Proceedings of the 12th International Congress of Speleology. Volume 1: Symposium 7, Physical Speleology & Symposium 8, Karst Geomorphology

Proceedings of the 12th International Congress of Speleology. Volume 1: Symposium 7, Physical Speleology & Symposium 8, Karst Geomorphology

Material Information

Proceedings of the 12th International Congress of Speleology. Volume 1: Symposium 7, Physical Speleology & Symposium 8, Karst Geomorphology
Alternate title:
12th International Congress of Speleology. Volume 1
Alternate title:
Symposium 7, Physical Speleology & Symposium 8, Karst Geomorphology
Jeannin, Pierre-Yves (editor)
International Union of Speleology
Swiss Speleological Society
Speleo Projects
Publication Date:
Physical Description:
1 online resource


Subjects / Keywords:
Speleology ( lcsh )
Caves ( lcsh )
Karst ( lcsh )
Conference papers and proceedings ( TGN )

Record Information

Source Institution:
University of South Florida
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
K26-05419 ( USFLDC DOI )
k26.5419 ( USFLDC Handle )

USFLDC Membership

Added automatically
Karst Information Portal

Postcard Information



This item is only available as the following downloads:

Full Text




Swiss Speleological Society (SSS/SGH) Proceedings of the 12 th International Congress of Speleology Volume 1 Symposium 7 Physical Speleology & Symposium 8 Karst Geomorphology La Chaux-de-Fonds, Switzerland, 10-17.08.1997


Editor : Pierre-Yves Jeannin Symposium 7: Physical Speleology Symposium 8: Karst Geomorphology Scientific coordinators: Editorial board : Scientific committee: Thomas Arbenz Philippe Audra Ian Chandler Jean-Paul Graf Stephane Gogniat Thomas Gubler Thilo Herold Beat Hochli Nicole Hochli Thomas Bitterli Wanda Stryjenska Jorg Bauchle Volker Bauchle Chris Keller Kurt Keller Joelle Salamin George Huppert Prisca Marietta Karlin Meyers Michel Monbaron Daniel Muller Luc Perritaz Sybille Rex Dagmar Risen Nicole Ruder Daniela Spring Hans Stunzi George Veni Sophie Verheyden Michaela Wessicken Andres Wildberger William B White Proceedings of the 12 '" International Congress of Speleology, 1997, Switzerland Volume 1 II/


I, IGU UGI E@ International Union of Speleology Union lnternationale de Speleologie lnternationale Union fur Spelaologie Swiss Speleological Society (SSS) Societe Suisse de Speleologie (SSS) Schweizerische Gesellschaft fur Hohlenforschung (SGH) United Nations Educational, Scientific and Cultural Organisation Organisation des Nations Unies pour !'education, la science et la culture United Nations Educational Scientific and Cultural Organisation International Association of Hydrogeologists (IAH) Association Internationals des Hydrogeologues (AIH) International Association of Hydrological Sciences (IAHS) Association lnternationale des Sciences Hydrologiques (AISH) Swiss Academy of Sciences (SAS) Academia Suisse des Sciences Naturelles (ASSN) Schweizerische Akademie fur Naturwissenschaften (SANW) International Geographic Union Union Geographique lnternationale International Association of Geomorphologists (I.AG.) Association lnternationale des Geomorphologues (A.LG.) ISBN 2 88374 006-2 (Vol. 1) ISBN 2-88374-012-7 (Vol. 1-6) Publisher: Cover: Sales: Speleo Projects, Therwilerstr. 43, CH-4054 BASEL, Switzerland Titanengang, Holloch, Switzerland (Photo Ballmann/Widmer) Bibliotheque de la Societe Suisse de Speleologie, c/o Bibliotheque de la Ville, Rue du Progres 33, CH-2300 La Chaux-de-Fonds, Fax: 021 947 53 78, email: Printed in Switzerland Papers published from the camera ready copies, prepared by the authors after reviewing by the members of the scientific committee. Despite this, the scientific board wishes to make clear that it shall take no responsability for any mistakes and omissions, or for the opinions stated by the authors. IV Proceed i ngs of the 12 th I nt ernat i onal Congress of Spe l eo l ogy 1997 Sw i tzer l and -Vo l ume 1


Table of Contents Symposium 7: Physical speleology Session 1 a: Karstic sediments and paleoclimate, sedimentology Choppy Jacques & Dubois P Pavuza Rudolf Forgeot Olivier Kadlec Jaroslav Allison Cara & Smart P. L. Felisiak lreneusz Krekeler Mark P. S. et al. Summers Engel Annette et al Maltsev Vladimir A. et al Zupan Hajna Nadja Etude scientifique de la grotte de Clamouse (Herault, France)............................................ ..... 3 Hohlensedimentologie in 6sterreich. ........ ............ ... .. .. .... ............... ...... .......... .. .. .... ................ ..... 7 Les mega-remplissages detritiques de la Cueva Fresca ............................................................. 9 Shape of fluvial pebbles in surface and subsurface streams from Moravian karst Czech Republic.................. .............................................................. 1 3 The value of quarry records and monitoring excavation progress in the interpretation of sediment-filled caves in Eldon Hill Quarry Derbyshire, England ........ 16 Age and distribution of some karstic sediments from the Polish Jura and their implications for the age of the relief ........................................................... 1 7 Sedimentology, clay mineralogy and geochemistry of cave sediment from Hard Baker Cave Rockcastle County Kentucky USA ..................................... 21 A study of cave sediment from Movile Cave, Southern Dobrogea Romania ......................... 2 5 Cave chemolithotrophic soils ................... ..... ..................................................................... .......... ....... 2 9 Mineral composition of elastic material in fault zones and open fissures in karst rocks, examples from SW Slovenia .................................... ...... ........ 33 Session 1 b: Karstic sediments and paleoclimate, dating, isotopes, paleomagnetism Tuccimei Paola et al. Th/U dating of sea level controlled phreatic speleothems from caostal caves in Mallorca (Western Mediterranean) ........................................................... 3 7 Labau V et al Speleothems dating using the thermoluminescence method .................................................... 41 Ford Derek & Zambo L. U series dating of phases of speleothem deposition and erosion in Baradla Cave, Aggtelek National Park, Hungary............ .......... ...... ...... ...... ...... 44 Hercman Helena et al Uranium-series dating of speleothems from Amaterska and Holstejnska caves Moravian karst Czech Republic.... ......................................................... 45 Linge Henriette & Lauritzen S .E. Potential use of uranium series dating of calcareous algae as a tool for dating paleo-watertables in maritime karst settings ................................................ 49 Berstad Ida Malene et al. U-series dating and stable isotope analysis of some last interglacial speleothems from north Norway..... ........................................................... 53 Holmgren K. et al. Speleochronology stable isotopes and laminae analysis of stalagmites from southern Africa.. ....................................................................... .......... 5 5 Mihevc Andrej & Lauritzen S .E. Absolute datations of speleothems and its speleomorphological significance from Divaska jama and Jazbina caves; Kras plateau Sloven i a ................................................. 57 MacDonald William D. & Yonge C J. The isotope systematics of perennial cave ice in Northwestern Canada.... ........... ....... ........... 60 Perrette Yves et al Shopov Yavor Y et al. Barredo Silvia P. et al. Scherrer Nadim C. et al. Enregistrement de l'activite charbonniere dans les speleothemes de Choranche (Vercors, France) ...................................................... .. ... 61 Influence of the bedrock C0 2 on stable isotope records in cave calcites ................................. 65 Paleomagetic study in the Cuchillo Cura System, Neuquen Province Argentina............ .... 69 The palaeomagnetism of speleothems of Middle Quaternary age from northern Australia..... ..................................................................................... 73 Proceedings of the 12 th International Congress of Speleology, 1997, Switzerland Volume 1 V


Session 1 c: Karstic sediments and paleoclimate, paleoclimate Frisia Silvia et al. Gradzinski Michal et al Hercman Helena et al Hercman Helena & Walanus A Ford Derek C. et al. White William B Williams Paul W. Verheyden Sophie et al. Paj6n Morejon Jesus M. Kempe Stephan & Hartmann J Lundberg Joyce Shopov Yavor Y Shopov Yavor Y et al Shopov Yavor Y et al Shopov Yavor Y et al Shopov Yavor Y et al Holocene paleoclimatic fluctuations recordedby stalagmite : Grotta di Ernesto (N Italy) ..... 77 Environmental controls of origin of the annually varved calcite speleothems.................. ...... 81 The antiquity of the famous Demianowska caves (Slovakia)...................................................... 85 Speleothem growth frequencies (PDF) as a climate record : problem of significance.... .. ... 87 A long interglacial and interstadial record in the North American mid-continent from vadose and phreatic speleothems Jewel Cave and Wind Cave South Dakota, U.S.A. ..... 88 Precise measurement of luminescence banding profiles in speleothems for paleoclimatic interpretation.................. ..................... ........................ 8 9 Evidence of glacial advances from New Zealand cave deposits..... .......................................... 9 2 L'oxygene et le carbone dans les concretions nous informent-ils sur les climats passes ? .. ........... ... ... ........ ... ....................................................... 93 Paleoclimate of the quaternary processes in the Western karst of Cuba : a quantitative characterization ................................................................ 97 Excentriques capillaries : do they carry paleoclimatic information?....... .................................... 100 Paleoclimatic reconstruction and timing of sea level rise at the end of the Penultimate Glaciation, from detailed stable isotope study and TIMS dating of submerged Bahamian speleothem .............................................................. 101 Speleothem records of environmental changes in the past Potential in comparison with the other paleoenvironmental archives and related UIS international programs 103 Speleothems as natural climatic stations with annual to daily resolution... ............................. 105 Speleothem luminescence proxy records of annual rainfall in the past. Evidences for 'the Deluge' in speleothems. .................................................................................... 107 Speleothem luminescence proxy records of geomagnetic field intensity................................ 11 0 Speleothem records of processes beyond the Solar System (supernova eruptions).. ....... 111 Session 2a: Cave morphology and genesis, geological factors Sebela Stanka Leel-Ossy Szabolcs Sustersic France Turchinov Igor Ruggieri Rosario Martin Philippe Terlau Craig A. & Day M J. Hof Alex Frumkin Amos Farina Daniele & Gallerini G. Development of cave passages according to geological structure ; example from Jama pod Pecno rebrijo, Slovenia ................................................... ...................... 113 Genesis of Jozsefhegy hydrothermal cave, Budapest.. ................................................................ 116 Cave patterns North of the Planinsko polje (Slovenia)................................................................. 117 Lithological control of speleogenesis in the Pre-Carpathian region.. ...................................... 1 21 The Karst in the area of Matumbi Hills (Kilwa district, Tanzania) Hydrogeological characteristics and relationships between morphostructures and tectonic phases .............. 125 Structures hierarchiques dans le karst de la St Baume (Bouches du Rhone, Var, France) ............................. ......................... ........ ...... ........ .... .................. ... 129 A comparison of the orientation of cave passages and surface tributary valleys in the karst of Southwestern Wisconsin U S.A. ................................ 133 Labyrinthes et Sieben Hengste ........................................................................................................... 137 Classification and some morphometric features of salt caves ..................................................... 139 A karstic system in a semiallochtonous gypsum unit : the Legnanone caves (Marecchia valley, Italy): preliminary geological and hydrogeological features ..... .. .......... 143 VI Proceedings of the 12 '" International Congress of Speleology, 1997, Switzerland Volume 1


Tarhule-Lips Rozemarijn F A. & Ford D. C Garasic Mladen Niggemann Stefan White Elizabeth L. & White W B Knez Martin Morphologic studies of Bell hole development on Cayman Brae .............................................. 1 46 Some types of speleogenesis and speleohydrogeology in D inaric karst area (Croatia Europe) ............................................................................................. 14 7 Or igin and development of caves in the Devonian massive limestone of the Rheinisches Schiefergebirge (Germany)......................................................... 151 Mechanics of cave breakdown : relative importance of shear strength and fracture toughness.. .................................................................. ....... .............. 155 Phreatic Channels in Velika dolina Collapse Doline (Skocjanske jame Caves Slovenia) ...................................................................................... ........ .... 156 Session 2b: Cave morphology and genesis, hydrogeological factors Klimchouk Alexander Klimchouk Alexander Audra Philippe Kicinska Ditta Foltete Jacques Jaillet Stephane Vasileva Danica Lauritzen Stein-Erik et al Rossi Carlos et al. Rossi Carlos et al. Galdenzi Sandro et al. Artesian speleogenetic setting ................. ..... .. .... ....... ........................................................................ 15 7 Speleogenetic effects of water density differences.......... ............................................................. 1 61 Le role de la zone epinoyee dans la speleogenese ..................................................................... 165 Reconstruction of paleoccurents in caves of the Bystra Valley (Tatra Mountains, Poland) on the basis of scallops and deposits analyses ........... .... ........... 168 Hypothese sur la genese d'un siphon....... ........................................................................................ 169 Hydrologie du systeme karstique du Rupt du Puits (Lorraine / Champagne, France) : fonctionnement du siphon aval .. .. ..... ....... ......... ..... ........ ...... 171 The role of groundwater in the genesis of Resava cave................ .. .. ........................ ... ............... 175 Bell hole morphometry of a flank margin cave and possible genetic models: Lighthouse Cave, San Salvador The Bahamas ............................................. 178 Cave development along the water table in Corbe System (Sierra de Pefialabra, Cantabrian Mountains Northern Spain) ............................................... 179 Multiple paleo-water tables in Agujas Cave System (Sierra de Pefialabra Cantabrian Mountains Northern Spain) : Criteria for recognition and model for vertical evolution ........ 183 La corrosione di placchette calcaree ad opera di acque sulfuree : dati sperimentali in ambiente ipogeo ............................................................. 187 Session 3: Cave climatology Mavlyudov Bulat R Jeannin Pierre-Yves et al. Halliday William R De Paola Marco et al Cigna Arrigo A. Stoev Alexey et al. Pechhold Eberhard Mueller Robert J & Day M. J Caves climatic system ................................................................................... ... ....... ............................... 1 91 Some concepts about heat transfer in karstic systems....... ......... ..... ................... ... ......... ....... ....... 195 Hyperthermic caves of the United States...... ....... ............... .... .......... ................................ .... ............ 199 Misure di temperatura in due cavita carsiche nella formazione Gessoso-Solfifera dell'alto Crotonese (Italia meridionale) ......................................................... 202 Monitoring results in Grotta Grande del Vento (Frasassi, Ancona Italy) and its visitors' capacity ...... ....... .............................................................. 203 Temperature anomalies formation and secular instabi lity research of ice of atmospheric origin in the karst caves of North Albanian Alps.. ...... .... .... .. .. ... ...... ....... 207 C0 2 Gehalte von Luft und Wasser im vadosen Karst und in der Bodenauflage... ............. 211 Daily atmospheric variation within caves in southwestern Wisconsin, U S A ............ .... .......... 215 Proceed i ngs of the 12 th Internat ional Congress of Speleology, 1997, Switzerland Volume 1 VII


Session 4a: Cave mineralogy, minerals Maltsev Vladimir A. Martini Jacques E J Hill Carol A & Forti P Viehmann losif et al. Ghergari Lucretia et al. Onac Bogdan Petroniu et al. Turchinov Igor Urbani Franco Overview of cave minerals onthogeny .... ........................................................................................ .. 21 9 Pyrocoproite (Mg (K Na) 2 PP 7 monoclinic), a new mineral from Arnhem Cave (Namibia) derived from bat guano combustion ................................................... 223 The classification of cave minerals and speleothems .. .... .......... .. .... ........ .... .. .. .......... .. .. ...... .. ....... 226 Crystallographical observations on calcite rafts from three Romanian caves ........................ 227 Mineralogy of crusts and efflorescences from Humpleu cave system .................. .. .................. 231 Deposition of black manganese and iron-rich sediments in Vantului Cave (Romania) ...... 235 Cave minerals of the Western Ukraine.. .. .. .. .... .... .... ...... ........ .. ............ .. ...... .. .... .. .... .. .... .. .. .. ...... .. ...... 239 Venezuelan cave minerals: a short overview.................................... .. ...... .. .... .. ...... .. .............. .. ...... 243 Session 4b: Cave mineralogy, speleothems Frisia Silvia et al. Niggemann Stefan et al. Perrette Yves et al. Tarhule-Lips Rozemarijn F. A & Ford D. C. Van Beynen P E et al Rowling Jill Maltsev Vladimir A. Dublyansky Yuri V & Pashenko S E. Gradzinski Michal et al. Song Linhua Aragonite precipitation at Grotte de Clamouse (Herault, France): role of magnesium and driprate .. .. ................ ... .... .. ............ .. .... .. .............. .. ...... 24 7 Aragonitic/calcitic coralloids in carbonate caves : evidence for solutions of different Mg influence...... .. .......... .... .. .. .. ............ .. .............. .. .......... .. .... .... 251 Characterisation of speleothem crystalline fabrics by spectroscopic and digital image processing methods (Choranche, Vercors, France) .... .. ...... .. .......... .. ......... 257 Studies of speleothem dissolution on Cayman Brae and Isla de Mona .... .. .. ........ .. ...... .. .. .... ... 261 Chemical differences between light and dark coloured speleothem .. ........ .... .......... .. .. .... .... .. .. 262 Ribbon Helictites from Jenolan Caves, NSW Australia.................... .. ............ .... .. ...... ........ .... .. .. ... 263 Stalactites, crystlactites, corlactites, tuflactites 4 types of stalactite-like formations generated from cristallisation environments with different physical properties.. .. ...... .... .... .... 267 Cave Popcorn An Aerosol speleothem ? .. ........................ ............ .. .... .... .... .... .. .. .... ............ .. ........ 271 Microbial agents of moonmilk calcification ...................... .............. .. ...... .... .... ........ ...... .. .... .. .. .. ........ 275 Classification of shields in Shihua Cave, Beijing........ ............ .... .. .............. .. ........ .. .... .. .......... ...... 279 Symposium 8: Karst geomorphology Session 1: Epikarstic zone: morphology, hydraulic behaviour and genesis Goldie Helen S Cox N J Tye Andrzej Puech Vincent & Jeannin P -Y Sbai Abdelkader et al. Liu Zaihua et al Geomorphology of limestone pavements of some British, Irish and Swiss sites ........ .......... 285 Epikarstic features in zones affected by periglacial processes example of the Silesian-Cracow Upland (Poland) ........ .. ..... . ... .... .. .. . .. .. . .... .. . ... . ... .. .. ........ .. 289 Contribution a la comprehension du fonctionnement hydraulique de l epikarst : experiences d'arrosage sur le site de Bure (Jura, Suisse) ................... .. ..... .... .. ........ .. .... ...... .. ... 293 lntensite de la corrosion dans les sols du Jura meridional (France) Note preliminaire .... 297 Carbon dioxide in soil and its drive to karst processes: A case study in transitional zone between North and South China .. ........ .. ........ .. .. .. ................ 300 VIII Proceed i ngs of the 12 th Internat i onal Congress o f Spe l eo l og y, 199 7, Sw i tzerland Vo lum e 1


Clemens T orsten et al. White William B. & White E. L. Klimchouk Alexander Harlacher Christof et al. Huntoon Peter W Martini Sergio Sauter Martin Song Linhua The influence of the Pco 2 on the chemical composition of karst spring water (Swabian Alb SW Germany) ................................................... .... ... ... .... ..... 301 A theoretical model for the distribution and transport of carbon dioxide in the epikarst.. .... 305 The nature and principal characteristics of epikarst.. .......................... ..... ..................................... 306 Changes in chemical composition and physical parameters of water in an alpine karst system (Totes Gebirge Steiermark Austria) ... ............................. 307 Definition and characteristics of stone forest epikarst aquifers in South China.. .... .. .... .. ........ 311 II ruolo dello strato limite turbolento nella morphogenesi dei rillenkarren.............................. 315 Determination of the hydraulic characteristics of the epikarst Local and regional approaches....... ................................................................... .... .............. .............. 318 Epikarst aquifer in Fengcong regions South China. ...... ....... ............................ .. ......... .... .. ..... .. .. 319 Session 2: Geotopes and human impact on karst Ek Camille & Closson D Le karst en tant que contrainte physique en amenagement du territoire. Exemple de la commune de Sprimont (Belgique) ........................................................................ 322 Pulina Marian et al. Human impact on karst environment of the Silesian-Cracow region in South Poland.... .... 323 Gillieson David Environmental change and human impact on the arid karst of the Nullarbor Plain, Australia.... ........................................................ .............. 327 Grandgirard Vincent & Spicher M. Les geotopes karstiques du canton de Fribourg (Suisse) .......................... .... ........................ ..... 331 Session 3: Karst history of alpine caves Audra Philippe Denneborg Michael Bini Alfredo et al Bitterli Thomas & Jeannin P -Y Loiseleur Bernard Vanara Nathalie & Maire R. Reflexion sur les facteurs controlant la karstification dans l'Arc alpin....... ................................ 337 Aufbau und Speleogenese eines hochalpinen Karstsystems (KolkblaserMonsterhohle-System Steinernes Meer, Osterreich ; L=43.4 km, HD= 711 m) .................. 341 Karst and glaciations in the Southern pre-alpine valleys.... ........................................................ 345 Entwicklungsgeschichte der Hohlen im Gebiet Hohgant Sieben Hengste Thunersee (Berner Oberland, Schweiz) ................................... 349 Le synclinal oriental de Chartreuse (France) Facteurs influant sur la genese des grands reseaux ........................... ...... ............................ .... .. 355 Neotectonique et speleogenese : application au massif pyreneen des Arbailles (France) ............................................................... 359 Session 4a: Karst geomorphology, karst & environment Furquim Scaleante Oscarlina A Pulina Marian & Glowacki P Choppy Jacques Singh Ramesh B Veni George Slabe Tadej Geography and speleology.. ........................................ .... ..... ... ... .. .. ............. .. .................................... 363 Reaction of the karst env i ronment for acid rains (example : the massif of Snieznik the Sudety Mountains, Polish and Czech sides) ............ 366 La tectonique et le karst.. .............................................................................. .............................. .... ...... 367 Monitoring tropical geosystem underground topography and potential water resources for sustainable development using geoinformatics Indian case study ............... 369 The effects of aridity and topography on limestone cave development.. .............. .................. 373 Cave rocky relief and its speleogenetical significance .................................................................. 377 Proceedings of the 12 th In ternational Congress of Speleology, 1997, Switzerland Volume 1 IX


Goggin Keith E. & Medville D. M. Zhu Xuewen Kadlec Jaroslav Hill Carol A Trofimova Elena Lin Junshu et al. Solutionally enlarged glacial str i ae on Ordovician dolomites in Western Wyoming USA ................................ .... .......... .. ..... .. ................ .. 381 Simultaneous systematic evolution of fenglin karst landforms ............ .. .. ... ...... .. .... .. .. .... .. .... .... .. 385 Reconstruction of the development of semiblind ponor valleys in Moravian karst based on geophysical surveying Czech Republic.. .............. .. .......... .... .... .... ... 387 Sulfuric acid karst: what i s it good for? .. .......... .... .... .. .... .. .. .. .. ...... .. .. ...... ...... .. .. .. .. .. ...... .. .................... 390 Karst denudation in Irkutsk Region.. .. .. .. ................ .. .......... .. ...... ...... .... .. .... .. .... .. .. ...... ................ .. .. ... 391 Karst geomorphological process responses to climate ........ .......... .... ....... .. ....... ......... ........ ..... .. 395 Session 4b: Karst geomorpholgy, regional studies Perna Giuliano Dimuccio Luca Antonio Choppy Jacques Glazek Jerzy Kostov Konstantin Tamas Tudor & Vremir M Otonicar Bojan Djurovic Predrag & Ljubojevic V Crochet Jean-Yves et al Benischke Ralf et al Hoblea Fabien & Audra P Song Linhua & Wang F Yonge Charles J et al Deep Messinian karst in Mediterranean area.. .. .. .. .. .... ..... ... ...... .. .... .. .. .. ..... .... .... .... .. .. .... .. .... .. ....... 397 Studio morphoevolutivo del fenomeno cars i co dell altopiano carbonatico di Cantanhede (a NW di Coimbra Portogallo) .. ............ ........................... .. ......... 400 Quelques observations sur les lap i az de Majorque ..................................................... .. ............... 401 Karst of the Tatra Mountains ... ..... ............. .. ... ............ .. ..... .. .......... ...................... .. ............................. 405 Karst morphology in Bazovski part of Vratsa mountain (Stara Planina NW Bulgaria) ....... 409 Karstological investigations in the Middle Basin of lada Valley (Padurea Craiului Mountains Romania) ................................................................................. .. ...... 413 Macroscopic Upper Cretaceous paleokarstic features from SW Slovenia .. .. ........................ .. 417 Caves and karst areas i n Serbia ......................................................................................................... 421 Le paleokarst polyphase du Quercy (sud de la France) .......... ... ....................................... .. ........ 424 Speleolog i cal investigations in Saudi Arabia ........ .. .... .. ...................... .. .. ...................... ............... 425 Etude speleo-karstologique du systeme Muruk-Berenice (Monts Nakanar ; Papouas i eNouvelle Guinee) : resultats de expedition de 1995 et perspectives pour 1998 ... ............ 429 Lunan Shilin landscape in China ............... .. .. ......... .. ...... .... ........... .. .......... .. ... ................................. 433 Speleogenesis of the coastal dune karst of SW Australia ...... .. ............................... .. .................. 43 6 Session 5: Vulcanospeleology and pseudokarst Middleton Gregory & Halliday W R Caves of the Republic of Mauritius Indian Ocean ................................................ .. ........ .. ............. 437 Loiseleur Bernard Genese de la depression des Soulages (Causse de Sauveterre) : une explication nouvelle ...... .. .................. .. .......................................... 441 Kempe Stephan Lava falls : a major factor for the enlargements of lava tubes of the K i lauea and Hualalai Hawaii. .... ..... .. .. .. ...... ... .... .. ... .. .. .. ... ................... ... ................................ 445 Kempe Stephan & Oberwinder M The upper Huehue flow (1801 eruption, Hualalai Hawaii) : an example of interacting lava flows yielding complex lava tube morphologies............. .. .. .. 449 Kempe Stephan et al. Mapping lava flows by following their tubes : the Keauhou Trail / Ainahou Ranch Flow Field Kilauea Hawaii. ........ .. ................................... 453 Medville Douglas M & Medville H E. Recent exploration of lava tube systems in Kona and on Mauna Loa Hawa i i ...................... 457 Halliday William R Gaal Ludovrt Filippov Andrey G Unusual volcanic caves of Hawari Island, Hawa'ii ........ .. .. ... ....... .. ... .. ...... .. ..................................... 461 The model of development of basalt caves by slope movements..... .. .. ... .. .. ... ... .. .... .. ..... .... ..... 464 Gravity caves of Siber i an platform ....... ............................ ......................... ..... ............. .. ..... .. ........... 465 X P rocee d i n gs of the 12 t h I n t ernationa l Cong r ess of S p eleology, 1997, Switzerland Volume 1


Eszterhas lstvan Str i ebel Thomas & Schaferjohann V W i llems Luc Hav li cek David & Tasler A. Konsequenzhbhlen i n vulkanischen Gesteinen .................. .. .. .. .. .. ............ .. .................................. 469 Karstification of sandstone in Central Europe : attempts to validate chemical solution by analyses of water and precipitates ........ .. .. .... ...... .. 4 73 Karsts siliceux au Niger occidenta l...... .. .......... .. .............. .. ........................ .... .. .. ...... .... .......... .. .... .. ..... 4 77 Pseudokarst process and speleothems in Bohem i a Caves ............... .... ........ .. .... .... ...... .. ........... 481 Session 6: Glacial speleology Badino Giovanni et al Eraso Adolfo Romero et al Leszkiewicz Jan & Pulina M Krawczyk Wieslawa Ewa et al. Tremblay Marc & G i etl D Present Status of Speleological Researches i nto the Patagonian Glaciers .......................... 483 Investigations on the endorreic drainage of the south east part of Vatnajokull glacier.. ..... 485 Hydrologic systems in carbonate karst and in subpolar glaciers S i milarit i es and d i fferences ................ .. .......................... .. ................................. 489 Similarity between karst hydrologic system of the Werenskiold Glacier (SW Sp i tsbergen) and a karst.... ...... .......... .. ............ ........................ 493 lnlandsis 1994 : glacial speleology i nto the Greenland icesheet.. .................. .. .... .. .... .. .......... ... 497 P roceedin g s of the 12 '" I ntern a tional Co n gress of Speleology, 1997, S witzerland Vo l ume 1 XI


Swiss Speleological Society (SSS/SGH) Proceedings of the 12 th International Congress of Speleology Volume 1 Symposium 7 Physical Speleology La Chaux-de-Fonds, Switzerland, 10-17.08.1997


Etude scientifique de la grotte de Clamouse (Herault, France) par Jacq u es Choppy et Paul D u bois 182 rue de Vaugirard, F-75015 Paris I rue des Grezes, Impasse des Merlets, F-34070 Montpellier A bst ra ct : Scient i fic study of the grotte de Clamouse (Heraul t, F r ance) Since the discovery of the Grotte de Clamouse in 1945, various observations had been made. In 1986 scientific research began to be organized. And in 1991 a pluridisciplinary study, mostly sedimentological, was started, and various complementary measures on air and water were made in the cave. Simultaneously, observations in various fields came from the numerous cavers visiting the cave. The Grotte de Clarnouse is now a place where a large range of data permits a synthetic visions. And it is now the cave that has been most studied in France as far as physical speleology is concerned. Resume Des la decouverte de la grotte de Clamouse, en 1945, furent faites diverses observation. En 1 986, l es recherches commencent d'etre organisees. Et en 1991 debute une etude pluridisciplinaire essentiellement sedimentologique; mais en comp l ement diverses mesures concernent l'eau et !'air dans la grotte. Simultanement, la frequentation active de la grotte a provoque des observations dans des domaines tres divers. La grotte de Clamouse est maintenant le lieu d'un eventail de connaissances, permettant des visions synthetiques. Et c'est actuellement la grotte la plus etudiee de France en ce qui concerne la speleologie physique La grotte de Clamouse (Herault, France) est l'une des quelques grottes europeennes dans lesquelles des etudes scientifiques de haut niveau sont en cours. La publication prevue (au moment ou nous redigeons cette note) des actes du Collogue tenu a Clamouse a !'occasion du cinquantenaire de la decouverte speleologique en fournira la bibliographie, qui aurait double le volume de cette note. L 'e poque des observations La decouverte de la grotte de Clamouse en 1945 a vivement frappe les esprits : C'etait en Europe a peu pres la premiere cavite presentant une telle richesse de concretionnement. Les trar;;ages ont prouve une liaison de plus de 14 kilometres (distance considerable pour l epoque) l eau franchissant l'importante faille du Larzac pour parvenir a !'emergence de Clamouse. Et une analyse du mondmilch de la grotte a montre qu'il etait essentiellement compose de huntite, carbonate double de calcium et de magnesium, mineral recemment decouvert Apres l'amenagement touristique et notarnment a partir de 1968 se font des observations nouvelles : -L emergence est situee aussi bas que possible a la sortie des gorges de l'Herault dans la plaine littorale et la grotte suit la limite de cette plaine ; elle apparait done comme un drain lateral du massif. -L examen des cristallisations de Clamouse a ete determinant pour decouvrir leur etagement en altitude; celui-ci confirme dans d autres grottes est apparemment en relation avec un etagement microclimatique qui defie encore nos moyens de mesure ; ceci dans de s condition s de confinement relativement poussees Toutefois des mouvements d air interviennent dans la localisation et I orientation des cristaux d 'aragonite En 1986 la Societe qui gere la grotte de Clamouse commence a promouvoir des recherches : mesures de CO , de temperatures. Dans la Cave de Vitalis (Herault) qui se trouve dans l'amont du reseau fut observee pour la premiere fois cette nouvelle forme de concretion que sont les baguettes de gours. Celles de la grotte de Clamouse, qui sont en aragonite, ont pour ame un filament organique probablement issu de la decomposition du guano de chauve-souris En 1990 debutent des datations uranium/thorium de concretions qui sont desormais au nombre de plusieurs dizaines ; elles sont particulierement nombreuses dans la zone d'entree (partie aval de la grotte), oil quatre niveaux de conduits de moins en moins eleves, perpendicu l aires a la direction de l'Herault, s'etagent en direction aval de ce fleuve Ces datations montrent que les concretions soot moins anciennes a mesure que !'on descend en altitude : Au niveau superieur leur age depasse souvent 350.000 ans ce qui est la limite de la methode uranium/thorium ; par contre dans les parties inondables !or s des crues, le concretionnement semble avoir debute ii y a 30 a 40 000 ans. Une recherche multidisciplinaire En 1991, la Societe prend la decision de mener une recherche pluridisciplinaire sur la grotte, a laquelle plusieurs des meilleurs specialistes europeens adherent spontanement : belges, italiens anglais, irlandais et franr;;ais Une materialisation des stations topographiques est realisee, qui permet desormais de localiser avec precision observations et points de mesure Cette recherche concerne surtout l'etude des remplissages toujours grace a des datations absolues, mais en mettant en oeuvre d'autres techniques: Cartographie des divers types de concretionnements. Par photographie de mondmilchs au microscope electronique a balayage (M E.B ) on a pu distinguer l 'hydromagnesite, dont Jes cristaux ont la forme de plaques approximativement carrees (figure I) tandis que la huntite est en plaques plus irregulieres. D'autres photographies au M E B. ont revele !'existence de cristaux courbes de calcite (figure 2). Ces photographies montrent encore la presence de bacteries enkystees dans les concretions, ce qui est desormais classique ; mais aussi celle de filaments, qui sont vraisemblablement de nature biologique Malgre leur aspect apparemment caracteristique, I 'analyse au rayons X d'aiguilles d'aragonite a montre qu'elles sont souvent en calcite : To u t d'abord grace a des circonstances climatiques favorables, l'eau aurait mis en solution assez de dolomite pour que la cristallisation puisse se faire en aragonite Sympos i um 7 : Phys i ca l Speleology 3


Figure J (photo Yves BODEUR au M.E.B.) : cristaux d'hydromagnesite, grossissement 5000 Figure 2 (photo Silvia FRJSIA au M.E.B.): cristaux courbes de calcite 4 Proceedings of the 12 th International Congress of Speleology, 1997, Switzerland Volume 1


Depuis plusieur s centaines d annees ces conditions ne seraient plus remplies a Clamouse ; et l'aragonite instable se serait transformee en calcite au mains dans certains cas Mesures de paleomagnetisme. Par ailleurs au mains quatre types de remplissages argileux etage s (figure 3) furent observes : Des placage s d halloysite ( argile blanche d origine volcanique ) etages sur 30 metres d altitude qui recouvrent une argile rouge en un point. -Une argile bruue a l eboulis terminal. Le depot d argile grise des Terrasses est en forte pente v ers le fond de la cavite ; il semble done independant du precedent. Entin une argile assez rouge culmine a une altitude inferieure a celle des depots precedents Des mesures en cours visent a etablir la chronologie relative de ces depots Mais c est surtout la premiere application d une demarche nouvelle et riche de promesses : Une argile deposee par decantation a une e x tension plus generale que les autres sediments de Ja grotte et sa signature resulte necessairement de la nature des sediments exterieurs qui furent sa principale origine Des mesures du clirnat souterraiu continuent d'etre realisees Les temperatures de air de la grotte sont remarquablement homogenes proche s de 14 6 C sauf dans les zones particulierement hautes (> 15 C) OU basse s (< 14 C) ; ces temperatures sont v raisemblablement un peu superieures a la temperature du lieu Les temperatures de s eaux stagnantes sont pratiquement egales a celles de !'air ; ce qui montre un exceptionnel equilibrage therrnique dans la grotte Les pressions partielles de CO sont decroissantes depuis le fond (0 53 % a l eboulis terminal 0 32 % en direction de l entree naturelle) avec un Jogique gradient decroissant en fonction de !'altitude dan s une section de galerie Le taux de radon est faible dans la grotte n atteignant en un point que 800 Bq / m 3 Une s tation de mesure en continu consacr e e a l'eau de percolation (vitesse de transit composition etc .), est installee tbouli tannin I 2-hlle plllM dans la grotte. La source du Drac est, a l'amont de Clamouse une emergence ascendante naturelle; lors des pompages de cette source, le rabattement est lineaire ce qui prouve qu une nappe captive contribue a son alimentation ; cela est confirme par observation de marees terrestres dans cet aquifere La transmission du signal entre le Drac et Clamouse est 25 fois mains rapide !ors d un trai;age que d'un pompage ; ce rapport relativement faible entre transit de flux et transfert de pression montre que la circulation n etait pas entierement en regime noye !ors du trai;age De nouvelles observations La frequentation active de la grotte provoque de nouvelles observations : De nombreux indices prouvent qu'une irnportante colonie de chauve-souris d au mains deux especes occupait la grotte mais qui a completement disparu longtemps a v ant la decou v erte de celle-ci Sortant d'un trou dans une paroi une v enue d'eau sous pression faisait un jet de 40 centimetres environ ; puis durant une periode venait un jet d'eau et d air et enfin d air seul avec sifflement ; le jet d eau reprenait au bout de 10 minutes environ. Une fracture courbe est visible pres de l eboulis terminal. Decouverte de nouveaux sites d aragonite coralloi'de et d'une stalagmite d'aragonite ; observation de fistuleuses et d excentriques d aragonite, tout ceci si on peut se tier a aspect macroscopique Decouverte de coupelles et de stalagmites a section triangulaire necessitant des conditions de confinement relativement severes et qui sont done posterieures a I etablissement de celles-ci. Dans une zone relativement proche de la paroi exterieure plusieurs stalactites proches ont ete cassees a peu pre s au meme 1 Coulolr de ntalJoyalla -----~18m 106m~.,.,.._ Grotte de Clamouae (H6rault) REMPLISSAGES ARGILEUX Coupe ach6matique cotu par rapport au nlveau da la mer 1 -Dlllclase ,J Figure 3: grotte de Clamouse (Herault), coupe longitudinale ultra-schematioque representant les niveaux de remplissages argileux Sympos i um 7 : Phys i cal Speleology 5


niveau, et ont repousse; leur fracture ne peut guere se comprendre que par l'ebranlement d'un tremblement de terre. Entin, la grotte de Clamouse est le prototype d'une cavite a priori paradoxale : son p l an est pour l'essentiel rectiligne, sans qu'a cette echelle une contrainte tectonique puisse etre invoquee alors que presque toutes les formes de creusement sont phreatiques II faut considerer cette grotte comme typique d un fonctionnement en regime noye durant les crues, avec une erosion assez active pour imposer le plan genera l de la grotte, la vitesse demeurant toutefois suffisamment moderee pour que la roche encaissante, dolomitique presente des formes de dissolution differentielle. A proximite de l'eboulis terminal la section de conduit est celle d un tube, resultat d'une action comparable dans une roche non dolomitique. En maigre, ii est probable que le creusement etait trop modere pour que la morphologie en ait conserve Jes indices Conclusion Les recherches dans la grotte de Clamouse ont ete commandees par des observations plus ou moins ponctuelles ou les preoccupations des chercheurs qui ont participe Par le volume de ces recherches, Clamouse est maintenant le lieu d un eventail de connaissances, qui en fait la grotte la plus etudiee de France en ce qui concerne la speleologie physique L'ensemble des resultats obtenus donne une vision tout a fait nouvelle de la grotte, liee pour une part a la transformation profonde des conceptions speleologiques depuis 25 ans. Les syntheses se font tres progressivement mais ii nous semble qu 'un protocole pre-etabli, probablement arbitraire n'aurait pas permis d'avancer aussi vite dans la connaissance de la grotte 6 Proceedings of the 12 'h International Congress of Speleology, 1997, Switzerland Volume 1


Hohlensedimentologie in Osterreic h von R ud o l f Pavuza Karstund hohlenkundliche Abteilung, Naturhistorisches Museum Wien Museumsplatz I /10, AI 070 Wien, Osterreich Abstract In the Austrian Eastern Alps extens i ve sedimentological investigations are conducted mainly in the course of a few paleontological and archeo l ogical excavations currently. Therefore the chance to encounter sediments of an age older than the late Pleistocene i s comparatively small thus giving an insight only to the very last phases of cave development. Two examples from bear caves in the Austrian Alps will document this timespan where in one case some indications on very much older sediments could be observed. More vertical sedimentological information would help to clear up the chronology of the cave development remarkably thus demanding for more "pure" sedimentological excavations i n the Alps Z u sa mm e nfa ss un g Umfangreichere hoh l ensedimentologische Untersuchungen werden in Osterreich fast immer nur im Zusammenhang mit palaontologischen oder urge sc hichtlicben Grabungen gemacht. Dadurcb ist auch die Wahrscheinlichkeit Hohlensedimente zu finden, die vor dem Jungpleistozan sedimentiert wurden, eher gering. An zwei Beispielen aus den Nordlichen Kalkalpen wird die sedimentologische Entwicklung von Barenhohlen kurz dokumentiert wobei wenigstens in einem Fall ein Hinweis auf bedeutend iiltere Sedimentanteile vorliegt. Bei einer grol3eren Zahl vertikaler Sedimentprofile

Lithologie Kiesfraktion Sandfraktion SCHUTTLAGE (NOP) AKLEI'.'/SAUGERSCHICHT ~~~~+~onJsU @@~ 1 ~t~:c~t 1 ~~~@ :t l EB < TY. i~'. ~== ~ tttt=~=tttttt=~H Hl ~I I i:;: i: i i ; ; : i : : 1 : ; rn U;; i:;; i: i; i; iii 1 i; I -I < -) B BRAUl'iER LEHt\l + SCHUTT I::;:;:::; : ;::! : : : ;: I l:!li:1:!:!:!: 1 1!1;1r :J ~': I B BRAUNER LEH'.\! I ; : 1 ; : ; 11 ::::::::::::::::::::::::::t:J 1 : 1 1 ; i 11 ::::::::::::::::::::::::::::r::: l C GELBER LEH!\! ~!!!!!!!!!!!!!!!!!!!I I 1: 1 ; 1 1 ; : ; 1 t :::::::::::::::: 1 ~ \:\I:\'::: = :::::::=::::::::::::::::::::::::::::t:I l :: l ~~'--'-'. 1 1.:: ....__ ;'.: I __, D AUGENSTEI:'700 000 a) zur Anwendung kommen. Fur dieses Ziel wird aber fraglos eine ungleich grol3ere Zahl an Vertikalprofilen erforderlich sein, als dies bei den palaontologischen Untersuchungen der Fall ist, da hier ja a priori keinerlei Hinweise auf die Erfolgsaussichten vorliegen. Freilich mull im Falle palaontologischer oder urgeschichtlicher Funde die Grabungsstrategie und -technik diesem Umstand Rechnung tra gen und der Bergung dieses Fundgutes Vorrang einriiumen. 6. Literatur PA VUZA, R. 1992: Die Sedimentologie des Nixloches bei Losen stein-Temberg (0 0.). Mitt.Komm.Quartiirforsch. Wien,8 : 17-19 PAVUZA R 1995:Die Sedimente der Gamssulzenhohle (War-schen eck, Oberlisterreich). Mitt. Komm. Quartiirforsch Wien 8 : 23-26 8 Proceed i ngs of the 12 'h International Congress of Speleology 1997, Sw i tzerland Volume 1


Les mega-remplissages detr i tiques de l a Cueva Fresca par O livier Forgeot 4 residence Gericault, 78150 Le Chesnay, France Abstract The Cueva Fresca (Cantabria Spain) is a cave with many big galeries in canyon shape. The size of its canyons put this cave among the first ones in the world. The present text will talk about huge amounts of sand we can see all along these canyons, and which have probably filled all the cave in the past. The chaotic floor reveals us the location of these fillings and undermining phenomenas. Resume La Cueva Fresca fait partie des grottes qui possedent les plus grandes galeries en forme de canyon au monde. Cet expose presente Jes gigantesques remplissages de sable qui ont probablement colmates la cavite dans des temps tres anciens. Le profil accidente du sol actuel nous renseigne sur la position de ces remplissages souvent situes sous des eboulis ou entrecoupes par d'enormes soutirages. Presentation geographique La Cueva Fresca se developpe dans le massif de Monte Llusias qui se trouve a !'Est de la cordillere cantabrique, elle meme situee au Nord de la peninsule iberique. Le sommet le plus proche de cette grotte est celui de Colina qui culmine a 1458 metres d'altitude. L'entree de la Cueva Fresca s'ouvre a 440 metres d'altitude, en rive gauche et a une centaine de metres au dessus du rio Asen. L'entree est au meme niveau que la voute des grandes galeries fossiles. Les eaux de la Cueva Fresca qui s'ecoulent a l'etage inferieur, rejoignent directement le lit du rio Asen au travers des eboulis de la vallee et finissent a une trentaine de kilometres au Nord dans !'ocean Atlantique. Le sud du massif de Monte Llusias sert de ligne de partage des eaux. Les pluies qui tombent au sud de cette ligne sont drainees par l'Ebre et evacuees vers la mer Mediterranee situee a plus de quatre cents kilometres au Sud-Est. Presentation geologique Un anticlinal dont !'axe est oriente du nord-est au sud-ouest fait apparaitre, pres d'Arredondo, le haut d'une epaisse serie argilo-greseuse wealdienne datee du Cretace inferieur qui realise un substratum impermeable. Au dessus de cette couche impermeable, nous trouvons les calcaires tres purs a facies urgonien du massif de Pena Lavalle, puis l'epaisse couche des gres d'Ason dans lequel s'ouvre le profond ravin de Rolacia. Au sud du ravin, le massif de Colina est compose d'une alternance de series greseuses et de calcaires a rudistes et polypiers. Cette altemance forme ce que !'on appelle de complexe calcairo greseux d'Ason date de l'Aptien superieur. Ce complexe repose sur le socle des gres d'Ason qui constitue un ecran impermeable entre les deux massifs. Le Val d'Ason a ete taille a contre pendage des massifs de Pena Lavalle et de Colina qui ont un pendage commun en pente douce vers le sud-est. La Cueva Fresca a pour particularite de s'ouvrir exactement a !'intersection de la vallee et du plan de contact entre le complexe calcairo-greseux superieur et l'epais socle impermeable des gres d'Ason. On remarquera qu'avec l'enfoncement du niveau hydrologique de base sous la contrainte du pendage du socle impermeable, l'entree fossile de la Cueva Fresca est plus au nord que la resurgence de l'actif au niveau de la source du Manantial de! Huerto del Rey. Presentation historique Les p r emieres reconnaissances spe l eologiques franyaises dans cette region ont ete effectuees des 1958 par le speleo-club de Dijon et dirigees par B. DE LORIOL. Les expeditions etaient placees sous l'autorite du Museum National d'Histoire Nature! de Paris et du Laboratoire de geologie de la Faculte des sciences de Dijon. En 1964, le laminoir qui fait suite a la salle d'entree fut desobstrue par l'equipe du Dr CASTIN, et le Canyon d'Exploration reconnu jusqu'au fosse du Bloc 64. L'immense reseau fossile fut explore par le S.C.D. les annees suivantes pour aboutir a la publication en 1975 d'une topographie globale. En 1968, C. MUGNIER soutint a l'universite de Dijon une these sur le karst de la region d'Ason qui deviendra la reference des travaux ulterieurs. L'exploration reprit un deuxieme scuffle lors des annees quatrevingts par l'arrivee de plusieurs groupes speleos sur le terrain tel que le Speleo-club de Chablis, Jes Speleos Grenoblois du Club Alpin Franyais et le Speleo-club de Paris du CAF Ile de France. La jonction entre la Sima Tibia (860m d'altitude) et la Cueva Fresca fut ainsi realisee en 199 0 par l'equipe de B. LISMO ND E du S.G.CAF A cote d e !'aspect sportif, le Speleo-club de Paris demarre, des 1988 sous !'impulsion de P h MORVERAND, le programme FRESCA d'etude scientifique sur ce reseau. Le developpement fit un bond et depassa Jes vingt cinq kilometres en mai 1994. Depuis, des observations morphologiques sont r egulierement rapportees, des datations U/Th sont faites, et la comprehension de la speleogenese progresse. Introduction C'est une vue globa l e mais non exhaustive des remplissages de la Cueva Fresca qui amene a presenter, dans cet article, l'hypothese du comblement total de ses canyons dans le passe. Ceci est observable aujourd'hui par la presence de mega remplissages detritiques. La premiere partie de cet expose consistera a decrire Jes remplissages que !'on rencontre ou suppose !o r s d'une promenade d'environ d eux kilometres dans Jes grandes galeries entre l'entree de la Cueva Fresca et le terminus du Canyon Rouge. La deuxieme partie discutera des remplissages observes, et des soutirages qui Jes separent, face a l'hypothese du colmatage de la Cueva Fresca dans des temps tres anciens. Symposium 7 : Physical Speleology 9


Observations de terrain Apres avoir franchi les deux laminoirs d'entree de la Cueva Fresca ii faut escalader le muret d'un massif stalagmitique imposant et genere par les ecoulements au travers des fractures de detente du massif calcaire a proximite de la vallee En marchant plus profondement sous une large voute rougeatre et basse, nous debouchons sous le plafond et dans l'axe du Canyon d'Exploration. Une descente d'environ quarante metres d'altitude sur une pente tres glissante, puis le passage d'une premiere vire au dessus d'un profond fosse, nous permettent d'atteindre le fond horizontal du canyon d'exploration Sur les parois du canyon on remarquera de larges banquettes subconcaves et metriques Pour cette partie du Canyon d'Exploration, le soutirage aurait ete effectue par le Reseau des Griffes qui s'ouvre en contrebas de la descente et sous la premiere vire. Les ecoulements ont encroute la pente d'une epaisse gangue de plusieurs couches de calcite comme l'a revele le sondage effectue a la pioche La calcification est superieure aux cinquante centimetres creuses. On notera toutes les montees et descentes qui s'observent le long des canyons. En poursuivant vers le fond, apres etre descendu au niveau bas du Bivouac, ii faut a nouveau gravir une coulee calcitique sur environ trente metres de denivele pour se retrouver pres du plafond du Canyon d'Exploration. Apres un court pa sage horizontal sous la voute, ii nous faut redescendre de plusieurs metres parmi un chaos de blocs pour atteindre la vire du bloc 64. Cette vire nous fait passer au dessus d'un profond fosse. La progression continue par une galerie boueuse de quelques metres de large qui permet de contoumer le colmatage du Canyon d'Exploration et d'arriver au pied du Reseau du Caviar. La galerie principale du Reseau du Caviar contient un enorme remplissage de sable fin qui est soutire a quelques endroits. Un leger depot de galets s'obs~rve dessus Le sommet de ce remplissage est remarquablement horizontal sous une voute presentant des formes de corrosion en chenal. En amont du Reseau du Caviar se trouve le Canyon de l'Eboulis, ou le Puits de l'Araignee a produit un soutirage. Apres avoir Franchi la Vire de l'Araignee au dessus du confluent du Canyon de l'Eboulis et la galerie du Tracastin, ii faut mooter sur un chaos de blocs, ou s'ouvre un peu plus loin le Puits du Sablier Ce puits est une demonstration de la puissance des soutirages le long des canyons En effet, ce gigantesque entonnoir, toujours actif par un ecoulement qui perce le plafond offre une coupe naturelle du remplissage et de la base rocheuse du canyon. Nous arrivons maintenant dans la Cinquieme Avenue La morphologie de cette large galerie a voute elliptique et la continuite topographique nous indiquent que nous sommes dans la suite du canyon qui est presque rempli jusqu'au plafond Toutefois l'analy e du remplissage est empechee par un tapis de blocs d'effondrement On remarquera les enormes lames de decollement le long des parois de la galerie qui trouvent probablement leur origine dans des phenomenes de detente lors de l'assechement du reseau. La Cinquieme Avenue debouche a mi-hauteur dans !'immense salle Rabelais qui mesure de 80 a 120 metres de diametre et I 08 metres de haut. (MARQUET, 1995) Le chaos de blocs sur lequel nous nous trouvons se poursuit tout autour de la salle. A la place d'un classique cone d'effondrement, on observe un enorme vide central qui suggere qu'un soutirage a eu lieu au centre de la salle. De la salle Rabelais, partent le Canyon Nord et le Canyon Rouge ous nous dirigeons vers ce demier et remarquons que la voute de celui-ci est sensiblement a la meme cote que celle de la Cinquieme A venue. En descendant l'eboulis de la salle Rabelais vers le Canyon Rouge, a proximite du Passage du Shunt nous pouvons observer les traces parietales d'une corrosion sous remplissage avec de larges cannelures verticales. Un peu plus loin dans le bas du Canyon Rouge, le passage peut se faire dans un surcreusement du canyon et sous un plancher stalagmitique reposant sur une epaisseur d'environ deux metres de sable siliceux blanc tres fin Toutefois, ce remplissage parait relativement jeune et fait probablement parti du demier cycle remplissage-surcreusement du Canyon Rouge car ii est lui meme contenu dans un surcreusement plus vaste. Dans le meme secteur, on trouve environ trente metres plus haut, un plancher stalagmitique date d'environ 330 000 ans (DELANNOY et MORVERAND 1989) Apres avoir franchi le chaos de bloc au bas duquel se trouve le Canyon des Cupules qui mene vers le Grand Racourci et que nous laissons a notre gauche sous une strate inclinee, ii nous faut gravir un deuxieme chaos qui nous permet de prendre pied sur un remplisssage incline et dont la forte pente verse dans le soutirage qui se developpe tout du long de la paroi de gauche Toujours en nous dirigeant vers le sud, a mi-longueur du Canyon Rouge le ressaut du Trou Blanc nous oblige a descendre un puits de dix metres environ dans un remplissage de breches greseuses. II nous faut alors passer sous le remplissage dans une galerie boueuse qui mene vers la suite du canyon Peu apres, l'escalade d'un talus d'une dizaine de metres nous permet de remonter sur le remplissage. Nous progressons rapidement sur le sol plat quoique parseme de blocs. Nous atteignons ensuite un chaos d'enormes blocs, ou la trace nous indique le passage entre la parois de gauche et l'enorme monolithe En avarn;:ant plus en avant dans le Canyon Rouge nous traversons !'immense et longiforme Salle du Monolithe ( 150 m x 50 m x 50m environ) qui doit probablement son existence a un phenomene d'effondrement de la voute et des parois du canyon On remarquera toutefois que ce chaos forme deux gigantesques banquettes d'une vingtaine de metres de large de part et d'autre d'un surcreusement axial decametrique. A l'autre extremite, on retrouve un conduit qui ressemble bien au sommet d'un canyon remplis jusqu'a une dizaine de metres de son plafond Sur le sol, de grands go u rs fossiles sont remplis d'un sable qui ne reagit pas a l'acide. La vot'.ite bien plus basse que celle de la Salle du Monolithe presente des coupoles metriques Au bout de cette galerie, la suite s'effectue par un laminoir sous un epais plancher stalagmitique, le passage du Mille-Feuille. Ce plancher materialise probablement un leger ecoulement d'eau sur le colmatage d'un ancien siphon. On ressort de ce laminoir dans un lac fossile ou l'eau est remplacee par le brillant d'une dalle de calcite noire parfaitement horizontale et d'une dizaine de centimetres d'epaisseur. La galerie reprend la fonne d'un canyon qui se poursuit jusqu'a la demiere salle dans laquelle descend sur la gauche un enorme eboulis d'une cinquantaine de metres de haut et alimente par une tremie suspendue dans le plafond. Revenons au bas de l'eboulis, nous retrouvons la suite du canyon dans !'axe du precedant. La promenade s'arrete malheureusement au bout d'une vingtaine de metres sous un pittoresque plancher stalagmitique suspendu a plus de trois metres de haut et soude au milieu des parois du canyon qui sont ecartees a cet endroit d'environ six metres. Le remplissage qui servi d'echafaudage au plancher, a ete partiellement deblaye. On trouve de petits galets prismatiques scelles sous le concretionnement. Le terminus actuel du Canyon Rouge se situe au niveau de ce second paleo-siphon derriere lequel la suite, indiquee par la presence d'un courant d'air, est interdite par une tremie. 10 Proceedings of the 12"' International Congress of Speleology, 1997, Switzerland Volume 1


Preambule Un remplissage peut s'observer de plusieurs manieres. On peut le remarquer par la presence d'une forme (talus banquette sol plat) ou par son absence (soutirage, plancher sta l agmitique suspendu trace parietale de corrosion sous remplissage). II peut etre visible (coupe naturelle sediments) ou suppose (altimetrie anormalement haute avec couverture de blocs ou de calcite) On peut aussi le mettre en evidence par exemple, en realisant un sondage profond sous un encrotltement calcitique. Discussion La coupe schematique presentee sur la figure I presente un profil estime des remplissages le long de !'axe qui va de l'entree de la Cueva Fresca jusqu'au terminus du Canyon Rouge A priori , en attendant les resultats des mesures fines d'altimetrie que le SCP a mis en chancier, on peut penser que les votltes des canyons le long de cet axe sont quasiment a la meme altitude, a une pente d'ecoulement pres Nous pouvons dresser la liste des talus en allant de l'entree vers le fond Nous rencontrons successivement le talus d entree, le talu s entre le Lac et le bloc 64 le Reseau du Caviar, le talus entre le Canyon de l'Eboulis et l'arrivee du Grand Racourci La Cinquieme Avenue et son altimetrie haute, le pourtour de Ja Salle Rabelais fonnant un talus entre son centre et le debut du Canyon Rouge, le remplissage peu apres le chaos de depart du Grand Racourci et qui va jusqu'au Trou Blanc, la suite du remplissage apres le Trou Blanc, Jes mega-banquettes de la salle du Monolithe le remplissage de la galerie des Gours jusqu'a l'autre cote du laminoir du Mille-Feuille et le remplissage du tronc,:on de canyon terminal. La caracteristique majeure de tous ces talus est qu'ils culminent a peu pres tous a la meme altitude et ce sur pres de deux mille metres de deve lopp ement. Ceci incite done a penser que ces talus sont le temoignage d'un rneme remplissage entrecoupe de creux provoques par des soutirages posterieurs a la mise en place de ce mega-remplissage Pour que des creux puissent etre consideres conune resultat d'un soutirage, ii nous faut y remarquer !'action d'une galerie ou d'un reseau inforieur. Ainsi le premier creux evoque celui entre la descente d'entree et la remontee apres le Bivouac peut s'expliquer par !'action du reseau des Griffes qui communique en plusieurs endroit avec le Canyon d'Exploration et avec l'actif situe une cinquantaine de metres plus bas. Le deuxieme creux decrit correspond au fosse du Bloc 64 qui lui aussi communique avec l'actif Le niveau bas du Canyon de l'Eboulis est probablement dtl a !'action du Puits de l'Araignee qui s'ecoule aussi vers l'actif. Le cone du Puits du Sablier cree par la chute d'eau qui perce le plafond mene aussi a l'actif. L'enorme creux de la Salle Rabelais reste encore une enigme quoique un court tronc,:on de galerie partant de la base de la salle vers le Nord-Est soit suspecte d'avoir un role dans un phenomene de soutirage II ne faut pas oublier d'autre part !'action de galeries inforieures qui s'arretent sur les chaos peripheriques de la Salle Rabelais. Ainsi on remarquera que la tremie amont du meandre du Grand Bourbier n'est autre que le bord enfoui de la Salle Rabelais et que la galerie de la riviere Tibia change de direction lorsqu'elle bute sur le bord de la salle Rabelais. On peut alors formuler l 'hypothese que la riviere Tibia lorsqu'elle etait a moindre profondeur traversait la Salle Rabelais pour rejoindre le meandre du Grand Bourbier et aurait participe au deblaiement central de la salle Puis son cours changea de direction pour emprunter le cheminement actuel plus au sud sous !'action conjointe de l'enfouissement du pendage de la fracturation et d'un eventuel effondrement dans la Salle Rabelais Le niveau bas de debut du Canyon Rouge s'expliquerait par !'action du Canyon des Cupules capture sous une strate au pied du Canyon Rouge, et qui mene au bas du Canyon du Grand Racourci Ce passage aurait pu drainer les eaux venant de l'ancienne riviere Tibia en les faisant remonter sur une centaine de metres de la salle Rabelais vers le Canyon Rouge L'evacuation aurait pu se faire par le Grand Racourci jusqu'au Canyon de l'Eboulis, contournant ainsi le remplissage de la Cinquieme Avenue, en formant le Carrefour de l'Araignee (BIGOT 1994) Plus en amont dans le Canyon Rouge nous avons le Trou Blanc qui presente avec son passage sous le s ediment un drain inferieur evident qui se dirige egalement vers le Grand Raccouci Notons que l'actif actuel passe a cinquante metres sous ce secteur. .. La tranchee axiale d e la salle du Monolithe fait penser a !'action d'un ecoulement, mais cette zone est ma! etudiee du fait de la difficulte a se deplacer dans ce gigantesque chaos Entin dans la zone terminale, le niveau du remplissage est materialise par les planchers stalagmitiques fossiles suspendus. En dessous le soutirage a ete deblaye sur plusieurs metres de haut sans qu'il y ait d'explication apparente Nous pouvon s seulement remarquer que cette zone du canyon se trou v e a !'intersection d'une fracture qui lui est perpendiculaire et qui a genere la salle de la grande tremie On peut done supposer que la karstification s'est egalement produite dans un etage inferieur et soutire le remplissage de cette zone Les autres remplissages de la Cueva Fresca On rencontre d'autres remplissages dans la Cueva Fresca, mais de nature bien differente de celle des remplissages sableux des grands canyons. Un premier type s'observe dans les etages superieurs. Par exemple, a la cote + 70m environ, le meandre au Plancher Perce comporte un remplissage de breches bien nivele. La Galerie du Marchand de Sable suspendue a plus de quarante metres au dessus du plafond des canyons contient un remplissage altemant sables et breches dont la coupe naturelle s'observe surplus de dix metres de hauteur (BIGOT 1992) Un deuxieme type de remplissage conceme les galets ovoi"des probalement d'origine fluviatile que !'on rencontre en grande quantile dans certains endroits de la Cueva Fresca (MORVERAND 1992). Un epais remplissage de galets existe dans le Reseau des Galets situe dans le plafond du Canyon d'Exploration (GISSELBRECHT 1991 ) Cette galerie relati v ment proche de la vallee et a une altitude voisine de l'entree de la Cueva Fresca permet d'etayer l'hypothese d'une injection a partir du Val d'Ason lorsque le Canyon d'Exploration etait rempli presque jusqu'au plafond Les caracteristiques des depots de galets en fonction de la distance de penetration dans la grotte demandent a etre etudiees. En effet, nous pensons avoir remarque une decroissance des depots en allant vers le fond, pour terminer par une epaisseur de quelques galets polis au dessus de la masse sableuse et homogene du Reseau de Caviar a plus de cinq cents metres de l'entree Les depots de galets melanges a une matrice sableuse que l'on observe dans le Reseau des Griffes au pied du Canyon d'Exploration resulteraient par contre du soutirage effectue par ce reseau sur le paleoremplissage sableux recouvert de galets du Canyon d'Exploration aujourd'hui disparu. Le Reseau des Galets suspendu pre s que a la verticale du Reseau des Griffes aurait ainsi conserve une partie du depot originel de galets Symposium 7 : Physical Speleology 11


Conclusion La Cueva Fresca est une cavite complexe, chargee d'une longue histoire, et qui est tres interessante a etudier. D une part pour la comprehension de la cavite elle-meme, et d'autre part pour la comprehension de !'evolution du massif entier et de son interaction avec le Val d'Ason. Cet article a presente le profil accidente du sol le long d'un axe majeur de la Cueva Fresca. L'observation des creux et des bosses qui se succedent permet de jeter un nouveau regard sur Jes gigantesques remplissages que l'on observe et d'en subodorer d'autres caches par l'usure du temps, blocs tombes, calcification et soutirages par exemple Cette demarche globale conforte l'hypothese d'un quasi colmatage en des temps tres anciens ( > 330 000 ans) d'un beau canyon souterrain (20m x 50m x 2000m) sur toute sa longueur par un sable siliceux fin et homogene. R efe r e n ces Articles parus dans les differents numeros de la revue 'Sous le Plancher' du Speleo club de Dijon en autres tome 3 fasc 4, tome 5 fasc. 1, tome 8 fasc. 2, et n-2. Articles parus dans Jes differents numeros de la revue 'Grottes & Goujfres' du Speleo-club de Paris, en autres nl l l, 115, 116, 117,119,120,122,123,124,126, 130,et 134 BIGOT, J.Y. 1992. Les remplissages des galerie bleue, lucarne, cinqieme avenue de la Cueva Fresca. Grottes & Goujfres + ---1--1 ------CUEVA FRESCA COUPE SCHBIATIQUE ESTIIIEE 126, 31-44. BIGOT, J Y. 1994. Les canyons de la Cueva Fresca en passant par le Grand Raccourci Grottes & Gouffres 134, 15-28 BIGOT, J Y. & MORVERAND, Ph. 1994. A propos des canyons de la Cueva Fresca en particulier et ceux de la Cantabrie en general. Actes de la quatrieme Rencontre d'Octobre, Pau 1994, Publication du SCP, 28-37. CASTIN, P. & KIEFFER, J P. 1975 Travaux dans le Val d'Ason, Spelunca, 3 (1975) p 3. DELANNOY, J.J. & MORVERAND, Ph 1989 Contribution a la connaissance de la karstogenese du massif de Pena Lavalle Grottes & Goujfres 111, 9-21 FORGEOT, 0. 1996. Mega-remplissages detritiques dans la Cueva Fresca, Actes de la sixieme Rencontre d' Octobre, Osselle 1996 Publication du SCP, 49-55. GISSELBRECHT, 0 1991. Le reseau des Galets. Grottes et Gouffres, 120, 23-25 MARQUET, P. 1995. Altimetrie a la Cueva Fresca. Rapport inedit. MORVERAND, Ph. 1992. A propos de certains galets de la Cueva Fresca, Actes des Journees Pierre Chevalier. Memoires du Speleo-club de Paris 16, 217-222. MUGNIER, C. 1969. Le karst de la region d'Ason et son evolution morphologique. These de troisieme cycle de la Faculte de Dijon. ., a ._, ....__.,__.1. _1., .. ............... ,....1T. Val d'Ason Figure I : Coupe sc h e ma tique de la Cueva Fresc a de l'entree au te rm inus du Canyon R ouge 12 Proceedings of the 12 1 h International Congress of Speleology 1997 Switzerland Volume 1


Shape of Fluvial Pebbles in Surface and Subsurface Streams from Moravian Karst Czech Republic Kadlec J a roslav Czech Geological Survey, 118 21 Prague, Czech Republic Abstract Streams draining karst areas often flow through caves which have significantly different hydraulic parameters than surface channels. Moravian Karst, Czech Republic, yields a good possibility to compare different degrees of reworking of pebbles transported by streams in surface and subsurface environments For the study of the reworking of fluvially deposited pebbles, modern sediments of two streams flowing through the Moravian Karst cave systems were chosen Almost 1700 greywacke pebbles (size interval 16-31.5 =) were measured in total. The changes in shape and roundness of pebbles in the modem stream channel persuasively document the dependence of reworking of clasts of fluvial sediments on the hydraulic conditions 1. Introduction Karst environment is typical for its unique hydrologic conditions which have dominant influence on the means of transport and reworking of pebb l es of fluvial sediments. Several authors studied the shape of pebbles in surface streams and the level of their reworking (see e g MILLS, 1979) Pebb l es deposited by subsurface karstic streams were also investigated (e.g BULL, 1978; KRANJC, 1981) l

o 1 z J i. 5km 1 cm z[z=J lOZ:) 'C0 Figure 2 : The northern and central parts of the Moravian Karst; 1 boundary between Devonian limestones and Proterozoic Brno massif, 2 boundary between Devonian limestones and Lower Carboniferous non-karstic sediments, 3 surface stream with a sampling point, 4 subsurface stream with a sampling point terranean river resurges on the bottom of the Pusty leb Valley The Punkva River then continues on the surface and meets the Svitava River The knowledge of greywacke pebble roundness gained by the study of the Bila voda Stream sediments was tested on pebbles of modern sediments of the Jedovnice Stream. This stream flows through cave systems in the central part of the Moravian Karst (see Fig 2) Its catchment area is located east of the karst area, being formed by greywackes, siltstones and shales. The catchment area is about a half of that of the Bila voda Stream. The length of the surface reach of the stream is 11.5 km. Close to the boundary between limestones and non karstic rocks, the stream vertically sinks by 90 m under the surface into the Rudicke Propadani Cave. From this cave the Jedovnice Stream proceeds to the Byei skala Cave and resurges to the surface close to its mouth. The underground reach of the stream is 18.5 km long and has a gradient of 1.2 %. From the resurgence downstream, the Jedovnice Stream flows in a surface channel and finally meets the Svitava River after4 km. 3. Roundness and shape determination of greywacke pebbles The Bila voda and Punkva streams Reworking of the clasts during transportation both in the surface and underground channels of the Bila voda and Punkva streams is expressed by the roundness and shape of the pebbles. Each sample consisted of 60 pebbles of greywacke from the size interval 16-31.5 = Roundness is determined by the ratio of the diameters of the largest and smallest circle drawn into the plane of the largest projection of a pebble (DOBKINS & FOLK, 1970) The resultant roundness value of the suite of 60 pebbles is expressed by the median of the set Figure 3-B shows that roundness in the surface flow gently rises downstream (samples I7). In places where short tributaries carrying less reworked pebbles enter the Bila voda Stream the roundness drops (samples 2, 4, 7) The profile of the flow rapidly changes in its character at the ponor of the Bila voda Stream (fig 3-A), where the stream sinks vertically 30 m into Rasovna Cave The hydrodynamic potential of the river increases and so does the rate of reworking of the pebbles. In the initial tract of the cave system, the stream flows through narrow places with long siphons often without open air surface at an increased velocity. The roundness of pebbles rapidly increases as a result of collisions with walls of narrow cave corridors. Also the more frequent collisions of pebbles with each other contribute to the rapidly increasing clast roundness Another increase in the pebble roundness occurs after the confluence with the Luha Stream which transports more reworked clasts. The pebbles collected in a surface channel closely upstream of the ponor at the village of Sloup (sample 100) show a better roundness than those found closely upstream of the ponor of the Bila voda Stream near the village of Holstejn (Fig 3-B). The steeper profile of the Luha Stream increases the hydrodynamic potential of the flow causing higher roundness of clasts It is not possible to reliably determine the roundness of pebbles in the surface flow of the Punkva Stream from its resurgence in the Pusty leb Valley to its confluence with the Svitava River Limestone and granitoid clasts dominate in the channel. One can assume that the roundness of pebbles in the surface flow of the Punkva Stream will not further increase, or will increase only very slightly (see KRUMBEIN, 1941) The shape of pebbles is determined by the ratio of all the three axes and calculated by a formula used by (DOBKINS & FOLK 1970) The resultant value from each sampling point is expressed once again as a median of a set of 60 pebbles of the fraction 16-31.5 = Positive values represent a rodlike pebble shape, negative values an oblate shape. An obvious trend in flattening of pebbles can be seen from the source of the Bila voda downstream (Fig. 3-C, samples 17). In places where short tributaries bring less reworked clasts, rodlike pebbles are more abundant (samples 2 4). Continuous flattening downstream results from shortening of the shortest axes This is caused by sliding of pebbles along the bottom of the channel. The flux pushes the clasts to the bottom by a 14 Proceedings of the 12 th International Congress of Speleology 1997 Sw i tzerland Vo l ume 1


force opposite to hydraulic lift ( B LATT et al., 1972 p 108) The same force causes imbrication of pebbles. ...... ,.., __________ ----------J? ... ... A ... ... l50 .... 8,.,. :;l ( ... ~,. ~ @I +--+---~----.;o"i--"1-~-+~ 1 al 20 ,. .. I I I I I I I I I I I I I .... Figure 3 : A gradient curves of the Luha Bila voda and Punkva streams with sampling points B roundness of the pebbles C shape of the pebbles On the rim of the Moravian Karst where the Bila voda Stream sinks into the karstic caverns, the hydraulic conditions change. The fluvial greywacke pebbles from the initial part of subsurface stream have significantly rodlike shapes (samples 9, 10,11) The increased velocity in the narrow underground channels and siphons lowers the pressure (DREYBRODT, 1988; KRANJC, 1989) Unlike the clasts in the surface channel these clasts are not being forced to the bottom of the channel. Pebbles in these parts of underground channels rotate around their long axes and obtain rodlike shapes. The next reach of the active stream channel is mostly unknown as it is situated 10-20 m below the Amaterska Cave. The Flood Corridor of the Amaterska Cave consists of vast passages active in the Holocene In some of the extremely spacious passages, the hydraulic conditions were similar to those of a surface channel. In such reaches of the underground stream the pebbles of greywackes were rapidly flattened due to the decrease in water velocity and increase in pressure (samples 12, 13 and 14) The last part of the underground flow of the Punkva River is formed by narrow channels and a 400 m long siphon between the Amaterska and Punkevni caves The pebbles become rodlike again due to the previously described mechanism (sample 15) T h e Je d ov nJc e Stre am The roundness of greywacke pebles in the surface part of the Jedovnice stream increases in the downstream direction (Fig. 4-B). The same diagram shows that the roundness of pebbles increases in the cave system. The shape of greywacke pebbles in the surface stream is clearly rodlike. Thinly bedded greywackes intersected by numerous joints are exposed in outcrops along the stream channel. Rodlike clasts get to the channel from the outcrops and no significant reworking occurs among them during the short transport. The first reach of the subsurface stream is formed by high, meandering corridors. The stream flows here with an open space above even during flood events. A gradual flattening of pebbles occurs in the sediments (Fig 4 C, samples 5 6,7,8,9). Narrow corridors and siphons several hundred metres long form the central part of the cave system. A different mode of pebble movement on stream bed occurs due to the above described changed hydraulic conditions of flowing water. As a result, the pebbles acquire a distinctly rodlike character (samples 10,11) Downstream of the siphon s, the underground flow runs through a large corridor in which similarly as in the large corridor of the Amaterska Cave the greywacke pebbles become flat in shape (samples 12, 13 14 ). ma.s L Svitan liver 75 0 surhce ---cnas I 650 550 A 450 ]50 25 0 ] 0,50 B o.i,o 0 30 0.211 0,10 o 6 5 4 ] C 0 -1 2 -] 10 15 I I I 12 141 I I I I 11,J ,,12 I I 1l I I I I I I I I I n I I I I I I 11 I 20 25k11 Figure 4 : A gradient curves of the Jedovn i ce Stream with sampling points B roundness of the pebbles C shape of the pebbles 4. Conclucion The changes in shape and roundness of modem pebbles in the two streams flowing through the cave system of the Moravian Karst persuasively document the dependence of reworking of clasts of fluvial sediments on hydraulic conditions of the environment. These conditions control water flow in the surface and underground channels, which in tum determines the mode of fluvial pebble movement on channel bottom as well as the reworking of the pebbles. Acknowledgments The study of sediments of the Moravian Karst is supported by the Grant Agency of the Czech Republic (grant No 205 / 93 / 0726) and U.S.-Czech Science and Technology Program (grant No 95 051 ). Sympos i um 7: P h ys i cal Spe l eo l og y 1 5


References BULL, P A. 1978. A study of stream gravels from a cave: Agen Allwedd, South Wales Zeitschrift .fiir Geomorfologie, 22, Heft 3: 275-296. BLATT, H.; MIDDLETON, G. & R. MURRAY. 1972 Origin of sedimentary rocks. Prentice-Hall, 634 p DOBKINS, J.E. & R.L. FOI..K. 1970 Shape development on Tahiti-Nui Journal of Sedimentary Petrology, Vol.40, No.4: 1167-1203 DREYBRODT, W 1988: Processes in karst systems Physics chemistry and geology: Springer-Verlag, 288 p KRANJC, A A 1981. Pebble investigation in Slovene caves (Yugoslavia). Proc 8th Int Congress of Speleology: 18-20. KRANJC, A A 1989 : Recent fluvial cave sediments, their origin and role in speleogenesis. SAZU, 27, 155 p. KRUMBEIN, W C. 1941. The effects of abrasion on the size, shape and roundness of rock fragments. Journal of Geology, Vol. XLIX, No. 5: 482-520. Mn.Ls, H. H. 1979. Downstream rounding of pebbles a quantitative review Journal of Sedimentary Petrology, Vol.49, No. I : 295-302. The value of quarry records and monitoring excavation progress in the interpretation of sediment-filled caves in Eldon Hill Quarry, Derbyshire, England C. Allison, P.L. Smart Department of Geography University of Bristol, Bristol BS8 1 SS Background Continued demand for aggregate is resulting in further expansion of limestone quarries in t he U K thus potentially increasing the destruction of caves and karst. A unique scheme for the recording and documentation of cave and karst features intersected by quarrying was established in 1993 at Eldon Hill Quarry and continues to date The Quarry is operated by RMC Roadstone Ltd North West who also provide the sponsorship for the scheme The Quarry is situated in the Peak District National Park Northern England and is included within the Castleton karst earth science Site of Special Scientific Interest. The quarry has intersected a large number of sediment-filled caves, fossil features representing an early stage in the evolution of the Castleton karst. The recording scheme The recording scheme includes three elements: 1) a procedure to alert scientists and cavers to features intersected by quarrying operations to enable inspection and study before their subsequent destruction; 2) the routine photographic recording by quarry staff of all faces prior to blasting ; 3) the archiving of rock quality logs obtained routinely from all blast drill-holes together with the location and elevation of each drill-hole Results The alert procedure has only operated once in a four year period (when an open void was intersected which enabled cavers to enter a 150m long cave system, Alsop's cave), but it has provided a framework allowing free access to sites of potential interest to both ca v ers and scientists, together with some financial and logistic support. It has also increased awareness of the significance of the cave sites among the quarry workforce. A near complete photographic record of the progressively retreating quarry faces (usually at I Orn intervals) has been obtained, although there are some problems with face identification and faces were not photographed prior to buffer blasting. The quality of this record is in some cases inadequate for detailed interpretation, often due to poor weather and light conditions or inclusion of too wide a field of view The use of marked drill holes for scale also proved inadequate. The photographic record enabled the gross morphology and distribution of intersected caves to be determined but it did not enable interpretation of the stratigraphy and nature of the sediment infill, even where faces were especially cleaned Data for more than 2000 drill holes at 4 5m centres were obtained, offering the potential for very high resolution of distribution and extent of caves by use of a 30 visualisation package However, many holes were deliberately deviated from the locations of known sediment filled caves or major joints to reduce drilling and blasting problems There was also a problem in the unambiguous conversion of rock quality terms such as "rough drilling" and "clay and rock" into karstically meaningful categories subcutaneous zone and gravel or boulder-filled cavity in these two cases 16 Proceedings of the 12 th International Congress of Speleology 1997, Switzerland Volume 1


Age and Distribution of some Karstic Sediments from the Polish Jura and their Implications for the Age of the Relief Ireneusz Felisiak University of Mining and Metallurgy Faculty of Geology, Geophysics and Environmental Sciences al Mickiewicz.a 30 30-059 Krak6w Poland Abstract Due to the lack of fossils, both the clay and heavy mineral assemblages were applied to the dating of both the Tertiary and Early Quaternary karst sediments in the Krak6w Upland Mineralogical criteria appeared to be the useful! age determination method Preliminary results compared with the available palaeontological datings allow to revise the age of drainage system in the studied area which appears to be younger than previously proposed The Late Pliocene-Early Eopleistocene was presumably the period of tectonic stabilization and planation whereas during Late Eopleistocene-Mesopleistocene quick uplift took place This event is connected with rejuvenation of faults and with the general uplift of the area north from the Krzeszowice Graben. Introduction Relics of fossils in Tertiary and Early Quaternary karst sediments are relatively rare Therefore, attempts are made to base their stratigraphy upon mineralogical criteria, particularly upon the composition of heavy and clay mineral assemblages Preliminary results obtained for the Polish Jura are encouraging. The Polish Jura (PJ) is a NW-SE-trending belt of outcrops over which Oxfordian and Lower Kimmeridgian limestones are exposed rn 02 l t.>.:::: ::J 3 E-=~J 4 lii!1I 5 6 Figure 1. Simplified map of studied area: 1-9 sampling sites for heavy minerals analyses. Lithostratigraphy: 1 pre-Upper Jurassic, 2 Upper Jurassic, 3 Cretaceous, 4 Miocene of the Carpathian Foredeep, 5 continental Neogene, 6 sampling sues. KG Kn.eswwice Graben. The belt extends from Krak6w through C~stochowa to Dzialoszyn (figure 1) and is the largest karst area in Poland Its southern part is called ''The Krak6w Upland ". It is cut b y the latitudinal Krz.esrowice Graben which marks the northern edge of the Carpathian Foredeep 1n the area of Forcdeep the karstified Oxfordian limestones are partly covered by marine Miocene clays The karst sediments in the studied area can be divided into the four age groups : I. Eocene, 2 Oligocene and early Miocene 3 Pliocene and Early Quaternary (Eopleistocene), 4 Mesoand, particularly, Neopleistocene The absence of Middle and Late Miocene deposits is caused by marine transgression 1n both the first and the second group sediments do not contain fossils and are locally covered by Middle Miocene strata (marine in SE and continental in NW parts of the studied area) The third group is palaeontologically dated but only in the NW part of the area Finally, the fourth group has full palaeontological dating Denudation of Miocene clays cover as well as the initiation of recent river system took place in the Pliocene and Eopleistocene Details of this process were studied in the SE part of the studied area (i.e in the Krak6w Upland) by DiUL YNSKI et a/ ,( 1966) The age limits proposed in their paper are rather hypothetical due to the absence of Pliocene and Early Quaternary fossils and had to be partly corrected (cf MADEYSKA, 1977) Hence applying the mineralogical criteria, the present author aims to ascertain which sediments in the vicinity of Krak6w belong in fact to Pliocene and Eopleistocene Although the study is based upon observations made in the NW part of the PJ, the uniform geological structure of pre-Tertiary basement allows to extend the conclusions to the SW part of the investigated area Summary of dated sediments Eocene The oldest sediments in the studied area are so called "moulding sands", i e red, clayey sands which crop out in the central part of the P J 1n the vicinity of CZl,stochowa these deposits fill numerous karstic depressions of dian1eters up to 1 km and depth over 45 m 1n some places the sands are covered with continental Miocene strata As the prevailing clay mineral of the moulding sands is kaolinite the sediments are regarded as products of warm and humid climat (GRADZ!NSKJ 1977) The heavy fraction is dominated by most resistive minerals : zircon, tourmaline and disthene, accompanied by rutile and staurolite Garnet and epidote are rare or absent. The heavy mineral assemblage originates from the Cretaceous source rocks (KRYSOWSKA-lWASZKIEWICZ 1974) Oiigocene early Miocene The second group of sediments includes green clays, rarely red and green clays with white sands and Oxfordian cherts (so called ''Rudawa Beds") occurring only in the southern part of the studied area (Krak6w Upland) The heavy mineral assemblage is identical with that of the first group but smectitic character of clay fraction implies more arid and somewhat cooler Oligocene clirnat (GRADZINSKI, 1962) Data from drill cores point out that deposition or, at least, redeposition of these clays proceeded up to Middle Miocene transgression (FELISIAK 1992) Sympos i um 7: Physical Speleology 17


Pliocene and Eopleistocene Sediments of the third group are red and brown clays and r e sidual cla y s with mammalian fauna which represents MN 15 Q2 time span (U pper Ruscinian-Upper Bibarian) (NADA C HOWSKJ et al ., 1 99 1 ; G LAZE K e r a l., 1994 ) 1n the oldest strata (MNl5 Raciszyn-2 La y er No 4 ; MNl6 Rc;bielice Kr6lewskie-l ) the hea vy fraction is dominated b y most resistive minerals : zircon rutile and disthene accompanied b y tourmaline and staurolite ( figure 2 ; samples 1 2 ). Garnet contents do not exceed 0 3% but increases in younger strata classified as MN 17 (from 5% in Przymilowice-3 Layer No 4 up to 36% in Kielniki-3B) (figure 2 ; samples 3 /1 3/2 4 1) 1n the Biharian (QI+Q2) sediments perecentage of garnets varies from 25 to 45% (Kielniki-1 and -3A Przymilowice-2 Layer No. 5) whereas disthene, staurolite and zircon are accompanying minerals (figure 2 ; samples 3/3 4/2,4/3) Unique component of these sediments is biotite found in clays from the Layer No 3 Przymilowice-2 beneath the silty clays of the La y er No 4 with the MN 17 zone fossils) Probably this mineral originates from redeposition of Miocene marine clays which contain both garnets and biotite The distinct increase of garnets content in continental sediments from the break of Pliocene and Pleistocene was observed also in the succession which fills the neighbouring Kleszczow Graben Strata dated as Reuverian C (corresponding to the MN I 5) reveal 0l 2% of garnets whereas in the overlying 4kinsko Formation their contents vary from 18 to 25% ( KRzYS Z KOWSKJ & SZUCHNI.K 1995) Red colours of sediments dated as Pliocene and Pleistocene alwa y s have a brownish tint distinctly darker than the red colour of the moulding sands Independently on colour the clay fraction is dominated by smectites accompanied by kaolinite Illite and mixed-layer kaolinite/smectite are present in trace amounts 1n the palaeontologically dated strata kaolinite predominates only in the Rc;bielice Krolewskie-1 site and the fossil assemblage from this site only points out to the more warmer and more humid climat in comparison with the recent conditions (NADACHOWSKJ unpub report 1995 ). Kaolioite also dominates in the brownish-red, fossil-free deposits from Dzibice which Pliocene age is inferred by position in stratigraphic sequence (BEDNAREK & LISZKOWSKI 1982) and confirmed by exceptionally low garnet concentration (below 0 6% Dzibice-3 and -6 ; figure 2, samples 5/1 512) It seems that red, kaolinitic cla y s from the discussed group were derived mostly from the washing out of Eocene moulding sands-type sediments because available data usually indicate warm but not very humid, mediterranean-type palaeo-climate (GLAZEK & S Z YNKIEWIC Z 1987 ; G LAZ EK et al. 1994) Mesoand Neopleistocene Sediments of this group show Largest extent in the studied area Lithologies include yellow and brown clayey sands with common Oxfordian cherts brown cla y s and residual (weathering) clays determined as terra fusca (BEDNAREK & LISZKOWSKJ 1982) and redeposited Loesses with Limestone debris Numerous cave sites are well-dated palaeontologically in the whole PJ area (e g MADEYSKA, I 981 ) Lithologically equivalent deposits cover the limestone outcrops and older karst sediments The sands originated from the erosion of fluvioglacial sediments which were the source of large volumes of garnets and non-resistive minerals amphiboles pyro x enes, epidote andalusite silimanite) ( KRYSOWSKA-lWAS Z KIEWICZ 1974) In the clay minerals assemblage smectites prevail over kaolinite whereas illite and mixed-layer smectite/kaolinite are trace components Chlorites which indicate the contribution from moraine deposits (op cit ) were found in X-ray patterns of clayey sands filling the sink-hole in Miocene gypsums in Krakow (Wola Duchacka, 58 ; figure 2 sample 9 ). Trace amounts of micas and feldspars were also found in this sample and in weathering clays from the Grzybowska Valley ( figure 2, sample 8) Although chlorite is absent in the latter site, such mineralogical composition can also be regarded as indicati v e of contribution from moraine cla y s w hich were ob s erved upslope over the studied outcrop w z w u o IUl w ...J ll. UJ z UJ g C:3 a. w z w u 0 IUl w ...J ll. := Vl LOCALITY Vl UJ Vl Z OQ U.. N GRZYBOWSKA VALLEY, AK 38 WOLA DUCHACKA 58 Sf\SPOW S1 JERZMANOWICE 23 JERZMANOWICE 23, depth 1.7m DZIBICE 6 DZIBICE 3 a2 KIELNIKI KIELNIKI 3a 01 PRZYMltOWICE 2 la er 5 KIELNIKI 3b MN 17 PRZYMltOWICE 3 layer 4 PRZYMlt.OWICE 2 la er 3 HEAVY MINERALS [%] 0 so D ZIRCON, !.!;-.!,! RUTILE. [Ilil) DISTHENE. Vl UJ ...J a. :l: 100 < Vl 5/2 5/1 4/1 3/2 3/1 2 s TOURMALINE, :: \ > I STAUROLITE, GARNETS. iii] BIOTITE, TITANITE. llIIIIlIIl AMPHIBOLE -EPIDOTE, D OTHER MINERALS Figure 2 Heavy mineral assemblages. Fossil zones marked exclusively for sediments with mammalian fauna. First character in sample number corresponds to sampling site in figure 1. Oldest Post-Miocene Sediments of the Krak6w Upland a discussion In the Krakow Upland only one locali ty has been noticed of brown, weathering clays with Earl y Pleistocene fauna the Low e r Wierzchowska (Mammoth) Cave (MAD E Y S KA 1981 ) Moreover the Pliocene age has been attributed (but not confirmed even b y th e heavy fraction analyses) to the brown and gr ey, arenaceous cla ys with cherts known from numerous other sites Red fossil-free clays were reported from the three caves located in the uppermos t, northern part of the Upland (Tune) Wielki Cave Ciasna Cave and Nad Jaskinic1. Niedos~p!lll Shelter MAD E YSKA 1977 ) The fourth site of red clays (and second with Eopleistocene fauna) was found close to the three mentioned above in the Wierzchowska Goma Cave. Apart from light-red silts the y ounger. brown clay gritstone was encountered with clasts of dark-red cla y s According to (NADACHOWSKJ unpub report 1985 ) the foss i l Gl is 18 P r oceed i ngs of t he 12 th I nt e rnati onal C ong r ess o f Spe leology 1 9 97 S witz e rla nd V olume 1


sackdilligensis (Heller 1930) point to the Biharian Red silts and clays can be older e v en Pliocene Analysis of heavy minerals is expected to provide solution The clays contain kaolinite and mixed-layer kaolinite / smectite in equal proportions whereas in sample from the Ciasna Cave smectite dominates over kaolinite Relatively old (Eopleistocene ? ) can also be the clays over 11 m thick which fill the dellen of the S7.klarka Valley in Jernnanowice23 The clays are brown and contain softened fragments of red clays 1n mineralogical composition smectites prevail over kaolinite similarly to the overlying brown, clayey sands with abundant Jurassic cherts The latter deposits undoubtly belong to Mesoor Neopleistocene and contain much more garnets and epidotes (34.5 and 7.5%, respectively versus 27 and 4% respectively in the clays ; figure 2, samples 6/1,6/2) Unexpected results were obtained for the clays from the adjacent S

Mi~ne transgression. It can also explain the probable initial age of discussed above clays from Sll5p6w which fill the dellen of the S11_sp6wka Valley. Summary The drainage system of the Krak6w Upland was initiated in Oligocene-early Miocene, whereas river valleys north from the Krzeszowice Graben have been significantly deepened after the Miocene (Felisiak, 1992, 1993) Preliminary results of the studies of oldest post-Miocene sediments in the Krak6w Upland do not support the presence of brown residual clays which would correspond to sediments from Raciszyn dated as MN 15 zone According to MAoEYSICA ( 1977), the age of removal of Miocene clays covering the valleys and deepening of the valleys is generally younger than that inferred by D2UL YNSKI et al ( 1966) Precisely : the c-terrace was fonned presumably in the former Late Pliocene (MN 16 and MN 17 zones, Early Eopleistocene according to the recent stratigraphy) Its broad, trough-like shape suggests the tectonic stabilization and planation of the relief, both the dand e-levels which are narrow ledges on the steep slopes of river valleys were developed in Pleistocene and the formation of e-level took place in the Eemian (MADEYSKA. 1977) Hence, it can be concluded that Late Eopleistocene and Mesopleistocene were the periods of quick uplift of the Krakow Upland and simultaneous deepening of the valleys This period roughly corresponds to the age of sediments in the lower, tectonically deformed, Quaternary structural stage of the Graben (KRzYSZKOWSKl, 1995) clays with cherts from Pasternik, still classified as Pliocene, are located much beneath the c-terrace (about 274 m as I.) (c.j D2ULYNSKI et al ., 1%6) If their Pliocene age is demonstrated, it would prove that the mentioned above uplift of the Krak6w Upland was connected with the rejuvenation of faults and with the uplift of the area north from the Krzeszowice Graben. Acknowledgements Sincere thanks are due to Prof. Adam Nadachowski (Polish Academy of Sciences, Krak6w) for kind assistance during field work and for identification of fossils, and to Prof. Teresa Madeyska (Polish Academy of Sciences, Warsaw) for providing samples. Help is appreciated also from Mr Adam Gawel and Dr Jerzy Czemy (University of Mining and Metallurgy, Krak6w) for identification of clays and heavy minerals Financial support of the research was provided by the University of Mining and Metallurgy Grants No 10.140.338 and 10 140.543 References BEDNAREK, J. & LISZKOWSKI J. 1982 The section of fossil weathering covers at Dzibice (ln Polish, English sum ) Biuletyn Geo/ogiczny UW 26 : 89-106 D2.UL YNSKI, S .; IiENKIEL, A. ; KLIMEK, K.; POKORNY, J. 1966 The development of valleys in the Southern Part of the Cracow Upland (ln Polish, English sum ) Rocz Pol. Tow Geo/ 36 : 329-343 FELISIAK, I. 1992 Oligocene-early Miocene karst deposits and their importance for recognition of the development of tectonics and relief in the Carpathian Foreland, Krak6w region, Southern Poland (In Polish, English sum ) Ann Soc. Geo/ Po/on 62 : I 73207 FEUSIAK, I. 1993 Application of paleokarst studies in dating of tectonic events and morphology formation in an intensively faulted area (an example from the Carpathian Foredeep, Southern Poland) Bull de la Soc geog de Liege, 29 : 129-133 GLAZEK J. & SZYNKIEWICZ, A. 1987 Stratigraphy of the Late Tertiary and Early Quaternary Karst Deposits in Poland and their paleogeographic implications (In Polish, English sum ). In : Problemy mlodszego neogenu i eoplejstocenu w Po/see p 113130 Ossolineum, Wroclaw GLAZEK, J ; NADACHOWSKI, A. ; SZYNKJEWICZ, A. 1994 Karst localities with Neogene and Quaternary Mammals on the Cracow C~tochowa Upland Neogene and Quaternary Mammals of the Pa/aearctic Post-Conference 'Excursion Guide-Book 22 May I 994 Krak6w GRADZINSKI, R. 1962 Origin and development of subterranean Karst in the Southern Part of the Cracow Upland (In Polish, English sum ) Rocz Pol Tow Geo/ 32 : 429-492 GRADznilSKI, R. I 977 Sedimentation of "Moulding Sands" on karstified limestones in the middle part of Krak6w-Wieluit Upland (In Polish, English sum ) Zesz Nauk US 183 Kras i spe/eologia l(X) : 59-70 Katowice KRYSOWSKA-IWASZKIEWICZ, M.1966 Mineralogical and petrographical study of Cenozoic continental deposits of the Cracowian Upland. Prace Mineral PAN 35, 69pp KRYSOWSKA-lWASZKJEWICZ M 1974 Heavy mineral assemblages in Miocene formation of the Silesian and Cracowian Regions Prace Geo/. PAN 36, 71 pp KRzYSZKOWSKI, D 1995 An outline of the Pleistocene stratigraphy of the Klesz.crow Graben, Belchat6w Outcrop Central Poland Quat & Rev 14 : 61-83 KRzYSZKOWSKl & SZUCHNIK 1995 Pliocene-Pleistocene boundary in the Klesz.crow Graben at Belchat6w, central Poland Jour of Quat & 10 : 45-58 MADEYSKA. T. 1977 The age differentiation of caves and their sediments of the SllSpoWska Valley (In Polish, English sum ) Zesz Hauk. US 183 Kras i spe/eo/ogia l(X) : 71-80 Katowice. MADEYSKA. T. 1981. Le milieu nature! de l'homme du Paleolithique Moyen et Superieur en Pologne a la lumiere des recherches geologiques (In Polish, French resume) Studia Geo/ Pol 69, 125pp. NADACHOWSKI, A 1985. [Palaeontological opinion on bone relics sampled in the Wiem:howska G6ma Cave]. Unpub report Archiv Univ Min. Metall., Krak6w, (in Polish) NADACHOWSKI, A. 1995 [Conclusions on palaeoclimate and palaeoenvironment of the selected Pliocene and Pleistocene sites in the Krak6w-C~tochowa Jura based on gastropods and vertebrates faunas]. Unpub report, Archiv Univ Min Metal!. Krak6w, (in Polish) . NADACHOWSKJ, A. ; WOLSAN M .; GOOAWA, J. 1991. New localities of Late Cenozoic faunas from Przymilowice in the Craco-Wieluit Upland, Poland Acta zoo/ cracov 34 : 425-435 RUTKOWSKI, J. 1996. On the polygenic character of the geomorphology of Cracow Region (S Poland) Acra Geogr Lodziensia, 71 : 207-217 20 Proceed ing s of the 12 th International Congress of Speleology, 1997 Switzerland Vo lume 1


Sedimentology clay mineralogy, and geochemistry of cave sediment from Hard Baker Cave, Rockcastle County, Kentucky, USA Mark P. S. Krekeler', Annette S u mmers E ng e l ', Scott Enge12, Da vi d Mixon 1, an d Michael Rags d a l e 1 University of Cincinnati, Department of Geology, Cincinnati, Ohio 45221-0013 USA 2 Dames & Moore, Inc., 644 Linn Street, Cincinnati, Ohio 45203 USA A b s tract An estimated I 0 12 m of elastic cave sediment is preserved in Hard Baker Cave, Rockcastle County Kentucky. The uppermost portion of the sediment is comprised ofa thick finely-laminated deposit, ranging from 30 cm to2+ m. A section of this deposit was studied to determine if there were any changes in texture mineralogy and chemistry of the unit, as well as to interpret what the controls of these variables were Whole rock chemistry as determined by X-ray fluorescence, shows no systematic enrichment or depletion of mobile elements. Chemical analyses of these sediments give a minimum chemica l alteration index of86-87, indicating that the sediment is highly evolved and has been through extensive chemical weathering prior to deposition in Hard Baker Cave Therefore the sediment in this cave can be used to investigate past weathering and soil development of the paleosurface before it was washed into the cave system Introduction Cave sediment has been studied extensively from cave systems throughout the world Basic reasoning behind sediment studies have been to determine I) how t h e sediment accumulated in the cave 2) from where the material came, and 3) what the sediment may reveal about the past. The present paper is concerned with these three points addressing them using sediment stratigraphy clay mineralogy and sediment geochemisty. These data will test for past weathering conditions and indications of landscape evolution as they may have influenced the deposition of large quantities of sediment in this karst area Hard Baker Cave (figure 1 ; total length survey: 442 5 m) occurs in the Crooked Creek Drainage Basin which occupie s about 8 % of Rockcastle County (O DELL, 1992). This area lies near the edge of the Cumber l and Plateau where ridges (up to 150 m) are separated by narrow valleys. Streams dissect through Lower to Middle Pennsylvanian sandstone and shales and soluble Middle to Upper Mississippian limestones of the Newman Formation Most of Hard Baker Cave is characterized by large trunk passages, ranging from 1 5 2 0 m in height and width from 2 I 0 m in width. The cave is main l y developed in the Ste. Genevieve Limestone Member of the Newman Formation The entrance pit (at 305 m above sea leve l ) is approximately 3 m deep ; at the bottom of this slope is the deepest point in the cave ( at 291 m above sea level refer to figure I) Sediment accumulations in the main trunk of Hard Baker cave are believed to be as much as I 0 m in thickness Investigations in Great Saltpetre Cave, which is located across the Crooked Creek valley at approximately the same elevation, suggest that sediment was deposited under similar flow regimes and that the materials in Great Saltpetre could correlate to those found in Hard Baker cave (KREKELER et al. in press) Materials and Methods A well-exposed section of the upper-most deposit in Hard Baker Cave was described and sampled (figure 2) Samples of cave sediment were collected whereby approximately I 00g of sediment was placed in 800 mL beakers and suspended in distilled water. The <2.0 m fraction from these beakers was collected and mounted on petrographic slides using the smear method of GIBBS (1968). Samples were X-rayed using a Siemens D 500 Diffractometer with Cu Ka radiation (XRD) with air-dried, ethylene glycol saturation, and heating at 375 C and 550 C treatments. Expandable clay minerals were modeled using the computer program NEWMOD (REYNOLDS, 1985) Whole rock chemistry was analyzed using a Rigaku Wavelength Dispersive X-ray Fluorescence Spectrometer (XRF) for major elements. Sediment Characteristics Se dim e ntolo gy Overall, the elastic sediment in Hard Baker cave can be grouped into two categories : material deposited during high flow regimes and sediments deposited under low flow regimes Materials capping the top of the sequence throughout the cave are typically silts and clays, whereas sediments at depth are characteristically coarser, containing sand and abundant gravel. Generally, cross-bedding in these deposits indicates paleoflow directions from north to south The top sediment (0-160 cm) of the deposit is characterized by finely laminated silts and clays (figure 2) Sediment consists of approximately 85% to 95% clay and 5% to 15% silt Silt grains are s u bangular to rounded Measurements of light and dark banding couplets in d icate that the bands are thicker towards the bottom Dark ban d s contain more organic material than the lighter ban d s No post-depositional burrowing or mechanical reworking are observe d Mudcracks are developed in only the top I 0 15 cm of the se qu ence, ranging from 8-38 cm in width Unidentifie d bones were found near the b ase of the sequence (M PORTER, pers. comm.) A grada t ional co n tact separates the laminated material from the lower unit. The sediment base is c h aracterized by grains and pebbles of varying litho l ogy (mostly quartz sand and quartzite pebble). Symposium 7: Physical Speleology 21


HARD BAKER CAVE ROCKCASTLE COUNTY, KENTUCKY @lnc:llcttH celling heigh! kl METERS M k\dlelt u el1v,tlofl ffl METERS bo SH Lt't'el. METERS 0 25 50 I I I r I r 1 r I '1 1 0 50 100 150 FEET N w E s Study Site Fig11re I. Ke11tucky. Map of Hard Baker Cave, Rockcastle Co1111ty, Map co11rtesy of the Greater Ci11ci1111ati Grotto, Ci11ci1111ati, Ohio. The mineralogy of the silt fraction is composed dominantly of quartz and K-feldspar, with trace amounts of plagioclase, pyroxene, amphibole, altered biotite, and zircon. Quartz:feldspar ratios range from 1.5 to 3.3 and K-feldspar:plagioclase ratios are on the order of 30: 1. Lithic fragments were observed occasionally and were identified as sandstone or chert. Clay Mineralogy Clay minerals in the sediment include kaolinite, illite (with minor interstratified smectite ), vermiculite, lepidocrocite, goethite, RO illite/smectite (J/S), and berthierine. Kaolinite is identified on the basis ofa 7.14 A (001) peak and a 3.56 A (002) peak in air-dried sample. These peaks collapse upon heating at 550C, confirming this mineral as kaolinite. Illite is identified on the basis of the -10 A peak which persisted through heating. Ethylene glycol treatment indicates that there is some smectite in the illite mineral structure. Using NEWMOD (REYNOLDS, 1985), the mineral is modeled as an R3 structure with 93% illite layers and 7% smectite. Vermiculite is identified based on the presence of a ~14A peak which expands to -14.5 A upon ethylene glycol treatment. After beating to 375 C, the 14 A peak collapses to a 12 to ~ 10 A peak, verifying that the mineral as vermiculite. Lepidocrocite and goethite are identified by the presence of 6.22 A and 4.17 A 20 40 6 0 (j) .. 8 0 IP .. IP E ;; 1 0 0 C: IP u 1 2 0 Mud cracks 1 4 0 B Laminated Clay 1 6 0 Darker Clay Band ( organic-rich) 1 8 0 m Sandy aay Fig11re 2. Stratigraphic col11m11 of cave sedime11t illvestigated for this paper. Total thick11ess is 180 cm. peaks, respectively. The peaks of these minerals collapse upon heating to 375 C. RO J/S is identified on the basis of a broad, poorly-defined peak at 16 .8 3 A. Traces of berthierine are identified by a -7 A peak which does not collapse upon heating to 550C. Clay mineralogy exhibits some variation with depth (figure 3). Throughout the section, as the amount of. Kaolinite is the 0 E 20 40 C g 60 ;;; 0 a.. 80 (.) :c 100 a. 0, 120 -~ ci5 140 160 0 10 20 30 40 50 60 Percentage of Clay Fig11re 3. Variatio11 of clay mi11eral perce11tages thro11gh /ami11ated clay 1111it. 22 Proceedings of the 12 m International Congress of Speleology 1997, Switzerland-Volume 1


most abundant clay in the sequence, varying from 44-51 %. Goethite and lepidocrocite show no systematic variation. Traces of berthierine and RO VS are detected in some samples throughout the section. Geoc h e m is tr y of Se dim e nt There is very little chemical variation within the sediment samples (table I) No systematic increase or decrease of elemental abundance occurs. Of the elements analyzed, with the exception of SiO,, none vary more than two per cent. Values for CaO, Na,O, ~O, with respect to the Al,O /TiO, ratio are constant. Additionally, the nearly constant Al O / TiO ratio values between 18 and 20 are indicative of a single source region for the sediment. We use the Chemica l Index of Alteration (CIA) (NESBITT & YOUNG, 1982; McLENNAN, 1993) as an estimate of chemical weathering intensity Typically, CIA values of world-wide surficial sediments range from essentially unweathered values on the order of 45 55 to completely weathered values near I 00 (McLE AN, 1993). The CIA is defined in molecular proportions as: CIA= [Al,0 / ( Al,0 + Cao +Na,O + ~0)] X 100 Cao is non-carbonate. Sediments from Hard Cave are carbonate free and permit the application of this equation directly to our chemical data. Remobilization of elements in the spelean environment has been demonstrated to be negligible in sediments for this area (KREKELER et al in press) Therefore, using the molecular proportions from each sample, we obtained CIA values of 86-87 which indicate extensive weathering. Interpretation of Sediment Pa l eo-co ndi t i o n s to crea t e se di ment c h emistry The deposit we studied is sedimentologically, mineralogically, and chemically uniform. Specifically the -2 m thick laminated clay sequence shows no significant changes in deposition, with the exception of the upper I 0-15 cm where dessication is evident. The uniformity of the Al,O / TiO ratios indicates that there was a single sediment source for the l aminated deposit in Hard Cave. This ratio implies that any changes in mobile elements within the cave sediment would be indicative of changes in weathering intensity and climate on the surface However, because no such changes are observed in the cave deposit, weathering conditions on the surface must have been relatively stable during the depositional sequence. S p e l eogenic i n fl u ences o n se d i m e n t d e p os i tion There are no apparent indications that the sediment was brought into the cave by any mechanism other than tluviokarstic drainage Occasional cut-and-till features in Hard cave sediment suggest that some of the material was deposited and reworked in open channels. Additionally, cross-bedding in stratigraphically lower (and older) sands and gravels indicates that paleoflow was from the north to the south. The sediment sequence investigated for this study is located at a passage junction where water and sediment would have been diverted westward, as delineated by elevations of modem passages The nature of the laminated clay sequence supports that flow regimes were much decreased (to the point of ponding) during deposition at this locatity and perhaps upstream and downstream, as well. The lack of lithic carbonate material in the sediment can be explained in two ways (KREKELER et al., in press). First the carbonate clasts could have been deposited and dissolved post depositionally. However, there are no indications that carbonate material has been leached from the sediment sequence Second it is possible that the drainage basin had not down-cut into the carbonate strata at the time of sediment transport and deposition in the cave system. The proposed scenario suggests that drainage at the time was not actively eroding carbonate material and thereby not contributing to the entrained sediment load of the streams (ENGEL & ENGEL, unpublished data). The present surface drainage, downcutting through the thick carbonate rock units bas developed baselevel caves that having completely different passage morphology than Hard Cave and similar caves at higher valley elevations (O'DELL 1992 ; E GEL, unpublished data). An extensive survey of the cave levels in the Mammoth Cave region, Kentucky, was conducted by PALMER ( 1987) All major passages above 180 m in the Mammoth cave area have thick sequences of elastic cave sediment. These sediment packages often till the passages to the ceilings and may be as thick as 25 m. These sequences are sands and gravels which are capped by silts and clays However, the entrance of Hard Baker Cave occurs at 310 m above sea level, while most of the passages range between 300 and 291 m. PALMER (1987) states that the high cave levels of Mammoth show various stages of passage development and sediment tilling, sometimes indicating slow fluvial entrenchment alternating with aggradation Nevertheless, it is difficult to draw correlations between the two cave systems at present due to limited data and differences in local controls (i e stratigraphy, structure, etc.). Presently the timing of sediment deposition is not well established for Hard Cave Although paleomagnetism data have been collected, results are not yet interpreted (ENGEL, unpublished data) Furthermore, large and frequent speleothems are being investigated for age-dates However, the sediment in Hard Baker Cave could be relatively old, predating continental glaciation for this region of the United States as suggested by PALMER (1987) SASOWSKY & WHITE (1992), and SASOWSKY et al. (1995). Significa n ce of cl a y m i n e ral ogy for pa leoc lim atic inferences The sediment in Hard cave suggests that there was extensive chemical weathering occurring on the surface prior to transport and deposition in the cave, as supported by CIA values of 86-87 and clay mineralogy. The extremely evolved nature of the sediment chemistry and clay mineralogy implies that the area presently occupied by the Crooked Creek drainage basin was in a humid, temperate climate and a topographically well-drained region The suite of clays are derived from pedogenic activty on sedimentary rocks, such as sandstone and shales, and are not the residue of carbonate dissolution. Goethite and lepidocrocite are indicative of reductomorphic soils in temperate to tropical zones (DIXON & WEED, 1989) However, the meaning or the lepidocrocite and goethite ratio is not well understood. Hematite is also a common Fe-oxide in soils ; however, the absence of hematite in the cave sediment suggests that goetbite is favored pedogenically due to low temperatures, high water activity, and high organic matter content (SCHWERTMANN 1971 ) The presence ofkaolinte implies that the landscape was older and mature, where the clay is inherited from primary Pennsylvanian-age shales (DIXON & WEED 1989 ; KREKELER et al., in press). GLENN ( 1960) suggests that kaolinte may be found in young glacially-derived soils where Symposium 7 : Physical Speleology 23


loess is the widespread surface deposit. However, in this area of Kentucky, there are no significant loess deposits on the surface and only traces of glacial material in post-glacial drainages. This raises significant implications as to the timing of surface weathering and sediment flux into the cave system. Acknowledgments We thank Bill Simpson, Bill Carr, and John Wisher of the Great Saltpetre Committee (Greater Cincinnati Grotto) for field support and access to the cave. Thanks also go to members and the executive committee of the Greater Cincinnati Grotto, of the National Speleological Society for financial and field support. Special thanks go to the Wittenberg University Speleological Society for the use of caving equipment. We especially thank P Mickler R Payn and M Porter for long hours of help in the cave, as well as D. Helfin for extensive time in the lab. References DIXON,J.B & S B WEED eds. 1989 : Minerals in Soil Environments. S S A Book Series: I Soil Sci Soc. Amer. Madison, l 244p. GIBBS R.J 1968 Clay mineral mounting techniques for ray diffraction analysis. J. of Sed. Petrology. 38: 242-244 GLE R C. 1960. Chemical weathering of layer silicate minerals in loess-derived Loring silt loam of Mississippi. Trans Int. Cong Soil Sci., 7th, 1960 (Madison, Wl) IV: 523-531. KREKELER, M P S.; SUMMERS ENGEL, A ., ENGEL, S MIXO D. & M RAGSDALE (in press) Mineralogy of cave sediment from Great Saltpetre Cave, Rockcastle County, Kentucky: detrital and authigenic minerals Clays and Cla y Minerals McLENNAN, S M. 1993 Weathering and Global Denudation J. a/Geology IOI : 295-303 NESBITT, H.W & G M YOUNG. 1982. Early Proterozoic climates and plate motions inferred from major element chemistry of lutites Nature. 299: 715717. O'DELL, G A. 1992 Field Trip : Hydrology of the Crooked Creek region. Karst-O-Rama Guidebook. Greater Cincinnati Grotto Cincinnati : 12-17. PALMER, A N 1987. Cave levels and their interpretation NSS Bull 49: 50-66 . REYNOLDS, R C (Jr.) 1985. NEWMOD : a computer program for the calculation of one-dimensional diffraction patterns of mixed layered clays SASOWSKY I & W.WHITE. 1992 The role of stress release fractures in the development of cavernous porosity in carbonate aquifers Water R es. R es. 30 : 3523-3530 SASOWSKY, I.; WHITE, W. & V. SCHMIDT. 1995 Determination of stream incision rate in the Appalachian Plateaus using cave-sediment magnetostratigraphy Geology. 23 (5):415418 SCHWERTMANN, U 1971. Transformation ofhematite to goethite in soils Nature (London) 232: 624-625 Table J. Chemical analyses as determined by XRF. Samples are delineated by depth in centimetersfrom the top of the described section. Chemical Index of Alteration (CIA) values are given below each sample respectively. 15 22.5 33 42 5 52 5 63 72.5 87.5 102.5 117 5 143.5 152.5 SiO 2 59 54 59.68 61.15 59.56 61.57 59.20 58.85 57 68 62.64 59 26 63.23 65.74 TiO 2 0.96 0 98 0.97 0 95 0 96 0 95 0.95 0.96 0.94 0.97 0 95 0.93 Alp 3 19 13 19 04 18.61 19.07 18.50 19 20 19.29 19 50 18.28 19 29 18 17 17.49 FeO(t) 5 15 4.87 4.75 4 97 4 83 5.04 5.09 5.08 4.85 5.12 4.74 4.59 MnO 0 09 0 08 0 08 0 05 0 01 0.14 0.07 0.1 I 0.08 0 06 0.09 0 09 MgO 4 90 4.92 4 88 4.94 4 89 4.95 4.92 4.94 4.87 4.93 4.85 4 .8 1 CaO 0.30 0.29 0.25 0.35 0 27 0.31 0 28 0.46 0.27 0 32 0.22 0 21 Na 2 O 0.13 0.14 0 13 0.14 0 14 0.13 0.12 0.15 0.14 0.14 0.13 0.13 ~o 2.43 2.44 2.40 2.41 2.39 2.46 2.40 2.46 2.36 2.45 2.27 2.23 P2OS 0.12 0.10 0 10 0.13 0.10 0.11 0 11 0 19 0.10 0.11 0.11 0.10 LOI 6 285 6 814 6 519 6.698 6.481 6.676 6.927 7.048 6.336 6 513 6 162 5.807 Total 99.035 99 354 99.839 99.268 100 .24 99.166 99 007 98 578 100.87 99 163 100 92 102.13 CIA 87 86 87 86 86 86 87 86 86 86 87 87 24 Proceed i ngs of the 12 th International Congress of Speleology, 1997, Switzerland Volume 1


A study of cave sediment from Movile Cave, Southern Dobrogea, Romania Annette Summers Engel1, Cristian Lascu2, Adrian Badescu1, Serban Sarbu2, Ira Sasowsky3, and Warren Huff 1 University of Cincinnati, Department of Geology, Cincinnati, Ohio 45221-0013 USA 2 "Emil Racovita" Speleological Institute (G.E.S.S.), Str. Frumoasa 11, Bucharest 12, Romania 'University of Akron, Geology Department, Akron Ohio 44325-4101 USA Abstract Toe study of karst systems has benefited recently by cave sediment investigations, allowing new avenues for studies in speleogenesis, landform evolution, and paleoclimate. While the Movile Cave, located in Southern Dobrogea, Romania, has been the center of attention for the biospeological community with the discovery of the unique chemoautotropically-based ecosystem, Movile Cave has a complex geological evolution, as evidenced through cave sediment studies. This paper presents the cave sediment stratigraphy, mineralogy, and ages in order to determine the history of the cave system. Sediment ofMovile Cave suggests fluviokarstic deposition, during the Brunhes Normal Polarity Chron (younger than 770 kya). Introduction Movile Cave, located several kilometers from the Black Sea coast in Southern Dobrogea, Romania, was discovered by accident in 1986 when an artificial shaft intercepted a cave passage at 18m below the surface (LASCU, 1989; figure 1). Much of the research done in the Movile has been to characterize the complex biotic system the first cave ecosystem that does not derive its energy from photosynthesis (SARBU et al., 1996). Instead, energy is derived from su!fide CTiiS)-laden waters and sulfur-oxidizing chemolithoautotrophic bacteria. Troglomorphic terrestrial and aquatic organisms, thirty-two of forty-seven being previously undescribed, have also been found in the Movile. It has been hypothesized that the cave has had no surface openings or surficial water contact since at least the end of the Late Miocene (5.2 5.5 mya) (LASCU, 1989; SARBU & POPA, 1992, SARBU &KANE, 1995). However, sediment studies and other geologic investigations suggest that the cave has not been isolated since the Late Miocene but has had more recent sediment input (ENGEL, 1996; ENGEL & LASCU, 1996; ENGEL et al.,1996; ENGEL, unpublished). Sediment studies in caves throughout the world have revealed a number of depositional environments that, with spatial and temporal correlation, can lead to reliable interpretations of geomorphic evolution of karst terrains and the reconstruction of climatic history {MILSKE et al. 1983; OSBORNE, 1984) Therefore, the present paper is concerned with defining the speleogenesis of the Movile by using sediment within the cave. All evidence for this study is derived from cave sediment stratigraphy, clay mineralogy, paleontology, and paleomagnetism studies. Geologic Setting Movile Cave occupies a small portion of the eastern flank of the "Obanul de la Movile" (Movile Sinkhole), containing approximately 200 m of explored, dry upper level passages and 40 m of submerged airbells and passages in the lower level (figure 2). Both levels' passage morphologies show that the cave has had an initial phreatic origin (CONSTANTINESCU, 1989). Toe cave is developed in Sarmatian-age (-12.5 mya) oolitic and fossiliferous limestones deposited on the Moesian Platform. In the Mangalia area, there are numerous sinkholes and karstified carbonate outcrops, indicating a severe karstification event in the past, particularly during the Messinian Crisis (5.55.2 mya) and possibly the Wurmian glacial regression (LASCU, 1989; LASCU et al., 1994). During the Quaternary, thick loess accumulations blanketed the region, thereby burying the karst and infilling sinkholes. The lower level of the cave is inundated by thermal (20.9 C) water flowing at 5 1/s (SARBU & KANE, 1995), containing an abundance of su!fide (0.3 mMol/1); ammonium, and methane (SARBU & LASCU, in press). Thermomineral sulfurous springs dot the landscape as the water emerges from Sarmatian limestones through fissures and faults into valleys and the Black Sea. Isotopic comparisons with this water and other aquifers throughout the Mangalia region indicate that the origin of the thermal water in the Movile is from a deep aquifer in Jurassic and Cretaceous units (SARBU & POPA, 1992) UKRAINE The Danube MOVILE CAVE .... ___ _j\ n 1........ BULGARIA \ '--.Mangala ..... ____ Figure 1: General location map for Movile Cave; the cave is located near the town of Mangalia. Symposium 7: Physical Speleology 25


0 5 10 m Main Passage A Stratigraphic Study Siles e Paleomagnetism Siles Access Passage ...... -" ... ,, L~::;~::~;:,,~ Upper Dry Passage Air Bell 4 Depth from the Surface m ::~ 120 Air Bell 2 ,.~----_}. ____ Air Bell 1 ""-=~.::::...-----,~ Lake :~~Z-=~:. ~ ~ ~ -, -~ --~ -~ -_ ; : ., ... ~~: ;; .~ __ _ _ 25 Lower Level Air Bell 3 Lower Submerged Passage Figure 2: Map of Mo vile Cave, adapted from SARBU & POPA (1992). Stratigraphic and paleomagnetism sites are indicated. Sediment and rock characteristics Laboratory analysis of sediment and rock samples Samples were collected from sites throughout the cave (floor and walls) where detailed stratigraphy and paleomagnetism were investigated (figure 2) Clay mineralogy was determined using X-ray diffraction techniques (XRD), starting with powder dispersions in 500 ml beakers Solutions were centrifuged to obtain the <2mm fraction of materials. Smear mounts were made and analyzed using a Siemens D-500 X-ray Diffractometer with Cu Ka radiation For each sample, XRD runs for air-dried, ethylene glycol treatment (to test for expandable clays), and heating to 375 C and 550 C (to determine kaolin group minerals and others that collapse with heat) were obtained. Scanning electron microscopy (SEM) with an electron dispersive system (EDAX or EDS) work was done with numerous samples throughout the cave in order to distinguish depositional characteristics of the material and sediment microfabric. A JEOL JSM-T220 SEM with EDAX was used and samples had to be coated with a severa l angstrom thick layer of gold. An Hitachi S-246N SEM with EDS, infrared chamberscope, and Robinson detector was also used and samples did not need to be coated (operated in environmental mode) Sediment mineralogy and fabric Cave sediment can be described as allochthonous, having been derived from outside the cave system and then transported mechanically into it, or autochthonous, originating from within the cave by internal breakdown/corrosion processes. This study will differentiate the two sediment types. Allochthonous clay-size sediment includes soil clay mineral assemblages similar to loess deposits: chlorite / vermiculite, (001) peak at -14 A (from XRD patterns) ; kaolinite/smectite, (001) peak at -7 A.; RO illite/smectite, (001) glycol treatment peak at 15 9-16.1 A.; R3 illite/smectite, (001) peak at 9.8-10.2 A, and goethite, peak at 4.16 A. These clay minerals do not result from carbonate dissolution. Sediment in larger size fractions (<2mm I 0cm) includes sub-rounded to angular carbonate clasts, clay clasts, sub-rounded to rounded quartz grains and pebbles, and fossil fragments. SEM photographs show that there is no compaction of clay flocks. The coarser sediment in the cave is distributed in gravel bed and bank deposits that mimic passage meandering. Slight clast imbrication and sediment sorting suggest water deposition; flow directions are generally from west to east in the upper level (refer to "Main Passage" on map) Autochthonous sediment consists of very small portions of illite (peak at -10.0 A.), kaolinite (peak at -7 A.), and quartz, with great percentages of dolomite (peak at 2 89 A.) Most of the sediment >2mm are dolomitic ooids. SEM reveals zoned dolomite rhombs (5-20 mm in width) on the outer surfaces of the ooids, showing no signs of dissolution or instability. Some crystal faces have an abundance of adhering, spherical single and chained bacteria. The average diameter of the chained bacteria is I mm with a connecting filament of approximately I mm Bacteria chains, up to 20 mm Jong rest in dents and depression on dolomite crystal faces. Sediment collected from the bottom of the lower level consists predominately of the low-pH clay mineral alunite (distinctive peak at 3.01-2.99 A.) Aunite forms from the 26 Proceedings of the 12 th Internat i onal Congress of Speleology, 1997, Switzerland Volume 1


alteration of pre-existing clay minerals (such as kaolinite and illite) in the presence of sulfuric acid (HILL, 1996) Bedrock mineralogy and fabric Rock samples were obtained from several localities inside the cave, as well as from surface outcrops in the surrounding region. The vast majority of the rocks from the surface consist of quartz, minor clays (<5% total volume), and calcite (peak at 3.04 A). However, some rocks were composed of about 98% high magnesium calcite (peak at 3.02 A). Rock samples collected from the cave shaft are composed of approximately 85% calcite and 10% dolomite (the remaining 5% being indistinguishable clays). Bedrock collected from inside the Movile behind thick clay accumulations, is calcitic. However bedrock sampled throughout the rest of the cave where there are no clay deposits reveal the main mineralogy to be dolomite. Bedrock samples~ 50 cm and 90 cm into the wall of the cave were analyzed by thin sections. Like HOROI (1994), the rock was found to be severely corroded. Where a considerable amount of porosity developed, dolomite filled the vugs. Whole ooids are shown also to be replaced by dolomite Age-dating Movile Cave sediment Previous paleontologlcaJ tnvestlgatlons ENGEL & BADESCU (unpublished data), LASCU (pers. comm.), and STIUCA & JLINCA (1995) found vertebrate fauna containing micromammalian and aquatic taxa from sediment collected at various places throughout the cave. STIUCA & JLINCA (1995) identified the animals to include birch mice,. wood mice, steppe lemmings, bats, frogs, and fish. These animals are characteristic of a dry, open environment in association with species indicative of grassland and marsh. This fauna could not have been derived from in situ weathering of the Sarmatian-age carbonates, as terrestrial animals are not s Horizontal: X Vertical: full scale: 1 h-41 KA/m u D w ASBIU ASE03& ASE040 ASE044 ASE060 N ._cl>llll1.91.ogl99 M/'291Sl.CMRA ~IIA Figure 3 : Zijderveld plots for sediment from Movile Cave. These samples show a normal horizontal component and a positive (downward) inclination, indicating deposition during Brunhes Normal Polarity C/iron (from 770 kya to present). found in the units (GRIGORESCU et al., 1986). Instead, the fossils were correlated to the late phase of the last glacial cycle (WilrmianNistulian), which is approximately 24 to 12 kya., or Late Pleistocene. Paleomagnettsm methodology and results As described by SCHMIDT (I 982), cave sediment can record paleomagnetic reversals in Earth's history. Oriented samples were collected at 10 sites (in the upper level of the cave at varying depths) using 2 cm plastic cubes. Two samples were taken from each locale. These samples were then subjected to a step-wise alternating-field (AF) demagnetization using a large bore ScT cryogenic magnetometer interfaced with a computer to determine characteristic paleomagnetic directions. Plots showing progressive removal of magnetic vectors, made after ZDDERVELD (1967), reveal that most of the sediment showed very little post sampling viscous magnetization. Almost all of the samples display nearly straight-line intensity decay (demagnetization) towards the origin, indicating absence of a secondary component of magnetization. Higher coercivity steps indicate a normal horizontal component and a positive (downward) inclination (figure 3). Therefore, most of the sediment in Movile Cave has been deposited during the Brunhes Normal Polarity Chron prior 770 kya, being constrained by the Matuyama paleomagnetic reversal (IZZETT & OBRADOVICH, 1994) Based on the lack of deviations in demagnetization vectors, it is unlikely that the sediment was deposited during an older normal polarity event. Interpretation of cave sediment There are two main unconformities present in the Movile Cave, one being marked by the contact of the bedrock and sediment (as a primary unconformity) and the other being between allochthonous material and autochthonous sediment of modern processes. Specifically, the sediment/bedrock contact indicates that there is a large gap in time (roughly 12 million years). The sediment that has infilled the cave was deposited in a fluviokarstic drainage system. The elastic components of the sediment do not correlate to the cave host lithology, but instead resemble surface deposits, specifically aeolian sediment in fine size-fraction fabric and mineralogy and channel gravels with coarse foreign lithofragments. The proposal that a free-flowing stream, having originated from the surface, could have occupied the Movile upper level is supported by evidence of slightly imbricated clasts, sediment sorting and grading, and typical channel facies associations. There is additional strong evidence that the Movile Cave (presently sealed) was in communication with the outside environment during certain times. Based on the paleomagnetic data, the cave was not receiving nor accumulating sediment prior to 770 kya, as constrained by the Matuyama-Brunhes geomagnetic boundary. If sediment was brought into the cave prior to this time, the material was either eroded or was not sampled. Furthermore, the vertebrate taxa found in the sediment indicate that there had to be a way for them to enter into the cave, probably by way of the associated sinkhole (STIUCA & JLINCA, 1995). These fossils have been constrained to 12 24 kya, and it is likely that the cave sediment is also from this time span. The upper level of the Movile has been abandoned by free flowing water and the lower level is presently occupied by sulfurous water. It is not known when sulfurous water inundated the lower level of the cave or how long the aquifer has existed. As indicated by the presence of alunite and isotopic Symposium 7 : Physical Speleology 27


investigations done on the water and surrounding rock (SARBU, unpublished data), the process of sulfuric acid speleogenesis is actively enlarging the lower level of the cave. The process is accelerated by the high concentration of sulfide in the water and microbial oxidation of that sulfide to produce sulfuric acid. This speleogenetic mechanism has been associated with the formation of Carlsbad and Lechuguilla Caves in the southwestern United States (EGEMEIER, 1987; HILL, 1996). Alunite or other low pH clays are absent in the upper level of the Movile, indicating that sulfurous water has not been in this portion of the cave. Autochthonous dolomitic clay and associated residue from corroded wall rock in the upper level is attributed to condensation corrosion (ENGEL et al., 1996; ENGEL & LASCU, 1996; ENGEL, unpublished data; SARBU & LASCU, in review). This speleogenetic process is poorly understood, but the presence of corroded walls in the upper passages, the absence of water infiltration from the surface into the cave, temperature difference between the water in the lower level and the walls of the upper level, and the high CO 2 concentrations throughout the Movile all suggest that condensation corrosion is presently affecting the morphology of the upper level of the cave (SARBU & LASCU, in review). It is uncertain as to what is causing the precipitation of dolomite in place of the original calcitic host bedrock, although biogenic process may be influencing the mechanism. Also of importance is that the bedrock covered by thick clay accumulations is not dolomitized and shows no evidence of being corroded. This suggests that the clay was in place before severe corrosion of the bedrock began. Acknowledgments The authors of this paper would like to thank the members of G.E.S.S (Group for Underwater and Speleology Exploration), Bucharest, Romania, for field support and assistance. Appreciation is given to Dr. Ken Bladh (Wittenberg Univesity, Springfield, Ohio) and Dr. J. Lian (University of Cincinnati, Cincinnati, Ohio) for access to and help with scanning electron microscopes. Thanks are extended to S. Engel, M. Menard, R. Popa, M. Porter, and L. Trump for technical support. This research was funded by grants from the National Speleological Society, the Geological Society of America, and the University of Cincinnati. References CONSTANTINESCU, T. 1989. Considerations sur la zone karstique de "La Movile' (Mangalia, Dobrogea de Sud, Roumanie). Misc. Speol. Rom. 1:7-12. EGEMEIER, S. 1987. A theory for the origin of Carlsbad Caverns. NSS Bull. 49: 73-76. ENGEL, A.S 1996. Scanning electron microscope images from fine-grained sediment in Movile Cave, Southern Dobrogea, Romania. NSS Guidebook. Salida, CO, 44. ENGEL, A.S. & C. LASCU. 1996. Clays of Movile Cave, Southern Dobrogea, Romania: speleogenetic indicators of cavern formation. Program and Abstracts of the 33" Annual Meeting of the Clay Minerals Society (Gatlinburg, TN), 53. ENGEL, A.S.; HUFF, W., BADESCU, A., LASCU, C & S. SARBU. 1996. Sediment studies from Movile Cave, South Dobrogea, Romania. Geol. Soc. of America Abstracts with Programs. Annual Meeting (Denver, CO), A-281. GRIGORESCU, D.; CONSTANTINESCU, V. & L. DRAGOMIRESCU. 1986. A paleoecologic analysis of the Bassarabian (Middle Sarmatian) vertebrate association from Credinta and Ciobanita (Southern Dobrogea) based on the taphonomic characters processed through methods of numerical taxonomy. Trav. Du Museum d' Hist Naturella Grigore Antipa" 28: 275-283. HILL, C. 1996: Geology of the Delaware Basin, Guadalupe, Apache, and Glass Mountains, West Texas and New Mexico. Midland: Permian Basin Section SEPM, 478p. HOROI, V. 1994. The corrosion process in ''Pestera de la Movile" Cave (Southern Dobrugja Romania). Theor. Appl. Karstol. 7: 187-191. IZZETT, G.A. & J.D. OBRADOVICH. 1994. 0 Ar/3'Ar age constraints for the Jaramillo Normal Subchron and the Matuyama-Brunhes geomagnetic boundary. J of Geophysical Res B99: 2925-2934. LASCU, C. 1989. Paleogeographical and hydrogeological hypothesis regarding the origin of a peculiar cave fauna. Misc Speol. Rom. 1:13-18. LASCU, C.; POPA, R. & S. SARBU 1994. Le karst de Movile (Dobrogea de Sud). Rev Roum. De Geographie 38: 8594. MILSKE, J.A.; ALEXANDER, E.C. (Jr ) & R. S. LIVELY. 1983 Clastic sediment in Mystery Cave, Southeastern Minnesota. NSS Bull. 45: 55-75. OSBORNE, R.A.L. 1984. Lateral facies changes, unconformities and stratigraphic reversals: their significance for cave sediment stratigraphy. Trans British Cave Res Assoc. 11(3): 175-184. SARBU, S. & C. LASCU. Condensation corrosion in Movile Cave, Romania. NSS Bull (in review). SARBU, S. & R. POPA. 1992 A unique chemoautotrophically based cave ecosystem. In: (CAMACHO, A., ed): The Natural History ofBiospeology. Monografias Museo Nacional de Ciencias Naturales: Madrid, 637-666. SARBU, S. & T. KANE. 1995. A subterranean chemoautotrophically based ecosystem. NSS Bulletin. 57:91-98. SARBU, S.; KANE, T. & B. KINKLE. 1996 A chemoautotrophically based groundwater ecosystem. Science 272: 1953-1955. SCHMIDT, V.A. 1982. Magnetostratigraphy of sediments in Mammoth Cave, Kentucky. Science. 217: 827-829. STIUCA, E. & H. ILINCA. 1995. Pleistocene mammals (Rodentia) from Movile Cave (Romania). Theor. Appl. Karst 8: 157-161. ZIJDERVELD, JD.A. 1967. A.C demagnetization of rocks: Analysis of results. In: (COLLINSON, D.W .; CREER, K.M. & S.K. RUNCORN, eds.): Methods in Paleomagnetism. Elsevier, Amsterdam. 2 8 Proceedings of the 12 th International Congress of Speleology 1997, Switzerland Volume 1


Cave chemolithotrophic soils Vladimir A. Maltsev', Viktor A. Korshunov2, Andrei A. Semikolennykh3, 'VNIIGEOSYSTEM institute. Moscow 121351, Yartsevskaya ul., 15-21, Russia 2 Geology Department Moscow State University Moscow, Tichvinsky pereulok, 9/12-4-86, Russia 3 Soils Department, Moscow State University Moscow, Pervomaiskaya str., 94-133, Russia Abstract Geochemical and microbiological studies in the Cupp-Coutunn cave (Turkmenistan) show, that the fluffy red-colored coverings on the cave walls and roof, 1-20 cm thick, are not a kind of a residual sediment, but a complex organomineral media. This media, named "okber", has a very high biochemical activity. Okber appears to be a result of both strong corrosion and weathering of the rock, and a formation of secondary minerals. It also appears to be a basis of the feeding chain for higher organized flora and fauna. Introduction The biological redox processes are already proved to be a significant karstification factor, and the reason for some specific features of caves mineralogy. The so far studied cases mostly concern phreatic areas in caves where the sulfate reduction proceeds in deep phreas Sometimes full bacterial sulfuric cycle was reported, with sulfatereduction in phreas, and sulfuroxidizing in dry areas [FORTI, 1988; BALL & JONES, 1990]. In spite of this, there are several caves known, that have no phreatic areas, but have active sulfuric processes in them. One of these caves i s the Cupp-Coutunn cave in Turkmenistan, and it was selected as an object of the study. Okher The term "okher" came from the caver's jargon, as a name for red fluffy coverings on the floors and roofs of the cave, and is based on an untranslatable game of wo rd s Up to 19 93 it was taken as a residual of atmospheric corrosion [MAL TSEV & SELF, 1992] but even the first studies shown, that it can't be so [SEMIKOLENNYKH & KORSHUNOV, 1994]. Ok.her vs. corrosion residues If we compare okher to classic corrosion residues (for example, from Lechuguilla cave), we'll easily see two main differences: a) In the corrosion residues we always see unaffected carcass of non-soluble minerals (especially alumosilicates), in which the original texture of the rock is saved recognizable. For the corrosion residues, known from Guadalopean caves, we can even see, that these residues are more likely simply weathered limestone the texture has completely survived, so the volume of these "resid ues is equal to the volume of the affected parent rock [CUNNINGHAM, 1993; M QUEEN, pers comm.]. On the contrary, okher never keeps unaffected carcass of the limestone, because of much more acid conditions. All minerals, even silicates, completely re-crystallize, mostly forming new minerals, which were never found in the parent limestone. For example, fig. I. shows rather large gibbsite crystals, generated in okber. The texture of the rock is lost completely, and the volume is changed greatly. For example, concentration of Fe O in different layers of okher is 5.7%-28.3%, and in the parent limestone it never exceeds 0.5% Even when converting volumes to weights (okber is fluffy, density of dried okher is about 0.20.6), the difference still remains great. The same is seen for silicates. b) The corrosion residues are mineralogically passive Okher, growing on pure carbonate substrate, is almost always covered by gypsum efflorescences, and even coatings. Disturbed okher noticeably smells of sulfuric gases, including H 2 S. In spite of this, there are no visible mechanisms for transportation of sulfur to the okber location along the cave (no water flows, and no sulfuric gases in the air). Figure 1. Gibbsite crystal in okher, 0.1 mm large, diagnosed via X-ray microzand. SEM photo by A.Semikolennykh. These two features already allow to separate residues (passive product) form okber (active strata with complete re-working of everything). Structure and morphology of okher Okber is spread almost all over the cave, and its morphology and structure is variable, depending on the limestone composition, humidity, air circulation, etc . But some generic features and typical structure exist. Okher has no definite boundary with the limestone Appearing as something fluffy and very porous on the surface, it becomes much more dense and sandy in the middle, then consequently transforms into altered limestone, and so on. The okher covering can be from a couple of millirneters thick up to 20cm and even more. Okber is always well structured and has several layers. The main layers, listed below, have specific properties and can be found in almost any kind of okber (fig. 2). a) The outer layer is thin (l-3mrn.), and consists of gypsum sand, rarely forming a crust. This layer is usually present in thin okher (up to 2-3cm) and absent in thicker one. In the last case gypsum sand doesn't form a separate layer, but exist in next Symposium 7 : Physical Speleology 29


layers. The boundary to the next layer is fuzzy rn places of strong seasonal humidity cycles, the gypsum may also be reworked into filamentary crystals [MAL TSEV 1996] b) The second layer is red, extremely fluffy and porous, Imm to I 0cm thick. In thick okher this layer has a color trend from brawny inside to bright red outside. The boundary to the next layer is sharp. pH in this layer is in the range of7-8. Figure 2. Ge11eral view of the okher cross-sectio11. Photo by V. Korshu11ov. c) The third layer is yellow or brawny-yellow colored Usually it's the thickest layer, and can be up to 20cm thick It's much more dense than the previous, sandy, and usually contains pieces of altered limestone. d) The fourth layer is dark gray up to black. This layer is thin and unstable, possibly seasonal (appearing due to overmoistening during corresponding seasons). It has a sharp boundary to the previous layer, and a fuzzy one to the next. e) The last layer consists of strongly altered limestone The limestone here keeps its structure, but looses density. In the outer part of this layer the limestone becomes very soft and strongly smells of sulfuric compounds when being broken Deeper it consequently tran its into normal limestone. pH on the contact of this layer with the previous is about 3.S-4 maybe even lower. The problem is that the spatial distribution of pH is irregular due to irregularity of bacterial communities and all the measurements are to be made in situ otherway the results are strongly distorted The pH data of the okher layers is to be compared to the pH data of the cave water. The meteoric water coming along fracture s, bas pH about 6.S in all the cases The water in the pools has pH 77 .S the same as the water in the springs and in the artesian basin, both fed from the karst. Okher covering is so delicate that sometimes a thick red layer disconnects from the roof and falls, forming thick deposits In these deposits only the top 10 cm look like the original okber and in the deeper layers it changes into some dense, fat and plastic brown clay and consequently into also dense fat and plastic blue clay. Microbiology of okher The key question the source of the H S gas in the okher in fact stays unresolved There are many indirect evidences of SRB (sulfate reducing bacteria) activity in the deep layers of okher, but no direct ones At the studied depth SRB (Desulfotamaculum) were found only in one sample. The source of the sulfur are bituminous inclusions in the limestone. Bacterial activity is the only known mechanism due to which the ulfur is partly returned (see below) Therefore SRB must be the main type of bacteria, utilizing the source sulfur The question mostly is whether the SRB activity proceeds in lime s tone themselve s, or only in the limestone near the cave. The first supposition has two Figure 3. Diatomea i11 the outer layer of okher. Completely re-crystallized surrou11ding shows, that it's 1101 residual. SEM photo by A.Semikolem,ykh. pieces of evidences against it: No traces of the sulfur was found in springs unloading from limestone; okher morphology too weakly correlates with fractures which collect the H S ga s, and this means that it mostly appears near the ca~e and is transported along the pores of the weathered limestone to deeper layers of the okber (porosity of the source limestone is extremely low) So, SRB must be mostly found in deeper layers than it was sampled Some evidences (mostly, studies of micro flora growing Figure 4. Simbios of mycoplasma a11d fu11gi (both u11diag11osed) i11 the outer layer of okher. SEM photo by A.Semikolennykh. on the candles left in caves), show, that utilizing of bituminous inclusions must be due not only to SRB but also to Artltobacter, Rhodococcus, Mycobacterium, found in all such samples On the contrary, another half of the sulfuric cycle is "opened", and proved From the samples taken from outer okher layers using the Vacksman Tausson and Birs media there were exposed sulfuroxidating bacteria particularly Thiobacil/us ferrooxidans, Th. thoioxidans and similar microorganism Their acti v ity is enough for partial return of s ulfur into th~ 30 Proceedings of the 12 th International Congress of Speleology 1997, Sw i tzerland Volume 1


corrosion cycle, and for enriching of the outer okher layers by Fe providing the red coloring. The sulfuroxidizing activity in the outer layer returning the significant part of sulfur into the cyclic sulfuric corrosion of limestone, is an additional indirect evidence of SRB activity somewhere inside because inside the okher pH falls to about 3, and free H S04 exists, reacting with limestone, that consequently results in gypsum generation inside Absence of this gypsum in middle layers of okher shows, that it must be destroyed by SRB The bacterial biomass from the okber is utilized by other organisms, giving a start to a long feeding chain. Immediately in okher several kinds of fungi ( detected Pe11icillu s sp., Aspergillus 11iger Scopulariopsis sp .), several protozoa diatomea (fig.3. 4), etc. were found. In the massifs of fallen okher several kinds of mites and insects cab be found. It's possible also that this chain I oxidation zone II reduction zone Figure 5. The seaso11a/ iro11 a11d su/fur 111igratio11. appears to be the chain, leading up to blind fishes found in the under-plane drainage of the cave system (Kugitangtou blind loach) but some elements of this chain are still absent. Looking at the details of the okher spatial distribution we can find out, that there is some optimum between general high humidity of the cave area, and the oscillations of the humidity, caused by the seasonal cave wind inversions. Okher developing within this optimum is about one order more active, than the okher, developing in other areas of the cave. The diagram on fig 5 illustrates the main relations between activities in the okher and the microclimatic conditions of the location Mineralogy of okher Due to the fact that the okher layers have very different pH and that the seasonal humidity cycle (and microbiology-driven cycles) result in cyclic migration of various ions through the pH barriers, all the matter is reworked This reworking has several variants, but the general scheme is like the following Silicates, present in limestone mostly as quartz and micas are transformed into kaolinite-like minerals in the middle layers of the okher and further to gibbsite and illite in the outer layers Gibbsite and illite co-existance may be explained with high-pH conditions, separating Si and Al. The same effect is known from the other parts of cave, where gibbsite and serpentine co-growth are known. Iron i s concentrated in the outer layer, mostly in the form of microcrystalline oxides hematite, goetite, etc .. Gypsum (fig.6) is generated both in deep layers (metastabil), and outer layers (final). Calcite mostly disappears in the middle layers, and appears again in the outer layers where it forms, together with gypsum, specific "sand". This sand, falling from Figure 6. Okher corrosion. The white stripes are calcite veills, co mpletely turned to gypsum in the outer layer of okher. The image height is about 40 cm. Photo by V.Maltsev. the roofs, sometimes makes up sediments up to severa l metres thick, appearing as the product of the lime stone corrosion by the okher. High acidity in the deep okher layers causes some side effects when the cave intersects veins, or the cave wall is covered by products which survived from the hydrothermal phase. Interactions between sulfuric acid, fluorite druses, silicate matter and ore veins leads to generation of minerals exotic for caves like saukonite, fraipontite, serpentine, etc Okher as a kind of soil So okher is an active strata, providing the corrosion of limestone, gas and water exchange with atmosphere, alteration of the residual material and bioma ss generation, that i s used by other organisms. Due to this okher certainly is to be clas ified not as a sediment, but as a special type of a soil, based on chemilititrophic bacteria Okher appears to be a significant corrosion factor. In some cave passages it provides up to 90 % of the total corrosion. This can be estimated through the total quantity of the Fe and Si in the okher, compared to their contents in the parent limestone in the full volume of the passage. In such localities the okher corrosion forms specific morphologic types of the cave landscape syr, red chinks, red tubes [SEMIKOLENNYKH, 1996]. Okher exists not only in the Cupp-Coutunn cave. There are evidences of it's existance in Snezhshnaya deep cave in Caucasus, and also in some minor caves within ore mine s, but this needs to be studied very carefully In the Guadalopean caves the visual similarity appeared to be a visual convergence of very different phenomenas Related effects in the Cupp-Coutunn cave The microbiological studies in the Cupp-Coutunn cave have not only shown the okhcr's microbiological origin. Some other corrosion and mineral generation factors definitely are also related to the activity of microorganisms For example, it 's the gypsum replacement by silicates in the Vodopadnyi chamber. Sympos ium 7: Physical Spe l eology 31


Figure 7. Gypsum replaceme11t from i11side by a foam-like silicate substa11ce. Photo by V.Maltsev. The gypsum gemmifonns (fig.7) are replaced from inside by some very porous foam-like silicate substance. Fig11re 8. Celestite glob11/es (those, where crystals are see11), silicate glob11/es a11d threads from the foam, replaci11g gyps11m gemmiforms from i11side. SEM photo by A.Semikole1111ykh. This substance, if studied closely, consists of celestite spherolites, amorphous silicate globules, amorphous silicate threads. The first of them may have any origin, but the last two can only be biological. lt can clearly be seen from the SEM photo on the fig. 8. There is no idea, what bacteria it could be. No growth in vitro was received, and no bacteria are known, leaving carcasses of Fe Mg silicate composition, very close to olivine. There are several more similar phenomenas known, but all are completely unstudied. Certainly, microbiology must become one of the main trends in studying the cave systems like Cupp Coutunn. References T.K.BALL & J.C.J0NES, 1990 Speleogenesis in the Limestone Outcrup Worth of the South Wales Caalfield/ / Cave Science, I 7( I) London K I.CUNNINGHAM., D .E.N0RTHUP, A .. M P0LLASTR0, W.J. WRIGHT, E.J.LAR0CK, 1995. Bacteria fungi and biokarst in Lechuguilla Cave, Carlsbad Cave ational Park NM // Environmental Geology, No. 25, p.2-8 P.F0RTI, I 988. The role of sultide-sulfat reaction in speleo genesi s//Firs t FEALC Belo Horizont, Brasil V K0RSHUN0V, A.SEMIK0LENNYKH, I 994 The model of speleogenetic processes connected with bacterial RedOx in sulfuric cycles in caves of Kugitangtou Ridge l /Breakthroughs in Karst Geochemistry and RedOx microbiology, Karst Waters lnst. NY, p.43. V.A.MALTSEV 1993. Minerals of the Cupp-Coutunn cave system, Southeast Turkmenistan // World of Stones, No.2, Moscow, p.5-30 V.A.MALTSEV 1996. Sulfate filamentary crystals and their aggregates: Proc. Univ. Bristol Spelaeol. Soc., 20(3), pp 171185 V.A.MALTSEV & C.A.SELF, 1992 Cupp-Coutunn cave syste m Turkmenia, Central Asia//Proc. Univ. Bristol Speleol.Soc., 19 (2)-29 A.A.SEMIK0LENNYKH, V A MALTSEV, V.V.K0RSHUN0V 1996 The processes of biogenic sulfuric weathering in the caves of the Kugitangtou Ridge Turkmenistan. ln: Mineralogy and Life: Biomineral Interactions. Syktyvkar. p 61. 32 Proceedings of the 12 th International Congress of Speleology 1997, Switzerland Volume 1


Mineral composition of elastic material in fault zones and open fissures in karst rocks, examples from SW Slovenia Nadja Zupan Hajna Karst Research Institute ZRC SAZU Titov trg 2, 6230 Postojna Slovenia Abstract Three different types of elastic material may be distinguished based on its origin The first type is infiltrated material into the open fissures from the surface. Mineral composition in that case depends on composition of rocks from were the weathering remains are originated The second type is elastic material which is filled into opened fissures during floods of the caves, composition of it corresponds to the rocks composition of underground water flow watershed. Sediments could be enriched in calcite or dolomite clasts these are actually small particles of limestone or dolomite from the walls of underground water passages The third type of elastic material is significant for inner part of fault zones This is clay size material, which was formed by tectonic compression of carbonate rocks Tectonic clay consists almost entirely of calcite or dolomite and their admixture depends on mineral composition of parent rocks Introduction Fissures open within the fault zones due to tension and the fissured zones may be washed off or filled up by the deposits of various mechanical sediments, deriving either from the surface or by cave flood material The fault zones are either opened or closed and their type controls whether there are infiltrated loams or tectonic clays. Tectonic clays are formed by compression of the carbonate rocks (PLACER, 1982). Tectonic clay has the same mineral composition as the rock comprising the fault zone there could be only a bit more of minerals which are insoluble residues of the carbonate rocks Mechanical sediments infiltrated into opened fissures may vary in respect to their mineral composition They are the residues of different rocks weathering Their mineral composition sometimes indicates their origin but usually it is very difficult because long lasting weathering disintegrates and shades a lot of primary minerals. Only the most resistant minerals to weathering remain such as quartz and heavy minerals. The less resistant minerals are replaced by secondary minerals like clay minerals and chlorites. Which are formed depends upon physical and chemical conditions in the environment. Sometimes within the same fissure minerals of different origin occur Mixing appears when the water from the surface finds its way along the existing deposit and transports in a new mineral. Figure 1: Position of 1 Divaca, 2 Postojna cave system and 3 Planina cave. Infiltrated material is presented by the sample from the old cave near Divaca samples of flood loams, silt and clay size material are from the same cave From Postojna cave system and Planina cave flood material was also analysed as was tectonic clay from Postojna cave system (Figure 1) All the samples were analysed by x ray diffraction method and some of them in thin section The quantity of the minerals is given in their respect to the height of the main peak of particular mineral and it is not absolute Infiltrated material By denudation processes and percolating water a lot of different material could be infiltrated through the open fissures into the karst (KOGOVSEK & ZUPAN, 1992) Mineral composition in that case depends on the composition of rocks from where the weathering remains originated The mineral composition of red loam infiltrated into the cave below Divaski hrib, from the surface is different from the mineral composition of flood loams in caves of the area (Figure 2) Cave below Divaski hrib red loam During the construction of the motorway across the Karst the works uncovered an old cave filled by two different types of elastic sediments (MIHEVC & ZUPAN HAJNA, 1996) In the upper part of open profile in the cave red loam was seen The sample consists of quartz 85 %, illite 5 %, gibbsite 4 %, kaolinite 2 %, hematite 2 % chlorite 1 %, turmaline 1 % and traces of plagioclase Due to mineral composition and position we know that this material is infiltrated into the cave from surface The lower part of the profile was different in colour and mineral composition, it is typical mechanical cave sediment. Legend for all figures Samples: Postojna cave system: P 1 Partizanski Rov, P 2 Sp. Tartarus; Planina cave: PI 1 Rudolfov rov, PI 2 entrance, PI 3 Pivka channel, PI 4, PI 5, PI 6 S/epic; Malni spring: Ml upper, M 2 lower. D 1, D 2 cave below Divaski hrib. Minerals: q quartz, ea calcite, ii illite, ka kaolinite, kl chlorite, mi microclin, pi plagioclase g goethite, h hematite, d dolomite, mu muscovite. Sympos ium 7: Phys ical Spe l eology 33


100 80 60 % 40 20 -0 q ii ka kl D 1red loam D 2-yellow loam pi g h -gi Figure 2: Comparison between mineral composition of infiltrated red loam and flood loam in the cave below Divaski hrib Flood loams The mineral composition of flood loams in the caves depends in their origin. The first case is represented in the cave by allochthonous elastic sediments from the cave below Divaski hrib and in the second case are represented by flood loams and sands from Postojna cave system and Planina cave In both cases the passages have developed in carbonate rocks but the waters flow from the non-karstic recharge area, where the rivers eroded Eocene flysch rocks and transported them into the caves (ZUPAN HAJNA 1992). Cave below Divaski hrib yellow sand and loam Cave filled with sediments was developed in Lower Paleocene limestone. according to the Basic geological map, sheet Gorica (BUSER 1968) Sample from the yellow part of the open profile consists from quartz 98 %, illite 1 %, chlorite I %; and traces of kaolinite, goethite, hornblende. This is the typical mineral composition of the cave sediment which has its origin in Eocene flysch rocks The way of sediments sedimentation is also typical for the cave environment. Postojna cave system and Planina cave The entrance to Postojnska Jama is in the eastern border of Pivka basin; the bottom of the basin consists of Eocene flysch rocks ; Postojnska Jama is developed in Upper Cretaceous carbonate rocks according to the Basic geological map, sheet Postojna (BUSER GRAD, PLENICAR 1967). The Pivka river transports into the cave sediments from the flysch recharge area The river flows through lower levels of Postojna cave system and disappears in a siphon in Pivka Jama. From there its route is unknown up to Planinska Jama a resurgence cave in the eastern part of Planinsko Polje and to spring ofMalni Figure 3: Position of the Postojna cave system, Planina cave and Malni spring. Legend E-flysch K limestone, dolomite The original rock of cave mechanical sediments is Eocene flysch ; the Nanoscica river transports the detritus into the Pivka and further into Postojna cave system In the Postojna cave system and in Planina cave, Pivka and Rakov channel, the mineral associations are very similar the quantity only varies a little. In some samples the quantity of carbonate minerals is unusually high because it is known that the origin of the caves flood loams and sands is from non-carbonate rocks Calcite can appear like cement but in analysed samples was presented in small grains, silt and clay size. These are small particles of limestone which were eroded from the cave walls % q ea ii ka kl mi pi g b d mu PI 86 5 1 1 1 I 1 4 P2 73 7 3 2 3 3 7 3 Pll 40 10 4 12 10 10 14 P12 81 I 4 2 6 1 5 PIJ 81 2 I 2 9 4 1 Pl4 81 1 3 9 4 2 PIS 78 2 1 2 8 4 3 2 Pl6 78 2 1 1 4 3 2 9 Pl7 79 2 3 13 2 Ml 85 2 3 3 3 4 M2 7 84 1 1 1 4 I Figure 4: Mineralogical compositions of flood loams and sands from Postojna cave system and Planina cave originated from Eocene flysch rocks, well expressed enrichment by calcite in same samples. Enrichment on calcite is especially well expressed in cases where the water flows through narrow passages or in siphons where the water has mechanically eroded cave walls. In cases where water flows through big channels no enrichment of carbonate minerals is detected. % q ea 100 80 60 40 20 0 .. .. .. .. .. Ill I I -...t N N M W N Figure 5: Comparison between content of q quartz and ea calcite in the samples from underground Pivka river system. Tectonic clay In the Postojna cave system, between Pivka Jama and Crna Jama significant fault zone is delineated by two fault planes. On the first look it is clear that this loam doesn't contain quartz grains which are characteristic of flood loams at this cave system. On the left side of the fault zone the limestone is undamaged and on the right side of it the limestone is crushed The tectonic undamaged limestone from the left side of the fault zone is biomicritic with fragments of shells but on the contact of the fault plane a 1 cm thick red belt is presented. This red border is parallel to the main direction of fault plane and all the biomicritic limestone is recrystallized into sparite The red colour is due to bigger concentration of less soluble iron and clay minerals which are concentrated on this recrystallized border 34 Proceed i ngs of the 12 t h Internat i onal Congress of Speleology, 1997 Switzerland Volume 1


% 100 80 60 40 20 0 q ea tectonic clay flood loam ----d ka kl pi g Figure 6: Comparison of the mineral composition of flood loam and tectonic clay from Postojna cave system. Between both final fault planes is a I meter wide fissured zone whi c h is consisted from several vertical laminas about I cm wide Between particular laminas yellow clay is situated (ZUP AN 1989 ). Particular laminae consist of sparitic limestone. The y ellow clay between them consists mostly of calcite Go e thite and kaolinite are presented onl y in tra c es (F igure 6 ). Figure 7: Fault zane from Postojna cave system a. tectonic laminated limestone b. tectonic clay c. undamaged limestone, d. te c tonic breccia C lose to the fault plane the v ery porous structure disappears Spariti c grains collapsed and clay is derived from the solid sparitic limestone it has the same mineral composition as the limestone On the right side of fault zone, close to the main fault plane where the tectonic sliding planes are seen a red clay was analysed By 90 % calcite prevails there are some kaolinite and goethite each 4 % illite and hematite are represented in traces In the red clay hematite is represented in traces and gives red colour to the clay The origin of the hematite is explained such that it is due to the tectonic pressure that squeezed the water out of the goethite and so hematite occurs The admixed minerals of the limestone like illite kaolinite and goethite concentrate especially at the tectonic plane where the carbonate solution is profoundly affected by the pressure Concl u sion Three different types of elastic material may be distinguished by their origin. In the caves opened by erosion processes or cut by dolines flood loam and sand originating from flysch sediments can be found Yellow brown colour is typical ofit but sometimes when the flood sediments are near the surface the colour is changed to red Due to diagenesis brown goethite is transformed to red hematite but the association of minerals is different like in red soils from the surface The mineral composition of red soils on karst can be totally different and depends on their origin An opinion prevailed for a long time that in our karst in all the fissures and fault zones red loams are either "terra rossa", infiltrated from the surface or yellow loams having their origin in flysch transported by water flow By microscope and x-ray analyses it was shown that this is not always the fact. Frequently within the fault zones tectonic clays are found originating within the inner fault zone and having almost entirely carbonate structure. The first type is infiltrated material into the open fissures by denudation processes and percolating water. Mineral composition in that case depends on composition of rocks from where the weathering remains are originated The second type is elastic material which was filled into opened fissures during floods the composition of it corresponds to the rocks composition of underground water flow watershed Quartz grains are characteristic of flood loams having the origin in non-carbonate rocks because of its properties Sometimes some significant heavy minerals can be found In that case the determination the origin is easier and better But man y mechanical sediments are composed of minerals which are the final product of weathering of different rocks but equal in respect to their mineral composition In such cases their origin can not be determined Clastic sediments could be enriched on calcite or dolomite These are actually small particles of limestone or dolomite from walls of the underground water passages The content of carbonate minerals in allohtonous material has been presumed to be low but the x-ray analyses and analyses in thin sections have shown that carbonate contents in a lot of samples could increased to high values from less then 5 % to more then 80 %. Very high contents of carbonate minerals can be detected in tectonic clays This third type of elastic material is significant for the inner part of fault zones in carbonate rocks This is clay size material formed by tectonic compression of carbonate rocks Tectonic clay consists almost entirely of calcite or dolomite and their admixture depends on mineral composition of the parent rocks Tectonic clay de v elops in limestone in such a way that under pressure the limestone recrystallizes Where sparitic limestone occurs is more porous especiall y at the active tectonic plane Directly at the main tectonic plane a collapse of solid limestone structure appears On the borders of the grains at first solution Sympo siu m 7: P hysical Speleo l og y 35


occurs, later sparitic grains collapse and soft, unconsolidated clay occurs. Due to the origin controlled by tectonic pressures, it is called tectonic clay This term is not yet entirely clear as in literature sometimes the expression milonite is used for technical crushed rock In tectonic clay not only calcite or dolomite are presented, depending in which carbonate rock the fault zone is developed, but also other minerals, which are admixture in them, may be found Tectonic clays may be either yellow or red. Goethite gives the yellow colour, red colour is due to hematite in the places where the water had been squeezed out of goethite and it was transformed into hematite. This clay impedes the drainage within the fault zone this is why in such fault zones no karstification was recorded. All of the described elastic sediments are characteristic for our karst but their colour is not distinctive for their mineral composition and their origin. The colour of all three types depend on their mineral composition and on physical an d chemical conditions in the environment. But the same colour ts not necessary the same mineral composition and the knowledge of the mineral composition is not enough to declare the origin of the elastic sediments. Because at the end just the most resistant minerals and also the distinctive secondary minerals, for certain environments are presented. References BUSER, S., 1968 Osnovna geoloska karta SFRJ Gorica I : 100000 Zvezni geoloski zavod Beograd, Beograd. BUSER, S GRAD, K & PLENICAR, M 1967. Osnovna geoloska karta SFRJ Postojna 1 : 100000. Zvezni geoloski zavod Beograd, Beograd KOGOVSEK, J. & ZUPAN, N. 1992 Prenos trdnih delcev pri vertikalni cirkulaciji v krasu Nase jame, 34, 13 19, Ljubljana MIHEVC, A. & ZUPAN HAJNA, N., 1996: Clastic sediments from dolines and caves found during the construction of the motorway near Diva~ on the classical Karst Acta carsologica SAZU, 25, Ljubljana. PLACER, L.,1982. Tektonski razvoj idrijskega rudBfa Geologija 25/1, 7 94, Ljubljana ZUPAN, N., 1989. Mineralogija tektonske gline v Pivki jami Acta carsologica SAZU 18 139156 Ljubljana. ZUPAN HAJNA, N., 1992 Mineralna sestava mehanskih sedimentov iz nekaterih delov slovenskega krasa Acta carsologica SAZU 21 115 130 Ljubljana 36 Proceedings of the 12 1h Internat i onal Congress of Speleo l ogy 1997 Switzerland Volume 1


Th / U dating of sea level-controlled phreatic speleothems from coastal caves in Mallorca (Western Mediterranean) by Paola Tuccimei 1 Joaq u i n Gines 2 Ange l G in es 3 a nd J oan J. Fornos2 1 Dip Scienze Geologiche, Universita "Roma Tre" Roma Italy; 2 Dept. Ciencies de la Terra, Univ. Illes Balears Palma de Mallorca, Spain; 3 Museu Balear de Ciencies Naturals, Soller, Mallorca Spain. A bstract Phreatic speleothems form today at the surface of the brackish pools existing inside littoral caves of Mallorca island (Western Mediterranean) Their occurrence directly identifies the height of the sea level at the time of carbonate deposition, because such pools are physically connected with sea waters. Therefore ancient phreatic speleothems can be used to reconstruct the Mediterranean's fluctuations during the Pleistocene, since many alignments of these deposits are present in several caves of the studied area Some phreatic speleothems have been dated using the Th/U method in order to determine the position of the sea level during the Late Quaternary Three high sea stands have been recognized at different heights between 1.4 and 2.5 meters above current sea level in caves located in various sectors of the eastern coast of Mallorca They date back to 83, I 08 and 124 ka B P and can be related to several minor events within stage 5 of the marine oxygen isotope record. Differences in height of about I meter can be observed in samples corresponding to the same sea paleolevel. Those which are particularly well-documented pertain to substage Se, with higher elevations towards the northernmost localities These altimetrical disturbances could be explained in terms of a tectonic tilt of the investigated area, and they can also be inferred from other stratigraphical and geological evidence. Res u m e n Es posible constatar la formaci6n actual de espeleotemas freaticos en la superficie de los lagos salobres de numerosas cuevas costeras en la isla de Mallorca (Mediterraneo occidental) Su presencia registra la altura de! nivel marina en el momenta de la deposici6n de estos carbonatos freaticos, habida cuenta de que dichos lagos subterraneos estan conectados fisicamente con !as aguas marinas Por lo tanto los dep6sitos antiguos de espeleotemas freaticos pueden ser utilizados para reconstruir !as fluctuaciones de! Mediterraneo durante el Pleistoceno ya que abundantes alineaciones de estos espeleotemas existen en diversas cuevas de! area en estudio. Se han datado algunos espeleotemas freaticos usando el metodo Th/U con la intenci6n de determinar la posici6n de! nivel marina durante el Pleistoceno Superior. Tres estabilizaciones altas de! Mediterraneo han sido reconocidas a diferentes alturas entre 1,4 y 2 5 metros por encima de! actual nivel marina en cuevas localizadas en distintos sectores de la costa oriental de Mallorca Dichas estabilizaciones se remontan a edades de 83 108 y 124 ka B.P pudiendo ser relacionadas con algunos eventos menores dentro de! estadio 5 de! registro marina de is6topos de oxigeno. Diferencias de altitud de hasta I metro pueden observarse en muestras pertenecientes al mismo paleonivel marina estando particularmente bien documentadas !as correspondientes al subestadio isot6pico Se, el cual presenta altitudes mayores en !as localidades mas septentrionales Estas anomalias a l timetricas deben ser explica d as en base a un basculamiento tect6nico de! area investigada, ta! como se desprende tambien a partir de otras evidencias estratigraficas y geol6gicas. 1. Introduction Mallorca is a mostly carbonate island of the western Mediterranean basin that shows very interesting coastal karst features in its different natural regions. A lot of karstic littoral caves are developed all around mainly along the eastern areas of the island. Such cavities have undergone a complex morphological evolution during Pleistocene times and contain deposits that include breccias with paleontological remains, and speleothems This materia l is of great speleo-chronological and paleoclimatic interest (GINES & GINES, 1986 ; 1995) due to the fact that its deposition was frequently affected by Quaternary sea level oscillations This paper deals with a set of chronological information supplied by sea-controlled phreatic speleothems related to Pleistocene marine paleolevels. In this respect a programme of isotopical studies of these speleothems has been developed over the last three years, including Th/U datings which are 2. Phreatic speleothems in Mallorcan caves The littoral karstic areas of Mallorca present many caves partially drowned by brackish waters (GINES, 1995) as a consequence of post-glacial sea level rising. Partial drowning of such coastal caverns creates subterranean brackish pools that sometimes reach great dimensions This happens to be a coastal phreatic environment controlled by the sea level where periodical water oscillations take p l ace adjusted to minor sea l evel changes like tides. A great number of different kinds of phreatic speleothems are associated to this singular geochemical environment. Today the formation of calcite rafts on the surface of these hypogean pools is often found as well as conspicuous bulky overgrowths developed on the cave walls and also around those stalagmites or stalactites situated in the current fluctuation range of the water table (POMAR et al. 1979). Phreatic crystallizations of calcite and aragonite located a few meters over the level of present day brackish pon d s are co mm on too (GINES et al. 198 1 b); through the bands of speleothems marked by them, former marine levels of stabilization corresponding to glacial eustatic sea risings are recorded Most of the phreatic crystallizations both the ancient and the Symposium 7 : Physical Speleo l ogy 37


present day ones, overgrow over pre-existent common vadose speleothems and adopt bulky morphologies. Sometimes these carbonate coatings are belt-like forms developed around stalagmites and columns In many cases, especially when the phreatic overgrowth affect the hanging tip of stalactites the original speleothem morphology can be greatly modified Occasionally, some floating calcite rafts are also trapped between the growth layers of these subhorizontal phreatic coatings. Phreatic speleothems of Mallorcan caves offer a great morphological textural and mineralogical variety as was showed in several publications (GIN E S et al. 1981 b ; POMAR et al 1976) These authors pointed out the paleoclimatic significance of their mineralogy by relating the presence of aragonite to warmer events which in beach sediments, results in the appearance of malacological terrnophile fauna typical of the Eutyrrhenian (Riss Wurm interglacial) The main interest in such processes of carbonate precipitation is that ancient positive fluctuations of the sea level, corresponding to interglacial periods, have been recorded inside numerous coastal caves by means of strictly horizontal alignments of phreatic speleothems The formation of this kind of crystalline deposits is related to paleolevels attained by the groundwater table as a result of glacial-eustatic sea oscillations (GINES et al ., 198 la ; POMAR et al 1987) In the case of Mallorca a great number of phreatic crystallization paleolevels have been observed between the current sea level and +40 meters above. In some papers (GINES & GINES, 1974), the feasibility of altimetrically correlating these deposits with ancient coastlines corresponding to the middle and upper Pleistocene has been considered These coast-lines have been identified by means of the stratigraphic and paleontological height u locality sample f' 234U/ 238U (m) (ppb) --~ Cv. C. Varques 8 l VB-D1 # +1.4 228 2 113 099 {:::v de Cala Falc6 FA-D2a# +1 9 542 1 378 0 003 FA-D2b ~ T + 1 9 651 I SE-D2a # +1 5 174 1 ~ 466.002 Cova des Serral SE-D2b + 1 5 200 I 1 453 0.032 I Coves del Pirata PI-D1 +2 1 j 262 1.663 0 062 ---Cova del Dimoni DI-D1 +2 5 1273 1 090 0 018 DI-D2 +2.5 1 2640 1.185.013 DI-D3 +2.5 1887 ~ 08 015 Cova de na Ban 300.7 Coves del Drac I DR-D4 1 +3 3 I 2888 ~ 039 016 inf. 1.038 0.040 > 350 --# mass spectrometry alpha counting 230ThJ232Th activity ratio is < 20, thus the true age is lower than the measured age (193 ka) Table/: Data on localities and obtained samples, as well as analytical results of the dating programme. 38 Proceed i ngs of the 12 1 h Internat i onal Congress of Speleology 1997 Switzerland Vo l ume 1


situated 30 meters above the present sea level should go back at least to the Mindel Riss interglacial. lo 1981 a preliminary isotopical dating programme on Mallorcan phreatic speleothems was started. The analyses were carried out using the Uranium series method, and confirmed the previously stablished chronologica l model, with regard to altimetric correlations between phreatic speleothems and Pleistocene beach deposits. The results of this absolute dating programme yield ages that range from 3.9 ka (clearly post-glacial crystallizations) to above 350 ka reaching the limits of this method (HENNIG et al ., I 981 ; GI ES & GINES, 1989 1993a 1993b). The sampled phreatic crystallizations younger than 250 ka show good correlation with the climatic events that involve a sea level similar or higher than the present-day one (stages I, 5 and 7, which correspond to warm pulsations); the samples that reveal ages above 300 ka (paleolevels higher than 30 meters a.s.l.) have to be assigned at least to the stages 9 or I I of the marine oxygen isotope record established by SHACKELTON & OPDYKE (1973) 3. Sampled localities Eight caves (Figure 1) were investigated in order to collect phreatic speleothems pertaining to several high sea-paleolevels Most of these caves are developed on the post-orogenic upper Miocene calcarenites from the Migjom platform area of Mallorca, whereas only two among the studied caves Cova de na Barxa and Cova de na Mitjana are located in folded Jurassic limestones that mainly integrate the Serres de Llevant mountain range Samples were obtained from phreatic speleothems that record several marine high-standings ranging from 1.4 to 5 8 meters above current sea level with most situated between 1.4 and 2 5 meters a s.l. Data about cave's names, samples obtained, and heights of paleo sea-levels represented in each locality are gathered in Table I. In spite of the fact that detailed mineralogical and textural studies of the dated samples are still lacking calcite seems the AGE (ka) lsotoplcal stages aso. . . . . ... . ... . ... . ... . . ... . . . . . . . . . . . ..... 0 . . 300. ... ..... ... .... .. . . . . .. ..... .. .. ... . 2so200.. . . . . . . . . . . . . . . . . ... . ... . . . . . . . . . . . . . . . . 150100500 0 9 7 Se Sc Sa o~----------------'----' UB FR SE PI DI BR Ml DR sampled localities Figure 2: Thill ages of phreatic speleothems studied in this dating programme. Caves are plotted in the same order as in Table I. main constituent of these speleothems; however some of them are aragonitic, like samples from Cova del Dimoni that are a spectacular phreatic coating of aragonite needles. Altimetrical information on Pleistocenic beach deposits in Mallorca (BUTZER, 1975; CUERDA, 1975) suggest that marine paleolevels recorded by the analyzed speleothems must belong to upper and middle Pleistocene high-standings fitting presumably in the range of applicability of Tb/U dating method In this context, the present dating programme of Mallorcan phreatic speleothems has been carried out in order to attain chronological information on the studied caves as well as more preci s ion on the late Pleistocenic sea level history of the Mediterranean meters (above present-day sea levet) 3 0-r------------------~ 2 5 -~ Se 2 0 1 5 1 0 ... . .. . . . . . . . .. . . .......... .. 0 5 0 SE South UB FR PI DI BR North Figure 3: Increasing elevation toward the north of dated phreatic speleothems pertaining to several substages of the last interglacial event. 4. Dating results Absolute dating of collected samples were performed by means of Uranium series methods (VESICA et al 1996) ; results are compiled in Table I. Calculated ages range from 83 ka to >350 ka, showing good agreement with stages 5 and 7 of the oxygen isotope record that correspond with events of high sea level (Figure 2). A sample from Coves del Drac which age exceeds the dating limits of the method, must pertain at least to climatic stage 9 Particularly well-represented are high-standings at ages of 83 ka, 108 ka and 124 ka B.P., dates that correlate quite well with the ages usually asigned to substages 5a 5c and Se of marine isotope record (SHACKELTON & OPDYKE 1973) These data document three main Mediterranean sea level stabilizations at heights between 1.4 and 2.5 meters a.s.l. corresponding to the last interglacial event. Our data are also in agreement with the ages of the marine terraces from Campo de Tiro in the Bay of Palma de Mallorca recently dated by U-series methods (HILLAIRE-M A RCEL et al., 1996). These units yielded ages of 135 and 117 ka and can be altimetrically correlated with our phreatic crystallizations paleolevels, since the oldest beach deposit culminates at + 3 meters a.s.l. The fact that the same paleolevel becomes recorded in diverse localities at different heights above current marine level, is evidence of a tectonic tilt of the investigated area In this sense Symposium 7 : Physical Speleology 39


phreatic s peleothem paleolevels belonging to stages 5a, 5c and 5e are recorded at higher elevations in northernmost localities (Figure 3 ) Data brought forward by the present dating programme confirm a previously estab l ished geomorphological assumption that correlates phreatic speleothems of coastal Mallorcan caves with s ea leve l high-standings related to interglacial events. Specially interesting are phreatic speleotherns in order to contribute to the knowledge of Mediterranean sea level history during upper and middle Pleistocene times 5. Conclusions Re s ults of these investigations on speleochronology of coastal caves in Mallorca island can be summarised as follows : The obtained ages of phreatic speleothem deposit i on in the investigated localities ranges from 83 ka to >350 ka B .P Speleothems with ages younger than 250 ka correlate well with isotopical stages 5 and 7 which are warm events with marine level higher than the present-day one Particularly well-documented are some minor events within isotopical stage 5 Dates of 83 ka I 08 ka and 124 ka correspond respectively to substages 5a 5c and 5e of the marine oxygen isotope record The increasing elevation toward the north of phreatic speleothems showing identical ages, gives evidence of a tectonic tilt of the studied area during upper P l eistocene times Acknowledgements This investigation has benefitted of two grants from the Mu se u Ba/ e ar d e Ci e nci e s Naturals (Soller Mallorca), during the years 1994 and 1995 Likewise this work is part of the DGICYT project number PB94-l l 75 References BUTZ E R K. W 1975 Pleistocene littora l -sedime n tary cycles of the Mediterranean basin: a Mallorq u in view In : (K.W Butzer & G L. Isaac, eds ) : After the Austra/opithecines : s tra t igraph y, e c ology and c ulture change in the Middl e Pl ei sto ce n e Chicago : 2571. Cu E RDA J. l 975 Los tiempos cuatemarios en Baleares. Instituto de Estudios Balearicos, Palma de Mallorca, 304 p. GI NE S A & J GrN E S 1974 Consideraciones sobre los mecanismos de fosilizacion de la Cova de sa Bassa Blanca y su paralelismo con formaciones marinas de! Cuatemario Boll. Soc Hist. Nat Ba/ears 19 : 11-28. GI N ES, A & J. GrN E S 1 986 On the interest of speleo chronological studies in karstified is l ands The case of Mallorca (Spain) Com 9 g Cong. Int Espel e ol Barcelona l: 297-300. GIN E S, A & J GI N ES 1989 Absolute dating of phreatic speleothems from coastal caves of Mallorca (Spain). Pro c. 10th Int Congress Speleol Budapest I : I 91-193 GrNES, A ; J. GINES & L. POMAR 1981a Phreatic speleothems in coastal caves of Majorca (Spain) as indicators of Mediterranean Pleistocene paleo l evels Pro c 8th Int Congress Speleol., Bowling Green, 2: 533-536 GrN E S J 1995. L'endocarst de Mallorca : els mecanismes espeleogenetics / Mallorca's endokarst: the speleogenetic mechanisms In : (A Gines & J Gines eds.): El carst i les coves de Mallorca / Karst and caves in Mallorca. Endins. 20 / Mon Soc Hist. Nat Ba/ears. 3 : 71-86 GrN E S J. & A. GrNES. 1993a Speleochronological approach to some coasta l caves from "Cap Verrnell" area in Mallorca island (Spain). Proc. XI Int Congress Sp e leol. Beijing, 5659. GINES J. & A. GrNES 1 993b Dataciones isotopicas de espeleotemas freaticos recolectados en cuevas costeras de Mallorca (Espana) Endins 19 : 9-15. GrNES J. & A GrNES. 1995 Aspectes espeleocronologics de! carst de Mallorca / Speleochronological aspects of karst in Mallorca In : (A. Gines & J. Gines, eds ): El carst i les coves de Mallorca / Karst and caves in Ma ll orca Endins. 20 / Mon Soc Hist Nat. Ba/ears. 3 : 99-112 GINES J.; GINES A & P OMAR, L. 1981b. Morphological and mineralogical features of phreatic speleothems occurring in coastal caves of Majorca (Spain) Pro c. 8th Int Congress Speleol B owling Green 2 : 529-532 HENNIG G. J.; A. GrNES ; J. GrNES & L. POMAR 1981. Avance de los res u ltados obte n i d os mediante datacion isotopica de algunos espeleotemas subacuaticos mallorquines Endins 8: 91-93. HILL A IRE-MARCEL, C. ; C GARIEPY ; B. GHALEB; J L. GOY ; c ZAZO & J CUERDA. 1996. U-series measurements in Tyrrhenian deposits from Mallorca Further evidence for two last-interglacial high sea levels in the Balearic islands Quaternary Science Reviews 15: 53-62 POMAR L.; A. GrNES & R FONT A R NAU. 1976 Las cristalizaciones freaticas Endin s. 3 : 3-25 POMAR L. ; A. GrNES & J. GrNES 1979. Morfologia, estructura y origen de los espeleotemas epiacuaticos Endins. 5 6: 3-17. POMAR L. ; A. RoDRiGUEZ ; J.J FORN6s; A GIN E S; J. GIN E S ; A FONT & A MORA 1987. Phreatic speleothems in coastal caves : a new method to determine sea-level fluctuations. In : (C Zazo, e d .): Late Quaternary sea-level changes in Spain Museo Nacional de Ciencias Naturales C.S.I.C Madrid Trabajos sob re Ne6geno-Cuaternario l 0: 197-224 SHACKELTON N. J. & N. D OPDYKE 1973 Oxygen isotope and palaeomagnetic stratigraphy of Equatorial Pacific core V28238 : Oxygen isotope temperatures and ice volumes on a J05 year an d 106 year scale Quaternary Research. 3: 39-55. V E SICA P ; P TuCCIMEI; B TuRI ; J.J. FORNOS ; A GINES & J. GIN E S. 1996 Th/U dating and C 0 isotope analyses of speleot h ems from coasta l caves in Mallorca (Spain) 30th JGC, Abstracts, Beijing 1: 87. 4 0 Pro c eedings of the 12 th International Congress of Speleo l ogy 1997 Switzerland Vo l ume 1


Speleothems dating using the thermoluminescence method V. Labau, E. Gaspar, T. Paunica Institute of Physics and Nuclear Engineering, Bucharest, Romania Abstract The thermoluminescence method was used to determine the age of a stalagmite from Wind Cave, Padurea Craiului Mountains. When certain crystallized inorganic substances are exposed to ionizing radiations and then warmed until a certain temperature characteristic of each material they emit a luminous radiation. In order to establish the geological age, the natural thermoluminescence of the calcite and the natural dose accumulated during geological time were measured. It was necessary also to determine the yearly dose due to the radiation emitted by the natural radioactive elements (U, Th, K) present in very small quantities in speleothems The accumulated natural dose was of 55 934 Gy The samples was warmed at 450 C. For the artificial irradiation of the calcite samples, a beta radiation source from 90 Sr 9 Y was used. The age of the stalagmite was 59,052 9,600 years. 1. Introduction The thermoluminescence (TL) method is based on the thermoluminescence phenomenon which consists in the fact that certain crystallized inorganic substances, when they are exposed to ionizing radiation and then warmed until a certain temperature characteristic of each material, they emit a luminous radiation The intensity of the emitted light is proportional to the radiation dose on which the substance has been exposed. The establishing of the age in geology by the TL method is based on the fact that the geological materials contain certain minerals which have thermoluminescent properties and may be used as an integrator dosimeter. Such a mineral is also the calcite which is the main constituent of the stalactites and stalagmites The event which is dated by the TL method is the moment of the crystallization of calcium carbonate From that moment the material is exposed to the radiation emitted by the natural radioactive elements (U, Th, K) which are found in the matrix. In the course of a year, because of the integrated character in the material is accumulated the dose (d) named the yearly dose. During the geological time, T, is accumulated the dose Do = d x T which represents the accumulated natural dose. From this relation of accumulation of the natural dose one can determine the geological age (T) as follows: T= Dn d (I) The age of a stalagmite fragment collected from Wind Cave Romania, was determined using the TL method. 2. The calcite thermoluminescence In order to establish the geological age, it is necessary to measure the natural thermoluminescence (NTL). For this determination, we have used a the thermoluminescence curve The TL curve represents the temperature variations of the luminous flux emitted by the matrix progressively warmed from the environmental temperature until a maximum pre-established temperature (~450 C). Figure 1 presents the natural thermoluminescence (NTL) and artificial thermoluminescence (A TL) curves obtained for the studied stalagmite The NTL curve is due to the calcite irra diation by the natural elements (U, Th, K) contained in very small quantities in the stalagmite. The NTL curve presents a lonely peak situated at a temperature of about 320 C. 1.1..l u z UJ~ u E V')~ 3~ 2 v, :::::>z ....Jw 012 z a::: 1.1.J I ID =720Gy 0 12 0 180 230 320 TEMPERATURE ( C ] Figure 1: Natural and artificial termol11minesce11ce curves for the studied stalagmite The artificial thermoluminescence curve (A TL) may be obtained by artificial irradiation using a beta or gamma radiation source. The A TL curve presents peaks in evidence at the temperatures of 120 C 180 C and 320 C It also presents a peak at the temperature of 230 C but less in evidence. The peak from the temperature of 120 C is unstable and disappears comparatively soon, when the irradiated sample is stored a long time at the room temperature During geological time, because of the thermic agitation, the peaks from the middle temperature (180 C and 230 C) present a decline (fading) which may lead to their disappearance. For these reasons, the peaks from the low temperatures (120 C) and middle ones {I 80 and 230 C) don't appear in the natural thermoluminescence curve (NTL) By irradiation with ionizing radiation the thermoluminescence sensitivity of the peaks from the low and middle temperatures increases. This is due to the supplimentary flaws caused by the irradiation of the calcite (ZELLER, 1968) Symposium 7 : Phys i cal Speleo l og y 41


3. The determination of the natural accumulated dose In order to determine geological age it is necessary to establish the natural dose (Dn), accumulated during the geological time, T The Dn dose is established by the measurement of the NTL and the dose conversion For this purpose it would be normal to use the peak from the high temperature (320 C). But this maxima although being stable, cannot be used at the measurement of the thermoluminescence in dosimetric purposes because its intensity depends not only on the absorbed radiation dose but also on other uncontrollable factors. In order to establish the Dn dose we used a method based on the combined action of the pre-irradiation (pre-dose) and warming (at 450 C) on the thermoluminescence at low and middle temperatures of the calcite. For the stalagmite we have studied, the effect was an increasing of the thermoluminescence sensitivity at low and middle temperatures, proportional to the dose (pre-dose) For the practical application of this method we have proceeded in the following way : a calcite sample was warmed at 450 C and after the cooling was established the thermoluminescence sensitivity (SJ by using a test dose (D, = JG ,); for other sample it has been applied the pre-dose (D), it has been warmed at 450 C and after the cooling, it has been tested again the sensitivity as having the value (SN+ D). In all the cases the measurement of the TL sensitivity was made on the peak of the temperature at 120 C which was proved to be reproducible By theoretical reasons and from the analysis of the data obtained it results that the relation S = f(D) between the thermoluminescence sensitivity (S) of the calcite and the dose (pre-dose) D is, in a first approximation, a straight line in the form : S=a+ bD (2) By the extension of the straight line obtained, towards the negative side of the abscissa, we determine the accumulated natural dose, Dn (fig. 2). w u~ Z> .E. V) z~ ~> ::::> _J ..... 0 V) z a: w w V) I t/i xEXPERtMENTAL VALUES oCOMPUTED VALUE~ Dn 0 ___ _. ___ _._ ___ __._ __ __J 200 400 600 800 EXPOSURE DOSE I Gy] Figure 2: The thermo/11mi11esce11ce se11sitivity versus th e dose 4. The determination of the yearly dose As it results from the relation (I) in order to establish the age it is also necessary to determine the yearly dose (d) due to the radiation emitted by natural radioactive elements (U, Th, K) present in very small quantities in speleothems The yearly dose of the stalagmite was determined indirectly from the natural radioactive elements content using a conversion table concentration dose (AITKEN & all, 1968) The radioactive natural elements content was established using the neutron activation analysis method. Thus, the radioactive elements content was : U = 0.3 ppm; Th = 0.0 ppm and K = 0 01 %. From the conversion of this radioactive content, the yearly dose resulted: d = 94.72 x 10' Gy / y. (3) 5. The experiment For the thermoluminescence measurement the stalagmite fragment was transformed by crushing, grinding and sieving in a microcrystalline powder. The diameter of the microcrystals used in this experiment varried between 80 and 160 m. For natural (NTL) and artificial (A TL) thermoluminescence measurements, the same quantity of microcrystaline power was used. The dosimetrical information was put in evidence at the TL using a conventional apparatus for detector TL measurements, type Mark IV TLD Reader. The thermoluminescence curve was determined by using a X-Y recorder. As a measure of the TL intensity one can take the peak height from 120C temperature. For the artificial irradiation of the calcite samples, a beta radiation source from 90 Sr 90 Y was used TL curve was recorded aft er calcite samples irradiation, directly. For the determination of the accumulated natural dose (Dn) we have used the "predose" method described above. As a measure of the TL sensitivity we used the heigth of the peak from 120 C obtained after calcite irradiation at a dose test of I Gy. The samples warming at 450 C was made with the same Mark IV apparatus. The samples warming was repeated 4 times in order to eliminate the residual TL influence on calcite sensitivity. 6. Experimental results In order to establish the growth of the calcite TL sensitivity depending on the dose (predose), separately calcite samples were irradiated at growing doses. The irradiation doses used (D) and the TL sensitiviy resulting are presented in table no I. The irradiation dose (D) is expressed in Gy and the thermoluminescence sensitivity (S) in m V D(Gv) S (mV) 0 45 105 90 270 210 570 380 720 500 Table 1: Irradiation doses and thermolumi11esce11ce se n sitivity 42 Proceedings of the 12 th Internat i onal Congress of Speleo l ogy 1997 Sw i tzerland Vo lum e 1


The thennoluminescence sensitivity versus the dose (predo e) Dis presented in fig 2 From these data one can see that the relation S = f(D) between the calcite thennoluminescence sensitivity and the dose (predose) is a straight line. The calculation of this traight line is given by the relation: S = 35 233 + 0 6299 D (4) The accumulated natural dose is obtained extending the straight line until the X axis, where S = 0: On= 55 934 Gy (5) The geological age of the studied speleotheme was detennined using the natural accumulated dose established as above and the yearly dose. From these data it results the age of the stalagmite: 59,052 9 600 years References AITKEN, M J. ZIMMERMAN D W & FLEMING S J 1968. Thennoluminescent dating of Ancient Pottery N atur e. 219 : 442-445 ZELLER, E J 1968. Geologic Age Determination by Thenno luminescence. In: (D. J Mc Dougall) : Thennoluminescence of Geological Materials Academic Press London & ew York, 311 p. Symposium 7: Physical Speleology 43


U series dating of phases of speleothem deposition and erosion in Baradla Cave, Aggtelek National Park, Hungary D e r e k F ord D e p a rtment of G e ograph y M c M ast er U ni ve r s i ty, Hamilton O N, C anad a L 8 S 4Kl L a sz lo Zambo D e partm e n t o f G e ogr a ph y E o tv o s Lorand U ni ve r s it y, Budap es t 1024 Hungar y The Domica-Baradla System The cave system of D omica (in S l ovakia) and Aggtelek-Barad l a (in H u ngary) is comp r ised p r i n cipally of a large, low gradient river passage The rocks is densely fractured limestone and dolomitic limestones of Perno-Triassic age that became kartisfied hills in the late Cretaceous and early Tertiary and were then partially buried by elastic s ediment s. D uring exhumation a stable watertab l e was established in the karst ; the r iver passage formed along it fed by tributary streams collecting on surviving elastics to west. In its downstream section (Baradla Cave) the stream flow for most of the hydrologic year is now captured to younger passages at a lower elevatio n but there is still significant flooding occasion The passage is well decorated with speleot h em including (i) large columns and bosses active l y growing in rece s ses sheltered from modem floods ; (ii) large to very large columns that were toppled, broken and eroded b y ancient floods in t h e bedrock channel ; (iii) think flowstones and small sta l agmites on bedrock and detrital terraces and (iv) small stalagmites growing at the edges of t h e modem b edrock chan n el. U series dating Twelve small calcite samples were taken from points in toppled or broken speleothems between O and 3 m abo v e the m odem channe l. They have bee n d a t ed by the 230 Th /23 4 U met h od using a lph a spectrometry The uranium content is v ery low (0 04-0 1 ppm). This is typical where the speleothem feedwaters are derived only from pure limestone s (a s at Baradla) rather than from limestones with shales or shaly interbeds; it makes accurate dating difficult. In addition the speleothems contain much detrital t h o r ium from the floodwaters as in d icated by 230 Th/ 232 Th ratios generally < 1 O ; a s a consequence it is not possible to determine precisely how much 23 Th has been created by decay of 134 U since depo s iti o n Despite these difficulties twenty uncorrected 230 Th U ages display a c l ear pattern of successive growth period s that correlates we ll with the published records of periodicity for a set of 500 dates from spe l eothems in NW Europe. Conclusions Our conclusions are (I) little or no speleothem in the Barad la river passage is o l der than the Last lnterglacial (I s otop e Stage 5 Riss-Wurm); (2) most rapi d and abunda n t growt h occurred during the Last Interglacial and the Holocene There were few floods during the Last Interg l acial. Th i s permitted large speleothems to grow in the channel. The Holocene deposits are comparatively small in volume except where t h ey are overgrowths onto the Last Interglacial bosse s and columns ; (e) d uring the Lower Wurm stadials an d interstadia l s (isotope Stages 4 and 3) the passage was aggraded by ri v er laid pebbles, sands and clays, forming terraces There was calciate deposition at diminished rates, on the terrace s and on columns and bosses standing in the channel. (4) There was little or no speleothem deposition during the main Wurm cold stage (isotope stage 2) (4) The passage was swe p t by major floods that removed most elastic terrace deposits and toppled the large speleothem columns and bosses. This flood period probably correlates with the cold climate of the main Wurm but could be a consequence of primary forest removal by early human sett l ers 44 Proceedings of the 12 th I nternational Congress of Speleology 1997, Sw i tzer l and Vo l ume 1


Uranium-Series Dating of Speleothems from Amaterska and Holstejnska Caves, Moravian Karst, Czech Republic Helena Hercman 1, Stein-Erik Lauritzen\ Jerzy Glazek3, Jan Vit' 1 Institute of Geological Sciences, Polish Academy of Sciences, ul.Zwirk.i i Wigury 93, 02-089 Warszawa, Poland 2 Department of Geology, Bergen University, Allegaten 4 I, N-5007 Bergen, Norway 3 Institute of Geology, Adam Mickiewicz University, ul. Makow Polnych 16, Poznan, Poland Czech Geological Survey, K.larov 3, 118 21 Praha I, Czech Republic Abstract Speleothems were sampled from Holstynska and Amaterska caves, The Moravian Karst, for Uranium-series dating and stratigraphic analysis. Dates spanning from present-day (i.e actively growing speleothems) to beyond the limit of the method (350 ka. a-particle counting) have been identified. In all, 4 phases of elastic sedimentation, interrupted by massive speleothem deposition was found in the paleoponor cave. Holsteijnska cave 1. Introduction The Moravian Karst is located north of Brno, in the Czech Republic Geologically, it belongs to the Bohemian Massif, geomorphologically to the Western part of Drohawska Upland Well developed karst features are widespread in the Devonian limestones. The classical karst forms are concentrated within an area of about I 00 km '. The most abundant karst formations can be found in the northern part of the Moravian Karst between the karst canyons of Suchy and Pusty Zleb 2. Holsteijnska Cave This part of the Moravian Karst contains the Holstejnska paleoponor Cave, which was discovered in 1966. The cave was almost completly choked with sediments, and all the presently known galleries (600 m) have been excavated by members of the local caving society The cave passage itself consists of a very wide trunk passage with an almost totally flat roof. Walls and ceiling are only sporadically exposed along the excavated galleries The passage cross-section suggests that the profile is in part developed as a classical water-table cave passage, possibly by paragenesis. Stratigraphic sections of the infilling sediments are well exposed on the walls of the artificial galleries We have identified 3 sequences of sediments (I, II and III) The oldest consists of strongly weathered gravels This gravel deposit probably filled the passage completely, but was eroded at a later stage The second sequence consists of red coloured silt and gravel. The youngest, uppermost deposit consists almost entirely of silt The three sequences are separated by speleothem These speleothem layers consists of broken flowstones blocks, sometimes overturned Twenty-eight subsamples were collected from the galleries in Holstejnska Cave and were dated with the U-Series method (Fig. 2). Three periods of speleothem deposition may be recognized from the data (Table I) 3. Amaterska Cave Amaterska Cave is one of the most famous caves in Moravian Karst. It is a horizontal river cave, several kilometers long and is a part of the underground course of the Punkva river. Eleven speleothem samples were dated with the U-Series method We can distinguish 4 periods of speleothems deposition (Table I). The results are compiled in Figure I. 4. Discussion Based on the dating results we may suggest a time scale for the development of the Holstejnska and Amaterska Caves (Fig 2). The speleogenetic history of the Holstejnska Cave began more than 350 ka ago. After development of the main passage there was a first episode of elastic sedimentation. The passage was most probably completely filled with fluvial gravels and sands. Later these sediments were partially eroded. Into the empty space, before 350 ka ago, the first recognizable speleothem deposition began. Erosion episodes probably occurred in between speleothem deposition After deposition of phase 2H (speleothems), the strongest erosional and depositional cycle occurred. Most of the sediments and speleothems which filled the cave passage were destroyed and redeposited. Then, the youngest speleothem deposition (IH) started. Quite recently yet another depositional phase occurred, when a layer of gray silt covered Table 1: Speleothem growth zones, based on U-series dating of flowstones and stalagmites in Holstejnska (H) and Amaterska (A) caves Period 1H 2H 3H 4H IA 2A 3A 4A Time Range, ka 9-16 100-145 150-300 > 350 0-10 45-80 90-140 180-200 Comments Possibly 2 phases of growth Marine Isotope Stage 1&2 2 7 and/or interstadials of 8 9 or older I interstadials in 3 & 4, und 5a? 5 7 Symposium 7 : Physical Speleology 45


I 0.8 0.6 0.4 0.2 -0.0 -0.2 -0.4 -0.6 -0 8 -1.0 -1.2 Holsteijnska and Amaterska Amaterska Holstejnska 0 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 50000 100000 150000 200000 250000 300000 350000 400000 Age [yrs] Fig. 1: Comparison between t/re U-series dating results from Amaterska and Holstejnska caves with t/re marine oxygen isotope stages. Arrows indicate infinite ages, with lower error limit (1 a) shown. All error bars are J a. Holocene speleothems (Fig 3). The maximum age of speleothem s so far found in Amater s ka cave i s 250-300 ka (Fig. 2) After a phase of large rockfalls, the fir s t recognizable phase of s peleothem deposition began (4A) These speleothems cement the breakdown blocks together. The next phase of speleothem deposition occured at 140-90 ka (3A) after a period of s and sedimentation This was the main phase of speleothem deposition and correspond to the last interglacial. The massive flowstones from this phase are 0.5-2 m thick A period of erosion followed by deposition of sands and gravels occured after this. The underground river (Punkva) reached its present level by cutting through about 4 m of older sediments. Then the youngest speleothem deposition started ( 1 A). At this time speleothem deposition was rapid Columnar stalagmites which are more than 3m tall have grown on top of the older sediments Chronologically they cover all of the Holocene. At the beginning the growth rate was ea 240 mm / ka but after about 5 ka it increased to 750 mm/ka Fig. 2: Development of Holstejnska (H) and Amaterska (A) caves during t/re last 1 350 ka. First column: Oxygen Isotope stages. Second Column: Growth range of speleot/rems. Legend: J) Flowstone/stalagmite deposition, 2) Clay/silt, 3) Sand, 4) Gravel, 5) Breakdown, 6) Hiatus. Semicircle symbol: Placement of passage formation (speleogenesis sensu stricto). 0 fCO 20Q lOO 400 .. 0 H 2 I 3 s 6 ) a fO I .. "' 46 Proceedings of the 12 th Interna ti onal Congress of Speleo l ogy 1997 Switze rl and Volume 1 A H A ------15:511 ~l I 1 03 ~'I I rBJS D:]6 I. 0 .., O I


Conclusions The two caves of the Moravian karst that we have investigated, display proliferent speleothem deposits that are suitable for Uranium-series dating The samples are also suitable for stable isotope work, as Holocene stalagmites may grow up to more than 3m in height corresponding to an average growth rate of 0.3 mm/year. maximum speleothem ages go beyond the limit of the dating method (> 350 ka by a-particle counting). Further work will focus on attaining higher time-resolution (TIMS-dating), stable isotope stratigraphy and palaeomagnetic investigation of the older parts of the sequences. Fig. 3: Late-Glacial to holocene stalagmite co/11m11 that was covered by later (Holocene) sediments. The sample was exposed during co11str11ctio11 of the artificial gallery. JH Z Symposium 7: Physical Speleology 47


Potential use of uranium series dating of calcareous algae as a tool for dating paleo-watertables in maritime karst settings He nri ette L i nge & Stei n -Erik Lauritzen Department of Geology, Bergen University Allegt. 41, 5007 Bergen, Norway Ab s tr a ct Calcareous algae which are sessile, marine organisms occur sometimes on cave walls in coastal karst and indicate minimum high sea stands. We investigate the possibility for applying uranium series dating to such deposits. Calcareous algae from Kapp Ekholm, Svalbard, has been evaluated for dating potential. The material display quite high uranium content which means that precise ages may be obtained from small samples, provided that they are unaffected by weathering processes. Because of difficulties in the accurate identification of the fossil algae only generic names (i.e. Lithothamnium) are used. This technique is also applied to algae situated in the twilight zone of a paleospring cave, north Norway. Introduction Numerous morphological indicators of paleo-watertables may be identified in karst caves (e.g. watertable corrosion notches, vadose-phreatic boundaries, and elastic sediment deposits and erosion levels in them). Dating of paleo watertables in coastal karst can be used for construction of shoreline displacement diagrams, on the assumption that the cave system is situated below the upper marine limit in the area. Such watertable markers may be dated providing that they are associated with datab l e deposits like speleothems, calcareous concretions, bone and shell fragments, or cap mud containing organic material. In situ, vadose speleothem deposits gives a maximum level of the paleo-watertable. Precipitation of calcareous concretions occurs in water-saturated sediments shortly after deposition of the sediments (HILLAIRE-MARCEL & CAUSSE, 1989) and gives a minimum level for paleo-watertables If the precipitation occurs in the epiphreatic zone this might result in stratigraphically inverted ages under falling watertable conditions. Bone and shell fragments predate the last sedimentation cycle; the relation between the level of deposition and the watertable depends on the type of transport mechanism. Deposition of cap mud is related to stagnant phreatic/epiphreatic conditions (WORTIUNGTON, 1991) and dating gives an age to a minimum level of the watertable We have investigated the possibility of using ca l careous algae from cave walls as datable (minimum) watertable indicators. They have the advantage over other biogenic deposits in caves (i.e. bones and shells) that they are sessile organisms and often preserved in situ. Calcareous algae (buildups) have been used as biological indicators of sea level fluctuations by several authors, e.g. LABOREL et al. (1994) 14 C dated of Lilhophyllum lichenoides, and estimated sea level fluctuations during the last 4500 years in southwest France and Corsica, BOYD (1986) interpreted paleoenvironment from the occurrence of Lilhothamnium sp. rhodoliths in late Quaternary raised coastal sediments, Irvine, Ayrshire, AKPAN and FARROW (1984) estimated depth of deposition and SARTORETIO et al. (1996) "C dated several calcareous algae species, which they correlate with bathymetric position Age estimates of high sea level stands can be compared with paleo watertables in the adjacent cave system. This is a way of connecting paleo watertables to sea level and also a test of what processes that governed changes in former watertables. Alternatives to high sea level stands are bedrock control, ice contact damming and local sediment-choked passages/springs. Uncertain factors associated with the use of calcareous algae in the estimation of sea level are at what depth the algae grew and whether dating is possi bl e with known methods. The depth of growth depends on several factors, and the growth depth today in an area is not necessarily r epresentative for former periods. Temperature and light are the main controls, determining depth and geogra p hical distri b ution of calcareous algae (ADEY, 1970) One example is the species Lithothamnium glaciate (Kjellman) that i s foun d in n orthern Norway and in fjords in southern Norway (fig. l) It is common in the sublittoral zone but is also found in the lower littora l zone in shady pools and runnels (RUE ESS, 1 977; IRVINE & CHAMBERLAIN, 1994). F i g 1 L ithothamn i 11m s p ., aft e r Oltmann s (1 9 22). The upper growth limit is controlled by sea level but also competition between different floral-species (light conditions), substrate an d water quality. The lower growth limit is mainly determined by light inte n sity, in Norwegian waters a l gae growth is seldom present deeper than 40 50 m (RUENESS, 1977) In the case of twilight zone caves, encrusting calcareous algae may live near the low tide level due to the low light intensity. Site and sample descriptions a ) Kjep s vlk ca v e s The caves of Kj0psvik (fig. 2), northern Norway, have been studied intensive l y late l y b y LAURITSE (in prep ), NESE Symposium 7: Physical Speleology 4 9


(1996), NESE & LAURITZEN (1996), LAURITSEN & LAURITZEN (1996) and LAURITZEN & LAURITSEN (I 995) Nygrotta (fig 3) is a sediment-choked paleospring situated 60 m a .s l. and was until recently c o vered beneath a talus slope The cave entrance is filled with bones and large amounts of mollusc shells The bedrock walls around the entrance is covered with calcareous c rusts consisting of white calcareous algae. The main cave Storsteinshola, which once drained to the paleospring possesses several distinct paleo-watertables and also large paragenetic features developed up to a level corresponding to Nygrotta. 0 0 Jan Mayen 00 Fig. 2 M ap of tll e S v albard arc hi pe la go a 11d Sca 11 di11avia, arrows i11 d ic a te l ocatio 11 of tll e s i tes di sc r i b e d i11 the t exl b) Kapp Ekholm, Sval bard Kapp Ekholm is the only known section in the central part of the Svalbard archipelago (fig 2) that has several till beds interlayered with marine sediments (MANGERUD & SVENDSE 1992) and is therefore important for the interpretation of the glaciation history of Svalbard and the NW Barents Sea The oldest marine unit is thought to be Eemian and the chronology is mainly based on luminescence and radiocarbon dating together with amino acid estimates and biostratigraphy The chronology from MANGERUD and SV E NDSEN (1992) is used as reference when evaluating the obtained uranium series age estimates. Calcareous algae are found in the Eemian, Kapp Ekholm interstadial (ea 40-60 ka) and the Holocene sediments. Several types of calcareous algae (Lithothamnium) have been analysed from these three units, and we use their uranium series systematics as examples (LING E, 1996) c) Ge n e r a l sam p le d esc ription Living calcareous algae are pink but fossil specimens are white or light red brown Both encrusting and free living (rhodoliths) species have been analysed. The crust of white fossil algae is normally less then I O mm thick and less than 3 mm for the red brown type Encrusting species may envelope stones, shells and other calcareous algae. Microstructures has been studied with SEM. XRD analysis show calcite with small amount s o f aragonite Analytical results Lithothamnium sp. precipitates calcite and small amounts o f aragonite and usually have a high content of magnesium and iron. The reliability of dates obtained from calcareous algae is dependent of in what way uranium is incorporated in the calcareous crusts and which processes affect the material after burial/ subaerial exponation. It is therefore necessary to establish criteria for reliability and to test samples with known ages Analysis of recent ca l careous algae (from Krager0 south Norway) have a uranium content of 1 5 ppm and a 2 :UV f 38 U activity ratio of 1.15 03, indicating that the organisms have a high in viva U-content which is in isotopic equilibrium with seawater. The study of Holocene Weichselian and Eemian calcareous algae from Kapp Ekholm, Svalbard, shows a generally high uranium content and a l arge spread in 234 U t2" U activity ratio values Nevertheless there is no distinct trend in the 21 "u /2 38 U values. Together with the observed deviation from the global marine 234 U/ 238 U activity ratio this indicates that the algae represents open systems with post mortem uranium uptake from the pore water in the surrounding sediments Table 1 s hows the obtained results Investigation of the individual samples with SEM reveals traces of early biogenic cementation, but interpretation of the microstructures is difficult due to the Jack of species control. With the observed enrichment of uranium compared with recent calcareous algae, it is expected that the ca l culated U / Th ages would be younger than the "true ages Discussion and preliminary conclusion Calcareous algae from Kapp Ekholm yields overestimated ages compared with the chronology developed from MANGERUD and SVENDSEN (1992) (fig 4). This can be explained with loss of uranium w i th time, but contradicts the observed enrichment relative to the modem algae. The uranium content differs strongly between samples and subsamples and it is therefore a need for evaluating the material. Full evaluation requires a larger set of data than present. Post mortem uranium uptake is evident from the 2 1 "u / 23 "u activity ratio and is most probably associated with organic matrix in the calcite ske l eton. Loss of uranium is due to normal weathering processes. The material often displays early biogenic cementation (ALEXANDERSSON 1974; MOBERLY 1968 1970 1973) which allow testing of the degree of preservation This biogenic cementation can be observed with SEM and is usua ll y believed to be of short duration in normal calcite undersaturated marine environment. Here sho r eline displacement and burial of the material comes convenient. The residence time the in marine environment may have been Jess than a few thousand years and hence the cementation is preserved. If the material can be dated safely, age estimates of former high sea stands during deglaciation can be obtained Twilight zone caves implies severe light conditions for growth of calcareous red algae, and growth might only have been possible close to the low tide level. The difference between high and low tide can be considered to be site specific and time independent. Linking of former sea level with paleo-watertables might therefore be both possible and useful combining both speleogenesis and deglaciation history of an area 5 0 P roc eed i ngs of the 12 th International Congr ess of Speleol ogy, 1997 Sw i tzerland V ol ume 1


Tab. 1 Uranium series results from calcareous algae from Kapp Ekholm, Svalbard. As shown in figure 4, the estimated ages are older than the assumed formation ages at the site This discrepancy might be due to loss of uranium, but the general 'Jt '38 hiJ!h uranium content toJ;!ether with the measured Ui U ratio indicates uranium uptake 'ram a meteoric water source. age sample no. lab. no. U-content (ppm) 2J4U/238U 23Thf234u 23ThfmTh calculated age (ka) Holocene 1988-523 1067 2.17 .04 1.19 .02 0.125 008 21.0 14.39 97 1988-318 1182 8.55 .10 1.62 .01 0.106 .006 >1000 12 03 .68 1988-714 1016 6.72 .11 1.79 .02 0.515 .014 >1000 73 77 .77 Kapp 1988-712 1169 10.4 .1 1.50 .01 0.561 009 >1000 84.42 .02 Ekholm 1988-713 1185 3.00.08 1.42 .04 interstadial 1988-094 1324 8 12.11 1 88 02 0.239.011 >1000 28.98 .54 (40-60 ka) 1988-713 1325 2.99 05 1 37 03 0.443 .017 >1000 61.67 .16 redeposited 1988-681 1015 26.7 +1.2 1.17 .02 0 784 063 450 155 95 .50 from 1988-681 1170 12.4 .2 1.29 .01 0.949 015 >10 00 245.01 .32 Eeemian 1988-681 1183 31 8 4 1.22 01 0 817 .007 >1000 168.31 .25 1988-679 1068 17 9 .3 1.57 01 1.400 .031 >1000 >350 1988-108 1184 17 0 .4 1.57 02 0.968 .042 >1000 233 70 91 Eemian 1988-688 1186 2.57 .05 1.55 03 1988-524 1326 4 14 .20 1.80 09 1.066 .262 >1000 >350 1988-108 1346 18.8 .4 1.53 .02 0 974 .021 >1000 241.27 .84 after the chronology published by MANGERUD and SVENDSEN (1992) 350 300 250 -;.c .!! .. E 200 ~ c 150 2 e formation F :, 100 ( Kapp Ekholm intcntadiaJ } 50 ~ --1324 0 Formation H (Holocene ) 0 20 40 60 80 assumed ages (ka) 100 =!= -l1 Fonnarion 8 (Ecmian ) 120 140 Fig. 4 Uranium series ages from calcareous algae from Kapp Ekltolm plotted againstformation ages from MANGERUD and SVENDSEN (1992). Tlte solid line represellls colterem ages. Algae crusts cover the walls and ceiling approximately one meter beyond the Nygrotta cave entrance and also large parts of the marble walls adjacent to the entrance Dating of the calcareous algae from this site are in process and will be presented at the symposium. Since the algae are relatively young (i.e. from the deglaciation period) and most probably well preserved due to the recently removed talus cover, precise uranium series ages might be obtained. The problem of possible overestimated ages calculated from Kapp Ekholm might also be absent. Hence, precise calendar ages of a former high sea stand at approximately 60 m a.s.l. will give the age of the corresponding paleo watertable in the Storsteinshola cave system in Kj0psvik. ma s l. 75 50 0 NYGROTTA paleospring 50m Fig. 5 Vertical cross section of Nygrotta. Box form indicates zone of algae growtlt. Modified from Nese (1996). References AKPAN, E.B. & FARROW, G.E. 1984. Depth of deposition of early Holocene raised sediments at Irvine dedused from algal borings in mollusc shells Scou. J. Geol. 20 (2): 237-247. ALEXANDERSSON, T. 1974 Carbonate cementation in coralline algal nodules in the Skagerrak, North Sea : biochemical precipitation in undersaturated waters. Journal of Sedimentary Petrology 44 (I): 7-26. BOYD, W.E. 1986. Fossil Lithothamnium (calcareous algae) rhodoliths from late Quaternary raised coastal sediments, Irvine, Ayrshire. Scott. J. Geol. 22 (2): 165-177. HILLAIRE-MARCEL, C. & CAUSSE, C. 1989. Chronologie Th/U des concretions calcaires des varves du lac glaciaire de Deschaillons (Wisconsinien inferieur). Canadian Journal of Earth Science 26: l 041-1052. Symposium 7: Physical Speleology 51


IRVINE, L.M. & CHAMBERLAIN, Y M. 1994 Seaweeds of the British Isles, Val.I Rhodophyta, Part 2B Corallinales, Hildenbrandiales. The Natural History Museum, London. 276 p. LABOREL, J., MORHANGE, C LAFONT, R., LE CAMPIO J LABOREL-DEGUEN, F. & SARTORETTO, S. 1994. Biological evidence of sea-level rise during the last 4500 years on the rocky coasts of continental southwestem France and Corsica Marine Geology 120 : 203-223 LAURITSEN, A (in prep ). Geomorfologisk utvikling av grottesystemene i Kj0psvik, Nordland. Cand.scient.-thesis. Department of Geology, Bergen University LAURITSE A & LAURITZEN S.-E 1996 Quaternary cave and landform development in the Tysfjord region, north Norway. Karst Waters Institute Special Publication 2 : 73-77 LAURITZEN, S -E & LAURITSEN A. 1995 Differential diagnosis of paragenetic and vadose canyons. Cave and Karst Science 21 :, 55-59 LINGE H C. 1996 Uranseriedatering av b i valver, kalkalger og torv fra noen arktiske lokaliteter Cand.scient.-thesis, Department of Geology, Bergen University 125 p MANGERUD, J. & SVENDSEN J.I. 1992 The Last Interglacial Glacial Period on Spitsbergen, Svalbard Quaternary Science Reviews 11 : 633-664 MOBERLY, R.JR. 1968. Composition of magnesian calcites of algae and pelecypods by electron microprobe analysis. Sedimentology 11: 61-82. MOBERLY R.JR. 1970. Microprobe study of diagenesis in calcareous algae. Sedimentology 14: 113-123. MOBERLY, R.JR 1973. Rapid chamber-filling growth of marine aragonite and Mg-calcite. Journal of Sedimentary Petrology 43 (3) : 634-635. NESE, H 1996. Sedimentologisk utvikling av grottesystemene i Kj0psvik, Nordland. Cand.scient.-thesis. Department of Geology, Bergen University, 157 p. NESE, H. & LAURITZEN, S.-E. 1996. Quaternary stratigraphy of the Storsteinshola cave system, Kj0psvik, north Norway Karst Waters Instilllte Special Publication 2: 116-120 OLTMANNS, F. 1922. Morphologie und Biologie der Algeo. Verlag van Gustav Fisher, Jena, 439 p RUENESS, J. 1977 Norsk algeflora. p. 1-23 and 54-63. Universitetsforlaget, Oslo. SARTORETTO, S., VERLAQUE, M & LABOREL, J. 1996 Age of settlement and accumulation rate of submarine "coralligene" (-10 to -60 m) of the northwestem Mediterranean Sea; relation to Holocene rise in sea level. Marine Geology 130 : 317-331. WORTIIINGTO S.R.H. 1991. Karst hydrology of the Candian Rocky Mountains Ph D.-thesis, McMaster University, 370 p. 52 Proceedings of the 12 th Internat i onal Congress of Speleo l ogy, 1997 Sw i tzerland Vo l ume 1


Useries dating and stable isotope analysis of some last interglacial speleothems from north Norway Ida Malene Berstad, Solvi Einevoll and Stein Erik Lauritzen Department of Geology, Bergen University Allegt. 41, 5007 Bergen, Norway Abstract Stalagmites from a karst cave in north Norway were dated by the U-series technique and analysed for isotopic equili"brium. The purpose of the study is to use stable isotopes for paleoenviroment reconstruction for North Norway, emphasizing Last Interglacial material. This study is in progress, so only preliminary results are reported here. Introduction Provided that the speleothem grew in isotopic equilibrium with the feeding dripwater, the 6 18 0 signal along the growth axis is a paleoclimatic proxy. Combined with U-series dating, a time series of paleoprecipitation changes can be made (Schwarcz 1996). This study aims at the last interglacial time-window Fig.I.Key map to the i11vestigated area. c.,. Strcamsink Cave developed within marble band in synfonn Fig. 2. Topographic and geologic setting of the investigated cave. Background The cave from which the speleothems were sampled, is located at Mo i Rana, at the Arctic Circle (66' N) (fig I) The thin band of steeply dipping marble was almost totally occupied by the 1,3 km long and 104m deep cave (fig 2) A sinking stream runs through the marble band, but the percolation water which precipitate speleothems has to penetrate trough l 00 m thick strata of mica schist and iron oxide ores on its way down to the marble band. The cave has previously been studied with respect to percolation water hydrology, cave microclimate, and stable isotopes (Einevoll & Lauritzen 1994). In addition, an almost complete time-series of Holocene 6 18 0 changes has been made. This 6 18 0 curve was calibrated against temperature changes during the last 250 years, and transformed into a temperature curve for the last 8500 years (Lauritzen 1996). Material Previous sampling by Einevoll & Gard provide material for screening with respect to U-series age and isotopic systematics. After preliminary screening, three specimens were selected, dated and tested in detail for isotopic equilibrium. According to the Hendy (1979) criteria, the isotopic composition within a single growth layer band must be practically invariant with respect to 6 18 0 and at the same time, 6"0 and 6"C values within the same layer must be uncorrelated. Previously, speleothem dates up to beyond 350 ka (the limit of aparticle dating ) have been found in the cave. The problem was to identify specimens that grew in isotopic equilibrium (i.e. Hendy, 1971 ), and at the same time covers the time window of the last interglacial. One sample which were dated to be from the last interglacial, was tested for 6 18 0 variation trough a cross section. "o 6 < Hendy test of speleothem calcite Travers 1 sample 210892 d .7 /""-. ' I ....- 0 ;' __,-o~O ,_,,,,,,, ......... ___ ___ .,,..... 0---0 O .9 o 2 & a 10 12 14 1& 18 20 22 -& cm Fig. 3 Test of isotopic equilibrium. Symposium 7: Physica l Speleology 53


Sample ULBNo U-cons. nnm i14U/ """U L>VTh/ '"U L>VTh/ '"T h Age (ka) +/1 cr (ka) 210892-1 top 1507 0,86 1,30 0,671 370 113,8 +5,83 /-5,55 210892-1 1508 0 83 1.35 0.712 664 124 8 +8.24 /-7.72 210893-1 1673 1.01 1.32 0 743 457 135.5 +5.23 /-5.01 210892-1 1674 1.26 1.34 0.727 998 129 9 +4.97 /-4.74 210892.1 1675 1.31 1.33 0.710 399 124.8 +4.74 /-4.56 210892-1 1555 1.92 1.38 0.748 48 136.0 + I0 .91 /-10.01 210892-1 bottom 1556 2,91 1,45 0,774 358 143,6 +l l ,4 1 /-10,44 Table. 1. Useries age of sample 210892-1 i11 Stratigraphic order 6 11 0 curve and age, sample 210892-1 0 f!~ ~~--j I :1-1 1-~-i -~ i :-:. ~.: Prlsent day value 2 .. .., 4 la a, ;: .! 6 e a, .. ., 0 1 warm 8 I I 1-i I I s E 0 ~_.-I 10.:: E u Fig. 4a. The i11vestigated stalagmite Results and discussion a. Useries dates aparticle dating { 23 "Th! 13 'U) of sample 210892-1 yielded the growth-rate curve shown in Fig. 4b. The sample grew through Termination II through to at least isotope stage Sc b. Stable isotopestest of isotopic equilibrium. The sample was tested for isotopic equilibrium according to the Hendy criteria, 6"0 along single growth bands varies by less than 0.5%0 in the vicinity of the drip-point. Also, the 6 18 0 vie. 6"C varies unsystematically and lacks correlation (Fig. 3.) We therefore infer that stratigraphic sampling along the two traverses shown in Fig. 4a, will represent calcite growth in isotopic equilibrium with the dripwater. c. d o variatio11 trough time Present day isotope composition of stalactite tips (6 18 0 = -7.33 .07o/oo) and calibration constants of the Holocene record presented elsewhere (Lauritzen 1996) both indicate that ~T and ~6 18 0 are negatively correlated, i.e shifts to lighter isotopic values indicate changes to warmer conditions. The sequence shows a distinct warming through Termination II, with indication of the two step deglaciation event {ZeifenKattegat, Seidenkrantz et al. 1996) which is also seen in the Norwegian FM-2 stalagmite (Lauritzen 1995). The remainder of the sequence suggests that isotope stages 5 and 5d are present (and for the better resolved traverse a) also possibly stages Sc and Sb are present. 12 11 -3 -7 6 0 cross section a Fig. 4b. Stable isotope variatio11 through two cross sectio11s combi11ed wit/, age calculatio11s i11 the same stalagmite. References EINEVOLL, S AND LAURITZEN, S.E 1994 Calibration of stable isotope and temperature signal in the percolation zone of a sub-arctic cave, northern Norway. Cave and karst Science 21,9 HENDY, 1971 The isotopic geochemistry of speleothemes. 1 The calculation of the effects of different modes of formation on the isotopic composition of speleothems and their applicability as paleoclimaic indicators. Geochemica Cosmochimica Acta 35 : 801-824 LAURITZEN; S:E 1995. High-Resolution Paleotemprature Proxy Record for the Last Interglacion Based on Norwegian Speleothemes. Quartemary Res/arch 43: 133-146 LAURITZEN, S.E 1996. Calibration of speleothem stable isotope signal against historical data: a Holocene temperature curve for north Norway?, Karst Waters Institute Special Publication 2: 78-80 SCHWARCZ, H.P. 1996, Paleoclimate inferences from stable isotopic studies of speleothem Karst Waters Institute Special Publication 2: 145-147 SEIDENKRANZ, M.S. AND NINE OTHERS 1996 Two-Step Deglaciation at the Oxygen Isotope Stage 6 / Se Transition: the Zeifen-Kattegat Climate Oscillation Qartemary Science Reviews, Vol.15, pp. 63-75 54 Proceedings of the 12 th International Congress of Speleology, 1997, Switzerland Volume 1


Speleochronology, stable isotopes and laminae analysis of stalagmites from southern Africa by K. Holmgren I1 W. Karlen 21 S. E. Lauritzen I >, J. Lee-Thorp 1 > T. C. Partridge 41 P.A. Shaw 5I and P. D. Tyson' l) Department of Geology, University of Bergen, Allegt. 41, 5007 Bergen, Norway 2) Department of Physical Geography, Stockholm University, S-10691 Stockholm, Sweden 3) Department of Archaeology, University of Cape Town, Private Bag, Rondebosch, Cape 7700, South Africa 4) Climatology Research Group, University of the Witwatersrand, Wits 2050, Johannesburg, South Africa 5) Department of Environmental Science, University of Luton, Park Square, Luton LUI 3JU, UK Abstract Holocene and Pleistocene speleothems from South Africa, Botswana and Tanzania have been uranium-series dated and analysed on their stable isotopic content. Most of the stalagmites sampled are composed of finely laminated dense calcium carbonate. The structure and periodicity of the growth laminae are presently being analysed. ew results from this study are presented and compared with data already available. 1. Introduction Speleothems are often suitable climate archives and provide an option to obtain terrestrial high-resolution palaeoclimatic records from all over the world (SCHWARCZ, 1986). During the last twenty years severa l highly resolved palaeorecords from ice cores from Antarctica and Greenland have been presented and inferences have been drawn regarding global Late Quaternary climatic changes, global symmetry and driving forces (LOWELL et al., 1995). There is an urgent need to obtain good quality terrestrial records of equally high resolution from low latitudes in order to test hypotheses about global climate changes and forcing factors. It is hoped that this recently initiated project will contribute such information through detailed analysis of speleothems from Tanzania, Botswana and South Africa. 20 30 1020 30 Figure 1: Map of sout h em Africa sho111ing the locations of caves mentioned in the text 2. Tanzania Tanzania is situated just south of the equator and has a tropical climate influenced by the seasonal latitudinal movements of the [ntertropical Convergence Zone and the monsoonal winds. Previous theories that climate in the tropics remained stable during the last ice ages have been proven incorrect (GUILDERSON et al ., 1994). Very few terrestrial high-resolution climate data have so far been presented from this region, but studies from caves from Somalia and Zaire (BROOK et al., 1990) glacier studies from Mt Kenya (KARLEN & ROSQVIST, 1988; MAHA EY, 1990), and lake studies from Kenya and Tanzania (BONNEFILLE et al., 1995) generally suggest wetter conditions during the Eemian interglacial, aridity and cooler temperatures during the last deglaciation, wetter conditions at the Pleistocene-Holocene transition, a climate optimum in the early Holocene and since then a trend toward drier climates We have sampled stalagmites from cave systems along the Tanzanian coast; the Amboni Caves and the Matumbi Caves, and from caves on a small coral island outside the coast; the Songo Songo Island (figure l ). Uranium-series dating so far has yielded ages from recent to the Last Glacial Maximum and from 38 to 34 ka (1000 years) (figure 2). The stalagmites are well laminated, with growth rates varying from 0.5 mm per I 00 years for the Holocene samples, to 7 mm per 100 years for one that grew between 38 and 34 ka. Considering the high growth rate on the latter it i s expected that a yearly resolution will be obtained from the laminae analysis and a 10 year resolution from the stable isotope analysis. This work is in progress. 3. Botswana and South Africa The speleothems studied from Botswana and South Africa are from cave systems in semi-arid central southern Africa (figure I) This region is of interest due to its transitional position between the summer rainfall zone to the north, the winter rainfall zone to the south, the semi-arid Kalahari Basin to the west and the more humid plateau to the east. Clearly, the climate and the borders of these zones have varied in response to past global climate changes. The limitations of most of the previous palaeoenvironmental data available are poor chronological control, low time resolution, and poor continuity (THOMAS & SHAW, 1991; PARTRIDGE 1995 ). Uranium-series dating, close-interval stable oxygen and carbon isotope measurements (8 1 k O, 8 13 C) and petrologic analyses carried out on one 33-cm-high stalagmite were reported Symposium 7 : Physical Speleology 55


to provide a palaeoclimatic and palaeoecologic record for the period 50-20 ka ( HOLMGR EN, 1995 ; HOLMGREN e t al. 1995) ( figure 2 ). The 8 18 0 record was interpreted as reflecting relative temperature variation s, and comparing it with data from other palaeoenvironmental sites in central southern Africa a regional decrease in temperatures from 50 ka to a minimum during the Last Glacial Maximum was confirmed The 8 13 C record in s peleothems i s governed by several factors among which is the relative influence of C 3 and c vegetation in the region (S C HWAR C Z 1986) This distribution is climatically controlled As a working hypothesis we propose that the Lobatse 8 13 C record reflects regional shifts in vegetation characteristics supporting a model that invol v es an enhancement of the westerlie s and an equatorward displacement of the subtropical convergence which leads to increased influence of the winter rainfall belt in the Southern Kalahari and to drier conditions in the Middle and Northern Kalahari ( COCKROFT e t al ., 1987) Subsequent dating and s table isotope analysis of core s from a 1 5-m-high stalagmite LII2 confirm parts of previous interpretations and contradicts others emphasizing the importance of cross-checking Further work on material from Lobatse to test t h e validity of the data is in progress Speleothem samples have recently been collected from Ficus Cave and Cold Air Cave in Transvaa l in South Africa These caves lie in the same climate region and the same bedrock formation as Lobatse Caves. Hence, the results from studies of speleothems from these caves are expected to test the robustness and regional validity of the Lobatse data The first uranium-series datings of the South African speleothems have yielded ages from Late Holocene to the present from the end of the last glacial and from isotope stage 5a (figu r e 2) Thus, material is available both for comparison with the Lobatse record and for extending the record in time. The young Holocene-to-recent stalagmite has grown at a rate of 15 mm per I 00 years and will be tested against the meteorologica l record 6 II) Cl Botswana C 5 :;; IV South Africa "C II) G) 4 II Tanzania ;:: G) 3 II) I :::::, .... 0 2 ... G) .c E :::, z 0 Ill 0 Ill 0 Ill 0 0 0 0 0 0 o N et"? $ (0 r-a, Ji o Ill o Ill 0 o o o N N I") Ill (0 r-CX) Ages (ka) Figure 2: Age frequency di st ribution of spe leothems from Bot swa na South Africa and Tanzania. 4. Discussion Speleothems are one of few potential archives for preserving palaeoenvironmental data in semi-arid regions their validity is not yet fully explored. However no archives are perfect and several problems are connected with semi-arid and tropical speleothems. They often possess parts that have suffered from re solution this can cause open system conditions and invalidate both the age dating and the stable isotope analyses They can be composed of aragonite or calcite or be a mixture of both It i s important to decide whether the structure wheter calcite or aragonite, is of primary or secondary origin before an y firm conclusio n s can be drawn regarding past environmental conditions These problems are best elucidated and overcome b y appropriate selection of samples by analysing several speleothems from the same locality and by performing a variety of analyses upon them. With this approach there may be a conflict between scientific interests and conservative intere s t s. However thanks to development of new techniques smaller samples are needed. 5. Conclusions Uranium-series dating of speleothems from Tanzania Botswana and South Africa has yielded ages from 90 ka to the present. Most of the speleothems analysed are compo s ed of compact laminated pure calcium carbonate The prospect of obtaining high-resolution palaeoclimatic records from the on going analyses is most e n couraging References BONNEFILLE R .; RIOLLET G. ; BUCHET G .; !COL E, M .; L A FO N T R .; ARNOLD M & D JOLLY I 995 Glacial / Interglacial Re c ord from lntertropical Africa High-Resolution Pollen and Carbon Data at Rusaka Burundi Quat Sci R ev. 14 : 917 936 BROOK. G A .; BURNEY D A & J B COW A RT 1990 De s ert paleoenvironmental data from cave speleothems with examples from the Chihuahuan, Somali-Chalbi and Kalahari deserts Pala eogeo gr .. Pala e oclim. Pala e oe c ol 76: 311-329 C OC KROFT M J .; WILKINSO N, M J & P D TY S O N. 1987 The application of a present-day climatic model to the late Quaternary in southern Africa Clim Chang e 10 : 161-191. GUILDERSON T P .; FAIRBANKS R G & J L. R U B E NSTO NE. 1994 Tropical Temperature Variations S i nce 20 000 Year s Ago : Modulatin g lnterhemispheric Climate Change S cie n ce 263 : 663 665 HOLMGREN K 1995 Late Pleistocene Climatic and En v ironmental Changes in Central Southern Africa Ph D ., The Department of Ph y sical Geography Stockholm University HOLMGREN K. ; KARLEN W & P A SH AW. 1995 Paleoclimatic significance of the stable isotopic composition and petrology of a late Pleistocene stalagmite from Botswana Quat R es. 4 3: 320328. KARL E N W & G ROSQVIST 1988 Glacier fluctuation s rec o rded in lacu s trine sediments on Mount Kenya N at G eog r. R es. 4 : 2 19232. LOWELL, T V .; H E USSER, C J. ; A N DERS EN, B G .; M O RE NO, P I. ; HA U SER, A ; HE U SSER L E .; SCHL 0 CHT E R, C .; M A R C HANT D R & G H DE N TON 1995 lnterhemispheric correlation of Late Plei s tocene glacial events. S c i e n ce 269 : 1541-1549 M A HANEY, W C 1990 Ice on the equator : Quaternary geo log y o f Mt Ken y a East Africa Wm Caxton Ltd Sister bay Wiscon s in P A RTRIDG E, T C 1995 Palaeoclimates of the arid and s e mia rid zone s of southern Africa during the last climatic cycle M e m S oc. geo l Fran ce 167 : 77-83 S C HWARCZ H P 1986 Geochronology and isotope ge o chemistry in speleothems In: (P Fritz & J Fontes, eds ) : Handbook of Environmental Isotope Geochemistry Elsevier Amsterdam : 271-30 3. THOMAS D S G & P A SHAW 1991. The Kalahari Environment Cambridge University Press 284 p 56 Proceedings of the 12 th Internat i onal Congress of Speleo l ogy 1997, Sw i tzerland Volume 1


Absolute datations of speleothems and its speleomorphological significance from Divaska jama and Jazbina caves; Kras plateau, Slovenia Andrej Mihevc Stein-Erik Lauritzen** *Institut za raziskovanje krasa ZRC SAZU, Titov trg 2 66230 Postojna Slovenia **In st itute of Geology, University of Bergen A llegaten 41, N-5007 Bergen, Norway Abstract Various speleothems, including tall stalagmites and massive flowstone deposits from the Kras plateau, Western Slovenia have been sampled and dated with the 23 Th r 'U method by a-particle counting Results of datations on speleothems in two caves, Divaska jama cave and Jazbina v Rovnah are presented. Sampling focused on the morphological evidence of the late Pleistocene and morphological impact of colder climate in caves In Divaska jama three periods of growth of flowstone were recorded The oldest flowstone was deposited between two phases of sedimentation of flood loam, before 350 Ka, and are out of range of dating method Intensive growth of thick large stalagmites is recordet at 240 -170 Ka. The newest deposition started on gravel after last glaciation, around 16 Ka B.P In Jazbina v Rovnah cave some flowstone crusts were deposited at about 240 Ka B.P. Flowstone was later fractured scattered and arranged in a circle due to cryoturbation. On that stalagmite start to grow at 43 Ka Growth was continuos till present, showing change of flowstone colour at 14,9 ,9 Ka. Another stalagmite analysed, started to grow at 16 ,6 Ka, having hiatuses in the bottom part, most distinct at 10,6 ,2 9,3 ,2 Ka, probably showing climatic changes of younger Dryas. Dating allows us to connect some morphological evidence from caves with climatic changes of the late Pleistocene Introduction Kras is about 50 km long and to 20 km wide karst surface, rising from elevations about 250 m on NW to elevations about 600 m in SE part Figure]: Position of two caves in tl,e W part of Slovenia. The karstification of mostly Cretaceous limestone started after its uplift in the Oligocene. There is about 300 m of the vadoze zone accessible and there are caves formed in all elevations from the surface to the sea level and below. Some caves have been filled with allogenic sediments, originating from the impermeable surroundings which were later partly eroded Different allochtonous sediment are showing different phases in the development of the karst but give no absolute datations of them. There are few absolute dates of speleothems available for the Kras. First datations, with 14 C were done by Gospodarit (1980) in Skocjanske jame and Vilenica cave, these were followed with U series dating on a stalagmite from Grotta Gigante by Cucchi and Forti (1989), and by Zupan (1991), datations from Skocjanske jame Lipiska jama cave, cave Vilenica and cave Mejame, but no more than 8 speleothems were dated by both methods. Here we present some results of the speleothem U series dating from Jazbina cave and Divaska jama Samples were dissolved in concentrated HNO A 218 Th/ 232 U spike was added as an internal standard for control of the chemical yields U and Th nuclides were separated by ion exchange chromatography in columns After purification steps both fractions were electroplated on steel disks, and counted to measure alpha particle activity. Ages were calculated from isotope activity ratios using standard algorithms (Ivanovich & Harmon 1982 1992 ; Lauritzen, 1991) The error is based on counting statistics For the samples contaminated with detritic 230 Th correction was done on the equation suggested by Schwarcz (1980) with correction factor B 0 = 1.5 Samples were prepared and dated in the U-series laboratory in Bergen. These results are preliminary, as there are not all the datations from larger area evaluated yet Work was conducted through collaboration between the Department of Geology, University of Bergen and the Institute of Karst Research from Postojna in a frame of PEP Ill, The Afro European Transect (PANASH) and with financial support of University in Bergen and Slovene Ministry for Science and Technology. Description of sampling points and samples Jazblna v Rovnah is 2000 m long and 130 m deep cave with the entrance at an altitude of 483 m The cave is reaching the lowest point in this part of the karst, 350 m respectively Cave is a three dimensional maze of phreatic channels and there is little flood loam, characteristic of many other caves in vicinity Samples, already broken stalagmites from two parts of the cave were taken A 238 cm long stalagmite C 228 grew in the depth of about 125 m, on the unstable debris The parts of broken column were collected and the foot of column removed from the rock The speleothem was split unhalves, showing, that it was composed of 224 cm tall stalagmite and a straw which joined the Sympos i um 7 : Ph y s i ca l Speleo logy 57


stalagmite and after that started to grow thicker. Thickness of the stalagmite varied from 7 cm at bottom to maximum 11 cm at 70 cm. Several subsamples of white or transparent flowstone were cut of the stalagmite and dated giving good results due to low 232 Th contamination and high uranium content. Stalagmite start to grow at 16 ,6 Ka (sample number 1396). Bottom parts show two distinctive hiatuses The more important is the second, at 17 cm, where two subsamples were taken. As the samples weight more than 30 grams, their centres were 3 cm apart. Lower (1288), was dated to 10 2 Ka, the one above hiatus (1385) to 9 2 Ka. The hiatus is distinctive, marked with thin red layer, below it flowstone is compact and translucent, above it turns to white, more thicker lamina Stalagmite became thicker too, which may be connected with increased water inflow. Growth was then disturbed several times, but only one hiatus at 151 cm is marked with thin layer of red soil A subsample was taken at 149 cm (1292), dated to 7,6 2 Ka The subsample at 211 cm (1398), showed age of 5 5 2 Ka All these data give the same growth rate of0 03 mm/y JBRT 1400 1392 1393 J/>2. KS -485m -325m A 1292 C228 1385 Figure 2: Cave Jazbi11a v Rov11al1. Scllematic cross sectio11, positio11 of tlle dated speleotllems a11d subsample 1111mbers 011 tlle speleotllems. The second stalactite was growing in the depth of about 75 m below the entrance. It grew on the loose rocks and broken flowstone crusts, which have been moved and arranged into polygon with flakes of crusts been in near vertical position, obviously due to cryoturbation Both broken parts of the stalagmite and ist base were sampled as well as the parts of the cryoturbated crusts. A sample of the crust dated (1373) gave age of 240 +40 -30 Ka. The result is not very reliable as there is some contamination with detritic Th. Stalactite has a dark brown crystalline laminated flowstone with a distinctive change to white flowstone 2 cm below the top Three samples were analysed showing age (1393) at bottom 43 Ka and (1392), age 14,9 ,9 Ka at 48 cm, below the hiatus. Growth rate of that part of stalagmite was 0.016 mm/y Above the hiatus the flowstone change to white and translucent. The top I cm thick layer of flowstone (1400)was dated to 4,9 .5 Ka The growth rate of this white flowstone is an order of magnitude lower, being only 0 0013 mm/y. DivaJka Jama is 670 m long cave, accessible parts of galleries are to 20 m wide and high The generations of fluvial sediments and flowstone in the cave were studied by Gospodarit (1984), but do absolute datations were made Its galleries are at elevation from 350 410 m, while surface is at elevations of 450 m and the actual underground rivers flowing in the nearest cave at elevation of 156 m a.s.l. (Mihevc 1984). Most extensive sediment in the cave is a more than 30 m thick laminated flood loam which filled up most of the cave. Loam was deposited at least in two phases, separated with a period of erosion and deposition of the flowstone On the top of flood loam sequence in some places angular gravel was deposited. Entrance to the cave is a collapse which split the cave gallery From it to SW the gallery is open towards NE gallery is filled, partly with rock blocks, partly with fluvial sediments, covering a massive flowstone too. 430m Fig. 3: Ctnie DivaJka jania. Scllematic Cross sectio11 of tlle cave a11d positio11 of tlle speleotlrems dated. A sample was taken from the massive flowstone at the collapse entrance of the cave The age of the sample (I 316) was 243 +81-47 Ka The flowstone from a top of small parasitic stalagmite growing on 12 m high, more than 3 m thick fallen stalagmite Harambasa was taken (1375) It showed the age of 176 +17 -15 Ka, which gives the time of its collapse In Zibernova dvorana 39 cm long stalagmite and flowstone crust grown between two phases of flood loam deposition. Subsample (1242) was taken from its base and dated to more than 350 Ka. In Pretnerjeva dvorana a stalagmite grew on the angular pebbles and boulders, which partly cover the upper laminated loam The surface of sediment is eroded making more than of I 0 mrelief. Base of the stalagmite (1243) gave age of 16 I Ka The top of the stalagmite (1244) was also dated, showing age of 7 ,8 Ka The growth rate of the speleothem was low 0 008 58 Pr oc eed in g s of the 1 2 1 h Internat i onal Congre ss of Speleology 1 997, Sw itz e rl a n d Vo l ume 1


mm/y only. The reason for that is probably a high ceiling, which caused the splitting of the droplets and spraying them around a process which a thin crust of flowstone covering the gravel in vicinity. Discussion These speleothems and ages calculated from them are enough to give us some instructions for future work and to show some events in development of the caves studied. Sampling concentrated on the tall stalagmites from caves, which were not influenced by sinking rivers floods or sedimentation during the end of Pleistocene The speleothems growing on such a deposits are in most cases badly contaminated with detritic 212 Tb, so datations are not very reliable. The datations we conducted and presented in this paper show, that the growth of tall stalagmites after last glaciation start at about 16 Ka bp. This ages we found in Jazbina cave at stalagmite C228. At stalagmite JBRT from the same cave, the change in colour of speleothem and lower growth rate after 14.9 ,9Ka can also mark the same climatic change. At stalagmite C228 a evident hiatus in its lower part is possibly marking the Younger Dryas. In Divaska jama same age, 16 I Ka, is expressed in the sample 1243, taken from the base stalagmite which grew on angular pebbles and loam indicating a cold period of washing away the sediments. Kras was not glaciated during the cold periods of Pleistocene Most of it was above the tree line, and the snow line was at the elevation of about 1200 m, but small glaciers were only 20 km away. Low temperatures, poor vegetation and thin soil didn't create good conditions for flowstone deposition in the caves, so in cold periods washing of older sediments from caves prevailed, undermining the flowstone growing on them too Through some entrances cold air could penetrate deep in the underground, shattering the older flowstone crusts or cause flaking of the cave walls or ceilings and causing frost movements. In Jazbina the JBRT stalagmite was growing on such a shattered crust arranged into frost polygon. As the crust was 250 Ka old it was deposited in MIR interglacial. After that, in Riss the crust was broken and cryoturbated Reason for that could very low temperatures of that glaciation or a larger entrance to the cave Later that entrance was closed, making possible the growth of the stalagmite JBRT during cold Wurm climate. Divaska jama has complex history In the cave there are two sedimentation phases, but older than the reach of the dating method. In the cave sinter was deposed between the two sedimentation phases before 350 Ka, but the fastest it seems was the deposition in MIR interglacial. The flowstone at what is now entrance to the cave grew at that time. To the same warm period also belongs the large stalagmite which collapsed at 176 Ka, which can be connected with new, Riss cold period which stopped deposition of flowstone and started the washing of the sediments from the base of stalagmite. Collapse of the ceiling and opening of the entrance occurred much later, probably in the Holocene References CUCCHI F ., P FORTI 1991 : The first absolute datation of a speleothem from Trieste Karst Acta carsologica 18 Ljubljana 53-64 GOSPODARJC R. 1985 : 0 Speleogenezi Divaske jame in Trhlovce ., Acta carsologica 13 Ljubljana, 5 32 IVANOVICH M. & HARMON R. S ., 1992 : Uranium series Disequilibrium Aplications to Earth, marine and Environmental sciences. Clarendon Press Oxford 1-910 KRANJC A.; J. KOGOVSEK & S SEBELA (1992) : les Concretionnements de la grotte de Skocjanske (Slovenie) et les changements climatiques. Karst et evolutions climatiques Bordeaux 355-361, LAURITZEN S E 1991: "Age4U2U" Program for Reading ADCAM energy 234 Uranium ages Turbo Pascal Code 5000 lines MlHEVC, A 1984 : Nova spoznanja o Kacni jami Nase Jame 26 Ljubljana 11-20 MlHEVC A 1994 : Contact karst of Brkini hill s. Acta carsologica 23 (1994) pp Ljubljana 100-109 SCHWARCZ H P ., 1980 : Absolute age determination of archaeological sites by uranium dating of travertines Archaeometry 22, 3-4 ZUPAN N 1991 : Flowstone datations in Slovenia. Acta carsologica 20 Ljubljana 187-204 Sampl Sample description Uppm 234U/238u 23 Th/ 234 U 230 Tht2 32 Th Age (Ka) CJ e. Jazbina cave 1396 C228, base of stalagmite 1.56 2 642 0.139 1084 16.1 .6 1288 C228, below hiatus at 17 cm 1.54 2.597 0.094 660.2 10.6 .2 1385 C228, abowe hiatus at 21 cm 1.6 2.48 0.083 10000 9.3 .2 1292 C228 stalagmite at 149 cm 1.8 2 418 0 068 10000 7 6 .2 1398 C228, stalagmite at 211 cm 1 6 2 432 0.050 10000 5.5 .2 1373 JAZ KS, cryoturbated crust 0.364 1.43 0.931 14.08 240 +40-30 1393 JBRT1, base of stalagmite 0.26 1.289 0.332 10000 43.1 1392 JBRT3, stalaQmite at 48,5 cm 0 62 1.095 0 128 10000 14.9 .9 1400 JBRT 4, stalagmite at 49,8 cm 0.586 1 29 0.044 10000 4.9 5 Divaska jama cave 1242 stalaQmite between loam 0.25 1026 1.019 10000 > 350 1243 base of small stalaQmite 0 24 0 976 0.140 258 0 16.42 -1 1244 top of small stalagmite 0 20 0 988 0.066 10000 7.48 .8 -0.8 1375 top of collapsed stalagmite 0.24 1.087 0.827 23 966 176.34 +17 -15 1316 flowstone at the entrance 0 06 1.09 0 928 10.372 243.70 +81 -47 Tabk 1. Uranium serie.f disequilibrium dates of.rpeleothem.f/rom caves Jazbi11a a11d Divaskajama. Symposium 7 : Phy sical Speleology 59


The Isotope Systematics of Perennial Cave Ice in Northwestern Canada W illiam D M a c D ona ld 2 and Charles J. Yonge 1 .1 1 EIUDP, Universitas Sam Ratulangi, PO Box 1357, Manado 95013, Sulawesi Utara, Indonesia. E-mail Yonge@manado.wasantara 2 Alberta Karst Institute, Box 8213, Canmore, Alberta, Canada. Resume Les glaces pennanant de quatre grottes a l'interieur et a l'est du territoire du "Canadian Great Divide" produisent un 6 'R O et un 60 plus positives en comparaison avec les precipitations annuelles significatives dans la region. De plus, dans un graphique 6"0 60, ces donnees se situent en dessous et le long des lignes ayant une pente moins forte que la Ligne d'Eau Meteorlogique Globale Pour expliquer ces phenomenes observes, nous proposons le processus suivant: un fractionnement isotopique entre la vapeur et la glace, qui change le signal isotopique de la vapeur d'eau en saison chaude, quand se condense le givre a l'entree des grottes. La chute du givre sur le sol de la grotte a travers une surcharge mecanique, a laquelle s'ajoute la glace qui provient du suintement de 1 'eau au sol (avec une composition isotopique annuelle significative), resulte en une formation massive de glace de composition multiple. Cette composition multiple est ce qui est observe dans les characteristiques et les liaisons trouvees. De telles decouvertes ont des implications pour !'interpretation climatique des grottes de glaces anciennes, et pourraient aussi contribuer a la comprehension des systemes d'evaporation et de precipitation qui traversent Jes chaines montagneuses. Cave Ice Isotopic Composition Which Does Not Fall on the Global Meteoric Water Line Stable isotope measurements, 6 0 and 60, were made on perennial ice from several ice caves in western orth America Three striking features of the data emerge, especially for cave sites west of the Great Divide: firstly, all 6 0 and 60 points lie on or to the right of the G l obal Meteoric Water Line (GMWL), secondly, those points lie on positive slopes less than the GMWL, and thirdly, the ice is more enriched in the "heavy" isotopes than expected. Caves from east of the Great Divide are compared to those on the coastal, eastern side, the latter exhibiting trends more concordant with the GMWL. The cave ice trend away from the GMWL is in general made up of three components The nearest (and most depleted in O and D) points to the GMWL are from seepage or ground water followed by those from massive floor ice. Finally, hoar ice formed on the walls and ceiling of the cave are most enriched in o and D and lie the furthest off the GMWL. The western Caves exhibit isotope enrichments more than that expected for vapour-water-ice fractionation (an enrichment in the ice of around 3%o in 6 1 0 and around 14%0 in 6D over a water condensate), whereas our hoar frost amples enrichments of up to 8%0 in 6 0 and 60%0 in 6D are observed Therefore, while vapour-ice fractionation must occur, there needs to be an additional enriched moisture source. Water condensed from the cave atmosphere is generally a mean annual average of precipitation falling at the site, and is positively correlated with the site's mean annual temperature, and we might expect the cave ice to be derived from this source except that it is too depleted. Hoar ice therefore appears to be growing mostly from warm moist air entering the cave during the summer months. This warm air being enriched in 1 0 and D over that of the mean annual average, and being able to carry more moisture than cooler air, supplies the additional moisture source required Ice Genesis in the Caves West of the Great Divide Moist, warm air enriched in O and D enters the cave subliming as hoar frost on the walls and ceiling. Over the warm part of the year the hoa r builds up to such an extent that it becomes overloaded and falls to the floor. A l ong with seepage, melting and recrystallization, it reforms as massive floor ice. As a result, this massive ice has a mixed composition between mean annual seepage and warm-season derived hoar ice. For example we determine that l0 C vapour entering the cave and sublimating as hoar ice would yield a mixing line slope less than the GMWL (around 7 1 ), with enrichments of about 8%0 in 6 0 and 55 %0 in 6D which tits our field data very well. Perennial Cave Ice as Proxy Climate Indicators How might these results be applied to paleoclimatology? Serendipity Cave in the Crows Nest Pass contains a 15-meter thick plug of massive ice, where guano trapped at a depth of 14.3 meters yields a carbon-14 date of970 years BP This site clearly offers the opportunity for an extremely detailed study of the last 1000 years or so, assuming that the isotopic composition of the ice can be interpreted climatologically. While the scenarios can be complex, we find that the caves encountering lower mean annual site temperatures contain massive ice that is more enriched in "O and D compared to caves sited in a higher temperature regime It therefore seems to be the extent of hoar ice from warm-season air masses that dominates the changes observed in the composition of massive cave ice. Swings in isotope composition in cave ice sections, interpreted as warming/coo l ing trends, would then be in the opposite direction to conventional glacial or polar cap cores Analysis of ice from Caves west of the Great Divide yields a very different picture; data generally fall close to the GMWL and unusual isotopic enrichments are not observed, actually the reverse. Here, winter snow falling into the cave entrances is preserved in ponded cold air. Recrystallization enhanced by summer rain results in massive ice, but not a complete homogenization of the layers. This means that a muted record of precipitation is recorded and this offers some potential as a proxy paleoclimate indicator. We would interpret swings in isotopic composition in the same way as for polar and glacial ice cores, and thus opposite to records from the eastern divide caves. That is to say that isotope depletion would be associated with a lowering of temperature and vice versa (although we recognize further complexity when considering glacial/interglacial climate transitions). la essence though it appears that we studying two trends; ice derived directly from a vapour source (east divide regime) and ice derived from a water / ice source(west divide regime). Our investigation of perennial cave as proxy climate indicators continues 60 Proceedings of the 12 1 h International Congress of Speleology 1997, Switzerland Volume 1


Enregistrement de l'activite charbonniere dans les speleothemes de Choranche (Vercors, France) par Yves Perrette, JeanJ a cqu es D e l a n n o y, D o m iniq u e Ge n ty, Jean Luc D estombes et Yves Quinif Universite Joseph Fourier Institut de Geographie Alpine Grenoble (France) U.RA. 723 du C N.R.S ., Laboratoire d hydrologie et de geochimie isotopique Orsay (France) U.RA. 249 du C N R S ., La b oratoire de Spectrometrie Hertzienne Litle (France) Centre d Etude et de Recherche Appliquee au Karst, Mons (Belgique) Abstract The study of the summital part of an active speleothem extracted from the Coufin-Chevaline cave (Choranche Vercors Fran c e) permits one to establish a link between the impact of human activities on the plateau des Coulmes karst and crystalline characteristic s. The application of different methods (digital image proce s sing laser reflectance and laser induced luminescence) helps to describe crystalline inclusion s (charcoals and clays) and crystalline fabrics The yearly alternantions allows us to validate the quality o f the speleothem recording. We show how the internal structure of speleothem keeps the mark of charcoal burning developement durin g the XIX' century This study points out the possibilities offered b y speleothems in the precise definition of environemental evolutions R esume L etude de la partie sommitale d une stalagmite acti v e de Coufin-Chevaline (Choranche Vercors France) a pennis d 'e tablir un e relation entre l'impact des occupations humaines sur le plateau karstique des Coulmes et diver se s caracteristique s de l a f abriqu e cristalline L'application de differentes methodes (traitement numerique d image mesure de la reflectance et de la luminescence l as er) a aide a la description fine des differentes fabriques cristallines et inclusions intercristallines (charbons de boi s, argile s ) Le comptage des alternances annuelles continues s ur cette partie a permis de valider la relation entre les modifications cristallines et les a c tivite s humaines notarnment le developpement des c harbonnieres durant le xix "" siecle Cette etude montre les po ss ibilite s qu offrent l es speleothemes dans la definition de s conditions environnementale s s ur un pas de temps historique mots des_: speleothemes !amines impacts anthropiques Vercors France Introduction L interet revele par les concretions endokarstiques speleothemes dans les etudes paleoenvironnementales est aujourd hui reconnu Au travers de leur presence/absence comme au travers des datations isotopiques U/Th au 14 C leur prise en compte est devenue habituelle Depuis le debut des annee s 1990 les informations contenues dans la structure interne des speleothemes sont de plus en plus etudiees ( WmTE W et BRENA N S., 1989 ; MAIRE R ., 1990 ; GENlY D ., 1992 1994 ; RoussEA U, L. 1992 ; BAKER, A et al ., 1993 ; SHoPOv Y -Y. & al. 1990 1991 1994) La meilleure comprehension du transfert des informations environnementales dans cette structure inteme permet aujourd hui d utiliser l outil speleotheme en vue de la connaissance des evolutions environnementales passees L objet de cet article est de presenter la demarche qui a permis de deceler l empreinte laissee par les activites humaines developpees sur le plateau des Coulmes (Vercors, France) depuis la periode de la Renaissance (acception historique) Apres avoir presente brievement les methodes utilisees !ors de cette etude nous poserons les hypotheses qui ant sous-tendu ce travail Enfm par le biais de l etude stratigraphique fme d un echantillon stalagmitique preleve dans la grotte de Coufm-Chevaline (Choranche Vercors France) nous verrons comment la calcite a enregistre evolution environnementale connue de ce massif 1. Presentation du site et de la methode Les differents aspects methodologiques de ce travail qui font l objet d une autre communication (PERRETIB, Y. & al ., 1997) ne seront pas decrits precisement ici 1 1. Site d'ec h antillonnage Le carottage a ete realise verticalement sur un edifice stalagmitique de type tarn-tarn de la salle de la Cathedrale du reseau de Coufm-Chevaline (Choranche Vercors France). Plusieurs raisons ant determinees le choix de cette salle et de cet edifice La raison necessaire a cette etude est l a c cueil et l interet porte a nos travaux par M' Mantovani et son equipe de guides Cette salle est situee dans la partie touristique du reseau speleologique. L edifice est done a proximite de l'escarpement qui forme le cirque de Choranche (fig. 1 ) La decompression importante dans cette partie du massif doit, en ouvrant le reseau de fi s suration laisser circuler des particules diverses qui seront piegees dans le depot carbonat e. L anciennete de cette s alle (DELA NN OY J.-J. & al ., 1988) sa stabilite mecanique permettent de considerer le fonctionnement de !'edifice comme etant stable durant la periode consideree L'edifice carotte est actif ce qui permet un calage chronologique par le haut ". Pour clore cette liste non-exhaustive des raisons ayant guide ce choix on peut rappeler que le plateau des Coulmes est etudie depuis le debut des annees 1980 par des karstologues (DELANNoY J -J. 1 9 81) mais aussi des palynologues (TliIEBA UL T S. 1991) prehistoriens (BIN1Z P, 1980 in DEI..ANNO Y J -J. & al. 1988) etc. L etude relatee ici s est done attachee a la lecture de s informations paleoenvironnementales contenues dans la partie sommitale de l echantillon (fig 1 ) Pour cela different s outils ant ete utilise s en vue de la stratigraphie fme de l'echantillon 1.2. Methode Apres sciage la section verticale a ete polie en vue de analyse stratigraphique Celle-ci s est attachee a la description petrographique de l echantillon en observant tout particulierement le type de fabrique cristalline la coalescence intercristalline la couleur et l opacite de la calcite ain s i que les diverses inclusions piegees dans le reseau cristallin Le s ob s ervations ant ete realisees a la loupe binoculaire avec un grossissement de 4 x a 40 x. De nombreux comptages d altemances ant ete realises ce qui a permis d evaluer l'iige approximatif des differentes informations. Ce comptage a ete manuel dans un premier temps puis assiste par ordinateur selon la methode developpee par GENIY D (1992). S ympo s i um 7: P hy s ic a l Spe l eo logy 61


fuL.ll : S i tuat i on du plateau des Cou l mes : B B o u rne ; PC plateau des Co u lmes. ; leqende / escarpement /4 reseau speleolog1que / S 1te de pr81&vement Choranche I I I N \ Bouma / ligne de f a ile f1fL.1.Q Croqu is du s y steme s peleologique de Coufin-Chevaline Figu r e 1 : Situation de Clw r and1e et site de preleveme n t. Afin de diversifier Jes donnees collecrees !ors de la stratigraphie, differentes methodes spectroscopiques ont ete rnises en oeuvre : mesure de la reflectance laser de la luminescence laser (excitation a A = 450 nm, mesure pour A.> 550 nm). Afin de valider certaines observations, des determinations a la sonde Raman ont ere realisees. L'utilisation de ces outils varies en vue de la stratigraphie fine de l'echantillon a necessite de clarifier Jes apports et lirnites de chacun d'eux dans ce type d'objectifs La demarche ayant conduit a limiter leurs apports, a envisager leur articulation au sein de cette etude fait l'objet d'une communication dans ce meme congres. 2. Hypotheses et connaissances sous tendant !'analyse stratigraphique Un certain nombre de connaissances acquises depuis le debut des annees 90, d'hypotheses relatives au site etudie nous ont permis de lire Jes informations speleothemiques La prerniere a deja ere enoncee; ii s'agit de l'homogeneite du mode de fonctionnement de edifice durant la periode consideree. Poser cette hypothese permet de considerer que Jes evolutions revelees dans la croissance, dans Jes couleurs de la calcite sont "directement" imputables a des modifications environnementales. La deuxieme hypothese est relative a la notion d'impluvium du speleotherne. On suppose que les informations collectees sont globales. Cette globalite est envisageable en raison du penclage generalement peu important des strates du massif. La troisierne hypothese est relative a la rnise en place des alternances et a !'incidence des acides hurniques dans la croissance des speleothernes. En 1992, D. Genty montre au travers de l'etude d'une stalagmite actuelle du tunnel de Goclarville (Belgique) que l'annualite des alternances !amine sombre et compacte (D C.L.) et !amine claire et poreuse (W.P.L ) est due a la differenciation entre deux phases de croissance : l'une rapide (W.P.L.) et l'autre lente (D.C.L.). Ces deux phases soot correlees, dans le cas du tunnel de Godarville, aux variations de l'excedent hydrique. D epuis, la variation de la luminescence dans !'ultra-violet des differentes !amines a ete etudie (BAKER, A. & al. 1993 ; WHTIE, W. et BRENAN, E -S., 1989 ; SHOPOV Y.-Y. & al., 1990, 1991, 1994) Cette luminescence est imputable a la presence d'acides hurniques Le role de ces macromolecules n est pas encore bien cerne A u cours de cette etude on considere que la rnise en place des alternances est causee par le couple excedent hydrique/presence d'acides organiques. Cette hypothese est essentielle dans la comprehension des evolutions de la couleur macroscopique ainsi q u e clans la domination de telle ou telle !amine dans l'alternance. La quatrieme des hypotheses majeures sous-tendant cette etude est liee a l'hyp o these precedente. On associe couleur macroscopiq u e 1 et porosite cristalline, porosite cristalline et presence du couple acides hurniques/excedent hydrique. Ces differentes hypotheses ont permis de mettre en evidence diverses informations relatives aux evo l utions environnementales du plateau des Coulmes depuis le xvem siecle jusqu a aujourd'hui 3 Observations et interpretations La figure 2 presente les principales observations collecrees !ors de !'analyse stratigraphique de l a partie somrnitale de l'echantillon stalagrnitiq u e p r eleve dans la salle de la Cathedrale du reseau de Coufin Chevaline. Avant de voir de quelle fas:on !'evolution historique environnementale du plateau des Coulmes a marque la cristallisation du speleotherne etudie, on peut decrire quelques resultats important. 3 .1. R es ult a t s g l o b a u x On peut dans un p r emier temps chercher a interpreter l'evolution globale de la couleur macroscopique de la calcite. De la zone (a4) vers le sommet on passe d'une calcite blanche laiteuse (dominante WP L) a une calcite pure transparente ( dominante D CL ). Le passage d'un facies a l'autre semble se faire progressivement au sein de la zone (a2). Cette evolution progressive est revelee par la m o dification continue de 1 On ne se refere ici qu aux speleothemes dont la coloration n'est pas imputable a la presence d'oxydes de fer de cuivre, etc 62 Proceedings of the 12 th Internat i onal Congress of Speleology 1997 Switzer l andVolume 1


Croquis stratigraphique Vitesse Inclusion Age Reflectance (rouge) Luminescence evaluee evalue Echelle : I I cm 0,14 mm/an 0,1 mm/an 0,11 mm/an 0,14 0,08 mm/an Legende: Charbons et argiles Idem Idem Argiles Lits d'argile distinct Idem Lit d'argile distinct [Z] 1996 / 1890 1890 / !SOO !SOO / 700 700 1 600 (1) (4) (2) .. ........ (3) (5) CJ (6) Figure 2: Principaux resultats de ['analyse stratigraphique de La partie sommitale de l'echantillon etudie Legende (1) Calcite blanche transparente a dominante WPL; (2) lvfauvaise coalescence cristalline avec grandes faces de crista11x et inclusions; (3) Zone a ma11vaise coalescence engendree par un nivea11 d'argile ; (4) Calcite blanche laiteuse opaque, dominante WPL ; (5) Zone a altemances ond11lees ; (6) Calcite pure et transparente, dominante DCL l'enregistrement de la reflectance dans le rouge Cette mesure donne des informations sur la couleur de la swface etudiee mais aussi sur sa structure optique. La zone (a2) dont la couleur grise provient des nombreuses inclusions de charbons (confirrne par analyse a la sonde Raman), masque cette transition. La connaissance de }'evolution environnementale du plateau nous permet de valider les hypotheses posees lors de ce travail. On envisage done cela comme une modification du couvert vegetal et/ou comme une modification globale de la valeur des excedents hydriques. Il n'est sans doute pas necessaire de trancher entre ces deux hypotheses qui semblent intimement liees. Quoiqu'il en soit, la connaissance historique de !'occupation du plateau des Coulmes permet d'opter pour ]'evolution du couvert vegetal Durant le developpement de l'activite des charbonniere, la foret en place majoritairement constituee d'essences resineuses a ete peu a peu remplacee par du taillis caducifolie permettant un rendement suffisant quantitativement et qualitativement. Les etudes palynologiques confinnent bien la dominance des especes resineuses dans cette partie du Vercors a la fin de l'holocene (THlEBAULT S 1991). L'optimum climatique de 1300 ans revele par Jes etudes historiques est lui aussi marque dans la calcite : la translucidite de la zone (a3.1 ) revele par !'observation des echantillons mais aussi par Jes mesures de reflectance et de luminescence est inteipretable comme une modification de la structure optique de l'echantillon. De fait, la baisse de signal enregistre dans cette zone n'est pas imputable a la couleur macroscopique qui y est homogene La modification serait ici due a la baisse de l'excedent hydrique On le voit, la structure cristalline de la calcite comme le type d'alternance revelent de nombreuses informations L'etude des vitesses de croissance tres globale ne permet pas d'affmer ces inteipretations. On peut simplement noter que Jes vitesses de croissances entre Jes zones (al) et (a3) sont assez proches bien que l'une soit constituee d'alternances a dominante DCL et l'autre d'alternances a dominante WPL. De plus l'etude fme des alternances de la zone (al) a montre que les sequences constituees de dominantes DCL correspondaient aux vitesses de croissance les plus importantes ( de 0,2 contre moins de 0,08 mm/an). L'exploitation de cette partie des informations speleothemiques est en cours de realisation. L'interet majeur de ce type d'etude ne se situe pas dans cette relecture des evolutions environnementales mais dans la lecture de leurs interactions avec !'occupation humaine du plateau des Coulmes Sympos ium 7: Physical Speleology 63


3.2. Empreinte de la dynamique historique de !'occupation humaine A la base de la partie etudiee, on observe de nombreux lits detritiques pertwbant la cristallisation et les alternances Occupation humaine ou desequilibre "naturel" du milieu on peut difficilement se prononcer Le contexte de la partie etudiee tend plutot vers la deuxieme lecture. De fait, la calcite presente au dessous un changement de facies lie sans doute a !'evolution du fonctionnement de !'edifice etudie : deplacement du point d'impact de la goutte. Information en soi, cela n'est pas exploitable au pas de temps retenu pour cette etude. On touche ici une limite de ce type de travail qui reside dans la recherche de continuite dans Jes facies etudies. On remonte done dans la calcite et le temps jusqu'aux zones (a3.2) et (a3 l ). Les depots d'argile intercristallins ne sont pas imputables a un arret du concretionnement cristaux continus mais plutot a une liberation d'argile du sol. On suppose que ces liberations correspondent a des evenements ponctuels dans le temps et dans l'espace ayant destabilise le sol. II s'agit certainement de defrichements sur le plateau Ceux-ci demarrent au sud-Vercors a la fin du x:rrmi siecle (CHEVALLIER, P. et CouAILHAc, M.-J., 1991) et gagnent progressivement le nord. Ainsi, ils atteignent le plateau des Coulmes vers 1300 selon Jes croissances evaluees On pourra confinner cette interpretation en decelant des charbons au sein des argiles : Jes defrichements se faisaient par essartage ; cette technique consiste a couper Jes arbres (taillis) puis a les bn11er a meme le sol afin de le rendre plus fertile. A partir de la date evaluee 1400 on note un accroissement des niveaux detritiques qui entrainent peu a peu l'ondulation des alternances zone (a 2.4) Cette periode d'accroissement de la pression humaine sur le milieu correspond au developpement des communes du contrefort Est du Vercors Celle de St Geivais va jouer un role particulierement important a partir de 1619 avec la creation de forges a canons. Ce developpement est net dans les observations stratigraphiques Les zones (a 2 3), (a 2.2) et (a 2.1) recelent de nombreuses inclusions de charbons. Celles-ci sont parfois accompagnees d'argiles. La zone (a 2.3) correspond au demarrage de l'activite charbonniere, Jes inclusions quoique suffisantes pour colorer la calcite sont presentes en plus faible quantile que dans la zone (a 2.1). Les vitesses de croissance evaluees pennettent de caler cette periode de 1400 a 1500 environ La zone (a 2 2) 1500 a 1600 revele une quantile assez faible de charbons ralentissement ou deplacement de l'activite? La zone (a 2.1) correspond au maximum de l'activite qui cesse brutalernent a la fin du xrx:mi siecle. Cette periode correspond de meme a la pente maximum de la reflectance et a la chute de la luminescence attribue a la modification du couvert vegetal. Tout cela tend a montrer !'incidence du fonctionnement des forges et autres industries ayant contribue a ecouler le charbon produit sur le plateau des Coulmes ; les forges de St Geivais cessent toute activite en 1869. La zone (a I) correspond a l'histoire de !'occupation humaine de ce siecle La zone (a 1.6) correspond a la relaxation du systeme face au desequilibre du a l'activite charbonniere ; celui-ci n'a pas ete etudie en detail. Nous sommes neanmoins conscients de l'interet de cette information : cette relaxation est-elle propre au systeme karstique, a cet edifice, est-il possible de retrouver ce type d'indicateurs dans tous les edifices !amines ? En ce qui concerne la croissance globalement plus rapide, on peut la correler a la fin du petit age glaciaire Les inclusions quasirnent inexistantes revelent l'histoire de la depression de l'homme sur ce milieu La purete de la calcite n est pas expliquee vraiment. Une hypothese plausible tend a l'irnputer a !'absence de sol et a la production tres discontinue d'acides humiques. Cette hypothese pose probleme face a l'etendue de ce phenomene qui est releve dans de nombreux autres massifs. 4. Discussion et ouvertures Cet exemple qui montre la richesse des informations speleotherniques entraine un certain nombre de questions Leurs reponses permettront de valider pleinement cet outil de recherche paleoenvironnemental. Nous ne reprendrons pas ces questions et pistes qui ont ete evoquees au cours de cet aperryu des resultats de ce travail. On peut rappeler celles qui nous semblent particulierement i.mportantes en vue de la mise au point de cet ouW : quel est le role des acides humiques dans la mise en place des alternances ? Quel est le type de drainage des particules collectees dans les concretions, global, ponctuel ? Quelle est !'incidence de la structure optique des speleothemes dans !'acquisition des informations paleoenvironnementales ? Bibliographie BAKER, A. & al 1993 Annual growth banding in a cave stalagmite. Nature. vol. 364 :.518-520. CHEv ALLIER, P. & CouAILHA c, M.-J. 1991 Les dauphinois et leur foret au XVIl et XIX siecle Publie avec le concours de l'U P.M.F. pour le C.R.H.E.S.I.. Grenoble. DELANNOY, J -J. 1981. Le Vercors septentrionnal : le karst de surface et le karst souterrain. These de ill cycle. Grenoble. Institut de Geographie Alpine, 537 p. DELANNOY, J.-J. 1984. Le Vercors: un massif de la moyenne montagne alpine. Karstologia. n :. 34-45 DELANNOY, J.-J. & al. 1988 Les remplissages speleologiques : un apport a la connaissance de la karstogenese du massif des Coulmes (Vercors A!pes). Annales de la societe geologique de belgique Tome 111 : 21-38. GENIY D 1992. Les speleothemes du tunnel de Godaiville (Belgique) un exemple exceptionnel de concretionnement moderne Interet pour l'etude de la cinetique de la precipitation de la calcite et de sa relation avec Jes variations environnementales. Speleochronos n :.3 -2 9. GENIY, D & al. 1994 Nouvel exemple d'alternances de !amines dans une stalagmite (Grotte de Dinan "La Merveilleuse ", Belgique). Speleochronos. n :.3-8. MAIRE, R 1990. La haute montagne calcaire These d'etat, Bordeaux. Karstologia memoires. vol..3 : 731 p. PERRETTE, Y. 1996 Contribution des speleothemes a la connaissance paleoenvironnementale : exemple de l'etude de la structure interne d une stalagmite du reseau de Coufin-Chevaline (Choranche Vercors, France) Memoire de maitrise dirige par Delannoy J -J.. Universite Joseph Fourier Grenoble I. Grenoble 188 p PERRETTE, Y & al 1997. Characterisation of speleothem crystalline fabrics by spectroscopic and digital image processing methods (Choranche, Vercors France). Proc of the Jntemationnal Speleological Congres, symposium 07. ROUSSEAU, L. 1992. Etude physico-chimique et mineralogique des planchers stalagmitiques du pleistocene moyen Doctorat du museum national d'histoire naturelle a l'institut de paleontologie humaine Paris SHOPOV Y.-Y. & al 1990 Microzonality of luminescence of cave flowstones as a new indirect index of solar activity. Compres rendus de l 'Academie bulgare des Sciences, Tome 437 : 9-12. Saorov Y.-Y & al. 1991. One new indirect solar activity index. Geomagnetism and Aeronomy, vol.31, n, 739740. Saorov, Y.-Y. & al 1994 Luminescent microbanding in speleothems : high-resolution chronology and paleoclimate. Geology. vol.22 :.404-410. TmEBAULT, S. 1991. Approche de l'environnement vegetal prehistorique pendant la fin du tardiglaciaire et !'Holocene entre Alpes et Jura par !'analyse anthracologique,. Quatemaire. vol. 2 : 49-58. WHITE, W & BRENNAN, E -S. 1989. Luminescence of speleothems due to fulvic acid and other activators Proc. of the Intemationnal Speleological Congress, vol.10: .212-214. 64 Proceedings of the 12 1 h International Congress of Speleology, 1997, Switzerland Volume 1


Influence of the Bedrock CO 2 on Stable Isotope Records in Cave Calcites B y Yavor Y. Shopov, Ludmil T. Tsankov Section Speleology & Faculty of Physics University of Sofia, James Baucher 5, Sofia 1164, Bulgaria, E-mail: YYShopov@Phys.Uni-Sofia.BG Charles J. Yonge, H.P.Roy Krouse Dept. of Phy sics & Astronomy, University of Calgary, Calgary,Alberta TIN 1N4, Canada A.J. Timothy Juli NSF-Arizona AMS Facility, Phy sics building 81, The University of Arizona, Tucson, Arizona 85721, USA Abstract This paper is a first study of the influence of variations of bedrock fraction in speleothem calcite on 0 1 8 0 and oC records in speleothems. It is demonstrated that this influence can explain all o C variations and can produce a substantial part of 0 18 0 variations in speleothems In both o"C and o"O records higher surface temperature produces higher values in the speleothem, because the bedrock itself has high positive values of both o"C and 0 1 1 0. Introduction Speleothems (stalactites, stalagmites, etc .) are secondary cave calcite aggregates growing continuously up to a million years. They incorporate in their structure number of impurities, elements and isotopes which held important paleoenvironmental information (SHOPOV ET AL., 1991, SHOPOV, 1996). Once formed, the speleothems preserve these records and they can be read out by different techniques. Speleothems are formed at almost constant temperature and humidity and remain unchanged with time. The speleothem records can be read with a remarkably high resolution (SHOPOV ET AL, 1994). So the speleothems appear to be one of the best paleoenvironmental archives known so far Paleoclimatic Records by Stable Isotopes Records of stable isotopes oD, 0 13 C and 0 18 0 in speleothems hold paleoclimatic information. oD and o"O values in precipitations are function of the air temperature (FAURE, 1987). OD from fluid inclusions in speleothems gives good records (YONGE ET AL ., 1985) of the paleotemperature during speleothem growth, because there is no isotopic exchange with the bedrock Far more complicated are the records of 8 13 C and o"O in speleothems, because they can be supplied both from the surface and from the bedrock. The influence of the bedrock fraction (which varies from 6 to 85%) in these records usually is underestimated because it is considered to be constant with the time (SCHWARCZ, 1986, HARMON ET AL, 1978, TALMA ET AL., 1992) But JAKUCS (l 977) demonstrated that the bedrock fraction in the karst waters varies several times with the climatic variations (during glacial-interglacial transitions), due to the variations of non-carbonate karst denudation produced by acids (mainly biogenic) and other inorganic minerals Other factor producing variations of the bedrock fraction arises from the fact that the saturation degree of karst solutions (by bedrock calcite) depends exponentially on the temperature, because the dissolving rate has the same functional dependence. Traditional explanation of o"C variations by C3-C4 type plants variations (T ALMA, 1992) cannot explain the observed temperature dependence of 0 13 C (SHOPOV ET AL., 1994, SHOPOV ET AL. in press) in the speleothems This paper is a first study of the influence of the variations of bedrock o"C and 0 1 8 0 fraction in speleothem calcite. Bedrock itself does not contain any ,.C because it is completely decayed due to the old age of bedrocks, so the bedrock fraction in speleothems is called "dead carbon" .The "C concentration is good tracer of the bedrock carbon fraction in speleothems. It can be determined experimentally. "C and C have similar isotopic behavior (fractionation) in the karst environment; so the dead carbon fraction of is identical to the bedrock fraction of C in speleothems. Determination of the fraction of modem and "dead" (bedrock) carbon in speleothem calcite can be achieved by two methods: I By "C and independent absolute dating: Measured "C activity from a speleothem of age t,is: Am = A 0 exp( Real (specific) "C activity which should be obtained from calcite of the same age, but containing 100% modern carbon (at times of its precipitation) is: A,= A 0 exp(-A.t,), where tm is the measured ,.C age and t is the real age of the speleothem, determined by an independent absolute dating method, A. is the decay constant of "c and A is the initial activity of the modern "C. Am =A,.M, (0) where M is fraction of modem carbon in the speleothem or seepage water Therefore: M = exp [-A.(tm t,) ], (1) and the fraction of "dead" carbon in the speleothem or seepage water is: D = 1 -M (2) Symposium 7 : Physical Speleology 65


or D = 1 exp [-A.(t m t,)) (3) Equation (4) can be used to calculate o C or o C and system of these equations (for several samples from one cave) to calculate both. A precise determination of participation of modem and "dead" carbon in spe l eothem calcite can be done using eq (l) and (3). For o O equation (4) has the form: o Om = M. o O + (1 M) o O (6) 2 By measurements of l5'3C of the bedrock (o 13 C,), soil water (o C .), cave seepage waters (o C .,.. ) or speleothem (o"C,.): Close to o"O is o"O for precipitation which can be calculated from average annual surface temperature from the equation : (4) o O = s. T i, p (7) and M = (o 13 Cm o 13 C,) I (o C o C,) (5) Determination of participation of modem and bedrock carbon in speleothem calcite (presuming that 0 1 3 C 1 did not changed significantly from present day va l ue) can be done using eqs (5) and (2) Obtained data can be used to derive a precise estimate for the absolute age of the speleothem by eq (0) from 1 C dating and a determination ofM (by measuring o C , o' 3 C, and o C.) of the same sample. where T is the air temperature [ 0 C], s is the slope of the linear dependence and iis its intercept. sand imust be determined experimentally for every region. Fo r Calgary region (including Rats Nest and Castleguard caves) this equation have the form : 0 11 0. = 0.38. T 19.5 (eq.(6) of YONGE ET AL (in press)). In the case when data on o C are not available its average value for marine sediments of 0 6 1 55 Per mi!. PDB can be used This equation is regional and using it here for calculation of Table 1 is just to illustrate the bedrock influence on speleothem o 'i om. Climate Glacial Periglacial Temperate Medlterr. Tropical Desert Averaf{e Annual Air Temperature (CJ 0 0 8 15 25 25 Participation of Bedrock Carbon in speleothems (calculated after JAKUCS 1977) dissolved by : Inorganic CO2 9 5 9 8 2.5 15 Acids 9 20 30 33 47 55 Total bedrock CO2 18 25 39 41 49.5 70 Acti v e ( modern ) Carbon Surface CO2 82 75 61 59 50.5 30 Additional CO in respect to bounded( %), for 300 mg/l CaCO 27 27 33.9 41.6 55.1 55.1 Calculated equilibrium participation in speleot h em calcite %) DeadC 48 52 57.6 56.6 58.5 73 Active C 52 48 42.4 43.4 41.5 27 Calculated equilibrium 8 1 C ( PDB) in speleothem calcite (for 8 1 C = 0 Per mil. PDBfor bedrock and 8 1 C =-10 for soil CO /' 1 -5.2 -4 8 -4 2 -4.4 -4.2 -2.7 (for 8 1 C = 0 Per mi!. PDB for bedrock and l1 1 C =-16 7 for soil co f 1 -8.6 -8 -7 1 -7 2 -6 9 -4.5 Calculated 8 8 0 (SMOW) in precipitation for f{iven temperature/ 1 -19 5 -19.5 -16.5 -13.8 -10 -10 Calculated equilibrium 11 0 ( SMOW) in speleothem calcite (for rock 8 8 0 = 28 6 Per mi! (SMOW)) in (SMOW) 3.6 5.5 9 5 10.2 12.6 18 2 in (PDB) -26.4 -24.6 -20.7 -20 -17 7 -12 3 Tabl e 1. Origi11 of C0 2 i11 kar s t d e 11udatio11 a11d speleotl,em calcite a11d calculated values of 8 11 C a11d 8 1 8 0 i11 speleotl,em calcite with give11 co 11te11d of b e drock a11d modem carbo11: [lfthis value of s oil 8 11 C is calculated from 8 13 C of cave see page waters a11d modem speleotl,ems i11 Rats Nest cave (our data) and Castleguard cave, A lberta Ca11ada (after GASCOYNE & NE LSO N, 1983) wit!, correction for "d ea d" carbo11 d e riv e d from 14 C and U/T I, date s [2]average value of so il 8 13 C measured by TALMA & V OGEL (1992) above Cango caves, South Africa. [3}calculated by equatio,i (6) of YONGE ET AL, (i11 pre ss) 66 Proceed i ngs of the 12 t h International Congress of Speleology 1997 Switze rl and Vo lu me 1


These equations were used to calculate Tables l and 2. To estimate the bedrock carbon fraction (equilibrium participation) in Table I we took into account, that the amount of additional CO 1 necessary to dissolve one unit of bedrock CaCO 1 is temperature dependent (J AKUCS, 1977). All additional CO 1 is modern, while 50% of bonded CO 1 (in dissolved CaCOJ is bedrock and 50% is modern carbon. Tables I and 2 suggest that variations of the fraction of the bedrock carbon in speleothems can explain all variations of 6uC, without involving any changes in C3 and C4 plant types. This influence can produce a substantial part of 6 11 0 variations in speleothems. 6"0 records probably are result of both variations of bedrock fraction in the speleothem and variations produced by temperature and rainfall variations. In both 6' 1 C and 6 11 0 records higher surface temperature produces higher bedrock participation (due to higher non carbonate denudation (JAKUCS, 1977) producing higher fraction of bedrock carbon) and higher values in the speleothem, because the bedrock has positive values of both 6uC and 6 18 0 Higher temperatures also produce higher degree of saturation of percolating solutions by bedrock carbonate, because the saturation rate depends exponentially on the temperature. In some semi-arid regions like Israel the observed negative correlation between paleotemperature and S 1 1 O m is due to local air circulation peculiarities (FRUMKIN ET AL., 1996) Some spe\eothems exhibit negative correlation between 6 u Cm and pa\eotemperature (SHOPOV ET AL, in press). It can be explained by the following way: Main part ofCO 1 in karst waters come from soils (JAKUCS, 1977) In soils there are two sources ofCO 1 (FAURE, 1987): (!) respiratory, which is enriched by 1 1 C, and (2) CO 1 formed by soil decomposition (which is depleted to uC). The rate of soil decomposition (producing CO 1 of type (2)) is exponentially dependent on temperature thus producing anticorrelation between 6 11 cm and paleotemperature. In average the amount ofCO 1 of types(!) and (2) in soils is equal (FAURE, 1987). But in some places type (2) can dominate if the region is covered only by grass and is heated directly by solar irradiation. Indeed the region with such negative correlation is covered by grass and soil temperature reach 55 C during summer times (SHOPOV ET AL in press). Evidently in such cases with negative correlation described dependencies prevail over bedrock fraction variations. Sample Modern C "Dead" C ouc. 6uC, 6llC, 6 11 0,. 6"0. 6"0 llI l"/o I IPDBI IPDBI IPDBI [SM0Wl [SM0Wl ISM0W] Temperate regionRats Nest cave, Alberta, Canada (I'.,= 2 8 C) Precif)itation {f o_ = -18.4 SMOW, bedrock 8 1 C = 2 82 PDB) RNC-Jw 93.2.5 6.8 5 -8 9 -9.8c 2 82m 28.7 -19.1 RNC-2w 92 5 8.0.5 -8.9 -9 9c 2 82m 32.I -19.5 RNC-3w 85.7.5 14.3+0.5 -8.7 -J0.6c 2.82m 26.0 -19.5 RNC-F3 53.9 46.J -3 66* N.A. 2.82m N.A. 14.51 RNC-O2 52.1 47.9 -3 .35* -10.l 2 82m N.A. 13.87 Periglacial Region:-Castleguard cave, Alberta, Canada (I' .. =-2.JC) (Precipitation 8 0 = -20 3 SMOW bedrock 8 0 = +28.6 SMOWpresumed) (calculated after GASCOYNE & NELSON, 1983) GGI 36 64 0 -5.9c 3.3c N.A. 14.5 11 GG2 29 5 70.5 -0 6 -10.0c 3.3c N.A. 14.6 14.2 GG2Ew 60 40 -4 1 -9.0c 3.3c N.A. N A. -0.7 Semi-desert Region:-Cango caves, South Africa, (T ., = 17 5C) (Precipitation 8 0 = 12 9 SMOW (calculated), presumed bedrock 8 1 C = 0 PDB) (calculated after TALMA & VOGEL,1992) V3 38.3x 61.7x -6.4 -16.7m 25.4 drin w -5.2 -5.8 Table 2. So11rces of Carbo11 a11d Oxyge11 i11 speleothem calcite. Colllelll of modem a11d "dead" carbo 11 in seepage water alld speleothems, calc11lated from precise 14 C a11d Ullh dati11g by eqs.(I) a11d (3) a11d x from meas11red 8 13 C., li3C, a11d 8 13 C,,, after eqs.(5) a11d (2). All isotope data are i11 Per. mi/. /11dices mea11: mmeas 11 red soil, rbedrock, ccalc11lated from eq.(4) or (6), or systems of these eq11atio11s, pprecipitatio11, reca/c11/ated bedrock co 11 te 11 r,' water sample, avaverage a111111al, tl l b d fi u us va 11e 1s o tame rom a place 11ear the C sampli11g place, but ,wt from the same material. a1 8 0P is calculated by equatio11 (6) of YONGE ET. AL., (i11 press). Conclusion The influence of variations of the bedrock fraction in speleothem calcite on 6 1 1 C and 8"0 records can explain all observable o u C variations and can produce a substantial part of 6 11 0 variations in speleothems Acknowledgements A part of this research was supported by Research Grant 439 to Y. Shopov from Bulgarian Science Research foundation. We thank to Dr D C Ford, Dr. H.P. Schwarcz and Dr J. Bland for the support of this research Symposium 7 : Phys i cal Speleology 67


References FAURE G 1987 : Principles of isotope geology, (sec edition), John Willey & Sons. FRUMKIN A., FORD D C ., SCHWARCZ H.P 1996: Paleoclimatic record from the Mediterranean desert border karst Israel. In Lauritzen S E (Ed ) : Climate Change. The Karst Record. : 35 KWI, Bergen. GACKOYNE M ., NELSON D.E 1983 : Growth mechanisms of recent speleothems from Castleguard cave, Columbus Icefields Alberta Canada, infered from a comparison ofU-series and C 14 Age Data Arctic and Alpine Research, 15, 4 : 537-542 HARMON R.S. THOMPSON P SCHWARCZ H FORD D C. 1978 : Late Pleistocene Paleoclimates of North America as inferred from stable isotope studies of speleothems Quaternary Research, 9, 5470 JAKUCS, L. 1977: Morphogenetics of Karst Regions Akademie Kiado 323pp. SCHWARCZ H.P 1986 : In : Handbook of Environmental Isotope Geochemistry, part B: 397-421. Elsevier, New York SHOPOV Y Y DERMENDJIEV V BUYUKLIEV G., 1991 : Astrophysical Effects and SolarTerrestrial Relationships Recorded by the Convencional Indirect Indexes and the new LLMZA index of the Solar Acti v ity in the Past in S Radicella and K. Serafimov (Eds ) : ICSU Round Table on Space and Solar influences on the E nvironment: 55-72 SHOPOV Y Y ., FORD D C ., SCHWARCZ H.P 1994: Luminescent Microbanding in speleothems : High resolution chronology and paleoclimate Geology, v.22: 407-410, May 1994. SHOPOV Y.Y., 1996: Speleothem records of Environmental Changes in the PastPotential in Comparison with the other Paleoenvironmental Archives and Related UIS International book "Climatic Change the Karst Record ", Ed by S.E.Lauritzen. KWI, Bergen: 148-149 SHOPOV Y Y., FORD D.C., YONGE C MACDONALD W ., GEORGIEV L., SANAMBRIA M DERMENDJIEV V., BENDEREV A., BUYUKLIEV G., GEORGIEV S ., DELCHEV M SIRAKOV A M. (in press): High Resolution Records of Climatic Variations and Solar Forcing from the Luminescence of Speleothems from Duhlata cave, Bosnek Bulgaria, Cold Water cave, lowa,US and Rats Nest cave,Calgary, Canada in Y Daoxian (Ed.): Geology, Climate and Karst Formation IGCP 299 final report TALMA A.S VOGEL J.C. 1992 : Late Quaternary paleotemperatures Derived from a Speleothem from Cango Ca v es, Cave Province, South Africa Quaternary Research, 37 : 203-213 YONGE C.J., FORD D.C GRAY J. SCHWARCS H.P 1985 : Stable isotope studies of cave seepage water Chemical Geology 58:97-105 YONGE C.J ., NORMAN A.L., KROUSE H R (in press): Isotopic composition of precipitation at Calgary Alberta, Canada 68 P r oceed in gs o f t h e 1 2 th Inter na t i onal Co n gress of Spe l eo l ogy 1997 Sw itz e rl and Vo lu me 1


Paleomagnetic study in the Cuchillo Cura System Neuquen Province, Argentina Barredo, S.P. 1 A. Balbi 2 y G. Re 3 1 CONICET. Universidad de Buenos Aires Facultad de Ciencias Exactas y Naturales. Pabell6n II. Grupo Espeleol6gico Argentino (GEA) E-mail : 2 Universidad de Buenos Aires Facultad de Ciencias Exactas y Naturales Pabell6n II. Grupo Espeleol6gico Argentino (GEA) 3 Laboratorio de Paleomagnetismo "Ing Valencio" Universidad de Buenos Aires Facultad de Ciencias Exactas y Naturales. Pabell6n II. E-mail: Abstract This work was focused on the evaluation of the Paleomagnetic Method applied to the speleological analysis of the Cuchillo Cura System This system is composed of a number of different caves partially disconnected and fully developed in the limestones of the Lower Oxfordian (La Manga Formation) Although the isolated primary components are widely dispersed, the preliminary results suggest a normal polarity for those sites situated in the El Arenal Cave It was also possible to establish the main responsible for this magnetization : Magnetite ; and to determine that the remanent magnetization of cave sediments might include secondary components of a possibly viscous origin This method could be essential to understand the origin and the evolution of these caves but it can not be considered as the most appropriate for those partially evolved karsts The devices used to pick the samples produce irreparable holes and the fact that the stalactites or stalagmites are removed makes this method one of the least advisable in terms of cave protection. Resumen El objetivo del trabajo es la evaluaci6n del metodo Paleomagnetico aplicado al analisis del sistema Cuchillo-Cura. Dicho sistema esta compuesto por varias cuevas parcialmente desconectadas entre si y totalmente desarrolladas en las calizas de la Formaci6n La Manga, de edad oxfordiana inferior A pesar de la dispersion de !as componentes primarias los resultados prelirninares indican polaridad normal para la cueva de El Arenal. Tambien fue posible deterrninar el principal mineral responsable de esta magnetizaci6n : magnetita y la presencia de componentes viscosas como componentes secundarias. Si bien este metodo podria ser esencial para entender el origen y evoluci6n de estas cuevas se estima que no es el mas adecuado si estas estan parcialmente desarrolladas. Los dispositivos que se utilizan para extraer las muestras producen orificios irrecuperables y la necesidad de extraer estalactitas y estalagrnitas para completar las secuencias hacen de esta herrarnienta una de las menos recomendables si se trata ademas de preservar las cuevas con toda su belleza .. Introduction The Cuchillo Cura System has one of the most important karst records of Argentina. In spite of this the system is not large and comprises an area of approximately 2 km 2 Composed of four caves, the system has exquisite speleothems and important fauna which make of this karst a valuable mean of understanding the related processes. The origin of the Cuchillo-Cura caves is likely to be due to solution processes that occurred during Pleistocene times under a favourable climate This climate was drastically changed to arid conditions during the last phases of the Andean Orogeny resulting in the present day inactive karst system. Regarding rock deformation induced by the tertiary tectonism the present upward growing corresponds to the collapse of blocks along the joints Periodically these caves are flooded by waters from nearby lakes but this had no influence in the karst evolution The paleomagnetic method has been widely applied in Southamerica for geological studies but rarely for speleological matters specially in Argentina Difficulties exist with samples extraction, which modifies essentially the natural features of the fragile environment in in such little caves A first study has been conducted in this country with the purpose of analysing the method efficiency for such rocks the probability of an accurate dating Furthermore, we wanted to know the needed sampling density and the resulting consequences for the caves The preliminary results are discussed in this paper Sympos i um 7 : Phys i cal Speleology 69


Geologic setting The cave system under study is located along the western side of the Neuquen basin in the Cuchillo Cura range 10 km south from Las Lajas town in the Neuquen Province Argentina (see Fig. 1) The geographical coord i nates are : latitude 38 37' 25" south, and longitude 70 23' 03" west. Situated at 900 m above sea level the climate is semiarid with a mean annual rainfall between 200 and 250 mm per year and temperatures that vary from 20 to 6 Celsius for the warmest and coldest months respectively. The cave air temperature is below 10 C, the humidity nearly 100 % and slight wind currents have been detected for some of them. The geology of this region has been described by GROEBER ( 1946) and many others Mainly Jurassic and Cretaceous rocks form part of the sedimentary infill of the basin which covers almost the whole province Alternating marine and continental deposits characterise the depositional story of the basin, developed over a Paleozoic to Triassic heterogeneous basement; only during tertiary times there were volcanic and pyroclastic events of basaltic and andesitic nature Light-coloured limestones sandstones and gypsum are Los Tabanos, Lotena, La Manga and Auquilco formations which are surrounded by a basaltic plateau These units comprise oolitic fossiliferous and medium grained limestones greenish-grey fine grained sandstones with carbonate cement and important fossils, and finally an evaporite sequence (gypsum) interfingered by thin bedded stromatolitic limestones All of them range in age from middle Callovian to Kimrneridgian The structure consists of a huge east-northeast dipping anticline cut by northeast trending faults and a number of minor high angled joints with northeast southwest and southeast northwest strike. The La Manga carbonatic facies (mainly packstones) is the primary cave-bearing unit which holds four caves known as El Gendarme, El Templo, El Arenal and Los Cabritos. They are situated in the northern flank of an anticline dipping 15 0 50 100km Fig. 1: Regional map of the Cuchillo Cura system. Speleologic setting The caves are Pleistocene in age as inferred by ELZEARD (1987) They were developed under the strong control of joints as suggested by the strike of the main and lateral galleries which are disposed in a reticulate array In spite of this only two caves are physically connected : El Gendarme-El Templo. The modem system is inactive due to the non favourable climate and bears no apparent relationship to the present topography or stream patterns. But in earlier times under a more humid and cold period, when La Manga was still undeformed the caves were developed in the phreatic zone and water-table zone The system evolved to its maturity building carbonate speleothems of a fine and wild beauty They occur as calcium carbonate stalactites, stalagmites and columns calcite cave pearls calcium cascades lavish decorations as crust deposits coating the bedrock, carbonate floors composed offlowstones with swirls and ripples which appear to represent the growth at former pond surfaces and fluvial 70 Proceedings of the 12 "' International Congress of Speleology 1997, Switzerland Volume 1


like travertine terraces. All these deposits suggest cyclicity in their evolution. Subsequently the last phases of the Andean orogeny led to an episode of tectonism resulting in the uplifting and tilting of the formation followed by a climatic change to current conditions The cave system is developed vertically, favoured by the collapse of blocks along the joints but bounded by bedding planes (ELZEARD 1987). Methodology and resulting data Speleothem deposits contain a weak but measurable magnetic remanence As they are free from the depositional effects which cause the shall owing of the inclinations provide a valuable mean of studying a part of the karst history For this study we used a complete stalactite of 40 cm long and a collection of only 6 samples picked in a sequential order which partially cover a whole determined sequence. They were used as a preliminary study focused mainly on the applicability of this tool and the direct consequences for the environment. The speleothems were collected in El Arena! cave where the existence of a profuse carbonate speleothem deposit gave the opportunity of a relatively easy sampling Basically, this latter was held using a Brunton compass for the orientation and a pickaxe for the main extraction The targets were always from hidden corners or the thickening backwards steps of the floor terraces to avoid the disgusting look of a sampled place. The magnetic susceptibility (JO) for each of the present minerals and the natural remanent magnetism {NRM) were measured for the whole collection Samples were then subjected to a stepwise demagnetization up to 700 (see adjoining table in fig 2). Some of them were also subjected to stepwise alternating field demagnetization up to 80 mT (see fig 2). The remanence was measured on a cryogenic magnetometer with the purpose of isolating the primary remanent magnetism which is the NRM SRM (secondary remanent magnetism) ; this was later added during the geological history of the rock (VALE c10, 1980 ; BARREDO 1992) The collection showed a hard and stable remanence and a median destructive temperature of 575 C (see fig 2) indicating magnetite as the main magnetic carrier of the remanence In addition hematite occurs as a secondary magnetic mineral After removing the viscous components it was possible to determine the polarity of these rocks negative which is normal for South Hemisphere The paleopole could not be established due to the insufficient number of samples Discussion As suggested above it is possible to constrain the age of the cave using this method But for the particular case of this study the number of samples was not enough for a complete paleomagnetic study ; therefore the paleopole could not be established Despite this the final data led us to arrive to a very important point. In caves of little development (2 km 2 for the whole system and only 383 m 2 for El Arena! Cave) a dense sampling oriented to a more statistically precise data implies a partial destruction of the little cavern and parallely the modification of the ecosystem This latter is produced by gases emanated from the oil drill in a quite reduced space with no air circulation what implies certain risk for the local fauna We could confirm it as some specimens died some days after the initial sampling work On the other hand if a pickaxe is used bigger holes are made to produce an oriented sample from where a cube can later be extracted This is still more destructive for the natural beauty of such concretions considering that they are poor in number little developed and quite separated one from each other Previous works like the one in Querey south east of France (LEVEQUE & SEVKET, 1991) arrived to the same conclusions; in the Cuchillo Cura case a complete study could imply the irrecoverable modification of the natural scene of one of the biggest systems of the country. Conclusions The speleothems contain key information about the characteristics of the earth s magnetic field during their deposition The main magnetic carrier is magnetite accompanied by hematite as a secondary mineral The study shows normal polarity for the analysed sequence It is possible to conclude that the age of the caves could be determined through this method if a more dense sampling is held For the case of Cuchillo Cura or any other poorly developed cavern this tool is not advisable. Sampling should take into consideration the preservation of the natural beauty of the cave environment. Sympos i um 7: Phys ical Speleolog y 71


Sample EAl EA2 EA3 EA4 M6 M5 References : EAn= Stalactitic sample Mn= Sequential sample JO 2 10 1.65 2.35 1.80 1.85 2 00 JO= Magnetic susceptibility (x 10 -6 Gauss/Oe) NRM=Natural Remanent Magnetism (Emu x 10 1 1cm 3 ) CT 0 = Curie temperature N I Core EAI B) O' 0" ocP ... ,. o .. OIO ... + O"'"" 011 0 Negative inclinatio Positive inclina NRM CT (::::: 575) 15 0 5 15 25 25 10 40 20 65 20 5 10 l q J/Jo S 10 20 )0 > ,o 60 10 ''"' D) Up,N rum REFERENCE -..i 0.445 emu/cm3 0 Horizontal D Vertical Fig. 2: A) Table of demagnetization; B) Stereographics projection; C) Normalized demagnetization curve; D) Zijderverg diagram. References BARREDO S P 1992. Estudio paleomagnetico de !as sedimentitas cretacicas de !as Formaciones Castillo y Matasiete en la comarca del caii.ad6n hom6nimo Thesis of the University of Buenos Aires Unpublished. ELZEARD L. 1987 Geologia de! sistema de cavernas de! cordon Cuchillo Cura Salamanca Bulletin GEA, v. III : 3-11. Argentina GROEBER, P 1946. Observaciones geol6gicas a lo largo del meridiano 70 Hoja Chos Malal Rev Soc Geol. Arg ., Buenos Aires 1 (2) : 177-208 LEVEQUE F & S SEVKET ,. 1991. Polarite, magnetique de remplissages karstiques paleogenes du Querey (Sud-Ouest de la France). C.R. Acad Paris t.312 Serie II : 1431-1438 VALENCIO D 1980 El magnetismo de las rocas Eudeba 351 p 72 Proceedings of the 12 th Internat ional Congress of Speleo logy, 19 97 Switzerland Vo l ume 1


The palaeomagnetism of speleothems of Middle Quaternary age from northern Australia N .C. Scherrer 1 M.J. Fischer2, S.J. Gale 2 R.N. Drysdale3 and H. Heijnis4 1 Department of Geography and Planning The University of New England, Armidale, New South Wales 2351, Australia; Current address: Mineralogisch-petrographisches Institut, Universitat Bern, Baltzerstrasse 1, CH-3012 Bern, Switzerland 2 Department of Geography, The University of Sydney, Sydney, New South Wales 2006, Australia 3 Department of Geography, The University of Newcastle, Callaghan, New South Wales 2308, Australia 4Environmental Radiochemistry Laboratory, Australian Nuclear Science and Technology Organisation, Private Mail Bag 1, Menai, New South Wales 2234, Australia Abstract Palaeomagnetic analysis of a flowstone interbedded with fluvial deposits in Ten Truck Cave in the Barkly karst of northwest Queensland, Australia, has provided a record of secular variations in the direction of the Earth's magnetic field. Dating of the speleothem by uranium-series methods has proven difficult because of detrital contamination and the preferential leaching of isotopes from the deposit. However, it is cautiously considered that deposition took place at around 140 500 55 500/40 500 before the present. This record represents the first evidence of palaeosecular variation for the Australian non-dipole field for any time prior to the Late Quaternary. The results demonstrate the considerable potential of speleothems as a source of datable and high resolution palaeomagnetic methods. Introduction The longest ex1stmg records of secular vanat10ns in palaeomagnetism from Australia extend back only ~16 200 years before present. One way of extending our knowledge of palaeosecular variation (PSV) beyond this is to exploit the potential of speleothems to preserve a reliable record of magnetic changes at a site The ability to date speleothems using uranium series methods allows such records to be placed in a reliable chronological framework. Orientated samples of two conformable flowstone units, TIC I (lower) and TTC2 (upper), were therefore obtained from an interbedded sequence of carbonate and elastic sediments located in Ten Truck Cave (18'S, 138'E) in the Barkly karst region of northwest Queensland Methods Seventy-five orientated specimens were obtained from eight stratigraphic levels of TIC! and a further 109 specimens were taken from seven stratigraphic levels of TIC2 The intensity and direction of the natural remanent magnetisation (NRM) of each specimen were measured using a spinner magnetometer. Eight pilot specimens from TIC I and nine from TIC2 were demagnetised using a biaxial, tumbling, alternating field (AF) demagnetiser. All remaining specimens were demagnetised to the optimum level of demagnetisation, determined according to the behaviour of the pilot specimens, and then remeasured. The age of the samples was assessed using the 23 Thl 234 U disequilibrium method Unfortunately, both TTCl and TIC2 contained detritus. This may have introduced uranium and thorium into the carbonate of the speleothem, thus invalidating the assumption of a system which is closed with respect to 23 'fh and 234 u and in which all the 23 'Th in the deposit is the product of the decay of 234 U. In the case of TIC! the detrital content was so high that the sample was considered unsuitable for analysis. Investigations were therefore restricted to nine specimens taken from the full depth ofTIC2. In order to minimise the amount of thorium derived from the non-calcareous fraction of the flowstone in the solutions made up for analysis, the specimens were prepared by dissolution in dilute nitric acid. The 23 'Th/ 232 To ratios of the digestions may be used to assess the extent of detrital contamination. Since thorium bonded onto detritus will contain isotopes which are not necessarily the product of radioactive decay, the presence of the long-lived parent isotope 232 To in a speleothem provides clear evidence of detrital contamination. With one exception, the 230 Th! 232 To ratios in TIC2 all lie below 5, that is, in the highly contaminated category of FORD & W1LLIAMS ( 1989, p. 360). The 234 ul2 32 To ratios of the weak leaches are also less than 5 By comparison, other contaminated speleothems (2 30 Th! 232 To < 20) from the Barkly karst possess 234 U!2 32 To ratios in excess of 600. The upper two specimens from TIC2 displayed 230 Th/ 2 34 u ratios > I Since uranium is much more soluble in freshwater than thorium, this suggests that some 234 U has been removed by post-depositional leaching. These two specimens were therefore omitted from further analyses The remaining specimens were assumed not to have been affected by leachin g. Plots of 238 U/ 232 To VS 234 Ul2 32 To and 23 4 U / 232 To VS 230Th/ 232 To provide a means of representing patterns of mixing between authigenic calcite (with a high uranium content relative to 232 To) and non-calcareous material (which generally has a low uranium content relative to 232 To). Plotting the results of the analyses in this fashion reveals straight-line relationships with a 234 u/2 38 u ratio of 2.038.243 and a 23 'Th/ 234 U ratio of 0.803.154. Both relationships are strongly indicative of a single source of uranium uptake which remained constant throughout the period of deposition of the speleothem. Results 1 Uranium-series analysis: the corrected 234 ut 23 Su and 23'fhl 2 3 4 U values yield an age of 144 500 55 500 / 40 500 years. 2. Palaeomagnetic analysis: all except four of the specimens from TIC! showed a normal polarity magnetisation, suggesting deposition during the present Brunhes normal polarity chron (<0.78 Ma). The demagnetisation process appears to have disrupted the magnetic signature of some specimens from TIC 1. Clustering significantly worsened with demagnetisation and the AF-cleaned dataset exhibited excessively large ag 5 values. The magnetic signal of TIC! Symposium 7 : Phys ical Speleology 73


cannot th erefore be considered to b e a r elia bl e indi cator of the PSV of the Earth's magnetic field Despite this, the data have been plotted in sequence with those of TTC2 (Figure I) Fisher mean directions and associated error bars have been plotted only where K > 2. Further interpretation of the data in terrn of geomagnetic PSV has not been undertaken As with TTC I. all the specime ns from TTC2 showed negative inclinations, s ug gest ing deposition in the last 0.78 Ma. TTC2 possessed r elatively high magnetic intensities, a minimum of 12 specimens per stra tigr aphic level, good clustering of magnetic direction s at each le vel and independent age calibration based on uranium-series dating I t therefore exhibited considerable potential for the provision of a reliable record of PSV The AF-cleaned declination and inclination data are plotted against depth in Figure I. Ten Truck Cave, northern Australia (18'S, 138'E) Declination (degrees) Inclination (degrees) 0 0 00 0 ~ 0 0 0 r-A0 0 0 f----\ I 50 50 0 0 ,__.......,. ax, I I / 1 1TC2 -+--i 1TC2 --I 1 o ...,, o ~ o 100 I I 100 I } 0 f--~ 0 ..... 0 0 i-h 0 .. .~ ,......_ I I I I '-' 150 hiatus separating lower and upper section 150 hiatus separati ng lower and upper sectio n -5 I I g. 0 I oo 0 0 0 0 o I 0 0 Cl I 200 200 I 0 I------+e-----t I 0 : .............. -+---4 f-o 0 0 J 0 0 250 0 0 o I oo 250 JTCJ TTCJ I 0 b 0 0 o l I I -~0 300 300 -80 -40 0 40 80 -90 -70 -50 -30 -10 10 Figure I The ma g neti c declination and inclination of TTCI and TTC2. The signal of each individual s pecimen ha s been ploft e d a s a s mall ope n circle, w hilst Fisher mean s for eac h lev e l have bee11 represellfed as so lid circles c o1111ect e d by a s mooth line. Circular sta ndard deviations are s ho,vn as horizanta/ er ror bars (~ 1 ,, c = "9 5 ; so,c = a9sfcos In c). Discussion Figure 1 represents the first r eco rd of PSV from prior to the Late Quaternary from anywhere in Australia Records of sim ilar age are available from ew Zealand (TUR ER & LYONS, 1986; PI LLANS & WR I GHT. I 990), but these show no similarities with those from Ten Truck Cave, probably because the sites lie in eparate region of the non-dipole field The resultant curves of inclination and declination exhibit a relatively smal l amplitude. This may be interp r eted in three possible ways: I The ample has grown very l owly and the pecimen height of 17.5 mm is insufficient to resolve the PSV of the geomagnetic field The signal is therefore an average approximating to the direction of an axia ll y-geocentric dipole field. 2 The sample has grown very rapidly, as is possible in tropical climates, but the record is too short to depict sign i ficant fluctuat i ons of the directional compone n ts of the geomagnetic fie ld 3. The sample recorded apparently stable geomagnetic field directions over a long time interval. Whilst suggestion I could be tested by decreasing the specimen size and measuring subpecimens on a more sensitive cryogenic magnetometer PSV records from other parts of Australia covering the last -16 ka show similarly low amplitudes of declination and incli n ation. If this is characteristic of the non dipole fie ld in the Australian r egion, possibility 3 cannot be ruled out. Option 2 seems unl ikely since the finely-laminated stratigraphy of the flowstone suggests that the samp l e grew ove r a relatively long time period Using the AF-cleaned dataset, a mean inclinatio n of -33.7 has been calculated for the whole of TTC2 This is close to the -34.1 expected for an axially geocentric dipole field at thi s latitude ( I 8'S) and may indicate that the sample was deposited over a period of several thousand years, the time required to equalise secular variations in the Earth's magnetic field. A significantly steeper mean inclination of -42.4 for the NRM dataset is mo t probably the consequence of a viscous overprint acq uir ed b y grains of l ow coercivity through exposure to a changing ambient fie ld Support for this interpretation comes from the 1995 inclination at the site of 48.3 (comp ut ed using the international geomagnetic reference field). This ap p ears to confirrn that demagnetisation of all specimens in an applie d field of I O mT has effective l y removed the viscous remanent m agnetisation and isolated the primary magnetisation. Plotting the Fisher mean directions of each stratum as virtual geomagnetic poles (pole po itions calculated on t h e assum ption that the palaeomagnetic direction of each stratum is the product of a dipolar geomagnetic field) reveals little movement, with position s closely clustered around the 7 4 Proceedings of the 12 1 h International Congress of Speleology 1997 Switzerland Volume 1


180 90w 90E o Figure 2. Virtual geomag11etic pole positio11s (1 is the most rece11t) based 011 the Fisher mea11 directio11s of each stratigraphic level for the AF-c/ea11ed dataset of sample TTC2. The positio11s are plotted 011 a polar equal area projectio11. Conclusions This paper reports the first investigation in Australia of speleothems as sources of PSV information. The results reveal the considerable potential of the method as a means of obtaining datable and high resolution palaeomagnetic records Figure I represents the first PSV curve which has been compiled for the Australian non-dipole field for any time prior to the Late Quaternary Such information is of immense value both in cerms of understanding geomagnetic field beha viour and in providing the foundations of a novel method of dating stably magnetised material. Acknowledgements We gratefully acknowledge logistical support for this work from the Musselbrook Reserve Scientific Study organised by the Royal Geographical Society of Queensland and in particular, from Mr K.N. Teys We should also like to thank the Australian Research Council and The University of New England for financial assistance. MJF acknowledges the support of an Australian Postgraduate Award and an Australian Institute of Nuclear Science and Engineering Postgraduate Research Supplement. This is contribution number 7 of the Barkly Karst Project. References FORD, D.C & WILLIAMS, P W 1989 Karst Geomorphology and Hydrology. Unwin Hyman, London 601 pp PILLANS B & WRIGHT, I. 1990 500 000-year paleo magnetic record from New Zealand loess. Quaternary Research 33, 178-187. TURNER, G.M. & LYONS, R.G. 1986 A palaeomagnetic secular variation record from c 120 000 yr-old New Zealand cave sediments. Geophysical Journal of the Royal Astronomical Society 87, I 181-1192. Symposium 7: Phys i cal Speleology 75


Holocene palaeoclimatic fluctuations recorded by stalagmites: Grotta di Ernesto (northeastern Italy) S ilvia F risia ', Andr e a Borsato ', Baruch Spiro 2, Tim H e aton 2, Y iming Huan g3, F rank M cD e rmott a n d G ianpa o lo Dalmeri 1. Museo Tri d entino di Scienze Naturali, via Calepina 14, 38100 Trento, Italy. 2 NERC Isotope Geoscience La bo ratory, K eyworth, Nottingham, N G12 5GG UK 3 D e p artment of Earth Sciences, O pen University Milto n Keynes M K 7 6 AA, UK. 4. University College Dub lin, Belfield D ublin 4, Ireland Abstract Speleothems from Gretta di Ernesto, located in the now temperate and humid Pre-Alps of Trentino (northeastern Italy) preserve records of c l imatic changes that occurred in the last 1 1,000 years and their effects on the cultural evolution of prehistoric man. The TIMS (thermal ionisation mass spectrometry) U-series dating of speleothem calcite, continuous stable isotope profile and calcite texture analyses on axial slabs of a stalagmite inform on the nature of H olocene climatic changes for the last 8 500 years The most marke d climatic events recorded in the cave spe l eothem are the Holocene h ypsithermal (7 650 to 5 300 year s B P ) the Medieval warming, and the Little Ice Age cooling. Rlassunto Gli speleotemi della Gretta di Ernesto, situata ne ll e Pre Alpi del Trentino, hanno registrato ii susseguirsi dei cambiamenti climatici avvenuti negli ultimi 11,000 anni. Questi hanno influito sull evoluzione culturale dell'uomo preistorico causando Ja frequentazi o ne o l 'abbandono dei siti di alta montagna Le datazioni col metodo U / Th unite alle analisi degli isotopi stabili in serie continua lungo l'asse di una stalagmite e correlate con le analisi tessiturali hanno permesso di identificare la natura dei cambiamenti climatici Olocenici per gli ultimi 8 500 anni, s eparando periodi fre d di e secchi da quelli ca l di e umidi Gli eventi piu marcati riconosciuti sono !'Optimum c l imatico olocenico (7 650 a 5,30 0 anni B. P ) la fase ca l da Medioeva l e e ii periodo freddo della Piccola Eta Glaciale. 1. Introduction P aleoclimatic reconstruction applied to archaeo l ogical sites inform on the environment and on the impact of climatic changes on human cu l ture One of the most common unknowns is the reason for the prehistoric abandonment of large areas such as Mesa Verde in the North American Southwest, which needs highly refined and precise information about the past climate to be solved. The present study is one such refined reconstruction for the Holocene through the physico-chemical characteristic of a stalagmite from a prehisto ri c site Gretta di Ernesto in northern Italy. Speleothems are climate-sensitive deposits which record pa l eoclimatic changes w i th annual reso l ution on decadal to millennial scales (FORD & WILLIAMS, 1989 ; GASCOYNE, 1992), and can be p recisely dated by U / Th thermal ionisation mass spectrometry (TIMS) ( I VANOVI C H & HARMON, 1993) With respect to other proxy data, speleothems have the advantage of carrying a continuous reco r d through isotope composition (8 11 C; 8 18 0) and textures (FRISIA & BoRSATO, 1997) of sequentia l growth layers The 8 18 0 of speleothem calcite precipitated in stable isotopic e qu ili b rium th eoretica ll y allows to calculate mean annual paleotemperatures by using the carbonate temperature equation ofFRIEDMANN & O NEILL (1977) : T (C) = 16 -4 14 (8 1 8 0 -8 1 8 0 .) + 0.13(8 18 O -8 1 8 O .)' (I) where the su b scripts c an d w refer respectively, to calcite and precipitating water By substituting TC with 6.5C (i.e the present temperature in Gretta di Ernesto) in equation (!), a difference of + I % 0 in 8 O c sho u ld corres p ond to a change in past temperature of -4. 0 6C provided that the 8 18 O w of the precipitating water remained the same. However this is not the case The 8 0 ,. of rainwater is also temperature dependent and combines its effect on the calcite-water fractionation equation Therefore, without knowing the original 8 18 0 it remains impossible to calculate the exact temperatures for the past through the sole 8 1 0 The 8 11 C values of spe!eothem carbonates provides a parallel record of climate controlled changes The carbon isotopic composition of seepage water is influenced by the uptake of soil CO ,. For most of the Ho l ocene in temperate preAlpine settings the vegetation like l y consisted wholly of C 3 type plants. Seepage water which has uptaken soil-CO derived from the decay of C 3 plant components commonly precipitates speleothem calcite with 8 1 1 C va l ues ranging between -13 to -II %0 (CERLING & HAY 1986). However, within C species, large variations in the isotopic signature are induced by fluctuations in water avai l ability and temperature. Under drought conditions C 3 plants show enrichment in 13 C in comparison to non-limiting water availability (BR U GNOLI & LAlITERI, 1991) which corresponds to a shift towards more positive values of the 8 13 C of seepage water and consequently of spe l eothem calcite Calcite textures are another proxy indicator of drip rate and degree of s u persaturation of the seepage waters (GONZALEZ e t al ., 1992, JONES & KAHLE 1993 ; FRISIA & BORSA TO 1997) which when correlated with isotope data underpin the interpretation of climate changes in terms of water avai l a b ility 2. Cave description Roadcut works in 1983 exposed the entrance to the Gretta di Ernesto at an elevation of I 167 m a.s.l. (Long. 11 39 28 Lat 45 58' 37") in Valsugana valley (Trentino, northeastern Italy). The first exploration revealed that the cave preserved important Mesolithic findings Consequently, the entrance was closed again by a door that inhibits air flow (DALMERI, 1985). Symposium 7 : Phys i cal Speleology 77


The cave is cut in the dolomitized limestones of the Calcari Grigi Formation (Lias) and consists of a single downdipping gallery, 2 to 5 m wide, up to 4 m high and 72 m long, which developed along a NW trending subvertical fault 20 m below the surface The original syngenetic morphology of the cave is now blurred by collapse phenomena and the whole floor is covered by angular blocks partially coated by flowstones Present-day environmental setting The baseline for past climate interpretation is the present climatic and environmental setting Climate in Valsugana Valley is of temperate humid sub-c o ntinental type. Mean annual precipitation rate varies between I 000 and 1500 mm and shows a bimodal distribution, with a maximum in May-June and a secondary maximum in October-November Snowfalls occur from December to March, and snowmelt takes place between March and the first half of April (BORSATO 1995). The vegetation above the cave consists of a C 3 association with Fagus, Carpinus, and subordinate Lari x decidua and Abies alba this laner being indicative of humid and relatively warm conditions The cave has several drips, with low (from 0.1 to 2 ml/minute) to medium discharges (up to 50 ml/minute) feeding small, muddy concretional pools. Pools show seasonal water level fluctuations: in summer some are dry, in autumn and spring all pools fill up and several overflow Palaeoanthropological data The floor of the first chamber is a 45 m 2 Mesolithic paleosurface with bones of Ibe x and red deer showing butchering marks A camp fireplace was located at the southwest wall. The fireplace was subsequently coated by a thin flowstone. A 14 C analysis on charcoals yielded a date of 8140 80 years BP (uncalibrated) (AWSIUK et al 1994), which sets the time span of human frequentation The cave entrance at that time was larger than present, and was blocked by a landslide immediately after human occupation (DALMERI, 1985, AWSIUK et al 1994) Apparently, this was not the cause for the abandonment of the site. In fact, all hunting activity on the area ceased at about 8,000 years B.P., when all the high mountain sites were abandoned for the valley bottoms. Hunters returned to dwell onto the plateaus from about 5,000 to 2,800 years B P .. The general desertion of high-mountain sites in the whole Trentino-Alto Adige region about 8,000 years ago poses the question whether this phenomenon was related to the cultural evolution or to climatic changes, or to a combination of both. 3. Methods The two-year monitoring of environmental parameters and chemico-physical characteristics of dripwater in the first chamber indicates that dripwater has a constant temperature of 6 5 + / 0 050C (BORSATO, this congress), and that the relative humidity is at saturation state, i.e there is no evaporation These are necessary prerequisites for calcite to precipitate in isotopic equilibrium (HENDY, 1971) A 368 mm long, cone-shaped, active stalagmite (ER 76) was removed from this chamber. ER76 was sliced along the axis and mapped to allow for precise correlation of sample location for all methods. In the axial part stable isotope samples were drilled at 2mm-interval; textures were identified through thin section series, and their changes recorded in a textural Log; and dating was performed by U / Th-TIMS on five samples selected on the basis of textural changes 8 13 C Poe (permit) 8 18 0 Poe (per mil) Calcite textures Columnar Microcrystalline Dendritic -11 10 -9 -8 -7 -6 -5 -8.4 -7 9 7 4 -6 9 -6.4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 N 0 0 0 0 0.: 0 N N ai 0 0 0 0 (") 0 0 0 0 0 (") (ll (") QJ >, 0 0 0 0 0 0 1/) 0 0 IJ) 0 0 QJ .,. 01 (ll 0 0 0 .c 0 0 I0 1/) :3 1/) 0 0 0 (!) 0 -0 0 0 0 0 0 (!) (ll (!) 0 0 0 0 r--. 0 a. 0 0 0 0 0 r-r-0 0 0 a:, 0 0 0 0 0 0 co a:, 0 0 0 Ol 0 0 0 0 0 0 Ol Ol Figure 1: tlle time series lf1C, S'O a11d textures for ER76 stalagmite plotted against time (Ullh i11terpolated ages). Nine episodes were defined 011 tlle basis of sllift i11 8 1 C record accompa11ied by covarialll S' 0 j111ct11ations Dashed lines identify tlle limit of eacl, climatic episode. Textural clla11ges aided i11 tlle recognition of dry vs. humid co11ditio11s (cf. Table I). 78 Proceedings of the 12 '" International Congress of Spe l eo l ogy 1 997 Switzerland Vo l u m e 1


4. Results The stalagmite commenced to grow at 8,581 +/150 years B.P ., i .e. when the cave was still open and hunters frequented it. ER76 grew, without depositional gaps to the present. However, it has frequent textural changes with a trend from dominant disequilibrium (dendritic) to equilibrium (columnar) fabrics (FRISIA, 1996; FRISIA & BORSATO 1997) toward the top. The synthesis of the analytical study is reported in Figure I, which correlates stable isotope and texture trends vs. time. The climatic episodes are individuated by rapid shifts in the 15 13 C record accompanied by covariant 15"0 fluctuations. The limit of the climatic episodes were set at the middle of the 15 13 C slope for each major shift. 5. Discussion The 15 18 0 values of calcite depend on both temperature and original composition of rainwater, and, consequently, its interpretation cannot be unequivocal. On the opposite, calcite 15 13 C depends directly on the vegetation type and soil turnover at the surface, whereby sudden shifts in 15 13 C can be interpreted in terms of climatic changes from humid (calcite 15"C = -11 %0) to dry ( calcite 15 13 C = -6 %0) conditions. Therefore in ER76 sudden shifts in 15"C mark real climatic changes occurred in the Holocene and identify climatic episodes, through the correlation with textural changes and 15 18 0 fluctuations, characterized by changes in temperature and water availability in the direction summarised by Table I. Climatic Calcite Calcite parameters 15 13 C 15 18 0 Textures Cold il' (*) il' Dendritic to microcrystalline Warm .(). Microcrystalline/Columnar Dry il' Microcrystalline to dendritic Humid .(). Microcrystalline/Co lumnar Evaporation il' il' Dendritic Table 1 : Expected stable isotopic and textural variations for Grotta di Emesto stalagmite calcite, as a ft111ctio11 of climatic clla11ges duri11g tire Holoce11e. (*) = 011/y i11 case of severe cold (i.e. mea11 a111111a/ temperature <2C tllat presellt-day), toward lack of vegetatio11. The followings nine episodes have thus been identified : I from the bottom (8,580 years 8.P.) to 7,650 years 8.P. This episode shows the least negative 15"C and 15 18 0 values, coupled with dendritic textures. These characteristics indicate a cold, dry environment and possible evaporation related to the fact that the cave entrance was still open. II from 7,650 to 5,300 years 8.P. This episode is warm and humid, but punctuated by cool and dry phases, the most marked of which occurred from 6,450 to 6,130 years B.P. This climatic episode shows the most negative 15 18 0 values for ER76, which probably indicate that it was the warmest period of the Holocene, i.e. the Hypsithermal. IIIfrom 5,300 to 4,260 years 8.P.This episode is cooler and characterized by dry conditions, with decreasing temperatures toward its end. It correlates well with the decrease of Abies in the area. IVfrom 4,260 to 3,200 years 8.P.This period is cold and dry, with dryness decreasing towards younger ages, as indicated by the slope of the 15 13 C curve (to negative values). Vfrom 3,200 to 2,700 years 8.P.This is a relatively cool and dry episode, as indicated by the presence of microcrystalline to dendritic textures and by a shift of 15C towards more positive values. VIfrom 2,700 to 1,100 8.P.This time span was characterized by humid conditions and a progressive increase of both temperature and rainfall. VIIfrom 1,100 to 400 8.P. (900 to 1600 A.D.) This is a warm and humid period, which corresponds very well to the historical Medieval warm phase. VIIIfrom 400 to 150 8.P. (1600 to 1850 A.D.)This time interval records a dramatic cooling, marked by a sudden negative shift of 8 18 0, which correlates with a shift towards more positive values of 15"C. The time span corresponds well with the Little Ice Age. The strong 15C shift may identify either a stress in the vegetation caused by cold/dry climate or extensive woodcutting above the cave for heating. IX from 150 8.P. to present-dayThis episode marks the recovery of warm and relatively humid conditions after the Little Ice Age (LIA) deterioration. It correlates well with the recovery of Abies and of present-day climate. The ER76 record indicates that Mesolithic hunters frequented Grotta di Ernesto in a dry episode, when mean temperatures were about I C lower than at present (Fig. 2). They abandoned the high sites during the hypsithermal, when the climate was warmer and more humid It is probable that the deciduous forest line shifted upwards and colonised the ecological niches of Ibex and red deer, which need open spaces and were the game of the Mesolithic hunters This hypothesis is supported by the warm and humid (Abies alba etc.) pollen associations ascribed to the Hypsithermal, from a peat-bog 200 m above Grotta di Ernesto (MARTELLO, 1993). It is also probable that increased humidity made inhospitable the rock shelters of the plateau. From the 15"0 values we infer (cf. Fig.2) that the hypsithermal had mean annual temperatures about 2C higher than today. This inference is supported by the l.5C increase calculated for the mid-Holocene warm period in the Alps by using pollen data {BURGA & OROMBELLI, 1996). The ER76 isotope curve shifts allow to set the end of the Hypsithermal at about 5,300 years B.P This limit is consistent with the onset of the Alpine neoglaciation as recorded by the Iceman of Hauslabjoch (3,250m a.s.l.). The Iceman's death occurred at 5,300 to 5,050 14 C ea!. years B.P. and corresponds to a sudden cooling as indicated by both the conditions of the mummy and its immediate burial under snow and ice (Baroni, 1996). Therefore, we can set with confidence the end of the Hypsithermal in the Alps at 5,300 yr B.P Hunters returned on the plateaux in Episode Ill, the subsequent cool and dry phase. As for historic times, ER76 records high temperatures for the Late Middle Ages and sets the beginning of the LIA at the year 1,600, which corresponds to the 1594 to 1608 severely cold winters historically recorded in Trentino. 6. Conclusions The climatic record extracted from ER76 gives useful indications to interpret cultural changes during prehistorical times. In particular, it appears probable that mesolithic hunting was related to cool and dry conditions, when game (Ibex, red deer) was bountiful and shelters were hospitable. Climate-related changes in vegetation associations explain major carbon isotope trends in ER76, which correlate well with pollen data from the same area. Therefore, 15 13 C appears to be a most reliable paleoclimatic indicator in Alpine speleothems, although useful information can be extracted from 15 18 0, when confronted with the other proxy data Symposium 7 : Physical Speleology 79


-8 20 10 5 -7 95 9 5 0 0 -7.70 8 5 :, ro ai 0.. ,...._ -7.45 E 7 5 E 2 'Q) 3 -7.20 ro :, 6 5 C C al 0 ro a. C 0 -6.95 co ro 5 5 Q) E c.o u -6.70 4 5

Environmental contro l s of origin of the annually varved calci t e speleothems Mic h al Gra d ziiiski, Mari u sz Rospondek, Joachim Szulc Institute of Geological Sciences, Jagiellonian University, Oleandry Str. 2a, 30-063 Cracow, Poland Abs tr a ct s Rhythmically laminated calcite speleothems have been studied in terms of the basic environmental factors controlling their growth mechanisms. The laminated microfacies occur in speleothems of different ages and various topographical and geological settings Rhythmical lamination depends on alternation of white, pure calcitic layer and brownish calcite layer composing elastic and organic components The alternation reflects seasonal changes in composition of percolating solutions. The rhythmically laminated calcite precipitates from very thin (> I mm) water film enriched in organo-clastic impurities during wet season (summer mostly in the tudied caves) while the solution stays pure during dry (mostly cold) season Precise paleoenvironmental analysis is however more complex and needs regard on topographic setting (high vs. low topography), climatic controls and geographical position of the studied site. Introduction Last decade brought significant progress in study of speleothems as a possible tool of paleoenvironmental reconstructions One of the most attractive type ofspeleothems is the rhythmically laminated calcite, found in speleothems of different age and sites (BAKER et al., I 993, 1996 ; SHOPOV e t al., 1994, 1996 ; GENTY er al., 1995 1996 ; GENTY & QUINIFF 1996) In contrast however to a wide employment of the laminated speleothems in paleoclimatic reconstructions their growth controls are poor recognized. In order to check the effects of environmental controls (e g. cave altitude and depth hydrological and (micro )climatic regime vegetation cover, age of speleothems etc.) we examined spe leothem sect ions from several caves of Poland and Slovakia Research objectives strategy and methods Speleothem from 15 caves from Poland and Slovakia have been chosen for detailed sedi mentological and geochemical analyses. Altitude of the explored caves ranges between 250 and 100km A investigated caves A l 350 kA collected in 5 caves (Figure I). Petrographical and UV luminiscence observations were carried out on the standard thin sections of calcite s peleothems The same specimens were examined by SEM method accompanied with microprobe analysator. A set of samples for geochemical analyses was selecte d based on microscopy. About 20 g of the crushed samp le s were treated repeatedly with acetate buffer at about 80C until carbonates were removed. Insoluble residuum was washed with water, centrifuged, dried and weighted. Powder X-ray diffractom etry (XRD) was performed on the so prepared material with a TUR M-62 diffractometer equipped with a hori zo ntal goniometer HZG-4B Cu-Ka radiation (Ni-filtered) was used Elemental composition was analysed with AAS (Table I). Results The rhythmical lamination i s an intrinsic feature of various types of speleotherns; stalactites, stalagmites and flowstone covers on cave walls. Rhythmical laminated microfacies (RLM) occurs as one of several microfacies (mf.) types building the speleothem sequences (DZIADZIO et al ., 1993 ; GRADZINSKI er al., 1996) RLM is particularly common feature of spelean columnar calcite. Other microfacies types of speleothems e.g. botryoidal or blocky mf are lacking of lamination The rhythmical lamination depends on alternation of white and dark laminae Thickness of the laminae ranges from severa l microns to I mm. The thicker laminae are visible with naked eye whereas the thinner ones are detectable only under microscopic study. Some of the discrete ultra-thin laminae are disc erni ble when one use other methods e.g fluore scence phenomenon White laminae are built by almost pure calcite while the dark laminae display higher contribution of elastic and organic components. RLM are composed of extremely pure calcite. Acetate buffer insoluble residua compose of-0.6% by wt. of these speleothems Only the sample from Zvoniva cave (Slovakian Karst) contains 3 6% by wt. of other minerals Amounts of hydrochloric acid insolubles varies from 0 08 % to 0.32% by wt. Amounts of the residua after HCI treatment are listed in Table I. Symposium 7: Physical Speleology 81


100% !:: CJ) z w 1z C!:: CJ) z w Iz t= ;;; --< N ..... ::l .... 0N C

Table 1 : Results of c h e m ical a11alysis ofi11vestigated samples HCI soluble [by weight] samples Ca Mg Fe Mn Sr P (%) [ppm) [ppm] [ppm] [ppm] [ppm] Lodova cave 39.4 1300 50 41 35 Wierna cave 39.8 172 80 177 Naciekowa 39.5 1600 250 26 cave P sia cave 39.6 1400 700 1 22 The longest found continuos sequence of rhythmic laminae is estimated as much as ea 2500 white-dark couplets Lamination displays lateral variation and changes from flat morphology to mall domal forms. This variation depends on contribution of elastic coarser material and inclusions iconcorporated within the calcite Discussion Nature and orig i n of lamination As sugge t the thickness, composition of the laminae and scarcity of erosional and depositional breaks within RLM this microfacies originated in thin water film under stable hydroclimatical regime. I 1 ,. r r t _-'7; ... I -"'lit, .. ... .. / ---/ ... 7 .. HCI insoluble [by weight] residuu Mg Fe Mn Al Si p m [ppm] [ppm] [ppm) [ppm] [ppm] [ppm] % 0.08 7 5 485 320 0.08 3 825 400 30 0.25 26 5 916 700 0.32 44 150 1120 1500 The lamination reflects changes of precipites and depends essentially on a non-carbonate contribution within spelean carbonates. The most common white-dark coupled band may be attributed to seasonal changes during the year White, pure calcite precipitates from solution free of other components while the coloured laminae originated in water, rich detrital and organic components. We interpret the white lamina as dryer-sea on increment and the darker one as deposited during wet period. Further interpretation could be more difficult since there are everal possible hydrodynamic control of such seasonality. For example the white laminae originate during the winter, in caves situated in lowland setting of the temperate climatic zone where, in tum, the summer rainfall promotes growth of the dark Fig 11 re 4 a-d: ormal" (a,b) a11d "reverse" (c d) l11m i11isce11ce of bro111 11 l ami 11 aefro m R LM. T h i 11 sectio 11 ofspe l eot h e m seq11e11ce fro m Z110 11 iva Cave (S l ovakiu) Left p h otos (a, c) tra 11 s mi tted lig ht rig h t p h o t os (b, d)-UV l11mill isce 11 ce Symposium 7 : Phys i cal Speleology 83


laminae On the other hand the dark laminae could form during wet winter months in the Mediterranean whereas the white calcite precipites during summertime. Regarding of the possible effects of the low vs high-relief topography of the studied objectives, the interpretation becomes even more complex. Lamination as a tool of paleoclimatic reconstruction and its limitations Annually stratified bands have been recognized as the most common climatic rhythm recorded in the laminated speleothems Some go even further and try to find finer cycles, up to diurnal growth of speleothems (SHOPOV e t al 1994, 1996) Most of the so far interpreted laminated fabrics were being studied in terms of luminescence microscopy The luminescence has been always assumed as evoked by humic and fulvic activators, encompassed in the dark laminae We have found however that the lamination could result both from the organic as well as inorganic contaminations in the calcite. Commonest inorganic components found in dark laminae are clay minerals siliciclastic ultra fine grains (quartz, micas) and amorphous ferrioxides (Figures 2 3). It is noteworthy that some of these impurities (e g. kaolinite) could give luminescence effect similar to fulvic / humic acids (GRANT 1962 ; TEI C HMULLER & Wou, 1977 ; DRAVIS & YUREVI C Z, 1985). Moreover, in some cases the dark laminae are high fluorescent and in some other cases the white laminae show higher luminescence as the darker ones (cf GENTY & QU!NIF 1996 ; GENTY et al 1996) We have found this phenomenon in one thin section (Figure 4)! All this suggests that interpretation of the laminated microfacies, based on luminescence microscopy may give very equivocal results. Optical microscopy observations show that the only one rhythm enabling further reasonable interpretations is the annual altemance. The lower-order cyclicity may be misinterpreted since the crystallographic properties of calcite crystals might strongly obliterate the visible features of lamination. Lateral vanishing or deformation of the laminae is commonplace in spelean calcite (Figure 5) Figure 5 : Vanishing of the lamination following the growth sector boundary Toe other important cause of uncertain interpretation of the high-frequency cyclicity is an information noise obtained for the rhythms shorter that the annual one. On the other hand the cycles longer then some thousands years could be significantly influenced by local factors. Transformations of plant cover, surface morphology or tectonic motion could substantially affect discharge of cave water and finally overshadow the climatic records. Conclusions 1 Rhythmically laminated microfacies occur in various types of speleothems formed in different geographical topographical and climatic settings. 2 Phenomenon of cyclical lamination depends on periodical changes in non-carbonate (elastic and organic) contribution in spelean calcite 3 Climatic, annual rhythm of the growth is the most common and obvious cycle in the RLM 4 Lower-and higher order of cyclicity might be modified or obliterated by other environmental factors Acknowledgements This study was supported by KBN (State Commitee for Scientific Research) grant no 0586/P04 / 95 / 09. References BAKER, A., BURNES, W L. & SMART, P.L., 1996 Luminescence and discharge in stalagmite drip waters, Bristol, England. In : E Lauritzen, ed.): Climate Change: The Karst Record Karst Water Institute Special Publication 2: 4-6 BAKER, A SMART, P. L., EDWARDS, R L., RICHARDS, D. A., 1993. Annual growth banding in a cave stalagmite. Nature 364: 518-520. DREVIS, J. J & YUREWICZ, D. A 1985 : Enhanced carbonate petrography using fluorescence microscopy } Sediment. Petrol. 55 : 795-804 DZIADZIO, P., R6ZNIAK, R. & SZULc, J., 1993 Origin of the Pleistocene calcite flowstones of two cave (Jaskinia Psia and Jaskinia Naciekowa) in the West Tatra Mts. Przegl'd Geologiczny 41 : 767-775. GENIY, D., BAKER, A BARNES, W. & MAsSAULT M 1996 Growth rate, grey level and luminescence of stalagmitic laminae. In: (S-E. Lauritzen,, ed.): Climate Change: The Karst Record. Karst Water Institute Special Publication 2: 36-39 GENIY, D. & QUINIF, Y., 1996 Annually laminated sequences in the internal structure of some Belgium stalagmites importance for paleoclimatology. J Sediment Res A66 : 275-288. GENIY, D QU!NIF, Y. & DEFLANDRE, G 1995 Microsequences de !amines annuelles dans deux stalagmites du massif de Han Sur-Lesse (Belgique). Speleochronos 6: 9-22 GRADZINSKI, M., ROSPONDEK, M. & SZULC, J., 1996. Microfacies types of calcite speleothem: hydrodynamical and chemical controls of their origin. In: (S-E Lauritzen,, ed.): Climate Change: The Karst Record Karst Water Institute Special Publication 2: 45. GRANT, J., 1962. Application of fluorescence analysis in sedimentary petrography. In: (H.B. Milner, ed.): Sedimentary Petrography, Allen & Unwin, London: 433-462 SHOPOV, Y. Y FORD, D C. & SCHWARTZ, H. P 1994. Luminescent microbanding in speleothems : High-resolution chronology and paleoclimate Geology 22: 407-410. SHOPOV, Y Y., TSANKOV, L., GEORGIEV, L. N DAMYANOVA, A FORD, D. C YONGE, C. J., MACDONALD, W & KROUSE, H P R 1996 Speleothems as natural climatic stations with annual to daily resolution. In: (S-E. Lauritzen,, ed ): Climate Change : The Karst Record. Karst Water Institute Special Publication 2 : 150151. TEICHMUllER, M & Wou, M., 1977 Application of fluorescence microscopy in coal petrology and oil exploration. J Microscopy 109 : 49-73. 84 Proceed i ngs of the 12 th International Congress of Speleology 1997, Sw i tzerland Volume 1


The antiquity of the famous Demianowska Caves (Slovakia ) Abst r ac t H. He r camn, P. Bella, M. Gradziiiski, J. Glazek, S.E. La u ritzen I. Institute of Geological Sciences Polish Academy of Sciences ul.Twarda 51 / 55 Warszawa Poland 2 Slovakian Karst Museum Skolska 4, Liptovsky Mikulas Slovakia 3. Institute of Geological Sciences Jagielonian University, ul. Oleandry 2a Krakow Poland 4 Institute of Geology Adam Mickiewicz University ul. Mak6w Polnych 16 Poznan, Poland 5 Institute of Geology Bergen University Allegtn 41 5000 Bergen Norway In a lot of books Droppa's scheme of Demianovska Karst System development was used as an example of multile v el karst system which may be correlated with a river tarrases on the surface. We have collected samples from the Lodova Cave which belongs to the IV level of Droppa's scheme. Basing on the dating results we can distinguish 4 periods of speleothems deposition : older then 350 ka ; 140-190 ka; 70-108 ka; younger then 6 ka The age of the oldest speleothems in this part of Lodova Cave allow us to estimate minimum age ofLodova Cave. Contemporary we will be able to estimate the age of IV level of Demianovska Karst System developped by Droppa The 234U / 238U method (RUBE method, IVANOVICH & HARMON, 1992) was used for estimation of the age for the oldest flowstones from Lodova Cave 234U/238U age of the basal part of the oldest flowstone from Lodova Cave showed that 700 ka ago Lodova Cave was dry and speleothems deposition was possible It means that the IV level of Demianowska System is much older then Droppa sugested basing on correlations with surface sediments Introduction The Demiano v ska Valley's Karst System ( D ROPPA 1957) is one of t h e most famous of the World In a l ot of books Droppa's scheme of its development is used as a example of multilevel Karst System which may be correlated with river tarraces on the surface. Until present no systematic work with radiometric dating of the speleothems from this system was done. Dating of speleothems from the Lodova Cave gave us the first oportunity to make a time scale for Droppa 's scheme (DROPP A, 1963 1964 1970 1972). We have collected 7 samples of speleothems from ,,Zruseny Dom" in the Lodova Cave This passage belongs to the IV level of Droppa 's scheme Method Standard radiometric dating procedure of 230 Th/ 234 U method were used (IVANIVICH & HARMON, 1992) Samples of 1530 g were dissolved in c a 6 M nitric acid Uranium and Thorium fractions were separated by chromatography method 234 U m u no Th and 132 Th activities were measured by using isotope dillution with 228 Th / 232 U spike. All measurements were done with alfa spectrometry The ages were calculated by standard algoritm (IVANOVICH & HARMON 1992) using program Age04 (LAURITZ EN, 1981 ). Reported errors are I sigma. Results Basing on the dating results we can distinguish 4 periods of speleothems deposi t io n (Fig. I). I Older then 350 ka 2 140-190 ka It may be correlated with the warm interstadials within 1 R O stage 6 3 70-108 ka which correlate with 1 R O stage 5. More precis l y it may be correlated with substage 5c and 5a 4 Younger then 6 ka which correlated with 1 R O stage I. The age of the oldest speleothems in this part ofLodova Cave is the m ost interesting. This age allow us to estimate the minimum age of Lodova Cave. Contemporary we will be able to estimate the age of IV level of Demianovska Valley Karst System developed by Droppa. The 234 U / 23 U method was used for estimation of the age for the oldest flowstones from Lodova Cave (Jlod 4 >350ka from 230 Th/ 234 U method). Calculated initial 234 U / 238 U ratios for the younger samples were used for estimation of initial 2).l u r u ratio in the flowstone Jlod 4 (RU B E method, l VANOVICH & HARMON 1992) We used the mean value of all the calculated initial 234 U / 3 R U ratios and obtained estimator of initial 234 U /m U ratio in the sample Jlod 4 equal 2 62.38. Results of 234 Uf m U dating of sample Jlod 4 are presented in the ta b le I. We can estimate that the period of deposition of the oldest speleothe m s in Lo d ova Cave was b etween 400 and 700 ka It may be correlated with the O stages 11-14. The calcite from the flowstone Jlod 4 and c.a. 8 cm of sediments covered by this flowstone have normal magnecity what confirm that t h ey are yo un ger then B run h es / Matuyama boundary. Symposium 7 : Physical Speleology 85


Lodo~ a Cave 1 \ I I I I I I l 100 200 300 -2.0 9 co ..... I 0.0 0 2.0 "'1----~--~--~--~--~------~----I 0 100 200 Age (ka) 300 Fig. 1. U-series dating results and theirs correlation with oxygen record Sample 234 LJf38U age fkal +60 Jlod 4/1 685 bottom -40 +73 Jlod 4/2top 410 -46 400 Table 1. Uranium-Uranium dating results of the oldesst flowstone from Lodova Cave Basing on the oldest data from Lodova Cave we can make a first test of Droppa's scheme. In the Droppa's scheme Lodova Cave is the part of rv level of Demianovska Karst System He suggested for this level the Mindel Age (DROPPA, 1970 1972). The age of the oldest speleothems from Lodova Cave show that 600 ka ago this cave was dry and the speleothems deposition was possible It means that the IV level of Demianovska Karst System is much older then Droppa sugested basing on correlations with surface sediments. References DROPPA A. 1957. Demanovske jaskyne Krasove javy Demanovskej doliny. SAY. Bratislava pp. 289 DROPPA A. 1963. Paralelizacia riecnych teras a horizontalnychjaskyn Geolog. prace GUDS c. 64: 93-96. DROPPA A. 1966. Correlation of some horizontal caves with river terraces Studies in Speleology. vol. l : 186-192 DROPPA A. 1970 Yyskum riecnych teras v zatopovej oblasti Liptovska Mara. Ylast. zbomik Liptov. I: 7-34. DROPPA A 1972. Geomorfologicke pomery Demanovskej Doliny. Slov. Kras X : 9-46. IVANOVICH, M & R. S HARMON 1982 Uranium Series Disequilibrium. Applications to environmental problems Clarendon, Oxford pp 571. LAURITZEN S. E 1981. "Age4U2U". Program for reading ADCAM energy spectra, integration peak-correction and calculation of 230 Thorium/ 234 Uranium ages. Computer Program Turbo Pascal Code, 5 000 lines Department of Geology Bergen. 86 Proceedings of the 12 th International Congress of Speleology, 1997, Sw itz erland Volume 1

PAGE 100

Speleothem growth frequencies (PDF) as a climate reco r d : problem of significance Helena Hercman Institute of Geological Sciences, Polish Academy of Sciences, ul. Twarda 51/55, O 1-116 Warszawa Poland Speleothems provide a sensitive tool for studies of past climatic changes as the periods of their growth coincide with relatively warm and humid episodes, while breaks in calcite deposition correspond to cool phases It means that e v en the presence or absence of the speleothems is the source of the paleoclimatic data. The growth of individual speleothems may be affected by local factors On the other hand the collective properties of speleothems growth may cancel out local effects If we have an enough large sample the distribution of the speleothems ages through time should aproximate the volume of calcite deposited through time. It was noted quite early that speleothems deposition at least in upper-mid latitudes is discontinuous in time clustering into distinct groups that correlate broadly with known warm and humid periods of the Upper Pleistocene. This discontinuity of the speleothem record has been explained by climatic control. Probability density functions (PDF) for speleothem growth are calculated from the scewed PDF for each date using the original 230Th/ 234 u and 234 u ; 2 38u ratios with the corresponding analytical errors. Filters may be used to screen out dates with low precision. As a result we obtain a continuous record which aproxirnate speleothem growth frequencies If we want to use PDF as a climatic record we should try to answer some questions. Maybe it is not a paleoclimatic record Maybe it is only the record of sample collection. We can test it by random elimination of a part of our original data. If the positions of the peaks are stable, PDF is more than record of sample collection. The most important question and the most difficult too, is about the significance of the record Which peaks are significant? P Smart and his team in an article about North-West European PDF used Monte-Carlo experiment for the estimation of the peaks significance They generated I 00 times a set of random data which had uniform distribution through time and quantity and accuracy of data similar for their original data For all data sets they constructed PDF and as an estimator of significance bands for their original PDF they used borders of 95 % of the range of the experimental PDFs changes The peaks which are above and minimas which are below these bands are significant. In fact this test may be only used as a test if distribution which is represented by our PDF is not a uniform distribution It doesn t tell us anything about significance of the individual piks In all regional age frequency distributions of speleothems the Holocene signal dominates and frequency decreases with time This is due to two effects First, Holocene speleothems dominate in caves. Second speleothems tend to be destroyed over time, and therefore older specimens are rarer than younger ones Then maybe we should use in our Monte Carlo experiment random data with a different distribution through time than uniform, for example exponential distribution. But the result will be the same We can only tell that our distribution is different then exponential. A little more information can be obtained from bootstraping of our data The bootstrap is a data-based simulation method If we assume that our PDF is a reflection of paleoclimatic conditions we can investigate how big will be the differences between the diffrent estimations. We replace our original data set by another data set which have the same quantity of data, and construct the PDF for these data. We repeat this procedure a large number of times say 500 or I 000 Then basing on the 500 ( 1000) PDFs we can estimate a confidence band for our original PDF. Another source of information is the analysis of subpopulations in PDF record. P Smart and his team used a simple algoritrn for estimation of subpopulation parameters. The same algoritms may be used for estimation of the quantity of subpopulations too If we try to estimate the parameters of subpopulations for a different quantity of maxima (I 2, 3 .. n) we can find the best solution : the quantity of subpopulations which give us the best fitting. The next source of information is multicorrelation between a PDF record and the other paleoclimatic records The result of these analyses may be used as a base for modeling of influence of different climatic factors on the PDF record All these calculations are in progress and the final results will be presented on the Congres sessions Sympos i um 7 : Physical Speleology 87

PAGE 101

A Long Interglacial and lnterstadial Record in th e Nor t h American Mid-Continent from Vadose and Phreatic Speleothems Jewel Cave and Wind Cave South Dakota U.S.A. Derek Ford, Henry Schwarcz, Hilary Stuart-Williams and Nic o la Swinb u rne, Geography and Geology, McMaster University, Hamilton, ON L8S 4KI and J o yce Lu n dberg, Geogra p hy, Ca rl e ton Un i vers i ty, Ottawa, O N KlS 5B 6 Geographic location Jewel Cave and Wind Cave are located in the Black Hills, an isolated domal structure in the centre of North America. The region was not glaciated but experienced cold winters. The modern climate is at the sub-humid/semi-arid transition with grasslands above Wind Cave ( m above sea level) and open pine forest at Jewel Cave ( m asl). Spe l eogenesis Jewel Cave and Wind Cave are multi-storey, rectilinear dissolutional maze caves formed by thermal waters rising through Mississipian (Lower Carboniferous) limestones and dolomites in the Black Hills, a domal structure in South Dakota. The caves are now drained and relict. Jewel Cave is encrusted with 6-15 ems of euhedral calcite spar precipitated from the thermal waters. Wind Cave has much tinner crusts, including its celebrated boxwork, and calcite raft deposits indicating paleo-watertable levels Vadose precipitates such as stalactites and stalagmites, are spare in both caves. Isotope records in sample precipitates have been studied intensively for their paleoclimate records. The thermal water calci t es U series and palaeomagnetic measurements have established that the great spar sheets of Jewel Cave are older than 2 million years and therefore not suitable for Quaternary paleoclimate studies. In Wind Cave the thickest sample WCMAJ, is only 2.4 ems Eight U series age determinations by thermal ionisation mass spectrometry (TIMS) that it great from 330 to 160 ka BP, when the water table fell below its position. 8 1 8 0 and 8 13 C profiles display two distinct halve: (i) 160-225 ka, where O and C are negatively correlated and there id 3%o variation in 18 0 record the climate of Isotope Stage 67 (i) between 225 and 325 ka BP both isotope signals are heavier and uniform. In contrast, the U content and m u ;m J ratios display a clear glacial-interglacial cycle over this time interval. The homogenising of the 18 0 and 13 C signals suggest an unknown geothermal effect which erases the paleoprecipitation signal. Core must be taken when interpreting paleoclimate records in thermal water calciates. Vadose calcite In Jewel Cave sample JC! I is a vadose calciate flowstone 12 cm thick, displaying strong colour banding. From twelve TIMS U series dates, the upper 4 ems grew at intervals between 92 and 470 ka BP. Eight JO-point growth layer analyses of O and C ratios were made to check for isotope kinetic fractionation effects. Four different profiles have been measured between 92 and 240 ka, with a highest resolution of 300-500 years Marine isotope Termination II is signalled by 18 0 increase of 3.0%0 between 131 and 129 ka. Isotope stage 5e lasted from 129 to 119 ka. At our scale of resolution it contains four or more excursions of 0.5%0 Growth in the previous interglacial (equivalent to marine isotope Stage 7) displays abrupt shifts of up to 4%o in the 18 0 signal; peaks occur at -210, -192 and -178 ka BP. All growth ceased during the coldest parts of the glacial stades. Between 300 and 470 ka BP, 8 1 8 0 interglacial peaks are also in good agreement with the SPECMAP isotope records from deep ocean cores; indicating that thin speleothem in the heart of the continent is recording global climate change in the same manner. The lower 8 ems of the sample are > 500 ka in age They display the same patterns of colour banding growth hiatuses and stable isotope cycles as occur in the younger calcite. We are attempting to date these patterns by RUBE and U/Pb techniques 88 Proceedings of the 12' h International Congress of Speleology, 1997, Switzerland Volume 1

PAGE 102

Precise measurement of luminescence banding profiles in speleothems for paleoclimatic interpretation William B. White Department of Geosciences and Materials Research Laboratory The Pennsylvania State Uni v ersity University Park PA 16802 USA Abstract It i s demon s trated that the microfocus Raman spectrometer is an ideal instrument for the measurement of luminescence banding in speleothems. The excitation source is an argon ion laser. The selected laser line is brought into a micro s cope through a beams pl i tter and brought to a focus on the microscope stage with a spatial resolution of 2 3 micrometers Luminesc e nc e radiation from th e illuminated area is collected by the microscope transmitted through the beam splitter to the entrance slit of a 0 75-meter focal length double monochromator. Luminescent intensity is measured with a photomultiplier interfaced with a computer for data proces s ing and analysis The luminescence wavelength can be set to any desired value and intensity measured as a function of tra v er s e distance by moving the specimen with a precision x-y translator on the microscope stage. Complete spectra for each band can al s o be obtain e d Records obtained with this de v ice are s uitable for input to mathematical processing such as spectral analysis and Fourier tran s form s. 1. Introduction There is a long and rather complicated history that has led over a period of more than 20 years to a realization that the well known luminescence of speleothems may contain a useful and very high resolution paleoclimatic record A starting point is the report by GASCOYNE ( 1977) at the Sheffield International Congress that he could find no correlation between the color of s peleothems and the iron content. GASCOYNE proposed that the pigmenting materials were organic molecules incorporated within the structure WHITE ( 1981) by reflectance spectroscopy was able to support GASCOYNE'S conjecture. Later LAURJTZE e t al ( I 986 ) actually extracted the organic component and showed that it was a mixture of humic and fulvic acids It was shown as expected that the blue-white luminescence of speleothems was also due to incorporated organic molecules predominately ful v ic acid (WHITE, 1986; WHITE A D BRENNAN 1989). Independently and in about the same time frame, YA VOR SHOPOY began reporting on his observation s of high resolution luminescence banding within the speleothem structure. This was a new idea ; previous workers had focused their attention on the bulk properties of speleothems Shopov s howed that the intere s ting and useful information was contained in the spatial distribution of the organic constituents With these observations in hand, many investigators have converged on this problem (SHOPOV et al 1994; BAKER et al. 1993 ; BAKER e t al I 995). Spatial transects along the axis of stalagmites of the alternating bands of bright luminescence could be combined with Uffh dating profiles to show that the width of the individual bands indeed represented one year of growth Twenty y ear s after GASCOYNE'S initial observations, it seems quite clear that the flowstone and stalagmite record does indeed have the potential of being a paleoclimate record of great value. The objecti v e of the present paper is to develop a methodology for obtaining a more quantitative record of the luminescence banding and one that would be suitab l e for more advanced forms of mathematical analysis 2. Instrumentation The proposal is to make use of a microfocus Raman spectrometer as a device for obtaining high quality records both from spatial resolution and in terms of spectroscopic accuracy The work ofSHOPOV e t al (1994) BAKER et al (1993) and BAKER AND SMART ( I 995) shows that the width of the luminescence bands is on the order of a few tens of micromet e r s. This means that the measurement device should have a s patial resolution of a few micrometers or about one tenth of tbe width of a band so that the luminescence profile s of indi v idual gro w th bands may be observed Figure I illustrates the optics of microfocus Raman spectrometer. The excitation source is an argon ion la s er The beam passes through a Pellin-Broca prism to filter extraneou s plasma lines through some focusing optic s, and then i s brought into an optical microscope It then passes through a beam s plitt e r which transmits about I 0 % of the beam inten s i ty down the optical path to the microscope objective. The beam emerge s from the microscope objective and is brought to a focu s on th e s ample with a focal spot I 2 micrometers in diameter. The luminescent radiation emitted from the a I 2 micrometer volume of the sample is collected by tbe objective lens and brought up through the microscope to the beam splitter which is so arranged that 90 % of the luminescent light travels to the entrance s lit of the monochromator. Tbe dispersing optics in thi s in s trument con s i s t s of two tandem 0 75-meter focal length Czern yTumer mounted grating monochromators which provides a w av elength re s olution of 0 05 nm The monochromator system provide s considerabl y better resolution than is required for the broad emission from fulvic acids. The light exits the monochromator through a slit and then is focused on a photomultiplier which converts it into an electrical signal which passes through some interface electronic s to a computer. The computer also controls the wavelength dri ve system so that data points can be collected by counting light intensity at wavelength intervals as small a s 0 1 nm This allow s the collection of data with very high accuracy in both wavelength and intensity. The data are also in digital form suitabl e for further manipulation The existing optical system provide s high s pectroscopic accuracy The additional modification that must be made is a translation stage on the microscope with the ability to control position of a slice of speleothem with an accuracy comparable to the existing spatial resolution of the microscope a few micrometers. This can be provided by a mechanical translator system with a precision worm drive. Data can be taken point by point by advancing a micrometer screw on the translator or as the next step replacing the hand adjustment w i th a stepper motor controlled by the computer With this arrangement the instrument can be used in either a wavelength mode where the beam is Symposium 7 : Physical Speleology 89

PAGE 103

Ob aerv tl on Screen Lena ,.., 'It N I I I I >: \ II I Cl) : z : Xe=417 WI I II I z. ' ID \ (J) ', N 250 '. I ------... 0 Cl) ,.., I 82-MM-004 FLOWSTONE PEACOCK CAVE, WV X 11 = 241 focused on a specific luminescence band and entire spectrum scanned or one can set the monochromator at the em.Jss10n wavelength of interest an d then translate the specimen to map the intensity of luminescence as a function of a x-coor d inate a l ong the translation direction The advantage of this type of instrumentation is in its great flexibility It's disadvantage is that it is not convenient to use with very large s pecimens so that if one wished to measure the l uminescence profile along a 2-meter segment of broom handle stalagmite it would be rather difficult. 3. Luminescence Spectroscopy As observed visually or recorded on photographic film, typical speleothem luminescence consists of a bluish white to greenish white emission sometimes extending to t h e orange yellow region of the spectrum When examined in detail there is much more to speleothem luminescence than alternating bands of bright and dim emission. The luminescence spectrum of a typica l speleothem consists of two parts (Fig 2) The emission spectrum is obtained by illuminating the sample with ultravio l et light of a specified wavelength a n d measuring the wavelength d istribution 600 Fig ur e 1. Op t i cs of mi c r o f oc u s R am a n s p ec trom e t e r Fig ur e 2. E m iss i o n a nd excit at io n spectra of speci m e n 82-MM-0 0 4, a flows t o n e s ampl e fr o m P e a cock Cave, West Virginia, USA of the emitted radiation The excitation spectrum is measured b y tracking the intensity of the luminescence as a function of the wavelengt h of the ultraviolet used to excite the luminescence The brightest and most efficient l uminescence is therefore obtained when the sample is illuminated with 243 nm UV with a secondary excitation at 296 run ln this s pecimen there is a single emission band peaking at 380 run The spectral line shapes are approximately gaussian The argon ion laser has lines at 457 488 and 514 5 nm all of which l ie at wavelengths greater than the wavelengths of the excitatio n bands However, the laser is highly inten s e and sufficient luminescence for measurement can be obtained by exciting into the long wavelength tail of the excitation band The laser lines are also at longer wavelengths than the emission band and so cannot excite the main portion of thi s band The emis s ion observed from Ar-ion laser excitation will be due to a different set of luminophors than those excited by UV radiation This is a difficulty that could be resolved by using a UV laser as an excitation source. Chromatographic separation of the organic constituents of dissolved speleothems (WHITE, SMAILER AND BRENNAN unpublis h e d data) show that the luminescent species in 90 Proceed i ngs of the 12 th International Congress of Speleo l ogy 1997 Sw i tzerland Volume 1

PAGE 104

,_ >1"' z_ WIc z uo z:: w ii u~ en o wa:: 0 :::, .J lL 460 500 100 540 5 80 6 20 6 60 700 WAVELENGTH ( nm I speleothems consists of a mixture of molecules each of which has its own characteristic emission and excitation spectrum. As a result the emission peak changes with wavelength of excitation. For the specific specimen used as an illustration, the main UV excited emission band has its peak intensity for the molecular weight fraction in the range of 5000 I 0,000 daltons (Fig. 3). The emission bands excited by the 457.9 nm line of the Ar ian laser in the microfocus spectrometer (Fig 4) occur in the range of 550 560 run much greater wavelength than the UV excited band shown in Fig. 2. The bands are broad but there are measurable shifts in the band positions Fig . 4 shows a sequence of spectra obtained from a sequence of luminescence bands in the same samples used for previous spectra There is some variability in the spectra from one band to another but because the bands are very broad, there is a great deal of overlap Because of the I 2 micrometer spatial resolution of the laser spot, the luminescence banding can be mapped at any chosen emission wavelength 740 140 120 100 10 so ,o ,o 4. Conclusions Figure 3. T h ree-di m ensional plot of l umin escence intensity as a function of bot/, emission wavelength and molecular weig h t. Eac h tick mark on the V, (elutriated volume) scale represents a sep a r a te chro m atographic cut and a separately measured emission spectrum. Figure 4. Laser excited (457.9 nm li n e) em,sswn spectra from individual growth bands in st a lactite 88-MM-004. The rnicrofocus Raman spectrometer is shown to be a useful device for the accurate mapping of the distribution of speleothem luminesce n ce. The spatial resolution is a few micrometers smaller than the annual growth bands. The emission wavelength to be mapped is an adjustable variable The instrumentation and technique could be easily extended to the use of a tunable UV laser as a source, thus adding the excitation wavelength as an additional adjustable variable Acknowledgment This work was supported by a grant from the Petroleum Research Fund administered by the American Chemical Society Symposium 7 : Physical Speleology 91

PAGE 105

References BAKER, A ., P. L. SMART, R. L. EDWARDS, & D A RICHARDS, 1993 Annual growth banding in a cave stalagmite Nature 364: 518-520. BAKER, A. & P L. SMART. 1995. Recent flowstone growth rates : Field measurements in comparison to theoretical predictions. Chem. Geol. 122 : 121-128. BAKER, A., P L. SMART W. L. BARNES, R. L. EDWARDS, & A. FARRANT, A 1995. The Hekla 3 volcanic eruption recorded in a Scottish speleothem. Holocene 5: 336-342 GASCOYNE, M 1977. Trace element geochemistry of speleothems Proc. 7th Internatl. Congress Speleol., Sheffield, 205-208 LAURITZEN, S -E., D C. FORD, & H. P SCHWARCZ, 1986. Hurnic substances in speleothem matrix Paleoclimatic significance. Pro c. 9th Internatl. Congress Speleol., Barcelona, 2: 77-79. SHOPOV, Y. Y., D. C. FORD, & H P. SCHWARCZ. 1994 Luminescent microbanding in speleothems: High-resolution chronology and paleoclimate. Geology 22: 407-410. WHITE, W. B. 1981. Reflectance spectra and color in speleothems. Natl. Speleol. Soc Bull. 43: 20-26. WHITE, W. B 1986. Luminescence of cave calcite deposits: A current appraisal (abs.). Natl. Speleol. Soc Bull. 48: 40 WHITE, W. B & E. S BRENNAN. 1989. Luminescence of speleothems due to fulvic acid and other activators Proc 10th Internatl. Congress Spe/eol., Budapest, 212-214. Evidence of glacial advances from New Zealand cave deposits Paul W. Williams Department of Geography, University of Auckland, Auckland, New Zealand Abstract Caves overrun by glaciers are known to accumulate dateable evidence of past glacial and interglacial events. Results are reported from investigations of Aurora Cave in Fiordland, southern New Zealand, and from other caves. Aurora Cave is situated in the side of a glacial trough. The cave commenced to form before 230 ka B.P. Sequences of glacifluvial sediments interbedded with speleothems are evidence of the number and timing of glacial advances and the status of intervals between them. Twenty-six uranium series dates on speleothems underpin a chronology of seven glacial advances in the last 230 ka, with the peak of the last main glacial advance at ea. 19 ka BP. With five advances in the Last Glacial, it is more complex than previously recognised in New Zealand. Comparison of the evidence fom caves with that recorded in deep sea sediments 300 km offshore from DSDP Site 594 reveals little matching. The results from caves support polar cap ice-core evidence for abrupt climate changes, but do not support synchroneity of onshore / offshore events. 92 Proceedings of the 12 th International Congress of Speleology, 1997, Switzerland Volume 1

PAGE 106

L oxygene et le carbone dans les concretions nous informent-ils sur les climats passes? par Sophie Verheyden Aspirant du fonds National Beige de la recherche Scientifique (NFWO), Vrije Universiteit Brussel, WE-CHRO, Pleinlaan 2, 1050 Brussel, Belgie E. Keppens, Vrije Universiteit Brussel, Y. Q u inif, Faculte polytechnique de Mons Guy Deflandre Abstra ct: Does the oxygen and carbon isotope record in speleothems register climate? Stable isotopes of oxygen and carbon in speleothems theoretically are related to respectively the surface temperature and the vegetation present during calcite precipitation In fact, this relationship seems more complex. The study of recent calcite and seepage water, as well as in situ measurements of climatic parameters suggest that other processes e.g. evaporation degassing interaction of the host rock . affect dramatically the oxygen and carbon isotopic composition of the speleothems R e s ume Theoriquement l'oxygene et le carbone dans les concretions nous informent sur respectivement la temperature de surface et Ja vegetation presente !ors de la precipitation de la calcite En pratique, ces relations paraissent plus complexes. L'etude de concretions recentes et de l'eau d'infiltration ainsi que des mesures de parametres climatiques dans les grottes belges remettent en question Ja relation theorique entre le climat et l'oxygene et le carbone dans Jes concretions Cette etude montre que des processus plus locaux comme !'evaporation le degazage !'interaction de la roche environnante, .. peuvent influencer de far;on drarnatique le signal isotopique de l'oxygene et du carbone dans Jes concretions 1 Introduction Cette etude fait partie du projet europeen A high resolution reconstruction of terrestrial palaeoclimate from continental carbonates" qui a pour but de retracer le climat Holocene (10000 0 ans BP) grace a l'etude de J'oxygene et du carbone dans Jes concretions de divers sites europeens. Nous presentons ici Jes resultats du site beige. La grotte du Pere Noel Fait partie du systeme de Han-sur Lesse comprenant entre autres la grotte de Han et le Pere Noel. Elles sont situees dans la bande de calcaire Givetien la Calestienne, qui traverse la Belgique d Est en Quest (fig I). On y trouve la plupart des phenomenes karstiques belges Le Pere Noel est le resultat d un recoupement de meandre de la Lesse a travers le Massif de Boine (QUINIF 1977). Denos jours cette cavite est entierement fossile et unimportant concretionnement s est mis en place depuis plus de I 00000 ans. Une stalagmite de 65cm de Jong a ete prelevee afin d'etudier Jes changements de composition isotopique le long de !'axe longitudinal (fig 2) La datation Urfh indique que la concretion s est deposee entre 13000 et 2000 ans BP (Before Present). Elle couvre done en partie le Tardi-glaciaire et !'Holocene. 2. Les isotopes stables dans les etudes du climat passe Dans la nature ii existe trois oxygenes de masses differentes : un oxygene de masse 16 ( 0) (c a d. I 6 fois la masse de l'hydrogene), un oxygene de masse 17 ( O) et un oxygene de masse 18 ( 0). Ce sont Jes isotopes de l'oxygene Aucun de ces elements n'est radioactif ils soot done tous stables Le carbone ll ll IJ compte deux isotopes stables, C et C et un radioactif le C Ces isotopes soot presents en certaines proportions (on mesure II 16 13 12 generallement le rapport 0 / 0 et C / C) Chaque processus physique (comme !'evaporation) ou chimique (comme la precipitation de la calcite) change la proportion des isotopes Pour faciliter les calculs on emploie Jes notations 8 18 0 et 8 13 C (exprimes en %o) Ils traduisent la difference entre le O /" O de la matiere etudiee et une matiere de reference. La matiere de reference (ref) dans le cas de la calcite est le PDB le "Pee-Dee Belemoite", un fossile de la Formation de Pee-Dee en C a roline du Sud. Plus Jes o soot eleves plus ii y a d isotopes Jourds (c.a d O ou 11 C) dans la calcite (c) (HOEFS 1997) o Oc = [( O !' 0 O)c (" O /' 0 O)ref]* I 000 / ( O !' o)ref Grace a l'etude des isotopes de l'oxygene dan s Jes depots marins et Jes glaces polaires nous connaissons bien les grandes variations climatiques a l'echelle mondiale. Ces etudes nous ont Fait mieux comprendre Jes cycles froid-chaud ou glaciaire interglaciaire. L' analyse de l'oxygene dans Jes tourbieres et dans Jes depots Jacustres, ainsi que dans les concretions nous informent sur le climat continental plus local et a plus petite echelle de temps (cycles annuels ou saisonniers). 3 Le 0 18 0 et le 0 13 C dans les concretions Pour bien comprendre le mecanisme des isotopes stables dans les concretions, rappelons-nous la formation d'une concretion. L 'eau de pluie percole dans le sol et se charge de CO qui est originaire de la respiration des plantes Ensuite l eau s infiltre dans Jes interstices du calcaire sous-jacent et dis s out le CaCO ,. Rencontrant un vide l'eau degaze du C0 1 et precipite de la calcite Theoriquement le rapport de l'oxyge n e dans la calcite (CaC!b) dependra de la te mp erature a laquelle la calcite s'est deposee et de la composition isotopique initiale de J'eau d ou la calcite precipite Dans Jes deux cas le o O est directement lie a la temperature de surface : dans les grottes belges qui sont peu profondes la temperature en grotte est une moyenne annuelle de la temperature de surface (W1GLEY ET BROWN 1976) Symposium 7 : Physical Speleology 93

PAGE 107

la compos1!Ion isotopique de l 'eau de percolation est comparable a celle de l'eau de pluie. La composition de LEGENOE DJ CRETACE CRAYEIJX JURASSIOUE MOYEN LIAS TRIAS POUllNGUE PERMIEN CALCAflE CAR~IFERE 0 CALCAIRES OU OEYONIEN MOY ET SUP 25 l eau de pluie est en relation directe avec la temperature de surface (DANSGAARD, 1964). N Fig. J: Les formations carbo11atees de la Belgique. Han-sur-Lesse se trouve dans la Ca/estienne (fleche). D 'apres C. EK, 1976. On peut done en analysant le 6 1 8 0 dans la concretion retrouver la temperature qui regnait en surface !ors de la precipitation de la calcite Le rapport du carbone dans la concretion (Ca!:;.O3) depend beaucoup moins de la temperature mais est directement lie au type de vegetation present a la surface !ors de la precipitation de la calcite Selon le cycle de photosynthese utilise, Jes plantes ont un 6 13 C different. Le cycle Calvin est utilise par les plantes dites C (dans un premier temps elles produisent une molecule a trois J atomes de carbone) En Belgique, toutes les plantes appartiennent a ce groupe, sauf le ma"is. Le cycle Hatch-Slack est utilise par les plantes dites C ,. Ces plantes vivent dans des regions arides et chaudes Ies herbes de savannes sont souvent des p l antes C ,. Le 6 13 C de la calcite sera different selon le type de plantes present en surface lors de la precipitation de la calcite (CE RLING I 984). II y a neanmoins une condition pour pouvoir utiliser ces equations theoriques: c est l'equilibre isotopique Si la calcite se depo se de fai;on lente et sans perturbation s les changements isotopiques sont connus grace a des experimentations en laboratoir e et on peut done retrouver la temperature en faisant Ies mesures de 6 0 Si la precipitation est perturbee par un degazage trop rapide ou de I' evaporatio n ces relations theoriques n e sont plus valables (HENDY 1971 ) 4. La courbe paleoclimatique Une mesure de 6 18 0 et 6 13 C a ete faite tout les S mm le long de !'axe Iongitudinale de la stalagmite (fig. 3). La courbe d'oxygene (6 0) ainsi obtenue devrait done theoriquement retracer les changements de temperature entre 13000 et 2000 ans BP La courbe de carbone devrait nous indiquer le type de vegetation present durant cette periode. Une rapide observation de ces courbes (fig. 4) montre des variatio ns dans le 6 0 et le 6 1 3 C tout au long du Tardi-glaciaire et de !'Holocene. Ceux-ci pourraient etre en relation avec d es cbangements climatiques Les va riations sont neanmoins le s memes pour l e 6 0 et le 6 C II y a done co-variation Cela suggere qu'un seul parametre est a l'origine des variations et que ceux-ci ne traduisent done pas des cha n gements de temperatures ou de vegetation. De plus les variations importantes du 6 C sont tres inattendues puisque la Belgique n a pas connue de plantes C durant cette periode (sauf le ma"is dans les region s cultivees). Finalement, des mesures de 6 0 et 6 C le long d une meme !amine de croissance montrent que la majeure partie de la stalagmite a ete deposee en non-equilibre isotopique (HENDY, 1971) 94 Proceedings of the 12 th Internat ional Congress of Speleo l ogy 1997, Switzerland Volume 1

PAGE 108

Fig. 2: La atalagmite pre/evee dans la grotte du Pere Noel a Han-sur-Lesse est environ 65cm de /011g. Les premiers resultats d'une etude in-situ mettent en relation un 8 0 et 8 C eleve avec un debit de percolation lent. Le signal isotopique de l'eau de percolation reste toutefois inchange par rapport au signal isotopique de l'eau de pluie La perturbation se situe done pendant la precipitation de la calcite. Une evaporation accrue de l eau (H Q) pendant la precipitation de la calcite entrainera une augm~ntation du 8 13 C de la calcite mais aussi une augmentation du 8 13 C En effet une evaporation de l'eau va de pair avec un degazage de 2 plus rapide ce qui entraine une augmentation du 8 C de la calcite Puisqu il y a perturbation de l'equilibre isotopique Jes equations theoriques ne peuvent pas etre employees dans ce cas-ci On peut toutefois supposer que pendant Jes periodes plus humides le 8 0 ainsi que le 8 13 C diminuent alors que pendant Jes periodes plus seches, avec une evaporation plus importante its augmentent. Done, si la stalagmite ne peut pas nous foumir des indications sur les changements de temperature et de vegetation elle peut quand meme nous foumir des indications sur la pluviometrie dans le passe A grande echelle pourtant on peut supposer que ces perturbations ont mains d'effet puisque on considere la moyenne Fig. 3: La stalagmite de la grotte du Pere Noel a Ha11-sur-Lesse e11 Belgique est sciee afi11 d'obtenir une tra11che du milieu de la stalagmite. U11 echantillon est pre/eve tout les 5111111 le /011g de /'axe /011gitudi11a/ de la co11cretio11 aji11 d'etudier les changements de la compositio11 isotopique ({/ 8 0 et 8 13 C) de la co11cretio11. de longues periodes. Ainsi le 8 0 augmente de I % 0 tout au long de la croissance de la stalagmite En tenant compte du changement plus ou mains connu du 8 0 de l eau de pluie durant cette periode on peut estimer un changement de temperature de 5 a 6 C entre 13000 et 2000 ans BP 5. Conclusion Les etudes de la composition isotopique de l'oxygene et du carbone d'une stalagmite Holocene beige montre que dans ce cas Jes changements de 8 '" 0 et 8 C ne retracent pas Jes changements de temperature et de vegetation comme theoriquement attendu. Des mesures in situ laissent toutefois supposer que les changements de 8 0 et 8 13 C peuvent etre mis en relation avec des changements de pluviometrie. A long terme la courbe de 8 0 suggere quand meme une augmentation de la temperature de 5 a 6 C entre 13000 et 2000 ans BP. Symposium 7 : Phys i cal Speleology 95

PAGE 109

Pere Noel: dO and dC (smoothed) -3 ,------------------------------------~ 1 -4 5 = Q 7 0 -8 "CJ -9 10 II 2 3 -4 5 6 7 Cl, (j 8 i 9 -dO 12 1--------------------------------------__J10 0 2000 4000 6000 8000 I 0000 I 2000 14000 -dC Interpolated Uffh age (year BP) --genern trend Fig. 4: En representa11t la composition isotopique (8 18 0 et 8 13 C) d'une stalagmite en fonction de /'age de chaque echantillon, on obtient une courbe paleoclimatique. la stalagmite prelevee dans la grotte du Pere Noel a Han-sur-lesse, Belgique s'est deposee entre 13000 et 2000 ans BP. le 8 18 0 montre tout au long du Tardi-glaciaire et de /'Holocene des variations importantes. A plus long terme, une augmentation du 8 18 0 d'environ 1%o est observee. Le 8 13 C montre egalement des variations considerables dura,11 la croissa11ce de la stalagmite, mais pas de changeme11t a lo11g terme. Je remercie vivement les grottes de Han pour l'acces au site, Dominique Genty Yves Quinif et Eddy Keppens pour !'aide precieuse dans ce travaille, ainsi que ma famille et tous les copains pour les coups de pouce. Les datations U/Th par spectrometrie de masse sont realisees par F Mc Dermott, University College Dublin. Les mesures du 6 18 0 de I 'eau sont realisees par A. Longinelli, Universita di Trieste Ce travail a pu etre realise grace a la Communaute europeenne projet EV5V-CT91-0509 References CERLING TH.E. 1984. The stable isotopic composition of modem soil carbonate and its relationship to climate Earth & Planetary Letters 71 229-240. EK C. 1976 Les phenomenes karstiques 137-157 in Geomorphologie de la Belgique. Hommage au Prof. P Macar ouvrage coordonne par A Pissart Liege, 224p DANSGAARD W. 1964 Stable isotopes in precipitation Tellus 16 436-468 HENDY C H 1971 The isotopic geochemistery of speleothems. The calculation of the effects of different modes of formation on the isotopic composition of speleothems and their applicability as paleoclimatic indicators Geochimica et Cosmochimica acta, 35, 801-824 HOEFS, J. 1997 Stable isotopes geochemistry 4th ed. Springer Berlin 24lp QUINIF Y. 1977 Essai d'etude synthetique des cavites karstiques de Belgique. Revue Beige de Geographie IOI fasc. l a 3 115-173 WIGLEY T M L. BROWN M.C ., 1976. in : The Science of Speleology, eds Ford & Cullingford, Academic Press N Y 329358 96 Proceed i ngs of the 12 'h International Congress of Speleo l ogy 1997 Switzerland-Volume 1

PAGE 110

Paleoclimate of the quaternary processes i n the western karst of Cuba: preliminary results Jesus M. Pajon M. Instituto de Geofisica y Astronomia, Calle 212 # 2906, e/ 29 y 31 La Coronela, La Lisa, Ciudad de La Habana, Cuba CP: 13 600, FAX: (537) 33.94.97 Abstract The main goal of this research project, is to obtain quantitative assesments about the paleoclimatic evolution and the Quaternary processes of Cuba during the last one million years as a contribution to the scientific understanding of global climatic changes in the world, especially in the Caribbean region. It is impossible to understand the impact of human activity on world climate and the environment without knowledge about the natural trends in planetary evolution. In this respect, the study of the actual climate is not enough for such an evaluation and prognosis. The fact that, most of the paleoclimatic investigations have been conducted in countries from temperate or mid latitude zones in the northern hemisphere, with only few and non integrated studies from tropical areas, is one important reason to develop such research. At present, our Project includes: A Paleoclimatic Data Base System of the Cuban Quaternary. GIS of the Cuban Paleoclimate in key sectors of Western Cuba. Characterization of the Cavernary levels of the Western Karst of Cuba, and their relation with sea level flucuations during the Quaternary. Study of paleotemperatures, paleorainfall and paleosalinity in the Western Karst of Cuba, during the last 125,000 years and application to the models of climatic forecast. Knowledge of the Cuban Climate in the present millenium. Map of the Carbon reserves in the soils of Cuba, for the last 3,000 years. The experimental areas of the Project consist of: two karstic cavernary systems (including up to 8 cavern levels and 40 km of underground galleries), one lake in a sedimentation basin and the Marine Terraces System in the North-Western Coastal Area of Cuba. Very important quantitative estimations can be obtained from these studies for Cuba and the Caribbean, with incidence in the planing of a sostenaible development of the society and the preservation of the environment and the biodiversity taking account the prognostic of the tendencies in the temperatures and the sea level changes, of great importance for the development and protection of the marine and terrestrial ecosystems, and the constructive and touristic development, specially in the coastal areas. Resumen El principal objetivo de este Proyecto es, obtener una valoraci6n cuantitativa de la Evoluci6n Paleoclimatica y los Procesos Cuaternarios en Cuba durante el ultimo mill6n de aiios, como una contribuci6n al conocimiento de los Cambios Climaticos en el planeta y especialmente en la region del Caribe. Es imposible evaluar la Actividad e Impacto de! hombre en el clima mundial y en el medio ambiente, sin un conocimiento acerca de las tendencias naturales debido a la evoluci6n planetaria, y en este aspecto el estudio del clima actual resulta insuficiente para ta! evaluaci6n y pron6stico El hecho que la mayoria de las investigaciones paleoclimaticas han sido desarrolladas en paises de zonas templadas o latitudes medias, en el hemisferio norte, con solo pocos estudios integrados en areas tropicales, constituye otro importante aspecto que apunta a la necesidad de desarrollar tales investigaciones. Actualmente, el presente proyecto incluye: Un Sistema de Base de Datos Paleoclimaticos del Cuaternario Cubano. SIG del Paleoclima Cubano en Sectores Llaves de Cuba Occidental. Caracterizaci6n de los Niveles de Cavernamiento del Karst Occidental de Cuba, y su relaci6n con los Cambios de! Nivel de! Mar, durante el Cuaternario. Estudio de las Paleotemperaturas, Paleoprecipitaciones y Paleosalinidad en el Karst Occidental de Cuba, durante los ultimos 125 000 al'ios. Aplicaci6n a los Modelos de Pron6stico Climatico. Conocimiento de! Clima de Cuba en el presente milenio. Mapa de las Reservas de Carbono en los suelos de Cuba, para 3000 y O anos. Las areas experimentales del Proyecto son: Dos Sistemas Cavernarios Karsticos (con 8 niveles de cavernamiento y 40 km de galerias subterraneas); Un Lago en una Cuenca de Sedimentaci6n y el Sistema de Terrazas Marinas en la Costa Noroccidental de Cuba. Importantes estimaciones cuantitativas pudieran derivarse de tales estudios para Cuba y el Caribe, con gran incidencia en la planificaci6n de un desarrollo sostenible de la sociedad y la preservaci6n del Medio Ambiente y la Biodiversidad, pudiendo destacarse el pron6stico de !as tendencias de las temperaturas y las variaciones del nivel del mar, de gran importancia para el desarrollo y protecci6n de los ecosistemas marinos y terrestres y el desarrollo turistico y constructivo especialmente en !as zonas costeras. Symposium 7: Physical Speleology 97

PAGE 111

General Comments The importance of carrying this investigation in Cuba is given by s everal reasons. One of them is its geographical position berwe e n the north American continent and the tropics, specially those island from the Greater and Smaller Antilles, giving an exceptional opportunity to study the influence of continental glaciations on tropical and subtropical territories. Thus, the paleoclimatic information obtained in Cuba will be of regional relevance. Another reason is that, practically all the observed geomorphology in Cuba, was modeled during the Quaternary period in which the variation s in the general sea level, temperature humidity and rainfall ha v e played a mayor role in landscape genesis and evolution The results obtained will contribute to the understanding of the relations berween climate, environment, landscape and time Moreover in Cuba there are many different geological and geomorphologic features, amenable to be used as targets for paleoclimatic tudies as there are hundreds of kilometers of coast with limestone marine terraces at different levels a shallow and extensive submarine platform covered with different generation of fossil and active coral reefs, and different geological formations of quaternary limestone among others. Another peculiarity seldomly found in other countries in the world is the existence of underground ancient tluvial caves excavated at different levels over the actual valleys corresponding to the local base variations, due to the changes of the general sea level, motivated by the advances and recessions of the glaciers in the continents These caves represents fragments of ancients river courses with their bed load sediments perfectly preserved in many places, interlayered also with carbonate cave sediments of chemical origin (stalagmites sinters etc.) which enables their dating using isotope and paleomagnetic methods as well as their characterization from the temperature point of view From another hand, the study of both the granulometric spectrum of the bed load sediments and coarse materials (pebble etc ), as well as the study of the scalloped cave conduits allow estimations of water discharges, runoff and other parameters of those ancient river, and since their basin are well bounded, paleoprecipitations can be also estimated All these ources of evidences are located either just in the same or places not far from each other (remember also that Cuba is not a big country) so that the multivariate characterization of well dated epochs from the near geological past is po ible Therefore there are suitable conditions for field and laboratory investigations with a complex of quantitative methods oriented to different branches of geochemistry geophysics, geology and geography as geochronology atmospheric sciences geomorphology hydrogeology and others which can be integrated in order to characterize the paleoclimates of the Cuban territory during different periods along the Quaternary This characterization will involve paleotemperatures, paleorainfalls, paleosalinities and perhaps other parameters This information together with isotope and paleomagnetic dating will depict more precisely the regional trends on climatic changes, at lea s t for the Cuban territory and its surroundings, as a contribution to the study and understanding of one of the most challenging and important problems of modem science, with a great impact in modem society long term development planing and environment preservation. Some preliminary results: The Paleoclimatic Data Base Systems of the Cuban Quaternary consist in an interdisciplinary network of information about Geochemistry of Sediments and Waters; Paleomagnetism and Magnetic Susceptibility; Stable Isotope Geochemistry and Isotopic Dating ; Sea Level Changes and eotectonic Movement; Paleorainfa ll Paleotemperatures and Paleosalinity; Historic and Present C l imatology; Geo l ogy, Geomorphology, Speleology and Hidrogeology; Cavemary and Marine Terrace Levels. All this information is connected to GIS according the specific experimental area, in order to obtain an integrate and interconnected evaluation. The Majaguas-Cantera Cavernary Systems in the Western Karst of Cuba, show a development of eight underground levels from + SO to + 290 m a s.l. There are rwo active levels (+ 70 m a s l.) one seasonal level (+ 80 to + 90 m a s l.) one episodical level (+ 90 to+ 110 m a.s.1.) and five merofossil levels(+ 120 to + 290 m a.s l.) (MOLERJO 1992 FLORES, 199S). These cavemary levels and the karstification processes, are related with the regional phenomena of the glacioeustatic changes of the sea level during the Quaternary period, as a consequence of the climatic changes The table I show a good correlation berween the cavernary levels of the Sumidero Underground System (ACEVEDO, 1971) and the kaolinite contents of the quaternary sediments deposited inside the caves (PAJON, 1983) FAGUNDO et al. (1984) calculated from a multiple regression equation of absorbance relations in the infrared spectra. The paleoflow, paleodrainage and paleorainfall values was calculate for the Sumidero Region in the Cuyaguateje Basin River, from the scallops analysis in a fluvial fossil gallery of the Perfecto Cave (VALDES, 1974). The paleorainfall value of 1410 mm suggest that the climate of the region during the hydrological behavior of Perfecto Cave (Illinois age-Cuban pluvial Illinois) was more wet that today (700 mm). Epoch of Speleogenesis Cavemary Level Absolute Altitude (m) Kao l inite Content (%) w Pre-Quaternary (Pliocene) Cima 19S-2S0 z ebraska Q Soterraneos Cave 170-19S S 1-S3 w 0 Kansas Q 11 Pio Domingo Cave 130-180 3S-43 0 IIllinois Q 11 Perfecto Cave 100-138 17-22 en iii Wisconsin early (Iowa) Q Clara Cave 8S-132 7 ..J A. Wisconsin late Q" ,v Sumidero 80-130 6 Holoceno Tab l e 1 Correlation between the cavernary levels in the Sumidero Underground System and the kaolinite content of the Quaternary sediments inside the caves. Western Tropical Karst of Cuba K( o/o ) = 7 7882 A 3 I ,+ 46,7491 A 5 / A 6 2S,7192 Where : A 3 = A(92S cm 1 ), = A(808 cm 1 ) A 5 = A(l040 cm 1 ), A 6 = A(l 117 cm 1 ) 98 Proceedings of the 12 '" International Congress of Speleo l ogy 1997, Switzerland Volume 1

PAGE 112

Considering the carbonate, siliceous and gypsum relicts, as well as other geographical index ORTEGA & ARCIA (I 982), proposed a paleorainfall map for Cuba in the last glacial (Wisconsin or Wurm) suggesting an arid climate in that period. On the base of the coastal eolic formations of the Pleistocene in the western of Cuba, SHA ZER et al. (1975) considered that the arid climates coincided with the glaciations KART ASHOV & MA YO ( 1976) establish that in the Cuban Pleistocene took place alternate climatical oscillation of pluvial and arid phases. MA YO ( 1970) established a correlation between pluvial and glacial advance. Based in the weathering profiles of Guane and Guevara Quaternary Formations KARTASHOV et al. ( I 98 I) suggest the division of the Cuban Pleistocene in two stages: a lower wet stage and an upper dry stage. Using the Gates climatological model and the probable level of the periglacial zone, ORTEGA ( 1983) proposed a paleotemperature map on the base of the paleosea mean temperature. Respect to the hi s torical time s the records of rainfall between I 87 I and 1984 in Santiago de Cuba station, show a significative increase, and a decrease between I 750 and 1871. Respect the temperature there is an increment of 0,5 C from the XIX century and 0 2-0,3 in the last 40 year s, in the Casablanca Station of Havana City (CELEIRO 1996) About the sea level change s the Siboney Station (eastern of Cuba), remark a quantitative increase of 2 9 mm/year during 30 years of records (HERNANDEZ, pers comm.). Acknowledgemen t s I would like to express my greatitude to my colleague and friend Lie. Ismael Hernandez for v ery hepful discussions and the composition of this paper. B i bliography ACEVEDO G. M 1971. Geomorfologia de Sumidero y sus inmediaciones, Sierra de los Organos Pinar del Rio Cuba Rev Tee Vol. IX No 3-4 (1971) pp. 33-54 CELEIRO M 1996. Conocimiento de Clima de Cuba en el pre s ente milenio (En prensa) FAGUNDO J.R., VALDES J.J ., PAJON J.M. 1984 Estudio de los Sedimentos Cuaternarios de la Cuenca del rio Cuyaguateje mediante Espectroscopia lnfrarroja y Difracci6n de Rayos X Rev Vol. Hid 63 Ano XXI (1984) 53-61 pp FLORES V E 1995. Niveles de cavernamiento y fluctuacione s glacioeustaticas cuaternarias de Cuba Occidental. Cong lnt. L V Aniv de la SEC (En prensa ) KARTASHOV J.P & MA YO I 976. Esquema e tratigrafico y division del sistema cuaternario de Cuba [en ruso ]. En sedimentaci6n y formaci6n del relieve de Cuba en el Cuaternario, Nauka, Moscu, pp 5-33. KART ASHOV I.P ., CHERNY AJOVSKI A.G ., PENAL VER L. I 981. El Antrop6geno en Cuba [ en ruso ], auka Mo s cu 14 7 pp MA YO N A 1970 Dep6sitos pleistocenicos de lo s cauce s abandonados de la Sierra de los Organos : E v id e ncias de periodos pluviales ACC Ser. Esp y Cars 7, p. 88-89 MOLERIO L.L. 1992. Distribuci6n del cavernamiento en las Sierras del Pesquero San Carlos Resolladero y Mesa Pinar del Rio, Cuba II Cong Esp Lat. y el Caribe Pinar del Rio Cuba, 19-20 pp. ORTEGA S.F & ARCIA M.I. 1982. Determinaci6n de las llu v ias en Cuba durante la Glaciaci6n de Wisconsin mediante Ios relictos edaficos. Ciencia de la Tierra y el Espacio, No. 4, pp. 85-104. ORTEGA S.F 1983 Una hip6tesis sobre el clima de Cuba durante la Glaciaci6n de Wisconsin Ciencia s de la Tierra y el Espacio No. 7 pp. 57-68. PAJON J M. 1983 Distribuci6n de los Sedimentos Cuaternarios en las Cavernas del Karst de la Sierra de San Carlo s, Sierra de los Organos Pinar del Rio Cuba (Manuscrito [nedito Archivo del GEM de la SEC) SHA ZER E.V., PETROV O M. FRANCO G L. 1975 Sobr e las formaciones costeras del Holoceno en Cuba La s terrazas pleistocenicas de la Region Habana-Matanzas y lo s sedimentos asociados a ellas. ACC. Serie Geografica, No 21 pp 3-26. V ALOES J.J 1974 uevo elemento para el e s tudio cuantitativo de los carsos obtenidos mediante el analisi s dimensional y s u utilidad en el calculo de paleoprocesos geohidrol6gico s. Re v. Tee Vol. XI No 3 Mayo-Junio /1 974 pp 23-32 Symposium 7 : Physical Speleology 99

PAGE 113

Abstract Excentriques Capillaries, Do They Carry Paleoclimatic Information? Stephan Kempe & Jens Hartmann Techn. Univer. Darmstadt, Geol.-Paleontol. Inst., Schnittspahnstr. 9, D 64287 Darmstadt, Germany Excentriques (Helictites) are one of the most intriguing speleothems. They grow seemingly unimpeded by gravity. Apparently several different fonns exist. We have investigated the capillaries of monocrystalline clear calcitic excentriques collected from caves in the Winterberg limestone quarry, Harz, Germany, in 1972. We inspected the capillaries both visually and by casting them with epoxy and dissolving the calcite in dilute HCI. The capillaries casts were then investigated by scanning electron microscopy. Even though all of the crystals are from the same cave, the capillaries show various patterns. Some show crystal faces, others are rather smooth and still others have a clearly beaded morphology. It was this beaded morphology which led Kempe and Sand Spaeth (1977) to hypothesize that the individual knots are annual growth increments. We now test this hypothesis by measuring bead sizes and we will try to link these records with recent meteorological data on temperature, precipitation and potential evaporation from the Harz region. Zusammenfassung Excentriques (Helictiten) gehoren zu den merkwurdigsten Spelaothemen. Sie wachsen scheinbar unbeeinflu.Bt von der Gravitation. Offenbar gibt es verschiedene Formen. Wir haben die Kapillaren von monokristallinen, klaren, calcitischen Excentriques untersucht, die von einer Aufsammlung aus Steinbruchshohlen des Winterberges, Harz, Deutschland aus dem Jahr 1972 stammen untersucht. Die Kapillaren wurden mit Epoxydharz ausgegossen und der Kristall mit verdtinnter HCl aufgelost. Die Kapillarausgiisse wurden anschlieJ3end mit dem Elektronenmikroskop untersucht. Obwohl alle Kristalle aus der selben Hohle stammen, zeigen die Kapillaren ganz unterschiedliche Muster. Einige bilden Kristallflachen ab, andere sind glatt und wiederum andere haben eine Perlschnur-Morphologie. Diese Perlschnur-Morphologie lieJ3 Kempe & Spaeth ( 1977) vermuten, daJ3 die individuellen Knoten jahrlichen Zuwachs reprasentieren. Wir testen jetzt diese Hypothese, indem wir die KnotengroJ3e vermessen und versuchen, sie mit den rezenten Wetteraufzeichnungen iiber Temperatur, Niederschlag und potentieller Evaporation aus der Harz Region zu korrelieren. References KEMPE, S. & C. SPAETH (1977): Excentriques: Their capillaries and growth rates. Proc. 7th Intern. Spel. Congr. Sheffield, U.K.: 259-262. 1 00 Proceed i ngs of the 12 t h International Congress of Speleology 1997, Switzerland Volume 1

PAGE 114

Paleoclimatic reconstruction and timing of sea level rise at the end of the Penultimate Glaciation, from detailed stable isotopic study and TIMS dating of submerged Bahamian speleothem Joyce Lundberg Dept. of Geography, Carleton University, Ottawa, Ontario, K IS 5B6 Canada Abstract TIMS U-Th dating of a Baharnian stalagmite from -45 m ASL yielded a (preliminary) date for sea level rise at the end of the Penultimate Glaciation, past -45 m, at~ 131 ka Sea level rise from 45 m ASL to the Last Interglacial high stand of +6 m ASL (~130 ka) must have occurred in a short time frame of perhaps I kyr The stable isotope record from this sample, of higher resolution than most other proxy records, reveals high frequency spikes and troughs that mimic the GRIP ice core record Two preliminary dates place the isotopic curve almost exactly coincidental with GRIP, and almost mid-way between the orbitally-tuned marine forarniniferal record and OH 11, the Devil's Hole calcite vein from Nevada Introduction A stalagmite recovered from 45 m below present sea level in a blue hole on Andros Island, Bahamas, has been U-Th dated by thermal ionization mass spectrometry The cave was air-filled during glacial low sea levels and stalagmites grew Calcite deposition was stopped by rapid sea level rise at the end of the penultimate glacial (Termination II). The sample was then subject to dissolution giving it a broadly cuspate and asymmetric form (Figure I) After this some boring of the outer surfaces by marine sponges occurred and the sample was encased in 1-5 cm of marine packstone A similar sample was dated by alpha counting in the 1970s (GASCOYNEet al 1979) Those dates were unreliable for two reasons : the first is that the large samples size required for alpha counting reduces the resolution attainable; the second is that the sponge borings on the outer layers introduced a younger aragonitic component. Only the dates on the inner material can be considered reliable (within the large error inherent to the alpha counting system) and these dates are not of great interest. Materials and Methods For this study the stalagmite (78032) was sampled from close to the core out to the youngest layer still present at the tip of a dissolution cusp Samples for oxygen and carbon stable isotope analysis were taken approximately every millimeter using a binocular microscope to choose sites free from sponge borings Uranium content is ~o 2 ppm; for each date 0 2 g material was obtained (enough for 1-2 repeats) by drilling a slice parallel to the growth layers, then splitting this into I mm thin cleavage fragments Each fragment was ex:imined under binocular microscope and any boring spong~ holes were drilled out. Most of the holes were empty but some had a reddish filling of younger material Any slice with more than about 20% borings was rejected. Chemical purification of the dissolved samples was by ferric hydroxide co-precipitation of uranium and thorium and separation on anion exchange columns. Oxygen and carbon isotopic ratios were measured on a VG Isogas SIRA mass spectrometer in McMaster University, Canada Uranium and thorium isotopic ratios were measured on a Finnigan-Mat 262 thermal ionization mass spectrometer in The University of Bergen, Norway Dates were calculated using 23 'Thf3 8 U and 23 'Thf34u ratios JI cm Figure 1: Sample 78032: to the left is the origi11al speleothem cross sectio11 a11d the calcite remai11i11g after dissolution a11d marine packsto11e depositio11; to the right is the sampled slab showi11g isotope sampli11g holes a11d slices cut for dati11g. Results and discussion Date of sea level rise : Preliminary test dates gave 142 ka close to the core and 132 6 ka on an outer layer; these results place the Termination II sea level rise past -45 m ASL a thousand years or so after 132 ka TIMS dates on Atlantic coral reefs from isotope substage 5e extend back to 130-131 ka ( eg, BARD et al. 1990; CHEN et al. 1991; GALLUP et al. 1994) Thus there is a narrow window between 131 ka and 130-131 ka when sea level rose rapidly from -45 111 to+6 m ASL. Stable Isotope Profile: The high resolution stable isotope record (~6 points per kyr) is shown in Figure 2, the h e avy line for the two-point running mean and the lighter line for the raw data The resolution of this record for this time period is rivalled by few other proxy paleoclimatic records Moreover, this significant interval, from the full glacial (isotope stage 6) through the dramatic climatic changes of Termination II, is covered by few other terrestrial records with such a fine time and signal resolution Several features are apparent: (i) as might be expected for this oceanic site, d 18 Ois dominated by the meteoric signal, with a positive relationship of d 18 O and temperature; Symposium 7 : Physical Speleology 101

PAGE 115

-1 5 (D 6 -2 5 g -3.5 ii 0 N "' 0 -4 5 (D .... -5 5 110 120 78032 : mm from top -1.5 00 -2.5 6 !3 -3 5 ;g ;.. -4 5 .., ..... -5 5 110 120 78032 : mm lrom top 020406080 130 FM2 : age ka 0 40 80 140 DH11 ... I 150 16 .................. 120mASL 13.5 130 140 150 160 Ageka -5 5 -6.0 -6.5 "' 0 "' -7.0 -0 N :::; -7.5 u_ -8 -30 -42 Figure 2: d' 8 0 values for 78032 compared wit!, FM2, a Norwegai11 stalagmite (the upper graph) a11d wit!, the GRIP ice core a11d DHJ I ca/dte vei11 from Devil's Hole, Nevada (the lower graph). (ii) there is a double warming reminiscent of the end of the last glaciation (the Younger Dryas stage of Termination I) : this two step deglaciation has also been documented but with much lower precision from pollen diagrams, speleothem, ocean foraminifera, and pedostratigraphy by SEIDENKRANTZ et al ( 1996); (iii) in spite of being in a tropical oceanic environment where climatic changes might be well moderated, the record is spikey, showing rapid oscillations of climate rather than a steady progression; (iv) the range in d 18 0 is high at ,,,,3 /-: if the range in the ocean foraminiferal d1 8 0 is "'2 /_ then the extra "'1 /can be assumed to represent temperature dependent fractionation in the cave. In Figure 2, 78032 is first compared with FM2 {LAURITZEN 1995), another high resolution speleothem d1 8 0 record from this time period, but from Northern Norway, at the extreme north of the Atlantic and in an area close to the Fennoscandian ice sheet and sensitive to climatic changes. 78032 actually shows a greater overall range than does FM2. Note that the isotopic signal of FM2 during the period of very rapid growth at the end of Termination JI is inhibited because the sea water d 18 0 is changing rapidly (see LAURITZEN 1995) The peaks and troughs for the two records can be tentatively co1Telated but 78032 seems to pre-date FM2 by .,,3 kyr. However, this discrepancy is probably not real : the seven alpha counted dates on FM2 have 2s errors of 8-17 kyr Jn addition the dates on FM2 can be shifted back by at least 2000 years by correlating the very cold spike at"' 114 ka in FM2 with the spike in the GRIP ice core record (DANSGAARD et al 1993, GRJP MEMBERS 1993) at"' 116 ka In the lower graph of Figure 2 78032 is positioned in relation to sea level and age. It is superimposed on the GRIP record and OH 11, the calcite vein from Nevada {WINOGRAD et al 1992) The coincidence of 78032 and GRJP is remarkable, as is the extent of the offset ofDH 11. Conclusion The isotopic record of 78032 during Tremination II shows rapid oscillations similar to those shown by the GRIP ice core Overall there is a marked two-step pattern similar to that of Termination I. In terms of timing 78032 correlates well with the GRIP record and with FM2, post-dates DHl 1 by ,,,,5 kyr and predates the orbitally tuned SPECMAP marine record (and thus the insolation record) by ,.,4 kyr (MARTINSON et al 1987) Until further dating confirms the position of 78032, more detailed discussion is premature. References BARD, E., HAMELIN, B & FAIRBANKS, R G 1990 U-Th ages pbtained by mass spectrometry in corals from Barbados: Sea level during the past 130,000 years Nature 346 : 456-458. CHEN, 1 H., cuRRAN, H.A., WHJT~, B. & w AssERBHfGn G.1 1991 Precise chronology of the last mterglac1al penos : UTo data from fossil coral reefs in the Bahamas Geological Society of America Bulletin I 03: 82-97 DANSGAARD, w ., & ten others 1993 Evidence for general instability of past climate from a 250-kyr ice-core record Nature 364; 218-220. GALLUP, C.D ., EDWARDS, RL. & JOHNSON, R G 1994 The timing of high sea levels over the past 200,000 years. Science 263: 796-800. GASCOYNE, M BENJAMIN, G.J., SCHWARCZ, H P, & FORD D C 1979. Sea-level lowering during the Illinoian glaciation: evidence from a Bahamas "Blue Hole" Science 205 : 806-808 GRIP MEMBERS 1993. Climate instability during the last interglacial period recorded in the GRJP ice core Nature 364: 203207 LAURITZEN, S-E, 1995 High resolution Paleotemperature Proxy Record for the Last Interglacial based on Norwegian Speleothems Quaternary Research 4:, 133-146 MARTINSON, D G., PISIAS, N G., HA '.'S, J.D ., lMBRIE, J., MOORE, T.C & SHACKLETON, N.J 1987 Age dating and the orbital theory of the Ice Ages: development of a high-resolution 0 to 300,000 yr chronostr&!igraphy Quaternary Research 27 : I 29. SEIDENKRANTZ, M-S. & nine others 1996. Two-step deglaciation at the oxygen istotope stage 6/5e transition : the Zeifen-Kattegat Climate Oscillation Quaternary Science Reviews 15: 63-75 WINOGRAD I.J. COPLEN, T.B., KLANDWEHR, J.M ., RIGGS, A C., LUDWIG, K.R SZABO, BJ., KOLESAR, P .T ., & REVESZ, K M 1992 Continuous 500,000 year climate record from vein calcite in Devil 's Hole, Nevada Science 258: 255-260. 102 Proceed ings of the 12 '" Internat ional Congress of Speleology, 1997, Switzerland Volume 1

PAGE 116

Speleothem records of environmental changes in the past potential in comparison with the other paleoenvironmental archives and related UIS international programs by Y. Y.Shopov Section Speleology & Faculty of Physics, University of Sofia, James Baucher 5 Sofia 1126 Bulgaria E-mail : YYShopov @ Phys U ni-Sofia.Bulg Abstract Potential resolution and limitation of high resolution speleothem records of Paleotemperature, Paleosoils Paleoseismics Past Precipitations Rock displacement Solar Insolation Geomagnetic field Plants Populations, Chemical Pollutions Air Composition Sea Level advances Advances of Hydrothermal Waters Cosmic Rays Flux variations Cosmogenic Isotopes production and Supernova Eruptions in the Past are discussed (table 1) An international working group on "Speleothem Records of Environmental Changes in the Past" of the Commission on Physical Chemistry and Hydrogeology of Karst of International Union of Speleology is dedicated to study these records It is coordinated by Y.Shopov and has different topic leaders (table 1) Potential of speleothem records of environmental changes in the past is compared with other paleoenvironmental archives. It is demonstrated that speleothems are the best paleoenvironmental archives of many properties of the environment. Highest resolution of speleothem records (6 hours) is higher than that achieved from any other paleoenvi ronmental record Table 1. Spe/eothem Records of Environmental Chan1:es in the Past Type of the Process Obtainable information Time range [a] Method leader ; best time resolution I Changes beyond the Solar System 1 Past Supernova eru ptions primary cosmic ray flux variations in the 010000 past beyond the Solar System. Past Supernova eruptions Y Shopov (Bulg) ; 20a II Changes within the Solar System 1 Cosmic Rays Flux Variations ; cosmic rays flux and solar activity in the 040000 Cosmogenic Isotopes past Variations D Lal (USA) ; 20a 2 Solar Insolation ; Laser Quantitative reconstructions of Solar unlimited Luminescent Microzonal activity variations in the past speleothem Analysis (LLMZA) growth interruptions volcanic eruptions Y Shopov (Bulg) ; 6hours III Global Earth Processes 1 Paleomagnetism ; Magnepaleomagnetism rock orientation changes unlimited tometry of speleothems in the past ; less than 50 a A.Latham R.Gilson (UK) 2 Geomagnetic dipole intensity ; quantitative reconstructions of unlimited LLMZA Geomagnetic dipole intensity solar wind flux at Earths magnetosphere in the Past Y.Shopov (BuJg): 40a IV Regional Processes 1 Paleoclimate and Paleotemperature possible plant unlimited Paleotemperature lfl calcite population precipitations and climatic speleothems ; cycles in the past ; stable isotopes D Ford C Yonge (Can) T.Arakawa (Jap) ; 25a Symposium 7: P h ys i ca l Speleo l og y 103

PAGE 117

2 Past Precipitations; annual quantitative reconstructions of past annual unlimited growth rate observed by rainfall past floods cycles of draughts and LLMZA floods Y.Shooov (Bulg) ; la 3 Past Paleotemperature; quantitative reconstructions of air urilimited LLMZA paleotemperature during speleothem growth temperature cycles glaciations etc Y.Shopov (Bulg) ; 6 hours 4 Paleotemperature in speair paleotemperature air CO 2 air isotope unlimited leothem ice ; composition in the past Stable isotopes C Yonge, W MacDonald (Can) ; 25a 5 Paleotectonics and Papaleoseismics rock displacements and unlimited leoseismics ; Orientation of bending soeleothem growth P Forti (It) N Nori (Jap) ;
PAGE 118

Speleothems as natural climatic stations with annual to dayly resolution Abstract by Y. Y.Shopov, L. Tsankov, L.N Georgiev, A.Damyanova, Y. Damyanov Section Speleology & Faculty of Ph y sics University of Sofia James Baucher 5 Sofia 1126 Bulgaria E-mail : YYShopov @P hys Uni-Sofia BG ; D.C. Ford Geography Dept. McMaster University Hamilton Ontario L8S IK4 Canada ; C.J.Yonge, W. MacDonald, H.P.RKrouse Dept. of Ph y sics Uni v ersi ty of Calgary Calgary Alberta Canada Calcite speleothem luminescence depends exponentiall y upon soil temperatures that are det e rmined primaril y by solar infrared radiation in the case when that ca v e is covered onl y by grass or upon air temperatures in case that cave is covered by forest or bush In the first case microzonality of luminescence of speleothems can be used as an indirect Solar lnsolation (SI) index, but in the second as an paleotemperature proxy So in dependence on the ca v e site we may speak about "solar sensitive" and "temperature sensitive" Paleoluminescence speleothem r e cords like in tree ring records but in our case record may depend either onl y on temperature or on solar irradiation : In case of Cold Water cave Iowa, US we obtained high correlation coefficient of 0 9 between the luminescence record and Solar Luminosity Sunspot index (fig I) and reconstructed sunspot numbers since 1000 AD with precision within the experimental error of their measurements ; in case of Rats Nest cave Alberta Canada we measured correlation coefficient of 0 67 between luminescence intensity and air temperatures record for the last 100 y ears (fig 2) and reconstructed annual air temperatures for last 1500 years at the cave site with estimated error of 0 35 c while the error of the direct measurements is 0 1 c > :oo ---------------, ;; C V 10 ::::, !;: ISO -::; = JOO z lO } iog C C E -= 0 7 0 C C 1.6 -----------------, no JI 12 = 1.0 / J ~ _,' C 0 a, ~.:-CO ," ~ :; 2 2 00 c.. E ;!! I 00 r~ I g TTTnTrJrTrTmTrfT1ill lji llf i ilijiiiliililjlll(TTTT / "'J C t _.,,, 5t (I 00 wu..uiu.J.w-Luiliu.. w.J. '....!.1.J.uu .... : .1.l.luw.W.UJ.1..1W:.J. OS L..-=~1 -00 ___ \_S00 ___ \ _900 ___ :ooo :) 20 4:l 6C so ICC 120 1600 A 2 e (vrs A D \ i'ear s eP Fig.I. (left up) Twenty year average sunspot records since 1700 AD (left down) Optical density of luminescence of a speleothemfrom Coldwater Cave, Iowa (USA). Fig.2. (right up) Average annual luminescence density of a speleothemfrom Rats Nest Cave (Canada) (right down) Annual temperature Ban ff, Alberta (Canada). Sympo sium 7 : P hysical Spe l eo logy 105

PAGE 119

Intensity of luminescence was not dependent on actual precipitations and sunspot numbers (zero correlation). Speleothem growth rate variations represent mainly rainfall variations. Speleothem luminescence visualizes annual microbanding we used to derive prox-y records of annual precipitations for the cave site. In case of Rats Nest Cave Alberta Canada we reconstructed annual precipitations for last 280 years at the cave site with estimated error of 80 mm/year. By comparison of luminescent records with other solar proxy records we obtained a reconstruction of growth rates and precipitations in Bosnek karst region near Duhlata Cave, Bulgaria, for the last 50000 years, and for the last 6400 years (with averaged time step of 41 years) for Iowa near Cold Water Cave US Achieved time step of 6 hours (fig. 3) in speleothem luminescence records allows resolution of several days in some best speleothem samples. Annual luminescence microbanding was used very successfully for relative and absolute dating of speleothems by Autocalibration dating This dating method appear to be more precise than TAMS 14 C and AMS Uffh dating for relative dating of short time intervals and only dating method for speleothems with little uranium younger than 2000 years. It is demonstrated that speleothems can be used as natural climatic stations with annual resolution for purposes of climatology and agrometeorology for a time span far exceeding all historic records. A 0.04 C 0 8 Power Spectral Density 0.0 2 0.7 0 ODL 0 0 02 0.04 0.6 0.5 B Residual ODL 0 04 /\ 0 0 ~ ~ t!Y:iA r',:r 1 r~;tA -0.02 0.5 V 1 YEARS 1.5 2 -0.04 1-------'------'----..,__ __ __,_ ___ __, Fig.3. A: 2.4 -yr-long section of annual record of CWJ, resolved to 6 hlpx, with 300 px spline curve fitted to the record. B: Residual after subtracting spline curve from original one. C: Power spectrum of original data, presenting peaks with periods (38, 27, 21 and 13.5 days) in addition to the strong annual and seasonal cycles (out of the scale of this graph). Presence of 13. 5 days peak indicates presence of periodic process shorter than 1 weak. ODLoptical density of Luminescence =In([), (IIntensity of Luminescence). 106 Proceedings of the 12 '" Internat i onal Congress of Speleology 1997 Switzerland Volume 1

PAGE 120

Speleothem luminescence proxy records of annual rainfall in the past: Evidences for "The Deluge" in speleothems by Y.Y.Shopov, L.T.Tsankov, L.N.Georgiev, A.Damyanova, Y. Damyanov, E. Marinova Section Speleology & Faculty of Physics University of Sofia James Baucher 5 Sofia 1126 Bulgaria Abstract E-mail : YYShopov@Phys Uni-Sofia BG D.C. Ford Geography Dept. McMaster University Hamilton Ontario L8S 1K4 Canada C.J.Yonge, W. MacDonald, H.P.RKrouse Dept. of Physics University of Calgary Calgary Alberta Canada A good correlation between the growth rate of the cave speleothems and the annual precipitations at the cave site has been demonstrated Extrapolating the results back in time we obtained a quantitative reconstruction of the annual precipitations for the last 280 years at the Rast Nest Cave site (Alberta Canada) Measuring another sample (from Duhlata Cave Bulgaria) we found an enormously high growth rate at 5500 years B C. Its possible connection with the Bible Flood is discussed. 1. Method Speleothem growth rate variations represent mainly rainfall variations (Shopov et al 1992 1994 ) Speleothem luminescence visualizes annual microbanding (Shopov et al. 1991) We used it to derive proxy records of annual precipitations around the cave site by measuring the distance between all adjacent annual maxima of the intensity of luminescence The resultant growth rates correlate with the actual annual precipitation (summed from August to August) 2. Quantitative reconstructions of annual precipitations In the past We studied the top of a 35 mm long stalagmite from Rats Nest cave (RNC) Alberta Canada We obtained reasonably good correlation (correlation coefficient of 0 57) between the annual precipitations (recorded at the closest weather station Ban.ff located in the same valley 50 km northern of the cave) and the annual growth rate of the speleothem (Fig 1 ) We used obtained regression coefficients to reconstruct annual precipitations for the last 280 years at the cave site The estimated statistical error is 80 mm/ year. Annual speleothem growth rate was independent on the intensity of luminescence on annual temperature and on solar luminosity for the same time span (zero correlation) 3. Evidences for the Bible Flood By tuning the time scale of a luminescent record with a geomagnetic dipole intensity record (Mazaud et al ., 1991) we obtained a reconstruction of growth rates and precipitations in Bosnek karst region near Duhlata cave {DC) Bulgaria for the last 50000 years (Fig 2). This record shows a very prominent peak at 5500 years B C ., when the annual growth rate was over 53 times higher than today Considering that cave site is located in the region of the oldest civilizations (Mediterranean basin) this event can be related to the Bible Deluge ". The age of the recorded event is about the age of The Creation of the World" The duration of the recorded event is of 120 years because of the low resolution of the record In case that the real flood happened for a short time span it suggests enormous high rainfall. Present day precipitation at the cave site is 650 mm/yr This speleothem was dated with 8 TAMS 1 4 C dates Such event is described in the Bible Greek mythology and the Sumerian epic Gilgamesh (compiled during 3-rd millennium B C. on the base of more ancient legends) At that time human civilization had been concentrated around the Black and Mediterranean sea therefore the Flood hited also recent Bulgarian lands Such immemorial precipitation probably would lead to some temporary rising of the Black Sea level. Such rising at 5500 B.C with 150 meters (which had flooded 60000 square miles) was recently suggested by an international team of scientists lead by Dr. William Rayn and Walter Pittman from Columbia University, Palisades New York (their results are still presented only in a press conference see J Wilford 1996) The Black Sea level rising itself cannot be undoubtedly related to the Flood but combined with the never seen (during the human civilization) precipitations at that time definitely lead to the conclusion Sympos i um 7 : Phys i ca l Speleology 107

PAGE 121

that this phenomenon is namely the Bible Flood Here we suggest a hypotesis for one possible mechanism of the Flood, consisting of the following : It is known that the Ocean level rised from 10000 years to the time of the Flood as a result of the glaciers melting. Black Sea had been isolated from the Ocean and it's level had been much lower In one moment the narrow band of land between the Mediterranean and Black Sea had broken down like a dam wall. This had resulted in flowing of giant masses of sea water into the Black Sea basin When it reached the opposite cost, then a giant wave had been formed (which probably was incomparably bigger that biggest tsunami known so far). This wave had destroyed everything on lands around the Black Sea even beyond the regions flooded by the sea level rise. The never seen precipitations at that time had contributed to the rising of the sea level and maybe caused the final rising, which had turned the Mediterranean Sea over the edge to flood the Black Sea region Further studies of the Mediterranean sea level during the Flood and data for precipitations from stalagmites taken from other caves in the region are necessary to prove this hypothesis. The mechanism of generation of such an enormously high rainfall is even more interesting. To generate such a rainfall is necessary to have much stronger than usual water evaporation It is very unlikely to be due to much warmer climate for several years. At the same time few percent higher than normal solar radiation will produce much higher evaporation (without significant rise of the air temperature) Such higher solar radiation can be produced by an extraordinary solar eruption, or (perhaps more likely) by explosion of a comet or asteroide in the Solar atmosphere Such explosion (like Tungussian meteorite) can cause a major mixing of parts of Solar shells and appearance of warmer Solar matter from the depth to Solar surface. Probably several years would be necessary for recovery of the Sun from similar catastrophic event. 80 0 <., u 60.0 ,...._ X a... !! 40 0 :5 3 20.0 0, 0 :, 0 0 0 20 40 60 Years BP 80 100 120 ::: 800 n-rrn-rrrrrr,n-rn"TTT"t"TTT"l"TTT"l.,..,,...TT"l"TTT"l"Trn"r~~"TTT"I~ C 0 (Il E6oo E ....,, C 400 0 --= a... c:; 200 a. 0 ::, C C < 20 40 60 Years BP 80 100 120 Fig.I (Up) Annual growth rate ofa stalagmite from RNC, Alberta, Canada (Down) Annual precipitations (from August to August) for Ban:ff station, 50 km north of the cave 108 Proceedings of the 12 th International Congress of Speleology 1997, Switzerland Volume 1

PAGE 122

10 9 .... 8 .. >, 7 -)( B 6 G) ii 5 a: .r: 4 i 3 0 .. CJ 2 1 0 0 10000 20000 30000 40000 50000 Age (yrs B.P.] Fig.2. Reconstruction of growth rates (proxy of precipitations) of a flowstone from DC Bulgaria for the last 50000 years 4. Conclusion The Bible Flood (catastrophic rainfall) as recorded in a speleothem had probably happened 5500 years B C ., causing Black Sea level rise It is probably due to higher water evaporation caused by higher solar radiation resulting from a catastrophic event on the Sun References MAZAUD A. et al 1991:Geophys Res Lett ., 18 No 10 : 1885-1888 SHOPOV Y.Y. et al 1991 : IGCP299 Newsletter 3: 52-58 SHOPOV Y.Y. et al. 1992 : GSA Abstr ., 24, No 7 : A268 SHOPOV Y.Y. et al. 1994 : Geology 22 407-410 STUIVER M & KRA R.1986 : Radiocarbon Calibration issue 28 WILFORD J. 1996 : Black Sea Rising : Was it the Flood? Int. Herald Tribune (19. 12 1996) Symposium 7 : Phys ical Speleology 109

PAGE 123

Speleothem luminescence proxy records of geomagnetic field intensity Abstract by Y.Y.Shopov, A.Damyanova, Y. Damyanov, L.Tsankov Section Speleology & Faculty of Physics University of Sofia James Baucher 5, Sofia 1126, Bulgaria E-mail: YYShopov@Phys Uni-Sofia BG The geomagnetic field is an electromagnetic raU1er U1an a pure magnetic field, so it depends on the far stronger solar magnetic field and the solar wind magnetic field The solar wind determines dimensions of the Earths magnetosphere and modulates the intensity of the geomagnetic dipole Using a luminescent record from Dublata cave, Bulgaria we obtained a reconstruction of the geomagnetic dipole during the last 50000 years with re olution of 28 yrs The original luminescent record exhibited a correlation coefficient of 0.78 with an independent record of variations of intensity of the geomagnetic dipole for the same time span (Fig. I.). Observed dependencies of the geomagnetic dipole intensity on the orbital variations and solar lumino s ity variations correlates excellently with the established theoretical equations. Obtained reconstruction of the geomagnetic field is of vital importance for calibration of cosmogenic isotope ( 14 C, etc.) dating techniques NASA used a record of lumine sce nce of a flowstone from Dul1lata cave, Bulgaria to estab lish a standard record of variations of Solar lrradiance ("So lar constant") for U1e last 10000 years by calibration of the lumine cence record with satellite measurements C: ~ 7 e 1.0 a. u I .., 0 1.2 2 C ::J Ql 0.9 > :.:; 0 &! 0 7 0 1 0000 20000 30000 40000 50000 Age (Years B.P.) Fig.l.(11p) Optical density of l11minescence of a cave flowstone (down) Past variations of intensity of the geomagnetic dipole expressed by inverted prod11ction of 14 C (MAZAUD et al, 1991) 110 Proceed ings of the 12 th Internat io nal Congress of Speleology 1997, Switzerland Volume 1

PAGE 124

Speleothem records of processes beyond the solar system (Supernova eruptions ) Abstra c t by Y.Y Shopov, A. D amyanova, Y. Damyanov, L.Tsankov, Departments of General Physics Astronomy and Nuclear Physics Faculty of Physics University of Sofia James Baucher 5 Sofia 1126 Bulgaria. E-mail : YYShopov@Phys Uni-Sofia BG ; C.J.Yonge, J. Bland Dept. of Physics University of Calgary Calgary Alberta Canada D.C. Ford Geography Dept. McMaster University Hamilton Ontario, L8S 1K4 Canada We used the standard Calibration 14 C record (Stuiver et al. 1987) to derive a proxy record of Cosmic Rays Flux This 14 C record represents the Cosmic Ray Flux (CRF) and modulation of the CRF by the solar wind A striking correlation (with a correlation coefficient of 0.8) was demonstrated between the calibration residual delta 14 C record and a LLMZA speleothem record Using a luminescent record from Duhlata cave Bulgaria we obtained a reconstruction of the solar modulation of the CRF during the last 50000 years with time resolution of 28 yrs Luminescence rnicrozonality was used to reconstruct Galactic Cosmic Rays Flux (beyond the Solar System) during the last 6500 years with 20 yr resolution by subtracting of an inverted luminescent solar activity record (sample from Cold Water cave Iowa) from the residual 1 4 C record Last result (fig I) presents a picture of past Supernova explosions in our Galaxy It is a quantitative confirmation of recent views on origin of Cosmic Rays from superposition of Supernova explosions in our Galaxy and agrees with astrophysical observations It completely disapproves the hypothesis of origin of a significant part of Cosmic Rays Flux from relic rays of the Big Bang E ...; ... a. ll 00 l 00 --r I 00 -ll 00 -1.IIHI non IIHIII IHI 1 _ l __L_. I l -.L. 1000 no Jo110 oo 10/lll on Ae c IYr s ll P I 1000110 l _, 61\110110 700000 Fig ]. Relative variations of the primary cosmic rays flux, expressed by cosmogenic 14 C variations. Sympos i um 7 : Phys ica l Speleology 111

PAGE 126

Development of cave passages according to geological structure; Example from Jama pod Peeno rebrijo, Slovenia Stanka Sebela, Karst Research Institute ZRC SAZU, Titov trg 2, 6230 Postojna, Slovenia Abstract About 1,8 km east from the entrance of Postojnska jama lies 203 metres long cave Jama pod Peeno rebrijo The cave has two passages with two vertical entrance shafts. Northern passage phreatically developed in two layers of Upper Cretaceous limestone. Tectonic uplifting of the region caused that the cave lost its hydrological functions but a part of the cave kept primary phreatical shape. 1. Introduction Under top of hill Peena reber (763 m a.s.l.) above Postojna (Fig 1) lies 203 m long cave (Reg No. 1577). Vh lZI '4 =: :-----~ .. .. .. I ...... .________.16 .. : ~7 .. .. .. 0 --~3 S.SEBELA IZRK ZRC SAZU 1996 Figure J : The location of Jama pod Peeno rebrijo according to Postojnska jam a and surrounding caves. I-railway, 2-liighway, 3-town, 4-above sea level in metres, 5-ground plan of cave, 6-cave entrance, 7-Register number of the cave, 8-Eoceneflysch and Upper Cretaceous limestone. Sympos i um 7 : Phys i cal Speleology 113

PAGE 127

1~ 2D 340 sl\....\ I I 11 I -0 30m ............_ S. ~EBELA, IZRK ZRC SAZU, 1996 Figure 2 : Space projectio11 of phreatic passage i11 Jama pod Pee110 rebrijo. The thick11ess of /imesto11e ;,, block diagrams is 3 times overlarged. l-grou11d pla11 of cave with polygo11 poi11t, 2-upper layer of Upper Cretaceous /imesto11e, 3-lower layer of Upper Cretaceous limesto11e, 4-directio11 of old water flow, 5-strike and dip of limestone layers. Jama pod Peeno rebrijo and especially its northern passage has phreatical characteristics. Both passage ends are today closed by collapse blocks. In the cave we find remains of sediment and flowstone formations. Northern passage (Fig. 2) developed in two layers of Upper Cretaceous limestone which dip towards SW. 120 metres long passage goes from lower into higher layer that means that dipping angle of passage is smaller than dipping angle of limestone layers. The passage isn't all the time parallel to the direction of layer but is declined in eastern and middle part for 30 Just in western part is the direction of passage parallel with direction of layer. Development of phreatical channels in limestone layers was described in the case of Velika dolina collapse doline in Skocjanskejame (KNEZ, 1994). In our example first phase in development of Jama pod Peeno rebrijo is phreatical formation in strata plane with divergences 0-30 from the direction of strata plane. Dinaric tectonic activity with principal directions of faulting NW-SE and NE-SW affected the western part of the cave, which is today closed with collapse blocks. Fault zone in which entrance shaft developed is just outer part of wider fault zone. 114 Proceedings of the 12 h International Congress of Speleology 1997 Sw i tzerland Volume 1

PAGE 128

The cave has two passages with two vertical entrances. Shaft in southern passage is 11,5 m deep and lxl m wide. Principal shaft entrance (647 m a.s.1. and 12,5 m deep) is accesible by badly preserved steps. The vicinity of Postojnska jama was being explored in the 19th century already with a desire to find new parts of the cave (BERT ARELLI et al. 1926). According to CA VE REGISTER in vicinity of Postojnska jama there are more 1 0 caves, some are not longer than 2 m and others are longer than 100 m. Table 1 shows basic data of bigger caves near Postojnska jama. NAME OF THE REG. VGLONGER DEEP a.s.l. of type of the cave CAVE NO. REG (m) {m) the entrance .NO. (m) Postojnska jama 747 108 19.555 115 511 ponor horizontal cave Zgubajama 6290 563 122 4 561 horizontal cave Jama na poti 583 2436 65 32,5 570 horizontal cave Jama pod Peeno 1577 203 38 647 horizontal cave rebriio Javornisko brezno I. 1563 1945 20 86 725 shaft Jama pod Volejim 2716 1946 50 133,5 715 shaft vrhom Table 1 : List of tlte biggest caves 11ear Postoj11a (after CA VE REGISTER IZRK ZRC SAZU). 2. Geological conditions in Jama pod Peeno rebrijo The cave developed in Upper Cretaceous limestone (BUSER et al., 1967). Northern passage (Fig. 2) rises from the entrance shaft to the end for 38 m. We can observe one limestone layer which is very rich with rudist rernaines. The passage leaves this layer east from point 16 (Fig. 2). The direction of strata in eastern part of the passage is 230/30 and in other parts 260/30. The passage dips from east towards west what is also concordant with directions of water flow shown by rocky relief. Southern passage is almo t horizontal in Dinaric direction (NW-SE). The northern passage is much more characteristically developed in strata plane than the southern one. The northern passage developed in two principal directions (Fig.2). From point 24 to east there is first direction of cave passage (260) which declines from direction of strata plane for 30 towards W. Second direction of passage (290) is between points 22 and 24. It declines from direction of strata plane for 30 towards W. The third direction of passage has the same direction as the eastern part of passage (260), but passage is parallel to direction of strata plane. Tectonically crushed zones are badly expressed in northern passage being usually transverse to the passage direction. The most expres ed fault zone lies west from point 16 (Fig. 2). Its geological elements are 70/70-80. Inside this fault zone we ob erve 3 chimneys, one in northern passage and two others in a southern one. Second intensive fault zone crosses entrance shaft (10 metres west from point 16) which developed between two tectonic systems. The geological elements of the stronger are 20-30/50 and of the weaker 120 /3 0. In northern passage we find rests of flysch sediment which completely filled the cave. In some parts flowstone is deposited over sediments. In later periods edirnents and flowstone were eroded. Through almost all northern passage we observe 1,5 drn wide channel on the ceiling. It partly disappears only in chambers where stronger tectonical zones occur. Paragenetical channel probably developed after filling with sediments. 3. Conclusions The entrance of cave Jama pod Peeno rebrijo (Reg. No. 1577) lies 647 m a.s.l. To the SE there are two hafts, Javornisko brezno I (Reg. No. 1563) which is 86 m deep and 20 m long and Jama pod Volejim vrhom (Reg. No. 2716) 133,5 m deep and 50 m long (Fig. 1). NE from Postojnska jama there is the entrance to 122 m long Zguba jama, Reg. No. 6290, 561 m a.s.1. (SEBELA, 1994). Genetical relation between morphologically similar Jama pod Peeno rebrijo and Zguba jamo in air distance of 2 km has sense if we describe mechanics of movements of cros dinaric (NE-SW) Postojnska vrata fault zone, which lies between these two caves. Symposium 7 : Physica l Speleology 115

PAGE 129

With activity or reactivity of "Postojnska vrata" fault zone (NE-SW) hydrodinamical conditions changed changing also activity of Jama pod Peeno rebrijo. It was filled with sediments and later sediments were removed. With tectonic uplifting of the region the cave lost its function of water channel for ever, but a part of the cave kept primary phreatical shape. In the vicinity of Postojna entrances of todays water active caves as Postojnska jama and Lekinka are located in the contact of Eocene Flysch and Upper Cretaceous limestone. Refere n ces BERTARELLI, L.V. & BOEGAN, E., 1926. Duemila grotte. Milano, 494 p. BUSER, S., GRAD, K. & PLENIEAR, M., 1967. Osnovna geoloska karta SFRJ Postojna 1:100.000. Zvezni geoloski zavod, Beograd. CA VE REGISTER IZRK ZRC SAZU K.NEZ, M., 1994. Phreatic channels in Velika dolina, Skocjanske jame (Skocjanske jame caves, Slovenia). Acta carsologica 23, Proceedings of 1st international karstological school "Classical Karst", Lipica, September 20-23, 1993, Ljubljana, 63-72. SEBELA, S., 1994. The caves Jama na poti and Zgubajama Acta carsologica 23, International round table "E.A. Martel et le karst slovene", Postojna, 12-13 november 1993, Ljubljana, 233-243. Genesis of Jozsef hegy hydrothermal cave, Budapest by Szabolcs Leel-Ossy More than 100 caves and cave-indications are known from the Triassic and Eocene carbonatic sequences of the Rozsadomb region, Buda Hills, Budapest. Most of these caves are characterised by the absence of any natural outlet. Therefore, their discovery, mainly in the 20th century, was accidental or a result of systematic research or due to the creation of artifical outcrops. One third of the caves are merely indications, another 30 percent have the size of tens of metres, while the rest exceeds the extent of several tens of metres. Only the large cave systems of Szepvolgy and the Jozsef, Matyas, Ferenc and Szemlo Hills have the dimension of km's. Studying the geological setting of the area and the main features of the formation of the caves we can predict that substantial parts of the system are still undiscovered. Most of the indications can be found in Buda Marl, while the galleries and levels of the larger caves are situated in Szepvolgy Limestone. The currently known cavern-system, with a total length of30 km, can be regarded as the fossile source level of the present-day thermal springs at the foot of the Buda Hills. Their origin is interpreted as a result of mixed corrosion along tectonic fractures at the level of carstic water. The radiometric age of the syngenetic minerals gave good constraints on the time of the cave-formation. The internal size of the caves of Rozsadomb changes drastically. frequently, corridors and halls with a size of ten or hundred metres can be found. The walls of the galleries are often adorned by spherical niches. More than a dozen species of minerals were reported from these caves. Especially the variety and mass of carbonates (mainly calcite) and sulphates (mostly gypsum) are remarkable. On the base of dating (carried out in Bergen, in the laboratory of Stein-Erik Lauritzen) the Jozsefhegy Cave is very young. The age of the system is maximum 400.000-500.000 years. A few crystals are syngenetic with the cave-system, but a lot of crystals (for example aragonite crystals) are much younger, and there are some recent crystals, too (for example gypsum needles). 116 Proceedings of the 12 International Congress of Speleology, 1997, Switzerland Volume 1

PAGE 130

Cave patterns north of the Planinsko polje (Slovenia) France Sustersic University of Ljubljana, Department of Geology, SI-1000 Ljubljana, Askerceva 12, Slovenia Abstra ct Detailed study of cave patterns in the area of the Planinsko polje (Slovenia) revealed that rather than in levels, the caves are organized in flow corridors and tiers. The primary tubes are formed along a small number of bedding planes, predominantly on upper or lower sides of the dolomite packs. Though joints are abundant in some cases it is evident that many a passage was at least incepted before jointing. The absence of typical epiphreatic details indicates that transition from completely phreatic to pure vadose was fast. Preface The area of the classical karst in Slovenia is best considered as a triangle with vertices that just encompass the cities of Ljubljana, Rijeka and Gorica/Gorizia. Within the classical triangle, the sinking Ljubljanica river is the backbone of the catchment that drains towards the Danube. CLA SS I C A L TR I A NG LE Fig J: The classic a l tria n g l e: P Plan insko po lj e The Ljubljanica catchment area It is widely known as a string of surface and underground stream segments, the latter emerging into karst poljes. It might b e expected that caves in this area must be fundamentally of more or less epiphreatic origin, with their development bound intimately to the formation of the poljes and the presumed river terraces on their margins (GOSPODARIC & HABIC, 1979). During the last century and a half, extensive speleological and geomorphological work was carried out and the great amount of information that was gathered did not confirm it but, on the contrary, made it hardly acceptable. It was not until the nineteen-nineties that studies founded upon other paradigms began to appear (BRENCIC, 1992; SEBELA, 1994; SUSTERSIC, 1994, in press). This paper deals with a relatively small sector of the Ljubljanica catchment area (Fig 1), north and north-east of the Planinsko polje. It appears that the FORD & EWERS' (1978) four-state cave genesis model offers a framework within which the conditions of the Ljubljanica River can be presented better than in any earlier model. Thus their terminology is used extensively in this paper, as are more recent improvements (in the author's opinion) suggested by WORTHINGT O N (1991) and LOWE (1992). The calculated size of the Ljubljanica drainage basin is 1779 km 2 and about 1100 km 1 of this are composed of karstic rocks. The location of the water divide is approximate, and at several boundaries bifurcations have been proved by water tracing. According to studies during the complex water tracing experiments of the nineteen-seventies, the catchment area of the Vrhnika springs, where the main river definitively leaves karst terrain, covers 1108. 78 km 2 T he karstic rocks in are generally micritic, locally oolitic limestones and dominantly late-diagenetic dolomites, mostly of Mesozoic age. They were formed on the Dinaric platform by continuous sedimentation under very uniform shallow sea Symposium 7: Physical Speleology 1 17

PAGE 131

conditions that were responsible for their extremely high purity, with generally <5%, but as low as 0.1 % insoluble residue. The total thickness of the carbonate sequence is 6850m within the considered area (CAR, 1996) Among non-karstic rocks, only early Tertiary flysch, which was deposited directly upon the carbonate sequence, has a significant role. Younger sediments are absent, suggesting that the final emergence happened at the end of the Tertiary Structurally, the whole of the Lju bl janica basin b elongs to the Adriatic sub-plate. The area is composed of a number of nappes (PLACER, 1982; PREMRU, 1983) that were overthrust in a NE-SW direction as a result of the collision of the Adriatic sub-plate with the European continent. Gradual change of direction of the movements brought about the formation of the Idria dextral strike-slip fault, which runs through the area in a NW-SE direction. The surface between the Planinsko polje (445m) and the Ljubljansko Barje (Ljubljana Marsh, 294m), where the karst waters finally appear on the surface, is typically karstic. More recent research (HABIO, 1981; OUOTEROIO, 1987) has proved that the relief is essentially tectonic. The difference between the caves that have formed as system drains and the vadose caves is very clear On the basis of observations made predominantly by cavers, it appears that nearly a half of the caves explored within the Ljubljanica catchment area are of phreatic origin. In the following text these will be referred to as horiwntal" caves, though generally they are not horizontal in the true sense of the word. The situation north of the Planinsko polje The area about 1km north of the main ponors of the Planinsko polje is built up of alternations of lower Cretaceous limestone with two intercalations of "dolomite", all overlain by upper Cretaceous limestone. The lower Cretaceous limestone (90 95% CaCO3) is micritic, rich in organic matter, locally rich in fossil shells The "dolomite" is a micro sparry dolomitized limestone (90-95% CaCO3) of quite varia b le texture Thus, this traditional term, used elsewhere in this chapter, is in fact not strictly appropriate. Both "dolomite" layers are a b out 30m thick. The upper Cretaceous micritic limestone is very pure (95-99% CaCO3). The general dip is 30 towards the west, but l ocal distortions are common. A number of fragments of horizontal caves are known in the area, and even at first sight they appear to be ranged along the lower and upper contacts of both dolomite packages. Typically, none of the active swallow holes penetrates more than some 1 00m from the polje. Some hundreds of metres from the polje flood water appears in Najdena jama (259 /1, 5008m long, 121m deep), and Vranjajama (88, 326m long, 90m deep). There is no direct connection between the polje border and inland caves, as a vertical difference of about 20m between the flood levels on the polje and within the system remains more or less unchanged regardless of the absolute flood level (SUSTERSIC, 1982). Inside the system (i.e in Najdena jama), it is evident that about 3km of its main passages developed along the upper and lower contacts of the stratigraphically higher dolomite package, and along a third bedding plane within the lower Cretaceous limestones, parallel with and about 5m beneath the lower dolomite contact. The primary tubes are lenticular to ellipsoid but their cross-sections have been altered locally by entrenchment. Apparently the passages are orientated either down/up dip o r along the strike, but detailed inspection reveals that this is only an approximation. Passages, formed along joints are greatly su b ordinate and generally they developed at the locations of breakthrough that provided links from one major bedding plane to another. Phreatic risers, some more than 50m (downstream) upwards, are not uncommon Though some canyonisation has occurred most of the system is typically phreatic, with no features suggestive of water table formation. Many of these statements also hold true for other caves in the area. The Vodni rov (Water Passage) in Vranjajama (88) is of special interest. It was formed along the same plane as some other structural segments in the vertical span of more than 100m and a horizontal distance over of 500m Its form is completely phreatic, and at first glance it appears to be formed in the dip direction (SUSTERSIC, 1994,a) However, when measured properly, there are significant differences (Fig 2). Fig 2: Str u ctures co 11 tro llin g t h e Water pass a ge in V r a n ja jama. A beddi n g p l a 11 e; B passage axis; P,Q, R ,S joints, secondar/y penetrated b y water; T downdip joi n t, untouched by cha1111el for m atio n 11 8 Proceedings of the 12 '" International Congress of Speleology, 1997 Switzerland Volume 1

PAGE 132

Although the original plane is dissected by several joints and laminated zones, they did not serve as initial structures, but water evidently penetrated into them after the main passage had been formed. Thus, though the channel was formed along the bedding plane, it does not appear to have been influenced by any of the numerous geological structures or orientations that now affect the bedding plane Con equently, the passage must have begun to form before the structures were imposed This is best explained by LOWE's (1992) theory of speleo-inception before orogeny. In Najdena jama, passages are penetrable (by scuba diving) down to a level of 388.5m, while the highest known position in system is at 531.8m. They display phreatic or secondary features, and are almost always found on the upper or lower contacts of the "dolomite" layer. A number of fragments that exist on intermediate levels generally follow the same bedding planes. The most convincing interpretation of these observations is provided by WORTHINGTON'S (1991) findings, viewing all the caves in the area as representing fragments of a single tier. Similar geological conditions and speleological effects are found along an older dolomite layer, lying about 2km to the east and stratigraphically 500m lower. Situation at the Laski Ravnik The Laski Ravnik lies about 3km north-east of the Planinsko polje. The area is generally flat and extremely rich in solution do lines (SUSTERSIC, 1994,b ). Some main streams of the underground Ljubljanica must flow beneath the area, though no active stream cave has yet been found among a number of short fragments of evidently phreatic origin. A bed of coarse-grained secondary dolomite with a general dip of 2520 / 320 occurs on the junction between Jurassic and Cretaceous limestones. Around this outcrop, at Javorjev Gric, an area of about lkm 2 was mapped geologically and geomorphologically at 1 :5000 scale Special care was taken to observe and record all detectable originally underground karst phenomena that now appear at the surface due to denudation (Fig. 3). 5'45 500 \ 5082 \ 000 f\.. 1 @~ 2 3 =L 0 300m During the field work it became more evident that there is not a single dolomite package, but an interfingering of wider or narrower dolomite/limestone stripes (Fig. 3). Among the superficial karst phenomena, and true vertical (vadose) shafts being omitted, four types of underground ones were recorded (numbered as in Fig. 3): 1 Openings of phreatic tubes, not higher or larger than 0.5m. All of them become choked after some metres. 2. Small collapsed features (<3m), evidently continuing to a choked tube. 3. Segments of (sub-)vertical tubes, formed under phreatic conditions 4. Denuded cave passages, filled with loam and other sediments. Fig 3: Situation at Javorjev Gric. D dolo m ite; n umeratio n see t h e text! Detailed inspection of the figure reveals that despite the area being stro n gly reworked by denudation it was possible to observe that most traces of phreatic tubes lie close to either the upper or lower limestone/dolomite contact. At some locations this correlation is clearly visible to the naked eye in the field .. It appears that around Javorjev Gric an earlier sector of an abandoned tier is exposed. Again, the limestone/dolomite contact was the most prone to channel formation. In the NW quadrant a concentration of features lies directly within the extrapolation of a presently non-existent continuation of a dolomite stripe what hints that the driving mechanism of inception might be recalcitisation of the dolomite. Symposium 7 : Physical Speleology 119

PAGE 133

Conclusions Horizontal caves appear in well-expressed clusters, up to several kilometres in length, a few kilometres in width, and some hundred s of metres in depth. This pattern fits the notion of flow corridors within a single tier as defined by WORTHINGTON (1991). Caves within a given tier were formed under phreatic conditions and reworked in vadose ones (in the hydrogeological sense). This means either that the change from the primary to the present hydrological conditions was relative fast or that activity within the system was suspended during the transition. Though the preserved phreatic passages are numerous where studied in detail, they are concentrated along a small number of bedding planes. In some cases it is evident that their directions do not follow any current structural framework, and that any penetration into joints was secondary. These relatively few bedding planes play the exact role of inception horiwns as defined by LOWE (o.c.) Joint s and smaller faults are really important only as master structures to guide the formation of phreatic jumps within a tier. This probably mean s that they play an important role during the adaptation phase of a tier. More highly tectonically di s rupted zones define areas o f significant cavern collapse or local slab spalling, during the subsequent decay of the cave. Thu s, the general s cheme ofa cave corridor (trans)formation is: 1: initiation along bedding plane s; 2: reorganisation, penetration into joints ; 3 : expansion by collapse of crushed zones of faults. References Archi v es of the Speleological association of Slovenia and the Karst research institute, ZRC SAZU Postojna BRENCIC M. 1992. Koselevc (Summary 2 ) Nasejame 34:, 41 51. CAR, J. 1982 : Geologic setting of the Planina polje ponor area (Summary) Acta carsologica 10 75 105. CAR, J. 1996: Personal communication FORD D.C. & EWERS R.O 1978. The development oflimestone cave systems in the dimensions oflenghth and breadth C anadian journal of earth science 15 : 1783 -1798. GOSPODARIC R. & Habic P. (eds.) 1976. Underground water tracing. Institute for Karst Research SAZU Postojna 309 p GOSPODARIC R. & HABIC P 1979. Karst phenomena ofCerknisko polje (Summary). Acta carsologica 8 (1978): 7 162. HABIC P. 1981. Karst relief and tectonics (Summary). Acta carsologica 10 : 23 44 LOWE D J. 1992: The origin of limestone caverns : an inception horizon hypothesis. Unpubl. PhD thesis, Manchester Polytechnic 51 2 p. PLACER, L. 1981. Geologic structure ofsouthwestern Slovenia (Summary) Geologija 24 (1): 27 60 PREMRU U. 1982. Geologic structure of southern Slovenia (Summary). Geologija 25 (1): 95 126. SEB E LA S. 1994 The role of tectonic structures during the formation of cave channels and the karst surface phenomena (Abstract) Unpubl. PhD thesis University of Ljubljana 129 p. SUSTERSIC F. 1982.: Morphology and hydrography ofNajdenajama (Summary) Acta carsologica 10 : 127-155. SUSTERSIC F. 1987. The small scale surface karst and solution dolines at the north-eastern border of Planinsko polje (Summary). Ac ta c ar s ologica 16 : 51-82 S U STERSIC F. 1994 a The Kloka cave and speleo-inception (Summary) Nasejame 36: 9 30 S U STERSIC F. 1994 b. Classic dolines of classical site Acta carsologica 23: 123-154. SUSTERSIC F. in press. Caves and Poljes ofNotranjska Acta carsologica 25 (1996). SUSTERSIC F & Puc M. 1970. The karst underground by the NE nook of the Planinsko polje (Summary). Acta c a r so l o gica 5: 205 270. WORTHINGTON S.R.H. 1991.: Karst hydrogeology of the Canadian Rocky Mountains Unpubl. PhD thesis Mc Master University 227 p. FOOTNOTES Numeration according to the central register of caves of Slovenia, maintained by the Speleological association of Slovenia and the Karst research institute, ZRC SAZU, Postojna. 2 The titles of s ummaries / ab s tracts (if they exist) are given just to show the english reader the contents of the original texts which are however considered in the whole. 120 P r o c eed in gs o f t he 12 th Internat i onal Cong r e s s of Spe l eolog y, 1997, Sw i tze rl and Vo l ume 1

PAGE 134

Lithological control of speleogenesis in the Pre-Carpat h ian Region Igor Turchinov Lviv Geology Survey Expedition, ul.Turgeneva 33, UA-290018 Lviv, Ukraine Abs t rac t The influence of lithological features of the Badenian (Middle Miocene) gypsum upon the speleogenesis in these rocks in the Pre Carpathian region is shown on the examples of Optirnisticheskaya and Dzhurinskaya caves (Western Ukraine) and Skorochitska cave (Southern Poland). In the regions of each of these caves gypsum is characterized by specific lithological and textural peculiaritie s which have controlled development of speleo-initiating jointing, space location of cave passages, morphology of cavities and planned structure of cave maze networks 1 I ntroduction Sulphate sediments of the Middle Badenian (Middle Miocene) evaporite formation stretch as the belt of 40-60 km in width along the northern and northern-eastern borders of the Carpathians (fig. I) Extensive development of gypsum karst is connected with these rocks In the region the five largest in the world gypsum maze caves are located ( Optimisticheskaya 192 km, Ozernaya I 11 km, Zolushka 89 km Mlynki 24 km Kristalnaya 22 km) Karst development in general, and speleogenesis particularly are controlled by numerous factors (tectonic, lithological, hydrogeological etc.). ln the Pre-Carpathian region lithological factor takes an important place in the control of karst processes. The influence of lithological and textural peculiarities of the Miocene gypsum upon the speleogenesis in these rocks is shown on the examples of Optimisticheskaya and Dzhurinskaya caves (Western Ukraine) and Skorochitska cave (Southern Poland). In the regions of each of these caves gypsum is characterized by peculiaritie s, which have detennined features of speleogenesis. p 0 L A N D \ ,, \ i : ........ ,':>5;_ ..... i \.11--~'r,. I\ {\ .. // .... /\ {\ /\ I\/\. I\ i {\ ... ol vov 1 A ;;,-.. Terrt0pol .. .. -,. V K R A j "".;/.: . \~-{:--... ... 100km N E Figure] : Locatio11 of studied area. A sulphate facies, B i11vestigated caves (I Optimisticheskaya, 2 Dzhuri11skaya, 3 Skorochitska) 2 Li tho l ogical peculiar i ties of the Middle Baden i an gypsum In the Pre-Carpathian region the gypsum stratum of I 0-60 m in thickne ss is characterized by clear vertical structural and textural differentiation which reflects conditions of se dimen tation and diagenesis of sulphate sediments (fig. 2). In the region of Optimisticheskaya cave the gypsum stratum of l8-22 m thickness has a three-unit composition (fig. 2A) Fine-grained gypsum with stromatolitic texture occurs in the lower part of the stratum (unit I) The middle part of the stratum (unit II) is characterized by concentric occurence of fine-grained and coarse-crystalline gypsum, which form dome-like structures of 0 5-3 m in diameter. Coarse-crystalline gypsum is composed of tabular crystals of 10-15 cm in size. The upper part of the gypsum stratum (unit III) is separated from the middle part by a thin ( 1-30 cm< normally 5-10 cm) layer ofbentonitic clays, and is built by giantocrystalline sa breA. gi,nt gypsum intrrgrowlhr strarutoUiic i;;;;;;J gypsum t.,bu l,r gypsum S" Lelal ~gypsum Ybre-likP gypsum B I clutic gypsum ~ spzr gypsum { C=:::J benlonit, c:::::::J C l1 y i:;:::::::;:::: p~ lilomorp hic t::::::::::j limestone ~ cuy s t qal U mts(onfs (IDwfr hcl1niJn) c. \ \ \ \ \ '\ '\ \ '\) '\ ) ) '\ \ \ ) h )\\)\ g \\\\'I\ "''))) Figure 2: Sequences of tl,e Middle Badenia11 gypsum. Ai11 the region of Optimisticheskaya cave, 8in the regio11 of Dz/111ri11skaya cave, Ctypical Bade11ia11 gypsum sequence i11 the Southern Poland (Borkow Quarry) (PERYT & JASIONOWSKI, 1994) Symposium 7: Phys i cal Speleology 121

PAGE 135

like gypsum. The most distinctive feature of this part is an occurence of giant vertically elongated dome-like structures up to I O m in diameter (nonna ll y 4-6 m) (KLIMCHOUK et al. 1995; TURCHINOV A DREICHOUK, 1995) The core part of the domes is built by fine-grained gypsum and has a shape elongation in the vertical direction The transverse size of the core parts is nonnally I 0-50 cm and the height is 2-5 m In the very top part of the core a monocrystal of spar gypsum occurs. Peripheral part of the dome is composed of upward-curved and splitting sabre-like gypsum crystals of 0,2-1,5 m in length In the near-core part of the dome sabre-like crystals are cemented by fine-grained gypsum The dome-like structures are divided by the boundarie which create a polygonal network in the plan view In the region of Dzurinskaya cave (fig 28) fine-grained stromatolitic gypsum also occurs in the lower part of the gypsum stratum (unit I) The upper part (unit II) is built by sabre-like gypsum which fonns two sub-units (Ila and lib) Both of these sub-units have small dome-like structures of 0, 1-1 m in diameter in their bottom The length of the abre-like crystals is 15-50 cm ln the Western Ukraine the gypsum stratum is overlain by the Middle Badenian pelitomorphic limestones of 0,2-1,5 m in thickness ln the Southern Poland the gypsum stratum has a complicated sequence (fig 2C) Giant gypsum crystalline intergrowths (up to 4 m high) occur at the base of the sequence (unit ~) They are overlain by bedded tabular coar e-crystalline gypsum with intercalations of alabastrine and stromatolitic gypsum (units b-e), followed by skeletal gypsum (unit f, composed of chaotically oriented tabular crystals) which pass upwards into abre-like gypsum (unit s g:l) (KASPRZYK 1993) Sometime the gently sloping domes up to 12 m in diameter and 4 m in height occur in the sabre-like gyp um. Core parts of these domes are composed of skeletal gypsum The upper part of the gypsum sequence (units .i.:n) is built by elastic gypsum (gypsum breccias and laminated elastic gypsum), however this part is eroded in the region of Skorochitska cave. 3. Lithological conrol of joint networks Joint system in gypsum are basis to development of karst cavities. According to the latest data (KL I MCHOUK et al., 1995) the genesis of joints in the Miocene gypsum in the region is a result of simultaneous impact of outer (tectonic) and inner (lithogenetical) stresses. Gypsum with different textures, has different physical and mechanical features. This factor causes a different appearance of joints in different parts of the gypsum sequence Most clearly t he relations between jo i nting and lithological peculiarities of gypsum are expressed in Optimi ticheskaya cave (fig 3). In the upper part of the gypsum stratum (unit Ill), which is composed of giant dome-like structures splitting of the rock under influence of outer tectonic stresses realized a l o n g the most relaxed parts of rock-along the planes t h at divide t h e domes Sometimes the dome-like structures are split from the center to periphery When joints develop along the planes dividing gypsum domes, joint networks are characterized by 5-6 angular hape of polygons by predominance of 3-beams connection, by absence of clearly expressed main orientations (fig 3A). In the middle part of the stratum (unit II) the rock also splits along the planes, which divide small dome-like structures occuring here. so that joint s fonned attain zigzag-like confuguration The orientation of such joints is approximated by a sraight line corresponding to the axis of a passage (fig. 38) The cave maze in this part is characterized by quadrangular shape of polygons In t h e l ower part of the stratum homogeneous fine-grained gypsum is characterized by isotropy of physica l and mechanical features. Here, the speleo-initiating jointing is characterized by clearly expressed main directions (with 2 main and 1-2 minor peaks on t h e diagram), by quadrangular shape of polygons and predominance of 4-beams intersections of joints (fig 3C) 4. Lithological control of space position of carst cavities and their morphology Optimi s tiche s ka y a ca v e In the l ower part of the gypsum stratum (unit I) the cave passages nonnally have rhomb-like cross sections (fig 4A, I 0) Combinations of two rhombus occur frequently ( 4A, 11 ) Such sections fonn due to episodes of prolonged stay of water table at a certain position, within the stage of dewatering of the gypsum stratum. Fissure-like passages (4A, 12) are less frequent here Wide and low passages with plane ceiling occur along the contact of the gypsum stratum with the underlying fonnations (4A 13) Cavities of quadrangular shape with plane ceiling (4A. 8; 9) are common in t h e midd l e part of the stratum (unit 11 ). Its ceiling is contro ll ed by the bottom of the unit III. Contact surface between the units II and Ill is an important factor in the control of space position of cave maze networks Verteba cave, the greater part of Ozemaya cave and many regions of Optimistic h eskaya cave are timed to this contact. A. UNIT Ill the upper ra,rt of the gypsum stral B. U NI T II t f-. e m d c1 e Prt of tt'e gyp s ~rn str a ~ .1 c. U NI T I th~ lower pan of thegypsum stra t a Jointing appearanc e Labyrin t h p attern inherite d ]o . . : c : . . . l ~ r O 2 4 6 ,. F i g u re 3 : P ec ul i ar i t ies of jo i11 t 11 e t works i11 d ijfere 11t i 11t erva l s of th e gy p s um st r at um i11 Optim ist i c h eskaya c a ve 122 Proceedings of the 12 th International Congress of Speleology, 1997 Switzerland Volume 1

PAGE 136

In the upper part of the stratum (unit III) passages most commonly occur along joints along the planes dividing the dome-like structures. Such passages usually have fissure-like and triangular section {4A, 7), less frequently-rhomb-like section (4A 1). Sometimes cave passage developed along joints radially splitting domes {4A 6) There are also cavities developed along surfaces dividing concenters of the structures {4A, 2) Dome-like cavitites of 2-5 m in diameter (4A, 3) are common in the upper part of the stratum These cavities have been formed by the select dissolution of fine-grained gypsum in the core and near-core parts of the dome-like structures Some of the galleries and halls in Optimisticheskaya cave are a combination of these dome-like cavities and other speleoforms. A. m20 15' iO 5 0 B. m i5 1.0 5 .0 0 c. h120 1.5 iO 5 0 Cavities developed along joints dissecting the whole gypsum stratum, have more complicated morphology {4A 14) and are represented by combinations of several speleoforms described above Dzhurfnskaya cave Space position of passages and galleries of this cave i s controlled by the surface between the units and sub-units of the gypsum sequence (fig. 4B) Cave passages are developed commonly in the upper part of the stratum and are characterized by plane ceilings. Morphology of the passages is determined by textural peculiarities of gypsum {subhorizontal bedding of layers of sabre-like gypsum) In the lower part of the stratum cave passages seldom occur and have fissure-like cross section ill ii I 0 s 10m I I l 0 5 10m f t,-e a 0 5 10m Figure 4 : Peculiarities of space location and morphology of karst cavities (in the vertical section). A Optimisticheskaya cave, B Dzhurinskaya cave, C Skorocl1itska cave. Numbers are referred to in the text Sympos i um 7 : Ph y s i ca l Speleo l og y 123

PAGE 137

Skorochitska cave This cave is a canal of the underground flow, opening most ancient cavities. Localization ofkarst cavities is controlled by the borders between the units of the gypsum sequence (fig. 4C). Cave passages are developed along the borders between the units and .!! (4C, 5; 8; 10), and f (4C, 7; 9) f_and~ (4C, 1; 2). Within the units f and 2 original dome-like cavities (4C I) have been formed as a result of the dissolution of the core parts of gypsum domes Their morphology may be complicated by collapsed gypsum blocks (4C, 2) Within the sabre-like gypsum, wide and low galleries (4C, 3) occur along the bedding planes. Within the units b-e morphology of cavities is complicated by ridges and cornices, which have been formed due to the selective dissolution of tine-grained gypsum. Karst cavities developed in giant gypsum intergrowths (unit ) have original arched shapes (4C, 4; 6) Acknowledgments I wish to express my gratitude to Prof. Dr. Tadeush Peryt (Polish Geological Institute, Warsaw) for his help in my acquaintance with geology of the Miocene gypsum and gypsum karst in Southern Poland Also I wish to thank the cavers from the Lvov Caving Club for their help in underground investigations in Optimisticheskaya and Dzhurinskaya caves. References KASPRZYK A. 1993. Lithofacies and sedimentation of the Badenian (Middle Miocene) gypsum in the Northern part of the Carpathian foredeep, Southern Poland Annales Societatis Geologorum Polonide, vol.63: 33-84. KLIMCHOUK A.B., ANDREICHOUK V. & TURCHINOV I.I. 1995. Structural prerequisites of speleogenesis in gypsum in the Western Ukraine. Ukrainian Speleological Association Kiev I 04 p. PERYT T.M & JASIONOVSKI M 1994 In situ formed and redeposited gypsum breccias in the Middle Miocene Badenian of southern Poland. Sedimentary Geology, 94 : 153-163. TURCHINOV I.I. & ANDREICHOUK V.N 1995 Dome like structures in the Middle Miocene Badenian gypsum of the Dnestr River region (Western Ukraine). Przeglad Geologiczny, vol. 43 nr. 5 : 403-405. 124 Proceedings of the 12 th International Congress of Speleology, 1997, Switzerland Volume 1

PAGE 138

The Karst in the area of Matumbi Hills (Kilwa district Tanzan i a ) Hydrogeological characteristics and relationships between morpho structures and tectonic phases by Rosa r io Ruggieri (Centro Ibleo di Ricerche Speleo-ldrogeologiche di Ragusa, via Carducci 165, 97100 Ragusa, Italy) Abst r ac t Following the first German-Turkish expedition performed in 1994 in the area of the Matumbi Hills of Tanzania, during which 5,390 m of the andembo system was surveyed in september 1995 a new Italian-German expedition carried out both furth e r exploration of new caves and a series of geomorphological and hydrochemical researches With this exploration the Nandembo karst system became the thirteenth longest explored cave in Africa ; other caves were urveyd in the eastern sector of Mbinga Hills and all the geological hydrogeological and geomorphological observations the latter on the intraformational paleo-karst forms allowed a relational hypotesis to formulate between the speleogenetical evolution of the explored cave system and the tectonic pha s es which occured in the area of Matumbi Hills in the infra-middle Jurassic and post Callovian age R i assunto Facenda seguito ad una prima spedizione tedesco-turca realizzata nel 1994 nell'area delle Matumbi Hils della Tanzania durante la quale furono rilevati circa 5,390 m del sistema Nandembo nel settembre del 1995 una nuova spedizione italo-tedesca effettua nella stessa area ulteriori esplorazioni di nuove cavita ed esegue nel contempo una serie di rilevamenti idrochimici e geornorfologici A seguito di tali esplorazioni ii sistema Nandembo, con 1 'aggiunta de! Ramo Sicilia raggiunge i 7,5 I O m divenendo la tredicesima grotta piu lunga d 'Africa ; ulteriori cavita vengono inoltre, rilevate nel settore occidentale delle Mbinga Hills mentre l'insieme delle osservazioni geologiche idrogeologiche e, in particolar modo quelle geomorfologiche sulle paleo-forme intraformazionali riscontrate con s entono di formulare una ipote i relazionale fra l evoluzione speleogenetica che ha caratterizzato i sistemi carsici esplorati e le vicende tettoniche legate alle fasi di emersione verificatesi nell'area delle Maturnbi Hills in eta infra-media Giurassica e post Calloviano 1. Introduction In 1994 a German Turkish expedition surveyed 5,390 m of caves in the Matumbi Hills in the Kilwa District of Tanzania In order to continue the explorations a second expedition was carried out in September 1995 by CIRS of Ragusa with some gennan researches of the previous expedition The principle objectives of the expedition was that to find in the Matumbi Hills as many caves as possible with the aim to obtain a general view both of the morpho-structural and hydrogeological characteristics of the kar s t area The area of research included in the maps to scale I : 50 000 Kipatimu, Sheet 239 / 4 and Kandawale Sheet 255 /2 is characterized by a monotonous pattern of hills with elevation between 400 m and 570 m a .s. l. long and narrow with sides meeting to the bottom of the valley with gradients both steep around 30 and more gently sloping around I O 0 The superficial water flow is generated througt a hydrographyc network noticeably dendritic belonging to the basin of the Mtunbei river. 2. Geology and stratigraphic series In the area the Jurassic sediments of the Matumbi series outcrop (STOCKLEY I 943) This layers, outcroping on most of the Matumbi hills, are made up by two Formations : Kipatimu b e d s and Mtumb e i b e d s. The litologic terms of the Kipatimu b e d s are made up of a series of pink-buff sandstones, from massiv to flaggy, and red grey-green and green blue mudstones The sand s tones form the great part of the outcropping in the area while the presence of the mudstone is shown by the outcropping of clay colured soils The age of the Kipatimu bed s dates from medium Jurassic (Bathonian-Callovian) to the upper Jurassic and probably extend to the lower Cretaceous while the thickness of the series has been estimated to be about 300 m The underlying series called the Mtumb e i beds is made up of oolitic limestones, hard and well bedded from shallow sea, with interbedding of sandstones and calcareous sandstones, going toward the bottom becoming sandstones The estimated thickness is about 150 m. The age of the Mtumb e i Formation has been calculated, based on its fossil content to be between the Bathonian and the Bajocian. Mor ph o-st ru ctural elements The trend of the relief shows in the research area, the presence of a gentle anticline with an axis having a NE-SW stricke approximately parallel to the superficial waterdivide described before, and reaching an e l evation of about 700 m a s l. to the south west of Nwenge In the southern s ector the above mentioned structure is delimited instead by a fault escarpment with direction from NW-SE to WNW-SSE The s ides of the anticline are lowered by a system of faults, with a E SW direction, parallel to the direction of the main superficial hydrical flow A second system of faults approximately perpendicular to the first, with a NW-SE direction complete the morpho structural context of the area from which is put in evidence a general control carried out by the tectonic on the evolution of the relief both the make up of the hydrographical network and the morphology of the outcropping Symposium 7 : Physica l Speleo l ogy 1 2 5

PAGE 139

3. A general description of the principle caves surveyed Mpatawa cave The Mpatawa ca v e is mainly made up ofa large gallery 1 160 m long with a s tream flowing along its length. The cave is entered from a collapsed doline after a steep descend amongst rocks of various dimensions and a low pa s sage The gallery whose dimensions are about 20 x 20 m has along the first tract corroded calcitic deposits and elastic rocks due to collapse After about 350 m a big deposit of flowstone obstructs the middle section of the gallery and it is possible to by-pass this through a narrow conduit in the lower part of the deposit which leads to the underlyng stream The above mentioned flow s tone deposits shows in the upper part some big basins and in the lower part where the stream flows on the ceiling small domes clustered together. Proceding along the stream, the gallery curves toward NE due to a tectonic structure then it starts a thick layer of deposits of sand and red clay Finaly after a further tract of 500 m the gallery becomes lower following the dip of the strata, until it assumes the configuration of a low flat conduit where the water pre v ents further advance The cave runs from west to east towards the Nandembo system from which it is only 500 m and for this reason it appears probable that the two system are hydrologically linked amaingo cave The Namaingo cave which drains the large spring of Kihangambembe is made up of a fossil sector of 495 m and from a main gallery along which a stream flows and of a labyrinthine system of passages and small chambers which together form a system of 2390 m. The river after 450 m from the entrance becames a syphon a s the ceiling lowers following the dip of the layers Along the course of the river the cave shows an ample section like a nave" of a church with roof channels both of conical shape and with a flat base in the layer of calcareous sandstone On the walls of the gallery morphologies of corrosion type bear scratch are seen and forms of corrosion type scallops. An interesting corroded surfaces covered by a layer of siltstone are present on th bed of a calcareous bank in a fossil sector of the cave In this last sector, near ther ceiling, on a sandy limestone layers small elliptical paleo-karst conduits are present. The Namaingo cave is developed in Jurassic limestone and drains the water of infiltration of a sector of the Matumbi Hills with a main system of tectonic structures NNW-SSE and a secondary system EE-SSW The andembo system The andembo system surveyed for a total length of 7510 m, resulted in being made up of a main gallery, along which a stream flows from NNW to SSE and a tributary system located in the left sector. This last system is made up of the large Nangoma cave, the Nakitara cave and from the Sicily inlet In the whole system there are several macro and micro morphologies coming from the alternate karstic cycles of erosion and corrosion linked to the tectonic evolution of the area In particular in the main gallery, along the stream, large channels are present on the ceiling with vertical walls and the upper side being flat corresponding to the calcareous sandstones layers Still on the ceiling meandering channels due to the erosion of the rising water level, are present. The main gallery shows elliptical tracts formed along the bedding planes and higher tracts formed along the fractures with a direction N I 70 -180 In the Nakitara cave a first fossil level is present partiall y filled by clasty and sandy deposits This level is connected to a lower meandering system of conduits periodically active formed by the erosion of the rising water during the overflood phase s. On the entrance of the Nakitara cave a system of fractures were surveyed showing a direction of N 70 -80 with spacing of 2-3 m and with an interlayer extension of 2 5 m These fractures which appeared to be interrrupted in corresponding to the sandstone strata, are present in the layers with a dip direction of N 345 and dip of 8 A second system of fractures intersects the first system with an angle of 90 and together forms the network of phreatic conduits, now fossil, in the initial part of the cave Fig. 1 Area of research in the Matumbi Hills The Sicily Inl et The Sicily inlet, with a length of 1,700 m is made up of a main conduit with a stream, of a secondary fossil system located in an upper level and lastly of a fossil conduit which join s w ith the Nakitara cave. In this sector the cave shows a gallery with a "bell s hape and continues with a series of conduits having both domes" and bell sections Following up this first fossil conduit the section becomes narrower due to the raising of the floor caused by the deposits of sand On the walls of the lower conduit which resulted as being a ceiling channel particular morphologies of corrosion are present with the shape of a "scoop" "semi-domes" and bear scratches" 126 Proceedings of the 12 th International Congress of Speleology 1997 Switzerland Vo l ume 1

PAGE 140

I {2 : i'.5 source in Kitanangombre -va T lte-; 1 1 7 -r ,-r 1 = C 1 A =J~=EJ 1 = S =[I O =Fr=T:r=HJE=J M ==BL 1 =.rN=G 1 = A J=HJ[I= L [ L =Sr=( 1 M=Ja=tuJ[m=bc=l rM=o 1 1 u =J~=ta] i [ln=~[)=L i likolongomba f j I = = 0 ,--~ I I \ I ~ Klpatlmu divis i on and Mtondo wa Klmwaga dlvtslon, l
PAGE 141

4. Hydrogeology and karst-tectonic relationships of the Matumbi Hills The vadose underground hydric system of the Matumbi Hills has been initially formed along tectonic fractures, linking the outcrop of sandstones (Kipatimu bed s ) with the underlying limestones (Ntumbei beds), subsequentely evoluted as sinkholes and pseudo-dolines, with the progression of karst activity in the above mentioned limestone Formation The vadose hydric flow so concentrated in a few conduits was developed later in phreatic ambients both along the dip direction of the layers and locally de v iated to the intersection of secondary tectonic structure s The gentle dip of the layers ( 5 to 7 ) and the presence of a poorly karstificable roof represented by the calcareous sandstones subject to geostatic collapsing due to the effect of the hydrostatic oscillation has let the karst system (evolved from phreatic to vadose) again return to phreatic conditions, after a certain underground way with the formation of sumps, where the base piezometric level reach the overlying sandstones Consequently, in corresponding to the topographic incisions or for the effect of structural lowering springs of vauclusian type are formed down stream from the abo v e mentioned karst sy s tem This was in general the s peleogenetic e v olution of the N a nd e mbo sy st e m Mpatawa and Namaingo ca ves, which show in the eastern sub-basin a direction of hydric flow from NW to SE along the dip direction of the layers (and coinciding with the superficial flow impo s ed along the secondary tectonic structure) terminal sumps and a series of significant geomorphological elements ( sand-clay deposits differently bedded and dipped, paleo-karst internal surface and various morphologies) which denote the accurance in the area of a cyclic tectonics phases of isostatics lifting and seas trangressions Regarding this according MOORE ( 1961) the movement of lifting of the Matumbi Hills associated to a system of faults with a NNE-SSW and NW-SE direction should have started in the premium Jurassic age and to have continued in an interrnittant way, up to recent times This lifting, futhermore, could have been more intense in the western sector of the Matumbi Hills which shows more elevation component from the eastern sector. The soils of the medium and Lower Cretaceous, along the escarpement between Migaregere and Nantumbili overlie the Jurassic sandstone with a low di s continuity which denotes furthermore a phase of emersion of the Matumbi series followed by a tran s gression of the sea Furthermore the evidence of structural lifting of the infra medium Jurassic age which has emerged the limstone of the Mtumb e i Formation has been noted in the Namaingo cave where is visible a paleo-karst surface overlaid by a siltstone layer. The siltstone which covers in an irregular way the underlaying karstified surfaces show s a lentilform thickness from half meters to three meters and could signal the starting of the s edimentation of the Kipatimu Formation In the same cave the above mentioned sandy limestone, overlaying the layer of siltstone presents furhermore some paleo-karst conduits, which s how a successive phase oflifting and emersion of the sandstones series (post-Callovian ? ) which followed the transgression of the layers of medium Cretaceous (in the sector North Kilwa) Lastly, the non correspondence of the underground hydriflow with the direction of the superficial flow (which coincides the last one with the main tectonic structures of the area) can be attributed both to the already mentioned overl y ing sandstone series porly permeable and above all to the tamponing (blocking) action performed by the tectonic structure s in respect to underground waterflow This tamponing action is due to the poor permeability of the cataclastic sandstone belt and to the interbedded cla y le v els, compared to the greater permeability present in tlie limestone which depends on the secondary tectonic structure s uch as bedding plains and joints. Conclusions The explorations and the researches carried out with the German-Italian Speleological Project "Tanzania 95" allowed the increase both of the knowledge of the underground hydric circulation of the Matumbi Hills and that of the speleo genetic mechanisms which are generated in the Jurassic limestone of the Mtumbei Formation Morphostructural evidences found in some caves furthermore, allowed us to suppose on the occurance of a lifting episode with the begining of karstification of the limestone of the Mtumbei series, before to the sedimentation of the Kipatimu b e ds A following transgression of the sea should have recovered then, the karstified external surfaces. In other caves isolated and residual conglomerated deposits which included biocla s tic elements, were found and they could be associated to the above mentioned sea transgression that must ha v e filled evol v ed karstic systems. The possibility to undertake in the future other studies both on the cave deposits with eventual spectrometric dating determination and on the numerous and various morphologies, would surely allow the adding of new and important elements to the already outlined hydro-karst system of the Matumbi Hills. Bibliografia LAUMANNS, M & RUGGIERI R (1996 ): Caves of the Matumbi Hills The International Caver (I 6) I 996 Aven Inter Publications Swindon England MOORE W R (1974): Geological notes on quarter degree sheets 256 / 256E Kilwa and 255 Njinjo s outhern region . Rec.Geol.Survey Tanganyca Vol. IX 31 ; Dares Sdalaam STOCKLEY, G.M (1943) : The geology of the Rufiji district including a portion of northern Kilwa district (Matumbi Hills) Tanganyka Notes and Records 16 Dares Salaam 128 Proceed i ngs of the 12 th Internat i o n al Congress of Speleo l og y, 1 997 Sw i tze rl and V ol u m e 1

PAGE 142

Structures hierarchiques dans le karst de la St Baume (Bas du Rhin Var France) Martin Ph. Universite d'Artois UFR Hist-Geo 9 rue du Temple BP 665 62030 Arras cedex et URA 903 du CNRS Universite de Provence, Institut de geographie 29 av r. Schuman, 13621 Aix en Provence Abstract On the St Baume I 80 caves are opened The deepest is the cave of Petit Saint Cassien (-321 m ; 9 3km) The position and the dif ference in level D (gap between the highest and the lowest points) of every cave are used to check the organisation of this karst on all the montain If the Pt St Cassien is excluded we show : C=287* I 0 0 0 > C is the number of caves whose the difference in level is inferior to D Zipf Mandelbrot s pattern is fitted to the distribution of the number of caves by I 0 m classes The frequencies F<0 l is fitted to a power law (slope= g = 2,22 ; fractal dimension= Il g) An additional parameter p alowes modeling all the distribution p must translate the state of epikarst. P = 1,36 A power law allowes modeling the variation of the specific difference of level (Ds in m/km2; sum of the D of the caves which open in a disc with radius R, divided by the value of the area of the disc) according to the radius of the disc Around the Pt St Cassien : Ds = 221 R 1. 6 This pattern is also validated around others big caves but with shorter radius Therefore to scale of the mountain karstification corresponds to fitted together hierarchical structures Resume Sur la Sainte Baume la karstification est developpee : 180 cavites dont l'aven du Petit Saint Cassien (-321 m 9,3 km). La position et la denivellation D (ecart entre le point le plus haut et le plus bas) de chaque cavite sont utilisees pour verifier que ce karst est organise a l'echelle du massif. Si on exclut le Pt St Cassien on montre que : C = 287 IO ''Ol7>. C est le nombre de cavites d'une denivellation inferieure a D Le modele de Zipf Mandelbrot est ajuste a la distribution du nombre de cavites par classes de 10 m. Les frequences F < 0 1 correspondent a une loi de puissance de pente g = 2 22 et la dimension fractale : 1 / g. Un parametre supplementaire p permet de modeliser ensemble de la distribution p doit quantifier l etat de l'epikarst. p = 1 36. Une loi de puissance modelise la variation de la denivellation specifique (Ds en m/km : somrne des D des cavites qui s ouvrent dans un disque de rayon R divisee par la valeur de la surface du disque) en fonction du rayon du disque. Autour du Petit St Cassien : Ds = 221 R 1 6 Ce modele est aussi valide autour d autres avens importants mais pour des rayons plus courts A l'echelle du massif la karstification correspond done a des structures hierarchiques emboitees. 1. Introduction La karsti fication est un processus de redistribution spatiale de bicarbonates pour l'essentiel. Cette action de mobilisation puis de depot s 'effectue aussi bien a I' interieur du karst que entre un karst et son environnement. Cette dynamique permet d agreger a l interieur du karst des volumes emplis d'air et/ou d'eau dont la succession constitue le cavernement. Pour partie celui-ci est connu grace a des inventaires realises par des speleologues. Les morphologues ont essaye essentiellement pour des for mes de surface de decouvrir des regles traduisant I 'organisation spatiale des formes Ces regles etablies empiriquement par HOR TO ( 1945) par exemple sont aujourd hui mieux comprises car englobees dans la geometrie fractale (MANDELBROT, 1995) Ce type d approche est encore peu frequent dans les karsts tant pour les formes de surface (MARTIN, 1995) que pour les formes del endokarst(CURL, 1966, 1986; MARTIN 1996) La St Baume a beneficie de recherches speleologiques tres intenses, au moins depuis la 2 ""' guerre mondiale. Ces travaux ont abouti a la publication d'un inventaire des cavites tres com plet (C.A.F. & S.C.M., 1987 a, b). Les donnees utilisees en sont issues et ont ete completees (FRANCO, 1994) Sur la St Baume s ouvrent 185 cavites (fig I) de 0 a 321 m de denivellation (D) qui correspond a la valeur absolue de l ecart entre la cote mini male et la cote maximale mesurees par rapport a l'entree Cette donnee est relativement sure car les cotes ont generalement mesurees avec precision. Elle nous renseigne sur les epaisseur s de roches carbonatees affectees par la karstification. Nous faisons I 'hypothese que ce s donnees sont repre ,--7,--------:;~e:.c;;;:;;;.-------=::::;:::::::~~7',---------.. ;ii~-7 sentatives de la karsti0 2km ...._____. fication du massif. Figure I : Carte du massif de la Sai11te Baume A talweg B entree de cavite C source pri11cipale D polje E village. Symposium 7: Physical Speleology 129

PAGE 143

2. Nombre de cavites et denivellation Les distributions d'objets geographiques classes en fonction de leur taille s ont hyperboliques comme par exemple la surface de lac s ou d iles la longueur des cours d'eau (KORCAK, 1940 ; FRECHET, 1941 ) Elle s suivent la loi de Pareto : si N est le nombre d'observations ou la variable envisagee est superieure a un nombre X et A et B 2 constantes positives: N =A / X et done : N = A x OU encore A = N x B etant l'exposant de Pareto Les ca v it e s ont ete classees dans un ordre decroissant, puis affectees d'un rang. La totalite de cette distribution ne s'ajuste pas a la loi de Pareto et done au modi:le bilogarithmique rang / taille de ZIPF ( 1949) Cette distribution peut etre par contre ajustee a un modi:le exponentiel a condition d'exclure l'aven du Petit Saint Cassien ; cavite la plus profonde Pour cela nous l'avons di v isee en 13 classes de I 0 m et nous avons detennine le nombre de cavite s entrant dans chacune des classes En effectuant successivement la somme des effectifs de chacune de ces c l asses nous etablissons la relation : C = 287 I 0 '"'-"" 01 avec : r = 0 99 ; n = 13 ; D etant la denivellation et C le nombre de cavites d'une denivellation infe rieure a D. Ce type de relation a deja ete mis en evidence pour la frequence des dolines cla s see s par taille (WHITE & WHITE 1987) Ce type de distribution exponentielle traduit le caracti:re aleatoire de l'occurence des evenements. La denivellation du Petit St Cassien (321 m) est proche de la difference d'altitude entre l'entree de l'aven et le seuil de debor dement de la Foux de Nans ; source la plus importante et la plus proche (MARTrN I 99 I). Je pense que cet aven, et son long reseau (9 3 km) correspond a une phase plus profonde et peut etre plus recente de karstification. II convient done de le reinte grer dans la distribution quitte a faire evoluer le modi:le de refe rence Le modi:le de ZIPF ( 1949) etendu par MANDELBROT ( 1953, 1995) derive de la theorie de la gestion de !'information a l'interieur d'un systi:me complexe (FRONTlER & Pl CHOD-VlALE, 1991 ) Cette loi correspond, par exemple pour une langue vivante, a une optimisation des signaux. Par analogie nous pouvons penser que sa verification sur une distribution de denivellations de cavites correspond a une optimisation des investigations speleologiques e t/ ou a une optimisation des mor phologies necessaires au transit des precipitations Si l'on ordonne les classes de denivellation par ordre de fre quence d apparition decroissante (F) et si l'on porte le logarithme de cette frequence en fonction du logarithme du rang (R) nous obtenons un alignement de points qui pennet d'ajuster une droite. L'equation est done de la forme : Log F = -g Log R + k ; g est la pente ; k est une constante Puisque lorsque R = I Log F = k on peut ecrire : Log F = -g Log R + Log F et done : F = F R '. Pour la St Baume (fig. I) l'aju s tement n'est possible que pour Jes frequences inferieures a 0 I L'equation est la suivante : F = 2,07 R '"'. L'ajustement est bon : r = 0 99 et n = 12 mais la figure I montre une courbure pour les plus fortes fre quences c'est-a-dire pour les classes de cavites de faible deni vellation Ce relatif manque de petites cavites peut etre quanti fiee Cela pennettra de comparer di fferents karsts et de faire la part entre les 2 hypotheses avancees ci-dessus (exploration par tielle e t/ ou etat de l'epikarst et de l'endokarst) Le modi:le prend alors la fonne suivante : Log f" = -g Log(R+l3} + Log f.. Ce que l'on peut aussi ecrire : t" = f' (R+l3 r'; -13est un parami:tre qui pem1et "d'imposer" au rang R un decalage -f'est le point d'intersection entre l'axe des ordonnees et !'extrapolation de la droite de pente -g pour les fortes frequen ces Pour R = I on a : f = f 13' Cette courbe est done !'asymp tote de la droite d'equation : Log F = -g Log R + Log F ,. Avec R = I et 13 = 0 on a : F = f '. Nous pouvons ainsi calculer 13 a partir de !'equation suivante : Log f Log f = g Log ( I + 13) c.. ., V C ., ::, .g J.. c.. !OT-" _______________ ro I FI ,01 Modtlt C Frfqutnce Frfq.Rtgru. ,001 I Rang= R Couples de valeurs ulilis~s pour 10 efTectuer la rigression F = 2,07 R .2,22 r = 0,99 ; n = 12 100 Figure 2: Ajustement du modele de ZipfMandelbrot a la di strib ution des denive/lations me surees dans les cavites de la St Baume Pour la Sainte Baume en remplai;:ant les tennes par Jes va leurs nous obtenons : Log 2 07 Log 0,308 = 2 22 Log ( I + 13) soit : 0 8274 / 2 22 = Log ( I + 13) => 2 3589 = I + 13 et done 13 = 1 36 L'ajustement du modi:le a la di s tribution est assez bon : r = 0,95 et n = 14. -13qualifie aussi bien le niveau des in v estigations speleolo giques que l'etat de l'epikarst. II pennettra de comparer des karst s tri:s differents de fai;:on a voir s'il est influence par la man i i:re dont l'eau entre sous terre -fne caracterise pas la courbe ii est seulement choisi de telle sorte que la somme infinie I(!") soit egale a I. -gpermet d'apprecier la diversite de la distribution. Une forte pente ,;orrespond a des effectifs reduits pour de foibles profondeurs et done a des effectifs relativement importants pour de fortes profondeurs, et inversement. Ilg etant une dimension fractale (D ), nous avons : D = I / 2,22 = 0 45. Cette distribution de denivellations s'ajuste done a un modi:le hierarchique revelant une certaine organisation a l echelle du massif Si les foibles effectifs des classes regroupant Je s petites cavites traduisent un manque dans les investigations speleologi ques, nous pouvons penser que la distribution complete doit s'ajuster au modi:le de ZIPF (1949) Si au contraire le faible effectif de ces classes correspond a une morphologie specifique de l'epikarst, la valeur de -13doit pouvoir etre reliee a d'autres caracteristiques (fracturation, couverture etc.) Cette approche ne nous a pas pennis de prendre en compte la position des cavites et plus particulierement celle de leurs entrees et done de nous interesser a I organisation spatiale des reseaux de drainage de l'endokarst. Sont-ils repartis de fai;:on reguliere dans le massif ou au contraire traduisent-ils un autre niveau d organisation ? Ont-ils tous la meme organisation? Par analo gie et en prenant comme reference Jes reseaux de drainage ae riens, ii est logique de penser que Jes reseaux de drainage de l endokarst sont aussi spatialement hierarchises. 130 Proceedings of the 12 th Internat i onal Congress of Speleology 1997 Switzerland Volume 1

PAGE 144

3. Repartition spatiale des denivellations La carte (fig.I) montre qu'il l'ouest s'ouvre un nombre plus important de cavites qu 'ii I 'est. Les gradients altitudinaux y sont plus forts (Pie de Bertagne I 042 m ; source de Saint Pons 320 m) qu 'ii l'est (Mourre d 'Agnis 9 I 9 m; source de la Fig u iere 305 m) Localement nous observons des regroupements d'entrees de cavites, et clans tous Ies secteurs du massif, une cavite plus im portante est entouree d'autres moins profondes La repartition du cavemement n est pas egale. Ceci est conforme ii la division en sous bassins du ma sif etablie par !'analyse du fonctionnement (MARTIN, I 99 I). Ce drainage clans l 'endokarst est-ii localement organise? ous faisons l'hypothese qu'il l echelle de la dizaine de km se sont developpes des reseaux correspondant ii u ne structure arborescente semblable ii celle que nous observons en surface (MARTIN, 1995) mais qui est, clans le cas du karst, developpee clans Jes 3 dimensions Une source en constituant l'extremite aval ; des avens et des galeries en etant les entrees et les branches Si I anaJogie evoquee ci-dessus est correcte cela impliqu~ qu 'ii toutes Ies echelles se repetent de formes semblables Les reseaux explores ne sont que ceux qui sont de taille suffi~ante pour per mettre le passage des speleologues mais ils ne do1vent pas av01r une forme differente de ceux, beaucoup plus etroits qui resteront ii tout jamais inexplorables, ou de ceux qui n'ont pas encore ete decouverts Je pense done que les donnees acquises sur les re seaux connus sont representatives de !'ensemble de la structure karstique tout comrne les connaissances acquise~ sur le reseau d'un sous bas in sont representatives, avec certames hm1tes, de !'ensemble du reseau d'un bassin versant. La cavite presentant localement la plus forte denivellation se ra consideree comrne etant le drain principal avec lequel les autres cavites confluent car je considere que la decouverte est plus facile, pour les speleologues, lorsque la karstification a ete plus importante. La karstification est elle, d'autant plus unpor tante que le drain est plus ancien et qu 11 drarne vers IUI un maximum de debits Cela a pu etre etabli par tra<;:age pour l aven du Grand C l apier et I'aven du Petit Saint Cassien mais cela reste une information exceptionnelle sur ce massif Par contre de telles confluences sont souvent revelees au fur et ii mesure que les travaux speleologiques progressent. II arrive toutefois _qu'assez souvent un retrecissement rende impossible la poursutte de la progression du speleologue jusqu'au site de confluence C~s secteurs etroits semblent aller de pair d'une part avec une d1m1 nution croissante des poss i bilites de mise en solution de la roche avec I 'augmentation de la profondeur ou plus exactement avec !'augmentation du temps de sejour de l'eau chargee d' a~id e carbonique (pCO ), mais aussi d'autre part avec Ies c~r~ctens tt ques des connexions preexi stent~s entre le~ d'.sc?~tmu1tes '.mtta Ies (fracture joint de strat1ficat1on, etc.) a I ongme du reseau Ceci n 'exclut pas en outre des possibilites locales de comblement (concretion, remplissage, etc.) Dan~ tous les ea~, !'absence de confluences etablies par exploration ne s1gmfie nullement !'absence de reseau unitaire D apres Jes Iois de HORTO (1945) si !'organisation du drainage superficiel est de fom1e arborescente, nous devon~ observer que la longueur cumulee du reseau au~m~n~e au fur ~t _a mesure que nous integrons des drains d'ordre mfer~eur. Exp~n mentalement ii est possible de montrer que la cro1ssance d un reseau de drainage s'effectue jusqu'il ce qu'il emplisse l'espace du bassin versant ii !'exception des surfaces de concentration de I eau en amont de chaque talweg initial. Nous pouvons done etablir la surface drainee par chaque ordre dont la somrne est egale ii la surface du bassin versant La longueur cumulee specifique qui est le ra~port ~n tre la longueur cumulee des talwegs par la s~rface c~.m~lee ~ramee par ces talwegs, augmente done elle auss1 a~ec I mtegratt~n de tal wegs d'ordres inferieurs. Ces considerations peuvent ~tre resu mees dans un chiffre qui est la dimension fractale du reseau. Par exemple le reseau de drainage aerien du Caramy ii !'est de la St Baume a une dimension fractale de 1,52 (MARTIN 1996). F i g u re 3 : S c h e mat is atio11 d un r ese au e t m et h o d e e m p l oyee Dans notre approche nous ne pouvons disposer des memes informations pour resumer en un chiffre la qualite du reseau de drainage en raison de la meconnaissance de la totalite du reseau dans laquelle nous sommes. ous a ll ons done u tiliser une autre approche et comparer les denivellations cumulees des cavites ii une surface de reference ou plus exactement etablir la relation qui existe entre la variation des denivellations cumulees specifiques de cavites s'ouvrant dans un disque et le rayon de ce disque. Cette relation doit s'exprimer par un alignement de points en coordonnees bilogarithmiques si elle traduit une organisation fracta l e du reseau. ous pouvons verifie r cela en considerant la surface topo graphique comme un plan tangent au reseau qui supporte Jes entrees (fig.3). En pre n ant comrne point de depart l 'entree de la principale cavite d'un secteur du massif, nous avons dessine une serie de disques concentriques dont les diametres etaient aug mentes successivement d'un km (JUUEN I 992 ; BO EFOY 1993) Ndus avons ca l cule pour chacun d'eux la denivellation cumulee des cavites qui y debutent. Ces denivellations cumulees ont ensuite chacune ete divisees par la surface correspondante du disque. Le resultat est une denivellation specifique (Ds) en m!km '. Pour chaque secteur nous avons ensuite etabli une rela tion bilogarithmique de la forme : Ds = a R entre la denivella tion specifique et la va l eur du rayon (R) en km de chacun des disques Si le drainage s effectuait par des drains sub-paraUeles independa n ts et au moins aussi importants nous observenons soit une valeur identique de la denivellation specifique quelle que soit la su r face consideree et done l'exposant -btendrait vers zero soit une croissance de Ds en fonction de R -btraduit done ia hierarchisation du reseau. Plus la valeur de best im portante et plus le reseau est hierarchise plus celui-ci empli I espace de I' endokarst. . Le graphique ainsi etabli (fig.4) montre un tres bon ~ltg~e ment des points comme en atteste les coefficients de correlatton l ineaire calcules (tab!. I). -bprend des valeurs comprises entre -1 62 et -1 04 Que le secteur ou se developpe l'aven le plus important soit affecte de la valeur de -bla plus forte parait tout ii fait logique tout comrne la tres foible valeur de -bpour le secteur de l 'extremite ouest de la haute chaine ou l'Escandaou, cavite en forme de mono puits est toujours apparu comme une paleo forme deconnectee des dyna miques actuelles meme si cet aven est parcouru par un filet d eau (COULIER, 1985). La plus forte denivellation moyenne (De / Nb C) par cavite a d ailleurs ete ca l culee pour le secteur de I' Escandaou. -aexprime I importance des denivellations cumulees (De) avec : De = 3,2 a ; r = 0,993 ; n = 7. Une relation peut etre etablie entre aet la valeur maximale de R : a= 75,5 + 408,5 I ogRmax.; r = 0,92; n = 7. De croit en fonction du nombre de cavites tout comme -a-. Aucu n e re l ation n'a pu etre etab l ie entre altitude de l entree de la cavite constituant le centre des disques et les caracteristi ques de ces reseaux. Sachant que Jes sources qui sont l eur exutoi res supposes soot situees en peripherie du massif a des altit_udes voisines, ii ne semble done pas que la hauteur de chute so1t un facteur determinant dans la constitution de ces reseaux Symposium 7: Physical Speleology 131

PAGE 145

Au dela de certaines valeurs de R la relation n'est plus li neaire Une rupture de pente apparait. Ce changement e s t une indication qui permet de se faire une idee de extension ma x male de chaque reseau de drainage Elle traduit integration d un autre reseau contigu dans le calcul. Cette technique permet done de representer en surface par le cercle de plus grand rayon !'exten s ion consider e e comme maximale d un reseau endokarsti que C e rtains cercle s s e recoupent assez largement. Cela n est pas pour nou s surprendre car nou s n avons tenu compte ni de l'anisotropie des reseaux ni des structures geologiques Dans le cas de s Encanaux, du Pin de Simon et de l'Escandaou se dessine une large bande a intersection des 3 cercles qui est exempte de cavite II nous parait logique que la limite entre ces 3 reseaux passe a ce niveau Dans le cas de Castelette nous savons qu e cette cavite est morphologiquement en rapport avec le polje parfoi s fonctionnel du Plan d Aups (MARTIN 1991) Plusieurs tentati ves de tra1,age effectuees dans Jes ponors se sont soldees par des echecs L extension aussi loin vers l ouest du cercle circonscri vant une aire en rapport avec le Pt St Cassien nous conduit a penser que le drainage du polje pourrait bien se faire vers cet a v en. A l echelle du massif ces relations bilogarithrniques tradui sent done I' existence de structures de drainage dans I' endokar s t --+-Pt St Cassien Ill Castelette 0 Chiteau Renard 1000 Denivellation specifiq ue Ds en m /km 2 100 10 Ph M 97 X Encanaux Ponchin o Escandaou Pin de Simon Rayon du disque en km 10 Figure 4 : Relations bilogarithmiques entre la denivellation specijique Ds et la valeur du rayon R des disques pour 7 sec teurs de la St Baume hierarchisees plus OU moins developpees et emboitees Le type de fonction nous conduit a penser que cette organisation resulte de processus deterministes generant une morphologie fractale (LAVERTY 1987) 4. Conclusion Cette approche reste imparfaite sur plusieurs points II ne nous a pas ete possible de tenir compte du cheminement des cavites a travers la masse carbonatee tout comme de l'ampleur reelle des vides induits par la karstification II conviendra done de poursuivre cette recherche en essayant de rapporter Jes volu mes dtis a la karstification a un volume de reference comme une demi-sphere. ous ne disposons pas pour instant d'une mesure de l ampleur des vides mais nous pouvons imaginer Jes determi ner a partir d une approche fractale de chacune des cavites (SUSTERSIC 1983 CURL 1986) Par la suite ii sera necessaire Modele bilogarithmique Alti Nb C De Del Cavite= c Rmrui Coef. deC surR en Nb a b km Cor. enm max km C Pt St Cassien 221 -1,62 5,0 0,999 740 36 1361 38 Castelette 44 -1 55 2 0 0 999 600 11 187 17 ChateauR 39 -1 10 2 0 0 999 545 JO 243 24 Encanaux 98 -1,34 3,0 0 994 433 25 644 26 Escandaou 76 -l 04 2,0 0,991 892 l l 445 40 Pin de Simon 117 -1,21 2,5 0,996 540 26 755 20 Ponchin 51 -1 39 2 5 0 999 745 11 265 24 Tableau I : Modelisation de la decroissance des denivel/ations specijiques cumulees autour des principales cavites de la St Baume De = Denivel/ation cumulee. de tenir compte de la position dans l'espace des reseaux karsti ques References BONNEFOY J L. 1993 Frequentation des commerces et s ervice s. Mapp e mond e n 4 p.40-41 CLUB ALPIN FRANCAIS & SPELEO CLUB DE MARSEILLE 1987 a-b La St Baume souterraine Tome I 2 ed ., Je s B du Rh ., 110 p et Tome 11, le Var. 217 p COULIER C 1985 Hydrogeologie karst i que de la Sainte Baume occidentale B du Rh ., Var ; Fr These de 3 cycle U ni v ersit e de Pro v ence A i x-Marseille I 400 p CURL, R L. 1966 Caves as a measure ofkarst J G eo l ., v ol.74 n 5 part 2 p 798 830 CURL R L. 1986 Fractal dimensions and geometries of c a v es Marh e mari c al Geology vol.18 n 8 p 765783 FRANCO, A 1994. Cavites du Var de plus de 100 m de den iv el e. Sp e lun c a bull e rin n 53 p 13 FRECHET M 1941 Sur la Joi de repartition de cenaines grandeurs geographiques J. Soc Srar Paris vol.82 p 114-122 FRONTIER, S & PICHOD-VIALE D 1991. Ecosysteme s. Struc ture fonctionnement evolution Masson e diteur P a ris 392 p HORTON R E 1945. Erosional development of s tream s and their drainage ba s ins : hydrophysical approach to quantitati v e morpholo gy. 8111/ e rin ofrh e G e ologi c al So c i e ry o f Am e ri c a 56 p 275-370 JULIEN R 1992 The application of fractals c ol101dal aggre g ation Croati c a c h e mi c a a c ta n 2 p 215-235 KORCAK, J 1940 Deux fypes fondamentaux de distribution s s ta ti s tique s. Bull e tin d e /'lnstitut lntemarional d e Stati s tiqu e, vol.30 p 295-299 LAVERTY M 1987 Fractals in karst. Earth s u ,face pr ocesses and landf o rm s, vol.12 p.475-480 MANDELBROT B 1953 Contribution ii la theorie math e matique de s communications These Universit e de Paris, Pub!. Inst. Stat. Un iv e. Paris 2 121 p MANDELBROT B 1995 Les objets fractal s, 4 ed Edition s Flam marion Coll. Champ n 30 I 208 p MARTIN Ph 1991. Hydrogeomorphologie de s g eo s ysteme s ka rs ti que s de s versants N et W de la St Baume (B du Rh ., Var ; Fr ). Th es e Universit e d'Aix Marseille II 326 p MARTIN Ph 1995 Les formes tluviatiles du massif de la St Baume ( B du Rh Var ; Fr) Th e oreti c al and a ppl ie d kar s t o l ogy, v ol. 8, p 103-115 MARTIN Ph 1996 De ) organisation de s forme s s uperfici e lle s et souterraine s du massif karstique de la St Baume ( B du Rh Var ; Fr ) Ukp ic n 8 Univ. de Fribourg Suis s e M Monbaron et S Fierz e diteurs p.45-64. SUSTERSIC F 1983 Determination of unknown cave p as sa g es length by means of fractal analysis ew trends in speleology Jancarik A editeur Dobrihovice p 61-62 WHITE W B & WHITE, E L. 1987 Ordered and stocha s t i c arran gement s within regional sinkole populations in proceeding of 2 "" multi di s ciplinary conference on sinkholes and environmental impacts of karst, Orlando Floride ; sous la direction de Beck B F et Wilson W L. A A.Balkema editeur p 85-90 ZIPF G K. 1949 Human behavior and principe of lea s t effort Cambridge MA : Addison-Wesley 132 Proceedings of the 12 th International Congress of Speleology 1997 Switzerland Volume 1

PAGE 146

A comparison of the orientation of cave passages and surface tributary valleys in the karst of southwestern Wisconsin, U.S.A. by Craig A. Terlau and Michael J. Day Department of Geography, University of Wisconsin-Milwaukee, Milwaukee, Wisconsin 5320 I, U.S A Abstract It has been suggested independently that regional structure in the form of bedrock fractures exerts a control over the orientation both of cave passages and surface tributary valleys If this is so then these features in a given area might be expected to have statisticall y similar orientation data sets This hypothesis was tested by comparing orientation data for adjacent cave passages and tributary valley s in the karst of s outhwestem Wi s consin Cave passage orientation data was derived from published maps of 21 caves in Richland County Wisconsin Orientation data for adjacent tributary surface valleys was derived from I : 24 000 U S Geological Survey maps using computer-assisted techniques Rose diagrams generated from the data sets suggested visually a close relationship between cave passage and tributary valley orientations Of20 Kolmogorov-Smimov tests in only four cases were cave and valley orientation data sets dissimilar at an alpha level of 0 1 In 80 % of cases there was a strong relationship between cave and valley orientations, suggesting that the orientation of these features in the southwestem Wisconsin karst may indeed be influenced by a common set of bedrock fractures Zusammenfassung: Vergleich der Richtungen von Hohlengangen und oberflachlichen Talformen im Karstgebiet im sOdwestlichen Wisconsin USA. E s wurde verschiedentlich postuliert, dass regionalgeologische Strukturen niimlich Kliifte, einen Einfluss auf die Orientierung sowohl v on Hohlengiingen als auch von kleineren Talformen auf der Oberflache (Trockentiiler) haben Wenn dem s o ist, miissen solche Formen in einem gegebenen Gebiet statistisch iihnliche Richtungen aufweisen. Diese Hypothese wurde an Hand von Vergleichen der Richtungen von Hohlengiingen und von oberirdischen Seitentiilem im Karstgebiet im siidwestlichen Wisconsin gepriift Die Oaten zu den Gangrichtungen wurde den Pliinen von 21 Hohlen in Richland County entnommen jene der im gleichen Gebiet v orhandenen Obertl ii chen-Seitentiiler aus den topographischen Karten U S G S. I : 24 '000 mittels computerunterstiitzten Techniken Auf diesen beiden Datenpaketen basierende Rosetten-Diagrarnme zeigen visuell starke Richtungskonvergenzen zwischen Hohlengiingen und Oberflachentiilem Dies wurde mit 20 stati s tischen Vergleicben erhiirtet: Es gab nur 4 Fiille in denen Hohlenund Talorientierung mit einer lrrtumswahrscheinlichkeit von 0 1 voneinander abwichen Also deuten 80% der Fiille darauf bin dass die Orientierung dieser Formen im siidwestlichen Karstgebiet von Wisconsin durch gemeinsame Grundgesteinsfrakturen beeintlusst wird 1. Introduction Karst caves usually form through dissolutional enlargement of fractures within carbonate bedrock (TR U DGIL 1985 ). These fractures, however do not occur randomly rather they are the re s ult of directional stresses placed upon the rock Studies of bedrock fracture orientations show that they occur in preferred directions (PARI Z EK 1976) Thus, if fractures are the dominant directional control mechanism in the formation of a particular cave, the straight sections of cave pas s age would be expected to have orientations similar to those of local fractures Dry valleys are other karst landforms which can develop along existing bedrock fractures Drainage waters will follow the hydraulic gradient but may find the path of least resistance along surface fractures for at least part of their course A valley which is cut into bedrock should appear linear along sections where surface fractures were utilized (PARIZEK 1976) these linear sections of valleys may be expected to have an orientational distribution similar to that of local fractures If both ca v es and linear valleys in an area have developed under the influence of the same structural control data sets of their orientation s should be statistically similar. The purpose of this paper is to examine the orientational relationship between linear sections of cave passage and linear tributary v alleys and to measure the extent of this relationship Based on theory, bedrock fractures exert significant control over the development of linear sections of both caves and tributary valleys The hypothesis being tested is that their orientational distributions are similar Further this relationship is expected to occur locally The valleys to be measured for orientation are those occurring in the areas immediately adjacent to each cave location These valleys have been chosen for the following reasons: I) because of their proximity the valleys and the caves are likely to have developed along the same fractures ( LAPOINT E & HUDSON 1985), 2) the scale of the caves and the tributary valleys is similar 3) the caves and tributary valleys occur at similar stratigraphic levels, and 4) the valleys are cut into bedrock and have little superficial cover 2. Previous Studies of Cave Passage Orientations There have been many previous studies involving comparison of cave passage orientations with orientations of other types of features, although few comparing cave passage and dry valley orientations. For a survey of such studies, see TERLAU ( 1995) Symposium 7 : Physical Speleology 133

PAGE 147

3 Bedrock Geology of Richland County In Richland County the interfluves are capped by lower Ordovician limestones and dolostones of the Prairie du Chien group The main valleys are cut into the underlying Jordan sandstone formation Beds dip gently to the southwest at about one degree (HEYL et al 1959) Very gentle folding can be observed where beds are exposed The Oneota formation of the Prairie du Chien group is the major cave forming member north of the Wisconsin river (DAY I 986). It attains a thickness of about 50m (HEYL et al 1959). Vertical fractures cut through the Jordan sandstone and the Prairie du Chien dolostone. Single vertical joints are traceable for up to a mile and cut as much as I 00m of beds (HEYL et al 1959) The Upper Mississippi Valley has repeatedly experienced regional stresses although record of orogenic events during the Paleozoic and Mesozoic is incomplete and interpretation is speculative. Fracturing of the bedrock has been attributed to regional Paleozoic orogenic stresses (Dutch 1981 ), to post-depositional shrinkage (LEITH, I 932), and to glacial crustal bending (MCGINNIS, 1969) 4 Karst and Caves of Richland County Richland County consists of a karstic upland dissected by fluvial drainage and with a local relief of about I 00m. The karst is developed on the ridges. which are narrow. gently curved on top and steep-sided. Cliffs and bedrock outcrops are common. Kar s t features of Richland County inc l ude over 50 sinkholes, over 600 springs, thousands of dry valleys and 44 caves (DAY et al 1989). In the absence of accurate dating, the age of the caves in south western Wisconsin is unknown (DAY. 1986a. b ) Development commenced sometime between the Cretaceous and the Pleistocene. an earlier date being suggested by the fact that the caves were well integrated with the pre-Pleistocene drainage system There are 44 known caves in Richland County with a total of 21 00m of passage The caves are small in both passage dimension and length (DAY, I 986a,b). They are relatively inactive and show limited signs of current dissolutional activity. Their ridgetop position places them well above the present day water table and gives the caves a limited catchment area Of the 44 known caves in Richland County, 23 (52%) are formed in the Prairie du Chien dolostone, and 21 (48 % ) are sandstone caves. Cave lengths are from 3m to 335m. Of the caves 25111 or more in total length, 93% are formed in dolostone. The mean length of caves in dolostone is I 00m atural cave entrances occur where sinkhole collapses break through to the ceiling of the cave and as outcrop entrances where downcutting of valleys has intersected the cave passage. While Richland County has 44 known caves which have connections to the surface, doubtless there are many others that do not. One way of approximating the number of caves with no surface connection is to place known caves into classes according to their number of entrances. Of 40 of the 44 caves in Richland County there is on l y one cave with three entrances, there are three caves with two entrances and there are 36 caves with one entrance "Statistical distribution functions and geologic reasoning agree that a substantial class of caves with no entrances should exist." (WHIT E 1988 p 61 ) Furthennore the class sizes of caves according to their number of entrances is suggested to follow Poisson distribution (WHITE, 1988). Applying a likelihood function to the cave entrance data from Richland County. there are like l y some 187 caves with no entrance (T E RLA U, 1995) The caves of Richland County are not regularly distributed throughout the county, but occur in clusters reflecting outcrops of the Praire du Chien Dolostone. Nearestneighbor analysis (EBDON, 1981) resulted in a calculated value c of -2.096, and a 98% probability that the distribution of caves in the study area is significantly clustered (T E RLA U 1995) Cave clusters were identified by locating groups of poin ts which are reflexive in that all points in the group have their nearest-neighbor within the group and that each point within the group has a nearest-neighbor distance below the value for all points in the sample area. In Richland County, seven groups of caves fit this criteria. 5. Methodology All linear measurements in the orientation analysis are in English units, since these are common to the USGS topographic sheets and the 21 existing cave maps which could be located. These maps are to British Cave Research Association grade 3 to 7 (ELLIS 1984) "Caves of Richland County, Wisconsin" (PETERSON, 1968) was the source for 12 of these maps. Nine other published cave maps were obtained from sources listed by TERLAU ( 1995). Cave P assage O rie n tat i o n s For this study, orientation lines were drawn to parallel the walls of straight sections of passage. and to parallel linear passage features whe r e straight sect i ons do not exist Sections of the caves which could not be interpreted as linear were not measured The existing cave maps were scanned I : I as 600 dot per inch PICT files. Adobe Photoshop, a graphics program which performs plotting of l ines and measuring of angles, was used to obtain cave passage orientations The angle of the line being drawn is displayed in an information box Measurements had a high level of precision ; the repeatability of measurements taken from cave maps was found to be between 0.1 and 0 5 depending on the length of the feature being paralleled. For full details of the measurement techniques, see T E RLA U ( 1995) L in ear Va ll ey O rie nt ations 20 USGS I :24,000 topographic sheets provided full coverage of Richland County Caves in which passage orientations were measured were located and marked on the topographic sheets. The portions of the topographic sheets showing the cave location and the interfluvial ridge associated with the cave were scanned I : I as 600 dot per inch gray-scale PICT images. The mean area scanned from the topographic maps was 3 87 square miles Five of the scanned map sections contained more than one cave therefore the mean map area per cave used for analysis was 2. 71 square miles. Linear tributary valleys were identified on the maps by contour analysis (TERLAU 1995). Map measurements were found to be statistically similar to field measurements using a paired t-test. Adobe Photoshop was used to plot lines showing the trend of linear va ll eys and measure their orientation Repeatability of measurements was found to be generally between 0 5 and 2 0, varying with the length of the valley being measured; longer valleys could be remeasured with greater precision than short valleys. A minimum of four contours had to define a linear valley in order for it to be included in the data The contour interval on the maps is 20 134 Proceed i ngs of the 12 '" International Congress of Speleology, 1997, Switzerland Volume 1

PAGE 148

feet. The data collected from linear tributary valley orientations is shown in full by TERLAU ( 1995). Data C omp a ri s on s As stated in the hypothesis, it was expected that cave passage orientations wi ll be distributionally similar to the orientation of nearby linear tributary valleys. This relationship was tested by comparing visual representation of the data distributions and by usi n g statistical techniques. Selected comparisons were made using the origina l data sets, combined data sets and data subsets (TERLAU, 1995). Of the 12 scanned map sections used to measure valley orientations, 7 contained only one cave, and 5 contained multiple caves. For the cases in which there is one cave to a map section, comparisons were made between the cave data and the valley data from the correspo n ding map s h eet. How data comparisons were made for map section containing multiple caves was decided on a case-by case basis (TERLAU, 1995) A program in True BASIC programming l anguage produced ro e diagrams with class intervals of I 0 The data were plotted according to frequency of occurrence i n each of the classes 20 data comparisons were performed (TERLAU, 1995). S tati st ic a l C omp a ri s on o f t h e Data Sets: A Kolmogorov-Smimov twoided, two-sample test was used to compare the data sets. The test analyzes the amount of agreement between the distribution functions of two independent data samples to determine if the two samples were drawn from the same popu l ation (or populations with identical distributions), or were drawn from popu l ations with different distributions The test is non-parametric, making it useful when no assumptions can be made about the distribution of values in the population from which the samples were taken The distributions of data used in this study are like l y to be multi-modal. The K-S test can be app l ied to data without grouping them into classe and also has the advantage of allowing samples of unequal size to be compared 6. Results Kolmo go r ov-S mirno v A nal ys i s The 20 Kolmogorov-Smimov tests comparing cave data to valley data, led to a rejection of the null hypothesis in 4 cases Therefore, 4 out of 20, or 20 % of the tests showed the distribution of the sample values of cave passage orientations to be significantly different than those of linear valley orientations at alpha = 0.1 This implies a strong rela t ionship between the orientations of caves and valleys in the remaining 80% of the cases. Full results of the ana l ysis are presented by TERLA U ( 1995). Ro se Di agr am A n a l ys i s Rose diagrams were generated to visually represent the distribution of values in the cave and valley data sets. An attempt was made to quantify the similarity of class distributions shown in the rose diagrams (TERLAU, 1995) In 11 out of 20, or 55 % of the valley roses, more than half the classes corresponded to the cave roses. In visual comparison, there were eig ht instances where there appeared to be a strong re l ationship between cave orientations and valley orientations. Although comparing data sets using rose diagrams is subjective, 40% of the cave rose diagrams exhibit large classes which correspond with valley rose diagrams. The rose diagrams generated from the combined cave data and combi n ed valley data showed considerable scatter in the distributions. Prominent orientations did emerge from the plots of these large data sets, however. The cave passage orientations showed a peak in the distribution centered at 20 and a second overall trend centered at 95 The valley orientations had peaks in the distribution centered at 0 65 110 and 145 The plots of these two distribution are dissimi l ar which confirms results of the Kolmogorov-Smimov analysis. The rose diagram generated from the fracture data had far less scatter than the plots of combined cave or combined valley data. Peaks in this distribution occurred at 5 35 75 and 145 The distribution roughly correlates with plots of the combined cave data again supporting the results of the Kolmogorov-Smimov analysis. 7. Analysis and Conclusions In assessing the resu l ts, it is important to consider the stress and erosiona l history of the study area. The bedrock in which the caves and valleys have developed has been subjected to repeated stresses res u lting in fracturing Initial fracturing may have occurred as a result of shrinkage, and orogenic events d u ri n g the Pa l eozoic resulted in further fracturing. Caves developed prior to the Pleistocene, but valley incision by meltwater was particularly acute during the Pleistoce n e, and glaciation then again exerted stresses resulting in additional fracturing. Hence fractures produced by glacial stresses may have influenced linear valley formation but not prior cave formation. "When the jointing pattern present in a rock mass originates from multiple fracturing episodes, the first set of joints may be spatially less variable. Subsequent stress relief leading to new fracturing may be strongly influenced by these pre-existing fractures so that they terminate against the original joints and consequently exhibit more irregular frequencies and orientations and locations (LAPOINTE & HUDSON, I 985) The main valleys of Richland County may have developed along fractures which were produced by orogenic stresses du r ing the Paleozoic In the pre-glacial drainage network the caves may have been integrated with these stream valleys. Glacial stressing during the Pleistocene may have further fractured the rock mass producing new fracture sets which terminated against existing fractures as described by LAPOINT E & H UDSON ( 1985). Tributary valleys of the present day ridge-and-ravine drainage network developed having available all the fractures that were present when the caves formed, plus the new fracture produced by glacial crustal bending. The overall agreement between cave and valley orientations when compared locally was found to be 80% based on the statistica l analyses. This is a significant trend which suggests that both caves and valleys developed utilizing fractures with orientations that were locally homogeneous While it is likely that additional fractures were available to control valley formation, the 80% overall agreement between cave and valleys orientations suggests that these additional fractures were either of a different scale than those which controlled the features observed in this study, or these new fractures paralleled existing fractures .. However. the orientations of cave passages for the entire county were found to be statistically dissimilar to the orientations of linear valleys for the entire county. This suggests that the orientations of fractures of the scale which controlled the development of cave passages and tributary valleys are locally inhomogeneous Symposium 7 : Physical Speleology 135

PAGE 149

References DAY, M.J. I 986a. Caves in Southwestern Wisconsin, U.S.A. Proceedings 8th International Speleological Congress, 155-157 DAY, M.J 1986b Caves in the Driftless Area of Southwestern Wisconsin. The Wisconsin Geographer 2, 42-51 DAY, M.J., RE EDER, P P ., & OH, J.W 1989 Dolostone Karst in Southwestern Wisconsin The Wisconsin Geographer 5 29-40. D UTCH, S.l. 1981 Lineaments and faults of Wisconsin Minnesota and the Western part of the Upper Peninsula of Michigan. U.S. Geological Survey Open File Report 81-977, 29pp. EBDON, D 1981 Statistics in Geography Oxford England. Ba si l Blackwell I 95pp. ELLIS, B .M. 1984 Surveying a Cave. In : Caving Practice and Equipment. ed., Judson D David & Charles Pub l ishers, Newton Abbot l69ff HARD ER, E.C. 1906 The joint system in the rocks of southwestern Wisconsin and its relation to the drainage network Univ Wis ., Bull. (sci ser. 3) No.5, 207-246 H EYL. A.V., A GNEW, A.F. LY ONS, E F & BEHR, C.H. 1959 The geology of the upper Mississippi Valley zinc-lead di st rict. U S Geological Survey Prof. Paper 309, 31 Opp. LAPOINTE P R. & HUDSON, J I 985. Characterization and interpretation of rock mass joint patterns Special Paper 199 Geological Society of America, 37pp. LEITH C.K. 1932. Structures of the Wisconsin and Tri-State lead and zinc deposits Econ. Geo l ogy 27(5) 405-418 MCGINNIS, L.D. 1968 Glacial crustal bending G S .A., Bull. 79(6), 769-776 PARIZEK R .R. 1976. On the nature and significance of fracture traces and lineaments in carbonates and other terraines Karst Hydrology and Water Resources Water Resources Publications 62pp. PETERSON G M. 1968 The caves of Richland County, Wisconsin. The Wisconsin Speleologist 7(3) 78-109 TERLA U, C A 1995 An analysis of cave and valley orientations in Richland County, Wi sc onsin MS Thesis, University of Wisconsin-Milwaukee, I 62p. TRUDGILL, S 1985 Limestone Geomorphology Longman, London and New York 196pp WHITE, W 8 1988. Geomorphology and Hydrology of Kar s t Terrains Oxford University Press New York and Oxford 464pp. 1 3 6 Proceedings of the 12 t h International Congress of Speleo l ogy 1997, Switzerland Volume 1

PAGE 150

Labyrinthes et Sieben Hengste par Alex Hof Ch. du Laze B, CH-1806 St-Legier Abstrac t By caving in Sieben Hengste, we find many labyrinths The author discerns any types of labyrinths and explain s her genesis Her exploration necessitates perseverance and systematic when we want understand Good luck when you explore it. Resume En explorant aux Sieben Hengste on rencontre de nombreux labyrintbes L'auteur differrencie quelques types de labyrinthes et explique leur singuliere formation Leur exploration demande de la perseverance et de la systematique si on veut les comprendre L'auteur encourage les s peleologues qui se lancent dans une pareille aventure 1. Introduction En explorant la region Sieben Hengste-Hobgant cette zone karstique des prealpes suisses bien connue pour le grand Reseau qu'elle contient on rajoute inlassab l ement des conduits dans un enchevetrement parfois deja complexe La contemplation de certaines zones du plan d'ensemble souterrain donne ]'impression de grands labyrintbes En fait, comment definir un labyrinthe ? 2. Definition du labyrinthe Parmi quelques definitions du mot labyrinthe proposees par des dictionnaires retenons en deux qui s'appliquent bien a la speleologie: reseau comp l iq u e de galeries d o n t o n a peine a sortir ; ncbevetrement, complication inextricable Le terme evocateur labyrinthe convient done bien. Cependant, i i vaut la peine d'etablir quelques distinctions 3. Les pseudo-labyrinthes En examinant des plans certaines zones ressemb l ent a de touffus labyrinthes Par co n tre vu de sous-terre on ne partage pas cette impression On parcourt des galeries avec des carrefours espaces On n'a pas conscience de galeries procbes, mais sans connexions penetrables II peut aussi s'agir de plusie u rs etages sans liens directs dont les galeries ne s'encbevetrent que sur le plan complet par additions des differents etages. Dans ces cas ii ne s'agit pas de veritables labyrintbes. Une conjonction de facteurs favorables a la formation de galeries provoque une concentration de conduits d'age et de morpbologie divers Citons en particulier la tectonique l'abaissement par etapes du niveau phreatique, la presence de plusieurs horizons impermeables etc Pour plus de details on consultera HOF ( 1995) Dans la region Sieben Hengste Hohgant Jes exemples de pseudo-labyrinthes sont tres nombreux Citons le cas extreme du Blatersystem decouvert sous deux etages connus dont l'un etait auparavant considere comme le niveau de base. Par endroit l'etage du Blatersystem, pourtant pas bien epais, se subdivise en trois niveaux, dont deux separes par moins de de u x metres de roche Heureusement un atroce jeune meandre jonctionne ces deux niveaux de galeries phreatiques a n ciennes qui se croisent. Sans Jui nous aurions doute longtemps d e ] 'exactitude de noire topographie. Pour detecter les pseudo-labyrinthes ii faut abandonner la division par cavite selon des criteres de penetration humaine Seuls des plans d'ensemble de toutes l es cavites placees dans un meme systeme de coordonnees revelent toute leur etendue Aux Sieben Hengste ces pseudo-labyrinthes vont encore s'etendre et s'etoffer Au-dessus de certaines zones deja complexes la prospection systematique n est pas encore faite Sous-terre exploration devient auss i systematique et amene des decouvertes meme a des e n droits du plan ou ii n'y a plus de place pour tirer un trait. Zone d e p se udo-lab y rinth e du R ese au 4. Les enchevetrements de puits Un seu l ruisseau re n contrant une zone favorable de fractures peut deja creer des puits paralleles par capture de plus en plus precoce Ajo u tez-y un d euxieme voire un troisieme ecoulement p our developper une in t eressante densite de puits lntercalez maintenant quelques co u ches moins solubles pas trop inclinees ou mieux encore des joints mameux Ainsi vous creerez des lucames ou des galeries d e liaisons a plusieurs niveaux Arrosez le tout pendant de nombreux siecles d' une eau bien acide. Voila la recette q u i vous donnera un joli labyrinthe tridimensionnel. Nous avons appliq u e cette recette aux Puits Johny mais nous ne disposions que de deux ecoulements Le nombre de puits n'est done pas tres gran d P ar contre, nous avons bien reussi leurs connexions et en soignant particulierement la qualite de l'eau nous avons tout-de-meme obtenu deux P96 un P80 un P35 et une d emi-douzaine d'autres p lus petits. Au Gouffre de la P entcote, nous avons traverse au plus vite un interessant entrelac de puits Quel courageux topographe osera se lancer a les relever systematiquement? Symposium 7 : Phys i cal Speleo l ogy 13 7

PAGE 151

5. Les labyrinthes classiques Un vrai labyrinthe se peryoit deja sous terre. Les carrefours se suivent et se ressemblent. L'explorateur retrouve difficilernent son chemin Plusieurs modes de formation expliquent le developpement de labyrinthes. Distinguons d abord entrre ceux creuses en regime noye et ceux resultant d'ecoulements vadoses La premiere categorie s e subdivise selon la vitesse du courant et la chronologie. 5.1 Les labyrinthes locaux Parmi Jes plus beaux labyrinthes, on trouve ceux creuses par un ecoulement lent et noye Cet ecoulement rencontre des fractures ou un joint interstrate proposant de nombreuses amorces pour la formation de galeries La corrosion joue un role preponderant. Souvent ii n'en resulte pas un labyrinthe, mais une grande fissure ou un large laminoir. La presence de sediments insolubles favorise le developpement du dedale en bouchant Jes amorces peu ouvertes et en concentrant la dissolution de la roche. Ces labyrinthes peuvent etre tres
PAGE 152

Classification and some morphometric features of salt caves Amos Frumkin Cave Research Section, Department of Geography, The Hebrew University of Jerusalem, 91905, Israel Fax: 972-25820549 Email: Abstract General morphologic features of salt caves are presented in numeric form and used for classification and presentation of cave types. Comparison with surface streams is attempted through the Horton Laws. 1. Introduction Salt caves display a rather simple type of karstification, with physical dissolution of halite. Known caves are of vadose origin, with multilevel branchwork passages. Here some morphologic features of salt caves are discussed, as a basis for numeric classification that may help analyze the caves and compare them with other types ofkarst systems. Salt caves are best preserved in arid environments, because under more humid conditions the entire salt outcrop tends to be dissolved. Mount Sedom salt diapir (ZAK, 1967) at the south-western edge of the Dead Sea has experienced semi arid to hyper-arid climate during the Holocene (FRUMKIN et al., 1991). Since being exposed at the Early Holocene (FRUMKIN, 1996a), salt has been dissolved by runoff collected from small catchments over a relatively insoluble caprock and captured into fissures. Vadose caves, the largest known in salt, have developed as soon as the rock salt was uplifted above the Dead Sea base level. A detailed study of subaerial catchments was performed in order to detect the caves through which the runoff is drained towards base level. An active channel along each cave carries floodwater during short intense rainfall events, lasting typically several hours each year. In spite of the short rainfall duration, downcutting of cave channels is rapid (FRUMKIN and FORD, 1995), exceeding the rising rate of the diapir and Dead Sea level fluctuations (FRUMKIN, 1996b). 2. General morphological features Fig. 1 is a Zipf Plot showing caves with an active traversible channel longer than 10 m (cf FORD & WILLIAMS, 1989 p.245). Each cave is represented by its active channel length (as opposed to total surveyed length which includes inactive passages), and by its rank in the channel length list. Caves exceeding 50 m in length are ordered along a straight line, indicating that the entire population of these longer caves was included in the study. The inclusion of all longer caves is also indicated by air-photo field study: large caves associated with large catchments could not have been overlooked. However, caves shorter than 50 m deviate from the straight line (Fig. 1), suggesting that some shorter caves may have been overlooked and not included in the study. A single active salt cave stream passage is referred to as a conduit. Examining the entire length of a conduit from sink point to outlet, it consists of (some or all of) the following five morphological elements (Fig. 2): (A) stream sink; (B) sub horizontal passage in caprock; (C) vertical shaft; (D) sub-horizontal rock salt passage; (E) open outlet. 100 0 0 0 0~ ~o !' -.;, 10 10 Figure I: Zipf plot of Mount Sedam caves ~ ~ ,;._ :o; .;: ,. 100 .. 0 0 ,ooo 10000 TOTAL LENGTH OF CAVE CHANNELS (m) Symposium 7 : Physical Speleology 139

PAGE 153

11111 10111 .. .. .. .. .. E .. .. .. .. .. 10011 .. .. .. .. ~.E .. .. .. C .. .. .. .. .. D ... ... ... o. .. .. .. A 10110 .. .. .. .. .. .. ???11 .. r::= .. .. o .. .. __ ,._ .. 2;;;;;;._E .. .. ...... .. .. .. .. .. lllOO A ~"' "'---::"'"' V 11000 'r l"\f' .:..,-.;..i ~"-'"-'D--"-'"-'"'" ,;~~~"""" ~~,-....,,~~........., .""'"'"""'"""'""' ~Orn .. .. .. .. D .. .. .. .. .. .... .. .. .. .. .. .. Figure 2: Examples of different types of cave conduit profiles in Mount Sedam. Black triangles indicate rock salt. Curly lines indicate caprock. The 5 digit numbers indicate the existence of the following features: (A) stream sink; (B) sub-horizontal passage in caprock; (C) vertical shaft; (D) sub-horizontal rock salt passage; (E) open outlet. A binary system was adopted in order to classify the conduits according to the existence of the elements. Five attributes were assigned to each conduit, representing the five elements. Each attribute may receive one of the following values: 1 =the morphological element exists; O=the morphological element is missing; ?=no evidence indicating if the element exists or not; #=an attribute used only for discussion purposes in this text, for eliminating irrelevant elements (either existing or non-existing). Examples of profiles of each conduit type are shown in Fig. 2. The distribution of the morphological types is given in Fig. 3. ..J w z z 0 w > < 0 u.. 0 w a.. >ti .:: wii 0.:: t..i c,) ,c t; F 5 I .. ffi~ CO~ >CO 0
PAGE 154

Most studied conduits (72 % ) include a sub-horizontal rock salt passage (###1#). The larger caves in Mount Sedom are included in this group. About a half (48%) of the studied conduits have an open outlet (####1). One third (34%) of the studied conduits consist of all morphological elements except a caprock passage (10111). One quarter (25%) of the studied caves have no salt passage and no open outlet (###00). These usually drain a small subaerial catchment, and their underground length is often limited to less than a few m. Some 28% of the conduits have developed a salt passage, but no open outlet was found (1#110). These caves, often located in the central portions of the mountain, are accessed through their sinks, and are referred to as 'inlet' caves. Inlet caves appear to terminate several m up to tens of m above the apparent water table (FRUMKIN, 1994b). Flood waters flowing into these caves partly evaporate and partly infiltrate into their alluviated floors (FRUMKIN 1994a). 3 Hor t on Laws in sa l t caves Some conduits join underground to form branchwork caves with several tributaries. Each tributary drains a subaerial catchment through a single active stream sink. The underground drainage system is organized in a similar way to surface drainage systems, so the HORTON (1945) stream order system may be used for the underground network defined in the following way : a first order conduit begins at a sink where a surface stream is captured to the subsurface; two first order conduits join underground to form a second order conduit; two second order conduits join underground to form a third order conduit. A cave order equals the highest order of its conduits. Fifty eight first order caves, eight second order caves, and one third order cave were studied. For each cave order, the following features are shown in Fig. 4: (A) number of caves; (B) mean number of tributaries per cave; (C) mean conduit length; (D) mean catchment area. All features seem to increase exponentially with cave order (straight lines on the logarithmic scale of Fig. 4) This trend is comparable with the HORTON (1945) laws for subaerial streams in spite of structural and lithologic controls which are more dominant in caves than in subaerial streams (FORD et al. 1988) Karst regions with a larger number of caves may yield a better evaluation of the applicability of Horton laws The definition of stream order is more complicated in autogenic karst with many small tributaries. Autogenic cave streams may gain further recharge along their underground flow course while the Sedom cave streams have an allogenic nature where (almost all) discharge enters each conduit through a single input. Autogenic caves appear more likely to obey Horton Laws which were originally proposed for surface drainage with spatial contribution of precipitation The Horton Laws certainly do not apply to maze or distributary cave systems where different methods should be sought. w > 0 )> z z (") (") )> 0 (") z 1000 J: 0 ;:: C --D CATCHMENT AREA --m :::; z .... r m )> z :Il G) m :i! )> x B NUMBER OF TRIBUTARIES ~100 10 3 CAVE ORDER Symi:x,s i um 7 : Phys i cal Speleo l ogy 141

PAGE 155

References FORD, D. C., PALMER, A. N. & WHITE W. B., 1988. Landform development; Karst, in: (Back, W., Rosenshein J. S. and Seaber, P.R., eds.), The geology of North America. Geological Society of America, 401-412. FORD, D C. & WILLIAMS, P. W., 1989. Karst geomorphology and hydrology Unwin Hyman, London, 601 p. FRUMKIN, A 1994a. Hydrology and denudation rates of halite karst. Journal of Hydrology, 162: 171-189. FRUMKIN A., 1994b. Morphology and development of salt caves. Bulletin of the National Speleological Societ y, 56 : 82-95. FRUMKIN, A., 1996a. Determining the exposure age of a karst landscape. Quaternary Research, 46: 99-106. FRUMKIN, A. 1996b. Uplift rate relative to base level of a salt diapir (Dead Sea, Israel), as indicated by cave levels in: (Alsop, I., Blundell, D and Davison, eds.), Salt Tectonics, Special Publication no. 100. Geological Society, London, pp. 41-47. FRUMKIN, A. & FORD, D. C. 1995. Rapid entrenchment of stream profiles in the salt caves of Mount Sedom, Israel. Eanh Swface Processes and Landforms, 20: 139-152 FRUMKIN, A MAGARITZ, M., CARMI, I. & ZAK, I., 1991. The Holocene climatic record of the salt caves of Mount Sedom, Israel. The Holocene, l (3): 191 200. HORTON, R. E., 1945. Erosional development of streams and their drainage basins. B ulletin of the Geological Society of America, 56 : 275-370. ZAK, I., 1967. The geology of Mount Sedom. PhD (in Hebrew Eng l ish abstract), The Hebrew University Jerusalem, 207p. 142 Proceed i ngs of the 1 2 th Internat i onal Congress of Speleology 1997 Sw i tzer l and Volu m e 1

PAGE 156

A karstic system in a semiallochtonous gypsum unit : the Legnanone caves (Marecchia valley I taly ): preliminary geological and hydrogeological features Abst r act Daniele Fari n a, Galleri n i Gi u liano Geoinfo, Via Marco Polo 9, 61100 Pesaro, Italy The Rio Strazzano basin, developed on gypsum outcrops of upper Miocenic age, shows interesting geomorphological and hydrogeological features, due to the presence of a small yet well developed karstic system, the Legnanone caves. The caves basically represent the underground path of the Strazzano creek, which sinks below surface in the mid -lower part of the valley, and gushes out again near the confluence with the Marecchia river. Geological, geophysical, hydrometric and hydrochemical data have been collected, in order to characterize the basic hydrogeology of the karstic system and to suggest some hypothesis on its evolution Geological structural setting and geophysical i nformations Rio Strazzano is a small tributary stream of the Marecchia river (Marche Romagna regions). Its basin is characterized by a quite large outcrop of detritic" balatino" gypsum of upper Miocenic age ("Gessi" formation), underlain by a clayey formation (Casa i Gessi fin.), with lateral etheropical lenses of conglomerate, belonging to the Acquaviva formation. The latter was origined by the erosion of a carbonatic platform (S Marino / M. Fumaiolo fms., lower Miocene) and of the eocenic marly calcareous Flysch (M. Morello fm.). The mid upper Miocenic sequence was formed on the large Val Marecchia allochtonous sheet represented by a highly tectonized melange-like, mostly clayey unit (Argille Scagliose s.l. / Sillano fm.) of Cretaceous Eocenic age, including the former S. Marino and M.Morello fonnations. The sheet moved eastward starting from middle Miocene to lower Pliocene, forming a series of "piggy back" sedimentary basins where the particular Rio Strazzano sequence was deposited. After the sedimentation of the Pliocenic marine clays (Argille Plioceniche fin.) topping the Gessi fm., a new major tectonic phase caused the complete imbrication of the former structures and the faulting of the relatively more rigid gypsum layer (see cross-section A-A taken from CONTI, 1989). The microcristalline gypsum, derived from the erosion of primary selenitic cristals, forms massive strata (5 I O m thick) with local marlybituminous interca l ations (BERTOLAMI, M. ROSSI, A., 199 I). The total thickness of the evaporites exceeds 60 111, gradually decreasing toward the front of the thrusted unit. This trend detected by means of Vertical Electrical Soundings, could also outline the limits of the gypsum outcrop to the N.E. part of the Rio Strazzano watershed ; besides, the geoelectrical survey showed the presence of more resistive layers inside the conductive substratum, interpreted as conglomerate lenses within the clayey Casa i Gessi formation. Geomorphological and hydrogeological features The Rio Strazzano basin shows interesting geomorphological and hydrogeological features that can be summarized as follows : the typical asymmetrical section of the Rio Strazzano valley; the presence of sparse water springs gushing out the Acquaviva conglomerates and / or S. Marino M.Fumaiolo fractured biocalcarenites in the upper part of the valley (S. lgne Passo delta Biforca sites), drained into the Rio Strazzano; the karstic sink situated in the mid lower part of the va l ley, where the Rio Strazzano is swal l owed into the Legnanone caves; the "dead" course of the Rio Strazzano, situated downhill the sink which remains dry most of the year; a widespread system of dolinas (absorbing pits) where a large part of the precipitations is absorbed underground ; the main entrance of the Legnanone caves, representing the resurgence of the Rio Strazzano, situated approximately 300 111 uphill the confluence into the Marecchia river. The sink of the Rio Strazzano is a narrow tunnel like a grotto with a partially collapsed ceiling, locally comunicating with the outside The sink occurs near a fractured and faulted zone in the mid lower part of the valley The "dead" course o f the Rio Strazzano is characterized by a quite narrow path loca l ly showing canyon l ike features. Its bottom shows several cracks and holes expecially on the left hydrographic side; a major sump shows seasonally fluctuating water levels The dolinas are often aligned along major tectonic discontinuities and are nonnally covered with silty soil af various thicknes s {SO C IETA GEOLOGICA ITALIANA, 1992). The Legnanone caves, which a r e being explored by the S Marino Speleological Group, show at least two different levels with the lowest one still active. The main path stretches approximately 600 m. and is characterized by frequent "pockets of debris (boulders, sand and silt) carried by the stream inside the cave and deposited along its banks (AA VY ., 1993). Some other finer sediments seem to have penetrated from the ceiling and deposited in a cascade like fashion at the Sympos i um 7 : Phys i cal Speleology 143

PAGE 157

end of the upper level grotto. Part of the ceiling and walls are unstable and subject to rock falls. The lower level main grotto is situated I O to 20 m above the base of the gypsum formation, underlain by the "Casa i Gessi" clays. Hydrometric and hydrochemical data During this preliminary study water discharge has been measured, both in the rainy and in the dry season, at both ends of the grotto. During the winter season there is no significant difference of discharge between up-hill and down-hill the cave. The yeld varies from approximately 20 I/sec. to I 00 I/sec or more, right after heavy rains. Durin g the summer base flow the creek is almost dry (0.2 0.4 I / sec. up-hill, 0.1 or les s at the main lower entrance). The water chemistry is of the calcium su lphate type, with mid salinity values (0 8 1.3 gr / liter) with an increasing trend during the rainy season. Discussion The Legnanone caves show an intimate relationship with the surface hydrology of the Strazzano creek, due to the basin's evolution during the recent Quaternary. The interpretation of all preliminary data suggests that a deep karstification of the gypsum outcrop developed as the base-level of the Marecchia Strazzano hydrographic system kept lowering as a result of the recent uplift of the region (see the suspended alluvial terraces on the Marecchia river's bank). The erosion of the clayey Pliocenic cover, proceeding from N-E to S-W, exposed increasing surfaces of the gypsum layer to the meteoric dissolution, causing the formation of the surficial karst forms and, particularly, the demolition of the cavities previously formed by the course of the ancient Strazzano creek: the canyon like features of the "dead" valley in fact, suggest that it possibly derives from the collapse of the former grottoes as linear karstification kept migrating toward the S-W.The wide fluctuation of yeld, the moderate salinity of the waters and the presence of the debris fillings inside the grotto suggest that most of the karstification process nowadays is still originated by the action of running waters. The completion of the study should verify the formation of deeper lateral caves within the gypsum S-Wdipping monocline, topped by the PI iocenic clays. Such study should extend the use of the geoelectrical survey to model the buried geometry of the gypsum layer and should carry on extensive differential flow measurements to detect possible losses of waters seeping through major faults to a lower discharge site. References AA. VY. 1993. Quademi del Circondario di Rimini n.3 La riserva naturale di Onferno a cura di L.Casini: 9 16 BERTOLAMI, M.; Rossi, A. I 991. La petrografia della grotta di Onfemo e delle aree limitrofe "Naturalia Faventina" Boll. Mus. Civ. Sc. Nat. Faenza CONTI, S. I 989. Geologia dell'Appennino marchigiano romagnolo tra le valli del Savio e del Foglia. Boll. Soc. Geol. It. Vol.108: 435 490. SOCIETA' GEOLOGICA ITALIANA. 1992. Guide Geologiche Regionali n.4 -Appennino Tosco Emiliano: 48 51, 70 71, 89 90 295 298 Aknowledgement Thanks to Stefano De Angelis for the CAD drawings. 144 Proceedings of the 12 th International Congress of Speleology 1997, Switzerland Volume 1

PAGE 158

100 50 10 5 0 5 0.1 \ \ \ \ I"-"' ,\ I \ I \ I I \ I \. Co Mg No .,.K Cl S04 HC03 w'oter So.riple-Moiri Gl"Ot to E.C =1181/'S/cri Strozzono creek -'w'otershed divide Gro"ti:o Deod Volley Dolino 0 Spr i ng Debris Dei:. ~; ;! .:, Alluvial Te-rroce-s All. [=:J Cloy ond Sands AP Gypsur'I G ES] Cloys ACG l?F#ad Congl Or'lero tes ACQ cr5 Biocolcorenites CSM Argille-sabbie Pl i oceniche tl"l Gessi tl"l Argille di Casa I Gessi tl"l, Acquaviva tl"l San Marino / M.Fuf"1a i olo rf"1 A Rio Rio Strozzono A' Generol cross-sect i on scale 1 : 50000 B Local cross-section scale 1 1 10000 Sympos i um 7 : Phys i cal Speleo l ogy 145

PAGE 159

Morphologic studies of bell hole development on Cayman Brae Res u m e Rozemarijn F.A. Tar hu le-Li p s and Derek C F ord McMaster University, Hamilton Ontario Canada II y a des bell holes des formes cylindriques dans Jes voutes avec un axe vertical dans beaucoup des grottes a Cayman Brae (Cayman Islands BWI) II semble qu'ils ont forme independamment des characterisiques structurales de la roche. On les a trouves dans les regions tropicales hurnides Les conditions de formation sont inconnues mais ii semble qu'ils sont des formes de dissolutions developpees dans un environnement vadose Leur morphology donne !' ~ impression qu'ils s'elargent initiallement jusqu'un diametre critique et qu'apres ils continuent en profondeur. Tous les bell holes sont plus ou moins coniques ce qui est probablement le resultat de I' approfondissement. Introduction Bell holes are cylindrical indentations in cave ceilings with long axes that are vertical. Cross sections are close to perfect circles in form Their depth is in general less than 2 m and their radius less than 0 5 m They develop with complete disregard for any structural features in the bedrock. Sometimes two or more bell holes merge to become one Apparently, they are observed only in the humid tropics B e ll h ole form atio n The origin of bell holes is poorly understood Existing hypotheses invoke a range of mechanical-chemical-biological processes to account for their development but none appear satisfactory In particular the strictly vertical orientation of the long axis remains unexplained by all hypotheses, as does the observation that they may form in bedrock without fractures or apparent feedwater inputs at the top This suggests that the process acts from inside the caves B e ll hol e di s tribu ti on On Cayman Brae, Caymen Islands bell holes are found only in the entrance zones of coastal mixing zone dissolution caves. They occur at random in some caves but clustered in others. All bell holes extend vertically into the ceiling regardless of its angle and have a more or l ess tapered form. All show smooth wa ll s without any vertical grooves or flutings. The diameter and depth of 55 bell holes from five different caves were measured Bell hole profiles were drawn from these measurements and the volumes were calculated as series of stacked cylinders Results and discussion Their location in the entrance zones (in general the largest and highest rooms of the caves) suggests that cave configuration plays an important role in the bell hole formation There is, however, no relationship between the distance into the cliff and the depth, opening radius or volume of the individual measured bell holes. It can be concluded that the process(es) responsible for their formation operate uniformly over the zones where they are found In the case where two or more bell holes have merged, it is sti ll possible to identify the individuals clearly Enlargement of the bell holes therefore seems to continue in all previous directions after intersection has occurred. This suggests that there was either a water-filled passage (phreatic environment), or a thin water film on the wall (vadose environment) as may be expected in the case of condensation corrosion. Water in the phreatic zone can either be slow or fast flowing. On Cayman Brae slow flowing water is most likely, in whic h case cy l i n dr i cal in d entations should also exist in walls and floor but this has not been observed. Bell hole formation in a vadose environme n t is s u ggested. The slightly tapered form indicates that the water does not lose its d issolving capacity fully when flowing down the wall or that condensatio n (and thus d issolution) takes place evenly on all the walls. Since a bell hole is growing upwards, dissolution at the base of it has bee n going on for a longer period of time an d thus more d issolutio n has taken p l ace there than at the top This causes the tapered form to develop The measured degrees of taper (deviation from strictly vertical walls) range from 2 5 to 30 5 with an average of I 0 The mean radii depth and vol u me of the bell holes in one particular cave were significantly smaller (at 95% probability level Student t-test) than those of bell holes in the other four sample caves, forming a class of their own on the lower side of the scale The volume of these bell holes is most strongly correlated with t h e basal ra d ius, whereas volume is more strongly correlate d with depth in the other four caves. It is possible that these smaller bell holes represent the formative stage of development. If this is the case, then the bell holes first developed in width plus d epth until a critical radius was attained after which they continued to enlarge primarily in depth For t h e bell holes on Cayman Brae the critical radius appears to be 0.35 m. 146 Proceedings of the 12 '" Internat i onal Congress of Speleology 1997 Switzer l and Vol u me 1

PAGE 160

Some types of Speleogenesis and Speleohydrogeology in Dinaric Karst area (Croatia Europe) Prof. dr. Mladen GARASIC ,.z..1 1 Hrvatsko speleolosko drustvo (HSD), Nova Yes 66, HRV-10000 Zagreb, CROATIA 1 Drustvo za istrazivanja i snimanja krskih fenomena {DISKF), Alfireviceva 13, HRV-10000 Zagreb, CROATIA 3 Institut gradevinarstva Hrvatske (IGH), Rakusina I, HRV-10000 Zagreb, CROATIA Till the end the year 1996 in Croatian karst there are more than 7750 caves and pits known. They are mostly located in Dinaric karst located in the western and the southern parts of Croatia. Dinaric karst is one of the best-known typical karst area in the world. Different types ofkarstification processes have been observed in speleological features (caves, pits, estavelas, vruljas, ponors, etc.) of Croatia in many localities It is most marked in the Mesozoic rocks of the carbonate facies. These rocks are mostly limestones, dolomitic limestones lime dolomite of the Triassic, Jurassic and Cretaceous with all their varieties, from both lithologic and stratigraphic aspects. Among over 7750 speleological features that have been explored to date, about 78% are vertical structures (pits), 21% are horizontal structures (caves), while the remaining I% are combined speleological structures. The main types of karstification processes are: gravitational karstification. inverse karstification, complex karstification, secondary and tertiary sedimentation, etc. Depending on the intensity of these processes, many special type of speleogenesis and speleohydrogeology are observed there. The results obtained by trial borings in the Adriatic karst of Croatia show that karstification processes occur even at depths of over 4 km It should be noted that the thickness of the Mesozoic rocks found in this area amounts to 8 kilometres. In Dinaric karst area there are two pits deeper than 1000 meters; the most vertical cave in the world Lukina jama, 1392 m in depth, and Slovacka jama, IOI 7 m in depth Gravitational karsti fication is present in the e pits located only I 0 kilometres from the Adriatic Sea. Therefore, it is to be expected that speleogenesis, as a part of the process of karstification is also very intensive in this area. According to speleogenesis, features are classified into several groups (abrasial, tectonic, erosional, corrosional, polygenetic etc.). The process ofkarstification can be best observed in the subsurface where speleogenesis is one of its most inten ive mode of activity. The observed regularities in speleomorphology (of the whole speleological features or of only some of their elements) undoubtedly confirm hydrogeological elements that were present (paleohydrogeology) only at a certain time of peleogenesis or are also active today. Modem hydrogeology of karst is impossible without a good knowledge of the type of speleogenesis and the degree of karstification of the areas explored. 1. Introduction The Croatian karst is surely one of the best known "locus typicus" areas of the typical karst in the world. Subsurface forms of karst are represented, among other things, by more than 7750 speleological structures having different morphological and morphogenetic properties and hydrogeological functions (GARASIC, 1986). Therefore, it is important to become aware of comprehend the laws of speleogenesis in this very area where karsrification intensity is very high as they can be applied in a large number of other kar t areas in the world Out of a great number of s peleological structures that are known in Croatia, for special re sea rch on speleogenesis in the process of karstification, 46 were selected. They are formed in rocks which differ in their litostratigraphic, tectonic and hydrogeological properties. In addition, the speleological structures selected are of different stages of speleogenesis (the initial, main and fossil stages) differing in their morphological types (s imple, branched, level. knee-like and systems) and in their hydrogeologicaJ functions (springs, ponor s, estavelas, percolating structures, vruljas, e tc .). The classification of all the speleological structures was carried out according to the adopted categorisation (GARASIC, 1991 ). The mea uremcnts of ab olutc neotectonic displacements in speleological structures were made over the period of berween 1980 and 1996 Geomechanical properties of the rocks in which s peleological structures research e d into had been formed were tested in the laboratory. All the gcospeleological researches in s peleological structures were carried out in co-operation with the members of the Croatian Speleological Association (CSA) and the Society for Research Surveying a nd Photographing of the Karst Phenomena (DISKF) of Zagreb Thi s paper deals only with speleology within the context of karst hydrogeology and the process of karstification although the re su lts of research lasting for several years are more comprehensive and numerous. 2 A brief survey of the past research Almost every description of a cave or a pit include s the description of a part of speleogenesis. However the aim of most authors or papers was not to research into peleogenesis in particular but it was only mentioned incidentally. Therefore, only important papers that served as a basis for this paper are mentioned Tectogenesis of Panjkova and Muskinja cave in Kordun were written about by GARASIC ( 1984a) and the speleogenesis of structures in !stria was described by BOZICEVlC ( 1985) The morphogenetic approach and classification of speleological structures in the Croatian karst with regard to morphology and hydrogeology were dealt with by GARASIC (1986, 1991 ,1993). Based on a great number of speleologic structures in the We s tern Carpathians, BELLA (1994) showed genetic types of caves that could be partially applied in the Croatian karst Speleogenesis of some big and small speleologic structures of Lika, Gorski kotar Hrvatsko Zagorje and [stria was treated by MALEZ ( 1966) while the structures of the Krka National Park was researched into by Lukic. The neotectonic approach was given and the mea uring of absolute displacements was made by GA RAS IC ( 1981 1984b 1989). Special researches on a role of the karstification proces s in speleogenesis were done in 58 caverns on the route of the Karlovac Rijeka highway viz. on the sections from Ostrovica to Kupjak and in tunnels Hrasten Tuhobic Vrata, Sljeme, Sopac and Vrsek in Gorski kotar Symposium 7: Phys ical Speleology 147

PAGE 161

3 General forms of speleogenesis in Croatian karst In speleogenesis there are three basic and a few special stages of the formation of speleological structures. The basic stages are the following: I. Early (initial) stage (Phase I) ; 2. Main stage (Phase II); and 3 Late (fossil) stage (Phase III). In the Croatian karst all three stages of development have been observed. The ratio of the main stage and the late stage is about 3: I (according to the number of structures). The early (initial) stage of development of the speleological structures is present in every place where karstification is underway, but due to the narrowness (small width) of fissures and initial joints, the direct observation of this stage of the formation of speleological structures is not possible. Based on the results of researches and analyses of a large number of caves and pits, it can be co n c l uded that in Croatia most of them have not gone through all three development stages yet. This can be explained primarily by geological, hydrogeological and climatic factors; for example the structures located in high mountain-ranges have not by the process of gravitational karstification yet "reached" the less permeable or impermeable rock base A very fine example is found in several very deep pits in Biokovo (the area surrounding Lokvica and Ladena) that do not retain water in lakes or siphons even at depths bellow 500 meters, but water flows away deeper down through narrow fissures or sumps (Skolska pit, Vilimova pit, Pit on Kamenita Vrata, etc.). A very special example is a deep pit Lukina jama in Mt. Velebit in which a permanent ground water flow appea r s as deep as under I 200 meters (BOZIC, 1995). The depth of this pit is I 392 meters, which is the proof of high intensity and deep karstification. Neotectonic and hydrogeological researches into this pit are underway and it is expected that important results will be obtained. 3.1 Tectos p e l eoge n ese A predisposition to t ec ton ic disp l acements of the areas in which speleological structures develop is crucial for speleogenesis. (HEAAK, 1984; GARASIC, 1984,a). For example, it has been observed that karstification in horizontal or slightly inclined layers progresses more slowly than in steep or vertical layers. Such a process can be monitored in Volarica pit (vertical and very steep layers) and in Kojina pit (horizontal layers) which is a few meters away from Volarica pit. It is a question of Senonian limestones K., ', of the same thickness and composition. In Volarica Pit, karstification is quick (more intensive) and consequently speleogenesis has reached the depth of I 18 meters, and cave channels (chambers) are bigger; in Kojina jama, after the vertical entrance to the structure follows the horizontal extension of the structure of about 60 meters in depth and about 750 meters in width. In places of Kojina jama in which vertical layers are present again karstification is more intensive and reaches deepe r (105 meters). Volarica pit and Kojina pit are located in the region of Kordun, near a small town of Furjan, on Masvina Hill near Rakovica Even more interesting conclusions can be made from the results of systematic research into all cave rn s in the tunnels on the route of the Karlovac Rijeka highway, where intensities of karstification, and morphological and hydro l ogical data have been monitored and compared regularly L n some tunnels tectonics played a crucial role in speleogenesis. Similar results were obtained in localities in Mt. Dinara and Mt. Velebit (Hajducki kukovi, Lomska duliba) The largest number of speleological features in the main stage of development (Phase II) was registered in the near vicinity of karst poljes and uvalas These structures have even today, a role of periodic or continuous springs, ponors or estavelas, and the presence of water in them has a very important role in th i s stage of speleogenesis. In this paper only some less known caves along margins of the karst fields are given (GARASIC & CVIJANOVIC, 1986) These are the follow i ng: Vranova cave (Pisacusa, Lika), Bunjakova cave (Debelo brdo, Korenica, Lika), Vrsina cave near Ondic (Lika), Duman cave (Canak, Lika), Cave near Krcevine (Smiljan, Lika), Ostrovica cave (Licki Osik, Lika), Cave in Plesa (Dreznica, Gorski kotar), Cmacka cave (Jezerane, Lika), etc. In the Croatian karst there are less than 3% of speleological structures that were formed by tectogenesis, i e. without a significant influence of ground water. They are located in high mountains, such as Velebit and Dinara (Maia cave near Tulove grede on Mt. Velebit, Jama near Plazonica Unista on Mt. Dinara). They are present on the top of anticlines where water had, in later stage, a minor role in speleogenesis Examples of such structures are narrow tectonic pits, such as for example Cirova pit on Zavizan (Velebit), which is as deep as 69 meters, but very narrow. These are mostly structures developed in solid and massive dolomites 3.2 H y dro s p e l eoge n ese Hydrospeleogenesis is always a complex process in which water has a decisive role. In Croatia there are as much as 4% of structures formed by an abra s i ve effect of water along coasts and river and lake banks The l argest number of abrasion structures are found in b reccia-like ma t erials (about 71 %), but they also occur quite often in limestones (about 25%) The main characteristic of the structures developed by abrasion process that played the predominan t role is as fo ll ows as a rule they are always horizontal or very slightly inclined; they are small in size; their rocks are smooth; etc. They are also called abries or semi-caves There are recent abrasion structures and paleo abrasion speleologicical structures. Some speleo l ogicical features a l ong the Adriatic Sea, today submerged by the sea, were developed by abrasion between 100000 and 300000 years ago (MALEZ, 1966; MALEZ & BOZICEVIC, 1965), which means that they are relatively younger of more recent origin. There is no doubt that they had developed in our area even before Q u artenary, but they collapsed (Phase III of speleogenesis) because it was the question of the rocks of relatively l ow strengt h and carrying capacity. When, in addition to that chemical reactions occu rr ed due to the effect of sea water, the process of collapsing was still more intensive However, there are some examples of abrasion caves that are much older (Palaeogene, Neogene), e.g. near Nerezisce on the island of Brae. Here are some examples of the abrasion caves: Crvena cave on the Island of Murter (Jezera), Cave at Ivan Dolac (the island of Hvar), Strasna pee cave (Savar, the island of Dugi otok), Raca spilja (the island of Lastovo), Modra spilja (the is l and of Bisevo), Przinova polaca cave (the island ofSipun), Popovska cave (Baska, the island of Krk) etc. E ro s ion and c orro s ion effects (mechanical speleogenesis and chemical speleogenesis) ~~cur together in the forming of speleological structures (BOGLI, 1964). In the Croatian karst about 93% of the structures developed, to a varying extent, by the influence of erosion and corrosion has been registered so far In some structures, cave sca ll ops have been also noticed which indicates the inte n sive process of subsurface karstification. The examples are: Pit under Debe l a glava (Blagaj, Kordun), Bozica pit (Mawina, Kordun), Ponor on Grgin brijeg (Jadovno, Velebit), Stanina pit (Moravice, Gorski kotar), Podublog pit (Kmica Istria), etc. In some caves and pits some other forms of the erosion and corrosion effects of water in the cave channels have been developed as well The examples of erosion pots in Vetemica (Medvednica, Zagreb), Du l in ponor cave (Ogulin, Korski kotar), Medvedica cave (Ogulin, Gorski kotar), Jopiceva cave (Brebomica, Kordun) or the po l ished rocks in Novokracina cave (Rupa, !stria), Ponor at Bunjevac (Velebit), Cave near Pavlinovici (Zupa, Vrgorac) clearly demonstrate the intensity of chemical or mechanical effects A good example of an abrasio n structure is Cavle cave in the canyon of the Zrmanja river Limestones of different textures and ages react differently on erosion and corrosion on the ground surface ; thus it can be presumed that the situation is similar in the subsurface. ln addition to all the a b ove mentioned factors in the Phase II of speleology a neotectonic activity is very important too; namely, it has been noticed that tectonics is still active in some speleological features even today Measurements show that rock movements in the speleological structures located in neotectonic zones can be significant, for example in a big cave system of Muskinja and 148 Proceedings of the 12 th International Congress of Speleology, 1997, Switzerland Volume 1

PAGE 162

Fig I. Relictes in lmerse karstijicatio11 (photo: dr .. M. Gara.sic) Pankova cave in Kordun. or in the pit Ledenica in Lomska duliba (Vclebit). Morphogenesis of speleological features in the Croatian karst is closely connected with lithostratigraphic and tectonic factors of surroundings in which they are fonned. and with abrasion, erosion a nd corrosion effects of water as well. The neotectonic activity has even more favourable effect on karstification of the ground, and it also accelerates this process. Morphology, length, depth orientation. inclination and other characteristics of a cave or a pit depend on the factors of speleogenesis. In Cretaceous limestones of the Central and Inner kar st zones, according to hydrogeologic karst zoning (HERAK, BAHUN & MAGDALENIC, 1969), speleological feature of the so-called polygenetic type were developed i.e. the ones developed by the primary effects of tectonics, abrasion, erosion and corrosion. The examples are: Mijatova pit at Matesko selo in Kordun (GARASIC, 1976 ), Upper Cerovacka cave at Gracae in Lika Go podska cave near the source of the Cetina river in Dalmatia. In the Outer (Adriatic) karst belt along the coast, abrasion together with tectonics is crucial for speleogenesis, e g. Borina cave (the island of Prvic) Pit near Jajce (the island of Zirje), Banova Ljut cave(Dubrovnik), Vela cave (Vela Luka, the island of Korcula) The Adriatic islands and the mountains by the Adriatic are most effected by erosion and corrosion, along with intensive tectonics Jn tenns of morphogenesis Jurassic limestones and lime dolomites are carriers of speleogenesis accompanied by intensive tectonics and by the s trong influence of erosion and corrosion. The examples are: Livnjak pit (Jasenak. Gorski kotar), Pit on Kozica (Zmici, Gorski kotar), Janusinka pit (Dreznica. Gorski kotar), Briscinka (Biokovo), Radekina pit (Bijele stijene, Gorski kotar), Samara pit (Samarske stijene, Gorski kotar). It is in Jurassic limestones that the most intere ting fonns of corrosion in ubsurface have been observed. Karstification have been monitored in tunnels (Gorski kotar), in fissures and caverns of all possible sizes. Measurements of the intensity of karstification in the rocks of tunnels are proportional also to the strength of speleogenesis in the same lithostratigraphic surroundings. Paleogen sediments were, in tenns of speleogenesis, exposed to the influence of tectonics, abrasion, erosion and corrosion to an equal degree. The examples of big subsurface hollow s in the area of Cmopac (Gracac, Lika) in which the intensity of karstification is very high or of Golubinka Pit (Posedarje. Ravni kotari) in which a fossil phase of speleogenesis (Phase 111) occurs in a part of it are very interesting. The monitoring and measuring of the neotectonic activities in caves and pits helped a following conclusion to be made in areas of the relative ascending of block s (e.g. Pit on Jatarina orth Velebit) cave channels get relatively deeper (karstification is more intensive), while in the areas of neotectonic descending (e.g. Cmo vrelo near Kordunski Ljeskovac in Kordun) channels are constantly filled with water, and the effect of corrosion becomes dominant. 4. Karstification processes Karstification processes in the speleological features of the Croatian karst progress in two basic directions The first, and more frequent, gravitationally towards the deeper layers, and the second, relatively less frequent, inversely i.e. towards the surface. The examples of inverse karstification have been found in !stria (MAUCCI, 1952; BOZICEVTC, 1985), but also in the areas of Kordun, Lika, and Gorski kotar. 4.1 Gravitatio n al karstificatio n Gravitational or surface karstification is always influenced by gravitational water, i .e. precipitation. It is very easily observed. It is very well developed in Jurassic and Cretaceous carbonate sediments. Tectonic fracturing is favourable for its development. 4.2 I n ve r se ka r stificati on Inverse karstification is observable only in the subsurface and results from hydrogeologicical conditions (changes in the level of ground water, its constant presence in the subsurface during the Phase I and Phase II of speleogenesis. etc.) lithostratigraphic factors (it is frequent in soluble strata, in P rom in a a nd Je l ar lime breccias sediments, in fine grains breccias etc.), and neotectonic ascending of blocks ln the zones of neotectonic ascending it is unfailingly present, and its intensity is proportional to the intensity of ascending. In the zones of air cushions. on the ceilings of caves, a strong effect of inverse karstification has been noticed (Figure I); for example Rudnica cave (Kamenica, Kordun), Vrelo cave (Fuzine, Gorski kotar). The intensity of the inverse karstification compared to the gravitational karstification is less dependent on tectonic fracture, and more on special hydrogeological conditions (GARASIC, 1993). 4.3 Cco mpl ex ka r stific a ti on In the Croatian karst yet another karstification process, the so called complex karstification (GARASIC, 1995) has been found too. This is the process which also requires, in addition to the conditions mentioned above, special conditions in speleogenesis that will influence the morphology of speleologicical features. These conditions are for example, a smaller depth to the impermeable bedding in karst (the area of Kordun. Gorski kotar, etc.), the occurrence of lenses of various rocks within carbonate complexes such as cherts within limestones or flysh materials between a carbonate series, etc. (the regions of Istria, Mt. Velebit Mt. Mosor, etc.). For these reasons there are also fonned speleological features where karstification has been developed in various directions (Figure 2), for example close to flysh or cherts or close to watertight bedding. In these cases such speleological structures suddenly change the morphological type. inclination and shape of cave channels (pits become caves, and caves become pits; narrow channels become wide, and wide channels become completely narrow; etc.). Fig.1.-Resu l ts of Complex karstijicatio 11 (photo: dr.M. Garasic) Sympos i um 7: Physical Speleology 1 49

PAGE 163

5. Classification of caves and pits according to speleogenesis Speleological features in Croatia can be classified in several group Basically there are si n s p e l eoge n etic and e p is p e l eoge n et ic features. The sinspeleogenetic features have been developed during the formation of matrices There are not very frequent. They are fonned by travertine barriers and falls on karst rivers and lakes (on Plitvice Lakes, b y the Krka river, etc.). The epispeleogenetic features were developed by the effects of various processes in rock karstification, such as abrasion, ero ion. corrosion, collapsing, tectonics. etc. Most often it is the question of p o l ys p e l eoge n etic features. Nevertheless, only some types of caves and pits (epispeleogenetic) differing by genesis will be mentioned in these paper. The c types are the following: c o rro s ion a l ; corr os i o n a l and co ll aps i ng; flu v i o karst i c; flu v i o k a r s ti c a nd co ll a p s in g; flu v i o karstic, corrosional a n d co ll a p s in g; t ecto n ic ( faul te d ), flu vi o ka r stic (fl u vial) a nd c o ll a p s in g; tecto n ic a nd corros i o n a l ; tecto ni c, corros i o n a l a nd co ll a p s in g; eros i o n a l ; eros i o n a l a nd corros i ona l ; erosio n al a nd fl u v i o ka r st i c; se dim e nt ic; abrasia l ; cryoge n ic (only a few structures), v u lca n oge ni c (only one structure in Croatia). structures formed by the influence of plants from the ground urface etc. A classification in phreatic structures and vadose structures. etc. has been also made 6. Conclusion Ba ed on a great number of the speleological structures explored in the Croatian karst (more than 7750), their documentation. speleological surveying, hydrogeological observations and neotectonic measurements of absolute displacements in them. conclusion about the modes and type of speleogenesis, which is a part of a complex process of karstification. ha ve been made. The influence of tectonics (initial joints, fissures, faults, folds, layers nappes etc.) together with litostratigraphic hydrogeologic and climatic conditions have developed speleogenesi in several type of karstification. Gravitational or surface karstification is most frequent (above 80%); next in frequency is inverse karsti fication (about 19 %); and complex karstification is identifiable (about I%) too Inverse and complex karstification can be studied into exclusively in the underground channel and chambers Special attention has been given to researches into the karstification process in 46 caves and pits, and caverns in tunnels (Gorski kotar) In term of speleogenesis, caves and pits in Croatia can be classified into two ba ic groups sinspeleogenetic and epispeleogenetic structures. Sinspeleogenetic structures were developed during the formation of matrix rocks, and epispeleogenetic structures were developed by subsequent processes of tectonics abrasion, erosion, corrosion, collap ing. edimentation, etc. Vulcanogenic structures have been cla sified separately as they do not belong to the karst area. It can be concluded that a great number of speleological structures in the Croatian karst has different types of speleogenesis, which resulted in a diversity in speleomorphology (simple, knee like, level, branched structures and ystems) in the p r esent stage of speleogene is (Phase I II or Il l) and in hydrogeological function of caves and pits (ponors sinkholes, spring estave l as. vruljas, percolating caves, and dry features) Further researches into speleogenesis will result in new knowledge of the most intensive today's geological process karstificatio n References BELLA, P .( 1994): Genetic types caves spaces the West Carpathians. Slovensky kras, vol.XXXII, pp. 3-20, Liptovski Mikulasi. BOG LI, A ( I 964 ): Die Kalkkorrosion das zentrale Problem der Verkarstung. Steir. Beitr. Hydrogeol., vol.64, pp. 75-90, Graz. BOZICEVIC, S. ( 1985): Morfogeneza speleoloskih pojava I stre i njihova zavisnost o geoloskim i hidrogeolo~kim uvjetima Disertacija. str. 11 66, Sveuci l iste u Zagrebu BOZIC, V. ( 1995): L'expedition speleologique "lukina jama 94". Regards, Bulletin de !'Union Beige de Speleologie. no 20, pp.2-6, Liege. GARASIC, M ( 1981 ): Neotectonics in Some of the Speleo l ogical Objects in Yugoslavia. Proceedings of 8 th International Congress of Speleology Bowling Green, str. 148-149, fot. 3, st. I, USA GARASIC, M. ( 1986): Hidrogeologija i morfogeneza speleoloskih objekata s vodom u kr,u Hrvat ke. Disertacija Zajednicki studij geo l ogije PM i RGN fakulteta Sveucili,ta u Zagrebu, str. 1-155, sl. 53, Zagreb. GARASIC M. & CVIJA OVIC, D. (1986): Speleological phenomena and seismic activity in Dinaric karst area in Yugoslavia. 9. Congresso International de Espeleologia. Communicacions Barcelona, vol. I, str. 94-97, fot. 2, st. 8, Barcelona. Espagna. GA RAS IC, M. ( J 989): New conception of the morphogenesis and hydrogeology of the speleological objects in karst area in Croatia (Yugoslavia) I 0. International Congress of Speleology, Procceedings, vol. I, str 234-236, I. 8, Budapest, Hungary. GARAS I C, M. ( 1991 ): Morphological and Hydrogeological Classification of Speleological structures (Caves and Pits) in the Croatian Karst area. Geoloski vjesnik, vol. 44, str. 289300, fot. 3, sl. 4, Zagreb GARASIC, M. (1993): The Karstification 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. HERAK M.( 1984) : Tektonski okvir speleogeneze. Deveti jug. speleo. kong., Karlovac, str. 111-130, Zagreb. HERAK, M., BAHUN, S. & MAGDALE IC A. ( 1969) : Pozitivni i negativni u tjecaju na razvoj krsa u Hrvatskoj. Krs Jugoslavije, vol. 6, str. 45-78, JAZU, Zagreb MALEZ, M & BOZICEVIC, S. ( 1965) : The Medvjedja pecina (Bear Cave) on Losinj Island a rare Case of Submerged Cave. Problems of the Speleological Research pp 21 1-216 Prague MA UCC I W (1952) : L' ipotesi dell "Erosione in versa. Bolletino dellea Societa Adriatico di Scienze aturali, vol. XL VI Trieste. 150 Proceedings of the 12 '" International Congress of Speleology, 1997, Sw itz erland Volume 1

PAGE 164

Origin and Development of Caves in the Devonian Massiv e Limestones of the Rheinisches Schiefergebirge (Ge r ma ny) by Stefan Niggemann Institut fur Geologie, Ruhr-Universitat Bochum, Universitatsstr. 150, D-44780 Bochum, Germany Abstract Several hundred caves are located in the massive, Middle Devonian limestones of the Sauerland, Eifel and Westerwald. Twenty seven of these caves reach 500 m in length, and two of them reach even 5000 m. The Grilnerbach Valley (Iserlohn, Sauerland) serves as a model of the progressive development of cave levels. The Rheinisches Schiefergebirge has been subject of continued elevation since the Pliocene, with valley entrenchments and formation of river terraces predominantly controlled by climatic factors. Interglacial phases of stagnation resulted in the development of horizontal cave tunnels extending over 2 km. The indirect correlation of cave levels with the river terraces enables determination of the minimum age of the caves and the estimation of their real age, which is believed to be some 400,000 to 500,000 years. Zusammenfassung Mitteldevonische Massen.kalke enthalten im Sauerland, der Eifel und dem Westerwald mehrere hundert Hohlen. Insgesamt sind 27 Grol3und zwei Riesenhohlen bekannt. Am Beispiel des Grilnerbachtales (Iserlohn, Sauerland) wird die schrittweise Entwicklung iibereinanderliegender Gangsysteme dargestellt. Im Zuge der seit dem Jungtertiar anhaltenden Hebung des Rheinischen Schiefergebirges kam es zur klimagesteuerten Eintiefung der Taler und Bildung von Flul3terrassen. Stagnationsphasen wiihrend der Warmzeiten fii.hrten zur Herausbildung von horizontalen Tunnelgangen, die sich 2 km lang iiber mehrere Hohlensysteme verfolgen !assen. Eine indirekte Korrelation von Hohlenniveaus und Flul3terrassen erlaubt eine Bestimmung des Mindestund Abschatzung des Bildungsalters der Hohlen auf etwa 400.000 bis 500.000 Jahre. 1 Introduction Devonian massive limestones of the Rheinisches Schiefergebirge contain some of Germany's biggest cave systems. Twenty-seven of these caves reach 500 m in length, and two of them reach even 5000 m. Striking of these massive limestones is WSW/ENE following the Schiefergebirges folding structure. Valleys of allogenic rivers cut up these small karstic areas. The cave systems with a maximum horizontal range of 400 m are characterized by high gallery density and distinctive gallery levels. This article presents a preliminary study of speleogenesis and cave development based on detailed research in recently discovered caves of the Schiefergebirge and literature information. 2. Evolution of the landscape In the study area the origin of caves is closely connected with the Neogene to Pleistocene history of the landscape. Depression areas of massive limestones are mostly intercalated with siliciclastic rock series. A peneplain with 20 to 50 m thick tropic weathering layers encompassed the Schiefergebirge in the Paleogen. A climatic change beginning in the Oligocene led to the selective deepening of the limestone areas within the peneplain. These depressions were used by rivers as discharge conduits (indicated by gravel deposits; SCHMIDT, 1975). The corrosive deepening of the limestone surface and formation of a "Karstrandebene" was the result of thorough wetting below the valley flat (MORELL, 1993). In the Miocene/Pliocene the elevation of the Schiefergebirge set in with beginning erosion of weathering layers that were sometimes preserved in dolines, karstic pockets, etc. The cyclic climatic changes and continuous elevation caused valley formation in the Pleistocene. There is a simple succession of terracing and linear valley cutting (SCHREINER, 1992): I) Interglacial stagnation without substantial erosion and only little accumulation. 2) Early glacial erosion with maximum valley entrenchment. 3) Full glacial deposition because the resulting gravel detritus cannot be transported any more. Table I shows the river terraces and their supposed chronological range for the Ruhr and the tributary rivers of the Northern Rheinisches Schiefergebirge. 3 Examined caves Detailed mapping and research have been performed in caves of the Iserlohner ("limestone depression"). In a 3 km long section of the Grilnerbach Valley there are known I 6 km of cave galleries of which the Dechen-, Knitter-, Hiittenblaserschacht-, B7and Bunker-Ernst caves comprise some 13 km. Comprehensive treatises of this karstic area can be found by HAMMEllsCHMIDT et al. (1995) and NIGGEMANN (1995). The biggest cave system of the Ronne Valley (Friedrich Cave, 1300 m length, and Feldhof Cave, 200 m) is in Balve (KOLARIK & HAMMERSCHMIDT, 1987). With a total estimated length of 7 km the Attendomer Tropfsteinhohle (Atta Cave) in the Bigge Valley (50 km S' of Iserlohn) is the biggest cave of the Rheinisches Schiefergebirge (HAMMERsCHMIDT, 1995). Symposium 7: Physical Speleology 151

PAGE 165

Series Holocene Late Pleist Middle Pleist. Early Pleist Earliest Pleist. Pleistocene classification Post glacial Weichsel-glac Eem-Interglac Saale-Glacial Wartbe-Stade Main-Interst. Drenthe-Stade Holstein-Interglac Elster-GlaciaJ Cromer-Complex Menap-Complex Tegelen-Complex Terrace River marsh soil Lower terrace Low Middle-Terr High Middle-Terr Low Main-Terr High Main-Terr Height above valley flat [ml 1-2 -4m(be low flat) 8-13 17-23 32-40 43-53 Otiifel-Terrace 63-74 Table 1: Classijicatio11 of river terraces (Len11e River) after VON KAMP (1972) and correlation with the chronology of Ruhr river terraces of the geological map 4510 Witten (JANSEN, 1980). 4. Morphology and sedimentology The cave galleries mostly follow NNW / SSE striking transverse joints or WSW/ENE striking joints of bedding Deviations arise in the case of gently dipping folding structures. At least two levels of main galleries (fossil "collecteurs") can be distinguished even in different cave systems of the same area. The tubular main galleries are considerably filled with elastic loose sediments, so the original volume is mostly unknown. Galleries of one level also show similar morphological and sedimentological features All morphological indications (corrosive forms like anastomoses, dissolution pockets, etc.) prove a phreatic origin of caves by corrosion. Different generations of speleothems can be distinguished in the upper cave level situated 20 to 30 m above the valley flat: An old generation includes bulky, big and often brown coloured speleothems with a rough, warty or nodular surface The recent, mostly active speleothems are often white coloured, fragile and have a smooth surface. Stalactites, in particular, show transitions Speleothems of the lower cave level (5 to 10 m above the valley flat) are totally part of the young generation The qualitative and quantitative abundance of speleothems in the upper cave level indicate a long growth period and thus a higher maximum age of speleothems (and of cave galleries) than in the lower level. HENNIG (1979) reported maximum speleothem ages of 285,000 to 340,000 years ( 23 0Th! 234 u method) for the Schledde Cave in Iserlohn belonging to the upper level. In the Dechenhtihle for the same level HAUSMANN (1986) determined maximum Th/U ages of 244,000 (+83,000, -45,500) years. HOMANN (1979) mentioned maximum 14 c ages of9,035 175) years from Knitterhtihle belonging mainly to the lower level. Speleothems from the lower level are smaller and less widespread, explaining the muddy character of the galleries (muddy level). In the caves from the Gro.nerbach Valley granulometrical and mineralogical composition of the cave clay is similar. It could be shown that the origin of the cave clay is polygenetic (Tertiary fluvial deposits, Pliocene to early Pleistocene terra fusca, Pleistocene loess, gravels and slope wash alluvium). Bones of Stephanorhinus kirchbergensis, Ursus spelaeus, etc. are evidential of the Pleistocene age of cave clay, a fact that is also supported by a high content of silt material derived from loess On the Pleistocene surface a mixing of different elastic sediments seems to have occurred, which were brought in the caves by sinking rivers and/or by sliding at valley slopes. The sedimentary and morphological research mentioned above only refers to the Grilnerbach Valley caves but a similar trend can be seen in other caves (e.g. Friedrich Cave Atta Cave and caves near Warstein). A differentiation between an upper and a lower level seems to exist in many caves of the Schiefergebirge 5. Origin and development of the caves The direction of cave galleries follows the parting plane system just as the hydraulic gradient. There is no evidence for any stratigraphical or facial control of cave development Due to low hydraulic gradients the extended genesis of main galleries could not begin before the Drufel terrace level which is thought to be some 500,000 to 700,000 years old. Both cave levels are indirectly related to younger river terraces (upper level lower main terrace; lower level lower middle terrace) The accumulation of terraces took place in late glacial phases. Stagnation (no considerable accumulation or downcutting) has characterized the following interglacial phase. Stagnation phases concerning constant heights of aquifer surfaces are essential for the developing of horizontal cave passages with big gallery cross sections (PALMER, 1987). So these main galleries have developed in such long, stagnant interglacial phases by slowly streaming karstic ground water that was and still is flowing towards the bigger rivers (e g. Lenne, Bigge) below the surface. Figure 1 shows the development of caves situated in the Grilnerbach Valley The phreatic development of meter-sized tubular cave galleries can only last 10,000 years after the initial phase (DREYBRODT, 1990). The relief of the groundwater table is determined by the resistance of fluid flow which is low for high density of parting planes. This is demonstrated by the four-state-model (FORD & EWERS, 1978). State l is represented by a high resistance of fluid flow, whereas state 4 stands for low resistance (water table caves). Most caves of the Schiefergebirge belong to state 3 or 4 as can be recognized by high parting plane density and existence of cave levels. The development of caves took place in the epiphreatic zone. The vadose flowing of these water table caves occurred in a late stage because of decreasing water quantity or decreasing resistance of fluid flow as a result of increasing gallery cross sections. The 4-phases-epiphreatic model of DAVIES (1960) can help to illustrate the cave development. I) Deep phreatic random origin of solution pipes, pockets, etc. caused by local groundwater flow conditions. 2) Merging and mature development of these solution structures along the epiphreatic zone during a stagnant head of the groundwater which is referred to river terraces. 3) Partial filling of caves by elastic sediments within the phreatic zone. 4) Draining of the cave caused by valley downcutting; speleothem genesis; destruction of the cave by breakdown processes. Due to the lack of datable terrace sediments outside the caves, an absolute chronological classification of cave development phases is difficult. However, figure 2 shows first results. The lower cave level belongs to the Drenthe stadia! lower middle 152 Proceedings of the 12 th International Congress of Speleology, 1997, Switzerland Volume 1

PAGE 166

terrace and is of Eemian age, which is based on sinter datations with maximum holocene ages (HOMANN, 1979) Also, there is no proof for Warthe stadia! river terraces. The upper cave level which is correlated to the lower main terrace of Cromer age developed in a younger interglacial period of the Cromer complex because speleothems of Holstein age are found within the galleries. The active level of the Knitterhohle comprising a small undergound brook marks the actual water table and is correlated to the lower terrace of Weichselian age. So cave development continues under Holocene conditions. The bulky character of speleothems within the upper cave level is caused by stronger quantities of precipitation within the Holstein and Eem interglacials which therefore increase the growth rate of speleothems. Distinctive speleothem forms like disques or cave blisters originated by capillary water were found only in the upper level. Obviously and for the same reason the hydrostatic pressure on flat sinter formations (walls, ceiling) was higher than today. This pressure is thought to control the growth of capillary sinter. Furthermore, higher dripping rates in the cave cause larger quantities of aerosol water. This is represented by many occurences of subaerial corralloids and rough surfaces of sinter formations in the upper cave level. In the northern Rheinisches Schiefergebirge, the beginning of extended horizontal cave level development can be appointed to 400,000 to 500,000 years. Certainly, the real beginning of speleogenesis (random deep phreatic solution pockets) can be older. DREYBRODT (1990) has calculated ages between 0.4 and 2.5 million years for different karstic systems of the Schwabische Alb. The results of this study are in agreement with other German karstic areas. West East Lenne ,----Gruncrbach Vall~ ----;. Piczometric surface T ---------::;.Lower Main Terrace Dechenh6nle Interglacial stagnation {plu=l!t development of the DotbenhOhle Main Gallc,y ) -----r ----lEarly glacial erosion (drauung of Dotbcnhohlc) following High-glacial accumulation oflh< High Middle Terrace, Interglacial stagnation ( relaled cave l evel is not devel >ped) Early glacial erosion ---HishMT Low. MT Hish-~-r Lower Middle T errace 6 High jlac:ial accumulal!oo "' ----==--T .-.:;:t Acknowledgement This study is part of a thesis realized at the Institute of geology at the Ruhr University Bochum I wish to thank Prof. Richter and Dr. Brix for instruction and discussion. Furthermore, I thank all members of the Speliiogruppe Letmathe, especially Elmar Hammerschmidt, for practical support in the caves. References DAVIES, W.E. 1960 Origin of caves in folded limestone. NSS Bull 22: 5-18. DREYBRODT, W. 1990. Ein Modell der Entwicklung von Karst unter Berilcksichtigung der Losungskinetik auf Kalkstein: Anwendung auf die Verkarstung in der Schwiibischen Alb Laichinger H6hlenfreund, 25 (2) : 47-83 FORD, D.C. & EWERS, R .O. 1978. The development of limestone cave systems in the dimensions of length and breadth Can. J. Earth Sci., 15: 1783-1798. HAMMERSCHMIDT, E. 1996. Die neuen Teile der Attendomer Tropfsteinhohle. In: (Ahrweiler, R.): Die Hohlen der Attendom Elsper Doppelmulde. Karst u Hohle, 1991 / 92: 29-36. HAMMERSCHMIDT, E., NIGGEMANN, S., GREBE, W., OEIZE, R ., BRIX M.R. & RiclfIER, D.K. 1995: Hohlen in Iserlohn. Schriften z Karstund H6hlenk. in Westf., l: 154 S. HAUSMANN, R. 1986. Allgemeiner Ko=entar zu den Ergebnissen der Uranund Thorium-Isotopenanalysen. Unveroff. Manuskript; Koln. HENNIG, G. 1979. Beitrage zur Th-230/U-234Altersbesti=ung von Hohlensintern sowie ein Vergleich der erzielten Ergebnisse mit denen anderer Absolutdatierungsmethoden. Diss Univ. Koln: 171 S. HOMANN, W. 1979. Zurn Wachstum holoz.i:i.ner Grof3stalagmiten in der Knitterhcihle bei Letmathe / Sauerland und zur Methodik der Sinter-Probenentnahme durch Kembohrungen Dortmunder Beitr z. Landeskunde. 13: 45-63 JANSEN, F. 1980. Geologische Karte von Nordrhein Westfalen I :25.000. Erlauterungen zu Blatt 4510 Witten: 176 S. KOLARIK, T & HAMMERSCHMIDT, E. 1987. Die Friedrichshohle im Honoetal. Mitt u Ber. Speliiogr. Letmathe, 4 (2) : 26-31. MORELL, U. 1993. Das Massenkalkgebiet der Honoe eine Karstlandschaft Antiberg, 52/53: 66 S. NIGGEMANN, S. 1995. Geologische Kartierung der Dechenhohle und ihrer Umgebung im Massenkalk bei Iserlohn Diplom-Kartierung Ruhr-Universitat Bochum: 128 S. (unveroff.) PALMER, AN. 1987. Cave levels and their interpretation NSS Bull., 49: 50-66. SCHMIDT, K.H. 1975. Geomorphologische Untersuchungen in Karstgebieten des Bergisch-Sauerlandischen Gebirges Bochumer Geogr. Arb ., 22: 156 S. SCHREINER, A. 1992. Einfilhrung in die Quartiirgeologie Schweizerbart' sche Verlagsbuchhandlung, Stuttgart: 257 S VON KAMP, H. 1972. Geologische Karte von Nordrhein Westfalen 1:25.000. Erlauterungen zu Blatt 4611 Hohenlimburg: 182 s WINOGRAD, l J COPLEN, T.B., LANDWEHR, J.M., RIGGS, lnterglncialslagl18tion ( phraticdevelopmenloflhelowercave l evel A C ., LUDWIG, K.R., Szabo, B.J., Kolesar, P.T. & Revesz, K.M. 1992 Continous 500,000-year climate record from vein calcite in Fig. 1: Model representing the development of cave levels and Devils Hole, Nevada. Science, 258: 255-260 river terraces in the Griinerbach Valley. Sympos i um 7: Physical Speleology 153

PAGE 167

Holocene Ecm-lnl er glacial f : We, chxlI N Glac 1al L ~ I ~ I "i' I o "i' "" ;:;,o 0 100 .?? I ] .,:. r~ !] ::; > ,j Saale-ln lerstadial 200 I I I 300 Tim e [ka] :'.l \ ------0 u I 0 1l 3~ ~j ::E~ / 1-----f l!. 3 'll~ 1 Cromer C omp l e x 400 ~u s: ::,_ \ I I h 1j = ~;; ::; ;Early-Pleislo ccne Glacials/ ln lerglac 1al s 500 Sch leddehOh l c Dec hcnh
PAGE 168

Mechanics of cave breakdown: Relative importance of shear strength and fracture toughness Elizabeth L. White and William B. White Environmental Resources Research lnstitute and Department of Geosciences The Pennsylvania State University, University Park, PA 16802 USA Zusammenfassung Einsturz einer Hi:ihlendecke ist ein Problem von Brechmechanisrnen. Vorhandene Modelle fur Einsti.irze sind auf der Theorie von festen und i.iberhiingenden Balken begri.indet, rnit der Schubkraft des Daches als hauptsiichlicher Parameter. Die Entstehung von gewi:ilbten Dachprofilen vergri:issert die mechanische Starke. Breite (20-25 m), flache Deeken in den Hi:ihlen der Isla de Mona (Puerto Rico) ist grosser als die Stabilitiitsgrenzen, die auf zerbrechlichen Modellen begri.indet sind. Man muss die Brech-ziihe in Bilanz nehmen neben der Schubkraft, um angemessener weise die Stabilitiit der Hi:ihlendecken zu modellieren. Abstract Breakdown of cave roofs is essentially a problem in fracture mechanics. Existing models for breakdown are based on the theory of fixed and cantilever beams with the shear strength of the roof rock as the material parameter. Development of arched roof profiles increases the overall mechanical strength. Wide (20 25 m), flat ceilings in the caves of Isla de Mona (Puerto Rico) exceed stability limits based on the brittle fracture model. It is necessary to take account of the fracture toughness in addition to shear strength in order to adequately model the stability of cave ceilings. The Breakdown Problem The detailed mechanics of roof failure are based on a model proposed by DAVI ES ( 1951) who treated breakdown as a process of brittle fracture of either fixed or cantilever beams The analysis was simply a balance between the gravitational load and the shear strength of the bedrock leading to a very simple equation for cave roof stability tc = p A 2 / 2 S t c = 3 p A 2 / 2 S [I] Fixed Beam Cantilever Beam where tc is the critical thickness for beam instability, p = density of bedrock, A = beam length (equivalent to width of passage), and S = flexural stress in the extreme fiber In this model, cave passages would be stable against breakdown until geological proce ses modified the parameters in the equations [I] and produced conditions of instability. WHITE and WHITE (1969) identified seven geological processes that would destabilize cave passages and produce breakdown Evidence Against the Fixed Beam Model The fixed beam model implies a completely elastic response of the ceiling beds. It does not allow for inelastic deformation and long term creep that could lead to bed failure in the absence of any geologic changes. Jn the Mammoth Cave System (Kentucky USA) there have been three documented roof failures in the present Century As part of their investigation of breakdown in the 1960s WHITE and WHITE ( 1969) prepared a high precision topographic map of a plastically-deformed sagging bed in Pohl Avenue in the Unknown Cave portion of the System That bed collapsed in the 1970s Clearly, creep and plastic defonnation play a role in breakdown collapse There are also examples of cave ceilings that are stable when the fixed beam model predicts they should be unstable. The best examples known to the authors are the caves of Isla de Mona, a 6-km diameter limestone island located between Puerto Rico and the Dominican Republic. The Isla de Mona caves are characterized by wide chambers, typically I 5 m high but as much as 20 25 111 wide. The ceilings are typically flat often punctured by skylights, and rarely having a domal shape to provide mechanical stability Yet breakdown in these chambers is relatively sparse These chambers would not be mechanically stable according to the fixed beam model. Fracture Mechanics Modeling of Cave Breakdown The basis for a more comprehensive model was propo ed by THARP ( 1995). Materials break through a mechanism of crack propagation. Microcracking, which allows for inelastic deformation and features within the material that inhibit crack propagation can greatly increase the resistance to fracture. In addition to the shear strength, a critical parameter is the fracture toughness, K I c The Paleozoic limestones which contain most of the caves of Eastern United States are dense, fine-grained rock Coarse fracturing occurs as joints and bedding plane partings, but within the rock mass there is little to inhibit crack propagation and these rocks are subject to brittle fracture. It is for this reason that the fixed beam model has worked so well. However, microcracking is possible leading to creep and ultimate failure as in the Mammoth Cave examples. The Miocene limestones of Isla de Mona have never been deeply buried. They are porous and contain an interconnected arrangement of small vugs. Propagating cracks are blocked by the vugs and pore spaces Although the Isla de Mona rocks are softer than the Paleozoic limestones, they have a much higher toughness As a result, the Isla de Mona cave chambers are stable with greater roof spans than would be expected from the fixed beam model References DAVIES W. E. 1951. Mechanics of cavern breakdown. National Speleol. Soc. Bull. 13: 36-43. THARP, T. M. 1995. Design against collapse of karst caverns Jn (B F Beck ed.): Karst Geohazards. A.A. Balkema, Rotterdam : 397-406 WHITE, E. L. & W. B WHITE. 1969 Process of cavern breakdown. National Speleol. Soc. Bull. 31 : 83-96. Symposium 7 : Phys ical Speleology 155

PAGE 169

Phreatic Channels in Velika dolina Collapse Doline (Skocjanske jame Caves Slovenia ) Martin Knez Institute for Karst Research ZRC SAZU, SI-6230 Postojna, Slovenia Abstrac t Collapse doline might be an efficient spot to observe and to study phreatic channels Question which I wanted to answer during the research was originated from collapse doline Velika dolina which represents one of biggest collapses in Skocjanske jame Cave system After visual observation one can easily notice that underground karstification do not appear displaced in disorder in rocky walls but is gathered along small number of bedding-planes which author calls main bedding-planes (Figs. I 2) Because similar evidences noticed some other researchers in Slovene karst region too (SUSTERSIC 1994 ) there raised a suspicion : Is it possible to find in this facts widder lawfulness and where are the roots for it. Main working method was microscoping over 300 thin-sections and there done over 10.000 mesurements presented in numeric form. We did not find the concordance between the inception reasons (for instance "trans-bedding contrast low grade of organic substances in the rock, reduction environment in the sediment, the influence of strong acids, others) and the actual state in Velika Dolina collapse doline (EWERS, 1966; RAUCH & WHITE, 1970; FORD & EWERS 1978 ; WORTHINGTO I 991; LOWE, 1992) Primary phreatic channels are concentrated along only three "formative" bedding-planes among 62 observed; less than along 5 % of all bedding-planes (KNEZ, 1996) This concordance cannot be only apparent. But, it was clearly evidenced that the inception, although the rock was very pure was concentrated on few bedding-planes only Figure I. View of Ve/ika dolina coll a pse doline (Skocja11ske jame Caves). Fig u re 2. Bedding-p l anes in Velika dolina collapse doline References EWERS, R 0 1966 Bedding-plane Anastomoses and Their Relation to Cavern Pa s sage s. Bull at. Spel. Soc. 28 3 I 33140, Arlington FORD D C. & EWERS, R 0 1978 The development of limestone cave systems in the dimensions of lenght and depth Canadian Journal of Earth Sciences, 15, 11, 1783-1798 KNEZ M 1996 Vpliv lezik na razvoj kraskih jam (Primer Velike doline, Skocjanske jame) Znanstvenoraziskovalni center SAZU 14 186 str. Ljubljana LOWE D. J. 1992. The origin of limestone ca v erns: an inception horizon hypotesis Unpublished PhD thesis XIX+5 I 2 str ., Manche s ter Polytechnic Manchester. RAUCH H W & WHITE W B. 1970 Lithologic Controls on the Development of Solution Porosity in Carbonate Aquifers Water Resources Research 6, 4 1175-1192 Pennsylvania State University, Pennsylvania SUSTERSIC, F ., 1994, Jama Kloka in zacetje ase jame 36 930 Ljubljana WORTHINGTON, S. R. H 1991. Karst h y drogeology of Canadian Rocky Mountains Unpublished PhD thesis, XVII + 227 str., McMaster University Hamilton. Hamilton 1 56 Proceedings of the 12 1 h Internat i onal Congress of Speleology 1997 Sw i tzerland Vo l ume 1

PAGE 170

Artesian speleogenetic setting Alexander Klimcbouk Institute of Geological Sciences, National Academy of Sciences, P.O.Box 224 / 8 Kiev-30, 252030 Ukraine Abstract The traditional paradigm of karstology is based largely on the concepts of unconfined or "open" karst. Widely accepted definitions oftbe tenn "karst" have a largely geomorphological meaning and formally have left no space for deep-seated artesian karst that is not manifested at the surface Here the problem of artesian speleogenesis is overviewed and a new approach is developed based on the non classical concept of artesian flow. Major features of artesian speleogenetic settings and artesian cave systems are summarised Artesian speleogenetic settings have a number of peculiarities that distinguish them from traditionally defined phreatic conditions in unconfined karst. The questions of hydrodynamics dissolution mechanisms, development of conduit systems cave morphology, and sediments. 1. Introduction: Artesian karst-a conceptual problem The origin of the term karst and the history of karstological studies led to karst being most commonly treated as a specific landscape or terrain, with distinctive hydrology and landforms (e.g J E N N INGS 1985; WHITE I 988; FORD & WILLIAMS, I 989). Most karst/speleogenetic theories are concerned with unconfined karst settings and ultimately imply close hydrological and mor phogenetic relationships between the surface and subsurface. Strictly, such definitions, and the who l e traditional karst para digm either ignore deep-seated karst that has no apparent rela tionship with the visible landscape, or treat such features as pa laeokarst. Some authors distinguish the special category of in trastratal karst, formed within already buried rocks, where kar stification is younger than the cover (QUINLAN 1978; PALMER & PALMER 1989 ; BOSAK FORD & GLAZEK I 989) The latter authors emphasise that there is abundant modem (active) intra stratal karstification in progress Confined or artesian karst falls within this category but the whole concept of intra-strata! karst, as well as of artesian karst, does not seem to be an essential part of the traditional karstological paradigm A term "karst" is also used to describe particular landforms and subsurface fearures produced by a speci fie set of processes in which dissolution is the main one initiating or triggering other processes such as erosion collapse and subsidence (QUINLAN 1978 ; MtLANOVIC 1981 ; BONACCI 1987; JAMES & CHOQUETTE 1988) This allows reference to deep-seated dissolution features as karst but does not resolve the general conceptual problem. A wider approach, long accepted in the Soviet Union was that karst was regarded as a process, or a combination of process and resulting phenomena In western literature, Huntoon (1995) discussed a need for a process-oriented definition of karst that emphasizes its hydrological function and geobydrological uniqueness, rather then its ambiguous morphological character. Karst is defined as a geological environment containing soluble rocks with a permeability structure dominated by interconnected conduits dissolved from the host rock organised to facilitate fluid circulation in a downgradient direction and wherein the permeability structure evolved as a consequence of dissolution by the fluid (p.343). This definition, which is sufficiently broad to encompass circulation systems in the unsarurated zone as well as in unconfined and confined aquifer system is adopted here Specific works on artesian speleogenesis are scarce FORD ( 1988) distinguished artesian caves as a type, referring to maze examples However the common view is that "true artesian conditions with lateral basinal flow from distant recharge areas through separated aquifers offer limited hydrodynamic and chemical potential for speleogenesis Examples of karst in deep seated artesian settings are commonly treated as palaeokarst. Adoption of the non-classical concept of hydrodynamics of artesian basins, which implies hydraulic continuity in basins and close cross-formation communication between aquifers has al lowed new conceptual models concerning artesian speleogenesis to originate during the last decade. Further development of these ideas allows abundant provings of deep-seated intra-strata! karst occurring under present artesian conditions to be treated as a modern (active) phenomenon. A B D 1 [Ililil 2 m 3 [ill] 4 B s ~ @Is Figure I: Flow pattern in the multi-storey artesian system (From Shestopalov, 1989). 1 = aquifers, 2-3 = aquitards, 4-5 = flow directions, 6 = potentiometric surfaces of aquifers. A = area of de scending circulation, and B = area of ascending circulation. 2. Concepts of artesian flow The classic concept of artesian flow implies that recharge oc curs only on limited areas where aquifers crop out at the surface (usually at basin margins) and that groundwater flows laterally through separate aquifers in the confined area. This brings about a major problem in interpreting artesian speleogenesis With a large distance and travel time through a karst unit water should not be capable of further dissolution in the confined flow area. The origin of maze caves is most commonly attributed to artesian settings but strong arguments developed from hydrodynamic (PALMER, 1975) and kinetic (PALMER, 1991) considerations questioned the possibility of maze caves origin if lateral artesian flow is implied through a karst rock unit. There is a common view that "true" artesian conditions (in the sense of the classical concept) offer limited hydrodynamic and chemical potential for karstification. Numerous existing provings ofkarst cavernosity in deep-seated artesian settings are commonly treated as palaeo karst. The new approach to the problem is based upon non classical concepts of artesian basin hydrodynamics developed within mainstream hydrogeology in recent decades from aquifer and well hydraulics data (on a local scale), and from basin hy draulics and water resources evaluations (on a regional scale) Hydraulic continuity in basins and close cross-formational com munication between aquifers are implied (e g. MJATIEV, 1947 ; SHESTOPALOV, 1981 1989; TOTH 1995; figure I) According to these views, recharge to, and discharge from a given aquifer (or vertical groundwater exchange in a system) may take place across dividing beds throughout the whole confined flow area It is controlled by the head relationships in a system (which are in tum, controlled to a significant degree by surface topography; figure 2) and by the presence of areas (zones) of enhanced per meability in dividing beds (facial "windows", zones of enhanced fissuring, fault zones, etc ) The most important speleogenetic implication of this concept is that a karst rock unit can receive Sympos i um 7 : Physical Speleology 157

PAGE 171

areally dispersed aggressive recharge from adjacent formatio n s (KLIM C HOUK, 1994). Adoption of the non-classical concept of artesian basin hy drodynamics allowed new speleogenetic interpretations to emerge. It provides an explanation of the artesian origin of some of the world s largest cave systems in South Dakota and in the Western Ukraine (FORD 1989 ; KLIM C HOUK, 1990, 1992 1994) and suggests some approach to the problem (KLIMCHOUK 1994) 3. Major features of artesian speleogenetic settings H y drog e ologic a J s tructur e In a typical ba s in aquifer s are separated from each other and from any upper unconfined aquifer by poorely permeabile beds. Initial permeabilities of common aquifers (e g some elastic rocks) are normally greater then that of karstifiable units (such as massive limestones or sulphates) before speleogenesis Thus karst units may initially act as separating beds (aquitards) As late diagenesis and tectonism impose fissure permeability on a s edi mentary sequence karstifiable units increasingly transmit groundwater between "normal" (i.e non-karstic) aquifers in zones where sufficient head gradients exist. Such simplified hy drostratigraphy when a karstifiable unit is sandwiched between formations with initially more favourable aquifer properties seems to be ideal for speleogenetic development. It is noteworthy that when conduit permeability is developed within the karst unit, conventional karstological wisdom views such hydrostratigraphy in the opposite way: the karst aquifer sandwiched between aqui tards In many cases beds of higher initial porosity and relatively diffused permeability (such as oolitic beds) exist within a mas sive and scarcely fissured carbonate sequence (LOWE 1992) They will act as aquifers and hydraulic communication across dividing compact beds will improve with time through speleoge netic development. More complex relationships occur in thick lithologically inhomogeneous sequences composed, for instance, of intercalated carbonate, sulphate and elastic beds with con trasting permeabilities In artesian settings the inversion of hy drogeological functions of different beds in a sequence during the speleogenetic evolution of karstifiable beds is quite common (KLIM C HOUK 1992, 1994; LOWE 1992) This reflects the gener ally underestimated fact that karst permeability changes through time, while the permeability of non-karstic beds is a relatively s tatic property On the local scale, the actual flow paths through a karstifiable unit are strongly guided by the initial fissure configuration R egio n a l a nd l ocal h y dr o d y n amics Artesian basins vary considerably in size, configuration and hydrogeological structure Regions with substantial local topog raphy and a stratified sedimentary cover (such as high-relief plat forms and foreland basins) are characterised by complex flow architecture Besides marginal recharge areas and lateral flow components, this architecture includes: laterally alternating re charge and discharge areas (areas of correspondingly descending and ascending cross communication) occurring through the whole region of confined flow ; superimposition of recharge discharge regimes for particular aquifers in a system and flow systems at different scales (SHESTOPALOV 1981, 1989 ; TOTH 1995) Groundwater circulation in a basin tends to adjust to configu rations of maximum and minimum fluid potentials The latter may be due to compression compaction dilatation, thermal ef fects, chemical processes etc, although hydraulic head gradients become the increasingly prevailing driving force once the subsi dence trend of tectonic movement has had reversed. In a typical basin stratigraphically lower (older) aquiferous formations crop out at successively higher elevations along the basin margins imposing greater heads.Hence, the regional flow system is nor mally characterised by an increase in head with depth, so that ascending cross-formational communication predominates through the confined flow area However local topographic highs superimpose local flow patterns (recharge / descending communication) encompassing mainly the upper part of the hy drogeological sequence. Ascending cross-formational hydraulic communication is the most common pattern to fa v our spele ogenesis in artesian settings ( figure 2). In the interior parts of a basin lateral circulation in an aquifer is normally slow. Communication acros s confining beds occur s even more slowly commonly measurable only on the times cales of centuries or millennia However, considering the geologically lengthy time-scales of basinal development, such movement may account for initiation of proto-conduits or for speleo-inception in the context described by LOWE (1992) Conduit development oc curs later, when uplift and geomorphic differentiation cause sub stantial activation of groundwater circulation In favourable areas, where upward head gradients are maxi mised (topographic-potentiometric lows deeply incised valleys ) and coincide with zones of greatest permeability in confining beds particularly in the uppermost one (fault zones fault-cored anticlines structural "windows", etc ), groundwater circulation can be active enough to account for the full development of cave systems The type of cave pattern depends largely upon the ex isting structural features. Rectilinear fissure caves and small mazes with fissure-like ascending pits (eastern Missouri BROD 1964), or without such pits (Black Sea basin, south Ukraine) can develop as can extensive multi-storey mazes such as Wind and Jewel Caves in the Black Hills (FORD, 1989), the great gypsum mazes of the Western Ukraine (KLIMCHO U K, 1990 I 992) and Botovskaya Cave, Siberia (FIUPPOV 1994) Multi storey mazes form when fissures are distributed uniformly in the lateral direc tion and organised in extensive superimposed networks confined to specific horizons (KLIMCHOUK I 994; KLIMCHOUK ET AL., 1995) Maze patterns are quite characteristic, although not confined to artesian speleogenesis. The diagnostic feature is uniform pas sage size and morphology at a given locality regardless of the pattern type. The fundamental reason is that artesian speleogene sis is largely discharge-controlled, and no considerable increase in discharge occurs as an initial fissure widens by dissolution. The simplest architecture that can be considered is that of two normal" (non-karstic) aquifers separated by a carbonate unit with fissure permeability with the whole system confined by an upper, non-karstifiable, aquitard In this case upward circulation and slow discharge from the system through the upper confining bed will converge towards an incising valley (figure 2). According to PALMER S (1984) theoretical consideration non uniform enlargement of passages occurs at the early (laminar flow) stage when the rate at which any route enlarges depends upon the amount of flow rather than solution kinetics (discharge controlled development). Enlargement rates increase if a passage increases its discharge, either by capturing water from neigh bouring passages or by enlarging its primary catchment area This explains competitive development of flow routes and the importance of initial differences in hydraulic efficiency in deter mining successful conduits in unconfined karsts At the next (solution kinetics-controlled) stage, the enlargement rate ap proaches a constant value, at which successful conduits continue to develop. ln the artesian settings specified above the amount of upward flow is governed not by available recharge to the karstifiable unit (which is essentially uniform to all fissures contacting the lower aquifer) but by constrained discharge through the upper confin ing bed. Even a small dissolutional enlargement of the initial network in the karst unit will rapidly minimise the head differ ence in both adjoining aquifers. After that the flow rate will be governed solely by the upper confining bed's transmissivity which is roughly constant and normally relatively low unless the confinement is fully breached by an incising valley Thus, the enlargement rate for all fissures (conduits) in the network will also be constant through most of the artesian stage, regardless of their initial dimensions. Superimposed upon the general background of ascending cir culation in areas of potentiometric lows, local convection cells may develop These are driven by density gradients imposed by thermal or chemical contrasts between the bulk water in the aqui fer and recharge water from the underlying formation This ac tivity may cause pronounced medium-scale effects in an already formed system (such as the growth of ceiling cupolas, half-tube s etc) but can hardly account for the development of an entire cave system on its own 158 Proceedings of the 12 th International Congress of Speleology, 1997 Sw i tzerland Volume 1

PAGE 172

margina l recharge area of confined aquifers discharge area recharge area discharge area Figure 2. Flow architecture in a typical artesian basin of the platform type artesian spe l eogenesis Solutional potential of groundwaters in artesian settings The concept of cross-formational communication has impor tant implications when considering potential for dissolution. Re charge from an adjacent non-soluble formation can be highly ag gressive with respect to a karstifiable rock. This is mo st evident in the case of simple dissociation of gypsum, but holds true for carbonate karst in shallow artesian settings, where soil-generated CO is delivered by water from "classic" marginal recharge areas through non-carbonate aquifers. Waters in deep-seated horizons often have high concentrations of CO derived from degradation of hydrocarbons or from other sources. During the last decade it has been increasingly recognised that, in addition to the bicarbonate dissolutional mechanism. a number of redox reactions, particularly those involving sulphur compounds and sulphuric acid, may contribute to carbonate dis solution (FORD & WILLIAMS, 1989) Some of the related reaction s generate CO ,, leading to doubled solvency. Reactions involving strong acids re believed to be particularly important during the early stages of speleogenesis (WORTHINGTON, 1992; LOWE & GUNN, 1995) which, in the light of the above consideration, commonly occur under artesian conditions. Cross-formational hydraulic communication i s of paramount importance. It allows the involvement of different chemical mechanisms, as well as reinforcing mixing corrosion effects, as it causes migration of various reactants and reaction products between horizons bring ing contrasting geochemica l environments and waters of con trasting chemistries into interaction (KLIMCHOUK, 1996). More extreme thermobaric conditions occurring in deep settings further increase the range of possible speleogenetic effects. Moreover, some still more exotic, as yet poorly understood effects may have a role in speleogenesis. For instance, radiolysis of under ground waters produces free I1 and highly reactive oxidising agents (VovK, 1979) This and related effects await further evaluation. Stages of artesian speleogenesis (1) During continuing subsidence and burial groundwater circulation in a ba sin is driven mainly by pressure heads devel oped by compaction and dehydration of sediments. Some speleo inception effects may operate at this stage (2) As the tectonic regime reverses between subsidence and uplift, and geomorphic development locally exposes basinal for mations, fluid flow becomes increasingly driven by hydraulic head gradients Speleogenetic initiation occurs widely during this stage, even in deep-seated horizons ; in favourable conditions true conduits can develop Flow is slow, especially if a cross formational component is involved, and elapsed time spans are commonly l arge. (3) As continuing uplift brings artesian aquifers closer to the denudation surface, circulation in the entire system and the cross formational flow component are increasingly activated. The full development of cave syste ms occurs in a pattern of upward cross formational communication in zones of potentiometric low s where valleys incise into the upper confining bed. Deposition of fine elastic sediments, re-distributed from occasional cavern breakdown, may begin at this stage ----(4) Breaching of artesian confinement at local discharge areas dramatically increases upward flow through the system and en hances cave development along preferred routes. This is the last stage of artesian speleogenesis. Ongoing entrenchment of major valleys and areal disintegra tion of the capping aquitard lead to radical re-organisation of re charge disc h arge configurations and inversion of flow patterns; artesian cave systems become relict. A number of speleogenetic and karst morphogenetic models address these open karst set tings, but commonly they fail to explain relict artesian systems adequately. These models interpret the superimposed cave devel opment in terms of phreatic / water table / vadose flow conditions and tend to overlook the possibility that previous artesian history could have influenced such development The inception horizon concept of Lowe (l 992), although not directly refferring to con fined settings, provided an important step towards the recogni tion of processes operating deep within rock sequences during a more extensive geological history The succession of stages outlined above applies only to a given formation. On the regional (basinal) scale, situations repre senting different stages may occur simultaneously in different formations, at different depths or in different parts of a ba sin. 4. Major features of artesian speleogenesis These can be summarised as follows: I. The imposition of basic cave system frameworks show no genetic relationship to modern landscapes. However, in the con text of palaeo-geomorphology, active and significant cave growth is normally induced by and converges towards valleys incising into an upper aquitard. 2. Caves patterns are guided by fissuring, and passages within a given series (storey) are quite uniform in size and mor phology. Twoor three-dimensional (multi-storey) rectilinear mazes are typical, although neither confined to nor diagnostic of artesian speleogenesis. 3. Cei l ing cupolas or half-tubes originated by the action of convection circulation cells or currents are common medium scale morphological features. 4 Clastic cave sediments are represented mainly by fine clays and silts. These can be partly autochthonous (comprising insoluble residues) However, they are mainly allochthonous, intruded into artesian systems from overlying formations only during the late stages (stages 3-4). Breakdown processes are in duced by a decrease of hydrostatic pressure, and increasing flow velocities allow some transport and re-distribution of elastic ma terial. However gradient fields in the aquifer remain much more uniform, and the energy of water flows remains much lower than in unconfined settings, so that sediments display fine-grain com position. Whereas sediments in unconfined phreatic and, espe cially, vadose caves are characterised by great structural and lithological variation over short distances, and by rapid facies changes, artesian caves normally contain sediments that are much more uniform, and display simi lar facies even on a regional scale Symposium 7: Physical Speleology 159

PAGE 173

5 Speleothems commonly altho u gh not inev i tably include exotic" hydrothermal minerals or minerals deposited as the product s of various redox reactions on the physical and geo chemical barrier s that are typical in the case of ascending cross formational flow Most of the above features were suggested (D U BU ANS K Y, 1980 1990) and widely cited (e.g BAKALOWI C Z ET A L. 1987, FORD 1995) as being characteristic of hydrothermal speleogene sis However from the above considerations it is apparent that they are diagno s tic not of a particular set of physical (thermal) conditions but of a specific ( confined) type of flow system and regime Deep basinal flow i s commonly heated, and its upward discharge generally creates thermal anomalies. Realistically the hydrothermal karst concept is largely, although not completely encompassed within the artesian karst concept. 5 Conclusions I Artes i an karst/speleogenesis is widespread through conti nents and is the common rather than special case It accounts for the full development of some of the largest known caves in the world and many shorter caves It also has an immense importance in s peleo-inception By far the greate s t part of presently uncon fined karst rocks experienced more or less prolonged episodes of basinal development before undergoing major uplift Thus, arte sian karstification almost inevitably precedes unconfined karst development. 2 Adoption of the non-classical concept of artesian flow in karst hydrogeology gives a wider perspective for development of the theory of artesian karst/speleogenesis. The latter is adequate to explain the hydraulic and chemical mechanisms involved There are also numerous provings of modern deep-seated intra stratal karst, and major feature s of known relict artesian cave systems 3 Artesian speleogenesis occurs in different lithologies It involves various dissolution mechanisms that operate under dif ferent physical parameters, although resulting in similar cave features This indicates clearly that the main factor responsible for the genetic specifics of this kind of speleogenesis is the type of flow sy s tem not a single chemical or physical peculiarity of the genetic environment. 4 Recognition of the scale and importance of artesian karst/speleogenesis and of hydraulic continuity and cross formational communications between aquifers in artesian basins is indispensable for the correct interpretation of spe l eogenetic processe s (evolution ofkarst aquifers) and the re s ultant phenom ena regional karst water resource evaluations and the genesis of some karst-related mineral deposits It has numerous other theo retical and practical implication s, and ultimately requires in re placement of the traditional karstological paradigm which i s based largely upon concepts of unconfined kar st. References B A KALOWI CZ, M.J et. al. I 9 8 7 Thermal genesis of dis s olu tion caves in the Black Hills, South Dakota. Bull. Geol. Soc Amer. 99 : 729738 BO NACC I, 0 1987 Karst hydrology with special reference to the Dinaric Karst. Springer-Verlag New York : 184 BOSAK P ., F O RD, D C. & J GLAZ E K 1989 Terminology In : (P.Bosak D.Ford J Glazek & I.Horacek eds. ): Paleokarst: a sy s tematic and regional review Academia Praha: 25-32 BRODT L.G. 1964 Artesian origin of fissur e caves in Mis souri NSS Bull e zin 26 (3) : 83-114. D U BUANSK Y, Y V 1990 Regularities of the formation and modelling of hydrothermal karst. Nauka, Novosibirsk: 150 p D U BLJANSK Y, W.N. 1980. Hydrotherma l karst in the Alpine folded belt of southern parts of the USSR Kra s i Speleologia 3 ( 12) : 18-36 FILIPPO Y, A G 1994 Botovskaya cave in the East Siberia Problems of Physical Speleology. Moscow: MFTI Publ. 1021 1 0. (Russ ., res engl.). FORD, D C 1988. Characteristics of dissolutional cave sys tems in carbonate rocks In : (N P James & P W Choquette, eds.) : Paleokarst. Springer-Verlag New York : 24-57 FORD D C. 1989 Features of the genesis of Jewel Cave and Wind Ca v e, Black Hills South Dakota NSS Bull e tin 51: I 00110 FORD, D C & P.W WILLIAMS 1989 Karst geomorphology and hydrology Unwin Human London. 60 I p FORD T D 1995 Some thoughts on hydrothermal cave s Cave and Karst Science 22 (3): 107-118 H UN TOON, P W 1995 ls it appropriate to app l y porous me dia groundwater circulation models to karstic aquifers? In: (Aly I. El-Kadi ed ) : Groundwater models for resources analysis and management. Lewis Pub l ishers Boca R aton: 339-358 J A MES, N P & P W CHOQ UE TTE (eds ) 1988 Paleokarst. Springer-Ver l ag, New York : 416. JENNINGS J N 1985. Karst geomorphology Basil Blackwell Oxford : 293 p. KUMCHOUK, A B 1990 Artesian genesis of the large maze caves in the Miocene gypsum of the Western Ukraine Doklad y Akademii Nauk Ukrainskoj SSR ser.B 7 : 28-32 (Russ ., res Engl.) KLJMCHO U K A.B. 1992. Large gypsum caves in the Western Ukraine and their genesis. Ca ve S ci ence 19 (I) : 3-11 KLIMCHOUK A.B 1994 Speleogenesis under confined con ditions with recharge from adjacent formations Pub!. S e rv G e ol. Luxembourg v.XXV/1. Compt e s R e ndu s du C oll Int d e Karstol. a Luxembourg : 85-95 KLIMCHOUK A.B 1996 The role of karst in the genesis of sulfur deposits, Pre-Carpathian region, Ukraine Environmental Geology 28 (3). KLIMCHOUK, A.B V.N.ANDREJCHO U K, & l.l.T U R C HI N O V 1995 The structural prerequisites of speleogenesis in gypsum in the Western Ukraine Ukr Speleol.Assoc ., Kiev : 106 p. LOWE, D.J 1992 The origin of limestone caverns: an incep tion horizon hypothesis Ph.D Thesis Manchester Metropolitan University LOWE, D.J & J GUNN 1995. The role of strong acid in spe leo-inceptio n and subseq u ent cavern development. In : (I.Barany Kevel ed) : Environmental effects on karst terrains A c ta G e ographi c a vol.XXXIV Szeged Hungaria : 33-60 MILANOYIC P T 1981. Karst hydrology Water Resources Publications, Littleton CO : 434 p MJATIEV, A N 1947 Confined complex of underground wa ters and wells. l zv estija AN SSSR otd t e khni c h nauk 9 : 33-47 (russ ) PALMER A.N 1975. The origin of maze caves. NSS Bulletin 37 ( 3) : 56-76 PALMER A.N 1984 Geomorphic interpretation of karst fea tures. In : ( R G. Lafleur ed ) : Groundwater as a geomorphic agent. Allen & Unwin Boston : 173-209 PALMER, A.N. 1991. Origin and morphology of limestone caves Geol So c Am Bull 103: 1-21. PALMER, M.V & A.N PALMER 1989 Paleokarst of the United States In : (P Bosak, D Ford J.Glazek & I.Horacek eds ) : Paleokarst : a systematic and regional review. Academia Praha : 337 365 QUINLAN J F 1978 Types of karst, with emphasis on cover beds in their classification and development. Ph D Thesis Univ of Texas at Austin SHE S TOPALOV V.M 1981. Natural resources of underground water of platform artesian b asins of the Ukraine Nauk Durnka Kie v: 195 p (russ.) SH E STOPALOV V M ., ed. 1989 Water exchange in hydro geological structures of the Ukraine Water exchange under natu ral conditions Naukova durnka, Kiev: 288 p (russ ) TOTH, J. 1995 Hydraulic continuity in large sedimentary ba sins H y drogeology Journal 3 ( 4) : 4-15 VovK l.F. I 979 Radiolysis of underground waters and it s geochemical role. Nedra, Moscow: 231 p (Russ ) WHITE W.B. 1988 Geomorphology and hydrology of karst terraines. Oxford University P ress Oxford : 464 p. WORTHINGTON S R.H. 1992. Karst hydrogeology of the Ca nadian Rocky Mountains Ph D Thesis, McMaster U n iversity 160 P r oceedings of the 1 2 th Internationa l Congre ss of Speleolog y, 1997 Switzer l and Volume 1

PAGE 174

Speleogenetic effects of water density differences Alexan d er KJimchouk Institute of Geological Sciences, National Academy of Sciences, P.O.Box 224/8, Kiev-30, 252030, Ukraine Abstract High solubility of some karst rocks, particularly salt and gypsum, leads to a significant increase in the density of dissolving water. This can be up to 30% for salt and 0.1 % for gypsum, as compared to aggressive water entering a karst system. Gravitational separation of water may be well expressed as density stratification or as circulation cells and distinct buoyant currents. Such phenomena may occur in a single cave pool, or at aquifer scale. The most pronounced speleogenetic effects occur when aggressive water recharges a gypsum stratum continuously from an underlying formation under sluggish artesian flow conditions. ln this case dissolution by natural convection flow may affect the cave system layout. More localised effects include the formation of dissolution notches, bevels and facets (due to density stratification), and formation of ceiling half-tubes pockets, cupolas,and domepits (by buoyant currents) in confined settings. 1. Introduction As a karst rock is dissolved and the solute content of the dissolving water increases, the solution increases in density. The greater the solubility of a rock. the greater the density differences that may develop within an aquifer or water body, especially if there is a continuous or frequent pulse supply of contrasting relatively fresh water. This causes gravitational separation and convectional circulation of water, which are most pronounced under sluggish laminar flow conditions with continuous fresh water recharge. CURL ( 1966) provided a theoretical analysis of cave conduit enlargement by natural convection in a limestone aquifer, depicting transition conditions determining the prevalence of natural convection or forced flow regimes. He found that. with sufficiently slow water circulation, convection caused by density differences may be the primary flow mode for limestone removal. This is made possible by even extremely small compositional differences. Supposed morphological effects include the upward enlargement of anastomoses above bedding planes, development of vertical asymmetry of conduits, and upward growth of dome-like ceiling features. This paper discusses hydrological settings and types of water density difference phenomena. as well as the resulting morphological effects. 2. Water density stratification and local convection cells In shallow phreatic and water-table aquifers with sluggish flow, or in standing water bodies such as perched cave lakes marked water density stratification can develop. To maintain or reinforce such stratification, continuous or periodic inflow of relatively "fresh" water is needed. This may enter from above (vadose percolation). from the side (e.g. from sinking streams) or from a basal non-karstic aquifer (upward recharge). In all cases, less dense fresh water will tend to occupy the uppermost available position, fonning a more or less distinct layer. The hydrochemical stratification phenomenon is particularly pronounced, and has been well documented, in gypsum caves. In passages that originated (or significantly modified) at a shallow depth below the water-table, flat ceilings ("Laugdecke" in German) or bevels are common. They are formed by dissolution in the uppermost, aggressive, layer of water, where a pattern of small upand downwelling convection cells ("salt fingers") operates due to small density differences (KEMPE, 1972; KEMPE ET AL., 1975; figure I-A). Such flat ceilings are best displayed in gypsum caves in Germany, but also occur in the Western Ukraine, the Urals and Siberia. Flat ceilings, combined with inclined facets, are typical for many German caves, producing cross-section in the form of a tip-down triangle. This reflects dissolutional widening in low-energy shallow phreatic or artesian conditions, with successive fast draining, and without any significant morphological modification under water-table conditions. However, field observations from caves elsewhere than Germany suggest that not only flat ceilings, but also ceilings with cupola-like forms and gothic arch shapes can form under the same conditions (figure 1-C). A B c Kc,npc's model: d 0 convection of bulk water and small convection cells \ 'V' 'V' 'V' 'V' 'V' 'V' -V:/ ~~(]# I It \? P. V 06 facets~# :l:,=~dfuby rums 8 Figure 1. Wa ll notching and wall facet formation by dissolution caused by natural co11vectio11. A = Kempe's model; B,C,D = typical passage cross-sections from the gypsum caves of the Western Ukraine, showing varieties of morphological modification. In England, distinct water stratification has commonly been observed by divers in phreatic passages in the limestone caves of the Yorkshire Dales (CORDINGLEY, I 991 ). Brown "peaty" water is frequently seen occupying the upper part of the water body, whilst nearer the floor the water is virtually transparent. The interface between the two layers is commonly well defined. A possible explanation of the phenomenon is that relatively fresh (less dense) peat-stained water replaces most of the water in phreatic passages during a flood and is then trapped in the higher levels as denser autogenic percolation water accounts for an increasing percentage of flow in passages during a drought. Differential dissolution is believed to account for the formation of ceiling solution domes. It is also hypothesised that horizontal wall notches can be formed due to water stratification under phreatic conditions, instead of being formed at a water-table. as traditionally believed. Such notches have been observed in Symposium 7: Physical Speleology 161

PAGE 175

phreatic passages that have never been drained (CORDINGLEY, 1991). Hydrochernical stratification of water is well documented where water bodies with an open surface occur in gypsum caves, either as aquifer "windows or as perched pools. In the partly drained Miocene gypsum aquifer of the Western Ukraine (cave lakes with sluggish flow in Optimisticheskaya Cave) a distinct density stratification develops, with solute content rising downward from 1 1-1.3 to 2.0-2.5g r within a depth of 25-30cm. Subsequently, SI increases from -0.4 almost to saturation value, and dissolution rates (measured from tablet experiments) decrease from -6.4 to 7 2mg day cm in the uppermost la yer to 0.08 to -0 .06 mg day cm at depths below 15cm (KLIMCHOUK & AKSEM 1988) This leads to the formation of recent water level notche s in walls. Horizontal notching due to chemical stratification of water, with higher dissolution rates in the uppermost layer, is a common morphological effect for caves in all major karstifiable lithologies (FORD & WILLIAMS, 1989), although it is best displayed in salts (FRUMKIN, I 994) and gypsum. The typical morphological elements of notches are s teep inwardly inclined sidewalls, or facets ("Facetten" in German), described from gypsum caves in Germany (BIESE, 1931; PF EIFFER & HAH 1972) and the Urals (LUKIN, 1967) and explained as being formed due to conduit-scale convection circulation. Formation of facets in gypsum was studied in detail and modelled theoretically by KEMPE ( 1972) and KEMPE ET AL ( 1975). Dissolution causes layers of water adjacent to the walls to become denser than the bulk water in the upper horizon, so that films of water "slide" downwards along the walls (figure 1-A). Examples of such l-3mm-thick currents descending the walls were measured at 3-30cm min-1 in a cave pool by KEMPE ET AL ( 1975) The inclined plane of a facet forms because solute content increases and dissolution rate decrease s as the water in the layers moves downward. Morphological effects in the gypsum caves of the Western Ukraine The great maze caves in the Western Ukraine originated under artesian conditions, with dispersed upward recharge into the cave systems from the basal sandy-carbonate aquifer (KLIMCHOUK, 1992). The fissure-guided rectilinear passage networks are developed at several storeys, which are pre determined by the multi-storey occurrence of fissure networks (KLIMCHOUK, 1994 ; KLIMCHOUK ET AL., 1995). The multi-storey structure and different location of artesian cave systems in relation to modem, deeply entrenched, valleys, combined with differential regional uplift rate s, caused a variety of draining (dewatering) histories, and hence a varying degree of water-table modification of passage morphology Passages with different degrees of modification may occur at different levels, or in different areas of a single cave system. Artesian conduits that have experienced virtually no modification normally display fissureand cleft-like morphologies (figure 1-B). Less commonly they are tubular in shape. Many passages have one "leve l" of notching with inwardly inclined sidewalls; ceilings may be of different shapes (figure 1-C). Such formed under shallow artesian conditions, with dis olution by natural convection operating, and passed quickly to the fully drained state, so that no tme water-table notching developed. When more then one level of horizontal notching occur in the walls (figure 1-0), the additional notches signify more or less stabilized palaeo-positions of the water table. They can be traced continuously across large distances within a given series of a labyrinth. 3 Buoyant currents from basa l recharge u nde r artesian settin g s Cave development in artesian settings is commonly associated with recharge from an underlying aquifer (KLIMCHOUK, in press). Local hydraulics and details of the speleogenetic development are well illustrated by the gypsum caves of the Western Ukraine, although their features are neither unique to the region nor to gypsum. Head gradients are directed upwards but flow is slow due to the high resistance of the upper aquitard, and of the karstifiable unit itself during the early stages of speleogenetic development. As water in the underlying aquifer is less dense than water already within a karst system, it tends to move upwards by natural convection, so that forced flow and natural convection coincide. When forced flow through the karst system is negligible due to poor vertical connectivity and the high resistance of flow paths, the natural convection regime predominates and "c lo sed" cells develop. After dissolving some material and increasing in density, the water returns downwards into the underlying aquifer and outflows laterally with the regional flow. Conduits at the bottom of the gypsum stratum and along the lowermost available fissures grow by upward stoping due to natural convection dissolution. Solution cupolas and domepits develop along vertica l fissures and propagate upwards to, or even beyond, the edge of the fissure plane. In this way, connections can be established from two or more neighbouring domepits with open and interconnected fissures of the next higher storey of fissures. Thus, backward circulation loops can be extended through two storeys, continuing the speleogenetic development. As buoyant currents of aggressive recharge always tend to occupy the uppermost available space in a developing conduit system, further upward propagation of dissolutional forms driven by natural convection will proceed, and eventually lead to establishment of direct hydraulic connection with the overlying aquifer. When the high resistance to flow between the lower and upper aquifers is destroyed, the equipotential field is drastically re-organised and forced convection flow becomes predominant. Because high conductivity is now established through the gypsum stratum, the head gradient between the two aquifers is minimised. Flow remains small and slow due to the remaining relatively high resistivity of the upper aquitard. But the flow pattern changes from backward loops and "c losed cells to an unlooped ascending system. Natural convection still contributes to the overall flow, and gravitational separation of buoyant currents becomes perhaps even more distinct, as the general flow component pattern is now unlooped. The most common morphological effect of buoyant currents is the formation of keyhole cross-sections or distinct ceiling half tubes. Keyhole passages are usually assumed to be associated with vadose incision into the bottom of phreatic passages; half tubes are widely believed to originate from dissolution by surviving flow through a passage that has become choked with sediment (FORD & WILLIAMS, 1989). However, half-tubes are acknowledged as being controversial. It is suggested here that both features can be formed in artesian caves when natural buoyant convection currents are established within an unlooped and generally ascending bulk water flow, as described above, although other interpretations such as noted above, are not excluded Both types of feature are widespread in the gypsum caves of the Western Ukraine. In the most complete cases, these features can be traced through two, or even three, storeys of passages When a fissure receives recharge from below, the less dense aggressive water 162 Proceedings of the 12 111 International Congress of Speleology 1997, Switzerland Volume 1

PAGE 176

' CONFINING BED UPPER AQUIFER ________ ,,,, ~----~------------V~ / natural convection flow .,,,forced convection flow LOWER AQUIFER CONFINING BED Figure 2. Theformatio11 of upward so/utio11forms by buoya11t curre11ts. The diagram shows schematically the re/atio11ship betwee11 lilles of natural co11vection flow a11d forced flow on the mature stage of artesia11 speleogenesis, whe11 co11duit co11nectio11 has already bee11 established through the gypsum, but the forced flow is still slow due to the major constraillt of the upper aquitard. tends to occupy the uppermost available space and flows towards the nearest connection with the upper conduit. The rounded sections of keyhole passages are formed in this way (figure 2) The tube climbs the wall where the lower passage joins the upper one and continues as a ceiling half-tube in the upper passage. Half-tubes normally end in a domepit open to the next higher passage or at the bottom of the overlying aquifer. Though most convincingly displayed in the gypsum caves of the Western Ukraine, the suggested mechanism is likely to have broad applications. When continuous aggressive recharge takes place under artesian conditions circulation cells driven by the density gradients may develop at aquifer scale within salt deposits (ANDERSON AND KIRKLA D 1980) The density gradients that give rise to natural convection circulation can also be developed or reinforced by chemical mechanisms, by thermal differences or by suspended load injection Ascending flow due to thermal convection circulation is believed to contribute to the development of maze caves in the Black Hills (BAKALOWICZ ET AL. 1987) In particular natural convection accounts for the upward dissolution of the cupola-like ceiling forms typical of the highest parts of the multi-storey Wind Cave and Jewel Cave systems. Suspended load injected into the cave systems of the Western Ukraine due to occasional breakdown during the last stage of artesian speleogenesis is likely to have reinforced the density gradients between the bulk water in passages and buoyant ascending currents as described above 4. Conclusions 1. When continuous or periodic recharge of "fresh" water occurs rock dissolution sets up density gradients which cause gravitational separation (stratification) of water and drive natural convection circulation The phenomenon may operate at local scale or at aquifer scale, and may be expressed as looped circulation cells or unlooped directional currents Because dissolution always leads to density increase gravitational separation of water and natural convection due to this effect are inherently involved in and affect the process of, speleogenesis Their importance is relatively high in sluggish flow conditions (particularly in artesian settings) and relatively low to negligible in cases of high flow velocity The phenomenon is more pronounced and more important in gypsum and salt karsts, and in cases where fresh recharge occurs from below atural convection may also be caused by thermal anomalies. 2. Density stratification and natural con v ection circulation (currents) cause differential dissolution and related morphological effects in caves. The formation of flat ceilings cupolas and domepits ceiling half-tubes keyhole profiles inwardly inclined sidewalls and horizontal notches can be explained in this way 3 Natural convection circulation and dissolution by buoyant currents may not only modify passage morphology but may play an important role in building up the whole patterns of artesian cave systems during the early stages of ascending artesian speleogenesis. They cause upward stoping of dissolutional forms before the initial flow paths through a karstifiable unit are hydraulically interconnected and / or where they are highly resistant to through flow between the input and output boundaries (between the upper and lower aquifers) Symposium 7 : Phys i cal Speleology 163

PAGE 177

References ANDERSON, R Y & D W KIRKLAND. 1 980. Dissolution of salt deposits by brine density flow. Geology 8: 66-69. BAKALOWICZ, M.J. et. al. 1987 Thermal genesis of dissolution caves in the Black Hills, South Dakota. Bull. Geol.Soc. Amer. 99 : 729738 BIESE, W. 1931. Uber Hohlenbildung, I. Entstehung der Gipshohlen am sudlichen Harzrand und am Kyffhauser. Abh. Preuss Geol. Landesanstalt, Neue Fo l ge, v I 37 71 p C0RDINGLEY, N.J. 1991. Water stratification in active phreatic passages Cave Science 18 (3): I 59 CURL R.L. 1966. Cave conduit enlargement by natural convection. Cave Notes 8 (l): 4-8. FORD O C. & P W WILLIAMS. 1989 Karst Geomo,phology and H y drology Unwin Hyman, London. 601 p. FRUMKIN, A. l 994. Morphology and development of salt caves. NSS Bulletin 56 (2): 82-95. KEMPE, S. 1972 Cave genesis in gypsum with particular reference to underwater conditions Cave Science 49: 1-6. KEMPE, S ET AL. 1975 Facetten" and "Laugdecken", the typical morphologica l elements of caves developing m standing water. Ann Speleol 30 (4) : 705-708 KLIMCH0UK, A.B. 1992. Large gypsum caves in the Western Ukraine and their genesis. Cave Science I 9 ( 1 ): 3l l. KLIMCH0UK, A.B. 1994. Speleogenesis under confined conditions, with recharge from adjacent formations Pub!. Serv Geol. Luxembourg v.XXVII. Comptes Rendus du Coll. lnt. de Karstol. a Luxembourg: 85-95 KLIMCH0UK, A. In press. Artesian speleogenetic setting. This volume. KLIMCH0UK & AKSEM, 1988 Regime study of gypsum karst activity in the Western Ukraine. Inst. of Geol. Sci ., Kiev : 55 p (Russ., res. Engl.). KLIMCH0UK, A.8., V N.ANDREJCH0UK, & 1.1.TURCHIN0V I 995 The structural prerequisites of speleogenesis in gypsum in the Western Ukraine Ukr.Speleol.Assoc., Kiev : 106 p. LUKIN V S. 1967. On the origin of inclined platforms in karst cavities Zemlevedenie (Earth Research), v.VII. Moscow Univ., Moscow: 212-214. (Russ., res Engl.) PFEIFFER, D. & J. HAHN 1972. Karst of Germany In : (M Herak & V.T. Stringfield, eds.) : Karst important karst regions of the northern hemisphere. Elsevier Amsterdam : 189-223. 1 64 Proceedings of the 12 m International Congress of Speleology, 1997, Switzerland Volume 1

PAGE 178

Le role de la zone epinoyee dans la speleogenese par Phili pp e AUDRA Groupe de valorisation de l'environnement (GVE), URA D1476 du CNRS, Universite de Nice-Sophia-Antipolis, 98 boulevard Edouard Herriot, BP 209, 06204 Nice Cedex. & Cagep (URA 903), Aix-en-Provence Resume Lors des crues, le niveau de la zone noyee est susceptible de presenter d 'i mportantes variations. Des exemples actuels montrent que l'ampleur verticale de la zone epinoyee peut atteindre plusieurs centaines de metres. C'est dans la zone epinoyee que se cumulent les processus mecaniques et chimiques. La concentration de conduits de grande taille en est la marque Ces processus permettent notamment de comprendre l'origine de formes qui ne pouvaient s'expliquer par un creusement noye, comme les tubes inclines sur de grandes denivellation s. On insiste sur la pertinence de la conception du drainage de la zone noyee, par des conduit localises dans la zone e pinoyee Les grands karsts noyes (Vaucluse ... ) pourraient s'ex pliquer par un noyage posterieur a leur genese. Abstract : the ep i phreatic zone p art in t h e s p e l eoge n es i s. During tloodings, the phreatic zone can undergo important level variations. Some contemporary examples show that the vertical extension of the epiphreatic zone can reach several hundred of metres Mechanical and chemical processes cumulate in this zone, marked by a concentration of large galleries. These processes allow to understand the origin of some forms which could not be explained by a phreatic evolution, such as large sloped tubes. The phreatic zone drainage by conduits located in the epiphreatic zone seems to be the mo st adjusted conception. Large phreatic karsts (Vaucluse ... ) could be exp l ained by water-table rises occuring after their genesis. 1. L'ampleur des mises en charge dans la zone epinoyee La mise en charge des parties basses de certains reseaux karstiques est un foit conn u de longue date L'exemple le plus celebre est celui de la grotte de la Luire (Drome, France) ou l'eau remonte de 450 m avant de "c rever" par le porche Grace a l'avancee des connaissances liees aux explorations speleologiques (LI MONDE & al.}, le exemples de mises en charge sur plusieurs dizaine de metres de hauteur sont innombrables, ceux depassant I 00 m ne sont pas rares (fig. I). Cavite M i se en c h aree Grotte de la Luire (Drome) 451 m Perte du Calavon (Alpes Hte. Pro v.) 263 m Puits des Bans (Hautes-Alpes) 217m Rescau Fanges-Paradet (Pyr. orient.) 170 m Holloch (Schwvtz) 170 m Re seau de s Siebenhengste (Bern) 150m Reseau du Revest (Alpes maritimes) 145 m Trou Qui Souffle (!sere) 120 m Trou du Renard (Pvrenees atlantiQues) 120 m Aven Souffleur (Vaucluse) 120 m Fontaine de Crevecreur (Htes-Alpes) 100 m Figure 1 : q u e l ques exe m p l es d e mi se e 11 c ha rge de p assa 11 t J OO III de hauteur. Cette portion du karst soumise a des noyages et denoyages successifs en fonction des variations de niveau de la zone noyee est appelee "z one epinoyee". 2. Des origines climatiques, structurales et speleogenetiques Ces variations de niveau de la zone epinoyee sont liees a l'arrivee brutale d'importants volumes d'eau, depassant la capacite d'evacuation des galeries. Ces conduits de taille insuffisante fonctionnent comme un diaphragme Ils provoquent une mise en charge a l'amont de !'obstacle. On observe alors une remontee dans les conduits verticaux qui fonctionnent en cheminee d'equilibre et des circulations noyees dans l es conduits horizontaux q u i devienne n t temporairement actifs. De tels phe n omenes sont lies d'une part a un contexte climatique fovorisant des ecoulements tres contrastes et d'autre part a !'existence de secteurs de foible transmissivite a l'aval des reseaux Les types de climats particuliereme n t fovorables sont ceux ou l'ecart entre les ecoulements de crue et d'etiage est important. C'est le cas des regions de montagne ou une grande partie de l'ecoulement se produit durant la breve periode de fonte des neiges Les regions me d iterra n eennes e t tropicales sont egalement connues pour la bruta l ite et la concentration des epi odes pluvieux Neanmoins, des crues brutales sont possibles dans tous les types de zones climatiques, car des moyennes tres ponderees peuvent masquer des ep i sodes exceptionnels. La prese n ce de zones de conduits a foible permeabilite a l'aval des reseaux karstiques peut avoir plusieurs origines qui peuvent d'ailleurs se combiner (fig. 2) : qua n d l es couches ca l caires plangent a l'aval en-dessous du niveau de base loca l co n stitue de roc h es impermeables on a affaire a un karst barre Parfois, le barrage est constitue de roches peu permeables ou legerement fissurees (calcaires mameux dolomies gres ... ), ou des circulations aquiferes existent cependant (fig 2-A). II s agit dans ce cas de karst "semi-barre", d'origine structurale (AUDRA, 1994) quand on observe un etagement des conduits lie a un enfoncement saccade du niveau de base bien souvent le reseau inferieur le p l us recent ou s'effectuent les circulations perennes est peu evolue avec d es conduits d e petite taille parfois impenetrab l es (fig 2-B) L'origine est ici speleogenetique. Dans l es deux cas, le karst adopte deux types de comportements hydrody n amiques en fonction de son etat de charge A l'etiage, Jes eco ul ements transitent au travers des fissures et des conduits d e foible debit. L'eau ressort au contact du niveau de base local, par une source generalement impenetrable. En crue l'eau excedentaire ne peut franchir Jes secteurs peu transmissifs et enva h it la zone epinoyee. Elle franchit !'obstacle par des conduits superieurs et sort au niveau d'un trop-plein perche au-dessus du niveau de base dont les galeries sont generalement vastes et penetrables Sympos i um 7 : Physical Speleology 165

PAGE 179

@ Eliagc l Dmins trcs pcu lr.lllsmissifs nivcau de lxtsc actucl Figure 2 : zones epi11oyees liees a 1 111 karst semi-barre (A) O U a 1111e adaptation recente a l' abaisse 111 e11 t du 11iveau de base (BJ. Du point de vue morphologique, les deux types de conduits s'opposent radicalement. Les reseaux inferieurs sont constitues de fissures a peine elargies, bien souvent impenetrables et noyees. Les reseaux superieurs localises dans la zone epinoyee presentent quanta eux des puits et des galeries horizontales ou inclinees de belle taille en forme de tubes Ces reseaux sont frequemment tapisses d'argiles de decantation. abandonnees durant la phase de decrue. 3. Consequences sur le fonctionnement de la zone epinoyee et son role speleogenetique Tout d'abord, cette zone est soumise a des variations de pression hydrostatique considerables dont les effets mecaniques sont incontestables. ainsi qu'il des circulations temporaires extremement abondantes. La corrosion s'effectue non seulement pendant la phase de crue ou circulent des eaux agressives largement renouvelees, mais elle se poursuit en plus apres le denoyage, grace au film d'eau recouvrant les parois, au contact avec une atmosphere relativement riche en gaz carbonique. C'est le "creusement dans la zone inondable", particulierement efficace (CHOPPY, 1994). L'exceptionnclle concentration de conduits de grande tail le dans la zone epinoyee en est la traduction. Tous ces reseaux presentent des sections tubulaires organisees par les crues en labyrinthe anastomoses (PALMER, 1991 ) Ceux-ci sont disposes selon un plan horizontal, soil pour des causes structurales (utilisation des joints de stratification plats), soil au toil de la zone noyee, auquel cas les conduits sont indifferents a la trame structurale. II s'agit des water table caves des anglo-saxons (FORD & WILLIAMS, 1989). Les conduits peuvent egalement s'agencer selon les discontinuites structurales en toboggans fortement inclines, relies par des puits verticaux ou des galeries horizontales, leur section etant toujours tubulaire. Lis forment ainsi de grandes boucles en montagnes russes dont l'ampleur depend de la hauteur de la zone epinoyee et peut atteindre jusqu'il 250 m (fig 3); ce sont les loop caves 1 Coupe schtmatique ,, o ,~ ,"~~ / / \ Figure 3 : l e reseau de la Ta 11 ta/l,ij l, /e (Sa/zbourg, Autriche) prese11te des tubes e 11 m o11tag 11 es r u sses do11t /es boucles atte i g11a 11 t p l usieurs ce11tai11es de metres d'ampleur correspo11de11t a u11e a11cie1111e zone epi11oyee (topographie d'apres COURBON & CHABERT, 1986). Ainsi peut s'expliquer l'origine de ces tubes inclines sur plusieurs centaines de metres de denivellation, caracteristique de la base des reseaux de montagne Ils ne peuvent en effet qu'etre lies aux mises en charge de la zone epinoyee durant les crues et ne peuvent en aucun cas s'accorder avec une zone noyee dont le niveau serait stable et largement perche au-dessus du niveau de base, celle-ci ne pouvant se maintenir sans se vidanger rapidement a l'exterieur. Ces phenomenes de mise en charge impliquent done !'existence de circulations temporairement noyees bien au-dessus du niveau de base !ors des periodes de crues. Autrement dit, ii est necessaire d'etre prudent, surtout dans les karsts de montagne, lorsque l'on considere, comme c'est frequemment le cas, d'anciens niveaux de galeries en tube comme etant des marqueurs altitudinaux de niveaux de base passes Ils peuvent etre en effet consideres comme tels, a la precaution pres que le niveau de base pouvait se situer en-dessous des galeries. et parfois plusieurs centaines de metres plus bas. 4. Vers une remise en cause du creusement noye profond ? II est manifeste qu'il est difficile de theoriser ur la speleogenese profonde, sachant d'une part que l'on ne possede guere de donnees verifiables sur les phases initiales de karstification et en particulier sur la genese des proto-conduits et d'autre part que la connaissance des zones noyees, bien qu'ayant recemment Fait d'immenses avancees grace a la speleo-plongee, reste encore bien insuffisante Neanmoins, certains fails d'observation s'integrent mal a la theorie du creusement noye profond et impliquent, sinon une revision complete, tout au moins certaines remises en cause Depuis DAVIS ( 1931) et BRETZ ( 1942). cette theorie admettait qu'en !'absence d'ecran impermeable a la base de l'aquifere (autrement dit dans le cas des karsts barres), que les circulations s'enfonc;aient profondement dans la zone noyee, pour remonter a la surface au contact du mur impermeable sous la forme d'une emergence vauclusienne (fig. 4-A) (on exclut le cas particulier des aquiferes artesiens ou les circulations sont contraintes par la structure a un cheminement profond) Ainsi expliquait-t-on l'origine des karsts profondement noyes, draines par des 1 FORD attribue leur genese a un creusement toujours noye Selon cet auteur, la difference entre les water table caves et Jes loop caves serait liee il une fracturation decroissante A notre avis, la differenciation provient en Fait des variations du gradient hydraulique et de l'ampleur de la zone epinoyee (voir aussi CHOPPY, 1994) 166 Proceedings of the 12 1 h International Congress of Speleology, 1997, Switzerland Volume 1

PAGE 180

emergences remontantes dont la fontaine de Vaucluse est l'exemple le plus marquant. Paradoxalement, on utilisait simultanement les anciens reseaux horizontaux comme indicateurs de positions de niveaux de base passes, sous entendant implicitement que les circulations suivaient le toit de la zone noyee sans s'enfoncer en profondeur Or, ii est desormais admis que la structuration du drainage karstique hierarchise clans l e milieu calcaire heterogi:ne s'effectue sous le contr61e du gradient hydraulique, qui tend ii generer des drains ou le s pertes de charge sont minimales (SWINNERTON, I 932 ; PALMER, 1991 ). Autrement dit, le "chemin de drainage" (CHOPPY, 1994) tend ii se rapprocher de la ligne droite, en direction du point d'emergence. Ainsi, clans la zone vadose, Jes circulations s'enfoncent verticalement sous ]'influence d e la seule gravite. Arrivees au niveau de la zone noyee, elles prennent alors la direction de ]'emergence selon une pente faible. La ligne droite etant le plus court chemin, le conduit principal suivra le toit de la zone noyee (water-table cave), ou se developpera dans l a zone epinoyee, lor qu'elle existe, puisqu'elle offre la moindre resistance aux circulations, de la faible !'importance des cavites (fig. 4-8). L'utilisation des discontinuites structurales favorise ]'apparition de tubes en montagnes russes (loop cave). D ans ce cas, les circulations noyees profondes sont negligeables et ne peuvent donner naissance ii de conduits suffisamment importants pour concurrencer ceux de la zone epinoyee. Les re eaux noyes profonds, pour le quels ii est necessaire de foumir une explication coherente, auraient clans ce cas une origine distincte (on ne prend pas ici en consideration les karsts hydrothermaux dont l'energie de provenance endogi:ne suffit ii expliquer leur existence). Cette origine pourrait tri:s bien se trouver dans leur geni:se. Ces karsti fications profondes se seraient developpee lors de periodes d'abaissement du niveau de base, Figure 4 : disposition des conduits de drainage de la zone noyee. A) mode/e critique, avec 1111 cre11se111e11t ,wye profond et ,me emerge11ce va11cl11sie1111e. BJ mode/e avec 1111 creuse111e11t au toit de la w11e noyee, da11s la wne epi11oyee, avec des conduits en 111ontag11e russes. Les circulatio11s 11oyees profo11des so11t margina/es puis elles auraient ete noyees suite ii une remontee du niveau de base. Les reseaux crees ii grande profondeur, puis noyes, seraient toujours fonctionnels, car ils offrent les meilleures facilites de circulation du fait de leur anteriorite. De telles oscillations du niveau de base se sont effectivement produites durant le Quatemaire sous l'effet des glaciations, ainsi qu'au Messinien lors de l'assi:chement partiel de la Mediterranee. Les karsts littoraux, ainsi que ceux situes ii proximite des fleuves ayant profondement surcreuse leur lit parfois tri:s en amont sur plusieurs centaines de kilometres durant J es episodes regre sifs, ont connu une telle evolution. Des concretions actuellement noyees ii plus de 100 m de profondeur attestent de la realite de eel ennoyage (goul de la Tannerie au bard du Rhone en Ardi:che ANDRES & LISMONDE, 1995). Ce modi:le suggi:re ]'abandon d'anciennes theories, en minorant l'ampleur de la karstification noyee profonde, au profit d'un drainage localise au toit de la zone noyee et en particulier dans la zone epinoyee, car ii s'accorde mieux avec Jes observations accumulees par les progri:s recents des explorations. II constitue un cadre de reflexion susceptible de mieux comprendre la realite de ces phenomi:nes majeur qui conditionnent toute ]'organisation des reseaux que nous etudions. References A DRES, D. & LISMONDE, 8. 1995. Bertrand Leger, spe/eonaute Groupe speleo de La Tranche & Comite departemental de speleo logie Grenoble, 126 p. AUDRA, Ph. 1994 Karsts a/pins. Genese de grands reseaux souterrains. Exemples: le Tennengebirge (Autriche), / fie de Cremieu, la Chartreuse et le Vercors (France) Karstologia Memoires n 5 These ii l'Universite J Fourier Grenoble I. Federation francaise de speleologie, Paris & Association francaise de karstologie, Bordeaux. 280 p BRETZ, J. H. 1942. Vadose and phreatic features of limestone caverns. The Journal of geology. 50, 6 : 675-811. CHOPPY, J. 1994. La premii:re karsti fication ; Syntheses speleologiques et karstiques. Choppy, Paris, 72 p. COURBO P. & CHABERT, Cl. 1986. Atlas des grandes cavites mondiales. Federation francaise de speleologie, Paris & Union intemationale de speleologie : 255 p DAVIS, W. M. 193 l. L'origine des cavernes calcaires. Science. 73, 327-331. FORD, D. & WILLIAMS, P 1989 : Karst geomorphology and hydrology. Unwin Hyman, Landres : 60 I p. LISMO DE, B. & LES SPELEOS GRENOBLOIS DU CAF, 1995. Lacrue du 18 mai 1994 au Trou qui Souffle (Vercors). Karstologia, 25 Federation francaise de speleologie, Paris & Association francaise de karstologie, Bordeaux, p. 1-12. PALMER, A. N. 1991. Origin and morphology of limestone caves. Geological society of American bulletin. I 03 : 1-25 SWIN ERTO A C. 1932. Origin of lime tone caverns. Bulletin of the geological society of America 43 : 663-694 Symposium 7: Physical Speleology 167

PAGE 181

Reconstruction of paleocurrents in caves of the Bystra Valley (Tatra Mountains Poland), on the basis of scallops and deposits analyses Ditta Kicinska Hydroconsult, ul. Ratajcaka 10 / 12, 61 815 Poznan Insitute of Geology, Adam Mickiewicz University, ul. Mak6w Polnych 16, 61 686 Poznan, Poland Abstract Tatra Mountains are the highest and northernmost massif in the West Carpathian system (fig. I). The central and southern parts of the Tatra Mountains are built of Paleozoic crystalline rocks. The northern slope of the massif consists of Mesozoic sedimentary rocks Among them several overthrusted tectonic units are distinguished (which are again distinguished into shallower Hightatric Succesions and deeper Subtatric Succesions). The Bystra Valley is the easternmost valley of the West Tatra Mountains. The karst catchment area of the Bystra Stream is considerably greater than the surface catchment of the Bystra Valley. It resulted from underground water flows which created cave systems. The underground flows of big karst springs in the valley are supposed to come from the area lying east of the valley built up of granitoids rocks and from the west of the valley where sedimentary rocks dominate. The Southern border of the valley is built up of metamorphic rocks injected by granitoids and overthrusted over a sedimentary sequence. The analyses of paleocurrent directions were made in bigger horizontally developed caves (Magurska 1200 m of length, Goryczkowa 312 m, Kalacka 360 m) lying 50-150 mover the present valley bottom. The analyses of scallops asymetry and deposits (heavy minerals roundness grades, grain size) in these caves were made in order to reconstruct the direction of paleocurrents. Variability of the mineral-petrographic composition and observations of scallops enable to distinguish two opposite directions of underground flows in these caves (fig 2 and 3). The development of underground flows is much more complicated than it was supposed so far. The analyses of paleocurrents and isotopic dating of speleothems with U/Th method showed that the earlier directions are older than 210 ka (+I 6 / I 4 ka) and the younger ones are older than 89 ka ( + 10 / -9 ka). This work was supported by grant KBN 0888/P2 / 94 06. . Figure 2 Bysrra Valley Tatra Mountains :agurs~ Cav~ j Kalacka Cave Goryczkowa Cave tlow > direction older younger Figu r e 3 ridge Magurska Cave Goryczkowa Cave Kalacka Cave large springs permament streams intermittent streams underground flows 168 Proceedings of the 12 1 h International Congress of Speleology, 1997, Sw i tzerland Volume 1 0 !OOOm [JI] Subtatric Succesions C=1 Hightatric Succesions metamorphic rocks and overthrusted over F:::::::i sedimentary sequence t::==::::l crystalline rocks

PAGE 182

Hypothese sur la genese d un siphon par J. Fo l tete Diren de Franche-Comte, 5 Rue du General Sarrail 25014 Besan9on Abstract Mechanical assumptions on waterflows in a limestone massif try to explain the genesis of siphons in the underground networks. This paper trends to make them complete when groundig them on tectonical and hydraulical data. Res u me ~es h y potheses mecaniqu~s sur Ies eco_ule'.11ents dans Ies massifs calcaires contribuent ii expliquer la genese des siphons dans les reseaux souterrams. La presente note v1se a les completer en s'a ppuyant sur les donnees de la tectonique et de I'hydraulique. Introduction De nombreuse rivieres souterraines voient leur cours emprunter un trace en siphon juste ii l 'am ont de Ieur emergence et cela dans une zone ou precisement existent souvent des galeries s uperposees, nombreuses, plus ou moins anastomosees et colmatees, temoin s de circulations anciennes. La presence d'un siphon est, en speleologie, un phenomene s uffisamment general pour que !'on se pose la question de sa genese : Pourquoi en effet l'eau des massifs calcaires ne descend elle pa s tranquillement, ou par chutes successives. jusqu'au niveau de !'emergence pourquoi !'existence de cette section en conduite forcee juste avant la sortie ii l 'a ir libre ? Les explications mecaniques du phenomene, s'appuyant sur Ies bilans energetiques des ecoulements, ne repondent que partiellement aux questions posees; d'autres explications se trouvent dans la connaissance de la structure du massif mais cela revient a l'etude d'un cas d'espece ii partir de releves geologiques et la transposition a d'autres systemes n'est pas generalisable. Une hypothese ancienne P Petrequin s'est beaucoup interroge sur cette question notamment ii !'occasion de ses sorties et publications sur le Vemeau *; ii avance l'hypothese de la detente des calcaires en bordure des vallees. Tres succinctement, on peut resumer cette hypothese comme suit : Parallelement ii la vallee la bordure du plateau est affectee par des "plans de detente :'du calcaire qui entrainent par "appel au vide" une fissuration transversale au reseau karstique; ces plans affectent le plateau sur une frange de quelques dizaines a quelques centaines de metres On peut concevoir en effet que dans une vallee ou !'erosion a fait disparaitre une grande epaisseur de sediment, des reajustements isostatiques ou lies ii la decompression du sediment, entrainent des mouvements qui sont a l'origine des ruptures en bordures ; l'interface alors se traduit par un plan vertical, parallele ii la vallee, limitant deux domaines; ce plan correspond ii une surface de "decohesion" de "microfissuration" de la roche et peut etre multiple comme cela apparait dan s ce schema que l'on doit ii Petrequin Chacun a pu ve rifier la justesse de ce schema au cours d'expeditions speleologiques ou de releve de fissuration. Cette hypothese est compatible avec l'idee d'un effet qui se propage ii partir du bord de la vallee vers l'interieur du plateau au fur et ii mesure de l'elargissement de celle-ci. Dans ce cas les panneaux effondres, en bordure de falaise, ne sont que des reajustements d'equilibre dans la morphologie actuelle. HYPOTHESE datente des calcalres en bard de vallae et appal au vide Autre hypothese P Petrequin 1980 Soil un plateau tel que l'a imagine P Petrequin Iimite par une vallee; en bordure de celle-ci et parallelement ii son axe de developpement la detente des calcaires a occasionne de s plans de microfissuration ; dans ces plans verticaux la porosite est plus importante que dans le reste du massif entrainant une meilleure permeabilite verticale. Si en outre quelque accident tectonique vertical, incident sur la va llee recoupe l es plans precedents on remarquera que les conditions sont reunies pour la constitution d'un axe preferentiel de fissuration a l'intersection des deux plan s verticaux, veritable "puits potentiel" a l'interieur du massif. En cas de developpement d'un karst ii partir de ]'accident tectonique precedent le "puits potentiel" est appele ii jouer un role particulier dans !'apparition du siphon. L'eau de pluie tombant sur le ma ss if s' infiltre lentement dans les microfissures sous l 'effet combine de la pesanteur et des forces capillaires opposees. Si ce cheminement croise un segment plus permeable -joint de stratification, plan de fissurationet que ce segment conduise a une zone de libre ecoulement, l'eau empruntera ce parcours des que la pres s ion dans le massif sera suffisante, soit au moment de !'averse. Au voisinage du niveau de base, la presence permanente de l'e a u ameliore la dissolution par augmentation du temps de contact Symposium 7: Phys ical Speleology 169

PAGE 183

et le processus de creusement d'abri sous roche ou de vasque, embryon de reseau, pourra s'enclencher. A l'interieur du massif, dans un "puits potentiel" la karstification se developpe avec la meme logique de descente vers le niveau de base. En fond de puits, la ou la dissolution est la plus importante, !'absence de soutirage par le fond peut ralentir !'evacuation des produits de dissolution qui se fera neanmoins par diffusion dans la colonne d'eau puis renouvellement !ors de la montee de !'averse. Si, au cours de sa descente, la karstification rencontre un plan horizontal plus fissure pourront se creer des conditions favorables a !'utilisation du cheminement hydraulique le plus econome en perte de charge (voir schema). L'agrandissement du schema illustre les conditions dans le sque lles le "cheminement hydraulique minimum" pourra entrer en fonctionnement et etre a l'origine d'un ecoulement en siphon. Soit un puits P a penneabilite verticale infinie par rapport a la permeabilite horizontale au fond duquel s'est developpe un element de conduit karstique AB. En eau basse la charge hydraulique en A et en B est sensiblement la meme puisque la ligne piezomi:trique est tri:s proche de l'horizontale ; dans ces condition ii ne peut y avoir de mouvement la charge BB' consideree comme force motrice est equilibree par la charge AA' consideree comme la charge s'opposant au mouvement. En Periode de crue, les ecoulements deviennent turbulents dans les conduits situes au dessus du niveau statique et la ligne de charge devient OA"B". La charge en B s'est accrue considerablement B' etant devenu B" alors que !'augmentation en A est proportionnellement plus foible A' devenant A". Dans ce systi:me ii y a desequilibre des pressions, la charge en B qui se transmet immediatement en A n'est plus compensee par la charge s'opposant au deplacement AA"; notons que cette surpression en A est liee a la longueur du tronc;:on AB et a la tangente de la ligne de charge, laquelle depend des differences de perte s de charge liees aux ecoulements entre le puits P et l'exutoire d'une part et a l'amont du puits cote massif calcaire. Plus le niveau de crue est eleve dans le massif, plus cette surpression est importante et pour un niveau de rupture suffisant les ecoulements se feront dans les microfissures entre A et 0 Ainsi les conditions pour la creation d'un siphon sont: L'existence de puits verticaux lies a !'intersection de deux plans de microfissuration pennettant une circulation de haut en bas plus facile dans le massif; Une dissolution maximale sous le niveau de base par augmentation du temps de contact et une evacuation des produits de dissolution a la montee des eaux; L'existence de zone a meilleure permeabilite horizontale joint de stratification par exemple, pennettant par leur karstification, la creation de segments economes en perte de charge. puits : P I 1 I Dans cette hypothi:se, ii est admis que la premii:re fissure horizontale rencontree sous le niveau de base sera kar tifiee et se developpera en raison de la presence pennanen e d'eau, limitant d'autres developpements a un niveau inferieu et que !'extension vers l'interieur du massif est limitee par I absence de plan de decompression des que l'on se situe au d la de la frange de terrain affecte par la detente du calcaire. R efere n ces AUCANT, Y., C. SCHMITT & J.P.URLACHER. 1985. Le Vemeau souterrain, Ed. SHAG 170 Proceedings of the 12 111 International Congress of Speleology 1997, Switzerland Volume 1

PAGE 184

Hydrologie du systeme karstique du Rupt du Puits (Lorraine/Champagne, France) : Fonctionnement du siphon aval par Stephane Jaillet, Laboratoire de Geographie Physique Appliquee Universite Michel de Montaigne Bordeaux 3 33 405 Talence France avec !'aide du Laboratoire de Geographie Physique de l'Universite de Metz Ile du Saulcy 57 045 Metz Cedex 1 France Abstract Under the lowland plateaus located at the border line between Champagne and Lorraine we find a covered karst called the "Rupt du Puits", wich is an affluent of the river Saulx The subterranean main river arm alternately flows under and overground Downstream, a 450 m siphon runs into two points of emergences : a permanent one saturing at a flow of about 300 to 400 1/s and a temporary one getting active when the permanent one reaches a flow of 200 1/s. Globally the siphon's feeding pattern is complex. Infiltrating water via the main river arm feeds the springs though it does not contribute by more than two thirds to their flow. When reflecting on the origins of the hydrous complement we are led to believe that it might be provided by other phenomena, in particular by the Saulx itself. Both the morphological aspects and the varuous measuring campaigns carried through seem to be indicative of a karstic type water circulation beaneath base level adding to the downstream water flow. Key-words : karst system siphon water bearing bed/river communication emergence Rupt du Puits Barrois Lorraine. Resume Au contact de la Champagne et de la Lorraine se developpe un karst couvert de bas plateau : le Rupt du Puits affluent de la Saulx Le collecteur souterrain presente une alternance de zones tantot noyees tantot a surface libre. Le siphon a v al long de 450 m debouche sur deux emergences : l'une perenne saturant vers 300 a 400 1/s et l'autre temporaire se declenchant quand la perenne est a 200 1 / s. Son alimentation est complexe. L'infiltration rejoint via le collecteur la zone des emergences mais ne participe qu'auxx 2 / 3 de leur debit. La recherche d'un complement hydrique pousse a supposer une alimentation par d'autres drains et par la Saulx elle-meme Les donnees morphologiques et Jes differentes campagnes de mesures tendent a mettre en evidence une circulation karstique fonctionnant sous niveau de base et participant a l'alimentation du siphon aval. Mots-cles : systeme karstique siphon, echange nappe / riviere emergence Rupt du Puits Barrois, Lorraine 1. Un karst couvert de Bas-Plateau Un pays de contact Entre Lorraine et Champagne dans la partie est du Bassin Parisien se developpe le karst couvert du Rupt du Puits Plus important systeme karstique souterrain connu du nord de la France c'est aujourd'hui plus de 21 km de conduits qui ont ete reconnus Le plus important reseau de ce systeme developpe a Jui seul 11 800 m et fut reconnu en grande partie en exploration post-siphon au debut des annees 70 avant que ne soit perce un forage d'acces de 47 m qui permet aujourd'hui de prendre pied directement dans la riviere souterraine Cette unite hydrogeologique (mise en evidence par de nombreux tra9ages) assure en foret de Trois Fontaines la transition entre deux ensembles morphostructuraux A !'Quest c'est le Perthois et le Vallage ou Jes depots argilo-sableux du Cretace determinent un paysage de plaine a topographie molle Ils recouvrent en discordance vers Est Jes plateaux calcaires du Barrois. Armes par Jes depots carbonates du Portlandien ces derniers offrent un paysage plus sec ou Jes rivieres (la Saulx et l'Ornain) ont incise des vallees sur une centaine de metres de profondeur (Fig 1) C'est dans ce contexte que se developpe done le collecteur souterrain du Rupt du Puits Le drain principal long de 7 km est connu sur plus de 80 % de son parcours. C'est une succession de zones noyees et de zone a surface libre De l'amont vers l'aval la Donna, la Beva et la riviere souterraine du Rupt de Puits zone trois secteurs ou l'ecoulement rapide exonde, tranche avec Jes zones noyees qui Jes separent (Fig 2) Le dernier siphon le plus a l'aval presente une diffluence Une premiere branche (la plus au nord) rejoint !'emergence perenne du Rupt de Frainiau. La seconde (la plus au sud) file vers la vasque temporaire du Rupt du Puits. Cette derniere legerement perchee fait office de trop plein de la premiere. C'est cette vasque qui, plongee dans les annees 60 70 permit la decouverte des reseaux donnant son nom a tout le systeme. Nous ne nous interesserons dans le present article qu'au fonctionnement d'une de ces zones noyees : le siphon aval. Figure 1 : Sihwtion du knrst du Rupt du Puits au contact du Barrois calcaire et du Perthois argilo-sableux Sympos i um 7 : Ph y s i ca l Speleo l ogy 171

PAGE 185

\ I Cl i\ :ii C I ~ ' 0 \ I i i/ / / i i I i I' i ) /' ~ C,~lc:c1c11rC,nndC f .. .... ollcc1cur nnyc! AffiucnlS ~xomJ~.$ \ c: r---=--i scnsdcs ~.., 1 L.::....J ccoulc mcn L.; o;> r-:=-,Tr JJ<;:l'5U~ L..::......J Jo; 6.nufCmcn 1.; r::--::::l Tr~~c ou,111[ .,. UU KUpl (JU r"Ull'f i I/'. I Limitc du bo.ui n \. cJ'a11mcn1.;111on ,I 5.JAILLET 1996. Fig1t r e 2 : Syntltese des resea1tx, des trafages et des Li mites d1t bas sin d'ali111e11tation d1t Rupt du P1tits Le si phon aval Les premieres plongees dans le siphon aval du Rupt du Puits sont l'oeuvre de Bertrand LEGER Apres la lecture d'un article de DESCAVES publie en 1962 dans Spelunca il se rend a la fin des annees 60 a Robert Espagne dans la Meuse puis au bord de la vasque et penetre dans la galerie noyee. Plusieurs plongees seront necessaires pour le franchir et c'est finalement une equipe de jeunes meusiens qui le 16 novembre 1971 passe le siphon (LUCION, in JAILLET, 1994). II a ete decrit comme suit. C'est une galerie, d abord etroite et e b ouleuse qui adopte ensuite une section plus reguliere (3 a 4 m de large pour 2 a 3 m de haut). Trois cloches franchies a 97, 188 et 256 m et l ecoulement jusqu'alors temporaire devient perenne. A cet endroit, la fameuse diffluence permet aux ecoulements de filer au Nord vers le Rupt de Frainiau Les tentatives de plongees vers cette branche n'ont rien donne. Un fort remplissage occupe le fond de la galerie qui devenant de plus en plus basse devient infranchissable Cote Rupt de Frainiau (l'emergence perenne), les tentatives de plongees n'ont de meme rien donne. Un amas de blocs barre effectivement l'acces De la vasque du Rupt du Puits au collecteur souterrain c'est 445 m de siphon qui ont done ete franchit. C'est a l'epoque un veritable exploit et le Rupt du Puits restera un temps le plus long reseau au monde explore derriere siphon. Depuis que le forage d'acces au collecteur du Rupt du Puits a ete creuse en 1975 il semble que plus aucun plongeurs n'ait franchi le siphon (Fig. 3). Nous avons en projet de faire revisiter ce siphon pour en obtenir un leve morphologique le plus precis possible. En attendant, c'est comme une boite noire que nous avons decide d'aborder le fonctionnement de ce sous systeme du systeme karstique du Rupt du Puits C'est a dire en analysant ce qui y entre et ce qui en sort Pour cela plusieurs mesures de debits conductivites temperatures et quelques analyses physico-chimiques complementaires ont ete mises en place (DEVOS et al ., 1996). Ces mesures se repartissent dans l'annee de fa9on heterogene, mais touchent differentes situations hydrologiques. 2. Dix campagnes de mesures de debits sur le siphon aval du Rupt du Puits A dix reprises, nous avons pu mener des campagnes de mesures sur le siphon aval du Rupt du Puits. Chacun de ces points de jaugeage a ete mesure a la perche a integration de mesure de debits, tant sous terre qu'aux emergences. Trois points significatifs ont ete retenus po u r caracteriser le fonctionnement du systeme siphon aval. L'entree dans le systeme a ete mesure dans le collecteur souterrain au point Rl3 Les sorties du systeme sont Jes deux emergences : Rl4 la vasque temporaire du Rupt du Puits et Rl5, !'emergence perenne du Rupt de Frainiau. Les valeurs obtenues lors de ces dix campagnes ont ete rassemblees dans le tableau (Fig 4) Rupt de Frnlnlau Colltt1eur routerr..ln 500m Figure 3 : Bloc diag r amme d1t site du reseau d1t Rupt du Puits. 1 72 Proceedings of the 12 th Internat i onal Congress of Speleology, 1997, Switzerland Volume 1

PAGE 186

F' 12ure 3 BI d. : oc 1a2ramme u site u reseau d d u upt u u1ts. d R d P. Campagnes Collecteur Vasque RdP R. Frainiau Total Apport % par rapport situation de mesures R 13 enl/s R 14 en 1/s R 15 en 1/s Emenzences siphon (1/s)) au collecteur hvdrolo2iaue 22-Fev-95 139 0 39 232 232 39 93 39 67 19 Haute s eaux 23-Fev-95 224 72 272 344 120 53 57 Haut e s e aux 24-Fev-95 1029 1271 451 1722 693 67 35 Crue hi v ernale 9-Dec-95 19 0 15 15 -4 -21,05 Etiage ( R l5 douteu x) 25-Fev-96 316 171 219 390 74 23,42 Fonte ni v al e ( petit e cru e) 26-Fev-96 200 2 2 224 226,2 26, 2 13 10 Fonte ni v ale (decru e) 21-Mar-96 26 0 52 5 2 26 100,00 E tiage hivern a l 7-Avr-96 22 0 38 38 16 72 73 Etiage hivernal 7-Dec-96 186 99 185 284 98 52 69 D e crue 8-Dec-96 167 20 141 161 6 -3 59 D ec rue Figure 4 : Debits (en 1/s) en R13 (collecteur a l'amont du siphon) et aux deux emergences (RU et R15) et apport calcule du siphon a partir de ces debits, /ors de dix campagnes. 11 apparaH que dans tout les cas (sauf un) le debit awe emergences est superieur a celui du collecteur souterrain Le rapport moyen entre l entree dans le systeme (le collecteur R13) et la sortie du systeme (les emergences R14 + R15) est de 1 6 Selon la situation hydrologique l importance relative des dewc emergences varie, exprimant la tres grande variabilite des debits que peut supporter l'emergence temporaire (ici O a 1271 1/s) a la difference de l'emergence perenne (15 a 451 1/s) (JAILLET 1996) Le cas du 9 decembre 1995 est interessant a plus d un titre C est la campagne d etiage la plus basse. Sous terre le debit est de 19 1/s. L'emergence temporaire ne coule pas tandis qu'au Rupt de Frainiau (l'emergence perenne) il coule 15 1/s C'est le seul cas de deficit. 11 est tres tentant de mettre ce deficit en rapport avec une perte de debit dans le conduit souterrain a la faveur d'une emergence sous-alluviale aujourd'hui inconnue. L'existence d autres emergences sous alluviales mise en evidence par tra9age colorimetrique (Viewc Jeand Heurs Usine de Beurey sur Saulx) accredite cette hypothese. Cependant il est boo de noter les conditions de mesures mediocres qui regnent a l'emergence du Rupt de Frainiau (R15) : section large eau profonde fond vaseux et surtout vitesse d'ecoulement tres lente. Une erreur de quelques litres secondes dans la mesure modifie de maniere considerable le resultat. On se gardera done de toutes conclusions hatives tant que d'autres mesures n auront pas permis de confrrmer ce fait. Les campagnes ont alors ete classees selon les debits en Rl3 et representees dans le graphique suivant. 3000 Debits cumules en I / s 2500 2000 1500 1000 500 0 9 -D ec 95 S e uil d e declenc h emenr de l 'e mer ~ence r empo r ai r e ( la vasq u e clu R up r d u P uirs) 7-A IT -96 2I-Mar 96 2 2-FC\ 95 8 -Dcc-96 Saturation de /'emergence pe r e nn e: le Rupr de F rainiau 7D cc-% 2 6-Fc -% ll-FC\ -9 5 25 FC\ -% 24-FC\ -95 Rl4 Vasque Rupt du PuitsSRI5 Rupt de Frainiau R I J C ollecteur ii l echelle Figure 5 : Representation graphique des dix campagnes de mesures de debits aux trois points caracteristiques du siphon aval du Rupt du Puits. Sur les quatre premieres campagnes (Fig. 5) quand le debit augmente sous terre (au collecteur) il augmente (toute proportion gardee) a !'emergence perenne : le Rupt de Frainiau. La vasque (temporaire) du Rupt du Puits elle ne coule toujours pas A partir d un certain debit que nous qualifierons de debit seuil la vasque s'amorce et se met a couler A S ym po siu m 7 : P hysi ca l S pel eol og y 173

PAGE 187

partir des campagnes de mesures dont nous disposons il semble raisonnable de penser que ce debit seuil se situe entre 21 o et 230 1/s. Si le debit au collecteur continue d'augmenter la vasque temporaire du Rupt du Puits prend le relais et son debit augmente dans des proportions bien plus importantes que le Rupt de Frainiau qui voit son debit continuer de cro1tre mais semble vite saturer. Les mesures de conductivites, de temperatures et les differentes analyses physico-chimiques realisees sur les trois points caracteristiques du siphon aval du Rupt du Puits permettent de preciser la nature de ces ecoulements souterrains (Fig. 6). On constate ainsi que quelle que soit la situation hydrologique la conductivite, la temperature et la durete totale des emergences sont superieures aux valeurs du collecteur souterrain. Systematiquement la vasque temporaire du Rupt du Puits est la plus chargee et la plus chaude. Certes durant son parcours noye l'eau continue a se charger mais Jes 450 m de parcours souterrain ne suffisent pas a expliquer une telle variation du chimisme. Pertcs et gu uffr cs ne passa nl pas par le collec l e ur CO LLE C l H I R Sot TERR, \J ,'\ DL : Rl'PT Dl' PlTfS R 13 Drams paralleles au co ll ecteur : E~ I ERGENCEFOSSILE SL 1 PPO _ EE Dl 1 SYSTEME VASQUE TEl\ l PORAIRE Dll RlfPT Dll PlllTS R 14 EMERGENCE PERENNE Dl' Rl 1 P T DE FRA I N1Al 1 R 15 C onduit lnconnu EMERGENCE SO! 'S ALLl V I ALE SllPPOSEE Figure 6 Collductivites, temperatures et duretes totales lors de quelques campagnes. Ainsi !ors de la campagne du 24 fevrier 1996, la charge dissoute en carbonates double en moins de 500 m de parcours souterrain noye et ceci pour une vitesse de transit de l'ordre de 3 a 6 heures. Comme pour Jes debits, le chimisme montre un apport d'eau complementaire dont ii nous faut trouver l'origine. 3. Vers un schema de fonctionnement du siphon A partir de ce constat, ii convient de tenter de determiner la nature des apports qui permettent d'expliquer !'augmentation de debit et la variation de chimisme dans le siphon aval du Rupt du Puits (Fig. 7). Ils peuvent selon nous etre de trois types : -1Une serie de drains paralleles au collecteur du Rupt du Puits rejoignent le systeme dans le siphon aval. L un au moins est assez bien connu c'est le drain parallele des Meilleurs. Une coloration dans cette galerie a permis de montrer son appartenance au systeme. Campagne de mc~ures 9-Dec-95 2-1 -Fe -96 26 -Fev-96 '.!7 -Fc\'-96 21-~ far -96 7-. \H -96 Figure 7: Alimentation etfonctiomiement mppme du siphon aval du systeme karstique du Rupt du Puits. A partir d'un certain debit, ce drain se met de meme a deborder et rejoint le collecteur souterrain du Rupt du Puits complexifiant le systeme d'alimentation du siphon -2Une serie de pertes et de gouffres verticaux existent a !'aplomb du siphon aval du Rupt du Puits En periode de pluie et de hautes eaux, ils contribuent au moins pour partie a l'alimentation directe du secteur noye sans passer par le collecteur. -3Les deux precedents apports ne permettent a priori pas d'expliquer les centaines de litres a la seconde qui viennent grossir le Rupt du Puits et le Rupt de Frainiau en periode de crue. ils nous faut supposer !'existence d'un apport supplementaire qui pourrait etre la Saulx elle meme Effectivement, la Saulx decrit un meandre un peu plus en amont au niveau de Vieux Jeand'Heurs. Dans le lit meme de la Saulx, sourd une emergence sous-alluviale mise en evidence par les trai;:ages montrant !'intense karstification du secteur sous le niveau de base. On pourrait done supposer que des pertes de la Saulx suivraient un parcours submeridien et rejoindraient le siphon aval du Rupt du Puit s. Le remplissage tardiglaciaire des rivieres lorraines (WEISROCK, 1993) et done de la Saulx masquerait ces pertes sous quelques metres d'alluvions. Conclusion Le schema propose ne reste pour !'instant qu'a l'etat d'hypothese. Ces hypotheses et une serie d operations sont prevues qui permettront de confirmer ou d'infirmer comme par exemple : Visite et !eve morphologique en plongee du siphon du Rupt du Puits reperage des drains annexes Reperage (toujours en plongee) d'eventuelles pertes dans le secteur de Vieux Jeand'Heurs (dans la Saulx). Analyses physico-chimiques suivies et comparees entre la Saulx et le reseau karstique Suivis pousses en continu de !'evolution des debits aux trois points de mesure (pose de stations hydrometriques en cours) et sur la Saulx pour analyser les hydrogrammes des differents systemes. La Saulx ayant un tarissement beaucoup plus lent a priori que le systeme karstique du Rupt du Puits (rapport d'echelle et principe d'hysteresis) References DESCA YES F. 1962. Repertoire des cavites de la region de Bar le Due. Spelunca n. Oct-Dec pages 27 a 32. DEVOS A. DEPAQUIS J P. HERBILLON C. JAILLET S 1996. Une campagne de mesures de debits au Rupt du Puits. Echo des cavemes meusiennes n. CDS 55. Bar le Due. JAILLET S. 1994 Une classique revisitee : L'aventure du Rupt du Puits Speleo n, la speleo grand format. page 3 a 6 JAILLET S. 1996. Dynamique des karsts couverts de bas-plateaux : L'exemple du Rupt du Puits Morphologie Hydrologie Indicateurs de l'environnement. D E.A. Universite Michel de Montaigne Bordeaux III. 151 p WEISROCK A 1993 Le remplissage tardiglaciaire et holocene des vallees lorraines. L'eau, la terre et les hommes: au fil de l'eau hommage a R FRECAUT Presses universitaires de Nancy page 303 a 309 17 4 Proceed i ngs of the 12 '" Interna tion al Congress of Speleology, 1997, Switzerland Volume 1

PAGE 188

The Role of Groundwater in the Genes i s of Resava Cave by Danica Vasileva Federal Hydrometeorological Institute, Bircaninova 6, I 1000 Belgrade Yugoslavia Abst ra ct Resava Cave is situated in the northern marginal part of the Divljakovac Polje in the eastern part of Yugoslavia. One of the region's most beautiful caves (also called the Resava Jewel) or was created by a long geological history of primarily groundwater activity. The cave is impressive and rich in sinter decorations such as stalagmites and stalactites. The genesis of Resava Cave is very complex and its generation was determined by many factors, primaril y by tectonic and hydrogeological ones. Water impacts on the cave's genesis have been connected with Alpine orogenesis. In the geological past a stream flowed through the Divljakovac Polje At that time, the river bed was at the cave entrance level and water was discharged in part and infiltrated into minor and major fissures resulting in mechanical and chemical processes in the carbonate rocks which created the cave and its pre sent appearance. R e sum e La Caverne Resa va est situee au nord bord du champ calcaire Divljakovac a la part orientale de Yougoslavie. Une de la plus belle caverne (appelee aussi la Caverne Gemme) etait cree pendant une longue histoire geologique avec des eaux, en premier lieu, des eaux outerraines. La caverne semble impressive, rich avec les ornaments des concretions calcaires comme des stalactites et des sta lagmites. La genese de la Caverne Resava est rres complexe et sa forn1e est definie par beaucoup de facteurs, premierement tectoniqu es et hydrogeologiques Le but de ce travail est une representation de la role des eaux souterraines sur la genese de la Caverne Resava L 'ac tion de s eaux souterraines sur la creation de la caverne etait en liaison avec Orogenese alpine dramatique Dan s le passe geologique un cours d'eau etait ecoule en traversant le champ de Divljakovac. Autrefois le fond du lit etait au niveau avec l'entr ee de la caverne et une part d'eau etait debordee et penerree a travers les crevasses moins grandes et plus grandes, faisant au lon g de sa route des proces mecanique et chimique dans le roches calcaires ce pourquoi on est cree la forme de la caverne avec le temp s. 1. Introduction The aim of this paper is to attempt to explain the genesis of Resava Cave as the result of complex geologic factors and especially by groundwater activity The geology, tectonic development geomorphological features surrounding the cave were presented by Y UGOS LAV COMMITTEE FOR INTERNATIO AL HYDROLO G I CAL DECAD E ( 1976) and by CIRIC ( 1996) A general review of the cave's genesis relative to Balkan tectonism was given by VASILEVA (1995) 2. A general review The recent "static" phase of the Resava Cave's genesis has been preceded by sedimentation of limestone and intensive MESOZOIC T J 280 195 137 ----CARBONATE ALPINE tectonic development which continued for a long time on the Balkan Peninsula The large carbonate masses were deposited from Middle Triassic to Upper Cretaceous Time, beginning approximately 235 million years ago (Table I). Carbonate rocks of the Tertiary age occur sporadically cover smaller surfaces. Intensive tectonic deformation of carbonate rocks related to Alpine orogenesis, began during movements that created the chain of mountains such as the Alps, Carpatho-Balkanian Arc, Caucasus and some mountains in Asia. During that period of time, the majority of the initial conditions were created such as faults. fissure s and cracks. CENOZOIC -PERIOD Cr Tc Q Age 67 1.5 _JMillion _yea rs] -ROCKS Sedimentation OROGENESIS Neotectonic Tectonic I ----P RE PHASE (Sedimentation) and Recent phaCave's Genesis TECTONIC PHASE
PAGE 189

Alpine orogenesis continued for about 150 mi ll ion years from the Middle Jura sic During the Middle Oligocene, this stage calmed but the tectonic activity continued by neotectonic development of terrain which is present today The main stage of the cave's formation by groundwater activity was related to the late Tertiary and beginning of Quaternary when the majority of territory was covered by water. In the Quaternary, during the major climatic changes. water played an essential role in genesis of underground karst forms. 3. Position geological composition and tecton i c development of the terrain The Resava Cave is situated about 140km to the south-east of Belgrade (Figure I) in the valley of the Divljakovac River a southern tributary of the Re sava River. The cave entrance is 483m.a s.l. The cave was discovered more than three decades ago and was opened in 1972 for tourist visits 10 IHm t=-r .. ----=::== C"it..l::. I .\< 'l:.Ol S .\UI::. PFR\OA,._ ,--------, PALEOZOIC S{'HISTS n\l::.RIHklSI Figure I: Position and geological compositio11 of Resava Cave s u rrou11di11gs (Source: YUGOSLAV COMMITTEE FOR l TERNATIONAL HYDROLOGICAL DECADE, /976) The geological composition surrounding Resava Cave (Figure I) is complex (ANTO IJEVIC et al., 1962 ; BOGDA OVIC, J 968) The limestone rocks of Jurassic-Cretaceous age extend as a belt of I 0km width between Permian sandstone in the west and Paleozoic chist in the east. The complex geological composition of terrain is a consequence of tumultuou s geological past (CIRIC, 1996). Oubmn ,_Jmn 4 7 1LH J O 1 00 200m Figure 2: Position of Resava Cave i 11 the li 111esto11e of Kucaj's anticline The main stages of Resava Cave's development can be involved in the scheme mentioned above. The Permian sa ndstone in the west was dragged on the Mesozoic limestone as an overthrust, probably at the beginning of Tertiary During the Alpine orogene stage, the l imestone of Jurassic-Cretaceous age was folded with the axes generally oriented north-south. Resava Cave is situated in the western slopes of Ku-aj's anticline (Figure 2). By folding of carbonate rocks in Alpine orogenesis, faults and fissures were created, predominant directions of the future groundwater flows and underground karst development. 4. Geomorphological features of Resava Cave From the geomorphologic point of view, Resava Cave is a very complex karst phenomenon The ground plan of the cave is an irregular shape (Figure 3). This dry cave is a former ponor consisting of halls, canals and crevices situated at two levels The difference between the highest (482.6) and the lowest elevation ( 453. 7) is 28 9m. O ~ Om @ Entn11~llall ~ 4 111\t @ P onor @ \' ~ st ,b ulcoflllstory (I) c.,,1a111.i 1 @ C oncfflH a/1 (f) Co,. I Canal Figure 3: Tlie ground plan of R esava Cave (Source: LAZAREVI<; 1991) The entrance of the cave leads to the Entrance Hall, which has approximately 630m of the floor area, and reaches ceiling heights of 2-3m. From there, the path leads southern and down into the Bee-Hive which is 4-5m high and divides into Ponor to the south and to the Vestibule of History to the north which is 78m high The path continues north for 20m along a dug passage and enters the Crystal Hall which is rich in stalactites, and reaches ceiling heights of about 4m. The Crystal Hall has three exits: one leads into narrow canal that heads west for about 25m and opens 3m above the Coral Canal. The second exit is situated to the north-east with a network of canals and ponors heading different directions The final exit leads downstairs (through a dug canal) for a length of !Orn into the Concert Hall. The Concert Hall has 850m 1 of floor area, reaches a height of about 20m, and is rich in sinter decorations A ponor existed at the bottom of the Concert Hall which was filled when the cave was prepared for tourist visits. From here the path leads gradually up to the Coral Canal and afterwards, to the Entrance Hall where the trail finishes with a length of 402m through the cave Geomorphologic features of the cave, relate to groundwater flow movements and water action in its mechanical and chemical formation 5 Groundwater i mpact to Resava Cave genesis The rivers Resava and Resavica flow through the investigated terrain and meet at Despotovac (Figure I) Resava Cave is situated between the southern tributaries of Resava : Klocanica River and Divljakovacka River (Figure 2) By the end of the Tertiary and beginning of the Quaternary. large quantities of water penetrated this area and flowed into the 176 Proceedings of the 12 1 h International Congress of Speleology, 1997, Switzerland Volume 1

PAGE 190

Morava Neogene Basin. At that time, groundwater created the majority of underground karst phenomena Divljakovacka River played an essential role in the creation of Resava Cave. At that time this river flowed widely through the valley and its level was above or at the entrance level (Figure 2) where water directly infiltrated into faults, fissures, cracks and crevices. The width of infiltrated water flow was probably not less than the 20m width of the Entrance Hall. The head of a water was not less than 2-3m. The velocity of water was not enough to mechanically erode the carbonate rocks in height, but lateral water power was strong. At the bottom of the Entrance Hall, groundwater reached a certain velocity and divided into two directions: to the southeast the Bee-Hive, to the east the Coral Canal. The larger quantity of groundwater flowed southeast, judging by the dimension of the Bee-Hive, and drained out through the Ponor partly. With increased velocity, water flowed into the Vestibule of History and cut a canal afterwards with a height of about 7-8m. By increasing its velocity and power, the groundwater flow fell into the Concert Hall. The groundwater flow descended the Coral Canal with a rapid acceleration along the steep slope of the canal, and drained out to the beginning of the Concert Hall. At the same time, the water flow infiltrated into the system of canals in the Crystal Canal, from where it drained partly through 25m long canal and fell in the Coral Canal from the height of 3m The other part of the groundwater flow caused by the steep slope, entered the Concert Hall, joining the above mentioned water flows and with greater mechanical power, dug a cavity of large dimension (25x34x20m), and afterwards disappeared at to unknown depths through the ponor at the bottom of the hall, probably to the south-east direction In final stage of cave formation, the rivers cut into carbonate rocks and descended down gradually to the river bed (about 80m) so it remained out of the river range Except the temporary floods, the cave has become dry with permanent moist walls, by which calmer process of cave formation has continued The chemical impact of water on genesis of Resava Cave related, not only to dissolving of carbonate rocks, but also in its sedimentation in the form of cave ornaments such as stalactites, stalagmites, and columns (MIJATOVIC, 1984 ). The sinter decorations were generated during favourable conditions: a necessary concentration of calcium carbonate in the water the favourable temperature for its sedimentation, also the absence of mechanical destruction, such as cave floods and collapse of ceilings caused by earthquake. There are no cave ornaments in the Entrance Hall because it has been exposed to external climatic variations which did not allow creation of appropriate conditions The Crystal Hall is very rich in sinter decorations, there are, especially, numerous small stalactites. There are, also, the decorations in the Bee-Hive Vestibule of History and Coral Canal. In the Concert Hall are an abundance of stalactites and stalagmites, and the two massive columns. One is at the beginning of the hall and is I Orn in diameter. The other is at the bottom of the hall and has an irregular shape of 5l 5m in diameter Judging by the dimensions of these columns the time of their creation is very long and entered Upper Tertiary. That means the cave, with its system of canals and halls was also created during the Tertiary. Resava Cave had not been researched paleontological during its discovery During preparation of the cave for visits, any remains of the flora and fauna were probably eliminated The clayey and muddy traces founded on the walls and decorations testify that the cave floods temporarily. The important problem of adequate illumination of the cave has been solved professionaly (LAZAREVIC, 1991 ), by which the natural conditions have been protected 6. Discussion [n previous researches of Resava Cave's genesis, there is no evidence of the cave's age as underground karst phenomenon, the age of the oldest ornaments, and nor an age of possible remains of flora and fauna. Therefore an obligation remains to establish the ages mentioned above and, also the geomorphologic features of the cave area by which the data of Resava Cave's genesis would be complete. Thus the role of groundwater in Resava Cave's genesis will become more clear. Conclusion A general summary of cave's genesis as a result of the complex geological-tectonic development on the Balkan Peninsula has been presented in this paper The sedimentation of carbonate rocks began 235 million years ago, from Middle Triassic to Upper Cretaceous, and the duration of intensive tectonic development as a consequence of Alpine orogenesis was for 150 million years from Middle Jurassic to Middle Oligocene as a pre-phase to the recent "static" genesis of the cave. The genesis of Resava Cave as a result of numerous factors primarily ofhydrogeological one, has been presented. The Divljakova cka river played an important role in the Resava Cave's creation. At that time the river bed was at the cave entrance level and river water was discharged into faults, fissures, cracks and crevices, resulting in mechanical and chemical erosion of the carbonate rocks. The analysis of groundwater flow on cave development has been done indirectly using the cave's geomorphologic features It is believed that the cave with its system of canals and halls was created during the Tertiary Resava Cave is rich in stalactites stalagmites and columns The columns have a large thickness which gives an evidence of their long period of generation that probably entered the Upper Tertiary. Nevertheless, for more exact ly determination of Re sava Cave's age and the role of groundwater in its formation, data obtained by laboratory analysis are necessary Acknowledgement I would like to express my apprec1at1on to the Federal Hydrometeorological Institute in Belgrade for its support and also to Mr. Dejan Lekic, B .Sc. for graphical description Special thanks are due to Dr. Radenko Lazarevic for important oral information. References ANTONIJEVJc, I. & R MILOsAKOVJc. 1962. Geological column of Kucaj Geozavod (Yugoslavia). Vesnik XX, Belgrade : 93102. BOGDANOVJC P. 1968 Stratigraphy of Upper Cretaceous of Eastern Serbia. Geozavod (Yugoslavia). Vesnik XXVI Belgrade: 71-90. CIRJC B. 1996 Geology of Serbia. Geokarta (Yugoslavia), Belgrade: 273. VASILEVA, D 1965. Impressions from scientific excursion : Resava Cave Devil's Town Lecture. Belgrade : 1-8 Symposium 7 : Physical Speleology 177

PAGE 191

LAZAREVlc, R. 1991. Results of microclimatic measurements in the Resava Cave. Institute for Protection of Nature, (Yugoslavia). Prote c tion of Nature 43-44 Belgrade : 57-79. MJJATOVlc B. 1984 Hydrogeology of the Dinaric Karst. Yerlag Heinz Heise GmbH Germany International Contributions to Hydrogeology Vol. 4, IAH : 255 YUGOSL AV COMMITTEE FOR INT E RNATIO NAL HYDROLOGI CAL DECADE 1976. Hydrogeological Map, scale I :500000, Belgrade Bell hole morphometry of a flank margin cave and possible genetic models: Lighthouse Cave, San Salvador, The Bahamas Stei nE ri k La u ritzen' J o yce L undb e r g 2 John E. Mylroie 1 & To b y D ogw il er 1 Department of Geology University of Bergen, Allegaten 41. N-5007 Bergen, Norway, stein.lauritzen @ 2 Department of Geography, Carleton University, Ottawa, Canada KlS 5B6, jlundber@ccs 3 Department of geosciences, Mississippi State University, Mississippi State, MS 39762 USA, MYLROIE @ Geosci Lighthouse cave is a large flank-margin cave that is developed within eolian calcarenites of pre-last Interglacial age, whilst most of the speleogenesis might have taken place within a time window of about 15 ka during Oxygen Isotope stage 5 (MYLROIE & CAREW). The cave ceiling displays a vast amount of bell-holes Bell holes are vertical cyl i ndrical to paraboloidal holes that may penetrate up to several meters into the host rock, in most cases without evidence for any pre-existing guiding void or fracture. Morphologically, they are superimposed onto the primary (phreatic) cave ceiling They may occur isolated but are most often clustered in groups that at high density, display polygonal packing patterns In the main chamber of the cave, 30 bell holes were measured and accurately located relative to each other. Due to the distribution over the sloping ceiling of the cave, there appear to be no level (watertable) control in their distribution Almost without exception, each bellhole is paired by a floor counterpart (bell-pit) that is slightly wider and shallower, and aligned vertically beneath The shape characteristics of the bell holes, their relation to corresponding floor cavities (bell pits) and their relation to corroded calcite speleothems and evaporite crusts is suggestible of a vadose origin rather than a phreatic, convection-cell type of formation mechanism This will be discussed towards various genetic modes and mathematical models for bellhole development. 178 Proceed i ngs of the 12 1 h Internat ional Congress of Speleology 1997, Switzerland Volume 1

PAGE 192

Cave deve l opment along the water table i n Cobre System (Sierra de Peiialabra, Cantabrian Mountains, N Spain) by Carlos Rossi (1), Alfonso Munoz (2) an d A d ri ano C o rtel (1) Dpt. de Petrologia y Geoquimica, Facultad de CC Geol6gicas, Universidad Complutense, 28040 Madrid, Spain. (2) Dpt. de Geodinamica, Facultad de CC Geol6gicas, Universidad Complutense, 28040 Madrid, Spain. Abstract Cobre Cave System (CCS), with 10.6 km of surveyed passage, is developed in a 150-m-thick sequence of Carboniferous limestones dipping 45-60 2 against the valley slope. CCS has a multi-story structure, with abundant relict levels at different elevations in 225 m of vertical range. Drainage at the modem level is mainly strike-oriented and occurs along two types of passages: (I) a low-gradient canyon in equilibrium with the water table, (2) a system of phreatic tubes upstream. Six sumps are known in this area that have water at the same elevation and are thus exposures of the water table Along the strike, the water-table gradient is very low ( I .4%), having only minor steps caused by faults. Profiles perpendicular to bedding reveal a more steeper gradient (4 6%) as a result the lower hydraulic conductivity in this direction The low-gradient canyon has several levels of notches with corrosion bevels (flat solutional roofs regardless of geologic structure) above the modem channel. They can be correlated along great distances and represent paleo-water tables. Toe canyon may be a 50-m-high, single conduit, but it usually splits into several relict, beveled canyons with separate phreatic roofs, low-gradient floors and whose elevation ranges do not overlap. The existence of correlatable, beveled canyons in nearly all the vertical range of the CCS indicates a long history of base level changes characterized by periods of static base level interrupted by episodic downcutting. 1. Introduction The recognition of fossilized paleo-water table surfaces in multiphase caves can serve not only to study base level evolution but also to understand the present-day hydrology of the system. In addition. the knowledge of the geometry and gradient of the present water table will help to validate the paleo-water-table definition. Several morphologic criteria have been developed to define paleo-water tables inside multiphase cave systems, such as the piezometric limit concept (PALMER, 1987-1989), the distribution of isolated vadose trenches in the upper parts of phreatic loops (e.g. SMART & CHRISTOPHER, 1989) or the presence of extensive beveled corrosion notches (FORD & WILLIAMS, 1989;FRUMK1N, 1995). Cobre System is a multiphase cave, 10.6 km long, developed in steeply dipping Carboniferous limestones. The downstream part of the lowest active l evel consists of a subhorizontal river passage that extends for 1.7 km to the spring and is apparently adjusted to the present base level of the system. This disposition suggests that the river passage could reflect the water table of a true phreatic zone. If this phreatic zone really exists downdip the limestone, and a water table control for the low-gradient river could be demonstrated, then projected elevations of the stream will show the topography and gradients of the water table surface, and eventually the influence of local geological structures, lithology and recent base level evolution on this geometry. All these data will be crucial for the definition and correlation of water-table-controlled, relict features inside the system, which represents an ideal site for this kind of study, as: (I) the steeply dipping nature of the beds allows easy discrimination between lithologic and water-table controls upon passage development, (2) there is an exceptional development of beveled notches of great lateral extent that can mark paleo-water levels very precisely and (3) the system show a well-developed multi-story structure, with abundant relict vadose, phreatic and water table passages, over 225 m of vertical range above the present active level. Toe aim of this article is to discuss the possible water-table control for the low-gradient river, and eventually the precise geometry and controls of this water table. As accurate elevation data is essential for this analysis, a specific survey have been carried out along the lowest active level of the cave, with a mean vertical error estimated in 0.1 % of surveyed length. Elevations of the water surface, passage floors, notches and corrosion bevels have been measured in every survey station of the main passage (128 stations). All these features were correlated in situ and mapped in profiles A 3D frame of the survey of the whole cave system as well as some selected projected profiles have been produced and analyzed with CAD software In these profiles the survey line of the river passage has been transformed to represent the low water level of the stream. 2 Geological and hydrological setting Cobre Cave System (CCS) is located in the Sierra of Peii.alabra, in the central part of the Cantabrian Mountains (Province of Palencia, N Spain). The cave is developed in a 150m-thick sequence of Carboniferous limestones (the Agujas Limestone Member) sandwiched between impervious siliciclastic turbidites of the Vanes and Covarres Formations (VAN DE GRAAF, 1971). These formations crop out in the south-western flank of the Sierra, where they show an overturned dip of 45-60 2 against the slope (Fig. !). This structural pattern locates the position of the karst base level just at the outcrop of the lower contact between limestones and turbidites, i.e., at the intersection of this lithologic boundary with the topographic surface. Toe karst base level is thus stratigraphically perched above the regional base level, which is located I 00-500 m bellow at the Pisuerga river valley, built in non-karstifiable rocks. The outcrop of the lower limestone-turbidite boundary along the Sierra is marked by several permanent karst springs, each one being the outlet of a separate cave system. Toe karst is thus divided into several blocks, separated by major faults and/or transverse valleys incised into the limestone. Each block drains preferentially along the strike, discharging at a spring located at the lowest outcrop point of the limestone. Cobre cave entrance represents one of this springs, draining a major limestone block Symposium 7: Physical Speleology 1 79

PAGE 193

1 Pisuerga river I IMPERVIOUS TURBIDITES COVARR~S FM h...A. Vadose trench notch levels +flow VANES SHALE ... _ i Phreat ic ...,..,,._, conduit , water table ,, resistant ,,, beds base level 1 at time 1 Fig. 1: Schematic vertical section of CCS In the dip dJrecdon, showing the response of the cave system to a base-level lowering (of "a" magnitude) caused by an erosive slope retreat (of "b" magnitude). at the south-western corner of the Sierra and giving access to nearly I O km of surveyed passage. The Agujas limestone, whose outcrops range in altitude from 1600 to 1900 m as!. in the CCS area, is overlain by unconformable Triassic Buntsandstein facies (sandstones, shales and conglomerates), which form the summits of the Sierra at 2100-2300 m asl. As the limestone crop out in only a narrow strip, sinking surficial streams originated in the Buntsandstein highlands provide most of the recharge for the karst aquifers in the underlying limestone A significant part of the allogenic recharge to the CCS occurs via two major penetrable sinks: the Torc6n (explored for 0 9 km and 140 m in depth) and Se/ (explored for 0.7 km and 115 m in depth) sinks, both connected hydrologically, but not yet physically, with the main cave. The streams that sink in Torc6n and Se! drain the Covarres Cirque and the Sel de la Fuente valley, two major forms of glacial origin, built essentially in insoluble Buntsandstein, and partially filled with Pleistocene moraine deposits (HERNANDEZ-PACHECO, 1944). Torc6n, as other sinks or paleo-sinks explored along the system, are partial to entirely filled by till; wedged boulders of limestone and Buntsandstein conglomerates that can reach several meters in diameter can be found deep into the cave system, occasionally sealing some inlet passages. 3. Description of the cave system and its hydrology CCS is comprised of three different sectors (Fig. 2) in terms of morphology, hydrology and relative age: (I) the low-gradient Apia canyon; (2) the Sifones-Sel-Torc6n branch; (3) the south eastern branch. (I) The Apia canyon is a low-gradient river passage in the downstream part of the system The active channel is excavated in the limestone bedrock and shows relatively little elastic bed load Although the canyon may be a 50-m-high, single conduit guided by a mayor phreatic tube, it usually splits into several relict canyons with separate phreatic roofs, low-gradient floors and whose elevation ranges do not overlap Along its entire elevation range, the canyon and its sub-canyons develop extensive and correlatable levels of corrosion notches. (2) The Sifones-Sel-Torc6n branch is a mixture of looping phreatic tubes and active vadose conduits, which transmit to the Apia river most of the allogenic water from the Torc6n and Sel sinks. The entrance series of Torc6n and Sel consist of active vadose shafts and high-gradient active canyons whose water finally disappears in impenetrable fissures These canyons intersect a system of relict phreatic tubes that loop at elevations between 95 and 81 m above the spring level (abbreviated as a.s.). In both caves, narrow vadose canyons are found again below this depth, giving access to another system of looping tubes in the range of 62-44 m a.s. and ending in a stagnant sump at the lowest point. An obviously equivalent and nearby tube system is accessible from the upstream end of the Apia canyon in Cobre Cave. This tube system, known as the Sifones series, is developed in the same elevation range of Sel-Torc6n lowest tube system, and contains exclusively phreatic tubes that form remarkable loops of 25 m of maximum amplitude. Five sumps are known in this area, all having water at near the same elevation (43-45 m a.s). The more north-westerly of these sumps supply most of the flow of the low-gradient Apia river, whose floor is also coincident in elevation with the water levels in the sumps upstream. The water in the other sumps are apparently stagnant, but their water levels fluctuate seasonally at least I m. (3) The south-eastern branch, that will not be discussed in this paper consists of a series of upper level relict conduits (the oldest passages in the CCS) partially captured by modern small streams infiltrated from the moraine deposits that cover the limestone in this area. The vadose streams of this branch join directly the Apia river through the Mareo meander or become lost in impenetrable fissures towards the active Sifones-Sel Torc6n branch 4. Profiles and gradient of the river passage The Apia river extends for 1.7 km, from the active sump that supplies its water to the spring at the Cobre entrance The difference in elevation between the spring and the water-supplier sump is 44 m, which gives a mean gradient of 2.5%. Seen in extended profiles (Fig. 3), the gradient is not uniform along the river passage, which can be subdivided for this purpose into two separate sectors: the downstream quarter (450 m of passage), with a mean gradient of 4 6%, and the remaining 1630 m upstream, with a remarkably less steeper mean gradient of I .4%. Once again, along each sector the gradient is not uniform. The shorter and steeper downstream quarter comprises several lower gradient segments separated by short ramps whose location is controlled by dolomite zones concordant to the bedding (see further details in ROSSI & MUNOZ, 1993). Upstream, the 180 Proceedings of the 12 th International Congress of Speleology, 1997, Switzerland Volume 1

PAGE 194

remaining main portion or the river passage shows an almost uniform gradient, near to the mean value of I .4 % only interrupted by minor steps of decimetric magnitude associated to mineralized faults. If the survey of the river surface is analyzed on projected profiles it becomes evident that the gradient thresho l d correlates with a significant change in passage structural orientation: (I) the downstream quarter of the river is broadly oriented parallel to the dip direction of the bedding, and in this sector the cave rapidly penetrates most of the stratigraphic section of the Agujas limestone following a series of faults (2) Upsteam, the main portion or the river passage follow either faults or bedding planes (both steeply dipping) but in this case there is a clear general tendency to drain along the strike of the limestone. 5. Morphology of the active channel Close to the active channel, the lowest portion of the Apia canyon contains three extensive levels of solutional sub horizontal notches along the walls (Fig 3). These notches extend several meters into the rock, are typically I m deep and show remarkably flat solutional roofs that bevel the dipping strata (corrosion bevels, according to FORD & WILLIAMS, 1989 nomenclature) These beveled notches are partly filled with fining-upwards sand and gravel sequences capped by flowstone. In the downstream quarter of the Apia canyon, the present stream occupies an incised trench under the lowest notch. The depth of the trench under this notch is maximum near to the entrance (about 6 m), decreasing rapidly and progressively in the upstream direction. 560 m away from the entrance this trench disappears, and the underground river is at the level of the lowest notch This lowest notch, which develop a very sharp bevel, can reach 20 m in width near the entrance, but progressively reduces upstream until vanishing as a notch approximately I km away from the spring The second notch is composed of two sub notches with corresponding bevels located at constant relative elevations of I and 2 m respectively, above the bevel of the lowest notch. This composite notch, especially the upper sub notch, can b e correlated along the entire Apia river. This notch is however interrupted when the phreatic roof descends to its leve l or just below representing paleo-sump sections that represent de facto piezometric limits (sensu PALMER, 1987) which demonstrates the strong water-table control in notch origin 6. Discussion The interpretation of the Apia river in CCS as an exposure of the water table is supported by (I) the low-gradient and perennial character of the stream, (2) the upstream origin of the water supply, directly from flooded phreatic condu i ts, (3) the coincidence in elevation between the sumps of the Sifones-Sel 1820ma s l Upper lim i t of limestone outcrop ---------------~-~-~-~------_ 1720ma s.l ______ Torc6n branch upstream and the water level in the Apia river downstream and (4) the broad parallelism between the profile of the present stream in Apia canyon and of those of the recent beveled notches, interpreted in turn as stable water table features The peculiar structural pattern of the limestone, steeply dipping against va ll ey s l ope an d with unde r lying impervious rocks, favors wate r impoundment and thus produces a phreatic zone downdip the limestone The lowest outcrop point of the limestone-turbidite contact controls the vertical position of the spring, which in turn controls the elevation of the water table at the output of the aquifer. Wate r tabl e geo m etry If the water table is defined by the level of the active passage floor in the low-gradient river and by sump levels in the phreatic segments upstream (in the Sifones-Sel-Torc6n branch), then their projected profile will show the topography of the present water table surface Along the strike, the water-table gradient is very low (I .4%), having minor steps caused by faults that give a smooth, stair-like geometry to the water table. Profiles in the dip direction of the bedding, reveal a more steeper gradient of the water table (above 4 2 ). This geometry is caused by the lower hydraulic conductivity in the dip direction, as the flow has to cut across the b edding; moreover, the limestone has several chert rich beds that act as minor hydrologic b arriers Along the strike, the hydraulic conductivity is enhanced as underground drainage is favored b y the bedding p lanes and abundant fa u lts parallel to bedding. The topography of the stab l e paleo-water tab l es deduced from projected profiles of the b eve l ed notches is roughly the same (with the above-expressed exceptions), which strongly supports this interpretation of the water table geometry Pr ese n t d eve l o pm e n t of ac ti ve phreatlc conduit s As the existence ofa phreatic zone requires effective porosity below the water table, and the Agujas limestone itself lacks intrinsic porosity, the phreatic develo p ment must be achieved via solution conduits, open fractures or both Upstream of the Apia canyon, in the Sifones sector, perhaps most of the flow occurs through active phreatic conduits that transmit to the Apia river the allogenic recharge that enters via the Torc6n and Sel sinks The existence of a we ll integrated tube system in this area is suggested by: (1) the water of Apia river is supplied by a flooded rising phreatic tube; (2) the Sifones sector (and also most of the lower sections of Sel and Torc6n) consists almost exclusively of dry looping phreatic tubes, generated when the water table was higher: a similar system could be active today in the same sector below the water table (3) several stagnant sumps with equivalent water levels are known in this area, which suggest the existence of an active system of liked tubes in which a steady phreatic flow is taking place today. '-: _______________________________________ ~ iu ~ki~ -_-_ ~ w_Aj~~ ~ j-~~~ ~ ~ -_ ~ : !~22 f11_ ~5.:1 spr,ng APIA CANYON SIFONES-SEL-TORCON BRANCH Fig. 2: Profile view of t h e the s u rvey f ra me of Co b re Cave Syste m a l o n g th e s tr ike of t h e lim esto n e. Sympos i um 7: Phys i ca l Speleo l ogy 181

PAGE 195

NOTCH LEVELS --------------------------------STREAM FLOW RESISTANT BEDS / FAULT Fig. 3: Schematic extended profile of the Apia Canyon (above) and correlation of some representative cross-sections of the lower part of the canyon (below). Horizontal scale arbitrary. The existence of phreatic conduits below the level of the stream in the Apia canyon is more difficult to prove, as no undoubted tubes have been detected under the shallow water. Nevertheless, along the whole height of the canyon there are abundant relict phreatic tubes that can be linked physically to former canyon levels and notches, distributed in an elevation range of 50 m above the stream. These relict levels and their associated tubes will not be discussed in this paper, but work in progress (ROSSI, CORTEL & KELLER, in prep.) suggests that the relict low-gradient canyon levels and stories are absolutely analogous in morphology and long profiles to the modern channel. The invariable presence of looping phreatic tubes under every relict level of the Apia canyon strongly suggest that similar phreatic conduits could be active actually under the stream level. Evolution of the water table after a recent base level faU The comparison of the active stream profile of Apia canyon and the profiles of the recent notches indicates that the river is actually downcutting: the trench under the lowest notch decrease progressively in the upstream direction, disappearance 0.5 km away from the entrance. From this point to its upstream end, the underground river is still at the level of the latest notch. This incision seems to be actually in progress, and it is expected that in some time in the future it will extend upstream to the limit of the active canyon. The disapparition of the lowest notch I km away from the spring suggests that this notch was not fully developed when a base level drop aborted it and caused its incision. The second notch, however, can be correlated along the entire Apia river, which indicate that this second notch had thus enough stable time to develop The present water table is thus relatively unequilibrated as a result of a recent base level drop, which has caused a steeper gradient than in stable situations The system will need some time until reach the equilibrium state, so a base level drop is not quickly followed by a water table drop, and apparently the system will downcut a trench along the entire Apia canyon to stabilize the water table and eventually begin to develop a new notch. The tube system that is inferred to exist under the Apia river was not able to divert the flow when base level dropped: this tube system must not be largeor integrated enough. Furthermore, as seen in the relict tube systems above in the canyon, the tubes under the river passage tend to be relatively short diversions that rejoin the low-gradient passage. Consequently, as the tube has to divert the water towards the same passage, the water level in the stream will not drop. Acknowledgments 0 Bemaldo, F.L. Quintanilla, A. Marcos, E. Gonzalez and A Gonzalez contributed to the production of the high-precision survey of the Apia river. This paper is a contribution to the IGPG project 379. The cooperation of the Consejeria de Medic Ambiente y Ordenaci6n de! Territorio (Junta de Castilla y Leon) is also gratefully acknowledged. References FORD, D.C. & WILLIAMS, P.W. 1989: Karst geomorphology and Hydrology. Unwin Hyman, London: 601 p. FRUMKIN, A. 1995: Morphology and development of salt caves. National Speleological Society Bullet in. 56, 82-95. HERNANDEZ-PACHECO, F 1944: Fisiografia, geologia y glaciarismo cuatemario de las montafias de Reinosa. Memorias Real Academia de Ciencias de Madrid, Sec. Ciencias Naturales, JO. PALMER, A N. 1987 : Cave level and their interpretation National Speleological Society Bulletin 49 (2): 50-66. PALMER, A N. 1989: Geomorphic history of the Mammoth Cave System. En: WHITE, W.B. y WHITE E.L (Eds.) : Karst Hydrology (Concepts from Mammoth Cave area), 317-337. ROSSI C & MUNOZ, A 1993: Geometry of piezometric surfaces in a perched karst: Redondo Valley Cantabrian Mountains, N Spain. Bulletin de la Societe Geographique de Liege 29, 153-161. SMART, P L. & CHRISTOPHER, N.S.J. 1989 : Ogof Flynnon Ddu. Jn T.D. Ford (ed): Limestones and Caves of Wales Cambridge University Press 177-189 VAN DE GRAAFF, W.J.E., 1971: Three upper carboniferous, limestone rich, high destructive, delta systems with submarine fan deposits, Cantabrian Mountains, Spain. Leidse Geologische Mededelingen 46, 157-215. 182 Proceedings of the 12 th International Congress of Speleology 1997 Switzerland Volume 1

PAGE 196

Multiple paleo-water tables in Agujas Cave System (Sierra de Penalabra, Cantabrian Mountains, N Spain): Criteria for recognition and model for vertical evolution by Carlos Ro ssi (*), A driano Cortel and Ro cio Arcenegui (*) Dpt. de Petrologia y Geoquimica, Facultad de CC Geol6gicas Universidad Complutense, 28040 Madrid, Sp a in. Abstract Agujas System, a multi-phase cave with 9 km of passages is developed in a 350-m-thick sequence of Carboniferou s Limestones dipping 50 2 to NE. Ten distinct levels of abandoned sub-horizontal water-table trunks are known, each confined in a specific ele v ation range covering together 300 m of depth The present base level drainage must be about 100 m below the lowest abandoned trunk but to date exploration has failed to find access to the lower levels. The relict trunks are wide meandering canyons, 20-40 m high, with low gradient floors and extensive levels of notches with corrosion bevels. These notches become wider near the base of the canyons, extending as far as tens of meters into the rock. Looping phreatic tubes develop just below the canyons emerging from the floor of a given notch. The extensive notches that bevel dipping strata, their low-gradient character and the multiple vadose-phreatic transitions at their bases demonstrate that the beveled canyons are relict water-table trunks. They represent an exceptionally complete record of paleo water tables that can be restored in geometry and evolution. Extended periods of constant base level were followed by abrupt drops (15-60m) producing a new canyon at a lower elevation Canyon downcutting is very rapid first, but later decelerates and reaches equilibrium creating wide notches at the water table and phreatic tubes below 1. Introduction Agujas Cave System (ACS), with 9 km of surveyed passage, has a complex multi-story structure developed in more than 300 m of vertical relief. Looping phreatic tubes and vadose high gradient canyons and shafts are present in the entire elevation range. Nevertheless, most of the explored passages are low gradient water-table trunks which are found almost in every part and at every elevation. This assemblage of relict conduits contains an exceptionally complete record of fossilized, stable paleo-water tables that can be defined by separate criteria and traced along large areas of the system. The aim of this article is to describe the general characteristics of the relict trunks and to show the relationships between these trunks and the genetically-linked systems ofphrea tic tubes. Based on high-quality survey information and detailed morphological observations, we will support a water-table origin for the trunks and discuss the water table evolution during trunk formation and also at the scale of the whole cave system ACS, whose explored conduits range in altitude from 1650 to 1950 m a s.l., is located in the Sierra of Pe.iialabra, in the central part of the Cantabrian Mountains (Province of Palencia, N Spain) The cave is developed the Upper Carboniferous Agujas Limestone (VAN DE GRAAF 1971), which exceeds 350 m in thickness in the ACS area. In the southern flank of the Sierra, the Agujas limestone has an overturned dip of 50 2 against the slope, and is sandwiched between conformable impervious turbidites. This structural pattern favors the deve l opment of a true phreatic zone downdip the limestone and places the karst base level (the springs) at the outcrop of the lower boundary between limestones and turbidites The geometry and gradient of the water table in an equivalent and nearby karst aquifer is discussed in detail in ROSSI, MUNOZ & CORTEL (this volume), where the reader can find additional data of the Pe.iialabra karst. 2. The cave levels Ten distinct levels of relict low-gradient canyons are known in ACS to date. These canyons are distributed in more than 300 m of vertical relief, each canyon being confined in a specific elevation range (Figs 1 and 2). The relict canyons are 20to 40-m-high meandering passages, with low-gradient floors and guided by looping phreatic tubes (Fig. 3). They all show extensive levels of subhorizontal corrosion notches with bevels (flat roofs regardless of geologic structure) These notches are usually restricted to the lower parts of the canyons, becoming progressively wider (as far as tens of meters) towards their bases The exploration of these beveled meandering canyons is not an easy job : the notches may be wide but are usually less than 1 m deep and separated by narrow trench intervals; frequently, the canyons are interrupted by breakdown (induced by notching) or by flowstone blockages; the canyons have a pronounced sinuosity and meanders o v erlap at different levels, creating complicated labyrinths where the explorer has to check regularly the paleo-flow from the scallops to avoid becoming lost. The guiding phreatic tube at the roof (devoid of meanders, notches and usually of breakdown) often represents an alternative and easier route except where the trench under the tube is too wide or where a younger vadose capture has widened the canyon walls. Systems of linked looping phreatic tubes can be found in the vertical space between each couple of successive low-gradient canyons. Each tube system develop below the floor of the upper canyon and the roof of the subsequent canyon These tubes are usually accessible from the floor of a notch in the lower part of the canyon. Parts of these phreatic systems become the guiding tubes (the roofs) of the following canyon underneath So, it seems that each canyon has associated a system of genetically related phreatic tubes just below its floor and usually above the following trunk Occasionally, some major tubes contain steep to-vertical segments that descend 80 m or more below the floor level of the parent trunk, covering the depth range of several trunks. There are some differences in the style of the trunk-tube systems between the upper, middle and lower parts of the explored cave system: (!) The two uppermost, and therefore oldest, relict trunks are characterized by the scarcity of notches except in the base of the canyon, where they acquire an exceptional development : here, two or three closely-spaced notches show remarkably sharp bevels that can reach as far as 30 m in width, the maximum observed in ACS. Below both canyons there are extensive systems of looping phreatic tubes with multiple connections to the notches of the canyon floor and even with some maze sections The loops of these tubes are usually restricted to a Sympos i um 7 : Ph y s i cal Speleo l og y 183

PAGE 197

1 ~ 100 m 0 legend: ~GUIDING PHREATIC TUBE ~~LOW-GRADIENT CANYON '\/ strike """-7dip Fig. 1: Profile view of the Agujas Cave system along the strike of the bedding. A schematic representation of the low-gradient canyons ls superimposed over over the survey frame. vertical range of a few tens of meters below the parent notch, but they can exceptionally reach maximum vertical ranges of 50 and 80 m, respectively, below the trunks. Most of these tubes remains undisturbed but some have been captured by younger vadose streams, that have cut narrow trenches in their floors. (2) The five levels of subhorizontal canyons situated at intermediate elevations inside ACS are characterized by the abundance of beveled notches along most of their vertical sections. The upper part of these canyons, close to the guiding tube, usually consists of a notch-free incised interval. The five canyons cover together about 150 m of vertical amplitude, each canyon being typically 20-30 m high The canyons of this part of ACS are well interconnected by major shafts, many of them active. With almost 2 km of surveyed passage, the topographically-lower of these five canyons is now the best known of ACS. (3) The two topographically lowest levels of ACS are poorly developed and also still poorly explored. Both are relatively narrow and show very few notches. The explored horizontal segments of these canyons frequently grade to phreatic conduits below the level of their floors The present base level drainage must be about 100 m below the lowest known low-gradient canyon, but to date exploration has failed to find access to the lower levels. The systems of relict low-gradient trunks and phreatic tubes may be affected by younger vadose captures that cut trenches in their floors. These trenches are narrow, with high-gradient floors, devoid of horizontal notching and usually linked to vertical shafts in both the upstream and downstream directions. Some of these vadose captures are relict features but many are active: small streams of aggressive water can be followed occasionally for more than 150 m in depth, creating shafts as far as 80 m deep and using parts of relict trunks and tubes, with or without trench development in them 3. Discussion and conclusions Origin of the low-gradient relict canyons The beveled low-gradient canyons represent relict water-table trunks, as demonstrated by: (I) the low-gradient character of the canyon floors, which transect steeply-dipping beds, cherty intervals and faults; (2) the presence of extensive levels of subhorizontal notches that bevel dipping strata (3) the abundant transitions to phreatic tubes at the canyon floors Furthermore, the deduced paleo-water tables show a general gradient towards the SW, i.e., towards the present base level. Along the strike the paleo-water-table gradients are usually very low, being remarkably higher in the dip direction, as expected by the lower hydraulic conductivity in this direction. Each relict canyon represents a period characterized by low rates of lowering of the water table (and thus of the base level). At the floor of the canyons, extended periods of constant base level are recorded, which allowed the formation of extensive corrosion notches at the water table and contemporaneous phreatic systems below the water table. The highly stable periods recorded at the canyon floors were followed by abrupt base level drops (15-60m), deep enough to cause the abandonment of the passage. Immediately after the drop, the flow was diverted to lower phreatic conduits that were rapidly downcut, producing a new canyon at a lower elevation. The downcutting of this new canyon is very rapid first, but later decelerates and reaches equilibrium again, creating wide notches at the water table and phreatic tubes below. This sequence of events repeated at least ten times, producing an exceptionally complete record of paleo-water tables that can be restored in geometry and evolution. 184 Proceedings of the 12 th Internat i onal Congress of Speleology 1997, Sw itz erland Volume 1

PAGE 198

Each relict trunk shows a characteristic cross section in terms of relative ele v ation and relative width of the notches and trenches. This canyon "fmgerprint" seems to be correlatable along the explored length of each trunk The vertical sequence of canyon fingerprints may be representative of the whole system and thus offer a basis for correlation with nearby systems Further exploration is needed, however, to link the known sequence with the lowest active level of the cave (the most obvious datum for correlation) and to clarify this sequence which at present shows some considerable gaps Coexistence of water-table passages with deep phreatlc loops The coexistence of water-table canyons and contemporaneous deep phreatic loops is a remarkable feature of ACS. In systems of low water-table relief such as ACS, the relatively low hydraulic head (the height of the water table above the spring) reduces the possibility of deep phreatic circulation and favors the development of water table passages Moreover in tightly folded rocks where fissure frequency is high, such as the Agujas Limestone water table caves are favored (FORD & WILLIAMS, 1989) These and other factors, such as the aggressiveness of the allogenic water or the siliciclastic (highly abrasive) composition of the abundant allogenic sediment s carried by the cave streams have possibly contributed to the predominance of water-table canyons in ACS Nevertheless abundant phreatic tubes some of them deep, were generated under the water-table canyons apparently at the same time, as the former are physically connected to the floors of the later The reason of this exceptional phreatic circulation under a low-relief water table must be the relatively high vertical hydraulic conductivity of the phreatic zone downdip the limestone: as the strata dip quite steeply bedding planes tend t o entrain groundwater to great depth. Other factor that could have favored deep phreatic circulation is the relatively high residence times of the water table in the same stable position (inferred by the well-developed beveled notches) coupled with the high erosive power of the allogenic water and sediments Significance The record of paleo-water tables, and thus of base level evolution, is exceptionally complete in ACS and represents a substantial vertical range (300 m) that further explorati o n will extend This complete record could be the reflect of ( I ) an exceptionally long history of evolution or (2) a highly-sensitive system to water table fossilization or (3) both factor s. Paleo-water tables are relatively easy t o defme and trace along the system. The combined presence of extensive sub horizontal beveled notches and ofvadose-phreatic morphological transitions (piezometric limits) at the same elevation are relatively conclusive criteria for paleo-water table definition. The levels to which vadose trenches are graded represent an additional criterion for paleo-water table definition but it should be carefuly employed to avoid consideration of trenches caused by younger captures. Moreover, the paleo-water tables defined exhibit a low gradient towards the base !eve~ cutting through structures and steeply-dipping strata, which strongly supports their interpretation as water-table features Acknowledgments ACS were discovered in 1993 and is actually being explored and mapped by a team of independent cavers which besides the authors, includes the following : R L. Anzuola, A Ortiz Salinas A Gonzalez and B. Rojo, with the aid of C Gibson R Close F.L. Quintanilla A Marcos, J. Matas, P Gomez, E. P e rez, P Keller and J. Borneman. This paper is a contribution to the IGPG project 379. The cooperation of the Consejeria de Media Ambiente y Ordenaci6n del Territorio (Junta de Castilla y Leon) is gratefully acknowledged. 380 Elevation in m above present base level 1 t 1 ____ 11' Elevation range of the ,&. /ow-gradient canyons 36 0 ~ o 320 i i ____ 300 t 3 __ __ 2 8 0 260 iA ___ 2 4 0 220 6 i 5 ____ L _____ r:fo1P 200 1 8 0 .!_0 1 4 0 120 lt9 ___ b.9 -Fig. 2: Proffle view of the the survey frame of Agujas Cave System In the dip direction of the limestone, showing the elevation range of the low-gradient canyons. Symposium 7 : Phys i cal Speleolog y 185

PAGE 199

References FORD, D.C. & WILLIAMS, P.W. 1989: Karst geomorphology and Hydrology Unwin Hyman London: 601 p VAN DE GRAAFF, W.J.E ., 1971 : Three upper carboniferous, limestone rich, high destructive, delta systems 400 0 with submarine fan deposits, Cantabrian Mountains, Spain. Leidse Geologische Mededelingen 46, 157-215 ROSSI, C. MUNOZ, A & CORTEL A (this volume) : Cave development along the water table in the Cobre System (Sierra de Peiialabra, Cantabrian Mountains, N Spain) STABLE PALEO WATER TABLES 380 Pauno~ 1 + 1 360 Oe1 0 340 ; Alambique o r~nico 320 300 280 260 240 220 o : .: a ---2 ___ -g=: :: _ Gusan~ o Alterofilia o --. [( 5 Vietnam/~ 6 _. oOrganico _,' '-o Ce 200 Boris o '' Paragenesis 5 --~ horn~~~~ ? 180 160 7 R i o Rojo 140 120 sR ----_ -9 ---Fo~ -9 -y ---sur o 0 . . .. \ ---Oso Richard o 2 tf o1P 7 ------:----? o 100 1 o s t ~ ~ ~---------1 .. .. .. Y ...... ~-o 0 PHREATIC TUBE 't? BEVELED CANYON PHREATIC CONNECTION 3 7 5 6 4 8 9 10 CANYON CONNECTION Fig. 3: Elevatlonal relationship for the various canyons and phreatlc tubes of ACS, showing the cross-sectional shapes of passages and the transitions between them. The diagram to the right Is a schematic plot of the relative Importance of each period of water table stabUlty deduced from the amplitude and persistence of the beveled notches and from the relative importance of associated phreatlc development below. 186 Proceedings of the 12 th International Congress of Speleology 1997 Sw i tzer l and Volume 1

PAGE 200

La corrosione di placchette calcaree ad opera di acque sulfuree: dati sperimentali in ambiente ipogeo Sa nd ro Gal d e n zi', Marco Menichetti 2 e Pa o lo Forti 3 1 lstituto Italiano di Speleologia, sede di Frasassi v.le Verdi,10 60035 Jesi Italia 2 Centro Escursionistico Naturalistico Speleologico Loe. Calcinaro 7/A 06021 Costacciaro Italia 3 lstituto Italiano di Speleo l ogia via Zamboni, 67 40127 Bologna Italia Abstract Sulphuric water corrosion on limestone in cave environment was measured through limestone tablets placed inside the Frasassi Caves. The room where the experience took place is interested by phreatic sulphuric water flow and shows many corrosion features and gypsum deposits due to mineralized water reaction with limestone walls. Some tablets were placed in the groundwater and others exposed to H 2 S vapours in the cave atmosphere; the experiment had a term of five years, and half of the tablets were weighed after two and half years. The surface corrosion resulted very evident in every tablet, and the weight loss reached values spreading from 15 to 20 mg / cm 2 / y; the limestone surface was covered by a thin layer of organic matter which surely plays an important role in H 2 S oxidation and in limestone corrosion In the groundwater the tablet surface was uniformly corroded; the weight loss resulted the same in the two different measurement stations and it remained constant during the whole period The tablets placed in the cave atmosphere were wholly covered with replacement gypsum, deriving from H S oxidation and reaction with limestone; the tablet surface under gypsum cover was deeply and quite irregularly corroded ; the weight loss shows an increasing velocity in time and reached the highest values, but it changes owing to the location of each tablet. lntroduzione Negli ultimi anni e stato fortemente rivalutato ii ruolo delle azioni speleogenetiche non collegate alla presenza di CO nelle acque carsiche : in diverse aree geografiche sono state segnalate importanti grotte la cui origine e stata imputata a fenomeni di ossidazione dell'H S circolante entro i massicci carsificati (EGEMEIER, I 981; HILL, 1987; GALDENZI & MENICHETTI 1995), mentre e stato ridiscusso ii ruolo dei fenomeni di ossidoriduzione dello zolfo (FORTI, 1988). La presente ricerca si proponeva appunto di verificare l'esistenza in grotta di azioni corrosive dovute alle acque sulfuree e di quantificarne l intensitil. Aspetti morfologic i Le Grotte di Frasassi (Appennino Centrale) rappresentano ii miglior esempio presente in Italia di grotte originate per effetto della circolazione di acque sulfuree all'interno di un massiccio calcareo (GALD E NZI, 1990); esse sono costituite da un reticolo molto ramificato, a tratti labirintico, di gallerie e sale disposte su piu livelli sub-orizzontali, ongmati m conseguenza dell 'approfondimento dell 'idrografia superficiale L 'evoluzione della grotta e stata piuttosto rapida, ed i piu antichi speleotemi trovati non superano i 200.000 anni di eta (TADDEU C CI et al., 1992) I livelli inferiori della grotta raggiungono la falda freatica, e qui possono essere effettuate inte1essanti osservazioni sui presenti processi morfogenetici. La speleogenesi e collegata principalmente all'ossidazione dell'H S nella parte superiore della falda freatica o nell'atmosfera secondo reazioni probabilmente catalizzate da batteri del tipo: H,S + 2 0 + CaCO + H 0 <====> CaSO, 2H 0 + CO Una caratteristica della grotta e costituita dalla presenza di depositi gessos1, ID tutti i livelli testimonianza della passata circolazione delle acque sulfuree II gesso risulta attualmente in formazione solo in ambiente aerato, dove le esalazioni di H S vengono a reagire con l'ossigeno atmosferico e con le pareti calcaree. Situazioni morfologiche favorevoli risultano le localitil in cui acque sulfuree risalgono da condotti con lunghi tratti sommersi. Le pareti direttamente esposte ai vapori sono interamente ricoperte da gesso, che si presenta per lo piu in forma microcristallina bianco, molle e lattiginoso ; nelle zone limitrofe ii fenomeno si attenua e sulle pareti possono comparire grumi gessosi piu isolati (Fig. l); queste azioni possono tuttavia rimanere attive anche a oltre 50 m di distanza dalle acque sulfuree secondo la locale circolazione dell'aria La formazione de! gesso e Jimitata agli ambienti aerati mentre sottofalda esso viene disciolto ed asportato. Le pareti calcaree in cui si e avuta la formazione di gesso risultano interessate da alveoli e fossette, di dimensioni centimetriche, ben evidenti dove ii gesso e stato dilavato dalle acque percolanti (Fig 2). Le zone soggette alla circolazione di acque sulfuree sono estremamente ricche di materiale organico presente in piccoli accumoli sapropelitici in falda o nelle diffuse vermicolazioni Mollo comuni anche organismi superiori, piu abbondanti che nel resto della grotta. Le modalita dell esperimento La misura dell'entita della corrosione in aree carsiche e stata oggetto di numerose ricerche condotte con differenti metodologie. La misura della perdita di peso di placchette calcaree esposte agli agenti atmosferici (TRUOOILL 1975) e Symposium 7: Physical Speleology 18 7

PAGE 201

F i g I : gesso i n for m azio n e su ll e pareti c a /caree sembrato un metodo idoneo ad essere impiegato anche in grotta, in condizioni ambientali difficili, per verificare l'esistenza di azioni corrosive dovute alla circolazione di acque sulfuree, e per quantificame l'intensita. L'esperimento si e protratto per 5 anni, ed a meta periodo e stata effettuata una prima serie di misure (GALDENZI & MENICHEITI 1994). Un altro vantaggio legato all'impiego di questa metodologia e costituito dalla disponibilita di una ricca documentazione sui risultati conseguiti in precedenti esperienze condotte in superficie (GAMS, 1985; STEFANINI et al., 1985). La stazione sperimentale e stata installata nel Crepaccio Sulfureo della Grotta de! Fiume; l'intensa circolazione di acque sulfuree, l'esistenza di un ampia interfaccia tra la falda freatica sulfurea e l'atrnosfera della grotta, la diffusa formazione di gesso sulle pareti hanno infatti consentito di riconoscere condizioni ottimali. el sito prescelto sono state realizzate tre stazioni di misura appendendo le placchette con filo di nylon (Fig. 3); nelle stazioni subacquee sono state collocate almeno 2 placchette, mentre in ognuna delle 2 stazioni aerate sono state appese 4 gesso Fig. 2 : mo r fol ogie corros i ve su ll a p a rete ca/c a rea evidenziate d a / dilava m e llf o de/ gesso placchette. Un'altra placchetta e stata seppellita sotto gesso molle in formazione. Le placchette, dalle dimensioni di circa cm 8 x 4 x I sono state levigate e misurate per calcolame la superficie laterale. II peso e stato determinato con precisione di 0.1 mg, previo essiccarnento in condizioni standard a 110 C. Esse sono state realizzate con Calcare Maiolica (Cretacico inferiore), litotipo affiorante nell'area: si tratta di calcare micritico assai puro, privo di impurita e di strutture inteme che avrebbero potuto complicare le operazioni di preparazione e misura delle placchette stesse. Le placchette sono state posizionate nel marzo 1990 e in ciascuna stazione sono state prelevate meta delle placchette dopo due anni e mezzo nel settembre 1992 Le rimanenti sono state ritirate nel marzo 1995, dopo 5 anni di / Fig 3 Grotta de/ Fiume ramo sulfureo schema s i to sperimentale 188 Proceedings of the 12 th International Congress of Speleology 1997, Switzerland Volume 1

PAGE 202

Fig. 4: Placchette esposte per 5 a1111i ai vapori di HiS, co11 gesso di 11eoformazio11e Fig. 5 : Superficie delle placchette dopo ii lavaggio de/ gesso di 11eoformazio11e Fig. 6 : Placcchetta dopo 5 a1111i di immersio11e in acque s11/f11ree; 11otare le ve11e di calcite in rilievo esposizione; una placchetta e andata perduta per la rottura del filo fortemente indebolito dalle esalazioni sulfuree Le placchette esposte all'aria erano interessate dalla fonnazione di gesso giil sei mesi dopo l'inizio dell'esperienza ; al momento de! recupero era presente uno spessore variabile di gesso, sia in fonna bianca microcristallina, che in fonna cristallina, generalmente in superficie o dove ii fenomeno era meno sviluppato (Fig. 4). Prima di pesare le placchette e stato necessario disciogliere ii gesso in acqua distillata saturata con polvere di calcare fino a scoprire la superficie calcarea, che appare corrosa in maniera molto eterogenea con "cariature" profonde alcuni millimetri (Fig. 5) Le placchette erano rivestite da patine scure, costituite da materia organica interposte tra calcare e gesso. Le placchette lasciate in falda erano invece coperte da uno spesso strato di materia organica, in cui prosperavano anche vermi in corso di studio. La superficie calcarea e unifonnemente corrosa, priva di gesso, con vene di calcite spatica in rilievo per corrosione differenziale (Fig. 6 ). La presenza di evidenti filamenti di materia organica sulla superficie calcarea delle placchette e poi stata confennata dalle osservazioni al microscopic elettronico (Fig 7) Risultati dell'esperimento I risultati delle misure relative alla perdita di peso riportati in tabella I secondo diverse unitil di misura giil utilizzate in bibliografia, abbinati alle dirette osservazioni effettuate, consentono di trarre alcune significative conclusioni Fig. 7: superficie delle placchette al microcoscopio elettronico (11otare la presenza difilamenti di materia organica) it1falda (a sinistra) 11el/'atmosfera del/a grotta (a destra) Symposium 7 : Physical Speleology 189

PAGE 203

Placchetta mg/cm /anno mg 10 /cm '/ di mm/I 000 anni anni Posizionamento 5 14 14 38, 75 55,6 2,5 falda (Staz B) 9 I 5,24 41,77 59,9 2,5 falda (Staz. A) 7 14,48 39 68 56,9 5 falda (Staz. A) 11 15,29 41,89 60, I 5 falda (Staz. A) 3 15 17 41 56 59 7 5 falda (Staz. B) 8 11 01 30, I 6 43,3 2,5 aria (Staz A) 10 6 97 19,09 27 4 2 5 aria (Staz. A) 13 8 59 23,55 33 8 2,5 aria (Staz. C) 14 16 27 44,56 64,0 2,5 aria (Staz. C) I 17 46 47,82 68, 7 5 aria (Staz. A) 12 13 ,89 38 06 54,6 5 aria (Staz A) 2 21 56 59,12 84 8 5 aria (Staz C) 15 17 6 48,23 69 2 2,5 in gesso Tab. 1risultati prove di corrosio11e delle acque sulfuree sulle placchette calcaree Le placcbette ubicate sottofalda presentano un uniforme grado di corrosione, intorno ai 15 mg/cm 2 / anno, indipendentemente dalla stazione in cui sono collocati; la perdita di peso per quelle esposte ai vapori di H,S risulta invece molto piu variabile in conformita a quanto riconoscibile anche macroscopicamente. La velocitil dei flussi freatici (maggiore nella stazione B che in A) non ha influenzato i risultati per le placchette ubicate in falda Al contrario nell'atmosfera i risultati variano in funzione de! tipo di circolazione idrica ; nella stazione C, ubicata allo sbocco di un canale sommerso con scorrimento rapido ed abbondanti esalazioni i valori risultano piu elevati II grado di corrosione in ambiente freatico e rimasto costante anche in rapporto al tempo di esposizione; le placcbette esposte ai vapori di H S banno subito azioni corrosive di intensitil crescente nel tempo ; i valori medi calcolati su base annua per le placcbette dopa 5 anni di esposizione risultano pari a circa ii 160 % di quelli calcolati dopa 2 5 anni In termini reali la perdita di peso per le placchette esposte 5 anni e risultata in media di 88 2 mg/cm 2, contra i 26 8 mg/cm delle placchette prelevate dopa 2 5 anni ; ipotizzando che ancbe le placchette esposte per 5 anni abbiano subito nel primo periodo una perdita di peso media pari a 26 8 mg/cm ', risulta che nella seconda frazione dell esperimento la corrosione e stata di 61.4 mg/cm cioe oltre due volte maggiore I presenti dati indicano che le reazioni di ossidoriduzione dello zolfo in atrnosfera si sono attivate piu lentamente che non in falda, verosimilmente per i tempi piu lunghi necessari allo sviluppo di cons i stenti colonie batteriche sulle placchette esposte all 'aria Gli alti valori misurati per la placcbetta collocata entro gesso di neoformazione giil esistente sulle pareti della grotta concordano con questa interpretazione Le azioni corrosive in atrnosfera ad opera di vapori di H,S sono risultate piu rapide che non in falda; e bene tuttavia ricordare che questi fenomeni banno una elevata intensitil solo in ristrette localitil, prossime alle polle di acqua sulfurea, e calano rapidamente allontanandosi da esse. Ringraziamenti Un sentito e doveroso ringraziamento ad ELISA ROMINI, che ha realizzato le foto al microscopio elettronico e collaborato alla loro interpretazione. Opere Citate EGEMEIER, S.J. 1981. Cavern development by thermal waters National Speleological Society Bulletin 43 : 31-51. FORTI, P. 1988. The role of the sulfide-sulfate reactions in the speleogenesis. Proc. 1st Congr ofFEALC, Ouro Preto, Brasil. GALDENZI, S 1990. Un modello genetico per la Grotta Grande de! Vento In : (S. Galdenzi & M. Menichetti ed.): II carsismo della Gola di Frasassi. Memorie Istituto Italiano di Speleologia s II, 4 : 123-142. GALDENZI, s & M. MENICHETTI. 1994 Misure sulla corrosione di placchette calcaree ad opera di acque sulfuree: primi dati sperimentali. Atti XXVII Congresso Nazionale di Speleologia. Castelnuovo Garfagnana, in stampa. G A LDENZI, S. & M. MENICHETTI. 1995 Occurrence of hypogenic caves in a karst region: examples in Central Italy En v ironmental G e ology 26: 39-47 GAMS I. 1985 International comparative measurement of surface solution by means of standard limestone tablets Ra z pra ve i v. razreda XXVI: 361-386 HILL, C.A 1987 Geology of Carlsbad Caverns and other caves of the Guadalupe Mountains, New Mexico and Texas. New Mexico Bureau Mines & Miner. Resources Bull 117: 150 p. STEFANINI, S ; ULCIGRAI, F.; FORTI, F & F. CUCCHI. 1985 Resultats experimentaux sur la degradation des principaux lithotypes du karst de Trieste. Actes XVI Congr. Nat. de Spel. (NancyMetz, 1985). Spelunca Memories 14 : 91-94. TADDEUCCI, A.; TUCCIMEI, P & M VOLTAGGIO. 1992. Studio geocronologico del complesso carsico "Grotta del Fiume-Grotta Grande de! Vento" (Gola di Frasassi, AN) e indicazioni paleoambientali. Il Quaternario (5) 2 : 213-222 TRUDGILL, S.T. 1975. Measurement of erosional weight loss of rock tablets.Brit Geomorph. Res Group Tech. Bull. 17: 13-19 190 Proceedings of the 12 th Internat i onal Congress of Speleology, 1 99 7, Switzerland Volume 1

PAGE 204

Caves climatic systems Bulat R. Mavlyu do v Institute of Geography Russian Academy of Sciences Staromonetny per. 2 9 Moscow 10901 7 Russia Abstract Cave climate is not homogeneous It is possible to choose three basic caves climatic systems in dependence on cave morphology and ventilation (for caves without water) : of horizontal caves (with entrances at different levels), of inclined descending and ascending caves In caves of each climatic system there are from 3 to 5 climatic zones Size and completeness of collection of climatic zones depends on outside climate and cavern morphological peculiarities Climate of caverns with complex morphology will have combinations of basic climatic systems Cave climatic system can be complicated by accumulation of perennial snow and ice or by flowing water or completely disturbed by them 1. Introduction Investigations of last years shows that we can't compare small caves climate with large caves climate. There are many caverns with very big length and big difference between entrances and its lowest parts. Some of these caverns dissect a few surface climatic belts. Climate of such caverns have big unhomogeneous. In some cases we need to say about microclimate of different parts of caverns. If we carefully consider climatic structure even of not very big cave we understand that it is not very simple. Traditional point of view on cave microclimate not allow us to do climatic description for big quantity of big caverns. In this report we discuss common principles of caves climatic structures (caves climatic systems) for caverns with different morphology and size. 2. Factors that define cave climate They are : climate of area where the cave is situated; cavern morphology that determines system of air circulation in it; geothermal conditions of the rock massif where cave is situated; system of water streams flows in caverns ; quantity of precipitation that penetrate in caves (first of all snow). We shall discuss influence of these factors on climatic conditions in caverns in brief. Air temperature and humidity wind velocity and direction, atmospheric pressure will be more important factors of external climate, that have influence on cave climate The warmer external climate will be the higher cave temperature will be. There are mean year air temperatures (MY AT) up to l2C in caves of south slope of the Big Caucasus (TINTILOZOV 1976) In the middle Ural cave MY AT is about 5-6C There is lower value of air humidity in the caves in the dry regions but in the wet regions humidity of cave air is higher Strong summer wind that blows in the entrance of small cave can be the reason of cave climate similar to external. Sudden change of atmospheric pressure leads to appearance of air flow in or out of the cave Air movement is the more significant agent for transmission of external climate influence inside the caves In is possible to choose two main mechanisms of air movement inside the caves (MAKSIMOVICH, 1963) : l) There is chimney effect of air movement in caverns with some (two or more) entrances at different altitudes The reason of it lays in difference of air columns weights near entrances. There are air movement from lower entrance to upper in winter and from upper to lower in summer (LISTOV, 1885) Air flow in the cave changes its direction when surface air temperature will be equal to one in the middle part of the cave (neutral zone DUBLJANSKI, 1977) 2) In inclined caverns air moves because of difference of air density in different parts of the cave Cold external air in winter ejects more warm cave air in inclined descending caves. Warm more light air in summer ejects more cold cave air in inclined ascending caves (MAKSIMOVICH, 1963). Air movement stops in both cases when external and cave temperatures will be equal. The air movement in caves can occur also when wind blow in cave entrances, when moving water takes air with it, when external atmospheric pressure suddenly changes. Temperature of rock massif where cave is situated also has big influence on cave climate External climate is transmitted inside the caves by cave wind It is superimposed on the climate that rock define in the cave Water streams also have influence on the rock temperature around the cave We can read in scientific literature two main opinions about temperature inside the caves: I) MY AT in neutral zone in caves is equal to external MY AT (PULINA, 1974) and 2) cave MYAT not depends on external MYAT ; it is determined by cave morphology and peculiarity of air changing with the surface (DUBLJANSKI & LOMAEV, 1980) We can t agree with second opinion because of it is well known that cave temperature decrease with elevation (MAKSIMOVICH, 1963). However we can observe equal values of cave MY AT and external MY AT not everywhere. From hydrogeothermy we know that rock temperature at the neutral layer of the earth (layer with minimum year amplitude of temperature oscillation) have value on some degrees higher ofMYAT (FROLOV, 1976). This layer usually situates on depth about 15-20 m lower day surface. For the area of the former USSR temperature difference changes from 2 to 6C. Rock temperature increases with depth on I degree on each 30-40 m If we shall imagine closed cavern that haven't contact with day surface air temperature will be equal to the rock temperature in it. The dipper cave will be the higher temperature will be in it. We can expect that temperature in deepest parts of real caves (without water) will be almost equal to surrounding rock temperature Investigations in some caves show that MY AT in their neutral zones are lower the rock temperature. For example, this difference in the Kungur Cave is about l,0C and in the Cave Sumgan-Kutuk (the Ural) is about l,5C. Caves usually situate in regions with not equal difference between external MYAT and rock temperature of neutral layer Caves locate also at different depths from day surface. Therefore it is difficult to expect equal regularities between external and cave temperatures For example, MY AT in caves neutral zones in Carpatians and Podolia is almost equal to theoretical value of Symposium 7 : Phys i ca l Speleo l ogy 191

PAGE 205

rock temperature. It exceeds external MY AT on 3.7 C. This difference in the Kungur Cave consist about 3 C for ventilated parts and 4 C for not ventilated parts of the cave In caves and mines of the Pamirs this difference is about 7 C. There are regularities of regional changing of cave MYAT in their neutral zones Some empirical equations for calculation of thermal conditions in the caves were received on basis of statistical analysis of cave temperatures data. For North America equation is next (MOORE & SILLIAN 1978): T, = 38 0.6L 0.002h, and for Europe it is (CHOPPY, 1977) : T, = 54.3 0.9L 0.006h, where L is geographical latitude, degrees; h is e le vation, metros. If we shall base on previous discussion it is possible to suppose that such equations have only estimative character. The fact that MY AT in big quantity of large caves are higher than external MYAT allows to say that mean year heat and moisture flows have constant direction oriented outside from these caves. Since heat capacity of water is approximately in 30 times higher than one of air it is possible to say that influence of water on cave climate is more intensive than influence of air on it. In many European caves in low mountain relief influence of water lead to equality of external and cave MY AT. On the contrary, in high mountains it is typical for caves with flowing melt water that caves MY AT are lower than external ones. The supply of caves by thermal water leads to exceeding cave MY AT above external one Seasonal water inputs into caves (floods) lead to warming of cold zones in some parts of the caves and co lding of warm zones in another. Snow is accumulated in some caves Snow is one of sources of coldness in the caves Rain as a rule warms cave climate ow we shall consider climatic systems of typical caverns (without water): "horizontal" (the caves with some entrances at different elevations) and inclined (bag type) caves with one entrance. We can divide second type into two groups : caves with entrances in its upper end and caves with entrances in its lower end. Climate of caves with other morphology we can ascribe to one of main types or their combinations. 3. Horizontal caves We can divide horizontal caves into 5 climatic zones (beginning from lower entrance) : I Lower transitional zone (L TZ); 2 Cold temperature anomaly zone (A-) ; 3. Constant temperature zone (neutral zone) (CTZ); 4. Warm temperature anomaly zone (A+); 5 Upper transitional zone (UTZ). The fist who said about zones A+ and Ain caves was Russian scientist V.S LUK.JN, 1965 Presence of zones Aand A+ in such caves come in contradiction with conception of existence of equalizing zone between external climate and cave conditions of V . DUBLJA SKI, 1977 and others. Lower transitional zone (L TZ) L TZ situates near lower cave entrance During the cold period of year (between two changings of air movement direction) external climate have influence on L TZ by air flow into the cave, by solar radiation and outside wind Air temperatures in L TZ in spring and autumn are positive, in winter they are negative and almost equal to external temperature In summer air temperature in L TZ is alittle above zero because of air moves here through cold zone A-. Size of L TZ is different in all caves and depends on dimensions of caves channels, intensity of air circulation, entrance orientation, directions of main winds in surrounding area, quantity of snow at entrance of cave Big part of year external air humidity is higher than caves one. I.e moisture flow has direction into the cave and only in first part of winter it has direction from the cave. There are typical values of MYAT : T >T and humidity : E >E Its means that during the year L TZ receives more coldness than warmth, and mean year moisture flow have direction into the cave. Cold temperature anomaly zone (A-) Zone Ahas influence of L TZ during the winter and influence of CTZ in summer. Boundary between zones Aand CTZ lay on the line with year temperature oscillations about 0 5 C. Air temperature in zone Achanges from the lowest value at the boundary with L TZ (T 2 T 4 =T 3 ) and outer boundary is on maximum va lue of cave MIAT (T,>T.>TJ Moisture flow on the boundary with CTZ has direction inside cave in all seasons of year (fig. I a). On the boundary with UTZ moisture flow in winter has direction out of cave but in summer it has direction inside cave MY AT in zone 192 Proceedings of the 12 h Internat i onal Congress of Speleology 1997, Switzerland Volume 1

PAGE 206

A+ are above rock temperature It connects with situation when heat flow from boundary with UTZ in summer not compensates heat flow from CTZ in winter. The zone A+ length depends on fissures and pits width This length, as a rule, is smaller than one of zone A-. Since maximum amplitude of rock massif cooling is measured from T 3 (but not from external MIA1) on absolute value it is higher than maximum rock massif warming in summer (Fig. 2). Because of it duration of periods of cave warming and cooling are different (also as caves wind velocities in summer and winter) For example, in the Kungur Cave duration of winter air circulation consists about 200 days (from 1969 to 1993) and duration of summer air circulation consists about 150 days (DOROFEEV & MA VL YUDOV, 1993). Melting and condense water from UTZ has influence on climate of zone A+ in winter. Flow of warm wet air from zone A+ through UTZ in atmosphere gives white jets in hard frost (they are like a smoke column) Warm wet air have contact with rock and snow. As a result melt water penetrate into UTZ and A+. Water move down and is warmed from air and rock In summer warm wet air from surface through UTZ moves into zone A+ where it has contact with more cold walls. In result of condensation of vapor the additional heat penetrates in the cave. U pp er tra n sitio n al zone (UTZ) UTZ is the equalizing zone between zone A+ and external climate Situation in UTZ is similar to one in L TZ with difference that it has contact with not Abut with A+ There are only some measurements of temperature of air that flow out from UTZ in autumn and winter. Temperature 6.8C was in October 1986 in air flow from Kungur Cave when external air temperature was 0.3C In February 1987 temperature of air flow from the cave was about 4.5C In Bzybskij Ridge (the Caucasus) in January 1980 temperature of air flow from the rock under snow cover consists 2 2C at elevation 2000 m. At the upper part of mountain Razvalka in the North Caucasus temperature of air flows from rocks consisted 6C but mean winter temperature was about 2.7C (LUK.IN, 1990) Cooling of UTZ is possible in winter by falling and blown snow. We can see moisture and heat flows on Fig. la. Next expressions are truth for winter: T T and E E , and for summer: T >T E T ,. In autumn air circulation is ended when T 2 >T Sympos i um 7 : Phys ic al Speleo l og y 193

PAGE 207

5 Vertical caverns Pits is IDC but with almost vertical incidence. Air movement, T and E in them are similar to ones in IDC Cold air usually penetrates in pits though center of shaft and more warm air lifts at perimeter of it However if pit cross sections on different depths are not equal replacing of cold air stream occurs to one of pit wall. Snow accumulates in pits in winter. Snow in pit can melt or not in dependence from pit depth, degree of its cooling, quantity of falling and blown snow in it, quantity of heat that penetrate in pit with rain. Snow is additional source of pits cooling. The more snow gets into pit the more long time snow remains in it Big quantity of snow in pits guarantees conservation of zero temperatures in pits during all warm period of year Very big depth of pits is not favorable for snow conservation in them because snow gets in warm part of zone and melts quickly In the dry dip caverns T3 can grow with depth in accordance with geothermal step of that area 6. Caves wi t h water Water streams have big influence on caves climate. The more water gets into the cavern the more homogeneous cave climate becomes All before mention climatic zones can disappear because water flows warm caverns in winter and cool them in summer. Through cave with constant big water flow will contain CTZ and seasonal small UTZ and LTZ Caverns lost all or part of cold accumulated in winter if they have small constant or episodic water flows in summer In these caves zone Awill be almost similar to L TZ Water in vertical caverns with big depth has especially big influence on caves climate. Melt water strong cools caverns and difference (T T.) becomes considerable So in the shaft Kujbyshevskaja (the Caucasus) at the depth 800 m from surface cave temperature is about 2C (DUBLJANSKI & KIKNADZE, 1984) (external MY AT at this elevation is about 7C) In the shaft Snezhnaja (the Caucasus) temperatures are : T =4.5C at depth 700 m (T =6C) and T =6 2C at depth 1370 m (T =10C). In the spring Mchishta where shaft Snezhnaja has discharge water temperature is about l0C (T =l4C) It is possible to calculate changing T with depth (i e. vertical temperature gradient VTG) for CTZ in climatic system of shaft Snezhnaja. It consists about 0.25C on each l 00 m of depth (let's compare it with VTG of Carpatians caves 0.4C / 100m and VTG of free atmosphere 0 6C / 100m). Peculiarity of climatic systems in big vertical caverns consists in air movement in them It depends on: difference of altitudes between entrances, capacity of water to take air with it, inversion of climatic zones in caverns when there is zone (because of snow) near upper entrance of "horizontal" cavern, but at lower entrance exists more warm zone. Sometimes descending air movement in upper part of cave system will be in opposition to ascending air movement from lower entrance in summer. Increasing of T in CTZ with depth leads to arising of ascending air movement under roofs of galleries (even when lower entrance is absent). As a whole climatic systems of large comp l ex caves still wait their investigators 7. Conclusion Cave climatic systems is not something constant. Changing of external climate has influence on caves climate First of all length of zones A+ and Achanges from it. External climate warming leads to decreasing or decay of zones Ain some caves. Calculations using mathematical model of air circulation in the Kungur Cave show that decreasing of external MY AT on I C lead to increasing of length of zone Ain the cave on 20-30 m (MA VL YUDOV, 1985). But not only natural changing have influence on cave climate. Anthropogenic factors have big influence on it in last years. Next factors have influence on air circulation in the caves : placing of entrance doors, changing of cross sections of caves galleries, changing of groundwater levels It leads to changing of cave climate and first of all dimensions of zones Aand A+. Since there are not regular meteorological measurements in majority of russian caves observations of dimensions of zones with ice can be good indicators of natural or anthropogenic changing of cave climate. We discussed climatic systems of main morpho l ogical types of caves only in common features. More detail study of them still wait own investigators R efe r e nc es CHOPPY, J. 1977. La temperature des cavites en fonction de la latitude et de !'altitude Spelunca, 3: 117-1 I 8. DOROFEEV, E.P. & B.R. MAVLYUDOV. 1993 Dynamic of Kungur Cave glaciation Peschery, Perm, 23-24: 131 140 (in russian) DUBUANSKI, V.N. 1977. Karst caves and shafts of the Mountain Crimea Nauka, Leningrad, 182 p (in russian) DUBLJANSKI, V.N. & AA. LOMAEV 1980 Karst caves of the Ukraine Naukova Dumka, Kiev, 180 p. (in russian) DUBLJANSKI, V.N & T.Z Krl
PAGE 208

Some concepts about heat transfer i n karstic systems by Pierre-Yves Jeannin 1 2 Rudolf Liedl 2 & Martin Sauter 'Center of Hydrogeology, University of Neuchatel, Rue Emile-Argand 11, CH-2007 Neuchatel, Switzerland 2 Applied Geology, University of Tuebingen, SigwartstraBe 10, D-72076 Tuebingen, Germany Abs t rac t The conceptual model of heat transfer in karst systems presented here is based on the existing models for the saturated zone and on a simple air flow model for the unsaturated zone. It is in good agreement with the fields observations. As a first approximation, it can be used to interpret the temperature responses of karst springs The shape of the temperature pulses provides information about the geometry of the conduit network within the saturated zone The amplitude of the temperature pulses provides informations about water flow velocities in the unsaturated zone The positive/negative temperature pulses ( warm floods in summer and cold floods in winter) provide information about the depth of the phreatic wne below the ground surface Resume Le present modele conceptuel des transferts de chaleur dans les systemes karstiques se base, pour la zone saturee sur les modeles existants et, pour la zone non saturee sur un nouveau modele simple des circulations d air Ce mdoele est conforme aux observations de terrain En premiere approximation ii peut etre utilise pour interpreter la reponse thermique des source s karstiques I..aforme des impulsions de temperature foumit des informations sur la geometrie du reseau de conduits karstiques dans la zone saturee L 'amplitude des impulsions de temperature donne des informations surtout sur la vitesse verticale de l eau dans la zone non saturee L existence d'impulsions positives ou negatives de la temperature a la source donne des informations sur la profondeur de la wne saturee sous la surface du bassin versant. 1. Introduction Temperature records of karst-spring waters can easily be measured with a high degree of precision Until recently in terpretations of this parameter have been restricted due to the absence of a physical concept of heat transfer in karst. 81:NDERnTER et al (1993) present a first theoretical concept and model for heat flow processes in carbonate sys tems. This model considers a single karst conduit which transmits waters rapidely to the spri n g and exchanges heat with the rock matrix. Results show that temperature pulses are only slightly modified by the flow through the conduit. RENNER ( 1996) has developed a similar approach for mo delling beat transfer in the saturated zo n e (flooded conduits). He also observed that temperature pulses due to recharge event s can propagate far across the saturated zone before they dissipate ; the shape of the pulse is modified along the under ground travel depending on the geometry of the conduit net work ( lIEDL et al 1997). Such a model allows the evaluation of the effect of different geometrical characteristics of the conduit network on spring temperature. Applied to a real karst system, volume, s u rface and openings of conduits could be estimated It is pointed out that the sensitivity of temperature responses on geometrical parameters of the conduit network is very high Nevertheless, in this model, the shape of the input temperature pulse is considered as a ca l ibrating parameter Heat processes within the saturated zone appear to be well described by the models above. However there is still no approach which allows the physical modelling of the beat transfer processes in the karst u nsatura t ed zone The aim of this paper is to present an overview of some general concepts and physical background which can be applied to model heat transfer in karst systems 2. The conceptual model Basic simplifications Several diffe r ent assumptions have to be made in order to be able to calculate heat flow transfer in karst systems : I ) Conduits are considered as a network of one dimensional pipes ( no change in fluid velocity at right angle to the flow ). 2 ) Flow velocity within the matrix is low i e it can be neglected ; heat tran s fer in the matrix can then be described b y conduction onl y. 3) Conduit walls have simp l e geometry (circular or rectangular cross sections) These assumptions apply to both the unsaturated zone and the saturated zone The phases that have to be considered are rock and water in the saturated zone and rock air and water in the unsaturated zone Other models can be considered depending on the aims of the studies Our model gives a first global approximation of the heat transfe r withi n complete karst systems It is not the aim of the st u dy to explain loca l microclimates in cave s Boundary conditions Two types of boundary conditions are considered (figure I). 1) The deep geother m al heat is coming from sources deeper than the karst system. In continental lithosphere geother mal heat flow ranges between 40 and 70 kW / km 2 depending on the position on the Earth surface The heat results of ra dioactive decay within the continental crust and of tempe rature gradient between crust and mantle This implie s that steady state conditions prevail independently from surface features like karst systems Here we assume that geother mal hea t flow is eq u al to 60 kW/km 2 It can be demonstrated that geothermal heat flow is very low with respect to water flow through most karst systems In temperate areas the specific discharge of karst springs ranges betwee n l and I 00 l/s/km 2, corresponding to 3 I 5 l 0 4 to 3 L 5 I 0 6 m 1 /km 2 / y On the one band, the geothermal energy is I 9 I 0 9 J per year and per km 2; on the other hand the energy needed to heat I m 1 of water by I C is 4 19 I o J meaning that the geot h ermal heat flow can hea t approximately 453 m 1 of water by I C per year. Applied to a karst specific discharge this flux can heat karst waters by less than 0 1 / y Similar values stem from BOF.GU ( 1980) and MATHEY ( 1974). The consequence of this low heat flow is that the lower li mit of karst systems can be considered as a no heat flow boundary as well as a constant heat flow boundary The model results will hardly be affected by the geothermal flow S ympo s iu m 7: P hysical Spel eology 1 95

PAGE 209

mass conservation: a aiJ p dQ + J p v o C :i: (6) with v=v 0 n is a nonnal vector to I:, l: is the wall surface of the conduit segment and Q its volume momentum conservation: :tJ p V dQ + J (p v V) n C :i: pn~ -JpgdQ= 0 :i: C (7) with g=gravitational acceleration and T =viscosity stress tensor energy conservation (see equ. 1 I for Qv,) : a v 2 atJ p(U + 2 )dQ-Jv(r n)-~ -J p v gdQo :i: Q Ji;~-JQ"' dQ0 :i: Q perfect gas equation: P=pRT with R=gas constant internal energy equation : 5 U=-RT 2 Such a problem has to be solved numerically. (8) (9) (I 0) and Q. in equ 8 can be considered as boundary condi tions of the conduit system. Meanwhile, to solve the problem correctly, one has to consider conduit and matrix as a system, meaning that has to be calculated using equ. l. The full system is then made up of six nonlinear equations. This is a very complicated problem to be solved by computer programs A further problem is introduced by air humidity (IJI). Humid air contains more energy than dry air at the same temperature. This is due to the energy needed to evaporate the amount of water enclosed in the air. It can easily be shown that a normal atmospheric vertical temperature gradient for dry air is 9.76C/k.m, and only 5.1 C/km for humid (saturated) air (see ROEDEL 1994) Air within karst conduits is almost saturated, but atmospheric air, entering the conduits is usually unsatura ted This has to be introduced into equation 8 in the last term : ( 11) It becomes apparent that equation 5 is a special case of equation 8, except for the hydrodynamical dispersion which is not considered within equation 8 First results and comparison with field data So far, the conceptual model described above has not been coded into a computer program However, two parts have been modelled separate[ y : the phreatic zone (RENNER 1996) and the unsaturated zone in a simplified manner (only air circulation using equations 6 to JO). Results from the unsaturated zone model show that air tem perature is mainly influenced by three factors : the elevation, the humidity (term 5 in equ. 8) and the heat exchanges between phases (term 4 in equ 8) Flow velocity or cross-sectional variations have only a limited effect on the air temperature Heat exchange between air and cave walls A change in value (energy) in term 4 of equation 8 induces a direct effect on temperature of the air. This shows that the air temperature is strongly influenced by the heat exchange with the walls (and/or water) WIGLEY & BROWN ( 1971) have shown that air temperature equilibrates with walls (and/or water) over a short distance in flow direction, the dis tance being dependent on the flow velocity and the contact surface between air and walls (and/or water) It can be observed at the cave entrances that the distance, along which tempera ture significantly varies along the year is generally quite short. It is known as the heterotherrnic zone This zone seldomly extends over more than 50 to 100 meters If the air flow is very high and the passages very wide, the equilibration distance can reach about 500 meters. Fields observations demonstrate that this principle can also be applied to water temperature, but so far a theoretical approach of this process is still lacking Evaporation/condensation processes may play an important role. Vertical air temperature gradients and humidity Vertical gradients found by our model are 9 76C/km for dry air Humid air gradients are calculated introducing equation 11 in term 5 of equation 8. The calculated value is then 4.80C/k.m what is not exactly the value expected from the literature (5 10C/km). This might be due to different physical constants used in both calculations. Temperatures have been measured in more than 50 vertical caves around the world Gradients ranges between 2 and 5C / km, most of them being very close to 5C/km (figure 2). Lower gradients observed can generally be related to local conditions like the presence of snow in cave entrances (shafts) which normally cools the cave down for some hundreds of meters. In this case, the heat exchange with the conduit's wall (the snow) is not equal to zero ( in equation 8), but is highly negative. In such cave sections, this effect leads to cooler air temperatures than the expected outside mean annual temperature, but this leads also to lower vertical gradients (less than 5C/km). If conduits are vertical cave temperature can be affected down to 200 or possibly 300 meters. Such a case has been observed by JEANNtN(1990). Heat exchanges within the phreatlc zone Within the phreatic zone heat exchange between water and solid rock is of the same type as for air and solid rock However with water having a larger heat capacity, any disequilibrium between water and rock temperature tends to extend over much larger distances (kilometers) BENDERITTER et al and RENNER's models are capable of si mulating pretty well the observed temperature variations at springs. However, in both models, temperature of water en tering the phreatic zone, which depends on the processes described above within the unsaturated zone is a calibrating parameter The model presented here provide independent information on this input temperature. Temperature of water entering the phreatic zone In both saturated models the input records were different: RENNER used only cold inputs and BENDERITT'ER cold in winter and warm in summer In fact these different inputs are in agreement with our conceptual model. Our model assumes that temperature of air and water de pends on the season down to a depth of about 50 m, possibly I 00 m below the surface (heterothermic zone). In the catch ment studied by BENDERflTER the water table is approximately 10 meters below the surface Therefore, it has to be expected that spring water temperature are season dependent. 196 Proceedings of the 1 2 1 h Internat i onal Congress of Speleo l ogy 1997 Sw i tzerland Volume 1

PAGE 210

400 m t Air flux >1 m3 s-1 km 2 + Water flux I I d t + ; Temperatures w i th i n ; ; karstic systems c ; t U---300 m 10 50 I s -1. km2 ; ; / 200 m / ; I 100 m I fI / ,,.,. ,,.,. sc / 7C ec t respect to the heat by only 0 1 c per High water t flow year Top of satu'rated zorie ., C 0 N u ., ::, iii "' C ::, u E ., .c 0 E 0 I Figure 1 : Conceptual model of water temperature distribution in karst systems. Air and water temperatures are generally almost equilibrated 2) The atmosphere exchanges air and water with the karst sys tem Air and water temperature of the atmosphere as well as water quantity mainly depend on meteorological processes Fortunately they can easily be measured Processes There are different types of heat transfer mechanisms wi thin the karst systems: heat conduction (mainly within the rock) heat convection in water heat convection in air heat exchanges (between the three phases), evaporation and con densation etc. Without having a physical model of each pro cess it is difficult to evaluate their respective importance HE.AT CONDUCTION IN lliE MATRIX ROCK It can be demonstrated that this process is of importance only if the respecti ve phase is immobile i.e. in the solid rock phase In fluids heat conduction is normally low compared to heat convection. Heat conduction in the rock can be described by the following equation: aT"' = a a 2 T,,. =a div(). grad T.. ) cl a z2 m m (I) where a is the thermic conductivity of the rock material and }..,,, is the heat conductivity of the solid rock: ).nl a=W p c where p is the density and c the heat capacit y of the rock. The heat flux et> across a surface A is : et> -A Am grad T,. (3) Equation I gives the relation between the variation of tempera ture in time at one point (of/at) and the variation of tempera ture in space (adiv('1> )). Assuming a geometry for the con duits heat fluxes across conduit walls can be calculated for a given temperature distribution in conduit and matrix rock CARSLAW & JAF.GER (1959) provide different analytical solutions for these calculations. HEAT CONVECTION This type of heat flux describes the heat transferred by fluid flow. It depends mainly on the fluid velocity v fl. The velocity is not constant across the section of a conduit we therefore have to account for hydrodynamic dispersion (D). We can then assume one dimensional conduit flow (in the direction of the conduit length) but with a dispersion D describing the effects of a 3-dimensional velocity field. The equation is : ar ar a 2 r -=-v -+D -
PAGE 211

Elevation ----~ ~ ~-=---'tst{ 1 l Kijahe Xontjoa Oaxaca, Mexico Kievskaya Klrktau, Uzbekistan 1450 1250 1050 gradient : 4 75C/1 000m 850 Elevation (m a s l.J ---2400 2200 2000 1800 gradient : 4 67C/1 000m 650 -+----+-~---1-~--~-+-~--~1600 ---+--+---+--+-----<>---I s temp ( CJ 11 12 13 14 temperature ( CJ Figure 2: Temperaure (air and water are not distinguished here) in two deep caves. Gradients are close to "humid air gradient" Below the homothennic zone, the water temperature is al most in equilibrium with rock and air temperature (homo therrnic zone), which implies that all three temperatures increase with depth, independent of the season. As water flows down quicker than air and rock (!), its temperature tends to be slightly colder than rock and air. This slight disequilibrium is a function of the vertical flow velocity of the water ln the catchment studied by RENNER the water table is about I 00 to 150 meters below ground surface. The phreatic conduits are located within the homotherrnic zone. The more rapid the water flows down, the colder it reaches the phreatic zone. A further example illustrating the almost exclusive relation between flow velocity (or discharge) and water tempe rature within the homothermic unsaturated zone has been des cribed by JEANNIN ( 1990): water temperature has been measured in a cave at a depth of 200 meters, once in summer during a summer stonn flood (warm infiltration) and a second time in early spring (April) during snowmelt The discharge in summer was about twice the one in spring. The water temperature of the river was about 0.3 C higher in April than in summer. Conclusion The conceptual model presented here assumes that the karst systems can be considered as composed of three main zones: the heterothermic zone close to the surface the unsaturated homothermic zone and the phreatic zone. 1n each zone air water and rock temperature tend to equilibrium. Heat exchange between air, water and walls is therefore the dominant process acting in all three zones. Meanwhile, temperature in each zone appears to depend on different dominating factors Air, water and rock temperature in the heterothennic zone are mainly influenced by the outside temperature. 1n the homothennic unsaturated zone the mean air, rock and water temperature seem to depend almost exclusively on the air temperature defining a vertical gradient of about 5C per I 000 m of elevation. Temporal variations can be induced by slight disequilibrium of water temperature with respect to rock and air due to velocity of water. Water is then slightly colder than rock and air, implying that the quicker the water flows down, the colder it reaches the phreatic zone. Our model implies therefore that average air discharge through karst systems are much higher than average water discharge. ln the saturated zone, water and rock temperature are hardly affected by geothennal heat flux. Heat pulses flow through the saturated zone almost conservatively. The shape of the pulses are somewhat changed, mainly as a results of the geometry of the conduit system in the saturated zone. The conceptual model presented here is in good agreement with many field observations and represent a valuable exten sion to the existing models. As a first approximation, it can be used to interpret the temperature responses of karst springs, providing different information about the structure of the systems: l) The shape of the temperature pulses provides infonna tion about the geometry of the conduit network within the saturated zone (see RENNER, 1996 or l1EDL et al 1997 for more details) ; 2) The amplitude of the temperature pulses provides infor mation about the importance of the unsaturated zone and mainly about flow velocities in it. 3) The positive/negative temperature pulses (warm floods in summer and cold floods in winter) provide infor mation about the depth of the phreatic zone below the ground surface lf the top of the phreatic zone is located within the heterothermic zone (probably less than 50 meters below ground), positive (in summer) and negative (in winter) pulses will be observed lf the saturated zone is located in the homo thermic zone, negative pulses will be observed all along the year This can be slightly modified by the presence of large concentrated infiltrations (large sinks) which may make the heterothermic zone deeper. References BEND E RITIER Y .; Rov B & A TABBAGH 1993 Flow characterization through heat transfer evidence in a carbonate fractured medium : first approach Water Res. Res 29 ( 11 ) : 3741 3747. BoEGu, A 1980 : Karst Hydrology and Physical Speleology Springer Verlag New York, 284 p CARsLAW H S & J. C. JAEGER 1959 Conduction of heat in solids 2nd ed. Oxford University Press London, 510 p JEANNIN P.-Y. 1990 Temperatures dans la zone vadose du karst. Bull. du Centre d'hydrogeologie de l'Univ. Neuchiirel No 9 1990: 89-102. LIED L, R; RENNER S & M SA t.rr ER 1997. Obtaining information on fracture geometry from heal flow data This issue. MATHEY, B. 1974 Gradient geothermique et hydraulique souterraine clans un aquifere karstique (Bassin de la source de l'Areuse I NE) Bull. Soc. neuclr Sci. nat ., Tome 97 : 301 314. RENNER S 19% Wannetransport in Einzelklilften und Kluftaquiferen Untersuchungen und Modellrechnungen am Beispiel eines Karstaquifers PhD Thesis, Tubinger Geowiss Arbeiren C30 89 p ROEDEL, W 1994 Pbysik unserer Umwelt: die Atmosphare. Springer Verlag, Berlin : 467 p WIGLEY T M. L. & M C. BROWN 1971. Geophysical applications of heat and mass transfer in turbulent pipe flow. Boundary-Layer Meteorology, D. Reidel Puhl Dordrecht (Holland) 1 .: 300 320 198 Proceedings of the 12 th International Congress of Speleology 1997 Switzerland Volume 1

PAGE 212

Hyperthermic caves of the United States by William R. Halliday Hawaii Speleological Survey of the National Speleological Society P.o. Box 1526 Hilo Hawaii, U.S.A. 96721 Abstract In the United States, hyperthermic caves exist in a variety of environments: hot spring terraces craters of dormant volcanoes, karsts in thermal areas, and others. Those in craters are of two types. One consists of ablation caves between fim or ice and the crater wall. The other includes a variety of hollow volcanic features discussed in a companion paper. Air temperatures are as high as 62C in a cave in a 1919 lava flow on the floor of Kilauea crater. Relative humidity is near I 00% and steam currents are common in these caves. New techniques have permitted limited exploration and study of caves with air temperatures as high as 51 C. Constant temperature monitoring and maintenance of escape routes is essential. 1 Introduction In various parts of the U.S.A. the temperature of some caves is influenced by warm or hot springs, cooling dikes or sills, diffuse areas of volcanic heat, or localized fumaroles. The two most celebrated hyperthermal caves are karstic solution caves: Warm River Cave, VA (DOUGLAS, 1964) and Lower Kane Cave, WY (HILL et al, 1976). Newly studied caves in Kilauea Crater HI have much higher temperatures and much greater variety of climatic conditions. 2. Hyperthermic karstic solution caves Air and water temperatures in Warm River Cave reach at least 28C. A warm stream enters a small terminal passage and equilibrates its air ln less constricted areas, the air temperature is essentially normothermic. This is a well-known thermal area including the famous resorts of Hot and Warm Springs, VA. Lower Kane Cave contains a seasonal resurgence of a thermal stream. In winter, the stream disappears and fumarolic gases enter the cave at the resurgence point. Summer air and water temperatures are about 27C A dry, detached upper level passage is less hyperthermic. Other hyperthermic karstic solution caves exist in Wyoming (HILL et al, 1976), in addition to what may be dissolution caves in hot spring terraces 3 Hyperthermic travertine caves Some travertine caves are formed by deposition from thermal waters, and some dissolution occurs within travertine terraces. Residual heat causes elevated temperatures in various caves in the terraces of Mammoth Hot Springs Yellowstone National Park, WY and elsewhere Steam emerges from the mouths of some of these caves. With about I 00 m of dendritic passages, Ron's Secret Cave, WA is an example of mildly hyperthermic caves formed by travertine deposition by a small spring at a temperature ofless than 25 C (HALLIO A Y 1982). 4. Geothermal ablation caves Geothermal ablation caves form at the interface of volcanic "hot spots" and glacial ice or fim Two of the five or six recorded caves of this type are in Washington state. Large, complex examples exist along the inner crater walls of Mount Rainier and Mount Baker (K.IVER, 1971 1975a, 1975b; HALUDA Y, 1978, 1980; HALLIDAY & COUGHLIN 1970) The morphology of these caves and their climatic conditions change rapidly with snowpack and variations in volcanic activity An old photograph shows the former existence of a similar cave high on the north slope of Mount St. Helens, WA at the site of the first phase of the 1980 eruptions. In the summer of 1975 only a patch of bare rock marked its former location. In these caves flakefall and other special hazards of glacier caves are added to those of hyperthermic caves per se (HALLIDAY, 1974) 5. Hyperthermic lava tube and other caves in pahoehoe basalt After the primary heat of their speleogenetic processes is dissipated, lava tube and related caves of pahoehoe basalt flows in Hawaii normally become normothermic within weeks or months. A few are affected by continued or renewed volcanic activity, or by s l ower cooling at depth An example of the latter is a rift tube cave extending lateral to the bottom of Mauna Ulu crater. This crater was active in 197276. In 1981 a residual temperature of 207 C existed in a fissure Sympos i um 7: Phys i cal Speleolog y 1 99

PAGE 213

intersecting the cave (FAVRE, 1995). Steamy lava tube caves with temperatures over 50C exist in and near the Puhimau steam vent field near the crater of Kilauea volcano These have not been studied Three lava tube caves located between the East Rift Zone of Kilauea volcano and the coastline are hyperthermic Best known is Hot Tub Cave where a bather died -almost certainly of hypertherrnia Its b ottom is a long, narrow pool with a temperature said to vary with rainfall. Before the entrance was obstructe d to prevent additional deaths I estimated the air and water temperatures as about 38 C Farther east are two other small lava tube caves with localized hyperthermic sections in a group of seven nearly parallel caves perhaps remnants of a braided system Except for ceiling drip all are dry The westernmost of the group has a 20 meter section where the dripping water is hot and the air temperature consistently measures about 34 C. A cutaround crawlway in another cave about 200 m farther east is estimated to have a localized air temperature of about 30 C but has not been studied Hot Tub Cave is about 4 km from the East Rift Zone and its hot water probably is the result of localized heating of ground water by residual heat of a 1950 eruption. The cause of the localized thermal zones of the other caves are less clear. Possibly small dikes extended from the rift zone to their vicinity in 1950 Solar radiation i s not a localized factor. 6. Hyperthermic caves in Kilauea crater, Hawaii The principal studies of hyperthermic caves in Hawaii have been on the floor of Kilauea crater. The last eruption in this crater was in 1982 but the Postal Rift lava flow of 1919 has been steaming since its deposition. All the hyperthermic caves known in this crater are in this flow but not all its caves are hyperthermic. These steamy hypertherrnic caves include small lava tube caves (three in the Posta l Rift System and others), boundary ridge caves of lava rises and other structures (WALKER 1991) hollow tumuli and others discussed in my companion paper. Measured air temperatures are as high as 62 C. Many have humidity approximating I 00 %. In some steam currents are constant or intermittent. Their elevated temperatures may be due to residual 1919 heat, to independent shallow magma bodies or both Using ordinary ca v ing clothing and low tech thermometric equipment, we have developed successful techniques for brief exploration and study of caves with air temperatures as high as 51 C. Our worst consequences have been faces reddened for a few hours (we plan to test face masks and shields soon). Despite the common presence of sweet-tasting fumes (probably hydrated sulfates) we ha v e had no respiratory difficulties Above 38C we have found light cotton clothing helpful in avoiding steam burns At lower temperatures clothing seems to be optional. Especially important is constant monitoring of air temperature In some caves it has shown microclimates where we can crawl or stoop beneath steam currents or pockets of very hot air We use an inexpensive "Radio Shack" digital thermometer with a response time of 15 seconds and a remote sensor. At all times we expect sudden changes in air flow and temperature and we are constantly aware of quick safe escape routes. Our learning period was lengthy and we urge similar caution for othe r s working in hot caves Increasingly, we are using sublingual digital thermometers as part of our learning On one occasion we found hyperthermia of 4 C in a caver after 5 minutes in a cave at 46 C This correlated with perceptible deterioration of physical and mental performance lasting a few hours Evidently there is considerab l e variation in tolerance to these con d itions. References DO U GLAS H. H. 1964 Caves of Virginia Falls Church VA. Virginia Cave Survey p. 81. FAVRE, G. 1993. Some observations on Hawaiian pit craters and relations with lava tubes. In : Proceedings of the 3rd International Symposium on Vulcanospeleology Bend Oregon 1982 p 37 HALLIDAY W R. 1974. American caves and caving New York Harper and Row, p. 58. (Aiso Barnes and Nob l e edition -, 1982 p. 58) HALLIDAY, W R 1978 Geothermal week in Washington Cascade Caver, 17: 13 March-April. HALLIDAY W R. 1980 Mt. Baker crater caves Cascade Caver 19:55 May HALLIDAY W R. 1982 Ron's Secret Cave Snohomish County WA Cascade Caver 21:18 April-May HALLIDAY W. R. & CHARLES L, COUGHLIN 1970. A preliminary reconnaissance of the Summit Steam Caves of Mount Rainier Washington Speleological Survey Bulletin 7 (Western Speleological Survey Serial #42 ) HILL C ., W SUTHERLAND & L. TIERNEY 1976. Caves of Wyoming Geo l ogical Survey of Wyoming Bulletin 54 p 89. K!VER E. P 1971. Summit fim caves of Mount Rainier. Science 173 : 320, 23 July KIVER E P 1975a Firn caves in the volcanic craters of Mount Rainier Washington. Nat. Speleol. Soc. Bull. 37:45 July K!V E R E. E P 1975b The first exploration of Mount Baker ice caves. Exp l orers Journal 53:84, June. WALKER G. P. L. 1991. Structure and origin by injection of lava under surface crust, of tumuli "lava rises", "lava rise pits" and "lava inflation clefts" in Hawaii Bull. Volcano!. 53:546. 2 0 0 Pr oceedi ng s o f th e 1 2 th Internat i onal Congress of Speleo l og y, 1997 S wi tze rl and Vo l ume 1

PAGE 214

LEGEND paug wa U v tlm of e ntrance ( Clrlp linel oO bleak.down boukfa,a $$ brH k Clown p i 'a t: cupola venlcal Clrop ce ilin g he i gh t ln mete11 t:\t~;:~ open a ir tr
PAGE 215

Misure di temperatura in due cavita carsiche nella formazione Gessoso-Solfifera dell alto Crotonese (Italia Meridionale) Marco DE PAOLA 1 -Luca Antonio DIMUCCIO 2 Paolo GIANNANDREA 3 V.le Japigia 84, 70126 Bari. 2 Dipartimento di Geologia e Geofisica, Universita di Bari, Via E.Orabona 4, 70125 Bari. 3 Dipartimento Geomineralogico, Universita di Bari, Via E Orabona 4, 70125 Bari. A bs t r act Air and water temperature measurements have been carried out within Grave Grubbo and Grotta dello Stige caves, alineated along N-S, karst system of the Gessoso-Solfifera Formation (Croton Basin Southern Ita l y). Temporal variations have been observed in each point of both caves, having different amplitude between air and water. The air temperatures are not stationary; the swingings are irregular and they not only depend on the entrances; probabily in both caves there are air of infiltration. Besides, Grotta dello Stige and Grave Grubbo probabily are in collection because there have been observed the same temperatures of the water. Mean annual temperatures of the water in both caves show a general increase according to the elevation and isopiezometric trends. l n troduzio n e Con ii presente lavoro si intende dare un contributo alla conoscenza del complesso sistema carsico che si sviluppa nella Formazione Gessoso-Solfifera affiorante nel Bacino Crotonese, situato sul bordo orientale della Sita (Mediterraneo Centrale), e in particolare delle cavita denominate Grave Grubbo e Grotta dello Stige. Le due grotte ricadono nella Tav I NE (p.p.) "Verzino" de! foglio 237, a sud dell'abitato di Verzino, e sono di facile accesso. Caratteristica fondamentale di queste due cavita e la presenza, in esse, di un torrente sotterraneo permanente che le rende estremamente attive. Sono state osservate le variazioni di temperatura sia nello spazio che nel tempo allo scopo di definime ii regime termico. Geolog i a e idrogeo l ogia dell area L'area in cui ricadono le due cavita oggetto di studio, fa parte de! Bacino Crotonese caratterizzato dalla presenza di una successione di sedimenti evaporitici di eta Messiniana (Ogniben, 1957; Roda 1964) su cui poggiano sabbie e conglomerati Pliocenici. Nell'area di affioramento delle evaporiti, l'altitudine dei rilievi varia da 200 m s.l.m., nella parte piu a Sud, a un massirno di 600 m nella porzione centrale (localita Vigne). In detta area prevalgono nettarnente le forme carsiche ed in particolare le doline, gli inghiottitoi e le valli cieche. La distribuzione e la forma delle doline e condizionata dall'esistenza di importanti dislocazioni tettoniche: sono cioe disposte secondo linee di frattura o di faglia e i loro contorni seguono, a tratti, tali lineamenti. Essendo la formazione evaporitica l'unica unita presente nella zona ad avere una permeabilita, per fratturazione e carsismo, piuttosto elevata, ne consegue che la circolazione idrica sotterranea si esplica fondamentalmente in essa, che funge quindi da "roccia serbatoio" (De Paola et alii, 1994); ii deflusso idrico, in generale, segue l'andamento della topografia, degli allineamenti tettonici e degli strati, evidenziando cosi un movimento delle acque in direzione NW SE. Le due cavita carsiche studiate, Grave Grubbo e Grotta dello Stige, sono percorse da un torrente permanente alimentato sia dalle acque di falda, sia dalle acque che scorrono nei valloni a seguito di eventi piovosi intensi. Discussione In base alle misurazioni fatte, nell'arco di quasi due anni, si possono tentare alcune considerazioni di carattere generale Per quanto riguarda la temperatura dell'aria si puo osservare che nessun tratto delle grotte e in condizioni stazionarie Le oscillazioni si presentano in maniera irregolare e sembrano non dipendere esclusivamente dalle comunicazioni con l'estemo; si deve cioe pensare che anche nei tratti intemi ci siano delle infiltrazioni di aria L'uguaglianza delle temperature dell'acqua del tratto finale di Grave Grubbo e di quello a monte di Grotta dello Stige conferma l'ipotesi della continuita delle due grotte, occluse da un tratto sifonante Considerando globalrnente le temperature medie dell'acqua di Grave Grubbo e di Grotta dello Stige si nota un generate aumento da Nord verso Sud che segue l'andamento della topografia e delle isopieze: sembra quindi che l'acqua sotterranea, di cui ii fiume presente nelle due cavita studiate ne e solo una minima evidenza, si riscaldi durante ii suo percorso: cio e c o nfermato anche dall'affl u sso di acqua calda proveniente dal sifone Sud di Grotta dello Stige. In generate si osserva che nelle due grotte, esiste una circolazione d'aria in ogni stagione dell'anno e sempre nello stesso verso, influenzata dalla temperatura estema. L'assenza di comunicazioni intermedie, nelle due grotte, dimostra inoltre quanto le discontinuita strutturali (faglie, fratture), comunicanti con l'estemo, possano influenzare ii regime terrnico di una cavita. 20 2 Proceedings of the 12 1 h International Congress of Speleology 1997, Switzerland-Volume 1

PAGE 216

Monitoring results in "Gretta Grande del Vento" {Frasassi Ancona, Italy) and its vis i tors' capaci t y Arrigo A. C i g n a Societa Speleologica Italiana, Fraz. Tuffo, 1-14023 Cocconato AT, Italy Abstract The "Grotta Grande de! Vento" is the Italian show cave with the highest number of visitors. For many years a monitoring network for the survey of some climatological parameters was been in operation Now the results obtained for the period 1991-1993 by such a monitoring network are here reported and examined The influence of the tourists flow in the cave on air temperature and carbon dioxide concentration of the cave atmosphere is investigated and an evaluation of the visitors' capacity is estimated Rlassun t o [Risultati de monitoraggio nella Grotta Grande de! Vento (Frasassi, Ancona, Italia) e la sua capacita ricettiva] La Grotta Grande de! Vento e la grotta turistica italiana con ii maggior numero di visitatori. Da parecchi anni e in funzione un sistema di monitoraggio di alcuni parametri climatologici. In questo lavoro vengono riportati i risultati ottenuti nel periodo 1991-1993 dalla sopra citata rete di monitoraggio e ne viene fatto un esame critico. E' stata inoltre studiata !'influenza de! flusso turistico sulla temperatura dell'aria e sulla concentrazione dell'anidride carbonica nell'atmosfera della grotta e, su questa base, e sta fatta una stima della capacita ricettiva della grotta stessa. 1. Introduction The Grotta Grande de! Vento is one of the most relevant Italian show caves both for the number of visitors and its intrinsic characteristcs Together with the Grotta de! Fiume, they form a karst system about 20 km long with an abundant cave fauna and many cave minerals. In fact the genesis of the system is due to the combined action of thermal water, rich in H2S, and karst water (CIGNA 1993). Just after the cave was discovered in 1971 air temperature within the cave was measured at different instances Since September 1974 the cave was opened to tourists who could reach a trail of 850 metres through an artificial tunnel. In 1982 a first automatic network consisting of two stations for the measurement of temperature, humidity, air velocity and CO2 concentration in air along the tourist trail was set up The system was rather unreliable and in 1989 it was replaced by a new network with six station The set-up of this network required a rather long time on account of troubles due to sensors and to storms resulting in frequent failures (BERTOLANI et al. 1991) On October 1993 a lightning badly injured the system and up to now (December 1996) the monitoring was interrupted. It is hoped that the operation of a more modem and reliable monitoring system should possibly start in 1997. In this paper the results obtained in the period 1991 1993 are reported and examined with the aim of investigating the influence of the tourists flow in the cave on air temperature and carbon dioxide concentration of the cave atmosphere In addition the visitors' capacity is estimated and the propagation of the seasonal thermal wave in different points of the cave is also studied 2. The data set As it was reported previuously, the automatic monitoring network gave a number of wrong results and had some period of failure. Therefore it was necessary to "clean" the original values from those which were unmistakable wrong Fortunately the wrong values were easily identifiable and therefore the cleaning did not introduce subjective modifications to the data set. The stations taken into consideration are reported in Table I. In addition some spot measurements of air temperature obtained in the period 1972-1978 were also considered to evaluate the trend of the air temperature in the last twenty years since the management of the show cave Tab l e I: Locatio n of sa mpl ing points Station Distance from Parameter the entrance (m) Outside Air temperature Ancona Hall 350 Air temperature Sala dei Duecento 400 CO2 concentration Lake Smeraldo 500 Air temperature L Orsa 585 Water temperature Cannella 655 Air temperature 3 A i r and water temperature trend Long term The plots of the air temperature data over a period of21 years which are available for two stations (Ancona Hall and Cannella) show an average increase of about 0.0345 _C per year. This trend has been calculated for the Cannella values which are less influenced by the seasonal thermal wave (Fig. 1). Then, the air temperature in the period 1972-1993 is given by the equation : where: T= T 0 + 0 0345 n TO= l 3. l 6_C (air temperature at Jan. 1972) n = number of years from 1972 It is interesting to observe that the air temperature calculated by this equation and extrapolated to Jan. 1996 has a good agreement with the experimental measurements Sympos i um 7 : Phys i cal Spe l eo l ogy 2 0 3

PAGE 217

"' N ..j "' r-,..; "' ,..; "' ('I ,..; !::! + "' r-"' 7 2 73 7 4 75 76 77 7 8 79 80 81 '82 8 3 84 85 '86 87 '8 8 89 90 '91 92 Figure l: Best fit of the air temperature in the period 19721993 for the "Cannella" station. But the accuracy and precision of the early measurements performed in the 70's is rather uncertain on account of possible systematic errors in the calibration of the thermometers used at that time and the influence of the seasonal thermal wave which cannot be taken into due account because of the paucity of data Short term The analysis of the three years ( I 991-1993) here considered gives results which are much more consistent (on the short term) on account of a much larger number of measurements and an intrinsic consistency of the data which were obtained by the same sensors throughout the whole period. The average increase of the air temperature is reported in Table 2 Table 2: Average increase of air temperature in the period 1991-1993. Station Distance from Increase the entrance (m) _C / year Ancona Hall 350 0.189 Lake Smeraldo 500 0.222 Cannella 655 0.042 These results show a correlation between the tourists flow and the increase of the air temperature inside the show cave Such an effect is quite larger for the first two stations which are closer to the tourist trail, while for the last station (Cannella) which is beyond such trail, the value of the increase is about 5 times lower and rather close to the value measured over a period of 21 years Seasonal thermal wave The air temperature outside the cave has obviously a seasonal variation which is found also inside the cave with a delay depending on the position and local characteristics The sinusoidal best fit of the data available for the period 1991-1993 is given by the equations reported in Table 3. It must be emphasized that the water temperature measured in a lake at "L'Orsa" was around I 4.5_C during 1991 and decreased in an interval of three months to a value around I 3.2_C (Fig. 2). Unfortunately it is not possible to ascertain if this pattern is due to a real change in the water circulation or to a breakdown of the monitoring network. For this reason the sinusoidal best fit of the water temperatures of "L Orsa" was carried out for the interval 19921993 only. 11")-r------------------.... II") c-i--r-------,.-------T--------1 1991 1992 1993 Figure 2 : Water temperature measured at "L 'Orsa" station. The propagation of the thermal wave from outside as detected from the measurements performed in the "Grotta Grande del Vento" cannot be through the rock because both the attenuation and the delay would have been much greater (CIGNA 1978) A careful examination of the data reported in Table 3 shows that the air temperature inside the cave is slightly less than the temperature outside. The attenuation factor of the wave amplitude is between 25 and 50 for the stations close to the tourist trail decreasing to 675 for the station (Cannella) beyond such a trail. ln Fig 3 and 4 the sinusoidal best fit of the air temperature measurements carried out outside the cave and in the Ancona Hall sare reported Table 3: Sinusoidal best fit of temperature values. Station Outside Ancona Hall Lake Smeraldo L'Orsa (water) Cannella Equation T(._C) = 14 23 9.45 sin [21t (d +57)) T(._C) = 13.73 0. I 92 sin [2n (d +27)) T(._C) = 13.51 0.302 sin [27t (d +33)) T(._C) = 13.19 0.386 sin [27t (d +57)) T(._C) = 13 84 0.014 sin [21t ( d+125)] The dates of the first maximum of the sinusoidal best fit applied to the temperature values measured in the different stations is reported in Table 4. The delay with respect to the maximum of the thermal wave of the outside air emperature is an evaluation of the speed of propagation of the thermal wave itself. For the stations Ancona Hall" and "Lake Smeraldo" there is a delay of about one month and there is no correlation with the distance from the entrance. Therefore it can be inferred that the heat tran s port is due to the tourist flow 204 P r oceedings of t he 1 2 1 h Internat i onal Congress of Spe l eo l og y, 19 9 7, S wi tze rl a n d Vo l ume 1

PAGE 218

Table 4: Dates oft/1efirst maximum of the sinusoidal best fit. Station Distance from Date (days Delay the entrance (m) since (days) Jan. I st 91) Outside 216 0 Ancona Hall 350 247 31 Lake Smeraldo 500 241 25 L'Orsa 585 217 I Cannella 655 510 294 -l 0 0 ,o 180 270 360 450 540 630 720 110 ,00 '' l080 Figure 3 : Sinusoidal best fit of the air temperature measurements carried out outside the cave. Air temperature LC) is reported on the Y-axis and the days since Ja11. 1st, 1991 are reported 011 the X-axis. 15 ~-,-----,----,---,--,...---,----,---,---,,---:----,--:1 14 1 12 J ____ ---l--'---'--'----.--:----:----,,---! 1 12 0 9\1 180 11. 360 450 540 ,30 TI( 810 ,00 HO 1080 Figure 4 : Sinusoidal best fit of the air temperature measurements carried out in the Ancona Hall Air temperature LC) is reported on the Y-axis and the days since Jan. 1st, 1991 are reported on the X-axis. It is also evident the long term i11crease of about 0.189 Clyear. Obviously such a flow operates according two mechanisms : first each person releases a given amount of heat to the cave environment resulting in the positive trend of the temperature reported in the previous two sub-chapters ; second the tourist flow causes a slow air circulation resulting in a delayed and subdued thermal wave along the show cave Beyond the part visited by tourists the thermal wave propagates much slowlier as it was found for the station "Cannella" reached by the wave nearly I O months later. The water temperature as measured at "L Orsa appears to be synchronous with the temperature outside But it is also possible that the delay corresponds to whole number of years ln fact the age of the water of some lakes in the same part of the cave was found to be around some years (CIGNA & GI0RCELLI 1989) 4. C02 concentration In February 1993 a sensor for CO2 was included in the monitoring network and its concentration was recorded hourly in the "Sala dei Duecento". The daily average values available are here reported. As it can be seen in Fig. 5 the Co 2 concentration is influenced by two different factors. One is of a nautural origin and it is responsible of the main mean term variations the other is due to the breathing of the visitors and results in the peaks synchronous with the sundays and the holidays periods E g. the peak in April corresponds to Easter ; the short peak at the end of February seems to be not related to a higher flow of visitors and therefore it could be due to a natural source. In 1983 some measurements of the CO2 concentration had already been carried out in the same station Unfortunately the instrument was not fit to operate in a cave environment and after some month it was no longer recording Nevertheless the data obtained at the time gave a first idea on the existence of two sources (natural and anthropogenic) as it was pointed out by CASTELLANI (1988) who suggested that there might be a natural reservoir of Co 2 elsewhere in the cave that was contributing to the build-up (CIGNA, 1993) The new set of data obtained at the end of 1993 confirm the previous hypothesis While the contribution of visitors to the Co 2 concentration in show caves has been studied rather extensively (e.g.: VILLAR et al. 1986) the mechanism and more in general the processes by which CO2 is naturally released in a cave environment are still unexplained. Obviously some hypothesis on possible sources connected with the seasonal cycle of vegetation and/or a contribution from thermal water circulation can be formulated Nevertheless this problem should be investigated with great attention because the identification of the processes involved could greatly contribute to the knowledge of the cycle of CO2 in caves and perhaps to some details concerning the speleogenesis. 5. Conclusion The results reported here suggest some further actions m different fields. With reference to the protection of the cave environment the temperature trend in the tourist section of the cave clearly points out that the present flow rate in the cave is slightly above the visitors' capacity of the cave itself {CIGNA 1989 ; CIGNA & FORTI 1990) The pattern of the CO2 concentration in the cave atmosphere seems to confirm a total recovery of the natural conditions throughout the years Therefore the overall effect due the visitors is probably only on the heat balance of the cave environment. An evaluation of the "static"heat balance of the cave has been made by MENI C H E TTI ( 1997) ; it would be rather interesting to develop an appropriate model to take into account the seasonal effect of the outside temperature as has been carried out recently for the Castellana caves {CIGNA & DINI 1997) In any case, for the managemnt of the Grona Grande de! Vento, it would be convenient to take into account the effect of the visitors flow on the air temperature of the cave. In particular a decrease of the thermal impact could be achieved through a Sympos i um 7 : Ph y s i ca l Speleolog y 205

PAGE 219

reduction of the heat released by both the electrical lighting and the tourists. A better choice of the lamps installed in the cave (with a higher efficiency of light output) and a decrease ot the lighting time (if compatible with the needs of the show cave) could be beneficial. A decrease of the number of visitors would be not acceptable from an economical point of view of the show cave management. Therefore the only possibility to decrease the heat released by the persons would the the reduction of time they spend into the cave. Such a result would be achieved by the opening of another tunnel to avoid the return through the same pathway covered when entering. The excavation of a new tunnel had already been considered in the past also for safety reasons: therefore a design already exists and the requirements to avoid changes in the air circulation inside the cave have been defined. Then it would be really worthwhile to take into account the opportunity to have another artificial access to the show cave. Acknowledgements I express my gratitude to my colleagues who contributed to this work and, in particular, to M. Menichetti who provided the data obtained by the automatic monitoring network and to F.G. Giorcelli who prepared some programmes for the elaboration and the calculation of the sinusoidal best fit of the values and assisted me with great patience during the development of thie whole work. References BERTOLANI M., CIGNA A.A. MACCIO S., MORBIDELLI L., SIGHrNOLFI G.P., I 991 The karst system "Grotta Grande del Vento Grotta de! Fiume" and the conservation of its environment. In : SAURO u., BONDESAN A. & MENEGHEL M (Eds.) Proc. Int. Con f. on Environmental Karst Areas (Italy I 527 Sept., 1991 ). Quad Dip Geografia, Univ Padova 13 I 991: 289-298; also as : Report ENEA RT / AMB / 92/19 CASTELLANI V., 1988 Frasassi e speleo-monitoraggio. Speleologia 9( 18) : 33-35. CIGNA A.A., 1978 Meteorologia ipogea In: Societa Speleologica Italiana Manuale di speleologia ., Longanesi Milano: 341 / 367. CIGNA A.A., 1989 La capacita ricettiva delle grotte turistiche quale parametro per la salvaguardia dell'ambiente sotterraneo. II caso delle Grotte di Castellana. Atti XV Congr. Naz Speleol ., Gruppo Puglia Grotte Amm. Comunale Castellana Grotte: 999/1012. CIGNA A A. 1993. Environmental management of tourist caves The examples of Grotta di Castellana and Grotta Grande de! Vento, Italy Environmental Geology, 21 (3): 173-180.Also as : Report ENEA RT / AMB/93/09 CIGNA A.A. & DrNI D., 1997 Qualche considerazione sul bilancio termico delle Grotte di Castellana Atti 17 Congr Naz di Speleol .. Castelnuovo di Garfagnana 8-11 Giugno 1994 (in press). CIGNA A.A & FORTI P ., 1990 La V I.A. delle grotte turistiche. The E.I.A. of a tourist cave. VIA I' Arca Edizioni Milano 4(16): 42 / 53. CIGNA A.A. & GIORCELLI F.G ., 1989 Underground water dating by tritium measurement. Proc 10th lnt Congr Speleology, Budapest 13-20 August 1989, 2: 562 / 563; also as : Report ENEA RT / PAS/89 / 24. MENICHETTI M., 1997 Bilancio energetico di una grande grotta turistica: la Grotta Grande de! Vento a Frasassi (An). Proc. lnt Symp Show Caves and Environmental Monitoring, Frabosa Soprana 24-26 March I 995, (in press). VILLAR E., FERNANDEZ P.L., GUTIERREZ I. QUrNDOS L.S & Soto J. 1986 Influence of visitors on carbon dioxide concentrations in Altamira Cave Cave Science, 13( I): 21-23 1500 ~-----------------------------------------, 1000 N 0 u E c. c. 500 + February March April May June July August Figure 5: CO 1 concentration in air measured at the station "Sala dei Duecento" in 1993. The triangles correspond to Sundays. 206 Proceedings of the 12 th International Congress of Speleology 1997 Switzerland Volume 1 September

PAGE 220

Temperature Anomalies Formation and Secular Instab i lity Research of Ice of Atmospheric Origin i n the Kars t Caves o f North Albanian A l ps A l exey Stoev Astronomical Observatory and Planetarium 62 Tsar Ivan Str., 6000 Stara Zagora Bulgaria Penka Muglova Solar-Terrestrial Influences Laboratory 6000 Stara Zagora, Bulgaria Dim i tar Stoev Sarnena Gora Touristic Society, 6000 Stara Zagora, Bulgaria Abs tr act The negative temperature anomalies in the karst caves of North Albanian Alps are a widespread phenomenon. The anomalies represent a decrease of the average anomalous temperature of the air and the main rock in the cave entrances in comparison with caves deeper parts in the presence of a lot of ice there. The difference in air exchange between caves with a different morphology and altitude is examined in the paper. An analysis of the temperature regime of thirty-eight caves has been made. The interaction between ice formations and karst massif geothermal field is shown. Three types of ca v es has been determined: !)warm caves with cold entrances and steady ice accumulations; 2)completely warm caves ; 3)cold caves with significant seasonal changes of the air flow in their entrances. The influence of snowfalls' quantitative balance on caves' temperature regime is investigated. The combination of three heat bearers : ground atmospheric air the karst massif and the falling water on it and within it forms a relatively stable temperature providing constant seasonal course of the temperature in the caves. It is shown that the microclimatic regime in caves as a whole is not determined by their origin, but depends on two factors : the angle of e negative bias of the caves from the entrance to the bottom (depth of the karst massif) and the surface of the entrance aperture cross-cut-section. It has been found that the accumulation of cold increases with the increasing of the two factors If the negative slope has a greater angle, and the entrance opening is comparatively small the cave is classified as "cold" More t h an 40 % from the invest i gated caves are glacial. The ice accumulates in the spring when a big part of the incoming waters freezes over the supercooled walls of the karst massif Another way of prolonged ice supply forming is due to the infiltration humidity, carr i ed up from the bottom to the cave surface, during the winter spring period. The negative temperature anomalies' formation in the karst caves of North A lb anian Alps is determined by their peculiarities as physico-geographical objects. They are interesting to study due to their situation in the caves and in the karst massif availability to express contrasts in the microclimate of the whole cave, and due to original processes of atmospheric circulation inside and outside the caves combined with the specificity of hydrological processes (HOTI MAHIR 1990). Temperature anoma l ies could be d efined as a result of microclirnatic processes in the caves that cause a decrease of the mean annua l temperature of the air and the main rock in the cave entrances in comparison with caves deeper parts. Most of the caves in this region of the Albanian Alps are cold during the whole year because of the constant mean day-time temperature in the interval of -1 +4 C, as well as of the existence of great amount of ice firn and snow in the entrance cliffs and halls. The basic factors that influence such temperature anomalies' formation are : I. Caves morphology 2 Pecu l iarities of the local relief. 3 Temperature regime of the underground waters. 4 Their thermal regime which d epends on the character and intensity of the air exc h ange The caves are grouped in a zone placed north west of Shcoder, in the region of the high mountain village of D rugomiri, at 800 to 2000m a l titude The mean annual air temperature is in the region of 8 to l 2 C and the mean annual rainfall is in t h e interval of 1600 to 3000mm. The rainfall minimum occurs duri n g the summer and the depth of the maximum seasonal freezing is near l y 120cm. The climate is s ub tropical and Mediterranean and it is cool and damp in the mountainous massif covered with limestone (STOEY 1996) Sympos i um 7 : Phys i cal Speleology 207

PAGE 221

The karst massif temperature is in the dynamic interval from 2.6 to 7.8C in the shade during the summer (SA VELEY 1985). The rock temperature in the stratum is a constant temperature of 2 6C, and its temperature in the first level carrying water is 2.2C. A field work on defining basic microclimatic elements' regime and the cave typology has been done as follows : depending on the snow-ice stock; depending on the temperature of zones with constant temperature (ZCT); determining of speleoatmosphere humidity; determining of the dynamics of the air exchange; determining of the dynamics of the water flowing from the glacier or the snowdrift Applying classical methods of microclimate investigation in the caves from the region, data for the basic microclimatic parameters have been obtained Comparatively low air temperatures in the interval of -0.8 to +4.4C have been registered. These temperatures as well as the availability of snow and ice mass of many years determine the examined caves as ice caves The zone of thermostatic air begins from the place of the first snow accumulation which imbalances the microclimate of the caves and they become colder than the others at the same sea level and climatic zone (SHVETSOV & KOY ALKOV, 1986). In ZCT the relative humidity is 60 to 80% and in most of the caves the flowing and dropping water is too little. Shift-flowing draughts with high velocities are not typical of the air exchange in caves. The standard equalization of both cave barometric pressure and that one of the outside ground air layer creates weak draughts outside and inside without anomalies of the velocity or their direction Taking into account that the caves in the region are developed mainly in zones of vertical-going down passing and horizontal circulation of the underground waters it is natural to associate part of the air exchange with karst massif clefts connected with the surface. Velocities of the order of 0.005 to 0.020m/s were registered. The air quantity coming in the caves from the plateau is 30m 3 /h for isostatic caves and 20 to 25m 3 /h for dynamic ones The analysis of the collected material connected with microclimatic studying permits making of following conclusions : I. Formation of karst, caves and cave systems contributes to accumulation of cold air masses. This determines the caves investigated as cold and some of them as icy caves. 2. Only the following basic conditions are completed the cold generation. Ice formation and its long period preserving in the caves from the region are possible because of: a) considerable height difference of the local relief forms; b) bag form of the karst systems or siphon available ; c) good cleft net with guaranteed opening on the surface; d) the bottom of the cave is screened by alluvial material -clay and rubble. 3 The sharp temperature boundary between cold and warm air masses which is on the way of infiltration waters contribute to detaining and piling up incoming water in hard state. 4. Preserving ice during summer is favoured by the northern slope the slow melting of the large volume snow stoppers and cold air standstill in the clefts and negative relief forms 5 The diurnal temperature amplitude becomes quiet at a depth of35 to 36m from the surface 6 Ice formations are localized at the cave entrances and on the walls and bottoms of the large vertical caves. 7 The cave ice is poly genetic. At the entrances the ice tongues, firn cones and gaskets are mainly of atmospheric origin and are formed after making tight snow masses entering from the surface. In the inner parts icy stalactites, stalagmites, columns and icy "armours" have mainly hydrogen origin due to both infiltration of atmospheric rainfalls through the cleft net of the karst massif and freezing of the condensation water. 8 The snow, fim and ice accumulation are favoured by the large initial precipices (>20m) The age of blocks of snow and ice could be estimated using measurements about their annual layers (SHUMSKI, 1955). In a section of a block from K25 cave nearly 356 layers has been registered For K25A cave they are 286. Probably this stratification shows limited age of the blocks about 400 years 9 The optimal water flow supp l ying levels carrying water is formed from the melting ice and the process of condensation in cave volume. Refe r ences HOTI, M. 1990. Particularites de la depression de mbishkodra et son region collineen,Tirana, Studime gjeographice 4: 159-172 RJEVSKI V & G. NOVIK.1978.Osnovi fizikii gornih porod, Moskva, Nedra SA VELEY, V 1985. Metodi izucheniya merzlih porod I ldov Moskva, Nedra SHUMSKI, P 1955 Osnovi strukturnogo ledovedeniya, Moskva, (AN SSSR ed.) SHVETSOV, p. & V. KOY ALSKI. 1986 Fizicheskaya geokriologiya Moskva, Nauka STOEY, A. 1996. Microclimatic parameters of the karst caves from the region of the village Drugomiri, North Albanian Alps, Annual reports of the National speleoexpeditions in Albania 1994-1996, Bulgarian Federation of Speleology 2 08 Proceed i ngs of the 12 th International Congress of Spe l eo logy, 1997, Switzerland Volume 1

PAGE 222

@ + SNOW, 'F!llAJ J,. tc e. Figure 1: Morphological typologies of s11ow-ice accumulatio11s. T .... R~lOAI exPLOR~O ~& ,-.._ '\llu..4 o~ pll,1)60MtR.l h [m] (900 ,1BW noo A&O 0 AL&AN!A @ !l tL'fJ Figure 2: Air temperature;,, the zo11e of co11sta11t temperatures depe11di11g on the altitude of the e11tra11ces. Symposium 7: Physical Speleology 209

PAGE 223

0 '10 .!O 30 "'tL'C] ,, 2 3 4 .5 h [wt1) Fi g ur e 3 : Diurnal te mp e ratur e o s cilation s amplitud e in d e pth. V l~ O)' OJI) 0,-CS 0.,(0 @ o,os 0 i 4 8 ~o {2 D ... 1, + 1.mJ [pJ] (. ,100 4 80 0 @ + ,o 0 4o 2 + i + .20 + 0 0 4 6 8 tl'c] Fi g ur e 4: Cav e di s tribution diagram 1w arm 2co ol 3 i cy, depending on the e ntran ce diam e t e r d and the in c lination i. i. L-cJ 5 4 3 2 1 14 18 iO R.!l !4 Fi g ur e 5: Maximal amplitude diff e r e n c e s of th e water s to c k from the s now blo c k 3 and dropping form 4 and t h e air t e mp er atur e in th e s pe/ e oatmo s ph e r e 210 Proceedings of the 12 th Internatio n al Congress of Speleology, 1997 Switzerland Volume 1

PAGE 224

C02 Gehalte von Luft und Wasser im vadosen Karst und in der Bodenauf I age von Eberhard Pechhold Am Ring 73, 71642 Ludwigsburg, Deutschland Abstract The co 2 available in the soil plays the most important part in determining where and to what extent corrosion of carbonate rock ta kes place The Co 2 content of the soil air is a dynamic equilibrium between biological C02 production and Co 2 losses by physical and chemical processes The Co 2 concentration, and thus the corrosion, in particular areas of karst rock is reduced mainly by convection induced ventilation of the rock and the overlying soil. By the lower Co 2 concentration measured below sloping soil surfaces, the for mation of dome reliefs on carbonate rock can be explained as a natural consequence of physicochemical processes without assuming any special conditions Zusammenfassung Das im Boden verfiigbare C02 bestimmt, wo und in welchem Ausmaf3 Korrosion des Karbonatgesteins stattfindet. Der COrGehalt der Bodenluft ergibt sich als dynamisches Gleichgewicht zwischen biologischer C02-Produktion und C02-Verlusten durch physikali sche und chemische Vorgange So setzt vor allem eine konvektionsbedingte Ventilation des Karstgesteins und des aufliegenden Bodens die COrKonzentration und damit die Korrosion im Gestein herab. Mit der durch Messungen belegten niedrigeren C02 Konzentration unter geneigt gelagerten Bodenflachen laf3t sich die Entstehung von Kuppenreliefs als natiirliche Folge physikalisch-chemischer Vor giinge erkliiren, ohne besondere Entstehungsbedingungen vorauszusetzen Lufts t romungen im Karst eben Einfluf3grof3en wie Temperatur, Niederschlagsmenge und Art und Dichte der Vegetation die in erster Linie die C02 Produktionsrate bestimmen, kommt der den Karst durchstromen den Luft eine besondere Bedeutung beim Verkarstungsprozef3 zu: Wie da Wasser die Hohlraume des Karstgesteins und die diesem aufliegende Bodenschicht durchdringt und dort die fiir die Korro sion des Karbonatgesteins maf3gebende C02-Konzentration beeinfluf3t o tut dies auch die atmospharische Luft Wiihrend Was er dabei der Schwerkraft nach, immer in derselben Richtung stromt, folgt die Luft einem zeitlich veriinderlichen Druckgefalle, das ihre jeweilige Stromungsrichtung und -intensitiit bestimmt. Meist wird dieses Druckgefalle
PAGE 225

nicht erkennen. Man kann hier nur, auf den Zufall hoffend, sta tionare Bodenluftsonden ausbringen und die CO2 -Konzentration der Bodenluft, die mit der der Karstluft vergleichbar ist, i1ber einen langeren Zeitraum hinweg messen. Wechsel der St:ro mungsrichtung, wie sie bei Konvektionsst:romungen bei einer Umkehr des Temperaturgefalles zwischen Auf3enluft und Karstluft auftreten, zeichnen sich im CO2 -Verlauf ah und erlau ben es, ventilierte Bereiche der Karstoberfliiche als solche zu identifizieren. Zurn Nachweis von in der Niihe von Karstquellen liegenden Lufteintritten kann ergiinzend zur CO2 -Messung in der Bodenluft das physikalisch geloste CO2 (sogenannte freie Kohlensiiure) in dem mit der Karstluft in Kontakt befindlichen Wasser bestimmt werden. C 0 2 La u bwald,3 0 u 5 0 cm venttllert/nlcht ven tlllert C02(Vol,;) 111mperatar( C) co, 1110 ......,. COJ WH FIL1 C02 NHO ........ co:a MW.JO T e111lea fi&.....! zeigt z.B. den Verlauf der CO2 -Konzentration der Bodenluft in jeweils 30 und 50 cm Tiefe im Laubwald (Schwiibische Alb, 750 m NN). Wiihrend ein erstes Sondenpaar (HS 30, HS 50) in nicht von Karstluft durchsetztem Boden das ganze Jahr Ober die i1bliche Zunahme der CO2 Konzentration mit der Tiefe zeigt, kehrt sich bei einem zweiten, etwa I km von dem ersten Sondenpaar entfernt ausgebrachten Sondenpaar (HW 30, HW 50) dieser Gradient um, sobald im Winter die AuJlen temperatur die Temperatur des Karstgesteins unterschreitet und dadurch eine Ventilation des Bodens aus dem Karstuntergrund heraus in Gang setzt. Die aus dem Karst aufsteigende Luft enthiilt hier relativ wenig co 2 ,was auf eine kurze, hindernisarme Zu fiihrungsstrecke und fehlenden Kontakt mit Karstwasser schlie f3en laf3t. -FILia .. .. Fig. 2 zeigt Co 2 -Gehalte der Bodenluft in 20 cm Tiefe im Bereich einer in Fig. 2a skizzierten Karstkuppe. Wie die Werte der Sanden WKI und WK2 in den Wintermonaten erkennen !assen, bildet sich hier eine Karstluftstromung zur Kuppenspitze hin aus, die die co 2 -Konzentrationen der Bodenluft im Bereich der Kuppenspitze weit ilher Sommerwerte hinaus ansteigen !as sen. Die wie in einem Kamin zur Kuppenspitze aufsteigende Karstluft enthiilt hier mehr co 2 als die Bodenluft, was auf fein C02 K a n t kuppe (L a a bwa l d ) C02(V o l1') T (G r ad C) 1.2..-------------------, 0.8 o ~ FILI -I l l .......,._ I U IU EU Tmllt l T lodell verteiltes Eindringen in den Karst durch die Bodenauflage hin durch schlief3en liillt. Kontakt mit Karstwasser ist hier nicht an zunehmen, da der Verlauf des CO 2 -Gehaltes des Wassers des das Gebiet entwiissernden Siebenbrunnens (Fig. 7) darauf keinen Hinweis gibt. Nachdem sich die Richtung der Karstluftst:romung mit dem Temperaturgradienten zwischen Karstluft und Auf3enluft grund siitzlich umkehrt, darf davon ausgegangen werden, daf3 im Som mer, der fi1r die Korrosion des Karbonatgesteins besonders wich tigen Jahreszeit mit hoher Co 2 Produktionsrate und hohen Niederschliigen, sowohl an dem im Zusammenhang mit Fig. I erwiihnten zweiten Sondenpaar als auch im Bereich der in Fig. 2a wiedergegebenen Karstkuppe Co 2 arme AuJlenluft in den Karst eindringt und dabei die co 2 Konzentration der Bodenluft herabsetzt. Zur COi-Messung in Luft gibt es heute genau und schnell arbeitende Mef3geriite. lch verwende z.B. ein von der Firma Drager AG entwickeltes, nach einem optischen Verfahren (Infrarotabsorption) arbeitendes Geriit (Typbezeichnung: Multi warn P co 2 ), das die Luft mittels einer kleinen Pumpe ansaugt und schon nach wenigen Sekunden einen sehr genauen Wert anzeigt. Mittels eines selbst gebauten Zusatzes liillt sich das Geriit auch zur Vor-Ort-Messung des im Wasser physikalisch gelosten Co 2 einsetzen. Auswirkung von Luftstromungen im Karst auf die Karstbildung Aufgrund von Ober sechs Jahren Messund Beobachtungstii tigkeit bin ich heute der Ansicht, daf3 alle wichtigen Einflilsse, welche die im Karst stromende Luft auf die Karstbildung ausilbt, letztendlich auf ihrer Fiihigkeit beruhen, gasfcirmiges CO2 zu transportieren. CO2 -Transport beeinfluJlt die Karstbildung wie derum dart am starksten, wo durch ihn hohe CO2 Konzentrationen, die hohe Korrosionsraten im Karbonatgestein verursachen konnen, herabgesetzt werden, wie dies zum Beispiel in einer dem Karstgestein aufliegenden, porosen Bodenschicht, durch die hindurch CO2 -arme Auf3enluft in den Karst eintritt der Fall ist. Durchsickerndes Niederschlagswasser nimmt in einer solchen Bodenschicht weniger CO2 auf und kann entsprechend weniger Karbonatgestein losen. Die Durchluftung des Bodens bewirkt hier eine gewisse Abtragungsresistenz des darun terliegenden Karstgesteins im Vergleich zu benachbarten, nicht oder weniger gut durchlilfteten Bereichen und, ilher langere Zeit hinweg, ein Herausheben des durchlOfteten Bereichs i1ber seine Umgebung, also die Entstehung einer Kuppe. Da/3 durch st:ro mende Luft zumindest zeitweilig auch CO2 in Karstbereiche mit urspri1nglich niedriger CO2 -Konzentration transportiert wird 2 1 2 Proceedings of the 12 th International Congress of Speleology, 1997, Switzerland Volume 1

PAGE 226

und dort zu erhohter Korrosion fii.hren kann, zeigt das oben im Zusammenhang mit Fig. 2 beschriebene Beispiel fur die Zeit der Wintermonate. He rab setzu n g de r C02 Konzentratio n du r ch K on v e ktion im Han g bo d e n Bei der grof3en Anzahl von co 2 -Messungen in der Boden luft, die ich zunachst mit Hilfe einer unmittelbar vor der Messung in den Boden getriebenen Sonde, bald jedoch mit stationiir einge setzten Sonden in verschiedensten Boden durchfllhrte, fiel mir bald auf, daf3 sich in starker geneigt gelagerten Boden Talhlingen, grundsatzlich ein Durchluftungseffekt zeigt, der die co 2 -Konzentration der Luft dort starker herabsetzt als in ent sprechenden, eben gelagerten Boden und nicht uberall durch in den Boden ausmundende Karstspalten hervorgerufen sein kann Nach genauerer Untersuchung dieser Erscheinung ergab sich als einzig sinnvolle Erkliirung hierfilr eine Eigenbeluftung des Rangbodens durch Konvektion infolge eines Temperaturunter schiedes zwischen Bodenluft und Auf3enluft. .... JIU ... --------Dank der Neigung der Bodenschicht, die den Aus -und Ein tritt von Luft aufunterschiedlicher Rohe ermoglicht, kommt hier, wie in Fig 3 schematisch dargestellt, eine Bodenluftstromung nach demselben Prinzip wie die konvektionsbedingte Karstven tilation im Crrof3en in Gang : 1st die Auf3enluft kiilter als die Luft innerhalb des Rangbodens, so steigt wiirmere und infolgedessen leichtere, co 2 -haltige Bodenluft innerhalb der porosen Boden schicht nach oben und saugt kiiltere, CO2 -arme Auf3enluft von der Seite her nach. 1st die AuBenluft warmer, so sinkt kiiltere Bodenluft in der Bodenschicht nach unten und saugt wiirmere, co 2 -arme Auf3enluft von der Seite her nach. In beiden Situatio nen wird Bodenluft durch Co 2 -arme Auf3enluft ersetzt und die co 2 -Konzentration im Rangboden gegenuber der CO2 Konzentration in vergleichbaren eben gelagerten Boden, in denen sich wegen der fehlenden Neigung keine Konvektion ausbilden kann, herabgesetzt In manchen Rangbliden bilden sich offenbar durch Aneinan derkoppeln einzelner Konvektionszellen ausgedehnte Bodenluft stromungen aus, die dazu filhren, dal3 im Bereich des RangfuBes bei gegenuber der Bodenluft kiilterer AuBenluft ( Winter) extrem niedrige Co 2 -Konzentrationen, bei wiirmerer Auf3enluft (z.B im Sommer) dagegen Co 2 -Konzentrationen, die mit denen der Bodenluft ebener Boden vergleichbar sind, in der Bodenluft gemessen werden Umgekehrt werden am oberen Rangende bei kiilterer AuBenluft vergleichsweise hohe und bei wiirmerer Au f3enluft niedrige co 2 -Konzentrationen der Bodenluft beobach tet. Fig. 4 zeigt den Verlauf der CO2-Konzentration der Boden luft an 30 und 50 cm tiefen Sonden im Laubwald, im Fuf3bereich eines mit 30-Grad geneigten Ranges und in fast ebenem Boden. Eine von mir aufgrund der Mef3ergebnisse eines Jahres vor genommene Abschiitzung der durch die Bodenluftkonvektion an diesem Hang zu erwartenden Korrosionsminderung ergab knapp Jahresver l au! Boden C02 la st e b e n e r und 30 ge n e lgt er Waldboden C02 (V o llL) Grcrd C 1-----~-~---------+IO I.Mat fl l.C I.Jui I QQ3 eben.0 3 m i I.Oki I.Jan I.Apr 20 I QQ4 _._ ao Hang o a m. -IiiSO Rang o e m 0 26 m ....._. Aulen .. mp 25%, bezogen auf die Korrosion unter dem entsprechenden, eben gelagerten Boden (Pechhold 1996). Es ist klar, daB eine solche Korrosionsminderung unter Hangboden im bedeckten Karst eine beachtliche Abtragungsresi stenz geneigter Fliichen gegenuber ebenen Fliichen schaffi, die sich im Relief der jeweiligen Karstoberfliiche dokumentieren muf3. Ober llingere Zeit betrachtet nimmt bei Abtragung einer Karstfliichesofem sich die Klimabedingungen nicht wesentlich lindem und Erosionseinflusse nicht uberhandnehmen der Anteil der abtragungsresistenteren geneigten Fliichen gegenuber dem der ebenen Fliichen zu, bis schliel3lich nur noch geneigte Fliichen vorhanden sind und sich die gesamte Karstlandschaft aus kegeli gen oder satteldachformigen Vollformen zusammensetzt. Da die Bodenluftkonvektion in Hangboden mit deren Schriiglage zu nimmt, neigen die Hangfliichen daruberhinaus zur Versteilerung bis es zum Abrutschen der Bodenauflage und zur Ausbildung nackter Felsflanken kommt, die nur noch in ihrem Fuf3bereich einer hoheren Korrosion durch Kontakt mit Bodenschichten ausgesetzt sind und sich dadurch zu senkrechten oder sogar uberhlingenden Steilwlinden weiterentwickeln konnen. Beeinf l ussung des Karstwasse r s durch Lu ft stromung im vadosen Bereich Hohe CO2 -Konzentrationen herrschen meist auch im Karstwasser, in dem sich in wenig durchlufteten Bereichen der vadosen Zone das physikalisch geloste CO2 nahezu im Gleich gewicht mit dem CO2 Gehalt der umgebenden Luft befindet. Da vor allem rasch flief3endes oder fallendes Wasser CO2 sehr schnell mit der Umgebungsluft austauscht entzieht uber die Wasseroberfliiche streichende oder Tropfwasserzonen durch stromende, co 2 arme Luft dem Wasser schnell einen grof3en Tei! des gelosten co 2 sodal3 zuvor kalkaggressives Wasser nicht mehr korrodierend wirkt, in manchen Fiillen sogar Kalk ablagert. Sympos i um 7 : Physical Speleology 213

PAGE 227

NN "' FI C. 5 "' I I .,. 5 HlSSl'U!Klt > 1000 m In der Falkensteiner Hi:ihle (Schwiibische Alb), ( Fig. 5 ), stromt im Winter co 2 -arme Aul3enluft vom Eingang her, entge gen der Fliel3richtung des Hi:ihlenbaches bis Zll einem 2200 m vom Eingang entfemten Siphon Sie verlal3t das Hi:ihlensystem fein verteilt durch die etwa I 00 m miichtige Hi:ihlenOberdeckung. Im Sommer dringt umgekehrt Co 2 arme Au13enluft durch die Hohlenilberdeckung in die Hi:ihle ein und stri:imt in Flie13richtung des Hi:ihlenbaches zum Eingang. C02-Gehalt in der Falkensteiner Hohle In der Lult und Im Wasser C02 Vol~ 2~---------------, o., o-te...----' ----'---...,__ ___,.___~--~--~-~ I 2 FIi.i 3 6 Meepantl-) 6 C02-W Wtnter C02-L Winter 7 .-C02-W Sommer -eC02-L Sommer 9 Fig. 6 zeigt an neun Mel3punkten in der Hi:ihle gemessene co 2 Gehalte der Hohlenluft (C02-L) und des Hoh!e~bach~as sers (C02-W) im Winter und im Sommer. Vor al!em 1fil Wmter ist das Hohlenbachwasser, das den gri:il3ten Tei! seines C02 an die entgegenstri:imende Luft abgibt, im vorderen Tei! der Hi:ihle stark kal.ki1bersiittigt, was zu Kalkablagerungen im Hohlenbach auf den ersten 1000 m der Hohlenstrecke gefilhrt hat. Der relativ niedrige co 2 Gehalt der von oben her zustromenden Luft unm.ittelbar vor dem Siphon (Mel3punkt 9) im Sommer, nimmt hohlenauswiirts rasch einen hoheren Wert an, was hier klar auf Kosten des co 2 -Gehaltes des Hohlenbachwassers erfolgt. Hochwiisser konnen durch Bildung eines zusiitzlichen Siphons etwa 400 m innerhalb der 1-lohle die Luftstromung ttnterbinden, was sich, insbesondere im Winter, durch starke Anstiege im C02 Gehalt des Hohlenbachwassers bemerkbar macht und in der co 2 Konzentration der Elsach-1-langquelle (s. Fig. 7) an der das Hi:ihlenbachwasser der Falkensteiner Hohle ins Freie tritt, deutlich wird. Die in der Falkensteiner Hohle nachweisbare Wirkung stro mender Luft auf das Karstwasser kann die Karstbildung prinz.ipi ell iiberall im vadosen Karst beeinflussen. Da das Wasser der vadosen Karstzone oft in phreatische Bereiche abf1.iel3t, kartn eine Durch!Oftung der vadosen Zone nicht nur dort selbst, sondem bis weit in die phreatische Zone hinein Korrosion herabsetzen und damit auch dort auf die Karstbildung einwirken. Bei Quellen im seichten Karst und im Grenzbereich zum tie fen Karst ist ei.n sicheres lndiz filr den Kontakt mit Karstluft ein ROckgang des physikalisch geli:isten co 2 mit der SchOttung. Durch Offnen von Siphonen und Freiwerden von dicht Ober dem Wasserspiegel gelegenen luftwegsamen Riiumen infolge des sinkenden Wasserspiegels wird niimlich der Kontakt zwischen Wasser und Karstluft und damit der Co 2 -Austritt in die Ober das Wasser streichende Luft verbessert Zusiitzlich werden bei weni ger Wasser und einer eher zunehmenden Menge zugefiihrter, Co 2 -iirmerer Luft sowohl das den C02 -Austritt bewirkende co 2 -Konzentrationsgefalle zwischen Wasser und Luft vergri:i13ert als auch, infolge verringerter Flie13geschwindigkeit, die Kontaktzeit zwischen Wasser und Luft verliingert und darnit die co 2 -Abgabe aus dem Wasser weiter verstiirkt. C02 EHQ Blautopt,7-Brunnen Tm(Grad C W1t Ulm)/SchCttg 1 2 0 O 0 ma x2 5 Blau ScbOtt BHQ m l'.! I Fig. 7 zeigt neben dem Verlauf der Schilttung und des physi kalisch gelosten C02 der o. g. Elsach-Hangquelle (EHQ) auch den Verlauf des Co 2 -Gehalts des Blautopfs (Schwiib.Alb) und des oben erwiihnten Siebenbrunnens, einer aus dem phreatischen Karstbereich e .uspringenden Que!le in der mittleren Schwiibi schen Alb. Aus dem Rilckgang des Co 2 Gehaltes des B!au topfwassers wiihrend der n.iederschlagsarmen Zeit van Januar bis Mitte Marz 1996 liil3t sich h.ier z. B schliel3en, dal3 auch der aus dem phreatischen Karstbereich austretende Blautopf von Karstluft beeinflul3tes Wasser aus dem vadosen Karstbereich enthiilt. Schriftenverzeichnis BOGLI, A. 1969. Co 2 Gehalte der Luft in alpinen Karstbo den und Hohlen. V. Int. Kongr. Spiiologie Stuttgart, Abh., 2: 28/2 28/9, 4 Tab. ; Stuttgart. PECHHOLD, E. 1996. Die Rolle der Gesteins und Boden bewetterung beim Verkarstungsprozel3 am Beispiel der Schwiibi schen A lb Laichinger Hohlenfreund, 31, (I): 25 38, I O Abb., I Tab. ; Laichingen 214 Proceedings of the 12 '" International Congress of Speleology, 1997, Sw i tzerland Volume 1

PAGE 228

Daily atmospheric variation within caves in southwestern Wisconsin, U.S.A by Robert J. Mueller and Michael J. Day Department of Geography, University of Wisconsin-Milwaukee Milwaukee Wisconsin 5320 I U.S.A Abst r ac t Although the general body of theory holds that typical cave atmospheres are relatively stable, with predictable temperatures reflecting the mean annual surface temperature, and consistently high relative humidities. this may not be the case where short, shallow, ridgetop caves in fractured rocks, particularly caves with multiple entrances, permit considerable air exchange between the cave passages and the surface. Caves in southwestern Wisconsin are typically short (less than 300m) and shallow (less than I Orn). Mo t are in ridgetop locations and have entrances in sinkhole bases or quarry faces Most have single entrances, but several have multiple entrances, some natural, others excavated. The Ordovician limestones and dolostones in which many of the caves are developed are typically thin-bedded and intensely fractured Over a two week period from May 26 through June 8, 1990 daily measurements were made of temperature and relative humidity at seven stations within John Gray Cave, a single sinkhole entrance ridgetop cave in Richland County The cave is 232m long, composed of an entrance lobby beneath the sinkhole and three radiating passages 49m, 72m and 111 m in length. Mean annual temperature in southwestern Wisconsin is about I 0 degrees C. During the study period mean external temperature wa 19.0 degrees C, ranging from I 0.0 to 23 9 degrees C. Lobby temperatures were similar, ranging from 12.5 to 23 .3 degrees C, with a mean of I 8.6 degrees C. Temperatures in the three passages varied less, but with ranges of 6.2 to 8.4 degrees C around station means between 14.1 and I 6.4 degrees C External relative humidity ranged from 32 0 to 90 0 % with a mean of 58 9%. Mean lobby RH was 52.2%, ranging from 41.0 to 69.2%. RH in the three passages ranged from 38.9 to 65 6 % with station means between 47.0 and 52.9%. The temperature data, considerably different from the mean annual surface temperature and with daily variations reflecting external temperature changes during the study period, coupled with the low and variable relative humidities within the cave, indicate a dynamic cave atmosphere suggestive of considerable air exchange with the surface Zusammenfassung Es wird in der Literatur allgemein dargelegt, dass die typische Atmosphare in Hohlen relativ stabil ist. Dies mit voraussagbarer, der Obertlachendurchschnittstemperatur entsprechender Hohlentemperatur und konsistent hoher relativer Luftfeuchtigkeit. Dies erscheint jedoch nicht der Fall zu sein in kurzen, niedrigen Hohlen, die im obertlachennahen, frakturierten Gestein gelegen sind. vor allem wenn sie mehrere Eingange besitzen, welche betrachtlichen Luftaustausch zwischen Aussenoberflache und Hohle zulassen. Hohlen im si.idwestlichen Wisconsin sind i.iberwiegend kurz (weniger als 300m) und niedrig (weniger als I 0m) Die Mehrzahl ist oberflachennah und besitzt Eingange
PAGE 229

1. Introduction The general bod y of theory as presented in both introductory and advanced texts (e.g. MOORE & SULLIVAN 1978; WIGLEY & BROWN, 1976) indicates that typical cave atmospheres are relatively stable with predictable temperatures reflecting mean annual surface temperatures, and consistently high relative humidities. Conversely detailed studies h ave shown that at least some cave atmospheres are more variable, particularly in short caves near entrances, and in caves with more than a single entrance (DA VlES, 1960 ; CROPLEY, 1965 ; ATKI SON ET AL, 1983 ; NEPST AD & PISAROWICZ, 1989; TRAPASSO & KELETSKY, 1994). On the basis of these previous studies, it is hypothesized here that cave atmospheric variabi l ity is particularly pronounced in the karst of southwestern Wisconsin where the caves are typically short, shallow, in ridgetop locations, and developed in thin bedded and intensely fractured carbonate rocks Under such conditions, it is anticipated that there is considerable exchange of air between the caves and the surface, resulting in cave atmospheric variability reflecting changing surface weather conditions. 2. Caves in southwestern Wisconsin The karst of southwestern Wisconsin is developed in the Ordovician-age Prairie du Chien and Platteville-Galena Formations, and constitutes part of the upper Mississippi Valley karst which is essentially coincident with the unglaciated Driftless Area of the upper Midwest. The carbonate rocks outcrop typically at or near the summits of interfluvial ridges, and karst landfonns with the exception of springs occur in disjunct ridgetop clusters. The karst includes a wide array of dry valleys sinkholes caves and springs (DAY et al, 1989). North of the Wisconsin River most caves are developed in the Oneota dolostone, which attains a maximum thickness of about 30m. South of the river caves are developed in the Platteville-Galena limestones and dolostones which attain a total thickness in excess of I OOrn The carbonate rocks are variably bedded and Paleozoic tectonic stresses have resulted in intense regiona l and local fracturing, accentuated near the surface by dissolution and by Pleistocene periglacial conditions Although there are more than 200 caves in southwestern Wisconsin, most are less than 300m in total length and none exceeds I OOOrn (DAY l 986a,b ). Dimensions are constrained by the slow dissolution rate of the dolostones, limited ridgetop catchment areas thin bedding (typically less than 0.5m), and the dismemberment of formerly more extensive systems by valley incision Most caves are shallow, typically at depths of I Orn or less, and rarely exceeding a depth of 25m (DAY et al, 1 989) 3. Hypotheses and methodology The research reported here represents a portion of a larger study of cave atmospheric conditions in southwestern Wisconsin. The larger study involves longer time periods and several caves, and is as yet incomplete (Mueller in progress). This report deals only with the results obtained from a single cave (John Gray Cave, in Richland County, Wisconsin) over a fourteen day period from May 26 through June 8, during the summer of 1990 John Gray Cave is representative of caves formed north of the Wisconsin River in the Prairie du Chien Formation Located on a ridgetop it has a single sinkhole entrance leading to an entrance lobby and three radiating passages, which are essentially the remnants of phreatic tubes, 49m (passage A) 72m (passage B) and I I Im (passage C) in length (ALEXANDER, 1980) All three passages are less than I O m below the surface of the ridgetop. Based on the general body of theory on the size and topographic l ocation of the caves in southwestern Wisconsin and on the seasonal surface atmospheric situation the following four hypotheses are tested here : a) the cave air temperature is relatively constant reflecting mean annual surface temperature, b) the relative humidity of the cave air is consistent and high, c) the air temperature decreases with distance into the cave and d) the relative humidity increases with distance into the cave Between May 26 and June 8 1990 temperature and relative humidity were measured daily at seven stations within the cave between 08.30 and 13.00 h ours Statio n LI was in the entrance l obby, adjacent to the talus pile beneath the sinkhole entrance; stations A l and A2 we r e Sm and 20m respectively into passage A; stations BI and 82 were !Orn and 55m respectively into passage B; and stations Cl and C2 were 20m and 25m respectively into passage C. Temperature and relative humidity were measured at between 12 and 15 centimeters a b ove lobby and passage floors using a Pacer Industries digital hygrothermometer with a resolution of 0.1 % RH and 0.1 degrees C. Reading accuracy is estimated conservatively to be within 1 0% RH and 0.5 degrees C. Surface temperatures and relative humidities were also measured with the same hygrothermometer, and were consistent with data for the same time periods obtai n ed from an independent National Meteorological Service weather station at the Tri-County airport, l ocated at Lo n e Rock, Wisconsin, 32km from the cave 4. Results Mean surface temperatu r e during the fourteen day monitoring period was 19.0 degrees C, ranging from a l ow of I 0 0 degrees C to a high of 23 9 deg r ees C Temperatures in the entrance lobby corresponded generally with the surface temperatures with a mean of 18.6 degrees C and a range of 1 1.2 degrees C, from 12 5 degrees C to 23 3 degrees C. Six of the fourteen lobby temperature readings were above those of the surface, seven were below and one was the same. Temperatures within the t h ree passages followed generally the same trends as the surface and the lobby, varying in similar patterns on a d aily basis, but with lower temperature means and smaller ranges (Table I; Figure I). Temperatures generally decreased with increasing distance into all three passages, and passage temperatures in general were lower than those in the lobby Passage A the shortest was consistently the warmest, and passage C, the longest, was consistently the coolest (Table I; Figure I ) Ta b le I. Pa ss a ge te mp er atur es. Station Mea n Minimum/Maximum Al 16.4 12.2 / 18.9 A2 15.8 I 1.9 / 18.1 BI 15.4 10.2 / 17.8 82 1 4.9 9.3 / 1 7 9 Cl 14.6 8.8 / 17.3 C2 14 1 8 8 / 1 7.1 Range 6.7 6 2 7 5 8.2 8.4 8.3 Surface re l ative humidities varied considerably ranging 58 % from 32 0% to 90.0%, wit h a mean of 58.9 %. Relative humidities within the entrance l obby were similar, but muted, with a range of27.2%, from 41.0% to 69 2% and a mean of 52 2 % Six of the fourteen lobby RH readings were above those of the surface, seven were be l ow, and one was the same. 21 6 Proceedings of the 12 th International Congress of Speleology, 1997, Switzerland Volume 1

PAGE 230

Relative humidities within the three passages followed generally the same trends as the surface and the lobby, varying in similar patterns on a daily basis, but with generally lower means and smaller ranges (Table 2; Figure 2). Relative humidities generally increased with increasing distance into the three passages, although passage humidities in general were lower than those in the entrance lobby Passage C, the longest, had the highest RH readings consistently, whilst passage A, the shortest, had the lowest. Tab l e 2. Passage rela t ive h u m i d i ti es Station Mean Minimum/Maximum Range Al 47 0 39.6 I 56 7 17 1 A2 48.1 40 0 I 58.1 18.1 BI 48.9 38.9 I 60.0 21.1 B2 50 2 41.0 I 61.3 20.3 Cl 51.5 40.7 I 64.8 24.1 C2 52 .9 41.1 / 65.6 24.5 5. Conclusions Analysis of the resu l ts demonstrates clearly that both temperature and relative humidity in John Gray Cave throughout the fourteen day study period fluctuate considerably and mimic changes in surface atmospheric conditions. This points unequivocally to a dynamic cave microclimatology linked to external stimuli via air exchange between the cave and the surface. Of the four hypotheses only the third is supported solidly by the data, and the fourth to a lesser degree. Cave air temperatures generally are not constant, and do not reflect the mean annual surface temperature. Likewise, the relative humidity of the cave air is variable and often falls below 50%. Cave air temperatures do generally decrease with distance in to the cave, and relative humidities tend to increase although not consistently. Variations in cave air temperature are particularly clear reflections of s urface air temperature changes. This is especially the case in the entrance lobby where the temperature range ( I 1 .2 degrees C) is similar to that at the surface ( 13.9 degrees C). Lobby temperatures were generally lower than surface temperatures but in six instances the lobby was warmer than the surface, perhaps re ulting from retention of heat in the rock walls. Air temperatures within the three passages were generally lower than those at the surface or in the lobby, and they varied less, although still by between 6.2 and 8.4 degrees C. As expected, temperatures within the lo n gest passage were the lowest, although this passage demonstrated conversely the greatest temperature range. Relative humidities within the cave demonstrated essentially similar patterns, with lobby RH clearly related to surface conditions, and passage air RH showing similar but depressed fluctuations Cave air RH was always less than 70% while surface RH ranged up to 90%. Although the results are suggestive of considerable air exchange between the surface and the cave, no measurements have yet been made of air movement either within the cave itself or through the entrance. Equipment available during this study was unable to detect air movement of less than 0.2 meters per second, and monitoring of cave air flows remains a project for the future. References ATKINSON, T .C., SMART P .L. and WI GLEY, T.M.L. 1983. Climate and natural radon levels in Castleguard Cave, Columbia Jcefields, Alberta, Canada. Arctic and Alpine Research 15: 487502 ALEXANDER, E.C. (Ed) 1980 An Introduction to Caves of Minnesota, Iowa and Wisconsin. .S.S., Huntsville, 190 p. CROPLEY, J B 1965 Influence of surface conditions on temperatures in large cave systems. Bulletin National Speleological Society 27: 1-9. DAY, M.J. 1986a. Caves in southwestern Wisconsin, U S A. Proceedings 9th International Congress of Speleology Volume I : 155-157. DAY, M.J. 1986b Cave studies in southwestern Wisconsin: Implications and importance. The Wisconsin Speleologist 19(3) : 1-21. DAVIES, W.E. 1960 Meteorological observations in Martens Cave, West Virginia. Bulletin National Speleological Society 27: 1-9. MOORE G.W and SULLIVAN, G.N. 1978. Speleolog y: The Study of Caves Cave Books, St. Louis, I 50p. EPS TAD, J and PISAROWICZ, J 1989 Wind Cave, South Dakota : Temperature and humidity variations. Bulletin National Speleological Society 41 : 125-128 TRAPASSO, L.M. and KALETSKY K. 1994 Food preparation activities and the microclimate within Mammoth Cave, Kentucky. Bulletin Natio n al Speleological Society 56: 64-69. WIGLEY, T M.L. and BROWN, M.C. 1976 The physics of caves. In : The Science of Speleology ed. T D Ford and C.H.D. Cullingford: 329-358. Academic Press, London Symposium 7: Physical Speleology 217

PAGE 231


PAGE 232

Overview of cave minerals onthogeny Vl a dimi r A M altsev VNIIGEOSYSTEM institute. Moscow 121351, Yartsevskaya ul., 15-21, Russia Abs tra c t The onthogenetic approach in mineralogy, mostly developed in the former USSR, covers the part of the earth science, connecting mineralogy and crystallography. The subject of study within this approach is the minor mineral bodies hierarchy their structures and textures, ~n_d the genetic con~rols for the separate features of the structures and textures. Onthogenetic tools, including the concept of characteristic symmetry, Cune symmetry principle, conception of mineral bodies with strict or partial interactivity, allow precise re construct10n of the genetic mechanisms for the minor mineral bodies. Caves, with their "clear-case" crystallization environments and ~!most unlimite~ sp_ace for free crystallization, allow to study the higher levels of the minor mineral bodies hierarchy, than are displayed in other geologic sites. The knowledge, received from this, may in its tum be applied to other cases, where the mechanisms are the same, but are not so clear, and thus significantly enriching the general knowledge of mineralogy Introductio n Onthogeny of minerals is one of the main parts of the mineralogical science in the former USSR, and is almost not studied abroad. Onthogenetic tools are rather important in studying speleothems. Expecially cave environments (crystallization in empty space) display higher organized objects, than other geologic environments, thus requesting more powerful tools for their study. Basic concepts The subject of the onthogenetic approach is the minor mineral bodies hierarchy. the minor mineral bodies interactions and the connection between minor mineral bodies and their parent morphogenetic environments. A "minor mineral body" is a term. separating the mineral bodies that are to be studied by means of mineralogy (starting with individual crystals. and raising to approximately veins). from the mineral bodies, that must be studied by mean of petrography (these are rocks). In cave environments minor mineral bodies include objects, starting with individual crystals, and raising up to whole speleothem ensembles, reflecting a whole cycle of the cave development history S t andard h ierarc h y In "us ual (non-cave) mineralogy the mineral bodies hierarchy is defined as following. According to DYMK0 V [ I 991 ], the term "mineral" is wide, and joins together mineral species, mineral variety, mineral individual mineral aggregate. Zhabin [ 1979] and Godovikov [ 1989] enchanted the list and now it's as below (excluding species and varieties. that have strictly the same sense that in Western science). 1-st order mineral individuals. These are simply crystals. 2-nd order individuals. These are objects, having an internal structure, but s till no texture They grow from a single crystal nucleo during a single-act crystallization process, and may appear as skeleton crystals, or splitted crystals. Skeleton crystals appear in environments with high oversaturations and poor feeding or poor mixing as a result of a disbalance between the rapid growth possibility and too poor supply for this [GRIG0RJEV, ZHABI 1975]. This leads to rapid growth on edges and tips without massive growth on faces. Splitted crystals appear as a result of mechanical or chemical admixtures in the parent solution, and, due to the splitting grade, may appear as partly faced cheaf-like crystals, or spherolites, or spherocrystals (top splitting grade with complete disappearance of any boundaries between sub individuals, and appearance of curved crystal faces and spherical cleavage). In Western science skeleton crystals are u s ually taken together with dendrites, and spherolites together with radial fibrous aggregates. This is unlikely, because both dendrite s and radial-fibrous aggregates belong to higher hierarchy levels, grow from multiple nucleos, and display selection between individuals. aggregates. These are minor min e ral bodies. displaying some regular co-growth of multiple mineral aggregates. On this organization level [MAL TSEV et. al. 1995] some kind of selection (competition) between individuals is nece ssary. defining the texture of the aggregate. Some explorers even take the terms "agg regates" and textures as synonyms. but this is truth only when considenng the lowest level of texture. By definition, texture is a set of geometric variations of the structures of mineral individuals and aggregates in the complete space of synchronous crystallization [STEPAN0V 1973], and therefore may apply to co growths of aggregates. The same kind of aggregate may be built from different individuals. For example, popcorn and frostwork appear as the same aggregate crystallictite [SLETOV 1985], built from spheroidolites in the first case or crystals in the second. higher organized minor mineral bodies usually they need not to be classified. Outside caves the synchronously generated minor mineral bodies, having through structure or through texture, and raising higher than aggregates are only veins or unique phenomenas Others convert into rocks losing through structure and texture Hierarc h y en h anceme n t for caves Caves (or separate rooms) by definition are "co mplete space of synchronous crystallization", and all the sinter has through texture, therefore matches the definition of a minor mineral body. In spite of this, several organization levels can be observed clearly higher, than aggregates. They can be observed in almost all the caves and so are to be classified. It 's not enough to simply add some levels Firstly, one must define the boundary between aggregates and higher-organized mineral bodies correctly, because currently it is not clear. Here we'll consider as aggregates only minor mineral bodies having a mono-mineral composition and built from homogenou s individuals. I. The first enhancement is needed on the individuals level. In caves we meet mineral individuals, having both features of skeleton growth and splitted growth, even other kinds of abnormal growth, at the same time These are selenite needles [MALTSEV 1996], nests [MALTSEV 199 7a], etc. They are rare Symposium 7: Physical Speleology 219

PAGE 233

enough not to be classified, but usual enough to defi n e a group of "complicated individuals" 2 Multi-aggregates [MALTSEV 1995). If an aggregate is something monomineral, mostly built from homogeneous individuals (both except the case of changing the environment between crystallization acts), multi-aggregate may consist of several minera l s, of severa l kinds of individuals, even may be an intergrowth of several aggregates, but they are always generated by the same crystallization environment. To understand the necessity of this conception, let's discuss two examples In many caves we can see the following sequence : calcite corallites, aragonite crystallictites, hydromagnesite efflorescences This sequence can be seen in each single bush or branch, and all three phases are simultaneous and singenetic The capillary solution film, while moving along the multi-aggregate and evaporating, gets enriched in magnesium, controlling mineral changes. The magnesium even goes into a cycle, being dissolved in the root part, where the solution is undersaturated in magnesium (drawing). Even more different minerals in the same state of the crystallization environment grow into different individuals and aggregates the crystals splitting grade is very different. This regular singenetic intergrowth of different minerals in different aggregates is what we call a multi-aggregate One can note that this example shows a phenomena, which is interesting for general mineralogy It is possible to have a complete picture of sequential crystallization ( even without appearance of induction surfaces) in a synchronous case. Another example is the syngenetic i n tergrowth of ca l cite druses and grainy calcite, desc r ibed by DYMK0V & SLET0V [ 1981). An interesting analogy is that the multi-aggregates conception seems like an onthogenetic reflection of the paragenesis conception (first observed by Stepanov [unpublished lectures] that described paragenetic sets of aggregates, that cover a half of possible multi-aggregates types), appearing as a rule, only in karstotypic [M0R0SHKIN, 1986] surroundings. Multi-aggregates appear as the last hierarchy level, where a through structure may be observed On the upper levels we can only speak about texture Also, multi-aggregates, as it can be seen from the first example. appear to be t h e first level, where we must speak not only about synchronous crystallization, but synchronous crystallization and re-crysta ll ization. 3 Crusts. The term was first suggested by STEPAN0V [1971] then concreted by several authors [MALTSEV, 1993, M0R0SHKIN, 1976, 1986). Now it is understood as a union of aggregates and multi-aggregates, generated within the same crystallization environment not only locally in the cave Good examples of crusts are the stalactite-stalagmite crust (union of products of gravitation-controlled dripping and flows), the corallite crust Uoining all the subaerial popcorn and frostwork, generated by capillary fi l ms evaporation), the antholite crust Uoining all the efflorescences, generate d from the inside-pores solutions evaporation). On all the preceding levels starting from indivi d uals, the hierarchy can break Individuals, not joining into aggregates are possible (calcite spherolites inclusions in gypsum crystals from Podolian caves [TURCHIN0V 1993], selenite needles [MALTSEV 1996]) Aggregates, not joining into crusts are even more usual (helictites) 4. The last hierarchy level appear in mineral ensembles. This concept, making a difference to previous ones, is not strictly determined, and 1t 1s not even transitive. The idea of such conception was first suggested by STEPAN0V [u n published lectures], but its final sense was concreted only in [MALTSEV 1995]. It can be understood as some generalization of the mineral association conception. As is shown in [MAXIM0VICH 1961 STEPAN0V 1971a], crystallization processes in each cave element change, due to microclimate changes, in linear or cyclic manner Crystallization environments also change, generating new types of individuals, new types of aggregates and multi-aggregates, new types of crusts, they dissolve existing ones Such sequences are mostly different in different caves and cave groups We'll call a mineral ensemble the sequence of sets of mineral individuals aggregates multi-aggregates, and crusts, generated on the mono tone part (to general inundation) of the last crystallization cycle, that is defining the "mineralogical landscape" of a cave element. The most important thing that we can get from this conception is to make the interaction between cave mineralogists and normal cavers when exploring a large cave much more simple. The idea is that in a cave in a group of caves even on a whole karst massif, we can find not only finite, but the very small quantity of ensembles For example in very variable Kugitangtou caves 8 [MALTSEV 1993). And in each ensemble a diagnostics set of 2-5 aggregates and multi-aggregates can be easily found. So, we have a compact convolution of mineralogical information, easy to be used by normal cavers, and informative enough for a mineralogist. It carries almost full description of the last crystallization cycle in an explored cave element, and some part of its modern geological history 5. Non-mineralogical formations. Understanding some logical impropertness, we'll continue considering them as minor mineral bodies, studied via means of mineralogy, residual formations like "gypsum flour" [ANDREICH0UK 1992], or "sulfide mirrors" [MALTSEV I 993], coming from dissolution or sub limation of speleothems. These formations are really rocks and theoretically are to be studied by means of petrography but there are few of them, and the cumulative effect of this error is not significant. Moreover really being rocks they still join into the above described hierarchy approximately on the level of multi aggregates, appearing as elements of crusts and ensembles Morphogenetic environments The morphogenetic environment concept (or crystallization environment) [STEPAN0V 1 97 la, M0R0SHKIN 1986] is the main tool, used in the onthogenetic approach Each individual, aggregate, or multi-aggregate is a product of some morphogenetic environment, and the central problem in genetic modelling is the problem of understanding which environments, and in which sequence produced the mineral object that is studied It's evident, that properties of the morphogenetic environments may be decoded from the mineral bodies features, but it is not very evident, in what manner. For example, the shape of individua l s give only restricted information on the environment only the information on the crystallization kinetics. It's theoretically impossible [STEPAN0V, 1973] to decode even the phase conditions of the environment from only the shape. Individuals and aggregates structure give a bit more information, but also a poor one. The bigger part of the genetic information, allowing the real decoding of genesis may be read from high-level object's texture, mostly through the characteristic symmetry tool. C haract e ri s tic sy mm e tr y By defi n ition, characteristic symmetry of some class of objects, or some media or some process (not a certain object) is a maximum possible symmetry, compatible with existence of this class media, or process [ZHABIN 1979] Characteristic symmetry is usually referred through some geometrical body having such symmetry. The most used kinds are cylindrical symmetry, cone 220 Proceedings of the 12 th International Congress of Speleology, 1997 Sw i tzerland-Volume 1

PAGE 234

symmetry, spherical symmetry. Another useful concept is dissymmetry, that is defined as some symmetry element missing to raise some phenomena to the next level of characteristic symmetry We can find an illustration to this when comparing a cave pearl and a stalagmite. They display the same kind of an aggregate (druse of spherolites), having spherical symmetry in their texture. But in the stalagmite case we have a linear dissymmetry caused by the gravitation control. It s v ery important that the main symmetry and the dissymetry appear on different examination levels in s pite of that their causes exist simultaneously This means that the texture symmetry features raise the contrast between effects caused by different properties of the morphogenetic environment and that s the usual thing Usage of the characteristic symmetry concept comes through the universal Curie symmetry principle: If some cause has some consequence, then the consequence s characteristic symmetry (dissymetry} is a projection of the cause's characteristic symmetry (dissymmetry)" [ST E PAN0V 1970]. When applied to the minor mineral bodies genesis, this mostly gives a one-to-one connection between the texture features and the mass transportation physics in the crystallization environment. For example, s olution flows have spherical mass-transportation physics on lower levels, overlapped by dissymetries on upper levels. Thin film e v aporation has cone symmetry of mass transportation aerosol precipitation cylindrical one etc In reality not only the mass-transportation physics are important. If we take a usual ice hoarfrost, on the bushes level we ll see the cone texture symmetry, caused by a thin film of evaporation processes, and not an inverse cone symmetry, that must appear within sublimation environments The reason of this is simple Ice has low thermal conductivity vapors condensation frees a lot of energy and so does crystallization. Therefore cristallisation is possible So, the areas of preferable condensation become forbidden for crystallization (no heat removal) Only the areas of extra heat removal may appear as crystallization areas and these are areas of higher evaporation. Some of the latest studies by other authors also show that here we deal with a more complicated process, than a direct crystallization from vapors For example P A R UN G0 [ 1983] speaks about quasi-liquid phase on the surface of growing hoarfrost. So the mass-transportation phy s ic s here is based not by the basic properties of the environment but by the energy-transportation physics in it. On the higher level (crusts) of the same hoarfrost we'll see the inverse cone symmetry and air flow geometry controlled dissymetries If we consider the physics of the symmetry features projection we must note that there are only two mechanism s for this the nucleation mechani s m and the individuals s election (competitions) mechanism The first one defines the low-level feature s, the second one high level ones The selection may be fortuitou s or geometric (both with aligning to the s ub s trate surface and both appearing only in spherical symmetry) or driven by the mass-transportation (energy-transportation) ph ys ics (for lower symmetry classes, with aligning to some definite direction or locally defined directions) A good example of fortuitous selection is seen in druses of spherolite s, geometric in dru s es of crystals definitely directed in soda-straws locally defined in fro s tworks and popcorns The high contrast and hierarchical construction of the symmetry and dissymmetry features in the mineral bodies texture allows clear decoding of the morphogenetic environment properties in most of the cases like shown in the example with the hoarfrost. Interactive aggregates and multi-aggregates Not all the mineral aggregates and multi-aggregates are passive products of some crystallization environment. Some of them may create the crystallization environment them s elve s, or significantly modify the parent environment. The grade of thi s effect can be different. The most clear case is shown by the helictites, described in [SLET0V 1985]. Several parallel spherolite bunches, growing together in some chemical starting conditions leave a channel between them At this moment the aggregate goes out of the controls of the parent en v ironment (that can be both gravitational or capillary) creating the special environment with capillary channel, taking the solution from the parent environment and feeding a very local capillary film spot at the end. This makes spherolites grow only in s harp s e ctor s, and this, together with differently located zones of main CO loss and of the final evaporation, keeps the aggregate together. We'll call such aggregates and multi-aggregates strictly interactive and will note that they never belong to some crust. For strictly interactive aggregates and multi-aggregates their definition using structure and texture is not complete We must speak about a new property behavior while interacting with obstacles For the aragonite helictites from Khaidarkan [SLET0V 1985] four types of behavior are found : a) breaking (when perpendicular contact) ; b) reflection ; c) rounding ; d} building together with further movement along the obstacle surface. Growing apart again is also typical. Some multi-aggregates don t create their own environment but modify the parent one So, the multi-aggregate alread y discussed as an example (called multi-corallite ), creates itself the disymmetry of magnesium mass-transportation in the capillary film that controls its differences from a regular corallite Their behavior is complicated enough not the same as beha v ior o f normal corallites the internal magnesium cycle is much more sensitive to the closeness of obstacle than growth itself so some elements of reflection are always seen Such aggregates and multi-aggregates are restrictedly interactive and usually belong to some crust. Some features of interactivity (less grade) may be found al s o for almost any dendritic aggregate and multi-aggregate The interactivity and behavior concepts have some historic roots STEP A NOV [ 1971 b] noted that the shapes of some complicated speleothems (corallites etc.) have a property of high individuality, known before only for organisms. Now we use the term behavior in the same sense as for organisms as a product of interaction with the environment al s o in the s ame sense as for organisms Discussion The two rather important concept s m the ca v e mineral s onthogeny both settled by Stepanov, but never publi s hed currently became the central points of modem s tudies The se are the concept of the hybride texture s, and the u s uall y undere s timated role of re-crystallization It was shown above that the minor mineral bodie s t exture sharpens the contrast between the effects caused by different properties of the morphogenetic environment. With thi s, a minor mineral body may be produced by two e nvironment s with different physics at the same time This lead s to their appearance of hybride textures possessing symmetry features of both environments on the same hierarchy le v el. Consequentl y thi s results in contiguous rows of formations For example these are contiguous lines between shields and helictites [SL ET0V 1985] Symposium 7 : Physical Speleology 221

PAGE 235

stalactites and frostwork [MALTSEV 1997b], etc. Some of these lines are usually interpreted as re-crystallized sequential growth [MAXIM0VICH 1961 ], but the author couldn't find any evidences of this. On the opposite, active growth of such hybride formations is usually seen. With this, the role of re-crystallization in the speleothems growth is much greater, than is usually considered. On high hierarchy levels interactions between the minor mineral bodies and the parent environments become strong. and the re crystallization is one of the main physical mechanisms for these interactions. We can take a usual carbonate soda-straw as a good example of this [MALTSEV I 997b]. The zone of active growth is controlled by hydrodynamic effects while dripping. This causes strikes of supersaturation (through C0 1 loss) without re-mixing of the solution ( on corresponding speed). This are conditions for the skeleton crystals growth, and in reality each soda-straw has a crown of skeleton crystals on the tip. The pressure pulsation inside the channel, caused by the came dripping, is the condition for rapid re-crystallization. And while any re-crystallization process, poorly balanced individuals, like skeleton crystals, re crystallize first [M0ROSHKIN 1976]. And this is the real reason of why oda-straws are monocrystalline. For a lot of other fonnations [MALTS EV 1989, 1993] re crystallization also plays the leading role, as a necessary result of self-organization of complicated systems. References V .ANDREICHOUCK 1992. A genetic classifications of karst cave sediments and deposits (in Russian)/ / ln book: Studies of Ural caves, Penn, p.95-98. A.A.GODOVIK0V, O.1.RYPE E V I.STEPAN0V I 989. Spherolites, spherocrystals, spheroidolites, corespherolites (in Russian) // ovye Dannye o Mineralakh (New Data on Minerals), vol. 36, Moscow," auka", p.82-89 D P.GRIGORJEV, A.G.ZHABI I 975. Onthogeny of Minerals. Individuals (in Russian). Moscow. Ju.M.DYMK0V 1991. Current problems of minerals onthogeny (in Russian) Moscow, "Nauka". V.A.Maltsev 1989. The influence of season changes of the cave microclimate to the gypsum genesis/ / Proc I 0th lnt.Cong.Spel.Vol. III .Budapest, p.8 I 3-814 V A MALTSEV 1993. Minerals of the Cupp-Coutunn cave system, Southeast Turkmenistan // World of Stones, No.2, Moscow, p.5-30. V .A.MAL TSEV 1996. Sul fate filamentary crystals and their aggregates: Proc. Univ. Bristol Spelaeol. Soc., 20(3) pp 171185 V A.MALTSEV., I 997a. Gypsum nests from caves complex mineral individuals: Litologiya i Poleznye Iskopaemye, I 987, no.2. In Russian. V.A.Maltsev, 1997b, Once more on stalactites with "internal" and "external" feeding: Natl. Speleol. Soc. Bull., in press. V A. MALTSEV V.V.K0RSHU 0V, A.A.SEMIK0LENNYKH 1995. Scientific report on the field studies in Kugitangtou caves on March 1995: Dep. RSIC N003-B95, 90p. In Russian G.A.MAXIM0VICH 1961. The genetic row of cave flowstone formatoins ( carbonate speleol ithogenesis ) // Peshchsery o.5(6). In Russian V.V.MoR0SHKIN 1976. On genesis of crystallictite type of aggregates (in Russian) // Novye Dannye o Mineralakh SSSR (New Data on Minerals in USSR). vol. 25. Moscow, auka", p.82-89. V.V.M0R0SHKI 1986. Karstotypic mineralisation (in Russian). // Minera l ogicheskiy Journal, Kiev, vol.8, No 5, p. l 0-20. F.P PARUNG0 I 983 Ice crystals growth at (-8+ / -2)C // Journ. ResearchAthmos Vol.17 o2, 139-156 V A SLET0V 1985. On onthogeny of crystallictite and helictite aggregates of calcite and aragonite from the caves of Southern Fergana (in Russian) // ovye Dannye o Mineralakh ( ew Data on Minerals), vol. 32, Moscow, auka", p 119-127. V I.STEPAN0V I 970 On genesis of so-named "collomorphous" mineral aggregates (in Russian) // In book: Ontogenetic methods of minerals studying. Moscow, "Nauka", p I 98206 V.I.STEPAN0V I 971 a. Crystallisation processes periodity in karst caves (in Russian) Trudy mineralogicheskogo muzeja imeny Fersmana. Moscow, 1971, No .2 0., p 161-171 V.I.STEPAN0V. I 97 I b, Studies of Khaidarkan caves, both hydrothermal and cold; in Velikii A S., Volgin V. J. Ivanov, V. S., and Stepanov, V. I. (eds.), Structural features and genetic sequences for some central Asian Hg-Sb depositions: Unpub Rept., IMGRE All-Russian Geol. Found., p 184-220 In Russian. V.I.STEPAN0V 1973. On aims and methods when studying crystallisation sequences in ore mineral aggregates ( in Russian) // In book: lssledovaniya v oblasty prikladnoy mineralogii i kristallohimii. Moscow, IMGRE, p.3-10. 1.1.TURCHIN0V 1993. Secondary mineral formations op the West Ukrainian gypsum caves (in Russian) // Svet, o.3(9). A.G ZHABIN 1979 Onthogeny of Minerals. Aggregates (in Russian). Moscow. 2 2 2 Proceedings of the 12 t h International Congress of Speleology, 1997, Switzerland Volume 1

PAGE 236

Pyrocoproite (Mg (K,Na) 2 P 2 O 1 monoclinic), a new mineral from Arnhem Cave (Namibia), derived from bat guano combustion B y Jac qu es E .J M artini Council for Geoscience, Private Bag X I 12, Pretoria, 00 I, South Africa Abstract The author describes a pyrophosphate mineral defined by chemical analyses, X-ray diffraction patterns, cell parameters mea ured and calculated densities. and optical characteristics It forms tiny crystals in slag produced by spontaneous combustion of bat guano in the deep part of a cave about 2000 years ago. Compari on is made with other minerals described elsewhere in the world and also derived from fire. Resume L'auteur decrit un nouveau mineral, un pyrophosphate, qui est defini par des analyses chimiques, des diagrammcs de diffraction X. la maille elementaire. les densitcs mesurees et calculee et les proprietes optiques. II forme des cristaux microscopiques dans de s scories produites par la combus t ion spontanee de guano de chauve-souris, dans la partie profonde d une grotte ii y a environ 2000 ans. Le mineral e t compare avec d'autres especes decrites ailleurs lesquelles sont aussi derivees de la combustion de matiere organique 1 Introduction During the course of a speleological investigation conducted by the South West African Karst Research Organisation in 1991, Arnhem Cave. situated 150 km east of Windhoek was studied for mineralogy The author collected a white. porous and friable slag forming a thin layer in a profile of the cave soil. It resulted from the melting of ashes during bat guano combustion about 2000 years ago and the fire is believed to have been pontaneous (MARTI I I 994b). Spontaneou ignition of bat guano became recently less hypothetical after the ob ervation of active burning in a pothole of Martinique Island (Mouret 1996) Two main mineral phases were identified (MARTINI 1994a) in this slag: pyrophosphite (K Ca P,O , monoclinic P2 / m) and arnhemitc ((K. a) Mg (P O ) ,. 58-O) Pyropho phite derived from the melting of orthophosp hates containing the cations ff or H beside K and Ca ,._ During the process of combustion, the oxides of the two fonner elements were lost as volatiles and the orthophosphate anion was depleted of half an oxygen. thus producing pyrophosphatcs Arnhemite is a hydrous pyrophosphate and was thought to have been derived by hydration of an anhydrous precursor (MARTI I 1 994a). A suitab l e primary pha e was found in the same cave in the course of a more thorough sampling in 1995 and is the aim of the paper. 2. Description Ph y.~ica l p roper ti es In the slag, the mineral fom1s grains a few I O microns across. without definite crystalline outline. always corroded and partly replaced by amhemite. It is colourless. The hardne s cou l d not be measured on account of small grain size. o cleavage was observed. The density could not be measured easily by ,, float and sink" in a mixture of tctrabromoethane and acetone due to the small grain size and intimate association with amhemite, which is lighter (2.35 g / cm'). everthe l ess, at density 2.96 g / cm only a few particles sank and this figure represents the density of the mineral free of impurities. This value is consistent with the density calcu l ated after the unit cell (2.98 g/cm '). O p ti ca l p rope r t i es. The mineral is transparent, has a re l atively high rcfringcnce. but a very weak birefringence: a= 1 558. 13 = ?, y = 1 560 No additiona l data could be obtained due to the small size of the grains and their impure nature C h e mi c al da t a Eight chemical analyses were carried out by means of an electron microprobe and gave the average va lue s presented at table I They yield the empirical fonnula. based on 0=7 : (K ., Na .,., ) , 11 (Mg. ,.,. Ca .,,~ ) 110 P ,,.,. 0 or ideally: (K. a) Mg P 0 The minera l is insoluble in water, but readil y soluble in strong acids. Pyrocoproite appears to hydrate more easily than pyrophosphite: among eight studied sample . in only one pyrocoproite was not completely altered into arnhcmite; this was the sample used for this study In contrast, pyrophosphite was always present. Tab l e 1 I 2 J a,0 4.44 1.92 2.51 K O 27.13 29 .2 6 28 20 MgO 14 44 12 .26 14 35 CaO 1.21 2.95 1.42 P O 53.09 53.28 53 87 TOTAL JOO .J I % 99 67 % 100 35 % I Pyrocoproite from Arnhem Cave, average of 8 analyses from the sample containing the anhydrous phase. 2 Pyrocoproite from oven melted amhcmite. from cave; average of 5 analyses. 3. Synthe t ic pyrocoproite, average of 2 analyses. Symposium 7: Phys ical Speleology 223

PAGE 237

Crystallography The X-ray diffraction powder pattern could be indexed by analogy with the pyrophosphite pattern, which shows comparable reflections but shifted towards larger spacings (table 2) Pyrocoproite is therefore its isomorphic magnesium equi v alent. The cell parametres are: a=9.410 b=5.424, c=I2.540 A !3=104 35 Z=4. By analogy with synthetic potassium-calcium pyrophosphate (BROWN et al 1963), the space group is P2 / m Table Z: Diffraction powder patterns: reflections relative intensities, observed and calculated d values, Miller indices. PYROCOPROITE ((K, Na) Mg P 0 ) 1 /111 d ..,, d ,., hkl 63 4.55 4 549 111 40 4 18 4 177 202 14 3.488 3.491 112 17 3.268 3.269 113 29 3.119 3.123 104 46 3 040 3 037 004 94 2.882 2 878 ; 2 882 204 ; 113 100 2 714 2.715 311 6 2.411 2.409 114 29 2 277 2.275 222 40 2.091 2.089; 2.088 222;404 9 2.021 2 025; 2.022 006 ; 315 17 1 976 1.974 ; 1.979 224 ; 402 6 1 697 1.700 206 9 1 656 1.654 133 14 1.623 1.623 026 ; 315 ; 317 11 1.598 1 596 ; 1 597 133;422 9 1 567 1.566 331 Name and type material The name is derived from its genesis by combustion of guano. Type material is deposited at the Geological Survey of Namibia, in Windhoek 3. Synthesis and genesis A concentrat e of arnhemite from material collected in 1991 wa s melted at 700 and produced a slag of pyrocoproite which yielded a X-ray diffraction pattern close to the natural mineral but with s lightly smaller cell parametres: a=9.37 l, b=5.4 I 9 c=I2 519 A, !3=104.48 Microprobe analyses (table I} indicate a lower a content but higher Ca In an attempt to s y nthetise pure K Mg P,0, an equimolar mi x ture of I<., H PO a nd Mg (NH ,) PO 6H ,0 was melted at 1100 C. Pyrophosphates formed but did not combine into pyrocoproite Th e experiment was repeated with the addition of some CaHPO ; 2H,0 and Na P O -I0H 0. Slag formed which after X-ray diffraction and microprobe analyses was found to consist of xenomorphic pyrophosphite and pyrocoproite, with euhedral K-Mg orthophosphate The chemical analysis of synthetic pyrocoproit e is given in table I The X-ray diffraction pattern could not be distinguished from the one of natural material. The succe s s of the synthesis with a and Ca added sugg e st s that the presence of these elements favour the formation of pyrocoproite In particular Na pyrophosphate con s iderably lower s the temper a ture o f fu s ion PYROPHOSPHITE (I<., Ca P 0 ) 1 / 1 100 d ..,. d c111c hkl 2 6 77 6 779 IOI 23 4 73 4 731 111 32 4.32 4 321 202 2 3 614 3 621 112 17 3 386 3.388 113 4 3 242 3 247 301 33 3.137 3 138 004 64 2.987 2.985 113 36 2.974 2 975 204 100 2.821 2 816 311 2 2.497 2.494 114 7 2.365 2.365 222 3 2.250 2 253 305 19 2 174 2.171 222 10 2 163 2 161 404 3 2.093 2 092; 2 093 006 ; 315 5 2.049 2.048 ; 2 049 402 ; 224 3 1.982 1.982 ; 1.981; 1.982 216 ; 124; 405 2 1.759 1.758 206 3 1.721 1.722 133 3 1.685 1 687 107 5 1.660 1.661 ; 1.658 133;422 5 1 631 1.630 331 During the combustion of bat guano, the process of fonnation of pyrocoproite may have been comparable with the laboratory experiments. The starting material may have contained archerite (KH PO ), biphosphammite ((NH.) H,PO J, dittmarite (Mg( H J PO4 -H,O) stercorite (Na (NH )(HPO ) -4 H,0) mundrabillaite and swaknoite (both Ca (NH ) (HPO ) -H,0). which may have been molten into a pyrophosphate slag All these minerals have been identified in the cave associated with bat guano ( MARTINI 1993). 4. The status of pyrocoproite as a mineral species It is relevant to briefly review the definition of a mineral species. Indeed since the majority of the mineralogists are involved with hard rocks only to them this occurrence would seem odd and therefore they may question the validity of pyrocoproite as a mineral species Minerals are defined as natural compounds or elements, implying that chemicals manufactured by man are excluded The boundary between minerals and artifacts however, is not sharp: there i s a transition zone of minerals, which formed by natural processes, but more or less with man s aid Although these border minerals are not accepted by e v erybody as genuine species they have been established by usage ; they are listed in textbooks, like for instance: LACROIX (1913 ), PALACHE e t al 224 Proceed i ngs of the 12 th Internat i onal Congress of Spe l eology, 1997, Sw i tzerland Vo lu me 1

PAGE 238

( 1944) and CAIRN CROSS & DIXON ( 1995) Some of them, like romarchite (ORGA & MA DARINO, 1970) have even been accepted by the International Commission for ew Minerals. These border minerals can be classified in three categories. The first one comprises minerals forming by a natural process, but from artificial material, like for instance the minerals resulting from weathering of artifacts (metal tools, slag) in archaeological sites. Romarchite (SnO) is an example. The second category includes the species resulting from a material and a process. which are both natural, but could not have formed without human intervention They comprise the numerous minerals forming in the mines An example is goslarite (ref. in PALACHE et al 1944), which was first described in a mine, where it resulted from a natural oxidation of ore, but was deposited in an artificial cavity The third category are the minerals due to combustion of natural material in a natural environment, but the fire may be natural or artificial. To the contrary of the first and second categories, it is generally not possible to ascertain if there was a human imput or not from the mineralogical associations or from the environment context alone. An example of such minerals is fairchildite (M IL TO & AXELROD, 1947), resulting from forest fire. P yrocoproite is to be classi tied in this third category, which is the least marked by the influence of man, if there is any. Acknowledgements Mrs E. Hattingh of the Council for Geoscience in Pretoria, performed the microprobe analyses. Mr J.C.E. Marais assisted the author with the sampling and Mr J Bekker, the owner of Arnhem Cave, kindly gave permission to enter and investigate the cave. References BROWN E H ., LEHR J R., SMITH J P. AND FRAZIER A W 1963. Preparation and characterization of some calcium pyrophosphates. Agricultural Food Chem. 11 : 214-222. CAIRNCROSS B AND DIXON R. 1995. Minerals of South Africa. Geol. Soc. South Africa, 290p. LACROIX A. 1913. Mineralogie de la France et ses colonies, 5 vol., Blanchard, Paris. MARTINI J.E.J. 1993 A concise review of the cave mineralogy of Southern Africa. Proc Xlth UIS lnt. Congr. Speleo, Beijing 1993: 72-75. -----------I 994a. Two new minerals originated from bat guano combustion in Arnhem Cave, Namibia. Bull South AJi-. Speleo Ass. 33: 66-69. ------------1994b The combustion of bat guano, a poorly known phenomenon Ibidem, 7072. MILTON C, AND AXELROD J. 1947 Fused wood-ash stone: fairchildite (n.sp) ((,CO ,. CaCO ,, biitschliite (n.sp) 31(, CO ,. 2Ca CO,. 6H 0 and calcite CaCO , their essential constituents Amer Miner 32: 607-624. MOURET C. 1996. Echo des profondeurs Antilles. Spelunca, no 64, 203 ORGAN R .M. AND MANDARINO J.A. 1970. Romarchite and hydroromarchite two new stannous minerals. Can. Miner. 10 : p 916. PALACHE C., BERMAN H AND FRONDEL C. 1944 Dana's system of mineralogy, vol. 2, Wiley, ew York, I 124p Symposium 7: Physical Speleology 225

PAGE 239

The Classification of Cave Minerals and Speleothems By Carol A. Hill and Paolo Forti Abstract The classification scheme of Hill and Forti. as used in the new second edition of Cave Minerals of the World. is presented as a "practical" solution to the problem of classification of cave minerals and speleothems Classification of cave minerals is by crystal class following nomenclature approved by the International Mineralogical Association These classes of minerals include the native elements sulfides. oxides-hydroxides. halides. arsenates. borates carbonates nitrates phosphates silicates. sulfates and vanadates No attempt is made to classify cave minerals by origin as was done in the first edition. Minerals in the outside world are not classified by origin due to the complexity of depositional environments (i e . a mineral may form by a number of mechanisms in a variety of different settings) and for the same reason cave minerals should not be classified according to origin In the past. classification of speleothems has been by (I) morphology. (2) origin or (3) crystallography (fabric) All three approaches have their problems The compromise approach of Hill and Forti is based primarily on morphology (i.e . the shapes one sees). plus whatever is known about origin and crystallography In the second edition of Cave Minerals of the World speleothems are divided into "official" types and subtypes and "unofficial" varieties A speleothem type is defined as a group or category of speleothems sharing one or more morphological characteristics and having a common origin different from other speleothem types (e g .. speleothem type = cave raft) A speleothem subtype is defined as a speleothem which has a structural identity similar to the type. but which has an origin different enough from the type so as to produce a deviant morphology with additional structural elements (e.g. speleothem subtype = cave cone. of type raft) A speleothem variety is defined as a slightly variant morphology produced by differences in water flow. mineral composition. color. crystallinity or otl1er factors. Speleothem varieties exist for both types and subtypes (e g .. "snowflakes"= variety of type raft "volcano cone"= variety of subtype cone) It is proposed that in the future new speleothem types. subtypes. and names be officially approved by a UIS Commission of cave mineralogists 226 Proceed i ngs of the 12 1 h Internat i onal Congress of Speleology, 1997 Sw i tzerland Volume 1

PAGE 240

Crystallographical observations on calc i te rafts from three Romanian caves losif Vie h mann' L u cretia Ghergari 2 Bogda n Petroniu O n ac 2 Speleological Institute "Emil Racovita", Clinicilor 5, 3400 Cluj, Romania "Babes-Bolyai" University, Deptartment of Mineralogy, Kogalniceanu I, 3400 Cluj, Romania Abs tr ac t The paper presents the crystallographical, morphological and mineralogical observations made on seventeen samp le s of calcite rafts collected from three Romanian caves (Hoanca Apei. Stanu Ciutii and Ciungi). The two sides of the rafts ex hibit different morphologies; the upper side is smooth while the underwater si de is rough. The 35 m of "cornflakes-like" depo sit (crumbed calcite raft ) and the filling-drainage mechanism of the pools from Ho anca Apei represent a unique case for the Romanian karst. 1 Introduction The paper emphasises the mineralogical and crystallographical features of the calcite rafts discovered in 19 87 in three pools and an artificial excavation (Ana's hole) in Ho anca Apei Cave (Bihor Mts.). The observations were compared with those made on simi lar speleothems from Stanu Ciutii (Padurea Craiului Mts.) an d Ciungi (Somesan Plateau) caves. In Romania such speleothcms were previously de scribe d from caves at Sura Mare (LASCU, pers comm.). Humpleu. Piatra Altarului (FRATI LA, I 996). Ponora s Batranului Osoi (ONAC. I 996). The calcite raft problem was world-wide treated (generally or in detail) by SALZER ( 1942). BLACK ( 1953) POMAR et al. (1975, 19 76), HILL! & FORTI (1986), VIEHMANN (1992), FORTI & CHIESI (1995). The pools from Hoanca Apei are located in a 15 per 9 m room (Sala Plutelor) at the far end of an upward si de passa ge, 220 m behind the cave entrance. These hollow basins (down to 45 cm) are placed on an important deposit that consists of millions of crumbed calcite rafts. The excavation done to a depth of 165 cm in the center of the room, revealed that the crumbed calcite rafts sed iment has a thickness of 90 cm. The rest of the deposit consist of moonmilk that subsequently was formed from accumulated calcite rafts In the eastern part of this room there i s a sma ll s pring ha v ing usually a discharge of8 cm '/ s. After heavy rain s, the di sc harge increase considerably so that the whole room is flooded. The pools have been found to be temporarily emptied or filled up with water. 2. Experiments Two types or experiments were periodicall y carried out in the last I O years: chromatic marking of the calcite rafts and tests concerning filling-drainage mechanism of the pools The two experiments proved that the calcite rafts lying on the bottom of an empty pool are brought back to the surface by a water that moves per ascensum within the pool (VIEHMA N, 1992) In Ana's hole, the water filling proces s lasted about 90 days. which gives an average value of 1.8 cm in 24 hour s. In parallel with these experiments we measured the carbon dioxide concentration in the air with a Draeger device (type WG-2M) and the water temperature in the spring, pools and cave atmosphere. While the carbon dioxide concentration have shown similar values everywhere in the cave, the temperature increased from 7. I C in the spring to 7.4 C in the pools and to 7.8C in the cave atmosphere. The relati ve humidity measured within the cave was 89%. The surface of most of these rafts range between 1.2 and 3.4 cm However. the largest calcite ran floating in one of the pools could reach. before breaking into pieces reached a remarkable area of 600 cm'. BLACK ( 1953) ha s described aragonite rafts in Carlsbad Caverns having more than 900 cm Following our calculation the calcite rafts deposit in Hoanca Apei Cave reaches 35 m '. 3 Mineralogical morphological and crystallographical observations Thirteen samples were collected in Hoanca Apei Cave as following: 9 samples (4-12) collected from different depths of the calcite rafts deposit (4, 8, 12 16. 20, 24, 28, 32 and 36 cm), 2 thin rafts formed in Ana's hole (samp l e 3 collected in 1995 a nd sample 13 collected in 1996), and 2 samples from the other two pools located in Sala Plutelor (sample I and 2). Three samp les were collected from the Stanu Ciutii Cave; one from the water surface (sample 14), another sample from the bottom or the same pool (sample 15) and the third one from the bottom of a dried pool. Only one sample was collected from Ciungi Cave (sample 17) The analyses made on these calcite rafts sa mple s consist of observations with scanning electron micro sco pe (SEM), transmission polarizing microscope and binocular. Several X-ray diffractions were carried-out in order to te s t the mineralogical composition. With respect to the s hape and size of the crystals that build the calcite raft, our samples can be ascribed to the following 3 groups: I) Calcite rafts made of large, well developed crystals (> I 00 mm). This type was observed in Stanu Ciutii, Ciungi and below 15 cm in the calcite rafts depo s it from Hoanca Apei Cave. 2) Calcite rafts consisting of medium size crystals ( I 0-100 mm) with many structural defects (imperfections) (e.g., natural pools from Hoanca Apei Cave and above 15 cm in the calcite rafts deposit from the ame cave). 3). Calcite rafts composed of small cry s tals having many structural defects (Ana 's hole). Mine r a l ogica l co mp os iti o n Optical properties and X-ray analyses showed the pre ence of calcite as main mineral in all samples te s ted The white rafts made up of almost transparent crystals, entirely consist of calcite while the brown ones are impregnated with clay mineral s. The diffraction patterns obtained of the rafts collected in the artificial pool from Hoanca Apei (sample 3, first collection) show 3 peaks of low intensity (3.24-15; 4.25-1; 1 .8 1-1) that were ascribed to Symposium 7: Physical Speleology 227

PAGE 241

quartz. Under the microscope this mineral cannot be seen due to the small grain size. In the sample 13 recently collected, the quartz is missing Morphology of the rafts The two sides of the rafts exhibit different morphologies. The upper side is smooth and has a granular structure, punched from place to place (small holes) (Fig. 2a). If the upper side was corroded than the crystals are clearly outlined (Fig. 2b). The crystals are mostly skeletic (hopper morphology) and grow with their c-axis at various angles against the water surface The underwater side is rough due to some prismatic crystals that grow perpendicular and at various angles against the raft surface (Fig. 2c, d). Other crystals develop around isolated crystals, resulting finally in radial shaped groups. On the upper side of some calcite rafts, in small holes between the crystals, we found a fine yellowish silky powder a b C I 1010 1010 a' a" b' Figure 1: The habit of crystals that composed the calcite rafts The calcite rafts samples framed into the first group consist mostly of euhedral crystals limited by well developed crystallographic faces. The crystals have prismatic habit the most common crystallographical forms being the first order hexagonal prism { I 010} that usually shows striations along it. Another frequent form is the positive rhombohedron with smooth faces. The prismatic crystals differ as the sixth pri sm faces have various degrees of development. Hexagonal prisms with equal developed faces are rare (Figs. I a, a'). As a rule, three of the faces are better developed when compared to the other three that are smaller (sample 14, 16 and less 15 from Stanu Ciutii, 17 from Ciungi) (Figs. 2d, I a"). The unequal growth of the faces leads to the development of two types of trigonal prisms: positive {IOIO} (samples I 5, I 6 from Stanu Ciutii, 17 from Ciungi) (Figs. I b b') and negative {O 110} (sample 15-Stanu Ciutii) (Figs. 2c, d and 2c, c'). Sometimes a steep rhombohedron {4041} or {0441} forms (samp le 14-S.!_anu Ciutii). When the calcite crystals grow parallel to the (I O 10) face a tabular-prismatic habit can be recognized. Among other crystals with different combinations of (sample 17 from Ciungi and samples I and 13 from Hoanca Apei). This powder consists of threadshaped calcite crystals having almost similar sizes (Fig. le). This type of crystals is characteristic for moonmilk deposits. The maximum thickness of the calcite rafts measured in Hoanca Apei was 0.288 mm, while the rafts that repeatedl y sink and float are between 0.067 and 0.21 mm. Crystal morphology The crystals have different sizes depending on the location and differs even in the same cave as a matter of presence or absence of the impurities The calcite impregnated with clay minerals always forms small crystals whereas the pure ones are larger. Sometimes the calcite crystals are covered by a thin layer of cryptocrystalline calcite (sample 17, Ciungi). d 0112 c' 1120 crystallographic forms, we want to mention: the { 1011} rhombohedron as a dominant form (Fig. I e) associated with sec~nd order hexagonal prism {1120} (subordinate) and the { 1011} and {01 12 } r~mbohedrons together with second order hexagonal prism { 1120 }(Fig. If). Crystals on which both types of prisms (first and second order) can be distinguished are rare Faces that belong to the second order prism show vertical striations due to alternating growth of the {1010} and {0110} faces. (sample 15, 16 from Stanu Ciutii). A second generation of crystals represented by first order prism and positive scalenohedron { 5491}, fill the holes left between the other crystals. Skeletal habit is a typical feature of the crystals that build up calcite rafts framed by us to the 2nd and 3rd group. The crystals may form parallel growth of tabular crystals developed parallel to the prism face {0110} or prismatic crystals with triangular habit due to the presence of the negative prism (Figs 2c, f) The prisms tip is either a corrosion surface 228 Proceedings of the 12 th International Congress of Speleology 1997 Switzerland Volume 1

PAGE 242

la lb le ld le 1f Figure 2: Calcite rafts from Hoa11ca Apei Cave. SEM: upper side of the raft: a) (sample 12); b) (sample 11); 1111denvater side of the raft: c) corroded crystals (sample 4); d) (sample 1 J); e) threadshaped calcite, grow11 betwee11 the crystals (sample /); j) corroded calcite prisms (sample 4). Symposium 7 : Physical Speleology 229

PAGE 243

generated by dissol~tion (Fig. 2f) or smooth faces of the positive rhombohedron { I O 11} (Fig. 2d) Calcite rafts affected by the dissolution processes were remarked in Hoanca Apei (both types of pools) and Ciungi caves. 3. Statistical remarks The statistical results are based on determinations made on more than I 00 crystals from all 17 calcite raft samples. The most common crystallographic forms are: {1011 }, { 1010}, { 1120 } {4041 }, {0441 }, { 0112 }, {5491} (Fig 3a) 100 80 60 ,, ... = ., ::, '=" 40 ., ... loo 20 0 0 0 N 1-t-MN I VIV l -10\ 0 0 V O V 0 V O 'l"'I Crystallographical forms 2 3 4 umber of crystallographical forms/cryslal Figure 3: Persistance and frequency of the crystallographica/ forms The persistence has high values for two of the crystallographical forms : {IOI I } and { IOI O } When analysing the combination of forms the highest frequency was found at crystals with two crystallographic forms (Fig 3b) This means that the growth conditions are similar even when the floating-drying process is periodically resumed Three types of contact twins were identified Their twinning planes are (IOI 0) (0112) and (0221 ) 4. The genesis of the calcite rafts The most important factor in calcite rafts genesis as in any other calcite speleothem is the supersaturation of the solution. This state can be achieved through one of the following processes : the decrease of CO partial pressure at the water-air interface; the loss of CO ,, by diffusion into open air will increase the s olution fluidity, speeding up the crystallisation; the increase of the solution temperature in pools will diminish the solubility of calcite bicarbonate and carbon dioxide. evaporation ; Very fine clay minerals or calcite crystals particles, as well as organic material (POMAR e t al ., 1975 1976) can act as nuclei around which calcite rafts will develop The size of the crystals is directly related to the number of nuclei ; large crystals will grow from pure solution, whereas an impure solution will encourage the growth of small crystals The presence of impurities and the CO release lead to the apparition of structure defects parallel growth and skeletal crystals The reason for calcite rafts floating was found to be the surface tension and the presence of skeletal crystals which contain holes filled with gas in between them 5. Conclusions The calcite rafts samples we studied were formed in two types of pools In Hoanca Apei all three basins lie on a huge calcite rafts deposit and are filled with water through a per ascensum process, while in the other two caves the pools are typical rimstone dams being filled by flowing or seeping waters Because of the different mechanisms the calcite rafts in Hoanca Apei will periodically be brought to the surface of the pool whereas in Stanu Ciutii and Ciungi the calcite rafts once formed will either be attached to the sides of the pool s or will s ink under their own weight Both sides of the rafts consist of calcite crystals that grow perpendicular and at various angles against the raft surface. The differences between the two sides are just morphologically ones ; the upper side is smooth while the underwater side is rough The dissolution process that affected the calcite rafts from Hoanca Apei and Ciungi is due to the aggressive waters which enter the cave after havy rains If in Ciungi only the upper side of the rafts show corroded crystals in Hoanca Apei both sides exhibit this phenomenon The explanation of this peculiar situation is that when the room become flooded, the aggressive water affects the upper side of the rafts but when the pools are filled up per ascensum only the crystals from the underwater s ide will be corroded The 35 m 1 deposit of "cornflakes-like" (crumbed calcite rafts) and the filling-drainage mechanism of the pools from Hoanca Apei represents a unique case for the Romanian karst References BLACK, D M 1953. Aragonite raft s in Carlsbad Cavern s, New Mexico. Science 117 : 84-85 FORTI, P CHIESI, M. 1995 A proposito di una particolare forma di calcite flottante osservata nella Grotta Grave Grubbo CB 258 (Verzino, Calabria). Atti e Memorie Comm Grotte "E Boegan" 32 : 43-53 FRA TILA, G R. 1996 Mineralogia si cristalografia speleotemelor din pesterile bazinului Somesului Cald, pp. 48-49. B.Sc Thesis "Babes-Bolyai" University HILL C A. FORTI P 1986. Cave minerals of the world pp 52-53 National Speleological Society, Huntsville, Alabama POMAR L. GINES A. FONT ARNAU R 1976 La s cristalizaciones freaticas Endis 3 : 3-25. POMAR, L., GINES, A GINES, J ., MOY A G ., RAMON G. I 975 Nota previa sobre la petrologia y mineralogia de la calcita tlotante de algunas cavidades de! Levante Mallorquin Endis 2 : 3-5 SALZER, H 1942 Ober das Vorkommen von Kalkhautchen in dem aufgellassenen Gipsbergwerk "Seegrotte" bei Modling (ND) Karst und Hohlenkiinde 43 : 66-74 VIEHMANN I. 1992 Experimental methods m s tudying the cave rafts Theor Appl. Karstol. 5 : 213-215 230 Proceed i ngs of the 12 1h Internat i onal Congress of Speleo l ogy 1997 Sw itz erla n d Volu m e 1

PAGE 244

Mineralogy of crusts and efflorescences from Humpleu cave system L uc re tia G h e rgari Bo g d a n P e troniu Onac G h e or g h e F r a til a B abes-Bo l yai" University, D epartment of Minera l ogy, Kogal n iceaou I, 3400 Cluj, Romania Abstract The paper presents the results of a mineralogic and crystallographic study d one on crusts and efflorescences collected from Humpleu cave system. Besides calcite, a r agonite and gypsum, a ric h association of phosphates (hydroxyapatite carbonatehydroxyapatite, b rushite, co ll opha n e), a l so the presence of amorphous silica (opal) are highlighted. Our results rely on observations made with scanning electron microscope, transmiss i on polarising microscopy, t h ermal and spectral analyses and X-ray diffraction. Introduction Humpleu cave system is composed of Pestera Mare din Valea Firii cave and Ave n u l din Poienita potho l e and it is located in the NW part of Apuseni Mountains, on Firii Valley, left side tributary of Somesul Ca l d R