Liptovsk Mikul pp 29. E SZTERH S I. (2003): A nemkarsztos barlangok kutatottsga Magyarorszgon Karsztfejlds VIII. ktet, Szombathely pp 347.

PAGE 219

Earth Sciences 1480 2009 ICS P roceedings 15th International Congress of Speleology ESZTERH S I. (2005): Magyarorszg nemkarsztos barlangjai kzirat a Vulknszpeleolgiai Kollektva vknyvben az MKBT s a BI adattrban, Budapest pp 162. E SZTERH S I. SZENTES Gy. (2004): Magyarorszg nemkarsztos barlangjainak katasztere http//: geogr.elte.hu/nonkarstic. JU H SZ (1987): vmillik emlkei Gondolat Kiad, Budapest pp 1. OZORAY Gy. (1952): e genesis of non-karstic natural cavities as elucidated by Hungarian examples Karszts Barlangkutats II. ktet, Budapest pp 127. SZENTES Gy. (1971): Caves formed in the volcanic rocks of Hungary Karszts Barlangkutats VI. ktet, Budapest pp 117.

PAGE 220

15th International Congress of Speleology Earth Sciences 1481 2009 ICS Proceedings FROM THE PLAINS OF ABRAHAM TO DODO CANYON: REMARKABLE DOLOMITE KARST IN PERMAFROST IN THE MACKENZIE MOUNTAINS, NORTHWEST TERRITORIES, CANADA.DEREDERE K FORD ORDE E meritus Professor of GG eography and EE arth Sciences McMaster UU niversity, HH amilton, ONON LL8S 4K1, CC anada Abstract Along a west-to-east transect across the Canyon Ranges at Lat. 66o N in the Mackenzie Mountains (Northwest Territories, Canada) there is a remarkable variety of karst landforms. e host rocks are Upper Proterozoic Devonian in age, consisting of two thick-to-massive platformal dolomite formations with salt and redbeds below and a dolomite-gypsum megabreccia above. Only the most easterly sector of the transect experienced any glacierisation (Laurentide Continental Icesheet), conditions in the mountainous interior being too arid. Modern ecoclimatic conditions range from polar desert at >1500 m above sea level in the west to taiga-boreal ecozone transitional at ~300 m asl in the east. Technically there is continuous permafrost >50 m deep along the transect but groundwater recharge is common wherever the geohydrologic conditions are favorable. Despite the cold and aridity, and the low solubility generally associated with dolomite in comparison with limestone, etc. karst groundwater circulation and landform development are widespread along the transect, including extensive solutional pavements, steephead valleys and cirques (recules), dolines, dry canyons, a large structural polje, fresh and salt springs, and a fantastical dissolution drape topography on the breccia. Protection of a representative corridor along the transect is proposed under the aegis of of the new UNESCO International Geopark program.

PAGE 221

Earth Sciences 1482 2009 ICS P roceedings 15th International Congress of Speleology NEW RESEARCH IN THE SOUTH NAHANNI KARST, MACKENZIE MOUNTAINS, NORTHWEST TERRITORIES, CANADADEREDERE K FORD ORD1 and STE E PHEN HEN WOR OR THING HING TON ON2.1Professor EE meritus of GG eography and GG eology, McMaster UU niversity,H H amilton, ONON LL8S 4K1, CC anada.2Worthington GG roundwater, 55 Mayfair AA ve., DD undas,ON ON LL9HH 3K9, CC anada. Abstract Karst developed in platformal limestones and dolomites of Devonian age in the Mackenzie Mountains appears to include some of the most rugged and distinctive topography known in any arctic or sub-arctic region. e karst straddles the basins of the South Nahanni and Ram rivers, centered at 61oo 45 N and 124o W. Elevations range from 300 m asl at the lowest springs (in a great antecedent canyon on the South Nahanni River) to 1800 m asl on sharply dissected plateaus to the NW. Mean annual temperatures range from -5 to -11o C or colder; permafrost is discontinuous at the lower elevations, becoming generally continuous above ~1000 m asl. Ancient karst is preserved at all elevations above 600 m asl in both limestone and dolomite. U series measurements of speleothems in relict caves in the 1970s established that most is greater or much greater than 350 ka in age. However, a proglacial lake dam burst of probable mid-Wisconsinan (Wrm) age has modied landforms in a central corridor into a composite of scablands and karst, with giant bogaz, shas and solution dolines, plus three small but beautiful poljes. Investigations were renewed in the summer of 2006 as part of a program to extend the area of the existing South Nahanni National Park Reserve. Using a helicopter it was possible to survey the highest and most remote karst in terrain that escaped any cover by owing glaciers because of its aridity under full glacial climatic conditions. e karst consists of plateau remnants that have been much reduced in extent by cli recession consequent upon canyon incision over deep permafrost. Nevertheless, there are many expanses of limestone pavement and some larger (>50 m diameter) dolines with subterranean drainage via taliks. e remnants convert into castellated tor topography where they become very narrow. Felsenmeer prevail on the less soluble dolomite plateaus but the more extensive remnants of them contain solution dolines that appear to be active. Several new karst springs with melt season discharges m3 s-1 were discovered. Dye traces of sinking streams determined that there was conduit groundwater ow through dolomites to the major springs at mean velocities >4000 m d-1, indicating that fully mature karst aquifers can be established and maintained in these adverse climatic conditions.

PAGE 222

15th International Congress of Speleology Earth Sciences 1483 2009 ICS Proceedings DNA Analysis NALYSIS of OF Fecal ECAL Bacteria ACTERIA to TO Augment UGMENT an AN Epikarst PIKARST Dye YE Trace RACE Study TUDY at AT Crumps RUMPS Ca A Ve E Kentucky ENTUCKY USARR ICk K FOw W LER1, BRIAN H H Am M1, C C HRIs S G G ROv V Es S1, and C C ARL BOLst ST ER2 1Western Kentucky UU niversity, 1906 CC ollege HH eights Bld. Bowling GG reen, KyY2UU nited States DDepartment of AA griculture AA gricultural RResearch Service 230 Bennett LL n. Bowling GG reen, Ky A rainfall simulation experiment was performed to investigate the transport behavior of fecal-derived bacteria through shallow karst soils and through the epikarst. e experiment was conducted at Cave Springs Cavern located south of Mammoth Cave National Park on the Sinkhole Plain of south central Kentucky. Using a rainfall simulator, water containing 514 ppm sulforhodamine B was applied at a rate of 6.6 cm/hr for 4 hrs to 150 kg cow manure spread over a 7.5 m2 plot on the surface. Water was then sampled from a waterfall within the cave predetermined to be hydrologically connected to the surface area where the manure was applied. Fecal and EE c oli MPN numbers were determined by standard culturing procedures and total DNA was extracted from each sample by direct lysis measured by uorometry. DNA was further analyzed by quantitative Real-Time PCR (qRT-PCR) with specic primers to specically amplify and quantify Bacteroides DNA (fecal-specic bacteria) and Eubacterial DNA (all bacteria) in the samples. Both methods show a bimodal distribution of fecal bacteria as it inltrated through the soil and epikarst. Fecal bacteria and DNA levels peaked in samples collected approximately 90 min. ahead of the tracer dye followed by a second peak of fecal bacteria and DNA which roughly corresponded to the dye peak. DNA analysis also revealed that a surge of non-fecal bacteria was carried along just ahead of the dye front. ese data suggest that a mobile population of non-fecal bacteria in the soil was carried along with the rain event, and that the fecal bacteria followed two routes of transport through the soil and epikarst; some fecal bacteria applied to the surface reached the waterfall quickly by way of preferential ow paths through the soil while other fecal bacteria inltrated through the soil matrix along with the dye front.1. IntroductionKarst aquifers are extremely vulnerable to microbial contamination due to thin soils, groundwater velocities that can exceed 200 meters per hour, and direct communication of groundwater with surface water. Microbial contamination of karst groundwater in agricultural areas from manure application is widespread and oen severe (Boyer and Pasquarell, 1999; Ryan and Meiman, 1996). An important component of many karst ow systems is the epikarst region. Epikarst is a specialized vadose zone with high capacity for storage and transport. Standard models of epikarst structure depict a soil horizon several meters deep with high CO2 concentrations due to microbial activity. e surface zone is inuenced by agricultural activities, acidic rainfall, wind, and weather. e lower zone is rich in dissolved CO2 causing large fractures to form at the bedrock interface resulting in aquifer recharge through ssures, oen emerging as shas in underlying cave systems (Bauer et al., 2005; Trek, 2007; Pronk et al., 2008). Because epikarst ow is concentrated into preferential ow paths to form a porous conduit network, water and contaminants can be rapidly transported downward through the active conduit network and into cave systems during storm events (Boyer et al., 2003; Groves et al., 2005; Pronk et al., 2008). Evaluating the impacts of epikarst ow and storage is critical for understanding the fate and transport of pathogens and indicator organisms through karst aquifers. Use of microorganisms as tracers has a long history that has become more sophisticated as our ability to detect, quantify, and discriminate them has advanced. As far back as the 1800s bacteria that produce red or yellow pigments were utilized, and later bacteriophages and inoculations of known bacteria were analyzed with classical micrbiological techniques. Today, culturing techniques for fecal coliforms and EE coli that use colorimetric and uorescent substrates to distinguish bacteria according to biochemical pathways are the standard techniques used to monitor presumptive fecal contamination and migration of pathogens in groundwater and for testing the purity of municipal drinking water (Harveya, 2006; Gppert and Goldscheider, 2007). Recent studies of bacterial transport through epikarst at karst springs according to particle size distributions substituted uorescent microspheres representing pathogens the size of bacteria (1 m) and CC r yptosporidium (5 m) to simulate transport and recoveries of bacterial cells (Pronk et al., 2007; Pronk et al., 2008). Modern techniques employing DNA analysis of bacteria with the polymerase chain reaction

PAGE 223

Earth Sciences 1484 2009 ICS P roceedings 15th International Congress of Speleology (PCR) have been used to demonstrate uctuations in epikarst microbial community composition in response to storm events with PCR followed by denaturing gradient gel electrophoresis and DNA sequencing (Pronk et al., 2009). e purpose of this study was to demonstrate the practical application of rapid and quantitative DNA analysis using quantitative Real-Time PCR (qRT-PCR) in parallel with conventional dye trace data to better understand transport of fecal bacteria and other constituents of cow manure through the epikarst. As part of the dye trace study reported in this volume by Ham et al., aliquots of the water discharged into the cave were assayed for total DNA concentration as well as specic bacterial DNA sequences using a quantitative and discriminatory assays for specic bacterial genes. Microbial source tracking studies have established that fecal bacteria of the genus Bacteroides are favored for fecal DNA analysis because they are the most abundant bacteria in the intestines of mammals. Another advantage is that dierent species and strains of Bacteroides tend to be associated with particular hosts therefore DNA analysis of Bacteroides can be used to identify the source from which the fecal contamination originated (Bernhard and Field, 2000; Field, 2002; Dick et al., 2005; Layton et al., 2006). Bacteroides are anaerobic and dicult to culture in the lab, while fecal coliforms and E E c oli are very easy to culture in the laboratory. Although fecal coliforms and EE c oli have been used as indicators of possible fecal contamination for over 100 years, the presence of Bacteroides in water is unequivocal evidence of fecal contamination. At temperatures typical of natural waters, Bacteroides persists in streams for four to ve days (Field, 2002). In our study, DNA sequences from fecal-specic bacteria (Bacteroides) correlate with conventional assays for fecal contamination plus they reveal additional information about epikarst bacterial transport: DNA analysis of Eubacteria in general suggests that a large component of non-fecal bacteria, probably resident in epikarst soil storage, was mobilized and discharged into the cave during the simulated storm event. Rapid parallel DNA analysis can augment the information retrieved during a conventional dye trace and contribute to a deeper understanding of epikarst bacterial transport.2. Experimental Procedures 2.1 Study Site Crumps Cave, A.K.A. Cave Spring Caverns, has been used as a eld site for research and education by Western Kentucky University (WKU) and the Homan Environmental Research Institute (HERI) to generate a large body of data including hydrological maps and chemical parameters during storm events (Groves et al 2005). e cave contains 2 km of large horizontal passage lying 25 m beneath mostly agricultural land with three percolation waterfalls 5 8 m in height numbered 1, 2, and 3 on the east side of a 160 m section of passage. Waterfall 1 is in a recessed dome 40 m from the entrance equipped with data loggers, a tipping bucket, and two ISCO fraction collectors. Sample collection bottles and caps were rinsed with 70% isopropanol and dried before transport and installation to minimize bacterial contamination. Samples were packed in ice while in the ISCO units and transferred on ice to WATERS Laboratory for analysis within 24 hours aer collection. Samples returned to the lab in these bottles were subdivided for chemical and biological analyze related to the dye trace. On the surface a small plot that was hydrologically mapped to Waterfall 1 was mowed and tted with a soil data logger and three lysimeters. At the time of the experiment land use around the plot was residential and immediately down slope from agricultural wheat production. Before application of manure to the surface, samples were taken every 30 min for 24 hrs during passive discharge at Waterfall 1 to establish background conditions. 2.2 Simulated storm event sample collectionAt the beginning of the experiment 150 kg of dairy cow manure was spread evenly over the 7.5 m2 site. A rainfall simulator consisting of an enclosed shower approximately 3 m tall by 2.5 m long by 3 m wide was placed over the plot. Previous comparisons of sulforhodamine B (SRB) with Fluorescein and Tinopal showed that SRB did not interfere with colorimetric or uorescent analysis. Water containing 514 ppm SRB was pumped at a constant rate of 6.6 cm/hr for 4 hrs (2000 L total volume). Cave samples were delivered from Waterfall 1 to ISCO fraction collectors using a ushing bucket, and fractions were collected at 15-minute intervals starting from the beginning of the simulated rainfall (t = 0). Background data was collected for 24 hours prior to the start of the experiment.2.3 Q uantication of fecal coliforms and EE coli Fecal coliforms and EE coli were measured with the Colilert method using uantitrays (IDEXX Corporation, Rockport, ME) to determine the most probable numbers (MPN) of viable cells per 100 ml (equivalent to cfu/100 ml). Trays were incubated at 44.5 C in order to inhibit total coliforms and limit growth to fecal coliforms and EE c oli. Tenfold dilutions were simultaneously analyzed when higher concentrations were expected.

PAGE 224

15th International Congress of Speleology Earth Sciences 1485 2009 ICS Proceedings 2.4 DNA extractions by direct lysisISCO fractions of 250 ml were centrifuged at 3000 x g for 10 min. to sediment all particulate matter including bacterial cells (Murray and Hampton, 1980). Supernatants were decanted and the pellets were resuspended in 30 ml (3 x 10 ml rinses) of 10 mM Tris-HCl, 1 mM EDTA pH 8.0 (TE buer) and transferred to 50 ml conical centrifuge tubes. Suspensions were centrifuged to recover particulate material and resuspended in 1.5 ml (3 x 0.5 ml rinses) of 0.1x TE buer and transferred to microcentrifuge tubes with volumetric graduations. ose suspensions were centrifuged at 10,000 x g for 5 min. and the nal volumes of the compacted pellets were noted aer decantation of the supernatant. Pellets were resuspended in 10 volumes (minimum 100 L) LyseNGo direct lysis solution (Pierce) and processed according to the manufacturers protocol to lyse cells and release a complex mixture of lysates containing environmental DNA. ese lysates were used to measure total DNA concentrations and for PCR amplications. 2.5 Measurement of total DNA concentrationsAliquots (10-20 uL) of lysates were assayed with the ubit uorimetric DNA quantication platform (Molecular Probes Division, Invitrogen Corporation, Eugene, Oregon) according to the manufacturers protocol. e ubit assay uses a selective DNA uorescent dye to quantify double stranded DNA with high sensitivity and specicity even in complex mixtures (http://probes.invitrogen.com/media/ pis/mp32851.pdf). Total DNA concentrations represent the bulk DNA from all cellular debris eluted through the epikarst, including fecal material applied to the surface as well as soil and interstitial organisms mobilized by the applied hydraulic gradient. 2.6 Q uantitative Real-Time PCR (qRT-PCR) analysisA Bio-Rad iCycler system was used to carry out thermocycling while monitoring SYBR Green uorescence at 490 nm to measure the increase in uorescence in proportion to a logarithmic increase in the concentration of specic double stranded fragments of Bacteroides 16S SSU-DNA (Malinen, et al., 2003; Ponchel et al., 2003). Fecal-specic Bacteroides 16S SSU-DNA was amplied from lysates with primers Bac-32F and Bac-708R as described by Bernhard and Field (2000). Similar conditions were used to amplify Eubacterial 16S SSU-DNA with universal Eubacterial primers 27F and 1492R (Fowler et al., 2003). Finished qRT-PCR products were characterized by melting point analysis and agarose gel electrophoresis. Standard curves were for quantication of Bacteroides or Eubacterial 16S SSU-DNA were generated with p-GEM plasmids carrying a copy of the homologous fragment from B. thetaiotaomicron or Eubacterial cave clone BCTP21 (Acidobacteria), respectively (both plasmids were constructed in this laboratory and conrmed by DNA sequencing). reshold uorescence was set at 100 relative uorescent units (RFU) above background, and the number of cycles required to reach 100 RFU (threshold cycle, Ct) was plotted against the log of the starting quantity (SQ) of DNA added to the PCR reaction. is relationship allowed calculation of the concentration of a specic sequence of 16S SSU-DNA (Bacteroides, Eubacteria) expressed as the number of 16S SSU-DNA gene copies (DNA molecules) per L of original sample. 2.7 Melting point analysisA cycle was added to the qRT-PCR thermocycling protocol to measure the disassociation temperature, or melting point, of the double stranded DNA fragments ~700 bp in length generated during q-RT-PCR. Temperature was increased from 65 C to 95 C in 0.1 C increments while emission at 490 nm was averaged over 10 sec. at each temperature. 3. Results and DiscussionA period of roughly two hours elapsed between the start of the simulated rainfall and the visible presence of SRB in the waterfall. e dye concentration increased over a period of two hours to a maximum in the 4.0 hr sample and steadily decreased over the following two hours to SRB concentrations comparable to those at the dye breakthrough. Monitoring of Waterfall 1 for fecal coliforms, E E c oli, and physicochemical parameters continued for two weeks (Ham et al., this volume). e SRB curve represents the behavior of solutes in the inltrated water as the rain event displaced epikarst storage water. Fecal coliforms (Fig. 1A) and EE c oli (Fig. 1B) were quantied using Idexx Colilert assays. Both graphs show two peaks in bacterial concentration representing two separate transport routes through the epikarst. A large proportion of the bacterial contamination on the surface was discharged at Waterfall 1 ahead of the inltrated water (2.25.0 hrs, lled arrows) while another large increase in fecal bacterial contamination was discharged along with the SRB (3.75.0 hrs, open arrows). Lower concentrations of fecal contamination were observed as the SRB concentration decreased. For extraction of total DNA, fractions collected from Waterfall 1were centrifuged at 3000 x g for 10 min, sucient to precipitate suspended material including bacteria and other colloids (Murray and Hampton, 1980).

PAGE 225

Earth Sciences 1486 2009 ICS P roceedings 15th International Congress of Speleology Figure 1. Graphs showing concentrations of fecal coliforms, E. coli, total DNA, Bacteroides DNA, and Eubacterial DNA in 15 min actions taken over a sixhour period during simulated storm conditions inside Crump Cave at Waterfall 1. Simulated rainfall containing 514 ppm sulforhodamine B (SRB) dye was applied at a rate of 6.6 cm/hr for 4 hrs to 150 kg cow manure spread over a 7.5 m2 plot on the surface aboe Crump Cave. SRB dye was visible in Waterfall 1 actions beginning two hours aer the start of the experiment. All methods agree that during storm conditions, surface application of cow manure results in rapid transport of fecal contamination to karst aquifers through two transport routes resulting in appearance of bacteria before (lled arrows) and during (open arrows) emergence of the tracer dye. DNA analysis also suggests that a large displacement of non-fecal soil bacteria accompanies the storm event. SRB () is shown on each graph for reference. A. Fecal coliform concentrations() measured using the Idexx Colilert assay show an early peak (2.25-2.50 hrs) followed by a later peak (3.75-4.0 hrs). A later peak of fecal coliforms (5.0 hrs) was also seen. B. E. coli concentrations () measured using the Idexx Colilert assay show an early peak (2.25.5 hrs) followed by a later peak (3.75 hrs). A later peak of E. coli (5.0 hrs) was also seen. C. Total DNA concentrations () measured by the Qubit uorimetric DNA assay show two early peaks (2.25-2.5 hrs, 3.25 hrs) followed by two later peaks (3.75.0 hrs, 4.75 hrs). e higher peaks in DNA concentration late in each phase (3.25 hrs, 4.75 hrs) do not correlate with any of the bacterial data. ose peaks presumably represent DNA contributed by other biological constituents of manure. D. Bacteroides 16S SSU-DNA concentrations () measured by qRT-PCR show an early peak (2.75 hrs) followed by a later peak (3.75 hrs). Later peaks in Bacteroides DNA concentration (4.5, 5.0, and 5.75 hrs) were also seen. E. Eubacterial 16S SSU-DNA concentrations () measured by qRT-PCR show an intense peak (3.5 hrs) early during the peak concentrations of SRB. is assay quanties any kind of bacterial 16S-SSU-DNA om fecal material as well as soil and epikarst storage water. Due to the large magnitude of the Eubacterial peak (3.5 hrs) and the fact that it does not correlate with the other indicators of fecal bacteria, we hypothesize that the peak represents mobilized indigenous soil bacteria displaced om storage in the epikarst zone by the simulated rain event.

PAGE 226

15th International Congress of Speleology Earth Sciences 1487 2009 ICS Proceedings Total environmental DNA from all particles suspended in the fractions (manure, soil organisms, colloids, sediments) was extracted from the precipitated material by a direct lysis procedure and quantied by a sensitive uorimetric assay for double stranded DNA. Similar to the results using standard culture assays, peak concentrations of total DNA (Figure 1C) were discharged ahead of the inltrated water (2.25.5 hrs, 3.25 hrs; lled arrows) and later concurrently with the SRB (3.75.0 hrs, 4.75 hrs; open arrows). e total DNA peaks at 3.25 hrs and 4.75 hrs do not correlate with any of the other four bacterial indicators plotted in Figure 1. ese results indicate that manure contaminants containing DNA in the fractions taken at 3.25 and 4.75 hrs also show two peaks representing two routes of transport. Manure constituents contributing this DNA discharge later than fecal bacteria thus they are inhibited during transport relative colloidal particles the size of bacteria (~1 m). Aliquots of DNA lysates were used as templates in qRTPCR reactions with primers specic for bacterial targets of interest in order to quantify the contribution of the target organism(s) to the total DNA. e concentration of fecal-specic target DNA is expressed as the number of Bacteroides 16S-SSU-DNA gene copies per liter. Bacteroides 16S SSU-DNA concentrations measured by qRT-PCR (Figure 1D) show an early peak (2.75 hrs) followed by a later peak (3.75 hrs). Later peaks in Bacteroides DNA concentration (4.5, 5.0, and 5.75 hrs) were also seen, in agreement with the fecal coliform and EE c oli Colilert assays above (Figures 1A, 1B). Note the highest numbers of Bacteroides are in the range of 104 copies per liter. DNA lysates were analyzed in another qRT-PCR reaction using universal primers known to amplify 16S SSU-DNA from Eubacteria (the subkingdom of true bacteria, excluding subkingdom Archaebacteria) to estimate the number of all bacterial cells in each fraction, represented by the number of Eubacterial 16S SSU-DNA genes per liter (Figure 1E). Eubacterial 16S SSU-DNA concentrations measured by qRT-PCR show an intense peak (3.5 hrs) early during the maximum concentrations of SRB and represent any kind of bacterial 16S SSU-DNA from fecal material as well as soil and epikarst storage water. Note the highest numbers of Eubacteria are in the range of 1010 copies per liter. Due to the large magnitude of the Eubacterial peak (3.5 hrs) and the fact that it does not correlate with the other indicators of fecal bacteria, we hypothesize that the peak represents mobilized indigenous soil bacteria displaced from storage in the epikarst zone by the simulated storm event. e dye trace will be repeated without the application of manure to test this hypothesis. 4. Conclusions A rapid DNA analysis technique was used to quantify specic bacteria in up to 72 samples simultaneously. DNA data agree with traditional culturing techniques for monitoring fecal coliforms and EE c oli. Collecting DNA data as part of a dye trace reveals information about bacterial transport that otherwise would not be apparent. Samples with peak concentrations may be analyzed in detail aer qRT-PCR to reveal detailed information by cloning and sequencing or other molecular techniques if warranted. e bimodal behavior of bacterial transport from the surface during high ow conditions suggests that a large proportion of fecal bacteria are transported rapidly through larger ssures and fractures due to size exclusion processes. ese bacterial cells interact very little with soil particles or other materials lining the porous network in the epikarst soil and become mixed with water stored in large ssures below the soil horizon. Other fecal bacteria are transported through the epikarst along with the inltrated water marked by SRB. ese cells could be retained in the porous network through interactions between the cells and soil constituents through electrostatic forces or structural features lining the pores. In addition to the fecal bacteria applied to the soil surface, high ow conditions elute a large number of nonspecic bacterial cells as a single peak along with the dye front. ese cells do not enter the fast transport route thus suggesting they were resident in the epikarst soil prior to the high ow event. Furthermore the quantitative signal is six orders of magnitude greater than that produced by Bacteroides suggesting the inltrated water mobilized those bacterial cells from storage in the large mass of epikarst soil between the surface and the bedrock interface. Soil is known to harbor a dense and complex population of bacteria responsible for nitrogen xation, bioremediation, and other processes important to agriculture.AcknowledgmentsWe thank Stacy Antle, Priscilla Baker, Benjamin Fowler, and Mark Tracy for valuable technical assistance. We also thank Dr. Kim Cook and John Sorrell for assistance with construction of the B. thetaiotaomicron plasmid.ReferencesBaAUeER, S., RR. LIeEDlL, and M. Sa AUteTER (2005) Water Resources Research 41,

PAGE 227

Earth Sciences 1488 2009 ICS P roceedings 15th International Congress of Speleology Bernhard ERNHARD A.E. and K.G. Field IELD (2000) A PCR Assay to discriminate human and ruminant feces on the basis of Bacteroides-Prevotella genes encoding 16S rRNA. AA pplied and EE nironmental Microbiology 66, 4571. Boyer O YER D.G. and G.C. Pasquarell ASUARELL (1999) Agricultural land use impacts on bacterial water quality in a karst groundwater aquifer. Journal of the A A m erican Water RResources AA ssociation, 35, 291. Boyer O YER D.G. and E. Kuczynska U CZYNSKA (2003) Storm and seasonal distributions of fecal coliforms and Cryptosporidium in a spring. Journal of the A A m erican Water RResources AA ssociation, December 2003, 1449. Brahana R AHANA J.V., T.E. Ting ING M. Al L Qinna INNA J.F. Murdoch U RDOCH R.K. Davis AVIS J. Lozef OZEF J. Killingbeck I LLINGBECK E. Szilvagyi ZILVAGYI M. Doheney OHENEY Skubik KUBIK I. Chaubey HAUBEY P.D. Hayes AYES and G. Thoma H OMA (2005) uantication of hydrologic budget parameters for the vadose zone and epikarst in mantled karst. P roceedings of the U U .S.G G eological Survey Karst I I nterest G G roup, Rapid City, SD, Scientic Investigations Report 2005-5150, 144152. Dick I CK L.K., Bernhard ERNHARD A.E., Brodeur RODEUR T.J., Santo A NTO Domingo OMINGO J.W., Simpson IMPSON J.M., Walters A LTERS S.P., and K.G. Field I ELD (2005) Host distributions of uncultivated fecal Bacteroides bacteria reveal genetic markers for fecal source identication. AA p plied and EE nironmental Microbiology 71, 3184. Field I ELD K. G. (2002) Fecal source tracking with Bacteroides. P roceedings of the UU .S. EE PA A Workshop on Microbial Source Tracking, Irvine, CA. http://oregonstate. edu/gradwater/sites/default/les/bio/eld.pdf Fode O DE Vaughn AUGHN K., Wimpee IMPEE C.F., Remsen EMSEN C.C., and M.L. Perille ERILLE Collins OLLINS (2001) Detection of environmental samples by direct PCR without DNA extraction. Biotechniques 31, 598. Fowler O WLER R., Roberson OBERSON E., and S. Sahi AHI uantitative Real-Time PCR assays of bacterial DNA in sediments of the Flint-Mammoth Cave system with evidence for NN i trospira Spp. at sites undergoing limestone dissolution and karst aquifer evolution. P roceedings of the NN ational CC ave and Karst Symposium, Gainesville, FL, October 2003:45. www.nckms.org/2003/pdf/fowler.pdf Groves R OVES C., C. Bolster OLSTER and J. Meiman EIMAN (2005) Spatial and temporal variations in epikarst storage and ow in South Central Kentuckys Pennyroyal Plateau Sinkhole Plain. Proceedings of t he UU .S.G G eological Survey Karst II nterest GG roup, R R apid CC ity, SD D, September 12-15 2005. Scientic Investigations Report 2005-5160, pp 64. Harveya A RVEYA R.W. (2006), Microorganisms as tracters in groundwater injection and recovery experiments: a review, FE E MS Microbiology RReviews, 20:461-472. Layton A YTON A., L. Mc C Kay AY D. Williams ILLIAMS V. Garrett ARRETT R. Gentry ENTRY and G. Sayler AYLER (2006) Development of a 16S rRNA gene TaqMan-based Real-Time PCR assay for estimation of total, human, and bovine fecal pollution in water. AA p plied and E E nironmental Microbiology 72:6, 4214. Malinen A LINEN E., A. Kassinen ASSINEN T. Rinttila INTTILA and A. Pal A L VA (2003), Comparison of real-time PCR with SYBR Green I or 5-nuclease assays and dot-blot hybridization with rDNA-targeted oligonucleotide probes in quantication of selected faecal bacteria, Microbiology 149:269 Murray U RRAY P.R. and C.M. Hampton AMPTON (1980) Recovery of pathogenic bacteria from cerebrospinal uid. J ournal of CC linical Microbiology 12:4, 554. Ponchel O NCHEL F., C. Toomes OOMES K. Bransfield RANSFIELD F.T. Leong EONG S.H. Douglas OUGLAS S.L. Field IELD S.M. Bell ELL V. Combaret OMBARET A. Puisieux UISIEUX A.J. Mighell IGHELL P.A. Robinson OBINSON C.F. Inglehearn NGLEHEARN J.D. Isaacs SAACS and A.F. Markham A RKHAM Real-time PCR based on SYBR-Green I uorescence: An alternative to the TaqMan assay for a relative quantication of gene rearrangements, gene amplications and micro gene deletions. BMC C Biotechnology, 3:18 (13 pp). http:// www.biomedcentral.com/1472/3/18 Pronk R ONK M., Goldscheider OLDSCHEIDER N. and J. Zopfi OPFI (2007) Particle-size distribution as indicator for fecal bacteria contamination of drinking water from karst springs, EE n iron. Sci. Technol., 41:24, pp 8400. http://pubs.acs.org/doi/pdf/10.1021/ es071976f

PAGE 228

15th International Congress of Speleology Earth Sciences 1489 2009 ICS Proceedings Pronk RONK M., N. Goldscheider OLDSCHEIDER and J. Zopfi OPFI (2009) Microbial communities in karst groundwater and their potential use for biomonitoring, HH yd rology Journal online, http://www.springerlink.com/ content/87747286v075587j/fulltext.pdf Ryan Y AN M. and J. Meiman E IMAN (1996) An examination of short-term variations in water quality at a karst spring in Kentucky. GG r ound Water 34, 23. Trrek R EK B. (2007), How can the epikarst zone inuence the karst aquifer hydraulic behavior? EE n ironmental G G eololgy 51:761.

PAGE 229

Earth Sciences 1490 2009 ICS P roceedings 15th International Congress of Speleology Dating ATING a A biblical BIBLICAL lady LADY : an AN unroofed UNROOFED SALT ca CA Ve E gi G I Ves ES birth BIRTH to TO Lot OT s S Wife IFEAmAM Os S FRUmk MK IND Department of GG eography, e HH ebrew UU niversity of Jerusalem, 91905, II srael msamos@mscc.huji.ac.il Trying to explain the origin of peculiar geological features has led ancient societies to develop narratives of mythical transformations. Dating geological features supposed to be related to ancient narratives has been virtually impossible until now, because of either the lack of identied, dateable eld evidence, or the doubted historicity of the narrative. For at least two thousand years historians and travellers attributed a relict salt pillar in the Dead Sea area to the biblical myth of Lots Wife, who became a pillar of salt (Genesis 19:26). e formation of the salt pillar and associated cave is analysed and radiocarbon dated, complemented by measurement of its rising rate. Contrary to earlier assumptions that the salt pillar was formed by direct rainfall, the current results attribute the salt pillar to the remains of a wall of a karstic salt cave that collapsed during the period commonly proposed for the Sodom and Gomorrah upheaval. e new evidences suggest that the myth of Lots Wife was originally based on one of the earliest geological observations: e sudden appearance of a salt pillar following a catastrophic earthquake which unroofed the cave. e continuous observation of the pillar by travellers must have been coupled with the formulation of the myth which was later documented in the book of Genesis. is is the rst eld evidence for the origin of such an ancient biblical narrative. 1. Geologic SettingMt. Sedom is the exposed top of an actively rising salt diapir (Zak AK 1967) at the southwestern shore of the Dead Sea (Fig. 1). Mt. Sedom comprises vertical layers of latest Miocene to Pliocene(?) salt, covered unconformably by residual caprock, composed mainly of anhydrite. e diapir has been rising throughout the uaternary along its marginal faults (Weinberger E INBERGER et al., 2006). Measured mean rising rate for the diapir has been 6 mm/yr during the Holocene, and 6.9+0.3 mm/yr for southern Mt. Sedom during the previous decade (Pe E eri E RI et al., 2004; Weinberger EINBERGER et al., 2006). is indicates that the present rising rate has persisted for at least thousands of years. e rising is driven by tectonics and buoyancy of low density salt, cutting through the overburden of denser sediments (Zak A K 1967). During the Holocene, salt dissolution has been concentrated mainly where ephemeral streams carrying runo water sink into the extremely soluble salt (Frumkin R UMKIN 2000), forming (ephemeral) river caves. ese underground voids usually comprise (1) vertical shas close to the streamsink; (2) active sub-horizontal channels, oen discharging at the foot of the marginal escarpment (Fig. 2); and (3) high, dry cave passages hanging above the active channel (Fig. 3). e oldest, uppermost cave level relates to the initial exposure of salt above base level. As the diapir rises, caves are downcutting, adjusting rapidly to base level changes (Frumkin RUMKIN and AND Ford ORD 1995). Subhorizontal dissolution notches are common features in Mt. Sedom caves, formed where owing aggressive water dissolves the salt wall (in addition to precisely horizontal notches formed by standing water). Low-gradient cave passages accumulate alluvium and some dateable plant remains, derived from the surface of Mount Sedom.2. Biblical EventsSeveral geologic events, some associated with caves, are described or alluded to in the Bible (Bentor E NTOR 1989). One of the earliest, the biblical story of Sodom and Gomorrah, Figure 1: (a) Location of Mt. Sedom salt diapir; (b) Plan of the study site, showing SSP as a relict between the unroofed cave and the marginal fault. Figure 3 section line is marked.

PAGE 230

15th International Congress of Speleology Earth Sciences 1491 2009 ICS Proceedings is argued by geologists to reect a natural calamity that took place in early historical times (Ben E N Menahem ENAHEM 1991; Bentor ENTOR 1989; Blanckenhorn LANCKENHORN 1898; Block LOCK 1975; Harris ARRIS and AND Beardow EARDOW 1995; Issar SSAR 1990; Neev EEV and AND Emery MERY 1995; Trifonov RIFONOV 2007). However, no clearly dated geologic evidence has been presented. Within the general framework of the Sodom and Gomorrah narrative, the myth that Lots wife became a pillar of salt is unique in several ways: (1) it is a very early description of both the morphology and lithology of a geologic feature (a pillar composed of salt); (2) if such a feature really exists, it can be located geographically, because the scene must be at the vicinity of the Dead Sea, and there is only one salt outcrop in the Dead Sea region, namely Mt. Sedom salt diapir. Mt. Sedom (Hebrew name), preserving the biblical name Sodom (Greek Sodoma; Arabic Usdum) is known near the Dead Sea at least since the 2nd century CE (e.g. Claudius Galenus, quoted and translated by Reinach E INACH (1895)). (3) Within the southeastern face of Mt. Sedom, a distinct salt pillar (termed here Sedom Salt Pillar SSP, Figs. 1-4) resembling a human statue in prole (Fig. 2d), has traditionally been called Lots Wife. A mid-19th Century pioneering expedition to the Dead Sea initiated a new era of realistic geologic observations: we found the pillar to be of solid salt... Its peculiar shape is doubtless attributable to the action of winter rains (Lynch YNCH 1849). SSP has been attributed to erosion by rainfall until recently (Bentor E NTOR 1989). Here the nature of SSP is discussed, following a wider analysis of this salt pillar published in the geologic literature Frumkin RUMKIN 2009).3. Results and Discussione following observations have been made during the present study: SSP is a 20 m high, 5-10 m wide pillar, standing within the upper part of the 87 m high eastern escarpment of Mt. Sedom. e upper third of SSP is composed of anhydrite caprock (Fig. 3); the lower two thirds comprise vertical salt layers. e caprock head of SSP Figure 2: (a) View of thestudy area om the east with the active outlet (AC) of Sedom Cave at the foot of Mt. Sedom below SSP and the upper cave (UC). Note people right of the cave entrance for scale; (b) a notche in salt wall of the active Sedom Cave; (c) paleo-nothches (N) in the salt wall of the unroofed upper cave. SSP is in the right; (d) SSP and its eastern walla marginal fault, as seen om the north, with SSP resembling a human statue. e original level of SSP, 4000 yr ago is indicated to emphasize the rising of Mt. Sedom. No horizontal moement is implied. Figure 3: Vertical section of the study site. e notches to the le of SSP correspond to the remains of the unroofed paleo-cave whose upstream uncollapsed northern branch was radiocarbon dated using deposited wood remains (Figure 5). Ephemeral runo enters the cave today through the Sedom Sha owing along the active channel to the Dead Sea base level. Leveling benchmarks and their numbers are shown in the active channel of Sedom Cave. eir measured rising rates (in mm/yr, relative to benchmark 1, are shown at the bottom, presented also by black bars. Blocks of rock, detached om the cave roof (shown schematically), litter both the active and unroofed levels of the cave. Altitude is given in m below Mediterranean Sea level. Figure 4: (a) the unroofed cave and SSP as seen om the west. Note paleo-notches on the right wall; (b) the active 70 m high Sedom Sha in rocksalt.

PAGE 231

Earth Sciences 1492 2009 ICS P roceedings 15th International Congress of Speleology protects it from being rapidly dissolved by rainfall. e eastern wall of SSP is an almost vertical bedding plane acting as a major marginal fault along the eastern border of the diapir (Figs. 1, 3). SSP is separated from the northern extension of the escarpment by a deep chasm. In the present study, subhorizontal paleo-notches were found in the walls of the chasm (Figs. 1, 3), similar to those observed in the active Sedom Cave passage. Collapsed boulders are piled on the bottom of the chasm. e base of SSP and its associated notches are hanging ~60 m above the associated active level of Sedom Cave. Closely associated with SSP, the 1799 m long Sedom Cave is one of the largest known salt caves (Figs. 2a, 3). On the western side of SSP is a continuation of the chasm, forming a 25-m deep rooess depression (Figs. 1, 3). Its south-western part is the active Sedom Sha (Fig. 4b), one of four active stream-sinks of Sedom Cave. e combined features clearly show that the depression is an unroofed cave, formed by the collapse of paleo-Sedom Cave, whose outlet had been at the chasm to the north of SSP. SSP is a remaining part of the eastern wall of the unroofed cave. As an integral part of the eastern escarpment of the diapir bedrock, the rising rate of SSP was monitored, for the rst time, by direct geodetic leveling across the eastern escarpment. A 100-m long leveling traverse, using an automatic tripod-mounted level (Bomford O MFORD 1985), has been measured yearly three times, over a set of benchmarks through the lower level bedrock of Sedom Cave (Fig. 3). Point 9 is located within the same vertical salt layer of SSP. e mean measured rising rate is 9.3+3.5 mm/yr across the entire traverse (Frumkin, 2009). e mean measured upli rate of SSP layer in respect to the foothills is 6+1.5 mm/yr. e rising rate measured in this study is relative to the foothills, thus it combines two components: (1) the absolute rising rate of the diapir, and (2) the absolute subsidence rate of the eastern foothills plain, associated with the subsidence of the southern Dead Sea basin. is combined eect is reected in the high relief of the eastern Sedom escarpment and border faults location (Figs. 2, 3). Cave level age is well represented by radiocarbon dates of wood fragment deposits, because the wood residence time on the surface before being swept underground, as well as the life span of a single cave level are smaller than the error of the date (Frumkin et al., 1991). Six radiocarbon dates, ranging from 4050 to 3580+80 yr BP (Fig. 5), were obtained from four wood fragments (SN1,2,3, and 5) in the upstream part of the upper levels of Sedom Cave, where the upper levels of the cave are well-preserved, with no major collapse. Aer calibration, the calendar dates within 95% probability indicate that the upper levels associated with SSP started forming between 2900 and 2350 calendar yr BCE. A probable (but not necessary) trigger of a cave collapse in cohesive salt within this tectonically active region is an earthquake. e dimensions and form of the studied unroofed cave suggest that SSP could become a distinct isolated pillar following a major earthquake. Other collapse dolines Mt. Sedom are discussed by Frumkin R UMKIN and Raz AZ (2001). e collapse terminated the functioning of the cave channel as an active stream. e maximal likely age for this event is given by the youngest calendar date of the dri wood, ranging between 2150 and 1730 BCE (95% probability, Fig. 5). e elapsed time (~3950+100 yr) multiplied by the measured rising rates suggest that SSP was ~38+14 m lower when it rst became an isolated pillar. is was much closer to the foot of Mt. Sedom, rendering it easily observable by anybody travelling along the foothill. e radiocarbondated disturbed sedimentary layers of the Dead Sea contain the best long-term record of Dead Sea earthquakes (Migowski I GOWSKI et al., 2004). Within this record, the largest cluster of relevant seismic events occured between ~2100 and 2000 BCE, represented by three disturbed sediment sequences. e intermediate one, ~2050 BCE, was suggested to reect a magnitude 8 earthquake, the strongest in the entire Holocene record of the Dead Sea (Migowski IGOWSKI et al., 2004). is date agrees with the latest Sedom Cave dates, suggesting that the same high-magnitude earthquake may have triggered the collapse of paleo-Sedom Figure 5: Radiocarbon dating of wood remains om Sedom Cave, upper level of Shual Natush branch. Conentional radiocarbon dates are reported with Calibration is determined by OxCal (Bronk-Ramsey, 2001). Age range of calibrated dates represents 95% probability, corresponding to the lower horizontal lines in each probability graph (right).

PAGE 232

15th International Congress of Speleology Earth Sciences 1493 2009 ICS Proceedings Cave, bringing about the formation of SSP. e present explanation and dating of Lots wife is independent of the explanation of the Sodom and Gomorrah narrative, although the two events may be related, as implied by the biblical narrative (Genesis 19) and the following discussion. Scholars who adopt the historical view of Genesis 19 usually relate it to the late Early Bronze or early Middle Bronze Age, ~2000 BCE, based on independent reasoning (e.g. Albright L BRIGHT et al., 1944; Grintz RINTZ 1983; Meitlis EITLIS 2006; Nissenbaum ISSENBAUM 1994; Rast AST 1987; Speiser PEISER 1964). is age is consistent with the present study. It may thus be suggested that the biblical narrative of Lots Wife originated from the collapse of the ancient Sedom Cave followed by an observation of the sudden appearance of a salt pillar ~2000 BCE. e measured rising rate of the diapir suggests that the salt pillar formed close to the eastern foothills plain, attracting travelers. Evoking an explanation, SSP was perceived as a testimony for Gods wrath on moral grounds. e early observers transferred the narrative as a living memory through time, until the evolved myth was recorded in the book of Genesis. eir morphologicallithological observation (pillar of salt) withstood the tide of time as it was continuously supported by new observations. is is probably the oldest narrative that can be geologically attributed to a particular dated geologic feature.Acknowledgementsis study beneted from grants of the Dead Sea Research and Development, Samuel D. Terner, Esquire, and the Science Ministry of Israel.ReferencesAlbright LBRIGHT W.F., J.L. Kelso ELSO and J.L. Thorley HORLEY (1944) Early Bronze pottery from Bab ed-Dra in Moab. Bull. of the AA merican Schools of OOriental R Research 95, 3. Ben E N Menahem ENAHEM A. (1991) Four thousand years of seismicity along the Dead Sea ri. Journal of g eophysical RResearch 96 B12, 20195. Bentor E NTOR Y. K. (1989) Geologic Events in the Bible. Terra N N o a 1, 326. Blanckenhorn L ANCKENHORN M.L.P. (1898) DD as tote meer und der untergang on Sodom und GG omorrah. Reimer, Berlin. Block L OCK J.W. (1975) Sodom and Gomorrah: A volcanic disaster. Journal of GG eological EEducation 23, 74. Bomford O MFORD G. (1985) GG eodesy. Oxford University Press, New York. p 855. Bronk R ONK Ramsey AMSEY C. (2001) Development of the Radiocarbon Prograam OxCal. RR ad iocarbon 43(2A), 355. Childs H ILDS B.S. (1974) e etiological tale re-examined. Vetus Testamentum 24(4), 387-397. Flavius L AVIUS J. (1961) CC omplete works; the antiquities of the Jews, 1,11,4. Kregel, Michigan. pp. 775. Frumkin R UMKIN A. (2000) Speleogenesis in saltthe Mount Sedom area, Israel. In Spe leogenesis: EE volution of Karst AAquifers (eds. A.B. Klimchouk, D.C. Ford, P.A. and W. Dreybrodt), pp 443. National Speleological Society, Huntsville. Frumkin R UMKIN A. (2009) Formation and dating of a salt pillar in Mount Sedom diapir, Israel. GG eo logical Society of A A merica Bulletin 121, 286. Frumkin R UMKIN A. and D.C. Ford (1995) Rapid entrenchment of stream proles in the salt caves of Mount Sedom, Israel. EE arth Surface Processes and LL andforms 20, 139. Frumkin R UMKIN A., M. Magaritz AGARITZ I. Carmi ARMI and I. Zak AK (1991) e Holocene climatic record of the salt caves of Mount Sedom, Israel. e HH olocene 1(3), 191. Frumkin R UMKIN A. and E. Raz A Z (2001) Collapse and subsidence associated with salt karstication along the Dead sea. CC ar bonates and evaporites 16(2), 117 Gilat I LAT A. and A. Vol OL (2000) Sodom and Gomorrah: a case of earthquake-eected hydrothermal explosion. In AA nnual meeting (eds. G. Baer and Y. Avni), p 49. Israel Geological Society, Maalot. Grintz R INTZ Y.M. (1983) e Book of GG enesis, its UU niqueness and AA ntiquity (HH ebrew). Magnes, Jerusalem. Harris A RRIS G.M. and A.P. Beardow EARDOW (1995) e destruction of Sodom and Gomorrah; a geotechnical perspective. uarterly Journal of E E n gineering GG eology 28(4), 349.

PAGE 233

Earth Sciences 1494 2009 ICS P roceedings 15th International Congress of Speleology Issar SSAR A.S. (1990) Water shall ow om the rock; hydrogeology and climate in the lands of the Bible. Springer Verlag, Heidelberg. p 213. Lynch Y NCH W.F. (1849) NN arr ative of the UU nitd States expedition to the RR iver Jordan and the DDead Sea. Lea and Blanchard, Philadelphia. p 508. Meitlis E ITLIS Y. (2006) DD ig ging the Bible (HH ebrew). Rubin Mass, Jerusalem. p 287. Migowski I GOWSKI C., A. Agnon GNON R. Bookman OOKMAN J.F.W. Negendank EGENDANK and M. Stein TEIN (2004) Recurrence pattern of Holocene earthquakes along the Dead Sea transform revealed by varve-counting and radicarbon dating of lacustrine sediments. EE art h and Planetary Science LLetters 222, 301. Neev E EV D. and K.O. Emery M ERY (1995) e destruction of S odom, GG omorrah, and Jericho. Oxford University Press, Oxford. p 175. Nissenbaum I SSENBAUM A. (1994) Sodom, Gomorrah and the other lost cities of the plain a climatic perspective. C C lima tic CC hange 26(4), 435. Pe E eri E RI S., H.A. Zebker EBKER Z. Ben EN Avraham VRAHAM A. Frumkin RUMKIN and J.K. Hall ALL (2004) Spatiallyresolved upli rate of the Mount Sedom (Dead Sea) salt diapir from InSAR observations. II s rael Journal of E E arth Sciences 2(2), 99. Rast A ST W.E. (1987) Bab edh-Dhra and the origin of the Sodom saga. In AA r chaeology and Biblical I I nterpretation (eds. L.G. Perdue, L.E. Toombs and G.L. Johnson), pp 185. John Knox Press, Atlanta. Reinach E INACH T. (1895) Textes dauteurs GG recs et RRomains relatifs au Judasme, p 163. Ernest Leroux, Paris. p 375. Speiser P EISER E. A. (1964) e AA nchor Bible GG enesis. Doubleday, New York. p 377. Trifonov R IFONOV V.G. (2007) e Bible and geology: Destruction of Sodom and Gomorrah. In GG eo logical Society Special Publication (eds. L. Piccardi and W.B. Masse), pp 133. Geological Society, London. Weinberger E INBERGER R., Z.B. Begin EGIN N. Waldmann ALDMANN M. Gardosh ARDOSH G. Baer AER A. Frumkin RUMKIN and S. Wdowinski DOWINSKI (2006) uaternary rise of the Sedom Diapir, Dead Sea basin. In NN ew ontiers in D Dead Sea paleoenironmental research, GGSA A Special Paper (eds. Y. Enzel, A. Agnon and M. Stein), pp 33. GSA, Boulder. Zak A K I. (1967) e geology of Mount Sedom, PhD thesis (in Hebrew, English abstract). PhD thesis e Hebrew University, Jerusalem.

PAGE 234

15th International Congress of Speleology Earth Sciences 1495 2009 ICS Proceedings GEOLOGIC AND HYDROGEOLOGIC CONDITIONS IN SPELEOGENESIS OF THE LONGEST AND DEEPEST CAV V ES IN KARST OF CROATIAMLADEN LADEN GARA GARA I I 1,2, DA DAVOR OR GARA GARA I I 2 1UU niversity of ZZ agreb, Kaciceva 26, HRHR -10000 ZZ agreb, CC roatia, mgarasic@grad.hr2CC roatian Speleological Federation, NN oa Ves 66, HRHR -10000 ZZ agreb, CC raotia, speleo@speleo.hr In Croatia about 54% of territory has characteristics of karst relief (with sea oor about 73% of territory). Between two basic kinds of speleological features in Croatia (caves and pits), pits prevail in about 70% cases. In Croatian karsts are 53 caves and pits with depth over -250 m (three pits with depth over -1000 m, one over -750 m, one over -650 m, nine over -500 m, ve over -400 m, sixteen over -300 m, eighteen over -250 m). ere were presented few speleological features with the largest depth in the Croatian karst. eir basic geomorphological characteristic were analyzed and compared: location, depth, kind, morphological type and basic conditions of speleogenesis (geology and hydrogeology conditions). e review of the most perspective areas for further speleological exploration in Croatia was made on the basic of above given facts. ere more than 11,000 caves explored in the Croatian karst areas, and it is assumed that several times more exist. In some kinds of calcareous breccia called Jalar from Paleocene geological period, mostly vertical and deep caves are present. For example System Lukina jama Trojama with depths of -1392 meters, Slovacka jama, depth of -1320 m, than System Velebita pit -1026 m, with worlds deepest underground free-fall vertical step, at -513 m, than Pit Patkov gust with entrance vertical of -553 m (the second in the World). e longest cave system ulin ponor -Medvedica near Ogulin is 16,396 m long. It had mostly developed through erosion and corrosion activities of the underground sinking Dobra river. e deepest karst springs are in Crveno jezero (Red lake) at -281 m, Una river spring at -205 m, Kupa River at -154 m, and Cetina river spring at -115 m, etc. All the Dinaric karst system is aected by neotectonic movements and this is one of the main reasons for the many vertical speleolological features found in the region.. 1. Types and Forms of Speleogenesis in Croatian Karste correlations between the initial, main and fossil phases of speleogenesis, the karstication types (gravity, regression, or complex karstication) and the place of occurrence (internal, central and peripheral karst zone), provide new information and scientic data, and point to the conclusion that the examination of caverns, which outnumber by far the speleological structures with natural entrance, can be regarded as the very foundation for the study of karst. e speleogenesis of caverns is directly related to their lithostratigraphic, hydrogeological and tectonic predispositions. ese occurrences are dominantly vertical speleological structures (pits), of knee or column shaped morphological type, assuming the hydrogeological function of intermittent or permanent springs, and are now going through the main phase of speleogenesis. Caverns are not aected by atmospheric inuences, or by surface waters via natural entrances, and so they reveal the veritable condition of underground cavities.2. Classication of Speleogenesise extensive study of genesis of speleological structures in Croatian karst, as based on a considerable number of examples, shows that such structures are located in the immediate vicinity of fault planes or paraclases, or along fault zones. Traditionally, speleologists have been using various techniques in the study of speleological structures, but were limited to structures having natural entrance at the ground surface (Beck, 2005). A signicant breakthrough was made by the study of caverns discovered during construction of motorways, bridges, tunnels viaducts etc., which has in fact enabled investigations along fault paraclases or in fault zones. It was noted that speleological

PAGE 235

Earth Sciences 1496 2009 ICS P roceedings 15th International Congress of Speleology structures always occur in several groups at a single fault (vertical or horizontal sequence), depending on the type of rocks, intensity of karstication, neotectonic activity, etc.2.1. Along the neotectonic outcrops of rock complexese correlation of absolute displacement (neotectonic activity) measurements shows that the frequency of speleological structures at the same fault (or fault zone) is the greatest in vertical direction in case of neotectonic outcrops, and when there is a smaller deviation from the direction of neotectonic impact in relation to the predisposed fault signicant for speleogenesis. 2.2. Along the neotectonic downli of rock complexesOn the other hand, the faults on which horizontal continuations of caverns are formed are generally found in the zone of neotectonic downli, or in the zone of change of direction of neotectonic impact in relation to the predisposed fault. e probability of occurrence of vertical or horizontal continuations of caverns was statistically calculated based on 3877 identied and registered faults and 845 caverns formed on their paraclases, or on paraclases of neighboring parallel or subparallel faults or fault zones. ese values range from 30 to 35 percent and, when neotectonic predispositions of faults are taken into account, the probability of occurrence of a speleological structure assumed in advance varies from 55 to 65 percent. e neotectonic activity was determined by direct measurement of absolute displacements in the surrounding speleological structures, or based on prognostic maps of neotectonic upli or downli. is method could be used to assume the spreading of many speleological structures in our karst and, of course, in the zones when construction activities are undertaken. By comparing the neotectonic (presenttime) orientation and intensity of displacement, it would be possible to compare, with a relatively high level of accuracy, the speleological structures in similar rock formations and, combined with comparison of their karstication rates, the resulting correlation data would be very indicative. e intensity of karstication is currently measured and compared in some caverns. e inuence of seismic waves is also measured and related to constantly measured neotectonic displacements.3. Regressive (Inverse) KarsticationIn the course of study of caverns in our karst, it was noted that, in all phases of speleogenesis, many structures tend to intensively develop towards the ground surface, which is due to the mentioned neotectonics, but also to the ground water inuence. is has resulted in formation of some large-size speleological structures which are internationally known as sinking valleys or sinkholes, and we have dened them as speleological structures formed by regressive karstication. In the rst two phases, the cavern is widening along the fault paraclases towards the ground surface. In the third phase the cavern is already a speleological structure with its entrance at the ground surface, while the fourth phase is characterized by the start of intensive rock weathering activity and backlling of the speleological structure. e tectonic predisposition of the area in which speleological structures develop is actually of highest signicance for speleogenesis (Herak, 1984; Garai, 1984a; 2002). It is known that karstication in horizontal or slightly inclined layers advances more slowly when compared to steep or vertical layers. More interesting conclusions can be derived from the results obtained by systematic study of all caverns in tunnels along the Rijeka Karlovac Motorway where the data about the intensity of karstication, as well as the morphospeleological and hydrogeological data, were continuously monitored and compared. In some tunnels (Tuhobi, Vrata, Sopa, Veliki Gloac) the tectonics has played a crucial role in speleogenesis. ese are speleological occurrences of mostly tectonic origin, and are exclusively dependant on the closeness to fault surfaces (Garai, 2002). ese occurrences are relatively rare. Less than 3 percent of structures registered so far in Croatian karst were formed exclusively by tectogenesis, i.e. without a signicant erosive or corrosive inuence of ground water (Garai et al., 1993, 1994). Other than in fault zones, they can also be found in the apexes of local anticlines (Sopa Tunnel) where ground water has subsequently exerted a secondary role in the speleogenesis. e hydrospeleogenesis is always a complex process in which water has a dominant role. e erosion and corrosion (mechanical and chemical speleogenesis) act together in the development of speleological structures (Trailkill, 1968; Garai, 2002). In fact, about 93 percent of structures registered in our karst have been formed, to a lesser of greater degree, through the inuences of erosion and corrosion.4. Other TypesSome other rocks present in our karst (particularly near the town of Knin) also have exceptional hydrogeological conditions. ese rocks are gypsums and evaporites. Other than on carbonate rocks (which are best known to us), the karst also develops on some other rock types which are

PAGE 236

15th International Congress of Speleology Earth Sciences 1497 2009 ICS Proceedings soluble and have a good secondary porosity. Evaporite rocks are a good example. ey are formed by chemical leaching from highly-concentrated solutions, as a result of a very strong evaporation. e most important and most frequent evaporite sediments are gypsum, anhydrite and salts. e chemical composition of the gypsum mineral is Ca(So4) x 2H20 and, just like other evaporites, it is highly soluble in water. e main dierence in the karstication of gypsum and carbonate rocks lies in the fact that the gypsum is more soluble. For that reason the surface denudation of gypsum is very intensive. e surface morphology of karst lying on gypsum is generally similar to morphology of carbonate terrains, although some specic features do exist. Concave forms are mostly less oen pronounced, i.e. they are wider and shallower. e occurrence of gypsum is related to the so called interstratal karstication as, when found in nature, gypsum layers are oen clenched between other sediments. ey can thus be protected with a less permeable or impermeable overburden formations. e karstication of gypsum underneath such overburden results in creation of underground cavities which are responsible for the cave-in ad settlement of terrain which is otherwise not susceptible to karstication. e study of speleogenesis in gypsum has revealed that such caves form if two basic conditions are met: a transverse ow of water between the roof and oor aquifers via the gypsum layer, and lateral discharge of water via the insoluble but porous aquifer situated underneath the gypsum layer (Klimchouk, 2000, 2007). Anhydrites and gypsums are sedimentary rocks of purely chemical genesis, unlike carbonates that can also be of organic genesis, while conglomerates and breccias may be of mechanical (clastic) genesis. As according to its specic weight, the gypsum is lighter than carbonate rocks, we have here the so called diapiric upli. In fact, the gypsum rocks rise from the carbonate or other massifs and form, during this movement, diapiric folds with numerous ssures (leptoclasms, diaclasms and diastroma). ey are sometimes interbedded with fault paraclases. ese movements are measured in hundreds of meters and the upli can sometimes measure several kilometres in length. In Croatia, diapiric gypsum deposits and gypsum-based speleological structures are found only in the Kninsko polje area. ese are mostly smaller-size speleological structures that have been degraded to present dimensions (20 to 50 m at the ground surface) through subsequent orogenetic movements and gypsum solubility. e upli occurrences are estimated at more than 1500 meters. at is why it is quite possible that bigger speleological structures in gypsum do exist, but at greater depths (caverns). e neotectonic activity has been registered in several structures (peripheral karst belt), e.g. in the tunnels of Hrasten, Sveti Rok, Tuhobi and Uka. ere some speleothems have been disrupted by neotectonic movement of blocks. ese displacements are measured in millimetres and are of local type, and thus not detrimental to the roadway. All morphological properties of speleological structures (form, length, depth, orientation, inclination, etc.) are closely related to the factors of speleogenesis, its type, phase, intensity and location (various karst belts). It is today a known fact that the most complex karstication processes (karstication, corrosion, erosion, abrasion, denudation, accumulation, secondary and tertiary sedimentation, etc.) act in dierent directions depending on hypsometric position, geological (lithostratigraphic) predispositions and hydrogeologic conditions. Although most speleological occurrences/structures have been formed through karstication by gravity, some twenty percent of them have in eect been formed through the so called reverse or regressive karstication. An example of the so called complex karstication has been identied in a cavern in the Uka and Vrata tunnel (inuence of impermeable rocks). e karstication by gravity advances quite intensively in the downward direction, the reverse or regressive karstication advances in the upward direction, and the complex type of karstication advances at a similar rate in all directions. e karstication by gravity or surface karstication is always inuenced by gravity water, i.e. by precipitation rate. It can be observed quite easily and is very pronounced in Jurassic and Cretaceous carbonate sediments. Tectonic fracturation is favorable for its development (Garai, 1986a). e reverse or regressive karstication can be observed only in the underground (in caverns) and is due to hydrogeological conditions (change in the ground water level, its continuous presence in the underground in the rst and second phases of speleogenesis, etc.) and lithostratigraphical factors (it develops intensively in easily soluble formations, in Jurassic and Cretaceous ne-grained breccias, in Plio-uaternary breccias, and in blocks with neotectonic upli). Its presence has been noted in the area of neotectonic upli (e.g. in the Jablan Suica zone and in the South Velebit area), and its intensity is proportional to the intensity of the upliing action. e signs that clearly point to the reverse or regressive karstication are, for instance, the so called air cushions found on cave ceilings (in the area of Fuine, Josipdol and Brinje). Unlike karstication by gravity, the intensity of reverse karstication is less dependent on tectonic fracturation, and more on special hydrogeological

PAGE 237

Earth Sciences 1498 2009 ICS P roceedings 15th International Congress of Speleology conditions (Garai, 1993, 2002). A complex type of karstication has been found in the cavern at the contact between the Palaeozoic clastic impermeable formations and the Mesozoic complex of dolomitic limestones in the Vrata Tunnel and at the contact with ysch at the Uka Tunnel. It is however quite rare to nd karstication which is advancing at a similar rate in all directions. e need to have the roadway and/or tunnel above a spring is the biggest possible engineering-geological, hydrogeological and civil engineering challenge. Signicant examples are those above the Jadro spring (Mravinci), in ysch materials or above the Zvir spring in Rijeka (Katarina Tunnel), and in fractured Mesozoic carbonates. 5. List of the Longest and the Deepest Caves in Croatiae longest caves are listed in Table 1. e deepest caves are listed in Table 2e biggest vertical shas:Patkov gust -553 meters (entrance vertical sha) and Velebita -518 meters (inner vertical sha)e deepest and longest siphons:Crveno jezero (depth 281 m /ROW/, 181 m cave diver), Una river spring (-205 m), Kupa river spring (-155 m), Sinac spring (151m ), Cetina river spring (-115 m), Gacka river spring (105 m, length 950 meters) ReferencesBaksic,D. and all (2008): Deep pits of Velebit. www. speleologija.hr Beck, B. ed. (2005): Sinkholes and the Engineering and Environmental Impacts of Karst. Geotechnical Special Publication no.144, 1 pp, ASCI, Reston. No.Name: Location: Length (m): 1. Sustav ulin ponor Medvedica Gorski kotar, Ogulin 16.396 2. Sustav Varicakova (Mukinjska) Panjkova spiljaKordun, Rakovica 12.385 3. Jama Kita Gacesina Velebit, juzni 10.603 4. pilja u kamenolomu Tounj Kordun, Tounj 8.487 5. Veternica Medvednica, Zagreb 7.128 6. Sustav Jopiceva spilja Bent Kordun, Brebornica 6.710 7. Jama Munizaba Velebit, juzni 5.993 8. Sustav Vilinska spilja Ombla Dalmacija, Dubrovnik 3.063 9. Gospodska spilja Cetinska krajina, Vrlika 3.060 10. Donja Cerovacka spilja Lika, Gracac 2.779Table 1: e longest caves in Croatia. No.Name: Location: Depth (m): 1. Sustav Lukina jama Trojama (Manual II) Velebit, sjeverni 1.392 2. Slovacka jama Velebit, sjeverni 1.320 3. Jamski sustav Velebita Velebit, sjeverni 1.026 4. Amfora Dalmacija, Biokovo 788 5. Meduza Velebit, sjeverni 679 6. Stara skola Dalmacija, Biokovo 576 7. Vilimova jama (A-2) Dalmacija, Biokovo 572 8. Patkov gust Velebit, sjeverni 553 9. Jama Olimp I Velebit, srednji 537 10. Ledena jama u Lomskoj dulibi Velebit, sjeverni, Lom 536Table 2: e deepest caves in Croatia.

PAGE 238

15th International Congress of Speleology Earth Sciences 1499 2009 ICS Proceedings Ford D. and P. Williams (2007): Karst HH ydrogeology and GG eomorphology. 1 pp John Wiley & Sons, New York. Forti, P.(1991): Processi ipercarsici e speleogenesi. Speleologia, vol12, no.24, pp 42, Milano. Garai, M.(1984a): Prouavanje tektogeneze u speleolokim objektima. Priroda, vol.LXXIII, no.1, str.20, Zagreb. Garai, M. (1984b): Neotektonske aktivnosti kao jedan od uzroka geneze i morfologije jednog od najveih spiljskih sistema u Hrvatskoj. Deveti jugoslavenski speleoloki kongres, Zbornik predavanja, str. 457 465, fot.3, sl.1, Zagreb. Garai, M. (1986a): Hidrogeologija i morfogeneza speleolokih objekata s vodom u kru Hrvatske. Disertacija. Zajedniki studij geologije PM i RGN fakulteta Sveuilita u Zagrebu, str.1, sl.53, Zagreb. Garai, M. (1986b): Dominantan utjecaj geolokih uvjeta na morfoloke i hidrogeoloke tipove speleolokih objekata u hrvatskom kru. Na kr, vol.21, str.57 63, sl.3, Sarajevo. Garai, M. (1989): New conception of the morphogenesis and hydrogeology of the speleological objects in karst area in Croatia (Yugoslavia). 10. International Congress of Speleology, Procceedings,vol.1, str.234 236, sl.8, Budapest, Hungary. Garai, M. (1991a): Morphological and Hydrogeological Classication of Speleological structures (Caves and Pits) in the Croatian Karst area. Geoloki vjesnik, vol 44, str.289 015, fot.3, sl.4, Zagreb. Garai, M. (1993): e Karstication processes and Hydrogeological features of the Mesozoic rocks in the Karst of Croatia (Europe). Proceedings of the XI International Congress of Speleology XI International Congress of Speleology, pp.1-4, Beijing, China. Garai, M. and N. Boi, (2006): Deep pits and caverns in Croatia. 2nd Middle-east speleology symposium, Proceedings, pp 16. Bejrut : MESS2 2006, Speleo-Club du Liban, 2006. Garai, M. (2007): e Longest and Deepest caves in Croatian karst. European Geosciences Union General Assembly 2007 Vienna, Austria, April 15 20, 2007, A-03002; GM7-1TU5P-0363, Vienna. Garai,M. (2008): Speleoloki objekti u kru. Hrvatska vodoprivreda, god.XVII, broj 181, na str.59, Zagreb. Hauselmann, P. (2005): Cross formational ow, diuence and transuence observed in St. Beatus Cave in Siebenhengste (Switzerland). International Journal of Speleology, vol 34, 65, Bologna. Herak, M. (1984): Tektonski okvir speleogeneze. Deveti jug. speleo. kong., Karlovac, str.111, Zagreb. Klimchouk, A. (2000): Speleogenesis in Gypsum. In: Speleogenesis Evolutions of Karst Aquifers, Klimchouk, A. (2007): Hypogene Speleogenesis: Hydrogeological and Morphogenetic Perspective. 1 pp, National Cave and Karst Research Institute, Special Paper no.1, Carlsbad. Maucci, W.(1952): L ipotesi dell Erosione inversa. Bolletino dellea Societa Adriatico di Scienze Naturali, vol.XLVI, Trieste. Palmer, A. (2007a): Cave Geology. 1 pp, Cave Books, Dayton. Palmer, A. (2007b): Cave geology and speleogenesis over the past 65 years: Role of the National Speleological Society in advancing the science. Journal of Cave and Karst Studies, vol 69, no.1, pp 3, Huntsville. raikill, J.(1968): Chemical and hydrologic factors in the excavation of limestone caves. Geol.Soc.Am.Bull., vol 79, pp19, Washington.

PAGE 239

Earth Sciences 1500 2009 ICS P roceedings 15th International Congress of Speleology PROBLEMS WITH CAV V ERNS WHICH WERE FOUND ON THE ROUTE OF THE HIGHWAYS IN CROATIAN KARST REGION DINARIC KARSTPROf F .DR. MLADEN LADEN GARA GARA I I U U niversity of ZZ agreb, Faculty of CC ivil EE ngineering, DDepartment of GG eology and GG eotechnics, Kaciceva 26, HRHR -10000 ZZ agreb, C C roatia, mgarasic@grad.hr and CC roatian Speleological Federation, NN oa Ves 66, HRHR -10000 ZZ agreb, CC raotia, While building highways in karst regions of Croatia during the last eighteen years, over 945 caverns (speleological features without natural entrance) were discovered and thoroughly explored. All of them were geologically mapped surveyed and photographed in details. Research were carried out systematically on sections of roads, highways, cuttings, slides, tunnels, bridge foundations, viaducts, etc. while building about 500 km of highways in Croatian karst region, e.g. Dinaric Classical Karst area (Zagreb-Rijeka highway, Zagreb-Split Highway, Epsilon of Istria semi-highway, Rijeka-Rupa highway, Zagreb-Zadar semi-highway, Rijeka by-pass, etc). Some of these caverns have big chambers like in tunnel Sveti Rok tunnel and in some of them like in tunnel Vrata there were necessary built bridge over cavern. It is the rst bridge over cavern in tunnel in the world. Special protection of the stability of cavern roofs has been done with hundreds of anchors and jet grouting. Groundwater is well protected in all of ne w founded caverns.  1. IntroductionRegardless of the quantity, type and quality of primary geologic, geophysical or geotechnical investigations, it is dicult to positively identify most of the caverns in karst areas. In many cases, caverns are discovered during the construction process, without prior indications pointing to their existence, or with indications based on qualitative forecast only. Preliminary investigations rarely provide information about locations of a cavern and even less so about its shape and size. Caves are natural holes in the lithosphere and according to their size they may be ranked among diaclases and/or diastromas. ey may have an entrance at the ground surface although, more oen, they do not as they are fully buried in the interior of the karst massif. Entrance-less caves at the ground surface are called caverns. Almost one thousand caverns have been investigated in Croatia so far, and the total number of such occurrences exceeds 11,000. In Croatian karst, speleological sites are dened as diaclases wide enough to be penetrated by a human at least 5 m in depth, or 10 m in the horizontal direction. As geologic processes are characterized by permanent activity in the course of time, changes in the number of speleological sites are also possible (Garai, 2008). Caverns are exclusively found during construction and/or mining activities namely in cuts, side cuts, quarries, during excavation of bridge and viaduct foundations or, most frequently, during tunneling works. During construction of infrastructure facilities, these occurrences are always regarded as technical problems but, on the other hand, they also provide us with new knowledge about complex karstication processes. Some commonalities in the occurrence of caverns have been noted and it may be concluded that they outnumber by 20:1 the caves with the opening at the ground surface. is means that in Croatia there may be as many as 100,000 caverns reaching down to several hundred meters in depth. In our country, the deepest traces of caverns were found in deep boreholes in the Adriatic, where dripstones (speleothems) were found in zones reaching down 3125 m. Karstication depth exceeds 7 km in this region, as all preconditions favoring such karstication depths are in place (i.e., solubility and fracturing of rocks, and groundwater). Here we also have a Mesozoic geosynclinal carbonate basin ranking among the thickest in the world. All these data lead us to believe that caverns are common in this part of the world, where we can nd many types and species of speleogenesis in all its phases. Since 1991, systematic speleological, speleogeological, and hydrogeological investigations of these occurrences are mandatory in Croatia, and are routinely carried out during construction of transport facilities in karst regions. So far, more than 945 caverns have been inspected in full detail along our motorway routes. Proper remedial measures are taken with respect to caverns encountered during such investigations, taking into account the safety of the transport facility, the need to preserve groundwater quality, and construction costs. A correlation has been noted in the present and shape of caverns, lithostratigraphy, neotectonics, and hydrogeology, which is the eect of various forms of speleogenesis and karstication.

PAGE 240

15th International Congress of Speleology Earth Sciences 1501 2009 ICS Proceedings 2. Karst zoning in CroatiaAccording to the Zoning and Classication of Karst, Croatias motorway areas should be considered separately, as the speleogenesis is dierent in each area (Fig.1). is has also been proven by detailed speleological investigations along motorway routes that have been conducted continuously since 1991. e Zone between Zagreb and Karlovac is a valley type no-karst terrain, mostly made of uaternary and Plio-uaternary formations. ere are no speleological sites in this area. In the area from Karlovac to Bosiljevo, the motorway runs through the Internal Karst Belt where speleological occurrences are mostly related to Plio-uaternary and Triassic formations. Rijeka route: On the stretch from Bosiljevo to Fuine the motorway passes through the Central Karst Belt, which is characterized by numerous speleological occurrences/ structures in Mesozoic carbonate formations. From Fuine to Rijeka the motorway is in the Peripheral Karst Belt, which is characterized by the exclusive occurrence of vertical speleological structures with signs of strong fault tectonics. e zone from Rijeka to Rupa is the transitional iarija zone, dominantly with vertical but with some horizontal speleological structures. Istria: In the hydrogeologic sense, the Istrian area is within the almost horizontal layers of the Vrsar Anticline, which has been fractured by local tectonics and which carries water vertically towards the rivers Raa and Mirna, or towards Istrias main karst sources. e iarija zone is related to reverse faults and overthrusts with water-impermeable ysch barriers, while the Uka zone is characterized by the contact between the ysch and carbonate sediments. In this area, all caverns transport groundwater into the Kvarner Bay where the water emerges as numerous sub-sea water springs. Split route: e zone from Bosiljevo to the Mala Kapela Tunnel, including the boundary area along karst elds in Lika, is characterized by horizontal speleological structures that are greatly inuenced by groundwater. From the Mala Kapela Tunnel to the Sveti Rok Tunnel, the road traverses the Central Karst Belt abounding in groundwater, with vertical but with dominantly horizontal or inclined movement of ground water. e area is characterized with many estavelles. e area from the Sveti Rok Tunnel to the Maslenica Interchange is characterized by rapid vertical circulation of groundwater and by slow movement in horizontal directions. e area of Ravni Kotari is extremely signicant in the vertical and also more importantly in the horizontal circulation of groundwater as sometimes the impermeable ysch substratum is located in the immediate vicinity of roadways. e Dalmatian Hinterland area (Dalmatinska Zagora) is similar to the area of Lika, groundwater is more intensive as the Adriatic drainage basin in this area receives an additional supply of groundwater from Herzegovina. At this point, special attention must be paid to the vertical circulation of groundwater. A speleological structure with the highest minimum to maximum groundwater amplitude of as many as 236 m, which ranks it among the highest in the world, was registered in the zone between Zagvozd and Vrgorac. It is expected that a considerable number of vertical caverns will be found in the tunnel underneath the Biokovo mountain. e area to the south of the Neretva in the direction of Dubrovnik is characterized by dominantly vertical (less oen inclined) speleological structures from which groundwater arriving from Herzegovina is eventually discharged into the sea, via the subsea springs and other water springs. Analysis of simplied speleogenetic peculiarities of the area traversed by main transport routes in Croatian karst shows that in the main phase of speleogenesis, the caves have a big potential for transporting groundwater, both in vertical and horizontal directions. e groundwater changes, through its karstication activity, the engineering geologic and geotechnical properties of rock formations it traverses, which should undoubtedly be taken into account during construction in karst zones in Croatia. e Croatian karst ranks among the better known locus tipicus of the worlds karst. e underground karst forms are represented with over 10,700 speleological structures Figure 1: Karst zones in Croatia.

PAGE 241

Earth Sciences 1502 2009 ICS P roceedings 15th International Congress of Speleology (Garai, 1991) with varied morphological and morphogenetic features and hydrogeological functions. e study of speleogenesis of this area can be an indicator of karstication index and can also be applied to many other karst areas in the world. Although very many caverns have been discovered and studied in this karst region, a relatively small number of papers have so far been published in this eld of speleology. e following authors have published papers about caverns (i.e., about speleological structures without a natural entrance at the ground surface): Malez (1956) Drenovac Cavern in the Gojak Tunnel; Boievi (1985) cavern in the Uka Tunnel, caverns in the Obrovac Tunnel; and Garai (1993; 2002) caverns in the Uka Tunnel (Milanovi, 2004) and in all tunnels along the Rijeka Karlovac Motorway and Zagreb Split Motorway (Sveti Rok Tunnel) (Garai, 2002, 2004, 2008). e latter author personally participated in the study of all caverns in Croatia since 1975 to this day. Caverns along motorway routes have been systematically studied since 1991 and, to this day, the total of exactly 945 caverns have been fully investigated. Although all of these caverns vary by their genesis, morphology, depth and length, they still conrm the scientic assumptions about dierentiation of karstication level, which is dependant on the level, type, quantity and aggressiveness of groundwater, and on tectonic predispositions and lithostratigraphic properties of environment in which they were formed. In all big road tunnels (Uka, Tuhobi, Sveti Rok, Mala Kapela), there was at least one large-size speleological structure that had to be remediated, and in some cases there were more than 70 of them (Brinje Tunnel). In some tunnels, the issue was solved by designing smaller bypasses above the caverns (Veliki Gloac, Plasina), while in some cases veritable bridges had to be built (Vrata) (Figs.2, 3). e groundwater from some caverns encountered during the tunnelling work is now used in water supply (Uka), and some tunnels actually pass above the well eld (Mravinci, Katarina, Gri, Vrata). Interestingly, sea water was found in some caverns (e.g., Peine Tunnel). All these and many other hydrogeological problems have successfully been solved by builders either during preliminary investigations, construction or use of the transport facilities. e piers or abutments of the following viaducts and bridges are actually situated above caverns: Aneli, Vela Draga, Delnice, Golubinjak, Hreljin, Severinske Drage, Osojnik, Zeeve drage, Bijakue, Maslenica Bridge, bridge over the Krka River, bridge over the rivers of Dobra, Bistrac, Kamanik, Rjeina, etc. All caverns discovered along our transport routes have been investigated in full detail, and these investigations were in all cases backed by an appropriate documentation (reports submitted to the employers). If we compare the above hydrogeological problems in karst encountered during construction of transport facilities in Croatia with similar ones in other Figure 2: Cave chamber in highway tunnel. Figure 3: Map of cavern in the tunnel Sveti Rok.

PAGE 242

15th International Congress of Speleology Earth Sciences 1503 2009 ICS Proceedings countries, we can be no less than pleased. In fact, we have had no environmental problems in Croatia due to incompetent handling of hydrogeological issues in karst and in speleological structures (caverns). On the other hand, many examples can be cited of road construction projects in Trieste hinterland, in Austria, in France, and in the USA, where some changes occurred in the movement, quantity, and quality of groundwater following construction of transport infrastructure projects in karst zones. Several levels can be dierentiated in speleogenesis. Various authors point to three basic and several special levels in the generation of speleological structures. ese three levels are: initial level (or phase I), principal level (or phase II) and fossil level (late level or phase III). All three levels of speleogenesis are present in areas traversed by the Rijeka Karlovac (Zagreb) Motorway. e ratio of principal to ssile level varies on individual parts of motorways, but an average value for the entire Croatia is 3.11:1. A completely dierent situation has been established for the internal karst belt, i.e. for the area from Karlovac to Bosiljevo. Here the ratio of principal to fossil level of speleogenesis is 1:6.89, which is an indication that the voids are in most cases lled with detached material, calcite, or most frequently with highly plastic, very rm, red to brown-red clay. In the Central Karst Zone, i.e., in the zone from Bosiljevo to Fuine, the motorway traverses the area in which the speleogenesis is characterized by the principal phase to fossil phase ratio ranging from 2.36:1 to 3.05:1. In the Peripheral Karst Zone, i.e., in the zone from Fuine to Rijeka, this ratio is 4.20:1. All known and investigated structures on the route of the mentioned motorway, and structures in its immediate vicinity were taken into account when calculating proportions of individual phases of speleogenesis (Garai, M. 2003a,b). All three levels of speleogenesis are present in the area traversed by the Zagreb Split Motorway. An approximate ratio of principal to fossil levels varies along individual parts of the motorway and generally amounts to 3.05:1. e situation is completely dierent in the Internal Karst Zone, i.e., in the zone from Karlovac to Bosiljevo. Here the ratio of principal to fossil levels of speleogenesis is 1:6.89, which is an indication that the voids are, in most cases, lled with detached material, calcite, or most frequently with highlyplastic, very rm, red to brown-red clay. In the Central Karst Zone, i.e., in the zone from Bosiljevo to the Sveti Rok Tunnel, the motorway traverses the area where speleogenesis is characterized by the principal phase to fossil phase ratio ranging from 2.21:1 to 3.22:1. In the Peripheral Karst Zone, i.e., in the zone from the Sveti Rok Tunnel to the Maslenica Interchange, this ratio is 4.44:1. All known and investigated structures on the route of the mentioned motorway, and structures in its immediate vicinity, were taken into account when calculating proportions of individual phases of speleogenesis. If we compare the above two most signicant roadways, a good correspondence can be established in the sphere of karst zoning, i.e., with respect to the dominant inuence of individual types of speleogenesis. e initial phase of development of speleological structures is present in all areas under study, and the karstication process is in progress in all these areas. However, it was impossible to directly monitor this phase of development of caverns because of the narrowness (low width) of cracks and initial ssures. Nevertheless, in the Internal Karst Belt, i.e., in the zone from Bosiljevo to Fuine (or to the Sveti Rok Tunnel on the Split roadway), the parallel presence of initial, principal, and fossil phases of speleogenesis in the ratio of 1:1:1 was noted in some y speleological structures. If this information is compared with the overall picture of all known speleological structures in Croatia, it can be concluded that this is a relatively rare occurrence. In Croatian karst, there are 183 speleological structures (or about 1.58%) for which it can be claimed with high level of certainty that the speleogenesis is in all levels of development, and this without great horizontal or vertical spacing among individual structures (no more than 100 m). Following analysis of a large number of studies and data processing results, it can be concluded that most speleological structures in Croatian karst have not as yet passed through all three levels of development. is is due to geologic, tectonic, hydrogeologic, and climatic factors, i.e., the structures in high mountainous massifs, although aected by the gravity karstication process, have not as yet reached less permeable or impermeable substratum, and so the fossil phase of speleogenesis has not as yet been attained. It is of special interest to compare speleological structures having natural entrance at the ground surface with those without such entrance, i.e., with caverns with entrances that were revealed in the course of construction works. In 825 samples, all three phases of speleogenesis have not been established in any of the studied structures. ey can be statistically classied as follows: 599 (or 64.75% of the studied caverns) are now undergoing the principal phase of speleogenesis, 223 (or 24.10%) are in the fossil phase of speleogenesis, and 103 (or 11.15%) of caverns have a similar ratio of initial and principal phases of speleogenesis. For

PAGE 243

Earth Sciences 1504 2009 ICS P roceedings 15th International Congress of Speleology caverns, the ratio of principal phase of speleogenesis to fossil phase is: 599:223 or 2.68:1. is can be observed in the zone of intensive occurrence of caverns in the Central Karst Belt, where the percentage of fossil phase is higher. e reason for this can be found in the intensive tectonics and the great quantity of precipitation replenishing the groundwater in the Central Karst Belt (Gorski kotar, Lika). e fossil phase is particularly present in caverns found in tunnels with a low overburden (Brinje, Hrasten, Sopa). If we consider the position of caverns with respect to the ground surface (thickness of overburden), it can be concluded that in the Internal Karst Belt, the fossil speleological structures are dominant down to about ten meters from the surface, while in greater depths this phase is replaced with the principal phase of speleogenesis. In the Central Karst Zone (Bosiljevo Fuine, Bosiljevo Sveti Rok Tunnel; Dalmatian Hinterland), the principal phase is encountered aer several meters in depth, while the fossil phase occurs between -100 to -300 m. In the peripheral karst zone (Fuine Rijeka, Sveti Rok Maslenica Interchange, and locally in Ravni kotari) this situation is less oen encountered and, when present, the overburden is thicker and ranges from 200 to 500 m deep. e following conclusions can be made by analyzing the above results and by taking into account the types of karstication processes in individual areas: Due to additional regressive karstication, the Central Karst Belt (Bosiljevo Fuine; Bosiljevo Sveti Rok Tunnel; Dalmatian Hinterland) is characterized, in addition to the principal phase of speleogenesis, by a higher number of structures that are currently in the fossil phase of speleogenesis. From the standpoint of civil engineering activity, this area is exposed to the greatest number of speleological occurrences that may hinder or prove detrimental to construction activity. is is why caverns are sometimes unstable, and may greatly vary in size. Because of dominant karstication by gravity in the Peripheral Karst Belt (Fuine Rijeka; Mala Kapela Tunnel Maslenica Interchange Ravni kotari), the speleological structures are generally vertical and are going through the principal phase of speleogenesis. From the standpoint of civil engineering activity, this area is exposed to a considerable number of speleological structures, but they may be quite big in size and cause some problems in remedial work. e simultaneous occurrence of the karstication by gravity and regressive karstication in the Internal Karst Belt (Karlovac Bosiljevo) brings about a higher incidence of fossil structures when compared to the principal phase of speleogenesis. From the standpoint of civil engineering activity, such speleological structures are likely to cause problems because of varied engineering geologic and geotechnical properties of rock formations in which they were generated or deposited. e number of such structures is relatively small when compared to other karst zones. Such occurrences are also dominant on the Zagreb Rijeka Motorway. Karrens have been found in structures situated in the Peripheral Karst Belt and, to a lesser extent, in the Internal Karst Belt. ese karrens are one of the indicators of an intensive underground karstication (Grabovek, 2000; Gunn, 2004). Some other forms of erosive or corrosive action of water in the underground have also been noted in the biggest caverns investigated so far (i.e., in the Tuhobi Tunnel and along the Vrbovsko Kupjak section) Such forms include turbulence pots as well as the hieroglyphs and leopard skin. e force of chemical or mechanical groundwater action has been proven beyond any doubt in places where the rock is highly polished (Sljeme Tunnel area, foundations of the Dobra Bridge, foundation of the Zeeve Drage Viaduct). Carbonate rocks (milestones, calcareous dolomites, dolomitic limestones and dolomites) of various textures and age react quite dierently to erosion and corrosion at the ground surface (Ford, 1970; 1971; 2007). e dierent reactions have been proven in the underground as well. e most intensive karstication along the mentioned roadway has been registered in the limestones of Jurassic to Lias origin, and then in the Dogger and Malm sediments, and in the Upper Cretaceous limestones. e Jelar and Promina formations are of particular interest in the dierentiation of various types of speleogenesis. Carbonate clastites of Lika, Velebit, Mala Kapela and Hrvatsko Primorje were initially described as facies with the Promina strata. However, their unclear correlation and the dierence in composition have resulted in the introduction of the name "Jelar formation." eir origin was later attributed to signicant tectonic disturbances (reverse faults and overthrusts), which was followed by vertical faults. Promina strata, found in Ravni Kotari area and at the Promina mountain, and also the area characterized by Jelar formation, are always in transgressive contact with

PAGE 244

15th International Congress of Speleology Earth Sciences 1505 2009 ICS Proceedings older formations. ese are the carbonate breccias that dated back to Paleogene Neogene. In these formations, karstication is highly intensive. It should be noted that the spring of the world's third largest sinking stream the Gacka River is actually situated in Jelar strata, and that the Zrmanja River partly cuts through Promina formations. In addition, the largest underground cavities in Croatian karst are situated precisely in Jelar formations in the South Velebit area. ese underground halls may sometimes reach even more than 100 meters in diameter. Also, the deepest speleological structures in this area pass, either partly or fully, through Jelar formations, which points to subsequent intensive karstication (sometimes more than 450 meters in depth). e karst is intensively developed in Jelar and Promina formations, and better solubility of breccias has enabled creation of a number of karst phenomena. Recent speleological and hydrogeological investigations, conducted in the scope of the Zagreb Split Motorway construction, have revealed a great intensity of karstication and a variety of speleological occurrences that could not have been detected in earlier times through investigations conducted from the ground surface. us, these occurrences were discovered during excavation of tunnels (Gri, Plasina, Sveti Rok, etc.) bridges (over the Gacka), cuttings, etc. According to the number of speleological structures per km2, Jelar formations come aer Lias Jurassic limestones, and locally aer Senonian Cretaceous limestones.3. ConclusionsDetailed speleological and speleogeologic investigations, topographic survey, and photographing of structures, and hydrogeological observations within the structures, were undertaken by the team of experts headed by Prof. Mladen Garai, Ph.D. during construction of the Zagreb Rijeka Motorway and Zagreb Split Dubrovnik Motorway in the period from 1991 to 2008. Out of 945 caverns subjected to detailed investigations, 791 or 85.5% are vertical speleological structures, and 134 or 14.50 percent are horizontal speleological structures. No combined or complex speleological structure has been registered. As the motorway routes along which the investigations were made pass mainly through the Peripheral and Central Karst Belts, it is logical that the data slightly dier when compared to general data for the whole of Croatia vertical : horizontal : complex (V : H : C = 78 : 21 : 1). An average cavern depth is 35 m, which is more than an average value for 11,000 speleological structures registered so far in Croatia (with the exception of min/max peaks) e average cavern length is 53.5 meters (without peak values). e deepest cave reaches -196 m and the longest one measures 1490 m (Garai, 2004). A big cavity is the cavern in the Sveti Rok Tunnel (registered in the le-side tube) at KM 200+525 where 1137 meters of channels, with the altitude dierence of 147 m, were subjected to speleological investigations and topographic surveys. Another signicant cavity is the cavern in the Uka Tunnel measuring 1490 m in total length, and 151 m in depth. e extensive study of speleological structures newly opened along the route of the Zagreb Rijeka Motorway and Zagreb Split Dubrovnik Motorway has contributed to the better understanding of the process of karstication, speleogenesis, and speleomorphology of karst areas. An invaluable experience and scientic knowledge gained in this way will certainly enable easier and safer remediation of caverns to be encountered during road construction in karst areas. e volume occupied by these caverns varies from several tens to several tens of thousands of cubic meters. According to their orientation, they mostly follow the Dinaric Line of Spreading (northwest southeast) with about 67% of caverns. Some 29% of structures are oriented almost perpendicular to that line of spreading (northeast southwest), while the remaining 4% of the caverns have no dominant spreading. During the detailed speleological investigations made in the Tuhobi Tunnel, a certain correspondence, i.e. a systematic occurrence of three vertical caverns, spaced at no more than 50 m from one another, was established. ese caverns are situated in Jurassic limestones. Similar occurrences were later on observed in the Sveti Rok Tunnel, also in Jurassic limestones (Garai, 2002). In the Internal Karst Belt, the karstication depth ranges from 50 to 100 m, in the Central Karst Belt from 150 to 600 m, and in the Peripheral Karst Belt from 500 to 1000 or more meters. Zones with inclined or even horizontal circulation, from which the water is transferred to various drainage areas, were also registered. e preparation of cavern remediation design solutions was greatly facilitated by speleological investigations. A certain regularity of occurrence, as related to types of karstication and speleogenesis, was noted. At some points, the roadway route had to be partly modied, but the groundwater ow patterns discovered during speleological investigations have in no case been altered. e results obtained in the course of these investigations will prove valuable in future infrastructure development activities (e.g., realization of tunnels, cuttings, side-cuts, viaducts, bridges) in our karst which occupies as much as 54% of territory of the Republic of Croatia.

PAGE 245

Earth Sciences 1506 2009 ICS P roceedings 15th International Congress of Speleology ReferencesBoievi, S.(1985): Peine, jame i ponori s vodom u podruju Dinarskog kra. Kr Jugoslavije, vol.6, str. 137, JAZU, Zagreb. Ford, D. and. P. Williams, (2007): Karst Hydrogeology and Geomorphology, 1, Willey. Ford, D.(1970): Geologic structure and theories of limestone cavern genesis. Trans. Cave Res. Group, vol 13, pp 81, Bridgewater. Ford, D.(1971): Structures in limestones aecting the initiations on caves. Trans. Cave Res. Group, vol 13, pp 65, London. Ford, D. and P. Williams (2007): K arst HH ydrogeology and G G eomorphology. 1 pp John Wiley & Sons, New York Garai, M.(1991): Morphological and Hydrogeological Classication of Speleological structures (Caves and Pits) in the Croatian Karst area. Geoloki vjesnik, vol 44, str. 289-300, fot.3, sl.4, Zagreb. Garai, M.(1993): e Karstication processes and Hydrogeological features of the Mesozoic rocks in the Karst of Croatia (Europe). Proceedings of the XI International Congress of Speleology XI International Congress of Speleology, pp 1, Beijing, China. Garai, M. (2002): Speleoloke pojave u tunelu Sveti Rok. Ceste i mostovi, vol.48, br. 1, str. 11, Zagreb. Garai, M. and B. Vivoda, (2003a): Speleoloke pojave (kaverne) u trasi autoceste Zagreb-Rijeka. Ceste i mostovi, vol.49, br. 5, str. 104, Zagreb. Garai, M. B. Vivoda, and Z. Karlovac, (2003b): Geoloke znaajke tunela Roman brdo.Ceste i mostovi, vol.49, br.11, str. 58. Zagreb. Garai, M. (2004): Istraivanje kaverni u tunelu Sveti Rok. Speleozin, vol. 12, br. 17, str. 21, Karlovac. Garai, M. (2007): Sveti Rok Velika kaverna u cestovnom tunelu na autocesto Zagreb-Split. Hrvatska vodoprivreda, vol.XVI, br. 180, str. 56, Zagreb. Garai,M. (2008): Speleoloki objekti u kru. Hrvatska vodoprivreda, god.XVII, broj 181, na str. 59, Zagreb. Grabovek, F. (2000): Evolution of early Karst Aquifers: From simple principles to Complex models. IZRK, pp 1, Postojna. Gunn, J. ed. (2004): Encyclopedia of Caves and Karst Science.1 pp, Fitzroy Dearnborn, Taylor & Francis Group. New York, London. Malez, M.(1956): Istraivanje peina Gorskog kotara. Ljetopis JAZU, vol.LXIII, str. 347, Zagreb. Milanovi, P. (2004): Water Resources Engineering in Karst. 1, CRC Press.

PAGE 246

15th International Congress of Speleology Earth Sciences 1507 2009 ICS Proceedings Dye YE Tracing RACING Oil IL and AND Gas AS Drilling RILLING Fluid LUID Migration IGRATION Through HROUGH Karst ARST Terrain ERRAIN : A Pilot ILOT Study TUDY to TO Determine ETERMINE Potential OTENTIAL Impacts M PACTS to TO Critical RITICAL Groundwater ROUNDWATER Supplies UPPLIES in IN Southeast O UTHEAST New EW Me E Xico ICO USAJAm M Es S R R G G OODb B ARU U .S. DDepartment of II nterior, Bureau of LL and Management 620 EE GG reene St. CC arlsbad, NN ew Mexico 88220, UU SA A Rapid oil and gas exploration, drilling, and production in karst terrains in southeastern New Mexico are posing increased potential for contamination of critical groundwater sources. e area of primary concern is the Capitan Reef aquifer and the gypsum karst lands of the Delaware Basin. ese aquifers provide critical drinking water supplies to the City of Carlsbad, Carlsbad Caverns National Park, Whites City, several ranching families, and water wells for domestic livestock in the area. e aquifers are also the source of water for numerous springs and resurgences in the area that provide the basis for critical riparian areas and wildlife habitat. ese fresh water sources are vital in sustaining life along the northern edge of the Chihuahuan Desert ecosystem. ere has not been a systematic analysis of karst ground water in this region. Gathering baseline data on ground water quality and proving the connectivity between oil and gas drilling operations and critical water supplies is crucial in understanding, detecting, and mitigating undesirable events associated with oil and gas drilling and production operations. A pilot study was initiated by the Bureau of Land Management in the summer of 2005 to begin gathering eld data. e study area was expanded in 2007 to keep up with the expanding oil and gas development. Oil and gas drilling operations are required to put water tracing dyes into their drilling uids before they start drilling and then again before they case and cement the well bore. Dye receptors are placed in several springs and water wells. Dye receptors are sent to a laboratory for analysis. e results of the dye tracing of drilling uids help ll some of the voids in our understanding of karst groundwater ow in the Southern Guadalupe Escarpment study area and place greater emphasis on developing better methods of karst groundwater protection. 1. IntroductionAs the price of oil and natural gas continued to rise in 2004 the demand for drilling of new oil and gas wells increased dramatically. e central portion of the Capitan Reef and the gypsum karst lands of the Delaware Basin in southeast New Mexico overlays rich deposits of oil and natural gas. is is the boundary of the study area (Fig. 1). Within these boundaries are critical groundwater recharge zones and numerous springs and resurgences. e aquifers in this area supply drinking water to the City of Carlsbad, Carlsbad Caverns National Park, Whites City, several ranching families, and water wells for domestic livestock in the area. ese aquifers are also the source of numerous springs that provide the basis for critical riparian areas and wildlife habitat. e largest is Blue Springs, with an outow of 10 to 15 cubic feet per second. Also included are numerous springs that give rise to the Black River. ese fresh water sources are critical in sustaining life along the northern edge of the Chihuahuan Desert ecosystem. Groundwater research conducted by G.E. Hendrickson and R. S. Jones (Hendrickson, 1952) for Eddy County, New Mexico indicate that recharge of the groundwater associated with the Capitan and Carlsbad limestones is largely through the joints and fractures in the bottom of gravel lled arroyos. Water that enters the gravel and boulders in the arroyo bottoms moves Figure 1: Oil and Gas drilling uid dye trace area, southeast New Mexico, USA.

PAGE 247

Earth Sciences 1508 2009 ICS P roceedings 15th International Congress of Speleology downward into the underlying bedrock. e amount of water that enters the underlying rock and into the aquifer depends on the permeability of those rocks. Where the gravels are underlain by limestones all the water probably enters the underlying bedrock. Movement of groundwater aer it reaches the bedrock is controlled chiey by fractures and bedding planes, more or less enlarged by solution in limestone and dolomite. ey further state that the ow of the Black River, Rattlesnake Springs and Blue Springs is sustained chiey by discharge near the base of the Capitan reef escarpment. e principal source of these springs is almost certainly discharged from the Guadalupe Mountains area, as the recharge in the area between the reef escarpment and the springs is not enough to provide their ow. In addition to the water discharged by the springs, groundwater probably moves from the Capitan limestone and other underlying limestones into the alluvium and underlying Castile formation and may supply water to several of the other springs located in the gypsum karst lands of the Delaware Basin. Perched aquifers may be present in the uaternary piedmont alluvial deposits and the uaternary alluvial deposits which are underlain by the Castile formation. Recent studies by Snow (Snow and Goodbar, unpublished paper, 2007) indicate the critical recharge areas of the Capitan Reef aquifer are within one mile of the reef front and along the contact zones of the Yates and Seven Rivers formations in addition to the recharge into fractures in the alluviated canyon bottoms.2. Potential Oil and Gas Impactse reasonable foreseeable development of the study area has yet to be determined. ere have been several geophysical studies conducted in the area that indicate the potential for oil and gas exploration is high. e 1997 Bureau of Land Management Carlsbad Resource Management Plan Amendment for Oil and Gas (DOI 1997) shows the area as having a high potential for oil and gas occurrence. To date there have been 103 wells drilled in the study area. Of these, 76 are oil or gas wells drilled on federal land. e remaining 27 wells were drilled on state or private land. Based on the maximum allowable number of wells per section (16 oil wells and 4 gas wells) the maximum number of wells that could be located in the study area is 1,680 wells if maximum production were achieved. Drilling in the study area could aect both the perched aquifers and the underlying Capitan aquifer. Potential groundwater impacts from drilling can be divided into those caused: (1) During drilling and cementing; Drilling and cementing uids may spill or leak into formations at any lost circulation zone. (2) During testing and production; If the inner and intermediate casing strings fail following installation due to inadequate cementing or long term (50 + years) corrosion, drilling uids, brine, oil,or gas could be released directly to the subsurface anywhere along the casing string. (3) Following plugging and abandonment of the well; Because the atmosphere in the unsaturated part of the aquifers contain elevated concentrations of carbon dioxide and trace amounts of sulfur compounds as well as oxygen, the steel well casing could slowly become corroded and eventually fail in zones not protected by cement. (DOI, 1993) A more complete description of oil and gas impacts on caves and karst can be found in the DD ark CC anyon EE nironmental I I mpact Statement, UU S DDepartment of II nterior, Bureau of L L and Management, 1993. (DOI, 1993)3. e Pilot StudyIdentication and monitoring of karst areas includes gaining a better understanding of the underlying groundwater ow paths and their associated erosional features. To monitor the integrity of subsurface groundwater a program capable of identifying potential sources of contaminates entering the aquifers needed to be established. To begin this study the Bureau of Land Management in cooperation with the oil and gas industry, local land owners, the City of Carlsbad, and the Center for Cave and Karst Studies at Western Kentucky University initiated a die tracing pilot study to help identify areas of potential concern. e purpose of the dye tracing study was to determine if contaminates could enter the ground water through drilling and cementing operations, or during later phases of production or abandonment in the event of casing failure. Any positive results from the dye tracing study would then indicate that the BLM in conjunction with the oil and gas industry needs to ensure that all possible down-hole mitigation measures are being taken to protect these vital water resources. As a pilot study a small area was selected to begin with. at area extends north from Whites City to Dark Canyon and east of Whites City to Black River. In August of 2005 activated charcoal dye traps (bugs) were place in six locations, three in perennial springs, two in domestic water wells, and one in the outow of a Carlsbad city water monitoring well. ese bugs were retrieved aer

PAGE 248

15th International Congress of Speleology Earth Sciences 1509 2009 ICS Proceedings one month to ensure that samples were taken before any dye was introduced into the system. New bugs were then installed prior to dye being added to drilling uids during oil and gas drilling operations. New oil or gas wells drilled in the gypsum karst planes were then required to add 16 ounces of Fluorescein dye (Acid Yellow 73) to their surface interval drilling uid. For wells drilled in the Capitan Massive or Carlsbad limestones 16 ounces of Orange (Eosin Y) dye were added. For wells that were drilled through the gypsum karst overlying the Capitan Massive both dyes were required to be added to the drilling uid. Due to the large number of wells being drilled and the complexity of the project it was decided to use only two types of dye, one for the gypsum karst and one for the Capitan Limestone group. e amount of dye to add to the drilling uid was calculated based on the amount of water needed to ll a standard reserve pit for drilling oil and gas wells in that area. is began during the fall of 2005. e bugs were changed out bi-monthly and sent to the Western Kentucky University Crawford Hydrology Lab for analysis. In addition to the dyes that were required to be added to the drilling uids, analysis was also run for Rhodamine WT. is dye is oen used by the industry as a marker dye when conducting pre-ushing operations of the well bore before casing is cemented. During the pre-ush the dyed water may also enter karst aquifers. In 2007 the amount of dye was increased to 32 ounces of 100% powered dye to be added during the initial spudding of the well and to the preush uids before casing was cemented. is was done to ensure the amount of dye was sucient for detection aer being diluted by the waters of the aquifers. Additionally, to ensure that dye would be added to both the top and bottom of the karst zones. If dye were only added to the initial spud uids it could be lost into the perched aquifers in the shallow surface interval and not reach the lower intermediate drilling interval of the Capitan Reef Aquifer.4. Initial ResultsTo date 23 wells have added dye to their initial drilling uids. Fourteen of these wells are in the Castile formation in the gypsum karst planes of the Delaware Basin. Five wells have been drilled in the transition area containing both gypsum karst and the Capitan Reef Aquifer and four wells have been drilled on the crest of the Guadalupe Ridge anticline. Lost circulation has been reported in ve of the wells drilled that were using dye. at is not to say that lost circulation zones were not encountered in the surface intervals of other wells, only that it was not reported to the Bureau of Land Management. All the dye receptors were sent to Western Kentucky University, Center for Cave and Karst Studies, Crawford Hydrology Lab for analysis. All the dye receptors sent in before dyes were introduced to the drilling uids came back with no Fluorescein detected and a weak background of Eosine and Rhodamine WT dyes detected. Aer 16 ounces of dye were introduced to the drilling uids the Able water well came back with a detectable concentration of Eosine of .952 ppb which is nearly 2 orders of magnitude greater than the previous back ground concentration of .042. e dye receptor for this reading was put in on 9/07/05 and taken out 11/20/05. During that time the Estell AD #3 gas well was drilled with both Eosine and Fluorescein dyes being added to the drilling uid. is would indicate a denite connection between that well and the Able water well but would not explain why there was not a similar recording of uorescein dye in the same dye receptor. e Carlsbad City Water Monitoring Well in Juniper Canyon showed a similar increase in detectable Eosine moving from a low back ground level up to .563 ppb. e dye receptors showing the increase was put in on 9/7/05 and taken out on 8/20/06. During that time, four wells were drilled through the Capitan Reef aquifer using Eosine dye. Subsequent dye receptor analysis produced concentrations of .508 ppb and .930 ppb. of Eosine dye in this location. Background levels of Rhodamine W/T have been consistently recorded. No detectable levels of Fluorescein dye were recorded. Another location that showed signicant increases in detectable uorescein dye was Blue Springs, moving from none detected, up to low background levels (.068), and then up to .601 ppb. Additionally, the detectable concentrations of Rhodamine WT increased from a background level of .049 ppb up to a concentration of 1.017 ppb. is occurred during the h sample period. e dye receptor was put in 10/20/06 and collected 5/02/07. It is not known what wells in the area were drilled on private or state lands and which of those wells may have used Rhodamine WT during their drilling operations. Another interesting hit came in June, 2008 with a Fluorescein (acid yellow-73) detection of 1.937 ppb. It was during that time the Hot Seat well #2 was drilled using uorescein dye. e Jurnigan Spring location showed none to very low background concentrations of Eosine and Fluorescein dyes during the rst three samplings. e fourth sample showed a possible positive detection of Fluorescein dye of .528 ppb. e Fluorescein dye could be from one of the wells drilled in the transition zone of the reef escarpment to basin margin.

PAGE 249

Earth Sciences 1510 2009 ICS P roceedings 15th International Congress of Speleology At this time the Beard home and Black River locations have only shown below detection levels or background levels of dye concentrations.5. Technical Q uestions and SolutionsIt appears that there are no solid positive (+), very positive (++), or extremely positive (+++) concentrations of dye detected in any of the dye receptor locations. is may be attributed to the increased dilution of the dyes as they move into the aquifers. A second possibility is that once the drilling operation looses circulation into the rst open zone all or most of the drilling uid and dye are lost into that zone. Any other lost circulation zones connecting to aquifers that were encountered below that point that may not receive any dye. To compensate for these two possible issues the dye amounts were doubled to 32 ounces of 100% powder and a second addition of dye was required to be added aer the completion of the surface drilling interval. e dye is added to the pre-ush uids prior to casing and cementing the well bore. In this way dyes can be pushed into the lower portions of the drilling section and enter the bottom levels of the lost circulation zones. Another unanswered question is what the residence time of the dye in the aquifers is. is question may be answered as the project progresses but more probably by the design of a dierent study.6. ConclusionsDye tracing oil and gas drilling uids in the Castile gypsum and Capitan Reef aquifers appears to be a viable way of determining if drilling uids can enter the aquifers. If dyes are detected, it suggests that failures in the production casing and cementing may also allow hydrocarbons to enter the aquifers. With this in mind, it then becomes incumbent on the land managing agencies and the oil and gas industry to ensure that the best possible drilling, casing, and cementing programs are put into practice. e initial results are conclusive that the drilling uids do enter the aquifers. e changes in procedures of adding additional dye during the initial spudding of the well and before the casing and cementing of the surface string may aid in producing more detectable concentrations of dye in the collection locations. e pilot study should be continued and built upon. A more denitive study should be designed and considered to monitor and document the potential impacts in the shadow of impending oil and gas development in the karst areas south of the pilot study area.ReferencesHedrickson, G.E. and R.S. Jones, 1952, Geology and Groundwater Resources of Eddy County, New Mexico: New Mexico Bureau of Mines and Mineral Resources Ground-Water Report 3. Snow, Scott Rice, and James R. Goodbar, Terrain Factors in Capitan Aquifer Recharge, Northeastern Guadalupe Escarpment, New Mexico, Unpublished paper, 2007. U.S. Department of the Interior, Bureau of land Management, Dark Canyon Environmental Impact Statement, 1993, pp 4. U.S. Department of the Interior, Bureau of Land Management, Carlsbad Resource Management Plan Amendment, 1997, P. AP3-1.

PAGE 250

15th International Congress of Speleology Earth Sciences 1511 2009 ICS Proceedings Ca A Ve E de DE Velopment ELOPMENT influenced INFLUENCED by BY hydrocarbon HYDROCARBON o O Xidation IDATION : an AN e E Xample AMPLE from FROM the THE Polish OLISH Tatra ATRA Mts TSMICHA G G RADZIsk SK I1, MAREk K D D ULIsk SK I2, H H ELENA H H ERCm M AN3, MICHA yw YW IECk K I4 J ANUs S Z BARy Y A5 1 II nstitute of GG eological Sciences, Jagiellonian UU niversity, Krakw, Poland2 Faculty of Physics and NN uclear Techniques, UU niversity of Mining and Metallurgy, Krakw, Poland3 II nstitute of GG eological Science, Polish AA cademy of Sciences, Warszawa, Poland4 II nstitute of GG eochemistry, Mineralogy and Petrology, Faculty of GG eology, Warsaw UU niversity, Poland5 CC aving Section of Krakw Mountaineering CC lub Abstract e development of caves inuenced by the deep circulation of water has received increasing interest for last thirty years. Deeply circulating waters are characterized by elevated temperature and chemical composition dierent from meteoric water, which results in its aggressiveness against carbonate rocks, and thus inuences the rate of karstication. In the cave Dziura Wynia located in the northern slopes of the Tatra Mountains, huge crystals of calcite spar have been found. Distribution of the crystals shows that they grow during the primary, phreatic stage of cave development. Analyses of the stable isotope composition of the crystals prove that successive growth zones were build up in dierent conditions. e values of 13C increase systematically upwards starting from values as low as -28.8 (vs. VPDB). Such a low value strongly suggests origin of carbonate deposits due to CO2 generated by oxidation of methane. It is in line with results of inclusion analyses, which show the trend in crystallization temperature of subsequent zones of the crystals and chemistry of their parent uids. At the beginning the temperature grew up to ca. 300 C, which is associated with the presence of liquid hydrocarbons in the inclusions. Subsequently the temperature rapidly decreased and oscillated between 162 C and 174 C. e facts listed above suggest that the parent solutions were composed of two components mixed in dierent ratios: (i) ascending component of deep circulation containing the carbonate molecules produced during oxidation of methane and, in some stages, also liquid hydrocarbons, and (ii) component of shallow origin which chemical and isotopic characteristics may have been related to meteoric water. e trend of the stable isotope content coupled with the record of temperature changes indicates decreasing inuence of the former component during the crystal growth. As the lower zone of the crystals postdates the origin of the cave itself it suggests that the isotopically light CO2 derived from methane oxidation played the crucial role in the origin of Dziura Wynia cave as well as neighboring Dziura cave with famous ceiling cupolas. e oxidation was driven by mixing of the above described two components of carbonate solution which in turn led to creation of karst caves in the mixing zones. Bearing in mind the common occurrence of methane and its ability to migration the above conclusion have widespread implication. Similar mechanism may be responsible for creation of other hypogene caves all around the world.

PAGE 251

Earth Sciences 1512 2009 ICS P roceedings 15th International Congress of Speleology STUDY OF TEMPERATURE AND AIRFLOW IN THE SCHELLENBERGER ICE CAV V E BERCHTESGADENER LIMESTONE ALPS, GERMANY CC GRE GRE BE E J. RINGEI RINGEI S, and A A PFLI LI TSCH CH Workgroup of CC ave & Subway CC limatology, DDepartment of GG eography, RR uhr-U U niversity Bochum, GG ermany, christiane_grebe@ gmx.de, julia.ringeis@rub.de, apitsch@aol.com Abstract In contrast to many other European countries, limited research activities in German ice caves have been carried out, notably by several scientists in the 19th century. Recently speleologists have again undertaken such work. e Schellenberger ice cave has been known for a long time and was rst mentioned in 1826 in the Bavarian ordnance map. It is the biggest accessible ice cave in Germany (total length: 2815 m) and is located in the Untersberg massif (1570 m a. s. l.), Since 1925 the cave has been run as a show cave. It is, to this day, only illuminated by carbide lamps, because there is no electricity in this part of mountains. A big entrance leads to the largest hall in the cave with a dimension of 70 x 40 m (Josef-Ritter-vonAngermayer-Halle). e entire oor of this hall consists of an approximately 30 m thick and 60000 m3 ice block. e block is surrounded by the show cave trail. At the deepest point of the show cave trail (Fuggerhalle) the ice was dated through a pollen analysis to an age of 3000 years b.p. In addition, to the 500 m portion of the cave with ice, there is also one main non-ice part. is non-ice part leads through several shas to the deepest point of the cave (-210 m). Although the cave is known since the late 19th Century, only few investigations about the cave climate, glaciology, etc. have been carried out. In this presentation we present a new study and the results of the rst year of investigations, which were part of a master thesis at the Ruhr-University Bochum, Germany. In October 2007 three temperature data loggers were installed in three dierent microclimate zones and dierent levels of the cave. Because there is no access for electricity and no possibility to enter the cave during the winter time because of danger of avalanches, it has not been possible for us to install sonic anemometers. In June 2008 two more temperature data logger were installed. In October 2008, another datalogger was installed directly on the surface of the big ice block; this logger will be freeze inside the ice during the winter season 2008. e aim of this campaign is to characterize dierent seasonal aspects of the temperature and airow regime, to dene the climatic behaviour of the cave and its interaction with the ice block, and to dene possible reasons for the strong melting of the ice block in some specic parts of the cave. ese measurements are the beginning of a long-term monitoring campaign in German ice caves, which shall help develop an idea or even a model of processes, dynamics and controls in German ice caves. e rst results show a partly dramatic overheating of the cave during long periods during the year, which cause a strong melting in several parts of the ice cave. ey show a warming tendency in the cave during the dierent seasons. Interior and exterior factors inuence the microclimate inside the cave in all zones.

PAGE 252

15th International Congress of Speleology Earth Sciences 1513 2009 ICS Proceedings DIFFERENTIATING KARSTIC AND PSEUDOKARSTIC CAV V ES AND CLOSED DEPRESSIONS IN THE AMERICAN SOUTHWEST, USAWILLIA ILLIA M R R HALLIDA HALLIDA YIU IU S CC ommission on Volcanic CC aves, 6530 CC ornwall CC ourt, NN ashville, TNN 37205 UU SA A wrhbna@bellsouth.net From Meteor Crater (Arizona) to Devils Hole (Nevada), home of the famous endangered pupsh, a perplexing variety of sinkholes and other closed depressions and related caves exists in the American Southwest (USA). Included are spectacular punched-out sheer-walled pits in alluvium, volcanic vents, craters and calderas, crevice caves and pits penetrating through carbonate, volcanic and clastic formations alike, ordinary-looking sinkholes in Pleistocene basalt ows underlain by cavernous formations, and many others. Dierentiation of karst from pseudokarst and non-karst oen is especially dicult because of (1) superposition of cavernous karstiable and pseudokarstiable rocks and (2) occurrence of large and small depressions and caves in various poorly consolidated, poorly soluble materials. In intensive followup of sporadic observations beginning in 1949, three week-long studies of such features were undertaken in contiguous parts of Arizona, California, Nevada and Utah in 2007 and 2008. Excluded were (1) volcanic craters and calderas, (2) obviously karstic features in carbonate spring deposits (e.g., Dianas Punchbowl, NV), (3) well-known karstic caves, (4) well-known karstic sinks clearly overlying salt or gypsum deposits (e.g., Chevelon Creek Sinks and others in the Holbrook Basin, Arizona), and (5) well-known, previously visited volcanic pseudokarsts in southern Utah and northern Arizona. Not all the areas and individual features planned for these studies could be visited within time constraints. Nevertheless, the variety of observed features initially of uncertain origin was much greater than expected. Among the ndings were one lava tube cave of pre-Holocene origin, and the nature of one open vertical volcanic conduit was conrmed. Some slot caves in alluvium were found to be analogous to headward migrating domepits in dense limestone in the southeastern USA. Alluvium Cave, Nevada, was found to be identical with karstic White Beds Cave, but shallow to deep piping has produced caves, punchedout sinks and pits in alluvium in widely dispersed areas. In other cases, dissolution of carbonate rocks has produced somewhat similar landforms. Some so-called sinkholes in or adjacent to Pleistocene basalt ows on the Arizona Strip appears to be merely surface drainage channels obstructed by the basalt ow, but others are the result of collapse into dissolution or piping cavities. e previously proposed origin by piping of shallow saucer-shaped depressions in outwash gravels of Utahs Cricket Mountains could not be conrmed. e pseudokarstic nature of extensive Bloomington Cave (Utah) was conrmed; it is a three dimensional maze of crevice passages aligned along the strike of a monocline in Permian Kaibab limestone. One small cave entirely in salt was visited; it is close to the location of Jedediah Smiths Salt Cave, destroyed by Lake Mead, but is unrelated to it. Additional studies in these and adjacent states are needed. 1. IntroductionScattered throughout the American Southwest are closed depressions of varying sizes and types. ey range from Devils Hole, Nevada (the much publicized home of the endangered pupsh, most of which is tectonic and, thus, pseudokarstic rather than solutional), to Meteor Crater (Arizona) and large punched out sinks along the margin of the Virgin River, Utah, and the Holbrook Basin, Arizona. Some of the largest and most intriguing sinkholes are in thick alluvium. In many cases, it is dicult to dierentiate karstic from pseudokarstic depressions (and even to dierentiate evaporate karsts from carbonate karst). I began to observe such landforms in 1949 and, with the assistance of several experienced western cavers in 2007 and 2008, I undertook three week-long eld investigations in the western part of the region (Fig. 1). A narrative account of these investigations has been published (Halliday, 2008). Here, specic ndings and conclusions are correlated with ndings and conclusions of George W. Billingsley and other geologists who have studied extensive gypsum karsts of the Arizona Strip (the part of Arizona north and west of the Colorado River) and the Hualapai Indian Reservation with particular reference to breccia pipes. Heretofore, breccia pipes generally have been considered karstic in origin.

PAGE 253

Earth Sciences 1514 2009 ICS P roceedings 15th International Congress of Speleology Billingsley and co-workers term them solution-collapse features (e..g., Billingsley et al, 1999: p.1). Because most of their volumes consist of or represent downward-piped clastic material, I include them here as pseudokarstic whether or not they are cavernous.2. Pits and Sinks in Alluvium and in inly Mantled GypsumInnumerable individual and compound sinks are present in widespread gypsum karst and in alluvium in this vast, semi-arid area. ey vary widely in size and form. Some of the largest and most spectacular are punched-out sinks seemingly entirely in alluvium (e.g., Devils roat, Nevada, and Mystery Hole, Utah). Yet, numerous publications of the U.S. Geological Survey (e.g., Billingsley et al, 1999) indicate that the diameter of the average breccia pipe in this area is still larger. Using Billingsleys criteria (e.g., Billingsley et al, 1999), we encountered three closed depressions that appear to be surface manifestations of breccia pipes: Eskdale Sink (Hole in the Ground), Utah, Dantes Descent, Arizona, and Littleeld Sink, Nevada. Eskdale Sink is a large, cupshaped cavity in bleached, limonite-stained limestone, with an incomplete terrace of ne-grained clastics that may represent reworking of alluvium by spillover from Lake Bonneville (Fig. 2). Located in a low ridge of limestone at the western edge of Utahs desert ranges, it is about 200 m in diameter and 35 m deep. Dantes Descent is a vertical sha about 30 m in diameter and 90 m deep, in dense basalt and underlying Permian red, mostly clastic units, with a funnelshaped collar of alluvium. It is in gently sloping terrain near Ash Fork, Arizona. Located in alluvium just west of a major fault that demarcates the Great Basin from the Colorado Plateau, Littleeld Sink (Fig. 3). is the smallest of the three examples. It is notable for a hanging plug of characteristic breccia. e rear wall of its terminal grotto is injected by mineral stringers. Spectacular Devils roat and Mystery Hole are not as clearly downward piping structures but it is dicult to hypothesize any alternative form of speleogenesis in the alluvium in which they descend. A 1975 map showed the freefall depth of Devils roat (Fig. 4) as 29 to 46 m, and its maximum width as 43 m (McLane, 1976). Its oor slopes westward (the direction of surface drainage) in an overhanging alcove to a total depth of 71 m. By 2008, block slumping and spalling had reduced its freefall depth to 20 to 38 m. Previously considered to be entirely in alluvium, favorable lighting conditions in 2008 showed that red sandstone is exposed in its overhanging alcove (Fig. 5). Any remnant of piping structures has been covered by sand and gravel. Large salt deposits are present farther west, but major faults associated with two mountain ranges exist in the interval and piping into cavities in local deep-lying gypsum (presumably in the Permian Muddy Creek Formation) appears more likely. Two similar but smaller pits in alluvium recently have been found nearby. Mystery Hole (Fig. 6) appears to be slightly smaller than Devils roat, and only alluvium is visible in its punched out walls. It is located in the mouth of an alluvium-lled canyon of the House Range and its oor slopes toward that range rather than toward Sevier Lake, the local base level. Extensive Paleozoic Figure 1: Study area in Arizona, California, Nevada, and Utah. Figure 2: Utahs Eskdale Sink (Hole in the Ground) looking westward into Nevada. For scale, Dale Green is barely perceptible, standing on the rim in the center. e ledge of ne-grained clastics may have resulted om overtopping and reworking by Lake Bonneville. e bedrock is altered limestone and the feature probably is the surface manifestation of a breccia pipe.

PAGE 254

15th International Congress of Speleology Earth Sciences 1515 2009 ICS Proceedings limestone crops out nearby and Mystery Hole presumably formed by downward piping toward buried dissolutional caverns in the House Range. Smaller closed depressions in alluvium also were investigated. In Nevadas Amargosa Desert, Alluvium Cave was found to be well-known White Beds Cave, a karstic dissolution cave in dark grey limestone entered through an overhanging funnel-shaped pit. It apparently was developed by ordinary karstic processes in a presumably Paleozoic limestone, now buried by outwash gravels. Heaton Knolls Cave in the Arizona Strip is at the downslope end of a sizeable sink in alluvium notable for its ll of tumbleweeds. e cave itself is a single low room in the gypsiferous Harrisburg member of the Permian Kaibab formation. Its burrow-sized extensions are utilized by local wildlife. A similar sink adjoining the Larimore Tank basalt ow (see below) ended in impassible channels. U.S. Geological Survey maps of the Cricket Range in southwestern Utah show numerous small closed depressions variously labeled sinkhole, sinkholes/pipes and Big Sink. All are in outwash gravels in both sides of the range (Hintze, 1984), close to or in the highest terrace of ancient Lake Bonneville. Dale Green (oral communications, 2007 and 2008) previously had investigated nearly all of these shallow features. Neither he nor I found any evidence of lateral or downward piping in this area but without drilling, the possibility of downward piping into two or more unique Figure 3: Littleeld Sink is located just west of a major fault demarcating the Basin and Range Proince om the Colorado Plateau. e large light gray formation in the center of the photograph is a hanging clastic breccia and the feature appears to be the surface manifestation of a comparatively small breccia pipe. Figure 4: e northeastern end of Nevadas Devils roat appears to be entirely in alluvium. Figure 5: Red sandstone is exposed in the deeper western end of the Devils roat, within an overhung alcoe.

PAGE 255

Earth Sciences 1516 2009 ICS P roceedings 15th International Congress of Speleology series of dissolution caverns along the buried anks of this range cannot be excluded. Another possibility is that they are compaction features of Lake Bonnevilles reworking of outwash gravels, but similar closed depressions have not been reported in Lake Bonneville terraces elsewhere. ey are much too small to be surface manifestations of breccia pipes as the term is generally used.3. Crevice and Talus caves Both karstic and pseudokarstic crevice and talus caves exist in this study area. Near Walnut Canyon National Monument, Arizona, in 1949, Dr. Edward Held and I descended into Bottomless Pit (Rio Frio Sink) in a calcareous facies of the Permian Kaibab formation., one of many so-called limestone solution cracks in this general area. In the 2007-2008 studies, I conrmed that well-known Bloomington Cave, in a similar facies south of St. George, Utah, is tectonic and, hence, pseudokarstic. It is an extensive three-dimensional maze of crevice passages oriented along the strike of a local monocline. During the 2007-2008 studies, we also encountered crevice and/or talus caves in travertine in Death Valley National Park, CA (Traverkeane Cave), and in basalt ows in the Arizona Strip (Mt. Trumbull Ice Cave and Hoarfrost Cave, a new discovery). Mt. Trumbull Ice Cave is an important biological site. 4. Stream-Cut and Eolian CavesTwo notable stream-cut grottos were visited during these studies and their precursors. Both are locally termed caves but do not t the scientic denition of this term. e larger is Cottonwood Cave in Death Valley National Park, in a cobble fanglomerate. Its single chamber is 29 m wide and 19 m from the drip line to the rear wall. Height was estimated at 10 m. Also in easternmost California, Whipple Wash Cave is a gaping grotto in a thick Tertiary pyroclastic sequence consisting mostly of welded volcanic ash but with a detached clast of mylonized gneiss along one wall. No special eort was made to investigate individual pocket caves in sandstone, but inverse spheroidal exfoliation was noted in one such sandstone pocket in Lake Mead National Recreation Area, Nevada. 5. Piping Caves in Poorly Consolidated MaterialImportant caves formed by lateral and other piping in poorly consolidated material were encountered in and near Death Valley National Park, California, and in Cathedral Gorge State Park, Nevada. While the best-known examples in and near Death Valley are the Mustard Canyon group and the Tecopa Caves, those in Cow Canyon near park headquarters are considerably more varied. In addition, important non-cavernous pseudokarstic features were photo-documented in Twenty Mule Team Canyon. ese consisted of blind gullies and pseudokarstic dimples. None of the Cow Canyon caves is as much as 50 m long, but they vary from horizontal to vertical, with notable vertical lapies (karren) in Mammoth Well Cave, located at the top of a comparatively steep pitch in the canyon oor. e entrance room of nearby Furnace Cave has an unusual fanglomerate roof. Also in Cow Canyon, No Snakes Cave is primarily a maze of breakdown from collapse of a stream-cut alcove. A few small halite stalactites are present. ree horizontal piping caves near Tecopa, California, previously were known to be up to about 50 m long, with some walking passage. In a nearby gully in borate-rich lakebed deposits, several smaller examples are together with radial crevices at its head, suggesting that piping speleogenesis is an important geomorphic process here. In Cathedral Gorge, similar but smaller piping caves are numerous. In addition, three extremely narrow, extremely contorted slot caves are tourist attractions of this park. eir up-slope ends are Figure 6: e walls of Utahs Mystery Hole appear to be entirely in alluvium. Sevier Lake is seen in the middle distance, but the features oor slopes in the opposite direction, toward limestones in the House Range.

PAGE 256

15th International Congress of Speleology Earth Sciences 1517 2009 ICS Proceedings partially roofed. Characteristically their alluvium walls have well-developed lapies and resemble small migrating domepits in dense limestones of the southeastern U.S. Welldeveloped but miniature subterranean drainage systems also are present in these alluvial pseudokarsts, with many swallets and well-formed resurgence channels (Fig. 7). Beneath a sloping rhyolite ow, Arizonas Lake Havasu Bat Cave is a wide, spacious piping cave in welded Tertiary volcanic ash. On its sloping ceiling is an unusual multicellular structure several centimeters in height and width. It has the appearance of a fossilized nest of Tertiary wasps.6. Caves and Closed Depressions in Pahoehoe Basalt Flows of pahoehoe and other basalts are widespread in the Arizona Strip and elsewhere in the study area Here, Pleistocene basalt ows characteristically overlie the Harrisburg member of the Permian Kaibab formation, notable for extensive gypsum karstication. U.S. Geological Survey geological maps (e.g., Billingsley and Priest, 2003) show innumerable individual sinkholes in a variety of Pleistocene basalt ows (and a few breccia pipes as discussed above). e greatest concentration of these features is shown on and adjacent to the Pleistocene Larimore Tank basalt along Mohave County Highway 5. At least three lava tube caves were reported here. We investigated all of these caves and a total of 15 features shown as individual sinkholes. e U.S. Geological Survey Hat Knoll topographic map also shows large, compound depressions within the boundary of this ow, largely unrelated to the relatively small individual sinkholes shown on the geological maps. We attributed the larger depressions to gypsum karstication and did not investigate them specically. We also investigated the smaller Heaton Knolls lava ow where another lava tube cave (Red Blanket Cave) was reported. All four of the supposed lava tube caves were found to lack characteristic features of such caves (e.g., Larson, 1993). Instead, they were interpreted as volcanic talus caves formed by collapse or subsidence of non-cavernous pahoehoe lava into cavernous spaces in the gypsum karst in the underlying Kaibab formation. e largest of these is Paiute Cave (not to be confused with a breccia pipe near Marble Canyon, also called Paiute Cave). It is an important site for aboriginal art. Most of the designated sinkholes also are collapse or subsidence features, but a few are complex depressions, formed when lava overran or surrounded sections of pre-existing stream gullies. One is the mouth of a deep crevice cave on the edge of a compound sink, and one appears to be a small volcanic vent. e sinkholes originally shown on the geological maps necessarily were identied by stereoscopy rather than by eld investigations (George Billingsley, email communication, 2008). e special concentration of such sinkholes on the Larimore Tank basalt may be due to preferential ow of its lava into pre-existing wide, shallow depressions in the gypsum karst. To determine if some other volcanic caves not hitherto visited in the study area also might be volcanic talus caves, the Tabernacle Caves and the Snow Canyon caves in southwestern Utah also were visited. ese were conrmed as lava tube caves.7. Other Caves and FeaturesGneiss Cave, a pre-Pleistocene lava tube cave in Death Valley National Park, California, is discussed in a companion paper. e caves of Old Caves Crater, Arizona, are cavernous remnants of a small multifocal eruption center in rhyolite breccia or tu, modied by prehistoric Native Americans (Barrett, 1923). Salt Cave, Nevada, is in the interface between karst and pseudokarst. Its main room is dissolutional but the entrance passage area is tectonic. In central Nevada, a vertical sha near the summit of the compound volcanic hill just south of Lunar Crater was conrmed as an open vertical volcanic conduit about 10 m deep and half as wide; the rst identied in Nevada. Its ejecta contain many xenoliths. Notable kaminitzas on Permian Toroweap sandstone were photo-documented at Toroweap Point, high above the Colorado River.AcknowledgmentsMy profound thanks to Ed and Kathy Block, Peter Druschke, David Ek, Greg Flores, Dale Green, Tom Madsen, Neil Marchington, Mary Beth Pierce, and Kyle Voyles. Without their able eld assistance, these studies would not have been possible. Figure 7: Most well-developed drainage systems in poorly consolidated material in Cathedral Gorge, Nevada are too small to penetrate. is small cavernous resurgence is alongside the entrance of one of its three slot caves.

PAGE 257

Earth Sciences 1518 2009 ICS P roceedings 15th International Congress of Speleology ReferencesBarrett, S.A. 1923. A Trip to Cave Hill, Arizona. Yearbook of the Public Library of the City of Milwaukee, 1922. Volume II, pp 176. Billingsley, George H. et al. 1999. Breccia-Pipe and Geologic Map of the Southwestern part of the Hualapai Indian Reservation and Vicinity, Arizona. US Geological Survey Miscellaneous Investigations Series Map I-2554. Billingsley, George H. and Susan S. Priest. 2003. Geologic Map of Upper Clayhole Valley and Vicinity, Mohave County, Northwestern Arizona. US Geological Survey Miscellaneous Field Studies Map :MF-2438. Halliday, William R. 2007 Pseudokarst Hunting in the American Southwest. NSS News, Volume 66, number 10, October, pp 4. Hintze, L.F. 1984. Geology of the Cricket Mountains, Utah. US Geological Survey Open-File Report 84-693. Larson, Charles V. 1993. An Illustrated Glossary of Lava Tube Features. Western Speleological Survey Bulletin 87, 56 pp. McLane, Alvin. 1976. Devils roat. Cave Lights, number 25, p 6.

PAGE 258

15th International Congress of Speleology Earth Sciences 1519 2009 ICS Proceedings USING CONSERV V ATIV V E AND BIOLOGICAL TRACERS TO BETTER UNDERSTAND THE TRANSPORT OF AGRICULTURAL CONTAMINANTS FROM SOIL WATER THROUGH THE EPIKARSTIC ZONE BRIAN RIAN HA HA M1, RIC RIC K FO O WLER LER1, CHRI CHRI S GRO GRO VE E S1, CARL CARL BOL OL STER ER2 1Western Kentucky UU niversity, 1906 CC ollege HH eights Bld. Bowling GG reen, KY, UU SA A2UU nited States DDepartment of AA griculture AA gricultural RResearch Service 230 Bennett LL ane, Bowling GG reen, KY, UU SA A Agriculture contamination is very common in karst systems due to the vulnerability of these aquifers. Animal waste is oen spread across crop land to enrich the soil with nitrates and phosphates. Herbicides and pesticides are also applied to the crops. e transport of these pollutants through the soil and epikarst is a dicult process to monitor due to the complex, heterogeneous behavior of the groundwater as it makes its way down to the aquifer below. An experimental site at Crumps Cave lends a unique opportunity to monitor the vadose zone at a waterfall in the cave below. A previous dye trace established the connection of a 11.15 m2 grass plot to the waterfall mentioned above. is eld design, accompanied by a rainfall simulator, allows researchers to control the input of precipitation in an eort to understand more about the movement of stormwater inltrating the soil and the dierences in transportation of solute particles and bacteria in the epikarstic zone. Two particle transport experiments were used to better understand these processes. e rst trace involved the rainfall simulated injection of uorescein dye and sodium chloride. A 2650 liter solution with an average concentration of 60 ppm uorescein and a conductivity of 10,000 microseimens was injected over a period of 3.6 hours at a rate of 6.6 cm/hr. An electrical resistivity traverse perpendicular to the straight-line path between the established dye trace connection showed a peak in lower resistance at the upper epikarst layers 4 hours and 15 minutes aer the beginning of the injection. Dye concentrations reached a peak of 1600 ppb 3 hours and 15 minutes aer the beginning of the injection. e conductivity also peaked at this time with a value of 814 S. is rst trace showed that rapid transportation of solutes takes place in localized conduits causing a peak of both solutes in the cave before the widespread mobilization of sodium chloride is seen in the epikarst by the resistivity images an hour later. e second trace involved a rainfall simulated injection of sulphorhodamine dye over 180 kg of dairy cattle manure spread on the 11.15 m2 plot of grass. e 2650 L solution with an average concentration of 240 ppm sulphorhodamine was injected over a period of 3.6 hours at a rate of 6.6cm/hr. Dye concentrations reached a peak of 27 ppm 4 hours and 10 minutes aer the beginning of the injection. Fecal coliform reaches its rst peak of 2755 MPN (most probable number of viable cells per 100 mL of water) 90 minutes prior to the dye peak and a second peak of 2481 MPN 15 minutes prior to the dye peak. ese results show that solutes travelling through the soil and epikarst follow similar paths while bacteria prefer conduits that oer more rapid transmission to the resurgence.1. Introductionis research aims to resolve some of the ambiguity associated with contaminant transport in the epikarstic zone. e uncertainty of capacity and ow within the epikarst is most certainly due to the fact that this realm straddles the disciplines of karst hydrogeology and soil science. Despite these hurdles, scientists have experimented with the intriguing nature of the epikarstic zone. Much attention has been given to the topic of epikarst hydrology and geomorphology. Researchers have recognized the dierences and overwhelming similarities between epikarst networks around the globe (WILLIAMS, 1983;VENI et al., 2001). Water on its way to the epikarstic zone inltrates the ground surface and travels through the soil matrix as well as preferential ow paths such as macropores (RAY et al., 1997). e preferential owpaths in the soil and epikarst enable ashy recharge of the aquifer below while the consolidated soil matrix along with the fractured and inundulating bedrock surface dene the

PAGE 259

Earth Sciences 1520 2009 ICS P roceedings 15th International Congress of Speleology storage capabilities within the vadose zone. e ow regimes and storage capabilities of epikarst has been examined by JENNINGS (1985). He emphasized the importance of soil properties and vegetation with respect to vadose zone ow. JENNINGS (1985) asserted the signicance of vertical inltration promoted by colluvial soil structures. Sediments are an important part of understanding groundwater hydrology in karst terrains. LEGRAND (1973) explained the inversion of soils in karst which occurs when the majority of the soils in an area are stored in sinks and the subsurface. e ability of the bedrock void to accommodate the sediment from above lends itself to the maturation of macropore networks. Component separation of spring hydrographs has become a useful approach to studying the factors that aect recharge within karst groundwater basins (EINSIEDL, 2005). is method has allowed researchers to distinguish between vadose and phreatic recharge waters. is important step was necessary for understanding the karst systems since springs are the most accessible and convenient place to monitor a groundwater basin. Many studies have involved monitoring water quality parameters at springs in order to characterize the karst system upstream (HESS and WHITE, 1988; RYAN and MEIMAN, 1996; RAEISI and KARAMI, 1997). Modeling of karst aquifers has been an important tool in understanding the transportation of contaminants within these systems. WHITE (2003) outlined the dierences in many types of karst systems with his conceptual models. Water quality monitoring and particle transport experiments have enabled scientists to evaluate the physical and chemical characteristics of karst aquifers (GROVES et al., 2005; FLOREA and WICKS, 2001; EINSIEDL, 2005). Recent improvements in monitoring techniques and equipment have allowed researchers to become more condent in their conclusions and spread their study areas to a wider radius encompassing many dierent types of karst systems. Research has also been conducted investigating the transportation of colloids such as fecal coliform (BOYER and PASUARELL, 1999; AUCKENTHALER et al., 2002; GPPERT and GOLDSCHEIDER, 2007). Cave Spring Caverns has been established as an experimental site to study particle transport and ow within the subcutaneous zone (GROVES et al., 2005). It was previously established that slow and rapid recharges of the epikarst aquifer depend primarily on antecedent moisture conditions and intensity of rainfall. A signicant amount of storage in the epikarst aquifer above Waterfall 1 was recognized due to the hydrochemical response of the water at this site.2. Site DescriptionCrumps Cave is a well-known cave located about 1.5 km northeast of downtown Smiths Grove, Kentucky. e property is situated in the sinkhole plain 4 km south of the Chester Uplands. e cave, formerly known as Crumps Cave, is a Native American historic site and also protects gray bats, an endangered species. It was featured on a PBS television program emphasizing the caves archeological signicance and praising the conservation eorts of cavers. e previous landowner had been running a respectful historic bed and breakfast operation also making the new name of the cave recognizable to many. e cave has now been purchased by Western Kentucky University with money from a state conservation grant. e cave will now become an outdoor classroom where students will gain hands-on experience with environmental research. Crumps Cave is contained within the Mississippian St. Louis Limestone Formation and the Lost River Chert Bed is estimated to be located between the land surface and the cave roof. e regional geology dips to the west at one to two degrees. Groundwater ow from the cave has been dye traced traveling to Wolf Sink, Grant-Palmore Cave, Mill Cave, and nally Wilkins Blue Hole in the Graham Springs Groundwater Basin. Vadose recharge within the cave appears mainly along the east wall inside the rst section of the cave. ese waterfalls will are referred to as Waterfall 1, 2, and 3. Waterfall 1 is the predetermined tributary draining the injection recharge area. is has been determined by previous dye tracing (GROVES et al., 2005). Soil surveys of the area show that the site is covered with two predominate types of soils. e Pembroke soils of the area are described as being moderately permeable, welldrained soils that formed in loess underlain by residuum of limestone. e soil is generally a reddish-brown silty clay loam which increases in acidity with depth. Chert fragments range from ve percent in the upper portions to een percent in the lower portions. Pembroke soils are oen found in the same setting as Baxter soils. is is the case at Crumps Cave. e area around the entrance sink is dominated by Baxter soils and the area of the injections is near the boundary of the two soil types. Baxter soils are also moderately permeable, well-drained soils. e parent material of this soil is a weathered cherty limestone. is type of soil is oen found in the sinkhole plain where sinkholes have breached the Lost River Chert Bed. e

PAGE 260

15th International Congress of Speleology Earth Sciences 1521 2009 ICS Proceedings subangular structure of the soil in this case is weak as seen in many colluvial soils. e color of the Baxter soils is yellowish red turning to red with increasing depth. Chert fragments range from about ten to thirty percent. ese soils are noted to be strongly acidic to very strongly acidic. Below thirtyeight centimeters, the soil survey notes that there are few ne continuous pores. 3. Methodologye rst trace was on March 5, 2008 (Julian day 64), following the initial background establishment, involving an injection of 1360 grams of uoresceine and 14.74 kilograms of sodium chloride, both dissolved in 2650 liters of tap water. e dye and sodium chloride was mixed in a large water tank mounted on a atbed trailer. A hose was mounted on the output drain of the tank and connected to a small water pump. is pump then led to the rainfall simulator where the pressure can be manipulated by the user. e solution was pumped from the tank at an estimated ow rate of 6.6 cm/hr. e injection began at 11:45 and ended at 15:20. A subsequent ush was also done with 757 liters of tap water from a clean tank. e 3407 liters were evenly distributed over a 11.15 m2 grass plot at the boundary of the Pembroke and Baxter soil types. Waterfall 1, the predetermined subsurface destination for the tracers was equipped with two automatic water samplers gathering twenty-four samples on a time interval basis of two hours. When these are oset, the combined samples allow one hour resolution over two days. is oset pattern also ensures that water samples were collected if one of the autosamplers failed to perform its program. e bottles were retrieved and replaced by clean ones when the forty-eight hours passed. A second location for detection of the conservative sodium chloride tracer was a resistivity traverse which was located at an estimated perpendicular slice of the subsurface between the injection location and Waterfall 1. A system of twenty eight electrodes and Sting R1 by Advanced Geosciences Inc. was used. e electrodes were placed every 4.6 meters for a total distance of 123 meters. is layout allowed for a vertical prole depth of 20 meters. Monitoring of this traverse at two hour intervals before, during, and aer the injection allows a unique visualization of the subsurface hydrologic conditions. e second trace on May 6, 2008 (Julian day 127), was another experiment analyzing the upper ow regimes of the system. e precipitation rate remained the same but this trace involved the injection of 1360 grams of sulphorhodamine b. Prior to the injection, 181,437 grams of dairy manure was spread on the ground surface. e amount of uorescent tracer and volumes of injection and ush water were preserved from the rst trace to allow for comparison of the results. e only independent variable was the antecedent soil moisture conditions. e amount of manure was calculated based on statistics for dierent crops needs for nitrogen. We used estimates for cool season grasses requirements for nitrogen and eventually calculated that 181,437 grams of manure was needed for the 11.15 m2 injection area. Water samples from waterfall 1 were analyzed for dye concentration and fecal coliform. ese samples were collected at Waterfall 1 with automatic water samplers using a sampling interval of een minutes for the initial portion of the trace. is interval was accomplished by osetting two samplers with sampling programs that collect every thirty minutes for twelve hours. e bottles were collected and replaced aer the program was complete and the autosamplers were programmed to continue this sampling regimen until the dye peak had clearly past. e interval was then adjusted to sample every thirty minutes between the two samplers. e bottles used for this trace were cleaned for dyes and sterilized with alcohol to ensure the absence of bacteria prior to the collection of the samples. is is necessary due to the fecal coliform analysis during the trace. Ice was also placed within the bottle carousel and fecal coliform IDEX analysis were ran within 30 hours of collection to follow standard protocol for this type of bacteria analysis.4. Laboratory AnalysisCharcoal dye receptors were brought to the Crawford Hydrology Lab and released into the custody of the technician that was working at the time. Each receptor is removed from the individual plastic bags and washed thoroughly with tap water. e receptor is shaken until the excess moisture is removed and placed on a drying rack covered with aluminum foil. When all of the receptors have been prepared in this fashion the rack full of receptors is placed in the convection oven for drying overnight. e next day the technician removes 0.5 grams of charcoal from each receptor and places the samples in small labeled cups. ese samples are then eluted with Smart solution prior to analysis in the Shimadzu Spectrauorophotometer. Water samples in small vials are ready for analysis aer rinsing and labeling. Results are reported for each dye in parts per billion.

PAGE 261

Earth Sciences 1522 2009 ICS P roceedings 15th International Congress of Speleology e IDEX fecal coliform analysis relies on the fact that certain uorophors will uoresce when combined and incubated with the fecal coliform. For analysis a uorophor named Coililert 18 was used which incubates with the water sample for 18 hours at 44.5 degrees Celsius before the sample may be visually analyzed. Samples with a limited volume were diluted in this process with deionized water and the alternate volume was recorded on the sample for later calculation of the concentration.5. Resultse rst trace began at 11:45 on Julian Day 64, 2008. e conductive salt was rst recognized on the resistivity images on Julian day 64 at 16:00. e resistivity images show a lower resistance over an area measuring approximately 50 meters. e lower resistance anomaly began to gain resistance in the image that followed at 18:00, until the image reached the approximate background resistance at 20:00. is anomaly was not seen as one localized conduit, but is instead represented by a large area within the estimated top, weathered portion of bedrock. e soil/ bedrock interface is estimated to be between 2 and 3 meters below the ground surface. A void is presumed to be beneath the traverse at a 54 meter distance from the beginning of the traverse and is estimated at its greatest height to be about 22 meters below the ground surface. e dimensions of the void cannot be assumed since the feature extends beyond the depth of the measured traverse. Conductivity measurements of the water samples inside the cave rst elevate at 13:00. e peak of the water sample conductivities was at 15:00 with a value of 814 S. e elevated conductivity curve lasted for approximately twelve hours, ending at 23:00. Fluorescein dye concentrations were very similar to the conductivity measurements. e rst breakthrough of dye in the cave was seen at 13:00 with the peak of dye concentrations at 15:00. is peak concentration of dye was 1660 ppb. e main dye curve reached a lower value of 47 ppb at 19:00 before a small rise in concentration to 69 ppb before falling again and resting at values around 10 ppb at 5:00 on Julian day 65. e second trace began at 12:20 on Julian day 127, 2008. e rst signicant rise of dye concentration occurred at 14:15 on Julian day 127 with a concentration of 305 ppb. e dye concentration in the water samples continued to steadily rise until 16:30 where the concentrations reached a peak of approximately 27000 ppb. e dye concentration then began to steadily fall to a concentration of 6093 ppb 2 hours later. e dye concentrations then rose to a peak value of 9359 ppb, 45 minutes later. e concentrations again fell steadily reaching values below 3000 ppb at 21:15, when the values were still consistently dropping but at a lower rate of decline for the rest of the monitoring period. Background levels of fecal coliform prior to the introduction of the manure average 40 MPN. e rst signicant increase in fecal coliform count occurred at 14:15 on Julian day 127 when the MPN count jumped to 373. e MPN counts continued to steadily rise to a peak of 2755 MPN, 45 minutes later (15:00). Fecal coliform reached a second peak of 2481 MPN, 75 minutes aer the rst peak (16:15). Another small peak in the fecal coliform with an MPN of 1860 occurred at 17:30. e last signicant peak in fecal coliform occurred at 20:15 with a value of 612 MPN. Aer this, the fecal coliform MPN values started to steadily retreat towards background levels. 6. Discussion e two traces conducted at Crumps Cave give insight to the transport of solutes and bacteria, as demonstrated by the introduction of salt, dye, and manure. e trace on Julian day 64, 2008, evaluated the epikarstic zone with the additional aid of a resistivity prole. is unique opportunity of geophysical imaging during the trace allowed visualization of the transport process. e tracing experiments have shown that particle transport through the epikarst is rapid when rainfall intensity is very high. While 6.6 cm/hr is a very high value for rainfall intensity, this is a good starting point for understanding how the system will react under high ow conditions. e rst trace results show that solute particles, dye and Figure 1. Fecal coliform and SRB concentrations at waterfall 1aer the injection on Julian Day 127. Notice the two peaks in fecal coliform prior to the intial dye concentration peak.

PAGE 262

15th International Congress of Speleology Earth Sciences 1523 2009 ICS Proceedings salt, are detected at Waterfall 1 seventy-ve minutes aer the beginning of the trace. A peak in these concentrations occurs two hours later at 15:00. e resistivity data collected during the 16:00 hour suggests a widespread mobilization of injected salt passing below the resistivity electrodes. Since the main peak of solutes at Waterfall 1 occurred 1 hour prior to this image, it is assumed that the majority of the sodium chloride mass was transported through localized conduits and preferential ow paths while spatial distribution of the solutes within the epikarst reached its peak aer small conduits and ssures had time to ll with the solution. e results from the second trace highlight the preferential ow paths within epikarst that bacteria travel within. e rst peak of bacteria occurred 90 minutes ahead of the initial dye peak. While the concentration curve of dye resembles a uniform recovery, the fecal coliform curve infers that the bacteria are able to choose dierent paths of greater transmissivity.Acknowledgementse authors would like to thank Stacy Antle, Mark Tracy, Jodi Linsey, the US Department of Agriculture, Homan Environmental Research Institute, and the WATERS lab for their technical support during these experiments.ReferencesAUCKENTHALER, A., G. RASO, and P. HUGGENBERGER, (2002) Particle transport in a karst aquifer: natural and articial tracer experiments with bacteria, bacteriophages and microspheres. Water Science Technology 46, 131 138. BOYER, D.G. and G.C. PASUARELL, (1999) Agricultural land use impacts on bacterial water quality in a karst groundwater aquifer. Journal of t he AA merican Water RResource AA ssociation 35(2), 291. EINSIEDL, F. (2005) Flow system dynamics and water storage of a ssured-porous karst aquifer characterized by articial and environmental tracers. J ournal of HH ydrology 312, 312. FLOREA, L. and C. WICKS, (2001) Solute transport through laboratory-scale karstic aquifers. Journal of C C a ve and Karst Studies 63(2), 59. GPPERT, N. and N. GOLDSCHEIDER, (2007) Solute and colloid transport in karst conduits under low and high ow conditions. GG r oundwater published online, http://info.ngwa.org/gwol/pdf/081882558. pdf. GROVES, C., C. BOLSTER, and J. MEIMAN (2005) Spatial and temporal variations in epikarst storage and ow in South Central Kentuckys Pennyroyal Plateau Sinkhole Plain. Proceedings of the U.S. Geological Survey Karst Interest Group, Rapid City, SD, September 12, 2005. Scientic Investigations Report 5160, 64. HESS, J.W. and W.B. WHITE, (1988) Storm response of the karstic carbonate aquifer of southcentral Kentucky. Journal of HH ydrology 99, 235. JENNINGS, J.N. (1985) K arst GG eomorphology. Basil Blackwell Inc., New York, NY, 293 pp. LEGRAND, H.E. (1973) Hydrological and ecological problems of karst regions. Science 179, 859. RAEISI, E. and G. KARAMI, (1997) Hydrochemographs of Berghan Karst Spring as indicators of aquifer characteristics. J ournal of CC ave and Karst Studies 59(3), 112. RAY, C., T.R. ELLSWORTH, A.J. VALOCCHI, C.W. BOAST, (1997) An improved dual porosity model for chemical transport in macroporous soils. Journal of H H ydrology 193, 270. RYAN, M. and J. MEIMAN, (1996) An examination of short-term variations in water quality at a karst spring in Kentucky. GG r oundwater 34, 23-30. VENI, G., H. DUCHENE, N. CRAWFORD, C. GROVES, G. HUPPERT, E. KASTNING, R. OLSON, and B. WHEELER, (2001) LL i ving with karst: AA agile enironment. American Geological Institute Environmental Awareness (Series 4). WHITE, W. (2003) Conceptual models for karstic aquifers. Speleogensis and EE volution of Karst AAquifers 1(1), 1. WILLIAMS, P.W. (1983) e role of the subcutaneous zone in karst hydrology. J ournal of HH ydrology 61, 45.

PAGE 263

Earth Sciences 1524 2009 ICS P roceedings 15th International Congress of Speleology THE KARST Hydrogeochemical YDROGEOCHEMICAL Features EATURES in IN the THE Catchment ATCHMENT of OF Baiyandong AIYANDONG Underground NDERGROUND Ri I Ver ER Bao AO Jing ING Hunan UNAN CHINAHH E SHIy Y I PEI JIANGUO, XIE YUNq Q IUI I nstitute of Karst GG eology, CAGCAGS and Karst DD ynamics Key LL aboratory, Ministry of LL and and RResource, GG uilin, CC hina 541004 Abstract e planning developing project to Baiyandong underground river system, Xiangxi, Hunan is to resolve the potential water supply for the county. Based on the large numbers of eld measured data on hydrochemistry, this paper gives a brief t analysis on the features and their dynamical changes of the karst springs water, surface water and underground river water. e results reveal that there are two kinds of karst spring waters in the catchment area, one is related to deep source CO2 and distribute along with the regional faults, and their main hydrochemical indexes are controlled by the CO2 concentration. Another is epikarst spring waters which are greatly inuenced by the surface environment and the rainfall, the dynamical change of their hydrochemical indexes are obvious in three dierent measuring periods. e features of hydrochemical indexes of the water in main underground conduit and its branches are in the interposition between of epikarst spring waters and surface water, the data shows that the HCO3 Ca2+ and Mg2+ content, electric conductivity, and so on is controlled by the diluting processes of rainfall, that is, the more the rainfall the low the concentrations of the indexes. Moreover, the series of methods oen used to study karst dynamical systems and also applied in this research have solved some factual problems, theses method character on obtaining large and exact data expediently.

PAGE 264

15th International Congress of Speleology Earth Sciences 1525 2009 ICS Proceedings RECENT OBSERV V ATIONS IN A REMARKABLY DYNAMIC, SULFIDERICH, HYPOGENIC CAV V E IN SOUTHERN MEX X ICOLOUILOUI SE E D D HO HO SE EG G ulf CC oast CECESU U NN ational Park Service 2261 Texas AA&M UU niversity, CC ollege Station, TX 77843-2261 UU SA A LLouise_HH ose@nps.go Observations over the last decade in a two-kilometer long cave in southern Mexico have revealed a remarkably dynamic hypogenic karst system. During this time, both mineral and microbial wall coverings markedly changed and, at two test sites where material had been completely removed, areas have repopulated to the point that the previous damage is no longer identiable. ese observations suggest tantalizing opportunities to directly observe, test, and quantify some forms of speleothem deposition, passage development, microbial participation in both processes, and hydrologic system dynamics. ese observations also document compelling evidence that unique sulfur folia form subaerially, a nding that contrasts with the hypothesized origins of more common calcite folia or mud folia. 1. IntroductionCueva de Villa Luz (aka Cueva de las Sardinas, Cueva del Azufre) is a two-kilometer long cave in Tabasco, Mexico, currently undergoing aggressive physical and biological speleogenetic processes that have resulted in rapid changes observable over the last decade. e mostly air-lled passages are fed by dozens of small hypogenic, suldic, brackish inlets that contribute to an anastomotic, shallow stream which ows through most passages of the cave (Fig. 1). Hose and Pisarowicz (1999) discuss the general characteristics of the cave, and Hose et al. (2000) provide more detailed information on the environment. Over 12 years of intermittent observations in the cave have Figure 1: Map of Cueva de Villa Luz.

PAGE 265

Earth Sciences 1526 2009 ICS P roceedings 15th International Congress of Speleology revealed many remarkable, and some unique, features. Perhaps most curious among these observations have been the rapid change in both mineral (i.e., sulfur crusts, sulfur folia, gypsum paste, and gypsum crusts) and biological (i.e., snottites, biovermiculations) wall deposits as well as changes in water inlet locations. ese observations suggest that subaerial speleogenesis in high-sulde environments may be measurable in a human time scale and that this cave oers an outstanding opportunity to observe and quantify this form of cave development.2. Hydrology and Atmospheric ChemistryTwelve years of intermittent observations in all seasons and rainfall conditions has demonstrated that the discharge from the cave to the surface does not uctuate much from an average base level of approximately 270 L/sec. e stream level will rise a few centimeters immediately following intense, local rain storms but drops within a few hours while the nearby surface river continues to rise ten or more meters. During extended seasonal droughts, the cave stream continues to ow at base level. Stream temperature remains nearly constant at 28o C, as was also reported by Gordon and Rosen (1962) for conditions in 1946. Short-lived temperature drops of several degrees immediately follow heavy rains when meteoric water enters the cave through skylights and ssures. Initial mapping identied about 20 inlets (subterranean springs) feeding the cave stream. Over the ensuing decade, more than twelve additional inlets have been identied. e mapping team probably failed to recognize some of these inlets. However, two large upwellings appeared within the last four years at the base of the stairs, in the Main Entrance room where teams commonly stage for entry and clean up following exploration. ese inlets have clearly changed since initial mapping in the latter half of the 1990s. Other inlets in the cave have decreased their discharge over the past 20 years. Two ephemeral environmental factors appear to impact many of the changes described below. First, water chemistry at the various inlets varies temporally. In particular, changes in dissolved gases, particularly hydrogen sulde, at the inlets result in changes in atmospheric hydrogen sulde levels throughout the cave. Second, increased rainfall almost immediately results in increased inltration of meteoric water into this shallow cave. is inltration provides lms of well-oxygenated, fresh water to the cave walls. e moisture mixes with the atmospheric hydrogen sulde and facilitates several physical and biological processes, resulting in seasonal or temporal changes in the interior of the cave.3. Changing SnottitesStalactite-like deposits comprising mostly microbial colonies that drip potent sulfuric acid (pH typically 0) were rst recognized and named snottites by Pisarowicz (1988a; 1988b). Initially considered unique, snottites have since been reported in a few locations elsewhere, including nearby Cueva de Luna Azufre and the Italian Frasassi and Grotta Nuova del Rio Garrafo Caves (Hose, 2005; Macalady, 2007). Typically suspended from selenite crystals and commonly following spider webs, the ephemeral snottites can grow several centimeters in 24 hours. While some specic locations in the cave commonly hold snottites, particularly the area called Snot Heaven, snottite distribution, abundance, location, and size vary dramatically from visit-to-visit, and even day-to-day. Empirical observations indicate that high atmospheric sulde levels throughout the cave as well as abundant surface water inltration (i.e., recent rains) promote a healthy crop of snottites.4. BiovermiculationsHose et al. (2000) coined the term biovermiculations to represent irregular, discontinuous wall mats of mostly organic material that resemble clay vermiculations (Hill and Forti, 1997). Northup (Hose and Northup, 2004) recognized fungi and different types of bacteria including A A c idobacterium and actinobacteria in Villa Luzs biovermiculations. Jones et al. (2008) found an extremely diverse community of 48 representative phylotypes in Frasassis biovermiculations. Observations in Villa Luz have shown biovermiculation deposits to be dynamic with significant, visually observable changes occurring over a few months. While changes in vermiculations have been noted in other caves (Jeannel and Racovitza 1929; Parenzan 1961; Jones et al., 2008), we now have a 10 year continuous photographic record of changes at one biovermiculation site in Villa Luz (Fig. 2). All surface coating was removed by scraping with a sterilized knife blade to expose a 3 cm by 3 cm square of bare bedrock limestone in 1999. The changes to that site and the surrounding growth of biovermiculations have been recorded on an approximately annual basis for the last 10 years. By spring 2008, the test square was no longer distinguishable. Biovermiculations appear to be pioneer organisms that colonize nearly all bare limestone in the cave that is not regularly washed by inltrating surface water or the cave stream. Biovermiculation excretions (pH ranging from 3.0 to 7.8, but typically in the lower, more acidic end) convert the limestone to gypsum. Initially, a veneer of gypsum

PAGE 266

15th International Congress of Speleology Earth Sciences 1527 2009 ICS Proceedings forms. Interestingly, the biovermiculations will shrink and, in some cases, disappear during dry seasonal conditions to leave subtle ghosts that appear to be gypsum powder and clay. Desiccated-looking biovermiculations at the test site during a particularly rainless period (spring 2007) appeared to be coated with a white powder that was interpreted as gypsum. When the biovermiculations plumped up the following year, the surface mineral deposit was no longer evident. Villa Luzs biovermiculations display a wide variety of colors and forms. Locally, dense biovermiculation sites grade into massive mats. When biovermiculations lie below a red clay seam in the bedrock limestone, or where surface clays and silts inltrate from the surface, the growths are commonly brilliant red and prompted the name ragu.5. Sulfur Crust/Mammillaries and FoliaA subaerial sulfur crust that resembles cave mammillaries and folia coat portions of the above-water walls in the most remote passage, named e Other Buzzing Passage. e sulfur deposits begin approximately 15 cm above the streams standard level and extend several meters up. Direct and instrumented observations along with indirect evidence (e.g., highest level of clastic material) provide compelling evidence that the stream in this area has never risen more than about 12 cm in recent times. e sulfur deposits are most dense on the wall within about one meter of the stream and become progressively sparse upward. e coatings display crystal faces, commonly about 1-1.5 cm across, but have a mushy consistency. P.J. Boston (personal communication, 2005) reported abundant microbial material within the deposits. e morphology of the sulfur folia strongly resemble calcite and mud folia (Davis, 1984; Green, 1996; Hill and Forti, 1997) but the sulfur form has not been reported outside of Villa Luz. Sulfur folia, like calcite and mud folia, are found on vertical and overhanging walls in hypogenic caves. Adjacent, upward facing slopes display mammillary-like crusts of the same composition. All sulfur and mud folia have relatively small diameters of one to ve centimeters, while some calcite versions are much larger. Sulfur folia occur in patches, particularly higher on the walls. is diers from the typically contiguous displays of calcite and mud folia. In addition, while there has been no agreement on the specic mechanism(s) of folia formation, there has been consensus that both calcite and mud folia form subaqueously. e sulfur folia are clearly subaerial deposits. e massive sulfur deposits grow on selenite crystals, which ubiquitously coat the walls in this area. Some of the folia engulf the selenite (Fig. 3). In 1999, a visitor unintentionally put his hand on the wall and broke o a large (approximately 15 cm x 20 cm) section of the sulfur deposit along with some of the selenite. Most of the chunk of sulfur broke up and (presumably) dissolved in the underlying stream within minutes. is careless, but ultimately fortuitous, act provided a test laboratory to demonstrate the subaerial origin of these sulfur folia. Over the following decade, the site developed a new coating of sulfur. By 2007, incipient sulfur folia were observed and by 2009, the entire area had recovered and distinct folia had developed (Fig. 4). e water inlets that rise in e Other Buzzing Passage have the highest concentrations of H2S in the cave, and there is no free oxygen (Hose, 2001). is relatively conned area also consistently contains the highest concentration of atmospheric hydrogen sulde (H2S), which has rarely been measured at less than 50 ppm and commonly exceeds 70 ppm when visitors are present. ese observations lead to the hypothesis that the warmer, suldic air rises from above the suldic inlet (Lake of the Yellow Roses), mixes with the higher and slightly cooler, oxygenated air Figure 2: Bioermiculations at a test and monitoring site in the Ragu Passage. a. Site in April 1999, immediately aer the 3 cm by 3 cm test patch was formed by scraping away all bioermiculations and any surcial gypsum coating with a sterile knife blade; b. June 2001 close-up photograph shows the test patch during the happy face stage of bioermiculation re-population; c. March 2009 photograph shows the test patch area has completely blended with the rest of the bioermiculation growths. It also shows dramatic changes to the general area over the past decade. (White arrows point to the test patch site in all photos.

PAGE 267

Earth Sciences 1528 2009 ICS P roceedings 15th International Congress of Speleology that circulates into the area from the surface, resulting in the subaerial precipitation of sulfur crystals. e sulfur may precipitate directly on the surface of selenite crystals on the walls or in the air, then dri down onto the walls to form the distinctive folia form. Whether the microbial inhabitants of the sulfur deposits participate in the process or exploit an attractive habitat is unclear with our present knowledge.6. Gypsum Paste and U-LoopsA gypsum paste forms as a weathering rind on some strata in Villa Luz cave. The material has the consistency, texture, and appearance of toothpaste with a typical pH of 3. This rind results from limestone bedrock converting to microcrystalline gypsum and is common in sulfidic caves worldwide. Villa Luzs gypsum coatings, both selenite and paste, are a step along a dynamic process of passage enlargement in which the gypsum regularly sloughs from the walls and ceiling (Palmer and Palmer, 1998; Hose et al., 2000). This process is so active that investigators have experienced spontaneous and, apparently, natural paste-falls from the cave ceiling on several occasions. e gypsum paste also ows as a viscous liquid and can form small stalactite-like features. Common spider webs in the cave can direct owing gypsum, resulting in U-loops of gypsum paste (Fig. 5). While previous researchers have proposed snottites as a possible model for the origin of calcite u-loop speleothems (Northup et al., 2000), these U-loops of gypsum paste provide an attractive alternative. e conversion of gypsum to calcite is well-documented in natural systems and seems more likely than conversion of an ephemeral, mostly microbial material to long-surviving calcite.7. ConclusionsTen years of observations have demonstrated that Cueva de Villa Luz is a dynamic, high energy system producing rapid changes that are unparalleled in most limestone Figure 3: Sulfur folia engulng selenite crystals. Figure 4: Sulfur folia and crust in March 2009. Ten years earlier, all sulfur was remoed om this site when a careless visitor broke o the previous deposit of sulfur. Only the underlying gypsum crystals remained at that time. Subaerial deposits of sulfur progressively developed until the damaged area is hard to identify. Sulfur folia were rst noted at this site in 2007. Figure 5: U-loop of gypsum paste. Microcrystalline gypsum has owed down and engulfed spider webs, forming uloop features.

PAGE 268

15th International Congress of Speleology Earth Sciences 1529 2009 ICS Proceedings cave systems. Preliminary records of growth and changes in folia deposits over only a decade provide tantalizing possibilities for better understanding the mechanisms and conditions leading to folia development. More detailed studies of dynamic biovermiculation growth may also lead to a better understanding of the role biovermiculations play in speleogenesis as well as possibly providing clues to the origin of more common clay vermiculations. More generally, this active, suldic cave system, balanced with an environment that allows direct exploration and study by humans, provides unparalleled opportunities to learn about hypogenic speleogenesis in a redox environment dominated by accelerated subaerial processes.Acknowledgmentsis work over the last twelve years has been supported by National Geographic Society, National Speleological Society, Richmond Area Speleological Society, Japanese Broadcasting Corporation, Chapman University, Westminster College, TLC, and Nick Davis Productions. Jim Pisarowicz deserves credit for rst recognizing the broad scientic importance of the cave and ensuring that a multidisciplinary team of scientists formed a project and explored the cave. Bob Richards serves as the projects excellent cartographer. Many cavers and other Villa Luz researchers assisted in these observations over the decade and their help has been appreciated, including Kathy Lavoie, Dave Lester, Laura Rosales-Lagarde, Doug Soroka, Lynn Kleina, Adam OConnor, and Will Heltsley. Correspondences with Donald G. Davis and Arthur N. Palmer about folia helped inform this paper.ReferencesDavis, D.G. (1984) Mysteries in mud: Ancient frost-crystal impressions and other curiosities in Cave of the Winds, Colorado. Rocky Mountain Caving, 1/3, p. 26. Gordon, M.S. and D.E. Rosen (1962) A cavernicolous form of the Poeciliid sh Poecilia sphenops from Tabasco, Mexico. Copeia, 2: 360. Green, D.J. (1996) e origin of folia [abst]. National Speleological Society Convention Program with Abstracts, Salida, Colorado, August 5, p 41. Hill, C.A., and P. Forti (1997) Cave minerals of the World, 2nd Edition. National Speleological Society, Huntsville, AL, 463 pp. Hose, L.D. (2001) Hydrochemical characteristics of a unique sulfur-rich karst system in southern Mexico [abst.]. GSA Abstracts with Programs, 33/6. Hose, L. D. (2005) ICS Pre-congress eld trip: H2S caves in central Italy. NSS News, December 2005, pp 10. Available at http://www.geosc.psu.edu/~jmacalad/ NSSNews_Dec05_article.pdf Hose, L.D., and D.E. Northup (2004) Biovermiculations: Living, vermiculation-like deposits in Cueva de Villa Luz, Mexico [abstract]. Journal of Cave and Karst Studies, 66/3, p 112. Available at http://www.caves. org/pub/journal/PDF/V66/v66n3-Abstracts.pdf Hose, L.D., and J.A. Pisarowicz (1999) Cueva de Villa Luz, Tabasco, Mxico: Reconnaissance study of an active sulfur spring cave. Journal of Cave and Karst Studies, 61/1, pp 13. Available at http://www.caves.org/ pub/journal/PDF/V61/v61n1-Hose.pdf Hose, L.D., A.N. Palmer, M.V. Palmer, D.E. Northup, P.J. Boston, and H.R. DuChene (2000) Eects of geomicrobiological processes in a hydrogen sulderich, karst environment. Chemical Geology: Special Geomicrobiology Issue, 169, pp 399. Jeannel, R., and E.G. Racovita (1929) Enumeration des grottes visitees 1918. Arch Zool. Exper. Gener., 68/ 2, pp 293. Jones, D.S, E.H. Lyons, and J.L. Macalady (2008) Geomicrobiology of biovermiculations from the Frasassi Cave System, Italy. Journal of Cave and Karst Studies, 70/ 2, p. 78-93. Available at http:// www.caves.org/pub/journal/PDF/v70/cave-70-0278.pdf Macalady, J.L., D.J. Jones, and E.H. Lyon (2007) Extremely acidic, pendulous cave wall biolms from the Frasassi cave system, Italy. Environmental Microbiology 9, 1402. Northup, D.E., C.N. Dahm, L.A. Melim, M.N. Spilde, L.J. Crossey, K.H. Lavoie, L.M. Mallory, P.J. Boston, K.I. Cunningham, and S.M. Barns (2000) Evidence for geomicrobiological interactions in Guadalupe caves. Journal of Cave and Karst Studies, 62/ 2, pp 80. Palmer, A.N. and M.V. Palmer (1998) Geochemistry of Cueva de Villa Luz, Mexico: An active H2S cave. Journal of Cave and Karst Studies, 60/3, p 88.

PAGE 269

Earth Sciences 1530 2009 ICS P roceedings 15th International Congress of Speleology Parenzan, P. (1961) Sulle formazioni argillose-limose dette vermicolari. Atti International Symposium, Varenna, 1, pp 120. Pisarowicz, J.A. (1988a) Revenge of Chac: 1988 in Tabasco. Association of Mexican Cave Studies Activities Newsletter 17: 129. Pisarowicz, J.A. (1988b) Southern Mexican caving-Tabasco 1988. Rocky Mountain Caving 5/3, pp 25.

PAGE 270

15th International Congress of Speleology Earth Sciences 1531 2009 ICS Proceedings THE BIG CAV V E AND CHAMBERS IN POYUE UNDERGROUND RIV V ER DRAINAGE IN SOUTH CHINA AND ITS CONTROLLING FACTORS BAO AO JIAN IAN HUANG HUANG YUANHAI UANHAI ZHANG ZHANG WEIHAI EIHAI CHEN CHEN I I nstitute of Karst GG eology ,C C hinese AA cademy of GG eological Sciences,G G uilin, P RR CC hina  Based on the statistics to the caves developed in Poyue underground river drainage, the authors found that there exist several large passages and 25 huge chambers with the area over 5000 m2 in the karstic terrain developed the Poyue underground river system. e authors analyzed the controlling factors, such as the geological aspects and the climate and hydrological factor. e results show that the allogenous water may play a key role in the formation of the big passage and huge chambers.1. IntroductionAbundant huge passages and chambers have been discovered in the karst terrain of the Poyue underground river system drainage. e features do not share the same developmental history with the large caves in the world. In order to explain the speleogenesis, we discuss the characteristics of the large passage and chambers and their developmental history to nd the controlling factors. 2. e World Class Big Caves and Chambers in Poyue Underground River DrainageAs one of the four largest underground river system drainage in Guangxi, South of China, Poyue underground river system drainage covers an area of 1484.55 km2, including the 708.35 km2 within the clastic rock mountains that surround the 776.20 km2 carbonate terrain. e subterranean river system is composed of several ramications and is recharged by inows that originate from the non-karstic rocks along the contact zone between the carbonate rock and clastic rocks, with a mean ow rate of 5796.8 L/s at its resurgence in dry season. As the source of the underground river, they disburse in the karst terrain of the Poyue underground river system drainage. e caves featured by their huge passages and big chambers may be among the largest caves in the world. According to the geologic and geomorphic analysis and the survey results of cave exploration made by several sino-foreign joint expeditions since 1988 ,the authors deem that more than 500 caves had been generated in the drainage, and most of them may be big caves or be developed with big chambers.2.1 e large passagesAt present, there are several classications of cave size. Generally, we can determine the size of a cave from some basic parameters, such as the length, width, height, and bearing, obliquity,etc. But, it is not easy to contrast the size between/ amongst two or more caves to determine which one is bigger from the survey parameters mentioned above. On scientic research, one must compare exactly cave size so as to contrast their dynamic condition of formation. We prefer to use the parameter specic volume (J.Gunn, 2003) to describe the size of a cave for it is more scientic for comparing dierent cave with dierent lengths. In addition, we still appreciate other parameters. But, in accordance with our primary research results (Zhang et al, 2008), the specic volume is the key parameter to describe the size of cave, especially the cave developed in the big drainage of subterranean river system. Here, we give the denition of the specic volume as the ratio of cave volume to horizontal length. Intuitively, the big passage of the caves in Poyue underground river system hit us between the eyes for most of the passages are much greater than 20m in wide and high. From the data on Table 1,we found that the specic volume of the caves varies from 57 to 13380 m3m-1 most of them is more than 600 m3m-1. In order to evaluate the size of these caves, owing to the paucity of the similar data, we select two gigantic caves to compare. One is the Skocjanske jame caves in Slovenia and another is Deer Cave in Malaysia. e total length of the former cave is 5800 m. Only two section Hankejev kanal and Mariniheva jama are the main big passage. Hankejev kanal is a gravitational canyon passage, 1 km long, with the maximum width of 55 m, height of 98 m, its volume is 2.000.000 m3. So, the specic volume is 2000 m3m-1. Mariniheva jama is a passage with the length of 200 m, maximum width of 65 m, maximum height of 37 m, and volume of 230,000 m3, then we can get the specic volume of 1150 m3m-1. Deer Cave in Malaysia is famous for the 162,700 m2 chamberthe largest chamber in the world. It is 2160m long with the volume of 13,170,000 m3, so its specic volume is 6097.2 m3m-1, the big digital is owing to the gigantic Sarawak Chamber, which is 600*300* 100 m. So, most of the caves in the Poyue underground river system drainage can compare favorably with the largest caves in the world.

PAGE 271

Earth Sciences 1532 2009 ICS P roceedings 15th International Congress of Speleology 2.2 e huge chambersBesides the big passages, there still exist lots of enormous halls developed in the caves. A cave chamber is a special passage in a cave that is prominently larger than the adjoining areas. e results of statistics on 50 chambers over 5000 m2 from 14 caves developed in the upriver reach of the Poyue underground river system drainage show that half of the chambers are over 10,000m2 in area. Four of them are more than 20,000m2 (Table 2) and six of them are over 25,900 m2, accounting for one-sixth of the 24 worldclass cave chambers over 25,900 m2 in area. So, we can call the terrain located in Fengshan county one of the hugechamber-area and the densest-distributing district of cave chamber in the world. For example, there are 24 chambers over 5,000 m2 in area distributed within the 37.9 km of passages in the Jiangzhou System, which is currently the third longest explored cave in China. Nine of the chambers are over 10,000 m2 in area.3. e Geological Setting of the Huge Passage of the Cavese research area is one of the isolated carbonate platforms in the Youjiang Depositional Basin formed during the HesianIndo-Sinoian Stage. e isolated carbonate platform was surrounded by big faults. Continuous deposition of the 3156 m thick sequence of carbonate rocks took place from Middle Devonian to Early Triassic. During the middle Triassic, the whole area was again submerged into a marine environment, then was covered by the 3000 m thick sequence of clastic rocks. During these stages (D2-T2), tectonic activity was not prominent, dominated by smallscale vertical activity. Since Late Triassic, all strata was folded Name of Cave or Section Horizontal Length /m Depth /m Mean Height /m Mean Width /m Slobe / Floor Area /m2Cave Volumn /m3Surface Area /m2Surface Volumn /m3Specic Volumn /m3.m-1Specic Length /km.km-2Area Ratio /% Volumn Ratio /% Jiangzhou System e whole37939.728149.727.69.71141188.928944498264123987421976952762.91.44.320.39 Section A11349.2212.423.12610.5334875.37671950.36760888.11436009686761.74.950.53 Section B4705.684.922.629.55.8124504.74100804.72170575.1184273804871.52.25.742.23 Section C333764.953.415.3774806.8783374.5407686.826470030.5234.88.218.352.96 Section D359276.659.5296.5127787.53024965.41356827.8103980874842.12.69.422.91 Section E4366.1179.540.341.211202327.17397150.31068647.81918172301694.24.118.933.86 Section F6822.2230.132.225.811.2216855.74549134.12033263.7467886195666.83.410.670.97 Section G252.922.67.39.612.41415.9232776261.8141294.69240.422.6116.47 Section H500.280.525.942.415.714363.6899790.934948.22813413.81798.914.341.131.98 Other Caves Dadong1092.911719.19.710.413006102488.6100354.411740816.193.810.912.960.87 Shendong673.533.620.317.85.413971213185.2148787.64998678.5316.54.59.394.26 Sifang 1532.85814.420.97.437087.3671091.3242554.214074975.4437.86.315.294.77 Dongli749.155.323.826.84.528659537855.3133171.173673334.97185.621.527.3 Mayoni1625.65821.343.56.268091.33076728.7544860.831625772.61892.7312.59.73 Maguai738.238.518.915.312.216200.9173076.8737642839192.4234.51021.9610 Green River3235.290.410.6376.8134639.14423095.916303171455112531367.228.363.04 Longshi Sha 841.5169 2023.918280.6336812.1117927.61930018.8400.37.115.51.69 Dashan1447.117222.536.213.964067.11892014.523134339795633.91307.56.327.694.75 Crystal794.440.219.320.3922006.9327930.988610.83560685.4412.8924.849.21 Cemetery1319.943.211.97.66.98502.175892.5401940.417351102.957.53.32.120.44 Feilong192.5 7055 109302575600 13379.7 Shegeng220 68110 242002584900 11749.5 Mawang6620 10050 294000032234400 4869.2 Yuanyang260 3090 225001071000 4119.2 Chuanlong370 42120 415001453200 3927.6 Yunfeng514 2533 386001007400 1959.9 Yulong 1350 4040 613502454000 1817 Siccily 1021 3036 27950839430 822.2 Poxia 1112 3420 22400727700 654.4 Table 1: Some Representative Passages of Caves Developed in Poyue Underground River System Drainage.

PAGE 272

15th International Congress of Speleology Earth Sciences 1533 2009 ICS Proceedings No. CHAMBER NAME CAVE NAME LENGTH /m WIDTH /m AREA /m2MAXIMUM HEIGHT /m MEAN HEIGHT /m VOLUMN /m31Fengshan museumChuanlong37296-14041,500 45 30 1,245,000 2NantianmenMawang 32088-26038,400 200 150 5,760,000 3ZhuanwanGantuan 32060-16031,200 87 50 1,560,000 4Luxi Green river240100-11026,220 105 90 2,359,800 5Yuanyang Yuanyang26045-13029,250 130 107 3,129,750 6Dongkou Gantuan 37545-13125,900 135 80 2,072,000 7ChuandongShegeng 22078-14224,200 76 70 1,694,000 8YangguangYunfeng 22053-12020,960 32 20 419,200 9DongdongMayoni 21080-12020,450 100 60 1,227,000 10Creaked Mud Jiangzhou system 26440-8318,500 11Maguai Maguai 34045-7417,000 50 30510,000 12Nanfang Mawang 24040-8816,800 165 150 2,520,000 13Longhuai zoulang Jiangzhou system 20050-9016,500 55 40 660,000 14Xiepo Jiangzhou system 22836-9616,461 65 40 658,440 15Moduan Gantuan 15512015,000 54 30 450,000 16S.Junction Jiangzhou system 16524-13014,920 17Dongfang Mawang 20060-8014,500 78 40 580,000 18Mud Dongli 26040-6514,300 19Zhonglian Jiangzhou system 21258-9513,992 95 60 839,520 20Xiagu Jiangzhou system 22531-6813,050 54 40522,000 21LiangdongLiang 27127-6712,500 47 35 437,500 22Dinner Time Jiangzhou system 28414-5410,792 23Wuyan Green river15040-8510,670 49 40 426,800 24Colossus Jiangzhou system 18832-7010,524 25Terragotta Jiangzhou system 16444-9010,168 26Houduan Liangfeng16040-609,650 27Gloopsville Junction Jiangzhou system 16024-579,600 28Minaret Stal Jiangzhou system 13038-869,320 29Hijack Junction Jiangzhou system 21418-569,202 30Longhuai Jiangzhou system 13548-949,108 48 30 273,240 31MonumentaleSiccily 13045-809,000 60 42378,000 32ZhongduanYunfeng 12050-1048,800 33CauliowerYulong 16030-808,800 34Stal Graveyard Jiangzhou system 12237-1108,549 35Gours I Yulong 16030-738,500 36Jianduan Liangfeng12060-808,500 37Football Pitch Jiangzhou system 13528-678,100 38Pink Pyjamas Jiangzhou system 11025-957,920 39Fishing Pitch Jiangzhou system 17614-707,392

PAGE 273

Earth Sciences 1534 2009 ICS P roceedings 15th International Congress of Speleology and faulted by the Indo-Sinoian Tectonic Movement and the whole structural framework was formed. Furthermore, it was lied as a land environment and suered from largescale weathering for eighty million years. Eventually, during the Middle Jurassic, the clastic cover above the isolated carbonate platforms was eroded and exposed the carbonate rocks. According to the paleo-geomagnetics research result, the latitude was one to two degrees lower than now. Karstication in the tropical climate began but the karst forms were later erased away. In the Cretaceous Period, a large lake basin formed in Guangxi in the arid or semi-arid climate conditions, depositing a set of red or morassic sediments, but it has since eroded in research area. From the Cenozoic on, a tectonic history of intermittent crustal upli and karstication may be divided into three stages: e rst stage is from the latest Cretaceous to Early Tertiary when the crust was in steady state, aer long-term erosion, an old land surface leaning southeastward was formed and only a few part of them can be found on the oldest the clastic peaks because most of it has been erased away. But, the present drainage system (including the surface and underground river system) had not been created at that time. e stream was in a disperse state. e second stage is called the Fengcong Stage (Peakcluster Stage) and dates back to neo-Tertiary to Eocene. e Qinghai-Tibetan Plateau rose rapidly due to the collision between the Indian Plate and the Euro-Asian Plate. Development of the lake basin in the whole Guangxi Region was nished and erosion restarted. e climatic condition was transformed into a regional monsoon climate by the planetary wind system, resulting in a humid tropical and sub-tropical climate. e activity of the surface stream was strengthened and the ancient hydrological net was generated. en the fengcong (peak-cluster) was formed in the bare carbonate area as downcutting led to the formation of underground rivers and caves. In this stage, the geomorphic forms were dominated by the fengcongdepression/valley owing to the vertical movement of water. e third stage begin in uarternary. Yunnan-Guizhou Plateau continued to be lied, Hongshuihe riverthe key erosive baseincised more deeply, the research area was relatively uplied, the caves formed previously were lied up gradiually, their passages were enlarged due to the synergetic balance amongst the velocity of downcutting surface and underground rivers, and erosion (including dissolution) took place.4. Controlling Factors to the Formation of the Big Caves and eir ChambersA cave is the combination of dissolution, erosion, and collapse (Bogli,1980). In south China, most caves belong to the rainwater-type of speleogenesis. It is generated by the rain water and its secondary seepage ow (Zhu, 1988). According to our research, the caves in the research area originated with underground rivers in the phreatic zone (Ford, 1988). So, the water may play an important role in the formation of the big passages of the caves. e controlling factors of the development of the huge passages and chambers may comply with those controlling the growth of relatively small caves, such as base level and dynamics condition. 40Zhongjian Mawang 12040-757,200 41Xuezhu Jiangzhou system 9056-937,146 39 20 142,920 42Piste Jiangzhou system 15616-567,020 43Entrance Siccily 9045-776,400 50 40256,000 44Shaobing Jiangzhou system 9426-806,288 55 40 251,520 45Shiqiao Jiangzhou system 15030-606,145 95 60 368,700 46Rukou Hei 10555-606,000 31 30180,000 47S.Herb Garden Jiangzhou system 9722-905,459 48Gours II Yulong 11520-805,400 49Cave Pearl Jiangzhou system 9045-645,035 50Tapehandle Junction Jiangzhou system 14823-545,030 Table 2: Some Cave Chambers in the Poyue Underground River System Drainage.

PAGE 274

15th International Congress of Speleology Earth Sciences 1535 2009 ICS Proceedings 4.1 Rock conditione bedrock of the cave in research area come mainly from Carboniferous and Permian strata. e old, pure, and solid carbonate rock are more favorable for karstication, with the consecutive thickness of 1077 m, than so, young, and high-void carbonate rocks, such as those along the Mediterranean Sea, London Basin, and the East Europe (most of them are Mesozoic), those in Nallorbour Plain, Paris Basin, and Caribbean area (most of them are Tertiary), so there exist a huge passage in the research area. e rocks are strong enough to support and maintain any landforms, even large scale, for a long time.4.2 Geologic structure and hydrogeological conditionFaults and joints are favorable for the transportation of water and dissolution of carbonate, facilitating cave passage development. Furthermore, a passage collects more and more water ow from the contiguous area, making the Poyue underground river more and more competent to carry more and more substance. Eventually, the passage get larger and larger. e junction of dierent ssures is favorable for the big chambers for the mixing-corrosion by two or more ow. e intermittent upli of the crust may be favorable for the velocity balance amongst the crust upli, dissolution, and erosion at some part of the passage to form the gigantic hall and enormous passage. e most important point must be emphasized that the structure of karst and non-karst area (i.e., the relative low density of karstic terrain) is surrounded by the clastic mountains comprising the materials bearing the pyrite and other sulfate minerals depositing mainly in deoxidizing environment in deep water .e oxidizing of sulfate may supply more acidic matter to make the water more acidic, the karstication may be strengthened. Besides, the contact zone between the carbonate rock and non-carbonate rock is always the regional fault where some sulfate-bearing hot liquids were easily transported and stored or enriched as ores. e sulfate can play the same role in the formation of big passage. In addition, the allogenic ow generated from clastic rocks is always more acidic and erosive, the analyzing result of water sample has attested to this. 4.3 e synergetic water heat condition of monsoon climatee major karst forms in the research area were formed consecutively from Pliocene to uaternary owing to the humid and warm, tropical and subtropical climate. According to the analyzing result of spore and pollen in neighboring areas, heat tolerant members account for over half of the sample, which show that it was a humid and rainy climate, and the temperature in the Pliocene was higher than present. In the uaternary, the climate was still humid subtropical and tropical in spite of the uctuation of ice period. At present, a subtropical monsoon climate with mean annual precipitation of 1400 mm and mean annual air temperature of 17oC prevails. e hot season corresponds with the rainy season. e synergetic water-heat climate condition is favorable for the dissolution of carbonate rock.5. Conclusionse Poyue underground river system drainage is the premier place for the development of the largest cave chambers and passages in the world. e enormous passages and gigantic chambers are a production of the Poyue underground river system under special geologic structural and hydrogeologic conditions. In the world, most of the largest passages developed in the areas with the most abundant rain precipitation. For instance, rain falls up to 5000 mm in Malasia. So, it is signicant for scientic research on the genesis of big caves and for cave expedition.AcknowledgementsWe are grateful for the survey data of the CHINA CAVE PROJECT lead by Mr. Andy Eavis and Professor Zhu Xuewen, especially the work of Mr. Bensly Bluce and Ged Campien. We also thank Mr Andrej Mihevec, who supplied the data of Skocjanske jame caves.ReferencesBogli, A (1980).Physical Speleology[M]. Berlin,Heidelberg,New York,Springer-Verlag. Ford, D.C. (1988).Characteristics of Dissolutional Cave System in Carbonate Rock.Paleokarst(James N.P. (ed.)[C].Berlin,Heidelberg,New York,SpringerVerlag ,25 Gunn, J(2003).EE n cyclopedia of CC aves and Karst Sciences[M] New York,London:e Taylor and Francis Group, 524 Zhang, Y.H., D.S. Han, and Y.D. Deng (2008).Explanation of Measurement to Cave Forms.CC a ve EE xploration,R Re search,E E xploitation and Protection(W.H. Chen, D.H. Zhu, and X.W. Zhu, ed.) [C].Geological Publishing House,Beijing ,87(in Chinese) Zhu, X.W., Y.H. Zhang, and R. Zhang, (1996).Systematic Evolution of the Underground Rriver Cave[J].C C ar sologica Sinica. 15(1-2):157

PAGE 275

Earth Sciences 1536 2009 ICS P roceedings 15th International Congress of Speleology Twenty WENTY years YEARS of OF monitoring MONITORING tests TESTS and AND e E Xperiments PERIMENTS in IN the THE Milandre ILANDRE ca CA Ve E Jura JURA Mountains OUNTAINS Switzerland WITZERLAND : Inputs N PUTS for FOR karst KARST hydrogeology HYDROGEOLOGY JEANNIN PIERREYv V Es S Swiss II nstitute of Speleology and Karstology, SI I SKA A PO O Box 818, CHCH -2301 LL a CC haux-de-Fonds, Switzerland (SI I SKA A), www.isska.ch Abstract e Milandre cave is a 10.5 km long cave with an underground stream, which can be followed from the spring upstream for 4.6 km (3 km as the crow ies) with an entrance at both ends. e cave is one of the most decorated ones in Switzerland and the spring is used as a drinking water supply by the community. e construction of a freeway on top of the cave implied a lot of studies concerning engineering issues, groundwater protection and cave protection. Fundamental research on karst hydrogeology and geophysics was also carried out. Concerning karst hydrogeology, this research makes this site one of the most studied in the World. A really interesting aspect of this test-site is that many experiments and measurements could be undertaken to verify in the eld, whether the real karst system behaves the way we expect it to from spring data interpretation of from models. For instance the respective eect of soil, epikarst, vadose zone, cave stream and phreatic zone on the generation of hydrographs, chemographs and isotopic variations could be investigated with a high degree of detail. Also, the realization of various experiments (e.g. tracer tests or inltration tests) and their related interpretation models could be compared to models derived from other types of data (e.g. chemographs or isotopic variations) and to direct observation of the expected processes in the cave. is leaded us to revise our own interpretation models several times along those 20 years of experience. A critique of many models found in the literature is being prepared with this sum of experience. For instance the storage component feeding the system during draught periods has been much better understood, as well as processes occurring inside the conduit system, which are responsible for many unexpected variations of the water quality. As an example, mechanisms controlling particle transport and water turbidity could be understood in a quite dierent way than what can be expected from the literature. Observation and development towards cave biology and its relation to the respective parts of the aquifer (soil, epikarst, vadose zone, underground stream and phreatic zone) is one topic we just started to develop. Collaboration in this domain is welcome. is cave could be considered as a real laboratory and could be used in the future for experiments at a more international and interdisciplinary scale.

PAGE 276

15th International Congress of Speleology Earth Sciences 1537 2009 ICS Proceedings Impact MPACT assessment ASSESSMENT of OF a A tunnel TUNNEL on ON two T WO karst KARST springs SPRINGS Flims LIMS Switzerland WITZERLANDPIERREYv V Es S JEANNIN1, PHILIpp PP H H Us S ELm M ANN N1, E E RIC WEb B ER1, A A NDREs S WILDb B ERGER2 1Swiss II nstitute of Speleology and Karstology, SI I SKA A PO O box 818, CHCH -2301 LL a CC haux-de-Fonds, Switzerland, www.isska.ch2Bureau DDr. on Moos, GG eotechnics, Bachofnerstrasse 5, CH-8037 Zrich Switzerland Abstract In October 2002 a karst conduit was cut by the road tunnel next to the Flims village in the eastern Swiss Alps. A discharge rate of 800 l/s was measured shortly aer. In winter the discharge rate decreased down to about 200 l/s, but increased again with the snow melt. Fieen months later (in February 2004) the Lag Tiert karstic spring located 2.2 km away dried up. For more than 40 years, the spring has been used for an electricity power plant and it never dried up. Is the tunnel responsible for this drying up? Why did the spring dry up only 15 month aer the tunnel hit the conduit? e year 2003 was extremely hot and dry in central Europe, could it be the reason for this drying up? How much water does the spring miss? e owner of the electricity power plant asked us to answer those questions. Another spring was later identied to belong to the same system. e Lag Prau Pult spring is an overow spring of the Lag Tiert feeding a beautiful lake which is very attractive for the tourists in summer. e community asked us to study the eect of the tunnel on the lake. We asked the civil engineers building the tunnel to make a dam in order to be able to raise the pressure of the karst conduit in the tunnel and to see if the springs respond to such pressure steps. A large series of pressure tests were run by low and high water situations between November 2004 and September 2008. During the rst tests (at low water stage) Lag Tiert spring responded 1 minute aer the pressure steps. e same tests made in 2008 evidenced a clear dierence in discharge rates in the tunnel, showing an increase in permeability towards the tunnel (erosion of inllings). However, the Lag Tiert spring as well as the Lag Prau Pult spring did not show any reaction (or very slight) during tests at high water stages. As these results could hardly be understood a numerical pipe ow model was applied. It could reproduce measured results with a fair precision and could explain why springs react at low water stage but not (or hardly) at high water stage. It was also possible to evaluate the impact of the tunnel on the springs discharge.

PAGE 277

Earth Sciences 1538 2009 ICS P roceedings 15th International Congress of Speleology TRACING GROUNDWATER FLOWPATHS IN THE EDWARDS AQ UIFER RECHARGE ZONE, PANTHER SPRINGS CREEK BASIN, N ORTHERN BEX X AR COUNTY, TEX X AS USASt T Ev V E JOHNs S ON1, G G EARy Y SCHINDEL1, and G G EORGE VENI2 1EEdwards AAquifer AA uthority, 1615 NN St. Marys Street, San AA ntonio, Texas UU SA A 78215; sjohnson@edwardsaquifer.org and gschindel@edwardsaquifer.org2NN ational CC ave and Karst RResearch II nstitute, 1400 CC ommerce DDr., CC arlsbad, NN ew Mexico, UU SA A 88220; gveni@nckri.org Abstract e Edwards Aquifer Authority injected non-toxic organic dyes into six caves within the recharge zone of the San Antonio Segment of the Edwards Aquifer to trace groundwater owpaths and measure groundwater ow velocities. e monitoring system consisted of 32 public and private wells, including irrigation wells completed in either the Edwards Aquifer or the underlying Trinity Aquifer. Results of the tracer tests revealed discrete groundwater owpaths associated with Panther Springs Creek. During most tests, dye was detected in well 68-28-608; however, dye was also detected in seven additional wells. Groundwater velocities to well 68-28-608 ranged from 1134 to 5283 m/day; velocities to the seven other wells where dye was detected ranged from 13 to 2330 m/day. ese are apparent or minimum groundwater velocities since velocities can only be calculated between injection and detection points. e results demonstrate the high groundwater velocities that are characteristic of karst aquifers. e results also indicate that groundwater ows freely between injection points in the upper member of the Glen Rose Formation, the stratigraphic unit that comprises the Upper Trinity Aquifer, and detection points in the Edwards Aquifer. Dye was injected into the upper member of the Glen Rose Formation through Boneyard Pit, Genesis Cave, and Poor Boy Baculum Cave, which penetrate the Edwards Aquifer. Dyes traveling along the owpaths between the caves and wells crossed several northeast-southwest trending faults in which members of the Edwards and Glen Rose formations are juxtaposed. Faults with up to 104 m of vertical displacement did not impede groundwater ow. Consequently, the tracer tests show excellent communication between groundwater in the Upper Trinity Aquifer and the Edwards Aquifer. One trace through soil, in a 1-m2 interstream upland area where no karst features were evident, was shown to accept 180,000 L (250 L per hour) of recharge over a one-month period that ushed dye to at least two Edwards Aquifer wells. is trace demonstrates that aquifer vulnerability to contamination is not limited to recognizable karst features. e study revealed the three-dimensional groundwater ow system in the Edwards Aquifer in the Panther Springs Creek area. Groundwater owpaths shi laterally and vertically in response to changing aquifer conditions. ese tests also highlighted the anisotropy such as discrete groundwater owpaths, aquifer characteristics that change with water levels, wide-ranging groundwater velocities, vertical groundwater ow, and rapid response to precipitation that exists in karst aquifers that is oen underrated or even ignored while characterizing groundwater systems. Finally, this study demonstrated the diverse data necessary to characterize a karst aquifer system, including tracer tests, hydrophysical surveys, continuous water level measurements, and cave mapping.

PAGE 278

15th International Congress of Speleology Earth Sciences 1539 2009 ICS Proceedings URBAN KARST DRAINAGE PROBLEMS IN THE ENSOR SINK TIRESTO SPARE SYSTEM, COOKEV V ILLE, TENNESSEE, USASIDNE IDNE Y W. JONE ONE S1, RICHARD RICHARD C C FINCH INCH2 1640 EE RRock uarry RR d., CC larkrange, TNN 38553 UU SA A. UU T Knoxville UU SA A2299 AA llen HH ollow RR d. CC ookeville, TNN 38501 UU SA A. Tenn. Tech. UU niversity UU SA A Ceiling failure in a cantilever dome produced a 30 m long, 13 m wide, 10 m deep collapse sink entrance to a cave in the Warsaw Fm. (Mississippian) beneath the city of Cookeville, Tennessee. Exploration and mapping of the 630 m long virgin cave revealed urban debris in prodigious quantity and variety, including 30+ automobile tires, hence the name Tires-to-Spare Cave (TTS). Prior to the collapse, the cave stream transported tires and other large debris items through the entrance room; aer the collapse, access to the upstream 380 m of passage was nearly blocked. A breakdown pile now acts as a strainer removing debris from cave oodwaters. A shopping cart and truck cargo seals of known provenance indicated the input point for the debris to be Ensor Sink (ES), 1.2 km straight-line distance from TTS, but access to most of this underground conduit is blocked by permanent sumps at the ES swallet and in the upstream section of TTS. ES, sinkpoint for Breedings Mill Branch (BMB), is the larger of two sinks which form a composite sinkhole. e surface drainage basin of ES, delineated with detailed contour maps (CI=2 ), is approximately 325 hectares and that of the combined sinks approximately 385 hectares. Dye traces were performed to elucidate subsurface owpaths in these basins. Using rainfall data, discharge data in BMB, and stage recorders in ES and other locations, we were able to dene a number of characteristics of the BMB/ES TTS drainage system. About 34% of the BMB/ES watershed is impervious surface and much of the natural channel of BMB has been converted into a concrete canal, resulting in rapid runo into ES. Extrapolation of our data yields a peak discharge of about 100 m3/s during the 100-year storm event, whereas drainage rates from ES are calculated to be an order of magnitude less when ooded to a depth of about 10 m. is disparity between input and drainage rates results in frequent ooding of ES, which may hold up to 123,350 m3 of water during the 100-year event. During our study period, ooding in ES damaged homes, submerged streets, and inundated recreational facilities in the Ensor Sink Natural Area. Rapid ooding at ES suggests a chokepoint in the underground conduit near the swallet. However, stage-discharge relationships for ES, for the swallet in the adjacent portion of the composite sink, and for ooding in the entrance chamber to TTS imply more complex controls on ow through the cave between ES and TTS. e breakdown plug at the collapse entrance to TTS is a known constriction in the conduit. Rapid, total ooding of a portion of the cave occurs upstream from this breakdown pile. Peak stage in ES occurs 30 minutes to two hours before peak stage in TTS, showing that at the time of this study the breakdown plug in TTS was not the cause of ooding in ES. is is expected to change as the interstices of the breakdown lter become increasingly plugged by debris, a process observed during our study. If the inux of urban debris at ES remains uncontrolled, Cookeville will lose the services of an important natural storm sewer.1. IntroductionOn April 25, 1994, ceiling failure in a cantilever dome produced a collapse sink 31 m long, 13 m wide and approximately 10 m deep. e collapse broke through at least 2.5 m of bedrock to give entrance to a hitherto unknown active stream cave developed in the Warsaw Fm. (Mississippian) beneath the city of Cookeville, TN. Albeit a virgin cave, exploration and mapping of the 630 m long cave revealed it to contain urban debris in prodigious quantity and variety: Cans, cigarette butts, a plastic milk crate, automobile parts, toys, a baseball bat, one full beer, a pot, shoes, a lawnmower (?) wheel, commercial truck cargo seals,

PAGE 279

Earth Sciences 1540 2009 ICS P roceedings 15th International Congress of Speleology a shopping cart basket, 5-gallon buckets, lumber, motorcycle tires and over 30 automobile tires (one still on its rim!). e plethora of tires (Fig. 1) inspired the name Tires-to-Spare Cave (TTS). Prior to the entrance collapse, the cave stream transported some large debris items (tires, buckets) through the cantilever dome room and on downstream. Aer the collapse, access to the upstream 380 m of passage was very nearly blocked. e breakdown pile now acts as a strainer removing all large and much small debris from cave oodwaters. A shopping cart (Fig. 2) and plastic truck cargo seals of known provenance indicated the input point for the debris to be Ensor Sink (ES), 1220 m straight-line distance from TTS, but access to most of the underground conduit between the two is blocked by permanent sumps in the upstream section of TTS and 10 m inside the ES swallet. e hydrologic connection between ES and TTS was reconrmed by a rhodamine-WT dye trace on April 8, 1995, with positive results. ES is the sinkpoint for Breedings Mill Branch (BMB), which drains approx. 325 hectares (800 acres) of Cookeville. e BMB drainage basin includes both residential and commercial areas. At the time of this study, about one third of the basin was covered by impervious surface. In 1984, the city of Cookeville spent over $600,000 to improve stormwater drainage in the BMB basin, hiring an engineering rm to (a) clean out the debris-choked ES swallet, (b) line 225 m of the stream channel in the sinkhole with riprap, and (c) convert 850 m of BMB into a concretelined canal. Today, due to the combination of widespread impervious cover and the canalization of BMB, stormwater runo into ES is very ashy and the sinkhole is subject to rapid ooding several times a year. Flooding is generally harmlessly conned to the sinkhole area proper, but upon occasion has overtopped two city streets and damaged one or two houses. In addition to damage and inconvenience, oods in ES and BMB pose a safety hazard to children in the area, who nd the BMB canal an attractive play area. A second sinkhole, Walmart Sink (WMS), draining approximately 60 hectares (150 acres), lies adjacent to ES. e sinkholes are currently separated by a divide, but when ES oods to a depth of approx.10 m, there is overow via a surface paleochannel from ES into WMS. e primary purpose of this study was to better characterize the ooding and drainage of ES and provide the city of Cookeville with recommendations for preserving this natural storm drainage so essential to the citys welfare. 2. Drainage Basin CharacterizationMost of the city of Cookeville is drained through sinkholes. Previous work in the area had established connections between a number of sink points and 6.6 km long Capshaw Cave, which resurges at a spring at the head of a short karst window stream known as e Canal. e Canal stream, which was thought to also be the resurgence point for ES, disappears underground to pass through Ament Cave before resurging for a nal time as the principal source for Pigeon Roost Creek, a tributary to the Caney Fork River. Tracing performed in 1995 established that drainage from ES passes through TTS. e Canal stream sinks to join the downstream end of the TTS cave stream and the combined waters ow through Ament Cave. A second dye trace showed that drainage from WMS also joins the TTS cave stream. A nal dye trace from a third sink lying between WMS and TTS indicated that water sinking here connects Figure 1: Four of over 30 tires in Tires-to-Spare Cave, accumulating at the base of the breakdown collapse. Figure 2: Basket of Roses shopping cart, rst indication of the connection om Tires-to-Spare Cave to Ensor Sink.

PAGE 280

15th International Congress of Speleology Earth Sciences 1541 2009 ICS Proceedings into both the TTS cave stream and into the spring at the head of e Canal. e BMB and WMS drainage basins were delineated using a topographic map with a 2contour interval. Drainage boundaries in these basins may not correspond exactly with topographic divides due to extensive stormwater drains and karstic owpaths in the subsurface, but tracing results in adjacent basins indicate that such errors are minimal. Rainfall within the BMB basin was estimated using a Rain Wise recording rain gauge deployed near the center of the basin, augmented at times by additional rain gauge measurements at other points in the drainage area. e recording rain gauge provided a record of rainfall amount and intensity at intervals of ve minutes, at least for the center of the 325-hectare BMB watershed. Flood stages in the BMB canal, ES, WMS, and the TTS entrance chamber were recorded through time with in-situ model 4000 and 8000 transducer/data loggers (Trolls). To guard against vandalism the Trolls were hidden or camouaged. Additional measurements were made by tape and clinometer surveys to high water levels observed and marked. Discharge downstream of the stage recorder in the BMB canal was measured at low ow and at ten dierent ood stages with discharges ranging from about 1 cubic meter per second (cms) to over ten cms to develop a stage discharge relationship for the canal. A stage-volume relationship for the composite Ensor-Walmart Sink was computed using the two-foot contour interval topographic map of the area. Stage measurements in the sinkholes and discharge rates from BMB canal were used with the sinkhole stage-volume relationships to compute average sinkhole drainage rates in ES over ve-minute intervals for various water depths (Fig. 3). We note that the drainage rates for the three largest oods for which we had complete data do not show the dependence on stage anticipated for a simple orice constriction at the Ensor Sink swallet. Using calculated drainage rates and the watershed characteristics, ood levels for storms not recorded during this study (e.g., the 100-year storm) were predicted.3. Characterisitics of Flooding at Ensor Sink and Walmart SinkDue to urbanization of the watershed and modication of the BMB channel, storm water runo in the BMB basin is routed quickly to ES. e lag time between maximum rainfall intensity and peak discharge into ES measured during this study was typically between 15 and 30 minutes. e BMB canal design allows brimfull discharge of approximately 28 cms (1000 cfs) near its terminus at ES. Hydrologic models incorporating our data indicate a peak discharge between 70 and 115 cms (about 2500 to 4000 cfs) during the 100-yr storm event, during which ES may hold up to about 125,000 m 3 (100 acre/feet) of water in temporary storage. In contrast, drainage rates from ES, as a function of stage, were calculated to be an order of magnitude less than the input: approximately 4.2.5 cms when ooded to a depth of 6.1 m (20 ), and approximately 7.4 7.9 cms (260-280 cfs) when ooded to a depth of 9.1 m (30 ). is disparity between input and drainage rates accounts for the frequent local ooding of ES. Aer this study was initiated, the city of Cookeville took several steps to mitigate the impacts of ooding in ES. One road was raised and a bridge was rebuilt. Two homes subject to ooding were purchased and removed, and much of the area around the ES swallet was designated a natural area and made into a small public park. During the time period spanned by this study, 1998, 18 ES ood events with water depth >6.1 m (20 ) were recorded (max recorded depth = 11.6 m). At least two events produced a merger of the ES and WMS ood ponds inundating recreational facilities in the Ensor Sink Natural Area, in addition to damaging homes and overtopping one bridge. 4. Causes of Flooding at Ensor Sink e rapidity with which ES oods suggests a chokepoint in the cave conduit not far downstream from the ES swallet. 3: Drainage rates at Ensor Sink as a function of stage.

PAGE 281

Earth Sciences 1542 2009 ICS P roceedings 15th International Congress of Speleology However, estimates of ES drainage rate do not display the functional dependence on stage (one-half power) anticipated for a single orice constriction. Simultaneous stage peaks in ES and adjacent WMS, recorded during a June 1998 ood event (which did not produce merger of the two sinkhole ood ponds), suggest that at least one ow restriction lies downstream from the WMS input, causing simultaneous backup in both sinks. During another ood event, which caused pond merger, prominent surface ow was observed from ES to WMS, suggesting that the ow restriction at the ES swallet and/ or additional constrictions between the two sinks create sucient gradient to cause overland ow to WMS. In contrast to the June 1998 ood event, during a December 1998 event stage synchronicity was less perfect, with ES remaining ooded to higher levels for a longer time. ere are at least two possible explanations for the disparity: (1) Rainfall intensity variations in the two drainage areas (we had no rain gauge in the WMS drainage basin); the WMS basin may have received lower intensity rainfall. (2) As previously suggested, ow constriction in the underground conduit between ES and WMS may result in some dierence between sinkhole water levels for less intense storms. Some increase in this suggested partial blockage (possibly due to the continued inux of trash items into the ES swallet) might cause ES to ood higher and drain more slowly. More data are needed to resolve this question. Another apparent restriction in the underground conduit occurs at the upstream sump in TTS where the cave passage size abruptly diminishes from walking passage to a low water-lled tube. e fourth and best known constriction in the subterranean drainage is formed by the breakdown plug at the collapse entrance to TTS. Rapid, total ooding of a portion of the cave upstream from this breakdown pile is known to occur. Does blockage in TTS cause ooding in ES? Not at the time of this study. Peak stage in ES occurs 30 min to two hours before peak stage in TTS, showing that the breakdown plug in TTS was not the principal cause of ooding in ES. is, however, might be expected to change as the interstices of the breakdown lter become increasingly plugged by debris (Fig. 4). During the course of this study an increase in ponding on the upstream side of the breakdown was observed, hinting at growing blockage in the lower levels of the breakdown pile. Additionally, an apparent increase in the degree of in-lling by trash of the interstices higher up in the breakdown lter was noted. If steps are not taken to control the inux of urban debris at ES, Cookeville may lose the services of an important natural storm sewer.5. Conclusions and RecommendationsBased on the results of our study, the following recommendations were made to the Cookeville City Council (which funded this study): at ES be thoroughly cleaned out: removal of some of the rip-rap that has washed into the swallet, and removal of all the trash currently in ES. Installation of a trash rack or strainer structure near the end of the BMB canal, at a point accessible for service. e structure would strain out trash from oodwaters entering ES and thereby reduce the inux of solid trash into the underground conduit, lessening the probability that this conduit will become totally blocked. Regular inspection and maintenance of the trash rack: removal of the collected debris aer oods, to keep the trash rack functioning properly, and to insure that it does not itself become a cause of ponding and ooding. Installation of one or more smaller, subsidiary trash racks elsewhere in the BMB drainage to strain out some of the debris before it reaches the main trash rack at ES. Installation of fencing along the upstream sections of the BMB canal, to reduce the amount of urban debris being washed into BMB, while simultaneously enhancing the safety of children living near the canal. Figure 4: Tire on rim, cans, sticks, sediment, and other debris lling the interstices of breakdown lter in TTS Cave.

PAGE 282

15th International Congress of Speleology Earth Sciences 1543 2009 ICS Proceedings Total plugging of the ES-TTS conduit at any of the several known and suspected conduit restrictions (or other undetected choke points) would necessitate construction of a 1.2 m (4 ) diameter storm sewer at least 1220 m (4000 ) long, blasted through bedrock, at great cost to the city of Cookeville. Fourteen years aer this problem was rst reported and ve years aer the results of this study were presented to the city, none of the above recommendations have been implemented. e city of Cookeville is considering the construction of a sump basin to extract bedload trash out of storm water runo entering ES, combined with a oating screen to remove oating trash. However, to date these plans remain tentative, with no nal design determined. e only concrete steps taken to reduce the amount of urban debris entering the ES-TTS conduit have been several trash clean-ups of ES performed by a local environmental activist group.

PAGE 283

Earth Sciences 1544 2009 ICS P roceedings 15th International Congress of Speleology The HE Importance MPORTANCE of OF Karst ARST Auifers UIFERS to TO Public UBLIC and AND Domestic OMESTIC Water ATER Supplies UPPLIES in IN the THE United NITED States TATES .WILLIAm M K. JONEs S Karst Waters II nstitute, PO O Box 356, Warm Springs, VA A 24484 UU SA A e purpose of this paper is to demonstrate the relative importance of karst aquifers to the supply of fresh water in the continental United States. e distinction between ground and surface waters is oen somewhat blurred in karst regions, with surface streams sinking to become ground water and ground water owing from large springs to become surface water. ere are many problems with quantifying the amount of water used from a particular aquifer type. e available records usually breakdown water usage by surface or ground water and classify by public supply, domestic, irrigation, aquaculture, industrial, mining, and thermoelectric. Total ground water usage gures are available for each state, but no breakdown by aquifers or regions within the states is readily available. Water supplies from the Edwards aquifer in Texas and virtually all fresh waters in Florida are from karst aquifers, but rough guesses must be made for most of the country. Using a state-by-state estimate of the area of karst aquifers and the total utilization of ground water for public and domestic water supplies, karst aquifers account for roughly 30 per cent of this total. Population densities are not uniform across the states, so some karst aquifers are relatively under utilized due to low demand.1. IntroductionKarst rocks, including evaporites, are exposed over about 20% of the earths dry ice-free land (Ford and Williams, 2007), but limestones and dolomites alone cover about 11%, increasing to around 14% if subsurface rocks involved in carbonate karst groundwater circulation are also included (Williams 2008). Karst landforms in the continental United States cover about 40% of the eastern half of the country. A brochure published by the Karst Waters Institute in 2001 had the statement that 40% of the drinking water supplied from ground water in the United States came from karst aquifers. is statement was unquantifed and based on the following assumptions: (1) karst aquifers are typically highly productive with hydraulic conductivity values several orders of magnitude greater than other consolidated bedrock aquifers; (2) the relative demand for ground water for potable water supplies was fairly uniform across the country and regions with 40% karst would supply at least 40% of the ground water; (3) in some cases, such as Florida, karst aquifers also sustain most of the surface streams and therefore represent a drinking water supply not usually listed under the heading of ground water withdrawals. e term drinking water was used to exclude the large pumpage from western aquifers for irrigation and agricultural uses. e utilization of water from karst aquifers is represented by withdrawals from both ground and surface water sources. is is due to the rapid interchange of water within karst aquifers and the storage characteristics of these aquifers. e ow duration curves for surface streams draining karst terrains are typically atter than for non carbonate terrains and show the role of karst aquifers in sustaining surface streams under base ow conditions. As an example, runo from karst aquifers in the Shenandoah Valley help maintain ows in the Potomac River and the water supply for the Washington DC area. ere are many problems with quantifying the amount of water used from a particular aquifer type. e available records usually breakdown water usage by surface or ground water and classify by public supply, domestic, irrigation, aquaculture, industrial, mining, thermoelectric (Hutson et al, 2004). Figures are available for each state, but no breakdown by aquifer types is readily available. Water supplies from the Edwards aquifer in Texas and virtually all fresh waters in Florida are from karst aquifers, but rough guesses must be made for most of the country. For this study, the public and domestic categories will be used for drinking water supply. In practice, less than two percent of the water used by a typical household is used for drinking (van der Leeden et al, 1990). All carbonate and evaporite aquifers are assumed to be karst aquifers and may be unconned or artesian. For most regions without specic data, the percentage of ground water from karst is assumed to be proportional to the amount of karst coverage per state or region. is may overestimate the karst water usage in a couple of the eastern states such as Virginia where large population centers on the coast result in high ground water

PAGE 284

15th International Congress of Speleology Earth Sciences 1545 2009 ICS Proceedings withdrawals from non-carbonate sources.2. Ground Water Use in the United StatesGround water supplies about half of the drinking water of the United States and about 40 per cent of public water systems (Reilly et al, 2008). A few numbers from the USGS (Hutson et al, 2004) showing ground water use in the US for 2000 serve as a starting point. Ground water withdrawals for the entire United States averaged 315 million cubic meters per day (Mm3/d). About 67 percent of this total is used for irrigation. e most heavily exploited aquifer is the high plains aquifer in the west-central United States drawing about 64 Mm3/d from ancient stream gravels. Ninety-seven per cent of this water is used for irrigation and about two per cent is used for public supply. Compare this to the carbonate Edwards aquifer of Texas where average withdrawals are 2.8 Mm3/d with 38 per cent used for irrigation and 56 per cent used for public supply. A study by Maupin and Barber (2005) presented gures for ground water withdrawals from thirty principal aquifers, but the classication of these aquifers was regional rather than by rock type or stratigraphic units. It is impossible to separate the carbonate aquifers from the regional groups and examine them individually. e study by Hutson et al (2004) presents ground-water withdrawals for 2000 by state grouped by the use categories of public supply, domestic, irrigation, aquaculture, industrial, mining, and thermoelectric power. Public and domestic withdrawals by state are presented in Table 1. A rough estimate of the per cent of outcrop area of soluble rocks (carbonates and gypsum) by state from karst maps by Davies et al (1977) and Veni et al (2001) is used in Table 1. e per cent of soluble rock aquifers times the ground water withdrawals for public and domestic supply gives an estimate of the amount of ground water from karst aquifers used for public and domestic water supply. ousand m3/d ousand m3/d StatePublic & Domestic Karst AreaTotal Karst Water % GW from Karst Alabama 1362.2215 0.3 408.88239 30 1884.5515 0.15 282.832095 15 607.4925 0.15 91.172025 15 11570.745 0.01 115.76859 1 456.0925 0.1 45.63335 10 462.527 0.01 4.627714 1 220.6655 0.01 2.207821 1 9080.215 1 9085.013 100 1468.58 0.5 734.678 50I I daho 1151.397 0.2 230.40108 20 1847.08 0.2 369.6112 20I I ndiana 1767.595 0.5 884.2645 50I I owa 1272.517 0.4 509.27576 40 732.776 0.2 146.63264 20 342.5425 0.4 137.0894 40 1476.907 0.05 73.88437 5 Maine 247.1605 0.1 24.72911 10 612.0345 0.2 122.47158 20 905.372 0.05 45.29252 5 Michigan 1839.51 0.05 92.0241 5 1551.093 0.05 77.59563 5 1469.7155 0.1 147.04921 10 1410.291 0.6 846.62172 60 Montana 277.819 0.05 13.89829 5

PAGE 285

Earth Sciences 1546 2009 ICS P roceedings 15th International Congress of Speleology e calculation of thirty-one per cent of the drinking water supply (public plus domestic) of the United States derived from ground water is very much an estimate and does not try to account for population distribution relative to the location of the karst aquifers. is overestimates the percentage of ground water from karst aquifers in some eastern states such as Virginia where large population densities along the east coast create high demand and withdrawals from the North Atlantic coastal plain aquifer. Karst aquifers probably are of little importance in water supplies for the coastal northeastern states, but most of this area relies primarily on surface water. e combining of ground water and surface water withdrawals for Florida may inate the numbers for this state, but there is some justication. Karst aquifers help sustain base ow for many surface water supplies and this is only reected in this study by the data for Florida. 3. ConclusionsIf half of the drinking water in the United States comes from ground water and thirty per cent of this is from karst aquifers, then een per cent of the drinking water comes from karst aquifers. is is at best a somewhat crude estimate based on the available data and was made to illustrate the importance of karst aquifers in the national situation in the continental United States. Ideally data should be collected by aquifer type for each public and domestic water system in the country, but this type of information is not currently available. e next step in rening the estimates presented in this paper should be to obtain more precise numbers for the exposure area of karst aquifers and incorporate an adjustment for population densities in the karst regions. 1190.004 0.05 59.53164 5 656.319 0.15 98.49987 15 279.7115 0.01 2.798593 1 1815.6645 0.05 90.831195 5 New Mexico 1110.519 0.3 333.33174 30 2744.125 0.1 274.5575 10 1343.675 0.05 67.21925 5 167.6755 0.01 1.677641 1 Ohio 2392.12 0.3 718.0152 30 524.2225 0.2 104.8999 20 705.1455 0.01 7.055181 1 1302.04 0.2 260.5456 20 97.99365 0 0 0 637.7725 0.4 255.2438 40 241.1802 0.15 36.196146 15 1338.376 0.5 669.5416 50 5264.935 0.7 3687.4019 70U U tah 1438.6785 0.15 215.915805 15 152.157 0.2 30.44748 20 771.0045 0.3 231.42357 30 2229.365 0.05 111.52715 5 307.342 0.2 61.50088 20 1613.5455 0.3 484.31943 30 Wyoming 241.36945 0.1 24.149699 10 Totals 72581.8413 22318.28687 31Table 1: Estimated use of karst water for public and domestic supply in the continental United States in 2000. Data on ground water withdrawals om Hutson et al (2004), page 9, table 4.

PAGE 286

15th International Congress of Speleology Earth Sciences 1547 2009 ICS Proceedings ReferencesDavies, W.E., J.H. Simpson, G.C. Ohimacher, W.E. Kirk, and E.G. Newton, 1984, Engineering aspects of karst (map): U.S. Geological Survey Open File Report, 2004-1352, map. FORD, D.C. and P.W. WILLIAMS, 2007, Karst hydrogeology and geomorphology: Wiley, New York, 562 pp. Hutson, S.S., N.L.Barber, J.F.Kenny, K.S. Linsey, D.S. Lumia, and M.A. Maupin, 2004, Estimated use of water in the United States in 2000: U.S. Geological Survey Circular 1268, 46 pp. Maupin, M.A., and N.L. Barber, 2005, Estimated withdrawals from principal aquifers in the United States, 2000: U.S. Geological Survey Circular 1279, 46 pp. Miller, J.A. ed., 2000, Ground water atlas of the United States: U.S. Geological Survey Hydrologic Atlas 730, Chapters B N published separately. Reilly, T.E., K.F. Dennehy, W.M. Alley, and W.L. Cunningham, 2008, Ground-Water Availability in the United States: U.S. Geological Survey Circular 1323, 70 pp. Van A N der DER Leeden EEDEN F., F.L. Troise ROISE and D.K., Todd ODD 1990, e water encyclopedia: Lewis Publishers, Chelsea, 808 pp. Veni E NI G., H. Du U Chene HENE N.C. Crawford RAWFORD C.G.Groves ROVES G.H Huppert UPPERT E.H. Kastning ASTNING R. Olson LSON and B.J. Wheeler HEELER 2001, LL iving with karst: American Geological Institute, 64 pp. Williams, P.W. 2008. World Heritage Karst and Caves: a ematic Study. IUCN World Heritage Studies 2, Gland, Switzerland, 26 p, +16 photos and 2 tables

PAGE 287

Earth Sciences 1548 2009 ICS P roceedings 15th International Congress of Speleology Some OME EVidence IDENCE of OF Hydrothermal YDROTHERMAL Speleogenesis PELEOGENESIS in IN the THE Akka KKA Limestone IMESTONE Northern ORTHERN Kitakami ITAKAMI Massif ASSIF Japan JAPANNN ARUHIk K O KAs S HIm M A1 H H IDEs S At T O KUw W AHARA2 1 Professor EE meritus, EE hime UU niversity, 10-13 DDogo-H H imata, Matsuyama, EE hime, 790-85772Speleological RResearch II nstitute of Japan, 1-1 Kannari, II waizumi-cho, II wate, 027-0501 Japan As is well known, many caves in the Trias-Jurassic Akka limestone were dissolved by the groundwater in contemporary drainage basins. However, the Akka limestone has been extensively subjected to contact-metamorphism by Early Cretaceous(120-110 Ma) granites. In this paper, the writers provide the following evidence that suggests hydrothermal activity -existence of a simple network of white calcite veins;-reconstructed to alternating and mixed black and white layers;-existence of hydrothermal crackle breccias;-identication of hydrothermal cave mineral (sepiolite)-occurrence of small skarn deposit; and-S and S(H2S type) springs are sporadically distributed. Auszug : Viele Hhlen im Trias-Jura Akka Kalkstein wurden durch Grundwasser in der zeitgenssischen Entwsserungbassins aufgelst. Jedoch wurde der Akka Kalkstein weitgehend In Verbindung tretenmetamorphismus durch frhe kreidige (120-110 Ma) Granite unterworfen. In diesem Papier schlgt der folgende Beweis hydrothermale Ttigkeit vor: Bestehen Adern eines der einzelnen Netzes des weien Kalzit, Kalkstein, der die weien und schwarzen Schichten wegen des Metasomatismus, Bestehen der hydrothermalen crackle breccias enthlt, Vorhandensein des hydrothermalen Hhleminerals (sepiolite), des Vorkommems der kleinen skarn Ablagerung und des sprlich verteilten Schwefellagers und des Suldlagers entspringt.1. IntroductionIt has been common knowledge that the limestone dissolution is being chemically due by the reaction kinetics of the CO2-H2O-CaCO3 system in a lot of karst areas distributed in Japanese Island. On the other hand, active hydrothermal Pirika limestone Cave (containing hot water) was the rst discovered in the Pirika karst area, Imakanecho, Hokkaido, Northern Japan (Kusaka et al., 1966). In addition, Lu (1999) points out with a large possibility that the compound thermal karstication has aected the Akiyoshi-dai (plateau) karst, Yamaguchi Prefecture, Southwest Japan. Recently, hydrothermal quartz crystals, calcite veins and silicied rocks due to paleo-geothermal activity (Early Cretaceous) were found in the Akiyoshi limestone (Akiyoshi-dai) (Taguchi et al., 2005). e purpose of this work was to investigate the evidence of the hydrothermal speleogenesis and the discovery of the hydrothermal cave in the Akka limestone, the Kitakami Massif, Northeast Japan. e hydrothermal cave cannot be found yet, to keep investigating in the future. e purpose of this work was to investigating evidence of the hydrothermal speleogenesis and the discovery of the hydrothermal cave in the Akka limestone, the Kitakami Massif, Northeast Japan. e hydrothermal cave cannot be found yet, to keep investigating in the future.2. Geological Settinge Northern Kitakami massif occupies the Pacic coast region of Northeast Japan. is region belongs to the North Kitakami Belt which the northeastern side of the Hayachine Tectonic Belt trending northwest is divided into two Subbelts, the Kuzumaki-Kamaishi Subbelt to the west side and the Akka-Tanohata Subbelt to the east side by the Iwaizumi Tectonic Line. ese two Subbelts are composed Jurassic accretionary complexes and commonly characterized by the steeply dipping structures with NE-SW or NNW-SSE trends (Sugimoto, 1974; Kitamura, 1990; Okami, 1999; Ehiro, et al., 2008) (Fig.1). e Akka limestone belongs in the Akka-Tanohata Subbelt, which is exposed in two lines of east and west in about 5km maximum-wide and approximately 60 km long, about 700 m in total thickness, and contains more than 200 caves. ere are activities of granites in the Early Cretaceous (120110 Ma) and a big and small granite bogies were intrusive in the North Kitakami Belt, and the Akka limestone have been contact metamorphosed to the recrystallized limestone (Kobayashi 1974).

PAGE 288

15th International Congress of Speleology Earth Sciences 1549 2009 ICS Proceedings 3. MajorEevidence of Hydrothermal Speleogenesis 3.1. Existence of white calcite veinWhite calcite veins are found in a simple network pattern of the Akka limestone all over the investigated area. Two examples of the white calcite veins are shown Figure 2. e white calcite veins in recrystallized limestone are carbonate precipitation along the rising hydrothermal uid paths.3.2. Existence of white and black layers e recrystalized limestone composed of the disturbed white and black layers as a ow-banded texture and layer texture due to metazomatism. in section studies indicate that white layer consists of pure calcite and black layer, a minute opaque black minerals are arranged. Figure 3 shows examples of the outcrops for white and black layers as a owbanded texture(le) and layer texture(right). 3.3. Existence of hydrothermal crackle brecciasExposures of crackle breccias are conned to a small area on cave entrance of Oana Cave, Sodeyama, Iwaizumi Town. e outcrop shown Figure 4, as easily visualized from the gure, the crackle breccia consists of brecciated wall-recrystallized limestone fragments and cemented by calcite. 3.4. Presence of hydrothermal cave mineralse hydrothermal cave minerals such as calcite, quartz, barite and sepiolite are described in Japanese Caves. Sepiolite (Mg4Si6 O15(OH)2 6H2O) is common found as cave mineral from Uchimagi-do Cave, Kuji City and rarely found from Ryusen-shindo Cave, Iwaizumi Town. In Uchimagi-do Cave, sepiolite forms white, exible sheets resembling leather hang down from open cracks in the recrystalized limestone result in corrosion of the bed rock (Kizaki, 2002). Sepiolite in Ryusen-shindo Cave showing vein-like occurrence in recrystalized limestone and resulted in condensation by the segregation from Mg and Si elements bearing bed rock (Honda, 1972). Figure1: Simplied geological map of study area, showing the approximate location of the Caves (: 1. Uchimagi-do Cave, 2. Shigawatari-do Cave, 3. Akka-do Cave, 4. Ryusendo Cave, 5. Ryusen-shindo Cave, 6. Oana Cave), Springs (1. Kawamata Spring, 2. Shinyamane Spring, 3. Shitatokusari Spring, 4. Momonokidaira Spring & 5.Sawamawari Sprig), Skarn Deposits (1. Okawame Deposit, 2. Kakukaketoge Deposit, 3. Torikaino Deposit & 4. Yamaguchi Deposit) and Outcrop points (1.Usaka-Toge, 2.Matsubayashi, 3.Shirando & 4. Sodeyama). Figure 2: Examples of white calcite veins in recrystallized limestone. (le) Usaka-toge, near Uchimagi-do Cave, Kuji City, (right) Akka-gawa river side, near Shigawatari-do Cave, Iwaizumi Town. Figure 3: Examples of white and black textures in recrystallized limestone, (le) layer texture of the wall an evacuation tunnel in Ryusenn-do Cave, (right) road cutting wall at Shirando, Iwaizumi Town Figure 4: Close-up photograph of the crackle breccia, near entrance of Oana Cave, Sodeyama, Iwaizumi Town.

PAGE 289

Earth Sciences 1550 2009 ICS P roceedings 15th International Congress of Speleology 3.5. Occurrence of small skarn depositSome skarn deposits are found in small bodies sporadically scattered in the Akka-Tanahata Subbelt. In the North area, Okawame molybdenum deposit and Kakukake-toge garnet skarn deposit in Kuji City formed in slate and recrystalized limestone (Yoshida, et al., 1987). In the South area, Yamaguchi deposit (Cu and W) and Makaino deposit (pyrrhotite, pyrite, hedenbergite, garnet) are distributed in Iwaizumi Town (Shimazu, et al., 1970). ese deposits were formed by the skarnization due to hydrothermal metazomatism. e skarn deposits in the studied area are now not mined entirely. 3.6. Sparsely distributed sulfur-bearing and sulde bearing springse chemical composition of spring waters in the AkkaTanohata Subbelt classied to NaHCO3, NaSO4 NaCl and intermediate types using a Trilinear diagram. e NaHCO3 type springs and hydrogen sulde (H2S)-containing springs are closely associated with the Akka limestone. Many of the springs in this area are non-thermal (cold water), these original paleo(fossil)-spring waters probably derived from granitic magma .4. Conclusionse study has resulted in the identication of hydrothermal speleogenesis heretofore undiscribed for the Akka-Tanohata Subbelt, the North Kitakami Belt region. Most of the above evidences were suggested and wide cited as being characteristic of hydrothermal speleogenesis. One of the research objectives, the hydrothermal cave has not been recognized yet in this area. Investigations in the future are planned. Acknowledgements We wish to express our thanks for the supply of a Grant in Aid for Speleological Research Institute of Japan (Iwaizumi). We also thank the 15th ICS Science Committee reviewers, whose useful comments helped to clarify the paper. References Ehiro, M., S. Yamakita, S. Takahashi, and M. Suzuki, 2008 J urassic accretionary complexes of the NN orth Kitakami Belt in the AA kka-Kuji area, NN ortheast Japan. e Journal of the Geological Society of Japan, 114Supplement : 121. (Japanese). Honda, S., 1972 With reference to discoery of sepiolite m ountain leather. CC ave study, 5 : 1. Cave Collaboration Group. (Japanese). Kobayashi, I. (Ed.) 1974-CC r etaceous GG ranitic RRocks in the Kitakami Mountains-Petrography and ZZ onal A A rrangement-. 139 pp. Geological Survey of Japan. (Japanese with English abstract). Kitamura, S.(Ed.) 1990 G G eo logical Map of II wate PrefectureMuseum edition-. Iwate Prefectural Museum. Kizaki, H., 2002 CC a ve minerals of UU chimagi-do CC ave. Research Report of Uchimagi-do cave, Yamagata Village, Iwate Prefecture, 131-137. Research committee of Uchimagi-do Cave and Board of Education of Yamagata Village. (Japanese). Kusaka, H., N. Kashima, N. Itoda, A. Nojo, and Pirika Sirenia Research Group, 1996 e rst nding of a hydrothermal karst in Japan ; Pirika limestone C C a ve, II makane, Southwestern HH okkaido. Earth Science(Chikyu Kagaku), 50 : 403. (Japanese) Lu Y., 1999 RR esearch on the EE volutions of Karst H H ydrogeological EE nironments and eir EE ngineering I I mpacts. 305 p Scientic Publishing House, Beijing. (Chinese with English title) Okami, K., 1992G G eology of II waizumi Town. Local History of Iwaizumi Region. 303p. Board of Education of Iwaizumi Town. (Japanese). Shimazu, M., K. Tanaka, and T. Yoshida, 1970 GG eo logy of the Taro district. With GG eological Sheet Map at 1:50,000. Geological Survey of Japan. 54p. (Japanese with English abstract 6p.). Sugimoto, M., 1974 St ratigraphical Study in the OO uter Belt of the Kitakami Massif, NN ortheast Japan. Contribution from the Institute of Geology and Paleontology Tohoku University, no.74 : 1, 24 gs., 7 tabs., 6 pls. (Japanese with English abstract). Taguchi, S., E. Kido, N. Tsukamoto, and T. Sugiyama, 2005 Paleo-geothermal activity in the AA kiyoshi LL imestone revealed by uid inclusions. Bulletin of the Akiyoshidai Museum of Natural History, 40 : 35, pls.1. (Japanese with English abstract) Yoshida, T., M. Yoshii, M. Katada, K. Tanaka, T. Sakamoto, and H. Satoh, 1987 GG eo logy of the RR ikuchu-O Ono district. With GG eological Sheet Map at 1:50,000. Geological Survey of Japan. 70 pp. (Japanese with English abstract 6 pp.).

PAGE 290

15th International Congress of Speleology Earth Sciences 1551 2009 ICS Proceedings EV V OLUTION OF THE BI NIK CAV V E ENV V IRONMENT KRAKOWWIELUN UPLAND, SOUTHERN POLANDKA A TARZ ARZ YNA NA KA A SPRO RO WSKA A VIACHE IACHE SLA LAV ANDRE ANDRE YCHOU CHOU KU U niversity of Silesia, Faculty of EE arth Sciences, Brdzifska 60, 41-200, Sosnowiec, Poland Binik Cave is considered to be one of the best investigated caves in Poland due to results of absolute dating of the sediments and thorough research of its evolutionary background. It is one of the oldest archeological sites in Poland and one of very few places in Europe where a prole of sediments formed during climatic phases starting with the Odra glaciation through the Holocene has been preserved. e data obtained af ter examining the sediments from the cave and their thorough analysis enable an attempt to reconstruct the evolution of the cave environment since its formation until today. An evolutionary model of Binik Cave has been presented. e cave environment is treated as a geocomplex of related elements (karstied rock, morphology, cave water, microclimate, vegetation, fauna and man) developing under the inuence of external factors (climatic changes, geomorphological processes, seismic and tectonic events etc.) and internal factors (speleogenesis, formation of cave sediments etc.). e main stages and events of the cave development since its formation through today have been marked on a geological time axis. e reconstruction is based on a detailed dataset and multi-aspect investigation of the cave sediments. 1. IntroductionBinik Cave is one of the numerous (non-active) relic caves in Krakow-Wielun limestone Upland, Southern Poland (Fig. 1). It is located near Strzegowa village, close to Smolen (the Wolbrom commune). e cave entrance is placed on the le slope of Wodaca Valley, in the West wall of Bisnik Rock (approx. 405 m asl 50 25.3N, 19 40.0E). e three entrances, directed NW and NE, are situated 6m above the bottom of the valley. Binik is a small cave, currently formed at three levels (approx 80 m long). Its system is tectonically predisposed by ssures, and consists of few horizontally developed passages and chambers (Fig. 2). e cave is considered to be one of the best-investigated caves in Poland, due to the fact that its evolutionary background has been thoroughly investigated and absolute dating of sediments has been carried out. Besides, it is considered to be one of the oldest archaeological sites in Poland and one of the very few ones in Europe where the prole of sediments formed during climatic phases from the period preceding the Odra glaciation till Holocene has been preserved (Cyrek Y REK 2002; Madeyska ADEYSKA AND Cyrek YREK 2002). Detailed and multi-aspect research on the sediments in the cave, their thorough analysis and interpretation, combined with the application of a special type of research methodology allow an attempt to reconstruct the evolution of its environment (treated as a geocomplex) since the its formation. 2. Research Concepte evolution of caves is the result of the development of karst processes (tectonic movements, climatic changes, geomorphological processes, inuence of human activities and others) and interactions between the formation of underground voids and their geological environment. Figure 1: Location of the research area: 1 the area of the Landscape Park Orlich Gniazd; 2 surrounding reserves of the park; 3 the area of Wodca Valley; 4 Upper Jurassic cuesta; 5 Krakw-Wielu Upland; 6 Silesian Upland; 7 Binik Cave.

PAGE 291

Earth Sciences 1552 2009 ICS P roceedings 15th International Congress of Speleology Generally, two main periods can be distinguished: speleogenetic (when the initial ssures are widened to bigger channels and passages) and speleoenvironmental (the period when these voids containing water, air, sediments, living organisms, etc., are forming the specic cave environment); the boundary between these periods is not well dened. e cave environment can be considered as a geocomplex; a set or system of connected elements (geocomponents), occurring in the rock void the cave. e geocomponents are: karstifying rock (L), morphology of caves (M), sediments (O), cave waters (H), underground atmosphere (A), vegetation (R), fauna (F), and sometimes a human being (C). Due to their nature the geocomponents can be grouped into: abiotic not alive, biotic alive and anthropica human being (Fig. 3). Elements of the cave environment as a geocomplex are linked by numerous connections: spatial, genetic, materialenergy, functional etc., creating a specic system. Various processes occurring in cave environment are nothing else but interactions of geocomponents. e formation of stage structures is due to interactions among elements of the cave environment. eir investigation allows reconstructing environmental conditions in the cave in dierent stages of its development. Multi-aspect investigation of cave sediments is especially important. 3. Methods of the Cave Environment Reconstruction Geological (sedimentological), geomorphological, paleozoological and archaeological methods are the most important among the methods applied to reconstruct the Bisnik Cave environment. Physico-chemical methods, especially uranium-thorium disequilibrium (U-), electron spin resonance (ESR), and thermoluminescence dating have been the most important in determination of the age of the sediments. e archaeological investigations and dating of the cave sediments were carried on by Pielisiak I ELISIAK (1993-1994), Cyrek Y REK (2002, 2006), Hercman E RCMAN AND Gorka O RKA ), Madeyska ADEYSKA (2002), Miros IROS aw AW Grabowska RABOWSKA (2001, 2002a, 2002b), Wiszniowska I SZNIOWSKA et al. (2002) and others. e cave has an almost complete and strongly dierentiated stratigraphical-chronological sediment cross-section, over 7m thick. Eighteen layers and 3 series of sediments (Fig. 4) have been identied in the open prole in front of the main entrance to the cave. ey probably formed during the Pliocene through the Wolstonian Stage, the Eemian interglacial, and the Vistulian through the Holocene period respectively (Miros I ROS aw AW Grabowska RABOWSKA 2002a, b). Particular parts dier from each other by their background, lithological formation, manner of sediment accumulation, their continuity (or the lack thereof), and their incorporation of dierent palaeontological specimens and artefacts. 4. Reconstruction of the Environment Evolution in Bisnik Cave Figure 2: Plan of Binik Cave (aer A. Polonius, 1991): 1 rock; 2 entrance; 3 the edge of the terrace in ont of the cave; 4 vertical rock sill; 5 the range of the hood; 6 the border of the area coered by the rock overhang. Figure 3: e structure of the cave geocomplex and its building elements (explanation in the text).

PAGE 292

15th International Congress of Speleology Earth Sciences 1553 2009 ICS Proceedings An attempt to reconstruct the environment (geocomplex) of Binik Cave from its formation until today has been made based on lithological, palaeozoological and archaeological data (Fig. 5). Reconstruction of the cave geocomplex has been correlated with palaeogeographic data and the processes that have taken place in the cave (corrosion, gravitation, sedimentation processes, activity of living organisms etc.), leaving material and morphological traces of their appearance. Two periods have been distinguished in the history of development of the cave geocomplex: speleogenetic (initialkarst Neogene) and cave-environmental (Pliocene/ Pleistocene Holocene) ones, and among them three stages: abiotic (karst), biotic and anthropogenic. e caveenvironmental period was completely recorded in the cave sediments (layers 18 a). Based on cave sediment analysis, the character of the cave environment has been reconstructed at dierent stages of its formation as well as the factors (external and internal ones) inuencing the processes (corrosion processes, frost disintegration, cryoturbation, climate oscillations, transgressions and regressions of land ice, activity of living organisms and humans etc.). Evolution of the cave environment has been characterised by the increase in the complexity of the geocomplex due to the formation of new environmental elements and the growing number of connections among them. e share of factors inuencing formation of the cave environment has been changing together with its development: the role of external factors has been decreasing, while the role of internal ones has been increasing (self-development).ReferencesCyrek, K. (2002) Rekonstrukcja zasiedlenia Jaskini Binik in Jaskinia Binik. RRekonstrukcja zasiedlenia jaskini na tle zmian rodowiska przyrodniczego, Krzysztof Cyrek (Ed.), Wyd. UMK, Toru, pp 9. Cyrek, K., (2006) rodkowopaleolityczne vistuliaskie zespoy wyrobw krzemiennych z Jaskini Binik in II n m emoriam Valdemari CC hmielewski. Ksirga powircona pamirci Profesora Waldemara C C h mielewskiego, Karol Szymczak, Micha Przedziecki (Eds.), wiatowit, Supplement Series P: Prehistory and Middle Ages, Wyd. Instytut Acheologii UW, Warszawa, pp 93. Figure 4: Cross-section of sediments aer the eld drawing by T. Madeyska (description of sediments aer J. MirosawGrabowska, 2002 a): 1 dark grey clay (a) and beige grey dust clay (b) with sharp-edged limestone debris; 2 light beige dust clay with single pieces of limestone; 3 beige mid-grained sand ; 4 beige-grey san clay and clay sand; 5 red-yellow, ne grained and laminated sand; 6, 7 dark yellow and yellow, ne grained sand with single pieces of limestone debris; 8 yellow-brown sand dust with excess of limestone debris; 9 grey-brown clay; 10-11 brown and dark brown sand clay with limestone debris; 11 sand clay; 12 beige-green san clay and clay sand in the form of lenses; 13 light brown and yellow-brown sand clay with debris; 14 grey-yellow dust clay; 15 yellowbrown, locally red-brown dust clay; 16-17 green-brown and green-grey dust clay with limestone debris; 18 red and red-brown residual loam. Figure 5: Evolution of the enironment (geocomplex) of Binik Cave. Elements of the cave enironment: L karstifying rock; M morphology; O sediments; H water; A underground atmosphere; R vegetation; F fauna; C man.

PAGE 293

Earth Sciences 1554 2009 ICS P roceedings 15th International Congress of Speleology Herman, H. and P. Gorka, (2002) Analizy metod uranowo-torow koci kopalnych z Jaskini Binik in J askinia Binik. RRekonstrukcja zasiedlenia jaskini na tle zmian rodowiska przyrodniczego Krzysztof Cyrek (Ed.), Wyd. UMK, Toru, pp 181. MADEYSKA, T., and K. CYREK (2002) Cave llings a chronicle of the past. An outline of the Younger Pleistocene cave sediments study in Poland. AA ct a G G eol. Pol. 52, 1, 75. Miros I ROS aw AW Grabowska RABOWSKA J. (2001) Litologia i stratygraa osadw Jaskini Binik in Badania naukowe w ponudniowej czrci Wytyny Krakowsko-C C z rstochowskiej, Jzef Partyka (Ed.), Materiay konferencyjne, referaty, postery, sesje terenowe, Ojcw, pp 529. Mirosaw-Grabowska, J. (2002a) Litologia i stratygraa osadw Jaskini Binik in Jaskinia Binik. R R ekonstrukcja zasiedlenia jaskini na tle zmian rodowiska przyrodniczego, Krzysztof Cyrek (Ed.), Wyd. UMK, Toru, pp143. M IROSA AW-GRABOWSKA, J. (2002b) Geological value of Binik Cave sediments (Cracow-Czstochowa Upland). AA cta GG eol. Polon., 52, 1, 97. PELISIAK, A. (1993-1994) Jaskinia Binik (gmina Wolbrom, woj. katowickie) badania z 1991 roku. A A ct a AA rchaeol. CC arp., 32, 125. Wiszniewska, T., P. Socha, and K. Stefaniak (2002) Czwartorzdowa fauna z osadw Jaskini Binik in Jaskinia Binik. RRekonstrukcja zasiedlenia jaskini na tle zmian rodowiska przyrodniczego, Krzysztof Cyrek (Ed.), Wyd. UMK, Toru, pp193.

PAGE 294

15th International Congress of Speleology Earth Sciences 1555 2009 ICS Proceedings MORPHOGENETIC CLASSIFICATION OF TALUS CAV V ES BASED ON GEOMETRY OF CLASTS AND SEQ UENTIAL DEV V ELOPMENTERNERN ST H H KA A STNING NING PH H .D D .N N ew HH ampshire GG eological Survey, NN ew HH ampshire DDepartment of EE nironmental Services, 29 HH azen DDrive, CC oncord, NN ew HH ampshire 03302, UU SA A ernst.h.kastning@des.nh.go Abstract Talus caves, also known as rockfall caves or boulder caves, consist of openings within an accumulation of broken or disaggregated rock (scree) on a steep mountainside, at the base of a cli, or along the oor of a valley. Particles (clasts or grains) within a talus deposit vary in size, sorting, and shape. Interstices are created among clasts during the processes of mass wasting and slope failure. ese caves are a subtype of pseudokarst that includes depositional, accidental, tectonic, and ri caves. Clasts that compose talus may be (1) boulders: rounded or angular ranging from high to low sphericity, (2) blocks: generally angular and having a wide range of sphericity, and (3) slabs: angular and generally low in sphericity. e size and extent of the openings (caves) among the clasts depends on the texture of the talus accumulation, namely size, shape, and sorting of the particles, much in the same way as these parameters determine the degree of porosity in a traditional ner grained sediment. Talus caves have three stages in development. (1) In the primary stage, the host rock is fractured, creating blocks bounded by joints. e nature of fracturing, including extent, density (spacing), and orientation of fractures will determine the size, shape, and sorting of the clasts. (2) In the secondary stage, rock fragments become dislocated and moved, creating enterable openings (caves). Processes in this step include movement under the force of gravity (rock falls, rock slides, toppling, creep, and other mechanisms), uid transport of clasts by uvial or glacial action, and packing of clasts into their nal orientation. (3) In the tertiary stage, diagenetic changes occur within the talus deposit, including physical and chemical weathering of clasts and changes in sizes of interstices through inlling of interstices with ner grained particles or removal of ll by erosion. Superlative talus caves (long, deep, or voluminous openings) occur where talus deposits are suciently thick, clast size is maximized, sorting of clasts is high, interstitial matrix of ll is initially minimal or has been erosionally removed, and openings remain large over time.

PAGE 295

Earth Sciences 1556 2009 ICS P roceedings 15th International Congress of Speleology The HE sinkholes SINKHOLES of OF Layla AYLA Lakes AKES ; Saudi AUDI Arabia RABIA and AND their THEIR singular SINGULAR sub SUB lacustrine LACUSTRINE gypsum GYPSUM tufa TUFASt T Ep P HAN KEmp MP E1, H H EIk K O D D IRks KS2, I I NGO BAUER3 and R R ANDOLf F R R AUs S CH4 1II nstitute of AA pplied GG eosciences, UU niversity of Technology DDarmstadt, Schnittspahnstr. 9, DD -64287 DDarmstadt, GG ermany; Kempe@geo.tu-darmstadt.de2DDornier CC onsulting, RR iyadh, Saudi AA rabia; HH eiko.D D irks@gtzdco-ksa.com3II nstitute of AA pplied GG eosciences, UU niversity of Technology DDarmstadt, Schnittspahnstr. 9, DD -64287 DDarmstadt, GG ermany; Bauer@geo.tu-darmstadt.de4GG TZ Z II nternational Services, RR iyadh, Saudi AA rabia; RR andolf.R R ausch@gtzdco-ksa.com e uplied basement of the Arabian Plate is overlain by east-dipping Mesozoic sediments. e more resistant of these strata form escarpments that curve throughout Saudi Arabia north to south. Level areas occur between these escarpments, one formed by the upper Jurassic Heeth Formation composed of about 150 m thick anhydrite. is formation is locally punctured by hypogene karst sinkholes. e most prominent sinkholes are the former Layla Lakes at 22.17N 46.70E. e lakes (17 formerly), that originally provided water for a sustainable agriculture in the area, have been drained in the late 1980s, revealing 19 sinkholes, some of them composites of several subsidence centers. e largest sinkhole is 1.1 km long, 0.4 km wide and about 40 m deep. Others are less than 10 m across and rather recent. e bottom of the former lakes and the ats around them are composed of thick layers of ne-grained lake chalks (Layla Formation). e most striking feature of these sinkholes is the several meter thick tufa covering the vertical walls of the sinkholes. It formed sub-aqueous and is entirely composed of large gypsum crystals. Morphologically, the tufa displays thick bulbous forms at the bottom of the walls changing to inverted conical forms at middle depth, and to gour-, gutter-, or shovel-like forms near to the former lake surface. e mineralogy and morphology of this tufa appear to be singular worldwide. eir growth apparently was caused by gypsum saturated water that formed by excessive evaporation in the shallow shore waters around the sinkholes. ese dense water layers then cascaded underwater down the walls of the sinkholes, triggering the growth of centimeter-sized crystals along the walls. e Layla Lake area should be protected by a natural reserve to prevent further dumping of refuse into them.1. Introductione Arabian plate is composed of crystalline basement in the west and of eastward dipping Phanerozoic strata in the east. In the center of the Kingdom of Saudi Arabia (KSA), a band of Jurassic and Cretaceous sediments crop out running nearly north-south throughout the entire country. Prominent escarpments are formed by middle Jurassic and lower Cretaceous limestones. e plain in between the escarpments is partly formed by the upper Jurassic Heeth (or Hith) Formation, which is composed of a >150 m thick sequence of laminated and autobrecciated anhydrite. is band of anhydrite can be followed on satellite images, such as those provided by Google Earth, as a string of bright, almost white areas. Ground inspection showed that these areas are marked by gypsum caliche, presumably formed by ascending waters that le gypsum upon evaporation. e Heeth Formation is locally punctured by hypogene karst sinkholes, which include the deepest cave in Saudi Arabia, A in Heeth, about 40  km south of Riyadh, and the former Layla Lakes, a series of sinkholes about 300  km south of Riyadh. e Heeth Formation is an aquitard, below which fossil groundwater is trapped that is extensively used for agricultural purposes. Consequently, the groundwater level h as dropped in many areas by >100  m in the last decades. is is illustrated by the history of the Ain Heeth cave. e cave formed by upward solution of the groundwater body in a hypogene setting sensu Klimchouk LIMCHOUK (2007), its walls show the typical morphology of a convective cave formation in a phreatic setting in gypsum (e.g., Kempe E MPE 2008). In the 1950s, the groundwater body reached the surface, forming a s pring lake. Nowadays, the groundwater level is 160  m below surface, revealing the cave. e drop of the groundwater level following agricultural activities is also the reason for the dry-up of the Layla Lakes.

PAGE 296

15th International Congress of Speleology Earth Sciences 1557 2009 ICS Proceedings 2. Layla Lakes SinkholesIn the early 1990s, the lakes dried up, revealing a series of textbook sinkholes. Before, the lakes were used as touristic spots for swimming and water sports. Google Earth provides a high resolution view of the northern sinkholes, while those in the south are barely perceptible once the ground situation is known (Fig. 1). All in all 23 sinkholes can be listed (Table 1), of which some have several subsidence centers (Numbers 2, 4, 18, and 22). We were able to visit most of the sinkholes on February 21 and 25, 2008, in order to inspect them for tufa occurrence (KEMPE & DIRKS, 2008). e largest sinkhole is No. 4 with an north-south axis of 1.1 km, a width of 0.4 km, and a depth of up to 40 m. e smallest ones have openings of less than 10 m across, but are bellowing out below, forming real sinkhole caves. Two of the sinkholes (18 and 19) are connected by a natural bridge. For a few sinkholes, we were not able to estimate the depth because the bottom was not visible. In general, the small overhanging sinkholes are extremely dangerous because their margins are composed of loose material (lake marls). Many appear to have opened rather recently because they lack signs of tufa. All in all, 19 of the structures (including the sub-sinkholes) have tufa, i.e., they are older and had standing water in them (Fig. 2). However, how many lakes these represented is not recorded. Apparently most of the sinkholes were originally covered by water and, thus, not known at all. e Water Atlas of Saudi Arabia lists 17 lakes (Ministry I NISTRY of OF Agriculture GRICULTURE and AND Water ATER 1984). e publication also lists electrical conductivities of the lake waters. ese ranged from 2510 to 8600 S/cm, i.e., the values are higher than to be expected in carbonate saturated waters. Some values appear to be even much higher than gypsum (or anhydrite) saturated waters. is may be due to ongoing evaporative concentration in the shallow lakes. us, the lake water most probably was high in sulfate as well as carbonate. is may explain why we only found two species of gastropods, quite a small number of mollusk species for a freshwater habitat. e shells of one of the species are ubiquitous while the second one, a cerithiid species, was found only in one place (half-way down in Sinkhole 2). e shallow lake bottoms (around the sinkholes) was occupied by reeds, the roots of which are still noticed everywhere. e bottoms of those sinkholes that contained lakes as well as the ats around the sinkholes are composed of very ne-grained marls or lake chalk (seekreide). Since the draining of the lakes, these deposits have dried out and shrunk. Numerous meter-deep shrinkage cracks crisscross Figure 1: Google Earth view of the Layla Lakes area and annotated map of sinkholes.

PAGE 297

Earth Sciences 1558 2009 ICS P roceedings 15th International Congress of Speleology the lake bottoms and surround the sinkholes. ey can not be dierentiated from cracks caused by ongoing subsidence. e shrinkage of the lake bottom sediments has caused their subsidence. In places, shrinkage amounted to more than a met er (Fig. 3) and opened over 10  m deep shrinkage caves between the sediment and the solid tufa (for example, at the southern end of Sinkhole 2). NoNorthEastSize (m) Depth (m) TypeRemarks (T = Tufa) Qanat 122.168946.7166200*50Ca. 3-10elongatedT, fresh circular cracks No 222.162646.7034400*100Ca. 5-30elongatedT, fresh cracks; pit caves No 2a22.162846.702410*10>20 narrow pit T No 2b22.162646.704550*50ca. 10circularT No 2c22.163246.404330*30ca. 10circularT, fresh cracks no 322.163446.701960*60ca. 15circularPost-lake no 422.168146.70871100*400ca. 30elongated T, former main lake, fresh cracks, terraces channel 4a22.168246.707750*50ca. 10circularT, sand dune no 4b22.168646.707315*15ca. 10circularT, sand dune at bottom No 522.170646.708750*40>20circularT ?, bottom not visible no 622.171546.708760*60ca. 30circularT, undercut, caves? no 722.174446.706860*60?circularnot visited no 822.177446.713370*70ca. 40circularT no 922.177146.71409*8ca. 10circularoverhanging, with sand pileno 1022.176946.713915*136circularoverhanging, with sand pileno 11 22.177646.71427*10>25circularoverhanging 1222.178546.714315*155circularhalf lled by dunes no 1322.178646.714616*146circularhalf lled by dunes no 1422.180746.7157105*90ca.40irregularT, sand dunes yes 1522.182046.717850*50ca.50circularT yes 1622.184046.7256100*11035irregularT, collapse blocks yes to 15 (80m) 1722.185346.724115*1530circularoverhanging no 1822.186546.7257315*110Up to 25elongatedT, composite of 3 sinkholes channel to 17 (130m) 18a22.185746.725495*70ca. 15elongatedT, terraces 18b22.187046.7257112*85ca. 15irregularT, terraces 18c22.187246.726895*80ca. 25irregularT, collapse blocks 1922.186946.727351*25ca. 20elongatedT, natural bridge with 15cyes 2022.188446.728038*38ca. 20irregularT, pit channel to 19 (160m) 2122.188046.728726*22?elongatednot visited yes 2222.189746.729460*35?elongatednot visited channel 22a22.189746.729220*18?circularnot visited 2322.209746.734548*38?circular not visited, half lled by dunes yes (?) Table 1: List of sinkholes near Layla. Sizes according to Google Earth and own eld inspection.

PAGE 298

15th International Congress of Speleology Earth Sciences 1559 2009 ICS Proceedings Field observation and Google Earth images also revealed that the lakes served as natural outlets of the deeper aquifer. A series of parallel channels and qanats (Arabic for subterraneous water channels) conducted water over about 5 km from the lakes to farms in the north. is usage of the lake water was sustainable, since the qanats did only allow for gravitational outow, thus only water that naturally owed out of the aquifer was consumed.3. Layla Lakes Gypsum Tufa 3.1 Observationse most important discovery made in Layla Lakes is the magnicent tufa that covers the vertical walls of the lake sinkholes (Fig. 4). Tufa here is understood as a sublacustrine low-temperature mineral deposit. Repeated test with 2NHCl, ngernail scratching and macroscopic inspection showed that the tufa is entirely composed of gypsum. In it, numerous gastropod shells can be found immured at places. Even though the tufa surface is covered with ne-grained, cream-colored calcareous dust, the tufa itself is a sparitic selenite or a massive, ne-grained gypsum. Since so far no such site has to our knowledge been described in the literature, the term tufa is used even though it has so far exclusively been used for calcareous deposits (compare Ford O RD & Pedley EDLEY 1996). e term travertine is avoided because it is more commonly used for hydrothermal and high pCO2 sources of calcareous deposits. e expectation to nd calcareous tufa or stromatolitic microbialites, as is common in CaCO3-supersaturated lakes such as Plitvice (Kempe EMPE & Emeis MEIS 1985) or Mono Lake, Pyramid Lake, Walker Lake, and Searles Lake in the western USA (compare Kempe E MPE & Kazmierczak AZMIERCZAK 2008), Lake Van in eastern Anatolia (Kempe E MPE et al., 1991), or in crater lakes such as Vai Lahi on Niuafoou (Kazmierczak A ZMIERCZAK & Kempe EMPE 2006), was not met. An overview over calcareous tufa deposition in low temperature environments was given by Ford O RD & Pedley EDLEY (1996). Apart from the singular mineralogy, the morphology of t he tufa is the most striking feature: e 20 to 30  m high walls of the sinkholes are covered with a several meter thick crust of tufa. ere is a distinct morphological change from bottom to top (example Sinkhole 2, eastern wall; Fig. 4): At the bottom, the crust forms a solid, overhanging panel with only shallow vertical groves. Above, the crust is more segmented into meter-sized bulbous, inverted cones (le on Fig. 4), and further up the tufa forms protruding, upwardoriented, shovel-like bowls or cups (Fig. 5). Some of the cups are more delicate than in other places and can form large shell-like basins. e rims are oen less than 10 cm thick and the cups are partially lled with loose sediment. Figure 2: Layla Lakes in the early 1980s. Figure 3: Shrinkage of lake-bottom chalk deposits. Figure 4: Panorama view of the eastern wall of Sinkhole 2.

PAGE 299

Earth Sciences 1560 2009 ICS P roceedings 15th International Congress of Speleology In a sense, they resemble speleothem gours. In places where the sinkhole does not display vertical walls, such as in the center of Sinkhole 4, we nd rather regular rows of cup-like structures forming steps. A specically regular tufa display was found at the northern wall of Sinkhole 4b. Overall, this triple division of the gross morphology of the tufa is found in all of the sinkholes.3.2 Genetic considerationsIt is clear that we will never be able to fully understand the genesis of this special form of tufa because the lakes and their physico-chemical environment in which they grew are gone forever. We, thus, can only speculate on these conditions, based on the few data that are available. e observation that the conductivities were high, i.e., in or even above the range of gypsum-saturation, and from the fact that these tufa forms exist, we must conclude, that the Layla Lakes were saturated with respect to gypsum. In contrast to calcite, gypsum cannot be highly supersaturated but precipitates at saturation. Apart from evaporation, two more processes govern gypsum saturation, best explained by Wigley IGLEY (1973) and exemplied for a gypsum karst setting by Brandt R ANDT et al. (1976). ese processes are: codissolution or co-precipitation of calcite and temperature alteration. us, several processes can be discussed that may have caused the gypsum precipitation along the walls of the lakes: Concentration by evaporation. It is conceivable that on the wide ats that surrounded the deeper sinkholes, evaporation would concentrate the water much faster than over the deep sinkhole sections of the lakes. Higher concentrated solutions may have run down along the oor and cascaded underwater over the lips of the walls of the sinkhole, bringing saturated solutions into contact with the crystals already growing on the walls (precipitation upon reaching nucleation sites). Calcite co-dissolution. Concentration on the ats by evaporation and degassing of CO2 causes precipitation of CaCO3 (calcite and/or aragonite). Downward cascading solutions bring the water into deeper water layers with a higher pCO2. ere, the co-transported ne-grained lake chalk is being dissolved, pushing (because of the common ion eect of the increasing calcium activity) the gypsum over the saturation limit and causing its crystallization along the walls. Temperature change. Ascending water from the deep aquifer, delivering gypsum saturated solutions warmer than the lake temperatures, cools in the lake and gypsum crystallizes along the walls. In any case, the continued delivery from the underlying aquifer of nearly gypsum saturated water is the key condition. ese solutions may either have risen slowly through the sediments of the lake bottoms or through distinct vents. One of such vents may have been adjacent to Sinkhole 2. Another vent may have been located at the northwest-end of Sinkhole 4 and in the western wall of Sinkhole 20. 4. Conclusionse situation seen today is the product of a long geologic history. At this point, one can only guess what its history. One of the clues is the fact that all of the sinkholes seem to occur in unconsolidated lake chalks, the Layla Lake Formation. us, the outow of deep aquifer water over long periods rells any subsidence holes with chalk. During times of wetter climate, the groundwater is recharged and the lakes expand. During dry climate periods, the groundwater ows out only sparingly and the lakes retreat to the immediate vicinity of the sinkholes. Where the gypsum Figure 5: Upward oriented, shoel-like cups of tufa (Sinkhole 2, western wall, size about 7 x 4 m).

PAGE 300

15th International Congress of Speleology Earth Sciences 1561 2009 ICS Proceedings tufa has collapsed from sinkhole walls, we nd over 10  m high proles of laminated or layered lacustrine chalks. ese oer the possibility to do paleoclimate research, an aim we will pursue next.References:Brandt RANDT A., S. Kempe EMPE M. Seeger EEGER and F. Vladi L ADI Geochemie, Hydrographie und Morphogenese des Gipskarstgebietes von Dna/Sdharz. Schweizerbartsche Verlagsbuchhandlung, Stuttgart, Geol. Jahrbuch, Reihe C, 45: 55 pp.Ford O RD T.D. & Pedley EDLEY H.M. A review of tufa and travertine deposits of the world. Earth Science Reviews, 41: 117. Kazmierczak A ZMIERCZAK J. and S. Kempe EMPE Modern analogues of Precambrian stromatolites from caldera lakes of Niuafoou Island, Tonga. Naturwissenschaen 93: 119. Kempe E MPE S. Vom Urkanal zur unterirdischen Kathedrale, Hhlenformen und ihre Entstehung. In: Kempe, S. & Rosendahl, W. (Hrsg.), Hhlen: verborgene Welten; Wissenschaliche Buchgesellscha Darmstadt, Darmstadt: 54. Kempe E MPE S. and H. Dirks IRKS Layla Lakes, Saudi Arabia: the world-wide largest lacustrine gypsum tufas. Acta Carsologica 37/1: 7 Kempe E MPE S. and K. Emeis MEIS Carbonate chemistry and the formation of Plitvice Lakes. In Transport of Carbon and Minerals in Major World Rivers, Pt. 3 (eds. E.T. Degens, S. Kempe & R. Herrera), Mitt. Geol.-Palont. Inst. Univ. Hamburg, SCOPE/ UNEP Sonderband 58: 351-383. Kempe E MPE S., J. Kazmierczak AZMIERCZAK G. Landmann ANDMANN T. Konuk ONUK A. Reimer EIMER and A. Lipp IPP Largest known microbialites discovered in Lake Van, Turkey. Nature 349: 605-608. Kempe E MPE S. and J. Kazmierczak AZMIERCZAK Soda lakes. In: J. Reitner, and V. iel, (eds.) Encyclopedia of Geobiology, Springer, submitted. Klimchouk L IMCHOUK A. ) Hypogene Speleogenesis, Hydrogeological and Morphogenetic Perspective. Nat. Cave and Karst Res. Inst. Spec. Pap. No 1, 106 pp. Ministry I NISTRY of OF Agriculture GRICULTURE and AND Water ATER of OF the THE Kingdom INGDOM of OF Saudi AUDI Arabia RABIA Water Atlas of Saudi Arabia. Saudi Arabian Printing Company, 112 pp. Wigley I GLEY T.M.L. Chemical solution of the system calcite-gypsum-water. Canadian J. Earth Sci. 10: 306.

PAGE 301

Earth Sciences 1562 2009 ICS P roceedings 15th International Congress of Speleology GEOMICROBIOLOGY AND HYDROLOGY OF POOL PRECIPATES IN THE GUADALUPE MOUNTAINS, NEW MEX X ICO, USAAA RA KOOs S ER1 L L AURA C C ROss SS Ey Y1, D D IANA N N ORt T HUp P2, MIk K E Sp P ILDE3,L L Es S LIE MELIm M4 1 EE arth and Planetary Science DDepartment, UU niv. of NN ew Mexico, AA lbuquerque, NN M 87131 UU SA A2 Biology DDepartment, UU niv. of NN ew Mexico, AA lbuquerque, NN M 87131 UU SA A3II nstitute of Meteoritics, UU niv. of NN ew Mexico, AA lbuquerque, NN M 87131 UU SA A4DDepartment of GG eology, Western II llinois UU niversity, Macomb, ILIL 61455 UU SA A e Guadalupe Mountains of southeastern New Mexico (USA) are home to hundreds of caves hosting a variety of speleothems. ese arid-land caves, as semi-closed systems stripped of the inuence of surface weathering, provide a particularly valuable window into the world of carbonate-precipitating microorganisms and the interaction of meteoric waters mixing with deep waters. e cave pool precipitates, many of which are biogenic, record microbial inuences, surface climate, and ecosystem changes. e cave pool precipitates researched in this study are pool ngers. Pool ngers are subaqueous pendant, nger-like speleothems. Previous work in Hidden and Cottonwood caves has shown microbial fossils associated with pool nger precipitates. is work expands on those observations by presenting data on biomarkers extracted from pool ngers. Preliminary analysis indicates that plant waxes (imported into the cave environment) dominate the organic residue in abiologic carbonates (pool spar), while evidence of hopenoids coupled with an absence of plant residues is found in carbonate deposits associated with biogenically-active environments (moonmilk). e connection between microbial communities and their chemical environment is especially strong in extreme environments (extreme pH, salinity, temperature, presence/absence of light, low nutrient conditions, etc.). Modern bacterial communities are known to utilize chemical species that are present in the cave environments for metabolic processes. rough the use of available water data from the literature and calculations of Gibbs free energy available from equilibrium considerations we identify energeticallyfavored metabolic pathways. Study of the fossil biomarkers combined with modern environments will lead to a better understanding of subsurface carbon cycling, characterization of microbial communities, and the input of deeply circulating waters. 1. IntroductionFundamental questions remain regarding formation of secondary calcite and the possible role of microbial precipitation. An interdisciplinary approach to eld and laboratory studies allows us to address some of these major geomicrobiological issues. Subsurface carbonate systems constitute a major arena of interaction between microorganisms, minerals, and water (Barton and Northup, 2007). e bioand geochemical signatures identied in carbonates may be generalized to other biogenic carbonate occurrences such as travertine mounds (Crossey et al., 2006) and deep sea hydrothermal vents, and have potential use in identifying the presence of life on other planets. is research focuses on a less well-studied speleothem that forms in subaqueous cave environments: pool ngers. In the Guadalupe Mountain (Fig. 1) cave pools host a wide variety of speolothems such as pool ngers, webulites, Uloops, and pool meringue (Davis et al. 1990; Hill & Forti 1997; ueen & Melim, 2006). Pool ngers are nger-like speleothems that hang down in cave pools. Pool ngers form entirely underwater and lack a central drip canal. e majority of pool ngers are 1-2 cm in diameter and 5-15 cm long (Davis et al. 1990). e giant pool ngers of Hidden Cave, NM are an exception at nearly 10 cm in diameter and well over 1 m in length. e internal layering of pool ngers indicates downward growth with lesser outward growth (Fig. 2). e current hypothesis is that microbial laments, hanging down from submerged surfaces, act as nuclei for the growth pool ngers. A previous study of fossil pool ngers in Hidden Cave found evidence of alternating microbial and abiotic layers (Melim et al. 2001). Further work in other caves has found a number of potentially biogenic carbonate precipitates (ueen & Melim, 2006). ese published studies suggest biogenic processes contributed to the formation of these cave paleo-pool carbonate deposits, and

PAGE 302

15th International Congress of Speleology Earth Sciences 1563 2009 ICS Proceedings thus pool ngers provide a record of paleo-hydrology and paleomicrobiology. Biomarker analysis is an important tool for understanding biogenic carbonates. Past and present bacterial communities utilize chemical species present in the cave environments for metabolic processes and may directly or indirectly contribute to carbonate production. e lithied communities leave behind ngerprints in the form of biomarkers. Biomarkers can include fossilized microbes, mineral fabrics, mineralogy, and preserved lipids. Lipid biomarkers le behind by microbes in the calcite can have a unique signature that can be used to identify dierent types of microorganisms. Several studies (Blyth et al., 2006, 2008) have shown that bacteria and plant biomarkers can be extracted from cave stalagmites. Earlier studies (raikill 1971; Forbes 2000; Turin & Plummer 2000, Palmer and Palmer, 2001) have examined the chemistry of pool waters or compared the chemistry to pool precipitates, but none have considered the role of microbes in modifying either the pool chemistry, pool speleothems, or linked the geochemistry, mineralogy, and microbiology of the cave pools. Bacterial communities utilize chemical species present in the cave environments for metabolic processes. Using available water data from the literature and calculations of Gibbs free energy available from equilibrium considerations, we identied the energetically favored metabolic pathways. Most metabolically important reactions involve oxidationreduction reactions among chemical (gas, aqueous and mineral) forms of hydrogen, nitrogen, carbon, sulfur, iron and manganese. e combination of thermodynamic analysis and eld observations (minerals forming in cave environments and identied microbial communities) further rene the list of potential metabolic reactions. ese predictions guide further investigation into the ways in which microbial species participate in the formation of cave precipitates. 2. MethodsSamples were collected from Hidden Cave and Carlsbad Cavern. A pool nger from Hidden Cave and pool spar were processed for lipid biomarker analysis. e biomarkers were extracted using a series of solvent washes in a Soxhlet Extractor; the products of each wash were analyzed using gas chromatography followed by gas chromatography/mass spectroscopy. Detecting other types of biomakers requires using various microscopy techniques. We used a on a JEOL 5800 scanning electron microscope (SEM) equipped with an Oxford (Link) Isis energy dispersive x-ray analyzer (EDX) to identify mineral fabrics and fossilized microbes; xray mapping provided mineral composition and on occasion identication of mineral via crystal habits, distribution of elements, and targets for the microprobe; back-scattered emission (BSE) supported the identication of chemical layering; cathode luminescence gave information about dierent trace element variations among layers. Analysis on a JEOL 8200 electron microprobe provided quantitative elemental composition, and x-ray diraction (XRD) was used for conrmation of mineral phases. Cave pool water samples were collected and analyzed using inductively-coupled plasma optical emission spectroscopy (Perkin Elmer Optima 5300 DV ICP-OES) and anion chromatography (Dionex 500x) to provide major ion chemistry for the active hydrologic system. Alkalinity was performed by standard titration methods using sulfuric acid. Analyses were performed in accordance with standard quality assurance quality control protocol. ese data provided information for hydrochemical mixing models, microbial metabolism models, and evaluation of stability of several important minerals (including gypsum and Figure 1: Regional and local map of the study area. Figure 2: Cross section of pool nger om Hidden Cave. Photo by Dr. Leslie Melim.

PAGE 303

Earth Sciences 1564 2009 ICS P roceedings 15th International Congress of Speleology carbonate). To understand trends in hydrochemistry, regional aquifer data were compiled from the literature and Piper diagrams were constructed to identify geologic units with chemically similar water and to dene the evolution in water chemistry along potential ow paths. Chloride/bromide plots were constructed to determine salt sources, and external carbon was computed (Chiodini et al., 2000) to identify the presence of carbon sources above that due strictly from carbonate dissolution. To provide reference points for active speleothem formation, we incorporated data from two cave systems, caves in Alaska (Cataract Cave and rush Cave) with active moonmilk formation and unnamed caves in La Madera, NM with active pool ngers, were examined as well. Several water analyses were selected for further thermodynamic modeling. e initial modeling looked at ve pools from Lechuguilla Cave: Lake of the Blue Giant, Sulfur Shores, Lake of the White Roses, Lake Chandelier, and Briny Pool. Waters that represent active biogenic formations as well as inactive waters were plotted and modeled to determine likely microbial metabolic pathways. e thermodynamic models were compared to each other to discuss dierences/similarities in potential pathways. 3. Results and DiscussionBiomarker analysis was done on a pool nger from Hidden Cave (Fig. 3) and a piece of pool spar from Carlsbad Cavern. e moonmilk portion of the large pool nger from Hidden Cave contained several short-chained fatty acids (C16-C22). e C16-C22 are considered generic lipid biomarkers. In the polar fraction unknown hopanes were detected. e presence of a hopanes with the short-chained fatty acids conrms the presence of bacterial biomarkers in the moonmilk portion of the pool nger. e pool spar sample (assumed to be abiotic) produced a dierent mass spectra pattern for the acid fraction and polar fraction. e acid fraction contains short chain fatty acids (C16-22), but there are no hopanes present in the other fractions. e polar fraction for the polar spar is dominated by plant biomarkers producing the rainbow spectra of C22 and higher chains. e initial x-ray map of one section of a pool nger showed spheres of magnesium carbonates (conrmed by EDS) (Fig. 4), but this is not conclusive evidence for microbial involvement and bears further investigation. e cathodoluminescence (CL) images of the pool nger showed very faint CL in the cores of two samples and plumes of brighter areas rimward (Fig. 5). e results from the CL imaging might be due to changes in trace elements or from re-precipitation (biotic or abiotic) of the calcite. Further investigation using high resolution imaging and microprobe data is planned. Fig. 5 Geochemical analysis was done on waters from three dierent cave systems: the caves in Alaska (in the Tongass National Forest); the La Madera waters (travertine mounds with caves); and nally, the cave pools in the Guadalupe Mountains e Piper Diagram shows a clear trend for the dierent cave systems (Fig. 6). e Alaskan caves, which have actively precipitating moonmilk, are clustered in the lower-le corner. e La Madera waters, which originate Figure 3: Preliminary biomarker data om moonmilk portion of a pool nger. Figure 4: Microprobe x-ray map of magnesium in pool nger. e light grey spheres are a magnesium carbonate.

PAGE 304

15th International Congress of Speleology Earth Sciences 1565 2009 ICS Proceedings from CO2 springs with other deeply sourced gases (Newell et al., ), plot roughly in the center, whereas the Guadalupe Mountain cave pool waters trend to the far le side. e sulfate + bicarbonate plotted against calcium + magnesium shows a clear trend of the Guadalupe Mountain waters trending on the 1:1 line with the Alaskan and La Madera waters are o to the right of the line. ere are a few exceptions for the Guadalupe Waters that do not follow the 1:1 line. e 1:1 line represents carbon sources from waterrock, soil, and atmosphere input (Crossey et. al., in press). e waters that plot o this line have a higher then expected carbon signal. e chloride/bromide plot shows that most of the salt present in the cave waters is from water-rock interactions. e preliminary thermodynamic modeling (using reaction quotient computations from activities calculated by the speciation program PHREEQc (Parkhurst, 1995) and methods of Meyer-Dombard et al., 2005 for ve cave pools from Lechuguilla Cave showed that microbial communities would potentially use oxygen as the primary electron acceptor, followed by nitrate, goethite, hematite, and sulfur. One pool, Sulfur Shores, was predicted to have communities that utilize goethite and hematite before elemental sulfur. 4. Conclusionse initial lipid biomarker analysis shows potential bacterial biomarkers in the outer portion of the pool nger from Hidden Cave. e spar sample (assumed to be abiotic) does not have bacterial lipid biomarkers, but instead there are preserved plant biomarkers, potentially conrming its abiogenic nature. ese results suggest a bacterial role in precipitation of the outer portion of the pool nger. e x-ray maps from the microprobe and the CL images from the SEM showed areas on the pool nger sample that will be targeted for analysis. e geochemical analysis shows that most of the cave pool water chemistry in the Guadalupe Mountains is derived from water-rock interactions. e Alaskan caves and La Madera waters are a combination of water-rock interactions and other processes. ere are several (nine pools) exceptions in the Guadalupe Mountains. e nine pools in the Guadalupe Mountains, along with the Alaskan and La Madera have an external source of carbon above what would be expected from water-rock, soil, and atmosphere interaction (Crossey et. al., 2003, 2006). Usually, this signal is accompanied by other gases such as H2 or H2S. In the case of La Madera, these gases are from deep sources (Newell et al., 2005). In the Guadalupe waters this might arise from active microbial communities. We will target these nine pools for further study. e thermodynamic proles predict that the ve pools that were modeled would have bacterial communities that use similar metabolic pathways except for Sulfur Shores. e thermodynamic proles are not complete until eldwork has been done to conrm mineral phases and microbial communities that may be present. e pool nger communities in the Guadalupe Mountains do not appear to be currently active based on the water chemistry and eld observations. However there are nine pools in the Guadalupe Mountains that need to be investigated based on their geochemistry to determine if there are active pool nger communities. e Alaskan caves and La Madera provide a speculative glance at what geochemical conditions would support a microbial role in pool nger formation. Evidence from previous studies by Melim et. al. (2001) and current work with lipid biomarkers provides increasing evidence of a biogenic origin for pool ngers. Figure 5: CL (le) and BSE (right) image of outer portion of pool nger. Figure 6: Geochemical analysis including Piper Diagram, Chloride/Bromide Plot, and External Carbon Plot.

PAGE 305

Earth Sciences 1566 2009 ICS P roceedings 15th International Congress of Speleology AcknowledgementsWe thank John Craig for analytical work; Paul Burger, Stan Allison, Pat Cicero, Aaron Stockton for gathering water and providing much needed data; Megan Curry for the AK water geochemistry; Kristen Mullen for the x-ray maps; Amy Williams for water chemistry help; Dr. Richard Pancost for the use of his lab; and Dr. Fiona Gill for helping with running the lipid biomarker analysis. is work would not have been possible without support from the Geosciences Directorate of the National Science Foundation, grants EAR 0719507 and EAR643364 .ReferencesBlyth, A.J., P. Farrimond, M. Jones, (2006), An optimized method for the extraction and analysis of lipid biomarkers from stalagmites, OO rganic GG eochemistry, vol 37, pp 882. Blyth, A.J., A. Baker, M.J. Collins, K.E.H. Penkman, M.A. J.S. Gilmour, Moss, D. Genty, R.N. Drysdale, (2008), Molecular organic matter in speleothems and its potential as an environmental proxy, ua ternary Science RReviews, vol 27, pp 905. Chiodini, G., F. Frondini, C. Cardellini, F. Parello, and L. Peruzzi, (2000) Rate of diuse carbon dioxide Earth degassing estimated from carbon balance of regional aquifers: e case of central Apennine, Italy, Journal of G G eophysical RResearch, vol 105, pp 8423. Crossey, L.J., T. Fischer, K.E. Karlstrom, J. Patchett, D. Hilton, W. Sharp, and A. Reynolds, (2003), Springs, gases and travertines of the Colorado Plateau: Xenowhis, links to tectonism, and view of a deeplydissected hydrologic system: GG eo logical Society of A A merica AA bstracts with Programs: vol 35, no. 6, p 200. Crossey, L.J., T.B. Fischer, P.J. Patchett, K.E. Karlstrom, D.R. Hilton, D.L. Newell, P. Huntoon, and A.C. Reynolds, (2006), Dissected hydrologic system at Grand Canyon: Interaction between upper world and lower world waters in modern springs and travertine: GG eo logy, vol 34, pp 25. Davis, D.G., M.V. Palmer, and A.N. Palmer, (1990) Extraordinary subaqueous speleothems in Lechuguilla Cave, New Mexico: NN a tional Speleological Society Bulletin, vol 52, pp 70. Forbes, J. R., (2000) Geochemistry of Carlsbad Cavern pool waters, Guadalupe Mountains, New Mexico: Journal of C C ave and Karst Studies, vol 62, pp 127. Hill, C.A. and P. Forti, (1997), CC a ve Minerals of the World, 2nd EEdition: Nat