9 789612 541613 C A R S O L O G I C A Karst rocK Features • Karren sculpturing editors: angel ginés, Martin Knez, tadej slabe, Wolfgang Dreybrodt KRF•1 • OK.indd 1 15.12.2009 10:41:35 Carsologica 9 Urednik zbirke / Series Editor Franci gabrovšek angel ginés, Martin Knez, tadej slabe, Wolfgang Dreybrodt (eds.) Karst rocK Features – Karren sculpturing Recenzenta / Reviewed by andrej Kranjc in/and rajko pavlovec Jezikovni pregled / Language review alenka Možina, trevor shaw, Waine tuttle Oblikovanje in prelom / Design and typesetting Brane Vidmar Oblikovanje ovitka / Cower design Barbara Hiti Risane priloge / Drawing iztok sajko Obdelava fotografij / Photo editing Marko Zaplatil Izdajatelj / Issued by inštitut za raziskovanje krasa Zrc saZu, postojna / Karst research institute Zrc saZu, postojna Zanj / Represented by tadej slabe Založnik / Published by Založba Zrc / Zrc publishing, ljubljana Za založnika / For the publisher oto luthar Glavni urednik / Editor-in-Chief Vojislav likar Tisk / Printed by littera picta d. o. o., ljubljana Naklada / Printrun 600 Izdajo knjige je podprla Javna agencija za raziskovalno dejavnost rs Subsidized by slovenian research agency cip - Kataložni zapis o publikaciji narodna in univerzitetna knjižnica, ljubljana 551.44(082) Karst rock features = Karren sculpturing / editors angel ginés . . [et al.] ; [risane priloge iztok sajko]. - ljubljana : Založba Zrc = Zrc publishing, 2009. - (carsologica ; 9) isBn 978-961-254-161-3 1. Vzp. stv. nasl. 2. ginés, angel 248380928 © 2009, Založba ZRC, ZRC SAZU. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the publisher. KRF•1 • OK.indd 2 15.12.2009 10:41:36 Karst rocK Features Karren sculpturing Editors: Angel Ginés Martin Knez Tadej Slabe Wolfgang Dreybrodt Postojna – Ljubljana 2009 KRF•1 • OK.indd 3 15.12.2009 10:41:36 KRF•1 • OK.indd 4 15.12.2009 10:41:36 ForeWorD rock forms are important traces of the formation and development of surface karst features. on various karren their record is especially rich, revealing to us the many factors that in diverse conditions formed the karst surface on various carbonate and other rock. We have tried to present the most characteristic rock forms and through them the most important factors and processes in the formation of the karst surface, the methods of studying them, and the most outstanding examples. Forty-nine contributing authors offer a wide spectrum of content and examples of rock forms from many karst regions around the world. During the preparation of the book, many new and interesting discoveries emerged that strongly encouraged further research into this extremely indicative and often aesthetically attractive part of our karst natural heritage. The editors 5 KRF•1 • OK.indd 5 15.12.2009 10:41:36 KRF•1 • OK.indd 6 15.12.2009 10:41:36 contents ROCK FORMS 1 Karrenfield landscapes and karren landforms ................................................................................................................... 13 Angel Ginés 2 Physics and chemistry of dissolution on subaerialy exposed soluble rocks by flowing water films ...........................................................................................................................................................................25 Wolfgang Dreybrodt and Georg Kaufmann 3 Biokarstic processes associated with karren development ..................................................................................... 37 Heather Viles 4 Karren simulation with plaster of Paris models .................................................................................................................. 47 Tadej Slabe 5 The problem of ril enkarren development: a modelling perspective ............................................................... 55 Matija Perne and Franci Gabrovšek 6 Some methodologies on karren research ..............................................................................................................................63 Gábor Tóth 7 Microril s ...........................................................................................................................................................................................................73 Lluís Gómez-Pujol and Joan J. Fornós 8 Cavernous weathering .........................................................................................................................................................................85 Andrew S. Goudie 9 Kluftkarren or grikes as fundamental karstic phenomena ........................................................................................89 Helen S. Goldie 10 Subsoil shaping ........................................................................................................................................................................................ 103 Anikó Zseni 11 Significant subsoil rock forms ........................................................................................................................................................123 Tadej Slabe and Hong Liu 12 Kamenitzas ...................................................................................................................................................................................................139 Franco Cucchi 13 Trittkarren ......................................................................................................................................................................................................151 Márton Veress 14 Corrosion terraces, a megaausgleichsfläche or a specific landform of bare glaciokarst ...................161 Jurij Kunaver 7 KRF•1 • OK.indd 7 15.12.2009 10:41:37 15 Rainpits: an outline of their characteristics and genesis ........................................................................................... 169 Angel Ginés and Joyce Lundberg 16 Ril enkarren ................................................................................................................................................................................................. 185 Joyce Lundberg and Angel Ginés 17 Rinnenkarren ..............................................................................................................................................................................................211 Márton Veress 18 Meanderkarren ........................................................................................................................................................................................223 Márton Veress 19 Wandkarren ................................................................................................................................................................................................ 237 Márton Veress 20 Coastal karren ........................................................................................................................................................................................... 249 Joyce Lundberg CASE STUDIES 21 Limestone pavements in the British Isles ............................................................................................................................ 267 Peter Vincent 22 Case studies of grikes in the British Isles ............................................................................................................................... 275 Helen S. Goldie 23 The karrenfields of the Muota val ey: type localities of the main karren types after the nomenclature by Alfred Bögli ....................................................................................................................................................... 291 Michel Monbaron and Andres Wildberger 24 The nature of limestone pavements in the central part of the southern Kanin plateau (Kaninski podi), Western Julian Alps ........................................................................................................................................299 Jurij Kunaver 25 Karren features in the Dachstein mountain ........................................................................................................................313 Gábor Tóth 26 Glaciokarst landforms of the lower Adige and Sarca val eys ................................................................................ 323 Ugo Sauro 27 Karren in Patagonia, a natural laboratory for hydroaeolian dissolution ....................................................... 329 Richard Maire, Stéphane Jail et and Fabien Hobléa 28 Cutters and pinnacles in the Salem limestone of Indiana .......................................................................................349 Arthur N. Palmer 29 Types of karren and their genesis on the Velebit mountain .................................................................................. 359 Dražen Perica and Tihomir Marjanac 30 Mid-mountain karrenfields at Serra de Tramuntana in Mal orca island ........................................................ 375 Joaquín Ginés and Angel Ginés 31 Tropical monsoon karren in Australia ..................................................................................................................................... 391 Ken G. Grimes 32 The tsingy karrenfields of Madagascar ...................................................................................................................................411 Jean-Noël Salomon 33 The pinnacle karrenfields of Mulu ............................................................................................................................................. 423 Mick Day and Tony Waltham 34 Arête and pinnacle karst of Mount Kaijende .................................................................................................................... 433 Paul W. Wil iams 8 KRF•1 • OK.indd 8 15.12.2009 10:41:37 35 Lithological characteristics, shape, and rock relief of the Lunan stone forests ....................................... 439 Martin Knez and Tadej Slabe 36 Two important evolution models of Lunan shilin karst............................................................................................. 453 Linhua Song and Fuyuan Liang 37 Solution rates of limestone tablets and climatic conditions in Japan ........................................................... 461 Kazuko Urushibara-Yoshino, Naruhiko Kashima, Hiroyuki Enomoto, Takehiko Haikawa, Masahiko Higa, Zenshin Tamashiro, Tokumatsu Sunagawa and Eisyo Ooshiro 38 The rock cities of Rosso Ammonitico in the Venetian Prealps ..............................................................................469 Ugo Sauro 39 Coastal eogenetic karren of San Salvador island ............................................................................................................475 John E. Mylroie and Joan R. Mylroie 40 Coastal karren in the Balearic islands ......................................................................................................................................487 Lluís Gómez-Pujol and Joan J. Fornós 41 Coastal and lacustrine karren in western Ireland ...........................................................................................................503 David Drew 42 Solution pipes and pinnacles in syngenetic karst ..........................................................................................................513 Ken G. Grimes 43 The karren landscapes in the evaporitic rocks of Sicily ............................................................................................. 525 Giuliana Madonia and Ugo Sauro REFEREnCES ........................................................................................................................................................................................................ 535 9 KRF•1 • OK.indd 9 15.12.2009 10:41:37 KRF•1 • OK.indd 10 15.12.2009 10:41:37 Rock FoRms KRF•1 • OK.indd 11 15.12.2009 10:41:37 KRF•1 • OK.indd 12 15.12.2009 10:41:37 KarrenFielD lanDscapes 1 anD Karren lanDForMs Angel GINÉS Karren is used today as a generic term for typical area of glattalp-Muotatal, the international lit- sculpturing features on surfaces of exposed solu- erature on karst geomorphology (trimmel, 1965; ble rock. as a whole, their outstanding variety of Monroe, 1970; Jennings, 1971; sweeting, 1972) shapes constitutes the kind of solutional exokarst- favoured progressively the change from lapiés to landforms of smaller size, spanning from milli- karren, even though introducing new connota- metres to more than 10 metres. But on the other tions in the term. indeed sweeting (1972) explains hand, quite large extensions of rocks sculpted by that “the name karren was originally used to de- karren spread over many square kilometres show- scribe solution runnels cut into limestones, but it ing in many cases a characteristic pattern of fur- is now used for the whole complex of micro-forms rows and grooves, separated by sharp ridges or that occur on outcrops of karst limestones”. More crests, that stands out in the landscape. precisely, White (1988) introduces his section on older references on karren description (Heim, solutional sculpturing pointing out that “direct 1878; chaix, 1895; eckert, 1902) were in the begin- rainfall, sheet wash, channelized flow, and perco- ning related geographically to the alpine karsts lating flow under various kinds of mantle materi- from glattalp-Muotatal (schwyz, switzerland), als produce a myriad of small sculpturings on the Désert de platé (savoie, France) and gottesack- bedrock surface of soluble rocks”, and then states erplateau (allgäu, germany) respectively. terms that “these forms, as a class, are called karren, a as Karrenfeld, Karren, lapiaz and lapiés, that are german term that has come into general usage today present in karst descriptions, appear there- in english”. looking back today over the evolu- fore in the literature, at the end of the 19th century, tion of karren terminology, the choice made dur- as a particular expression of high mountain karst. ing the seventies of an anglicized version of the later on, in parallel with growing speleologi- Bögli’s terms seems to be successful in avoiding cal research in europe, the words lapiés and kar- the uncontrolled burst of new terms. ren were early incorporated as synonyms through Karren landforms develop not only on lime- the karst literature after the Martel᾽s (1921) and stone outcrops. They occur in bare carbonate, gyp- cvijić᾽s (1924) publications, including also several sum and salt rocks (Figures 1, 2, 3) or under soil mid-mountain locations as Var (provence, France) cover. The presence of karren is widespread wher- and Velebit (Dalmatia, croatia). But subsequently, ever karstic rocks are exposed, but sometimes especially under the influence of new research many significant – but not especially conspicuous developed by Bögli (1951, 1960a) in the classical – forms are ignored or neglected. right up for the 13 KRF•1 • OK.indd 13 15.12.2009 10:41:37 Karst Rock Features • Karren Sculpturing present time, the history of karren and karrenfield description has been strongly biased towards the most striking karren landforms: for instance, ril- lenkarren and spectacular solution runnels (Fig- ure 4), as well as the scenic karrenfield landscapes of alpine and tropical environments (Figures 5, 6), are very much better documented than rainpits and monotonous karrenfield surfaces related to arid environments. according to a. ginés (1995) and Fornós and Figure 1: Rillenkarren features developing on micritic ginés (1996), karren features are, probably even limestones at Serra de Tramuntana karrenfields (Mal- more than dolines, the most widespread karstic lorca, Spain). landform. This is obviously unquestionable if the term karren turns to be applied to any small-sized solutional sculpturing, as it is done in the more recent literature, because nearly every karstic ter- rain exposed to surface erosion contains karren forms. some of them are even microscopic solu- tional and biokarstic features related to the highly specific limestone weathering processes. others remain hidden under soil or buried by clastic sed- iments, as happens with the typical subcutaneous karren features or cryptolapiaz. Finally, there is a great variety of karren forms – the better known, being exposed to open air so that their current growth and development remain basically under the control of atmospheric precipitations – which Figure 2: Rillenkarren features developing on gypsum blocks, after one century of exposure to rainfal , at the become integrated as complex karren assemblages Minoan Knossos Palace (Crete, Greece). Width of view into larger karstic features called karrenfields; that is 15 cm. is, the former Karrenfelder. in many cases the role of such karrenfields in the autogenic recharge of karst aquifers could be even more important than karst depressions or dolines. classification of karren landforms is obvious- ly an open debate. since it is a matter of natural forms – and therefore it is characterized by inher- ent difficult description and intermingling origin – several different criteria can be considered as sound bases for classification. probably the better option is to move from one classification to an- other depending on the approach to the problem that is being studied. one useful analysis of kar- ren landforms we are suggesting here relies on a Figure 3: Rillenkarren features developing on salt rock at three-level approach, based on the following scale Cardona mountain (Catalonia, Spain). of decreasing size and complexity: karrenfield 14 KRF•1 • OK.indd 14 15.12.2009 10:41:41 Angel Ginés, Karrenfield landscapes and karren landforms landscapes, karren assemblages and elementary karren features. Karrenfield landscapes interest in karrenfields started in the second half of 19th century, probably due to the increased ap- preciation of the wilderness and scenic values of alpine landscapes. significantly, the two oldest german descriptions of karren areas appeared in journals edited by alpine societies (schweizer al- penclub and Deutscher und Österreichischer al- Figure 4: Conspicuous solution runnels on the sides of a penverein) and both include the word Karrenfeld jagged limestone pinnacle, 8 metres tal , called Es Ca- in their titles. ancient literature on karren land- mel de Lluc (Serra de Tramuntana, Mallorca, Spain). forms was clearly caught by the description of bi- zarre shapes and impressive landscapes. The most conspicuous karren landforms were reported dur- ing this time in the geographical literature, with special emphasis on rillenkarren, rinnenkarren, kamenitzas and trittkarren-features. in this way, the term karrenfeld became progressively associ- ated with high mountain landscapes. additional descriptions of bare limestone out- crops gradually demonstrated the existence of large areas of karren landscapes in the form of the so-called limestone pavements of the British isles as well as by occupying many barren and defor- ested limestone mid-mountains around the Medi- Figure 5: A typical mountain-karrenfield, with plenty of terranean basin. Their particular scenic value fo- wal solution runnels related to snow melting, at 4,500 cused again the interest on the strange forms, but metres a.s.l. in Yulong mountain (Yunnan, China). at the same time these descriptive studies allowed to some kind of wider generalization promoting the use of a revisited concept of karrenfield as it is implicitly postulated in sweeting (1972). all through the seventies and the eighties a remarkable number of tropical and equato- rial karst settings have been described over the world; namely in new guinea (arête and pinna- cle karst of Mount Kaijende), northern australia (chillagoe, Queensland), Borneo (gunung Mulu, sarawak), southern china (stone forest of lunan) and Madagascar (Bemaraha tsingy), among oth- Figure 6: Extremely dissected pinnacles and pil ars in ers. Many of them are in fact giant karrenfields, as the subtropical karrenfield type-locality of Lunan Stone assumed in Ford and Williams (2007). once again Forest (Yunnan, China). 15 KRF•1 • OK.indd 15 15.12.2009 10:41:45 Karst Rock Features • Karren Sculpturing Table 1: Classification of karren forms. Yellow areas enclose elementary karren features. Green areas enclose complex large-scale landforms, namely karren assemblages and karrenfield types (after Ginés 2004, slightly modified). SOLUTIONAL KARREN FORMS SYNONYMS AGENT BIOKARSTIC BORINGS IRREGULAR WETTING ETCHING TINY WATER MICRORILLS RILLENSTEINE FILMS STORM RAINPITS SOLUTION PITS SHOWERS DIRECT RILLENKARREN SOLUTION FLUTES RAINFALL SOLUTION RINNENKARREN RUNNELS CHANELLED WALL KARREN WANDKARREN WATER FLOW DECANTATION RUNNELS MEANDERING RUNNELS MÄANDERKARREN STANDING KAMENITZAS SOLUTION PANS WATER SOLUTION BEVELS AUSGLEICHSFLÄCHEN TRITTKARREN HEELSTEPS SHEET WASH WATER FLOW COCKLING PATTERNS SOLUTION RIPPLES TRICHTERKARREN FUNNEL KARREN SHARPENED EDGES LAME DENTATES SNOW MELTING DECANTATION RUNNELS MEANDERING RUNNELS MEANDERING MÄANDERKARREN ICE MELTING RUNNELS INFILTRATION GRIKES KLUFTKARREN RUNDKARREN ROUNDED RUNNELS SOIL SMOOTH SURFACES PERCOLATION BODENKARREN, SUBSOIL TUBES WATER SUBCUTANEOUS SUBSOIL HOLLOWS KARREN CUTTERS UNDERCUT HOHLKARREN RUNNELS CLINTS FLACHKARREN PINNACLES SPITZKARREN COMPLEX PINNACLE KARRENFIELD PROCESSES LIMESTONE PAVEMENT STONE KARRENFELD FOREST ARÊTE KARST 100m- 0-1mm 1mm-1cm 1-10cm 10cm-1m 1m-10m 10-100m 1km- LAPIÉS 1km 16 KRF•1 • OK.indd 16 15.12.2009 10:41:45 Angel Ginés, Karrenfield landscapes and karren landforms these impressive labyrinths of pinnacles and verti- represented by limestone pavements and stone cal pillars are mainly appreciated in the geomor- forests (table 1). phological literature by their striking landscapes in his “glossary of Karst terminology”, Mon- rather than by being extreme examples of karren- roe (1970) defines Karrenfeld as “an area of lime- fields. nowadays, it is not enough to describe in a stone dominated by karren”. More especially, detailed manner such spectacular landscapes: fur- Bögli (1980) states that “karren fields appear as ther studies on the role of karrenfields in karst hy- bare karst and consist of the sum of exposed and drology as well as on the genesis and development half-exposed karren, and occasionally also of cov- of the continuum karrenfield-epikarst call for a ered karren which have become exposed”; moreo- new approach to this topic. Having this in mind, ver, he points out that “they attain the size of a probably it could be useful to redefine and enlarge few hectares to a few hundred square kilometers”. the term karrenfield in order to embrace whatever ginés (1999a) suggests attaining a minimum of large surface of exposed soluble rock, sculpted by 50% of exposed bare rock – extended over large karren and showing topographic characteristics limestone outcrops – to be considered karren- comprised between two significant end members fields; but obviously there exists a clear hydro- Figure 7: The most significant karren assemblages from the mid-mountain karst of Serra de Tramuntana (Mallor- ca, Spain) usual y include rillenkarren, trittkarren-heelsteps and cockling patterns from which collected water is drained into wider solution runnels. 17 KRF•1 • OK.indd 17 15.12.2009 10:41:46 Karst Rock Features • Karren Sculpturing logical continuity along exposed and soil-covered rocks and – following these pathways downwards – subsoil karren features appear as direct links to- wards the most efficient epikarst drains. since the complexity of karst subsurface poses many chal- lenges to understanding the transition between karren landforms and the epikarst (Williams, 1983; Klimchouk, 2000c; Jones et al., 2004; Wil- liams, 2008), special attention must be paid in the future to describe in detail this subcutaneous part of the exokarst. The growing interest on subsoil Figure 8: Funnel shaped trichterkarren from an alpine karren assemblages (slabe, 1999) could focus ad- limestone outcrop at 2,050 metres a.s.l. in the Vallon ditional research on the overall evolution of kar- des Morteys (Fribourg, Switzerland). Width of view is renfields and epikarst. 15 cm. Karren landform assemblages Karrenfields are in fact major recharge areas feed- ing the epikarst, but at the same time karrenfields are vast bare rock areas dominated by little-scale karren sculpturing. Filling this remarkable size- gap, in his explanation on “the genetic system of karren forms”, Bögli (1980) states an intermedi- ate stage of integration gathering the so called sin- gle karren forms into greater complex karren forms and indicating clints and pinnacles as examples of such level of classification. indeed, limestone pavements in alpine or glaciated karst are char- acterized by a clint and grike topography, defined initially as Flachkarren (Bögli, 1951, 1960a), in the same manner than intertropical stone forests are characterized by a pinnacle and trench topogra- phy, defined initially as spitzkarren (Bögli, 1951, 1960a). For this reason, i personally think that the old distinction between Flachkarren and spitz- karren yet remains a very useful criterion, espe- cially suitable for landform modelling. clints and/or pinnacles become dissected and sculpted by characteristic karren assemblages (Figure 7) whose distribution patterns over the rock faces have not been sufficiently studied until Figure 9: Sharpened edges, topped with their charac- teristic flat facets, are found in many alpine karrenfields recently (tóth, 2007). on the one hand, it can like this karren-wal at 1,750 metres a.s.l. in Bödmeren be expected that association of karren features (Schwyz, Switzerland). reflects the many factors that interact in the par- 18 KRF•1 • OK.indd 18 15.12.2009 10:41:50 Angel Ginés, Karrenfield landscapes and karren landforms ticular shaping of different homogeneous rock assemblages along climatic gradients and/or be- outcrops subjected to the agents that cause kar- tween different lithologies is therefore justified, ren development. on the other hand, increasing taking into account the complex frame of factors knowledge on karren assemblages could provide determining the array of forms able to result over- information on the genetical processes involved printed on the rock. in the formation of each type of elementary kar- ren features, particularly on the base of their frequency of appearance as well as on their cor- Elementary karren features relation (positive or negative) with other elemen- tary karren features (ginés, 1996a). For instance, Karren is a complex group of small to medium- there can be found in mountain or cold climate sized karstic landforms showing a great variety of areas typical karren assemblages related to snow characteristic shapes. some of them can be con- melting, that are characterized by trichterkarren sidered as elementary karren features, since they (Figure 8) and sharpened edges (Figure 9) as diag- seem associated to definite genetic factors and nostic features (perna and sauro, 1978). reshap- they become frequently integrated in wider-scale ing of subsoil karren after deforestation and rock karren assemblages. The bewildering diversity of exposure to direct rainfall also produces a definite karren is difficult to summarize. a brief descrip- assemblage of forms showing vertical gradation tion of the most documented elementary karren downwards, from the top of the rising rocks to features is stated below, and is presented as a sim- the soil surface. plified diagram in table 1: table 1 identifies as follows the main solutional Biokarstic borings: nanokarren features pro- agents responsible for karren growth: biokarstic duced or promoted by microorganisms (bac- corrosion, wetting by dew, tiny water films drawn teria, cianobacteria, algae and fungi), lichens by capillary tension, direct rainfall, channelled or roots. pits, boreholes, trenches and tunnels, water flow, sheet wash water flow, short but in- generally smaller than 1 mm in size are their tense storm showers, still-stand water, snow melt- most common appearances through micro- ing, glacier ice melting, direct infiltration and scopic examination with seM. slow soil water percolation. it appears that many Irregular etching: nanokarren and microkarren of them are strongly controlled by environmen- features constituted by a wide variety of pitting tal factors, mainly climate-dependent conditions. and differential, non-oriented, etching forms, This explains most of the hydrological integration, commonly showing protrusions and hollows from the upper part of the rock outcrops down- lesser than 1 cm. rock surfaces affected by wards, shown by several types of karren assem- these rough shapes show no clear directional blages (e.g. from the splashing of raindrops to the trends and are characterized by a chaotic and collection of water into channelled water flow). coarse microtopography. even, on the contrary, the apparent lack of hydro- Microril s (= Ril ensteine): Microkarren features logical connection between karren features that characterized by rock surfaces showing sever- is exhibited by some karren assemblages related al different patterns formed by tiny channels to arid or semiarid climates indicates juxtaposi- and/or micro-spikes, rarely surpassing 1 mm tion of etching elements rather than efficient water in width. They have been typically described drainage. coastal influence, wind orientation, sun as about 1 mm wide rills, round bottomed exposure, frost action, dampness, frequent dew and packed together with characteristic tight- formation, etc. could favour or hinder the devel- ly sinuous to anastomosing plan view patterns opment of specific karren assemblages. an envi- on gentle slopes, becoming more parallel and ronmental and statistical approach to the karren straighter with increasing slope. 19 KRF•1 • OK.indd 19 15.12.2009 10:41:50 Karst Rock Features • Karren Sculpturing Rainpits (= solution pits): small, hollowed cup- from a diffuse or linear source (e.g. a bedding like karren features, subcircular in plan and plane) situated upwards. generally their cross- nearly parabolic or tapering in cross section, sections are largest close to the input of water whose diameter ranges from 1 cm to 5 cm, and and diminish downslope. exceptionally exceeding 2 cm in depth. Fre- Meandering runnels (= Mäanderkarren): small quently appear clustered in groups and can co- winding channels that are cut directly into the alesce to give irregular and carious appearance rock surface or within a larger runnel. This spe- to the rock surfaces. cial kind of karren channels exhibits meander Ril enkarren (= solution flutes): small, straight, forms with typical undercutting and slip-off narrow, closely packed, parallel solutional fur- slopes. There is frequent overlapping between rows, that head at the crest of bare rock slopes meandering karren and some types of decanta- and extinguish downslope. Their dimensions tion runnels. in limestone outcrops are typically 1.2–2.5 Kamenitzas (= solution pans): Dish-shaped de- cm in width, 2–6 mm in depth and 10–30 cm pressions, 1 cm to 0.5 m deep, 5 cm to 5 m wide in length. individual flutes are parabolic in and mostly elliptical or circular to highly irreg- cross-section and are separated by sharply pro- ular in plan. usually they have flat and nearly nounced cusp lines. in plan view, they may horizontal bottoms that are floored by a thin form a simple suite of parallel flutes showing layer of soil, vegetation or algal remains which remarkable regularity of form and dimension. decay enhances further dissolution. Their bor- Their development to either side of a crest often ders are frequently overhanging and may have produces a typical herringbone pattern. small overspill outlets. Solution runnels (= Rinnenkarren): Mesokar- Solution bevels (= Ausgleichsflächen): Flat, smooth ren features consisting of linear channels or surfaces, 0.2 to 1 m long, usual y found as plane furrows that generally show increased width sub-horizontal belts developed below the level of and depth downslope. Threads of runoff water, ril enkarren-flutes extinction. pouring down the flanks of the rocks, are col- Trittkarren (= heelsteps): conspicuous karren fea- lected into channels to create solution runnels tures that form arcuates headwalls, which flat which width and depth range from 5 to 50 cm, floors are open in downslope direction. The sin- being very variable in length (commonly from 1 gle trittkarren consists of a flat tread-like sur- to 10 m, but in some cases exceeding 30 m long). face, 10 to 40 cm in diameter, and a sharp back- owing to the great diversity of topographic slope or riser, 3 to 30 cm in height. Their typical conditions and the kind of water supply feed- appearance is as groups of heel prints excavated ing their channelized flow, they have a remark- as steps on the rock outcrops. They seem to be able variety in cross section and plan pattern the result of complex solutional processes in- (including tributaries). volving both horizontal and headward corro- Wal karren (= Wandkarren): straight furrows sion generated by the thinning of water sheets that are often found on sub-vertical outcrops as flowing upon small slope falls. a result of water flowing down rock walls. They Cockling patterns: solutional concave shapes are in many cases real macrokarren features undulating the rock surfaces randomly or, in because they can attain lengths close to 100 me- some cases, producing a regularly crinkled pat- tres. There is frequent overlapping between wall tern and a characteristic horizontally-lined ap- karren and some types of decantation runnels. pearance to the rock slopes. Decantation runnels: channels generated by Solution ripples: Wave-like forms, transverse water released steadily, that start from an ups- to downward water movement under gravi- lope point-located store (e.g. a patch of moss) or ty. Their rhythmic forms suggest that periodic 20 KRF•1 • OK.indd 20 15.12.2009 10:41:50 Angel Ginés, Karrenfield landscapes and karren landforms pulses of flow or chemical changes are impor- Subsoil tubes (= subcutaneous karren pipes): per- tant in their development. forating tubes and rock holes, formerly filled Trichterkarren (= funnel karren): concave kar- with soil, which diameter range from a few cen- ren forms that resemble trittkarren in shape timetres to less than 1 m. They show rounded and dimensions, differing from such karren cross sections as well as branching and complex type in their much rounded cross section which tridimensional patterns that penetrate deep- lacks the characteristic horseshoe appearance ly into the rock. subsoil tubes are a significant and the stepped pattern. There is frequent over- part of the epikarst in many well developed lapping between genuine trichterkarren and karrenfields. trittkarren, but the former seem to be exclu- Subsoil hol ows (= subcutaneous hol ow-forms): sively associated with snow melting. subsoil hollows showing different shapes and Sharpened edges (= lame dentate): Mesokarren sizes are frequently exposed in road cuts and features delimiting a sharp and straight edge quarry walls. They are also common at the on the flanks of steep rocks. Their upper part foot of karren pinnacles. in addition to pock- shows a characteristic and almost triangular ets, niches and recesses, great subsoil wells and flat facet. They seem to be exclusively associated small subsoil pitting and scallop-like features with snow melting. are frequently found. Grikes (= Kluftkarren): Deep clefts, from 1 cm Cutters: great solution crevices similar to grikes to 0.5 m across and up to several metres deep. but generally larger than 1 m in wide and filled They are one of the most typical mesokarren with regolith or soil. They appear much wid- features, normally from 1 m to 10 m in length, ened at the top and tapering to narrow cracks formed through the simple solutional enlarge- with depth and are generated by vertical solu- ment of joints or cracks. Their linear trends tion along fractures beneath a thick soil. are determined by major structural directions Undercut runnels (= Hohlkarren): Mesokarren as joint sets or faulting. owing to the fact that furrows transformed by organic debris or par- such slots cut in the bedrock are merely the tial soil filling because their side walls have visible surface expression of the fissures criss- been hollowed under by enhanced biogenic crossing the karstifiable rocks, grikes constitute carbon dioxide concentration. typical bag-like a significant component of the epikarst. cross sections, wider at the bottom than at the Rundkarren (= rounded runnels): channels or top, are generated in this way. furrows developed beneath a soil cover, which Clints (= Flachkarren): tabular intervening troughs and ribs become smoothed by the blocks or slabs isolated by grikes. They are flat more active corrosion associated with soil wa- or gently inclined outcropping rocks which ters that produces their characteristic rounded become divided into straight-sided blocks by cross-sections. the solutional widening of fissures. These bare Subsoil smooth surfaces (= subcutaneous karren plane surfaces of limestone, generally parallel forms): a whole array of characteristic rounded to the bedding, are the main constituent of and smooth rock surfaces appear clearly relat- limestone pavements. ed to subsoil corrosion. smoothing and bleach- Pinnacles (= Spitzkarren): Vertical solution along ing of subsoil karren-forms is evident when it joints and fractures lowers the intervening rock is compared with the sharpening and grey ap- flanks and produces isolated spires or pinna- pearance of exhumed karren forms. subsoil cles that can reach a few metres or tens of me- smoothing can be considered a consequence of tres in height. usually the side walls are deep specific nanokarren features developed in con- grikes with runnels cutting across one anoth- tact with the sponge action of acidic soil water. er to form sharp ridges and peaks. They could 21 KRF•1 • OK.indd 21 15.12.2009 10:41:50 Karst Rock Features • Karren Sculpturing Figure 10: Winding rib remnants that are exclu- sively found between some rillenkarren flutes developed on salt rock (Cardona, Catalonia, Spain). Width of view is 12 cm. be considered as a particularly mature form of A broad discussion about so complex terminol- karren and in many cases are the result from ogy would be tedious in this merely introductory sharpening of subsoil pinnacles after being ex- chapter. During the seventies, the wide use of the humed by soil removal. terminology adapted from Bögli (1951, 1960a) This is necessarily a simplified overview on into the international karst literature proved to the elementary karren features that are enclosed in Table 1 (based in Ginés, 2004), and corre- sponds to some extent with the so called single forms from Bögli (1980). It stands out that some of these elementary karren features accumulate a very uneven amount of bibliographical references (for instance, many references about rillenkarren in comparison with very few on trichterkarren). Furthermore, some of them can be considered as the end members from a continuum of related forms (this is the reason for the choice of the com- plementary term runnel, which embraces a wide array of forms, including the formerly spelled Rinnenkarren, Wandkarren, Mäanderkarren, Rundkarren and Hohlkarren as well as many Figure 11: Eogenetic karren formed on Miocene reef decantation forms). Finally, there are elementary rocks of south-eastern Mallorca (Balearic islands, Spain). karren features easy to delimit (as, for example, Recent soil removal allows us to recognize the charac- the rainpits), but many others show clear overlap- teristic differences existing between the upper rough and grey formerly exposed rock and the smooth and ping (this seems to be the case for the transitional whitish appearance of freshly exhumed subsoil sculp- forms between trittkarren and trichterkarren). turing. Width of view is 1.5 m. 22 KRF•1 • OK.indd 22 15.12.2009 10:51:22 Angel Ginés, Karrenfield landscapes and karren landforms Figure 12: Great subsoil- shaped pinnacles arise after iron mining in the palaeokarst of Cerro del Hierro (Seville province, Andalusia, Spain). be a successful decision. But today, in my opinion, carried out by petterson (2001) and Dreybrodt and such tendency has outlived its usefulness. For ex- Kaufmann (2007) on the solution by flowing water ample, and from a strictly personal point of view, films on sloped blocks. experimental work in the let me suggest avoiding the future expansion of laboratory, using plaster of paris (artificial gyp- germanized new terms as recently occurred with sum) block-models, were previously initiated with the excellent paper from simms (2002) on lacus- success by glew and Ford (1980) for rillenkarren, trine röhrenkarren. and subsequently were developed by Dzulynski et al. (1988) and slabe (2005), trying to embrace also the kluftkarren, solution runnels and subcutane- Looking forward to karren research ous stone-teeth formation. Furthermore simula- strategies tion of rillenkarren flutes by slabe (2005) has been especially successful in obtaining by means of among the major issues faced by the researchers plaster models the expected downward sequence on karren landforms in the next years are: the use on the typical karren assemblage consisting of ril- of modelling techniques for a better explanation lenkarren, ausgleichsfläche and solution runnels. of the genesis of specific elementary karren fea- There is an increasing literature on morphom- tures; the accurate morphometrical description etry of elementary karren features though not all of the most significant karren landforms; and the the measurements of karren types are substanti- application of statistical methods in order to com- ated by large sampling nor are they easily com- pare karren locations differing in lithology and/or parable with another karren locations. Valuable climate conditions. field data on rillenkarren are gathered in Belloni The formation of the different types of karren and orombelli (1970), lundberg (1977a, b), Hein- features can be largely understood in terms of the emann et al. (1977), Dunkerley (1979), goudie et mobility of solutes along the rock surfaces. This al. (1989), stenson and Ford (1993), Mottershead approach constitutes the base for the experiments (1996a) and ginés (1996b), while measurements 23 KRF•1 • OK.indd 23 15.12.2009 10:41:55 Karst Rock Features • Karren Sculpturing of clints and grikes are documented in goldie and precipitation and temperature. correlation with cox (2000). The references on morphometry of macroflora is not quite congruent because karren rainpits are limited to lundberg (1977a) and ginés is obviously defined as predominantly bare rock- (1998b), and only scant quantitative data are re- surfaces and landscapes. But correlation with ported from meandering runnels in Zeller (1967), weathering and soil-forming processes could be Hutchinson (1996) and Veress and tóth (2004). useful. topography would be significant, espe- Morphometric characteristics of trittkarren-trich- cially regarding soil removal and exposure to sun terkarren features are discussed by Vincent (1983a) radiation. Furthermore, it can be expected that on the base of six morphological variables. Belloni lithology influences the nature and rate of kar- and orombelli (1970) provide also measurements ren growth in two major ways: mineralogy and of kamenitzas and solution runnels. in this way, texture. correlations between forms and environ- an important recent trend in exokarst studies lies mental factors along gradients as well as multifac- in the greater application of morphometric tech- tor analysis including rock properties (Figure 10) niques to karren descriptions. Further research have presumably good application to karren re- would be encouraged as long as available meas- search owing to such complex causation. statisti- urements of elementary karren forms on different cal methods (e.g. principal component analysis) karstic locations continue to grow. a combined and integrated geomorphological and ecologi- and cautious use of morphometry and terminolo- cal studies (e.g. transects) are being introduced gy is recommended in order to clarify and simplify gradually with this purpose, as suggest the con- the overwhelming amount of karren terms. tributions from goudie et al. (1989), ginés (1996a, it has long been recognized that several fac- 1999a) and Veress et al. (2001a). Finally, time is tors may control karren development: lithology, also important in relation to karren because these topography, climate, biological activity and time. landforms are subjected to evolution at different The effects of climatic differences on karren devel- chronological scales, ranging from decades or opment tend to be apparent at a global overlook, centuries of deforestation and subsequent soil re- both at the scale of karrenfields and at the scale moval (Figure 11) to very old burying under sedi- of individual elementary karren features. Most ments for millions of years until produce in some of the diversity in karren landforms can be easily cases palaeokarst ore-deposits (Figure 12). explained as a direct consequence of variation in 24 KRF•1 • OK.indd 24 15.12.2009 10:41:56 pHYsics anD cHeMistrY oF Dissolution 2 on suBaerialY eXposeD soluBle rocKs BY FloWing Water FilMs Wolfgang DREYBRODT and Georg KAUFMANN Karren is the generic term for dissolution features The fluid dynamics of water films on on exposed soluble rock surfaces. Because of their smooth and rough surfaces variety of shapes and also their regularity karren have been a fascinating object of interest for ge- When rain with intensity q (cms-1, 1 mm/hour = omorphologists. although a large body of obser- 2.8 ·10-5 cm s-1) falls onto an inclined smooth sur- vations and descriptions of karren has been accu- face with slope angle γ, a thin layer of water is es- mulated, knowledge on the physical and chemi- tablished (see Figure 1). its flow rate Q in cm3/s per cal processes on their formation by dissolution is unit width is given in cm2/s. after distance x´ = l scarce. in this paper we will focus on processes down the surface of the rock Q is occurring on bare rock surfaces such as limestone = ⋅ = γ (1) or gypsum exposed to the atmosphere, and cov- Q x q l q cos , ered by flowing water films. two basic ingredients The thickness h (in cm) of the water film is re- control the dissolution process, the hydrodynam- lated to flow Q (Myers, 2002) by ics of thin water films flowing down inclined sur- 3 ρ gh Q = sinγ , (2) faces, and the dissolution kinetics of the co -con- 3η 2 taining rainwater on limestone or gypsum rock surfaces. after discussion of these two topics, we will use this for an interpretation of the data of glew and Ford (1980), who performed experimen- tal simulations on the formation of rillenkarren on inclined surfaces of plaster of paris exposed to artificial rainfall. using these results an interpretation of existing field data on lengths of rillenkarren is presented. also recent data by petterson (2001) on dissolu- tion on rillenkarren from plaster of paris will be discussed. Finally the dissolution kinetics of lime- stone will be used to explain surface denudation on bare limestone surfaces. Figure 1: Water film on inclined rock surface. 25 KRF•1 • OK.indd 25 15.12.2009 10:41:56 Karst Rock Features • Karren Sculpturing where g is earth’s gravitational acceleration, ρ is Dissolution kinetics the density of water, and η its viscosity. By using eqns. 1 and 2, we obtain the film thickness Gypsum 3η q (3) h =  3 . ρ g tanγ By use of rotating disc experiments Jeschke et al. (2001) have found that the surface reaction rates of For rainfall intensities of 1 mm/hour onto a gypsum (in mmol cm-2 s-1) are given by surface sloping with 45° and at a distance l = 50 R = k − c c = α c − c (8) s s (1 s / eq ) s ( eq s ) cm, a fairly thin film of h = 3.6·10-3 cm develops. For 40 mm/hour rainfall intensity as used by glew with and Ford (1980), the film thickness h is 1.2·10-2 cm. -3 α = k c s s / eq =6.5·10 cm/s. The flow velocity u (in cm s-1) is obtained from u·h = Q by inserting eqns. 2 and 3 and one finds Here, c is the calcium-concentration at the s 2 2 2 2 ρ gQ sinγ ρ gq l cos γ sinγ u = (4) surface and the rate constant is k = 1.1·10-4 mmol 3 = 3 . s 3η 3η cm-2s-1. The equilibrium concentration c with re- eq note that the velocity increases with flow dis- spect to gypsum is 15.4·10-3 mmol cm-3. ca2+ and tance l. assuming a rainfall of 10 mm/hour, a flow so2--ions released from the mineral surface are 4 distance of 1 m, and a slope angle of 45°, the veloc- transported away from the surface into the solu- ity is 2.1 cm s-1. if rainfall is reduced to 1 mm/hour tion by molecular diffusion. Therefore concentra- one finds 0.5 cm s-1. These velocities are of impor- tion gradients exist and the surface concentration tance because they give the time of residence dur- c differs from the concentration c in the bulk. The s ing which a water parcel can dissolve bedrock. transport rate R by molecular diffusion is given D When the surface is rough a correction factor by must be introduced (Myers, 2002), which is given R = k − c c = α c − c (9) D D (1 / eq) D ( eq ) by 1.28   k  (5) with f  =  − c 1 0.25   ,   h    α = k c D D / eq where k is the roughness of the surface and h the where k is the transport constant and c is the D film thickness of the layer on a smooth surface, as average concentration of the bulk solution. since given by eqn. 3 (phelps, 1975). This dimensionless due to mass conservation R must be equal to R , S D factor relates the flow velocities u and u of the one finds an effective rate law (Dreybrodt, 1988). r smooth to the rough surfaces respectively: R = k − c c (10) eff (1 / eq) u = f ⋅ u (6) r c . with Because u·h = Q the film thickness values are k ⋅ k related by s D k = eff , k + k 1 s D h = h (7) r . f or c R = α c − c eff ( eq ) For k/h = 2, a reasonable number, we obtain f c ≈ 0.4, and flow velocities are lower. Film thickness with values are higher by a factor of 2.5. α ⋅α s D α = eff . α +α s D 26 KRF•1 • OK.indd 26 15.12.2009 10:41:56 Wolfgang Dreybrodt and Georg Kaufmann, Physics and chemistry of dissolution on subaerialy exposed soluble rocks by flowing water films When k >> k , k becomes close to k and rates s D eff D are controlled by diffusion. on the other hand if k << k , k becomes close to k and the rates are s D eff s surface controlled. in the region where k and k s D are of similar magnitudes both processes control dissolution. For a laminar water film of thickness h, k is D given by (Beek and Muttzall, 1975) k = Dc h (11) D 2 eq / or α = D h , D 2 / where D is the coefficient of diffusion (1·10-5 cm-2 s-1). For h = 0.01 cm one obtains α = 2·10-3 cm s-1 Figure 2: Dissolution rates of limestone by CO -contain- D 2 and the rates are controlled by diffusion. However, ing water. Three regimes of very fast (region 1), moder- raindrops impinging on the water film may cause ate (region 2), and inhibited dissolution rates (region 3) are clearly distinguishable. Only the fast dissolution rate mixing, which could increase the effective diffu- in region 1 is relevant in this paper. sion constant. only a factor of 10 suffices to obtain surface control and a value of α ≈ 7·10-3 cm s-1. eff to convert the rates from mmol cm-2 s-1 into retreat of rock in cm/year for gypsum one has to R = α c − c I ( app ) 1 (12) multiply by a factor of 2.3·106. for Limestone c ≤ 0.3c . eq Water films running down rock surfaces under For higher calcium concentrations a second lin- natural rainfall conditions have a comparative- ear region with significantly lower slope α arises, 2 ly small depth of a few tenths of a millimetre. in until close to equilibrium in region 3 for c ≥ c , sw contrast to gypsum, where dissolution rates are above the switch concentration c = 0.9 c inhibi- sw eq determined by surface reaction and molecular tion occurs and the rates are controlled by slow diffusion, the situation on limestone is more com- surface reactions. plex. Figure 2 schematically depicts three regimes The dissolution rates in regions 2 and 3 are well of dissolution rates. For highly undersaturated understood (plummer et al., 1978; Buhmann and solutions, 0 < c ≤ c , rates are high and decline Dreybrodt, 1985; svensson and Dreybrodt, 1992). app steeply with slope α to an apparent equilibrium Three basic chemical reactions control the dis- 1 concentration c = 0.3· c , where c is the true solution of caco : app eq eq 3 equilibrium concentration with respect to calcite. 1. + ← 2 H + CaCO + → Ca + HCO− 3 3 The values of α are almost independent on the 1 film thickness h for 0.005 cm < h < 0.03 cm, and α 2. ← 2 H CO + CaCO + → Ca + 2 HCO− 1 2 3 3 3 = 5·10-4 cm s-1 (Kaufmann and Dreybrodt, 2007). to a good approximation the rates found by 3. ← 2+ 2− ← 2 CaCO + H O + + + + → Ca CO H O→ Ca + HCO− OH − 3 2 3 2 3 theoretical modelling can be expressed by For all three reactions co dissolved in the solu- 2 27 KRF•1 • OK.indd 27 15.12.2009 10:41:56 Karst Rock Features • Karren Sculpturing tion must be hydrated into carbonic acid, which be equal with Q dc, where dc is the increase in total rapidly reacts to H + HCO− + 3 . concentration from x to x + dx, and Q is the total total flow rate in cm3 s-1. From this a differential 4. H O + CO ←→ H CO 2 2 2 3 equation is found dc α ⋅ W (15) 5. CO + OH − ← 1 = ( c − c). → HCO− 2 3 dx Q app total The pH-values of the solution in region 2 are be- its solution is tween 7.5 and 8.3. For such pH-values conversion   α  (16) of co is slow (usdowski 1982, Dreybrodt 1988) 1 c( x) = c  − −  x app 1 exp  , 2   Q and for thin films below 0.02 cm control by co -  2 conversion limits the rates. For film thickness where Q = Q / W is the amount of flow in one cm total between 0.01 cm up to 0.04 cm slope values are width of the film. about α = 3·10-5 cm s-1, lower by about one order We use eqn. 16 to determine α experimentally. 2 1 of magnitude than α = 5·10-4 cm s-1. to this end, we have constructed a channel of 5 1 The reason for the high rates in region 1 are re- cm width and 1.2 m length by employing acryl actions (1) and (3). When no calcite has yet been rims fixed to a plate of limestone. The inclination dissolved the initial pH of the solution in equi- is γ = 3.2°. at the end of the channel a funnel of librium with co in the atmosphere is 5.7. since acryl-glass channels the water into a hole from 2 reaction (1) is very fast protons are rapidly con- where it runs into a bottle. The experiment is il- sumed by dissolving calcite. lustrated in Figure 3 (at the top), which provides a Furthermore dissolution of calcite produces view from above. to guide the water into a stable oH- ions. Therefore pH increases to values of film the channel at its upper end is blocked by a about 11. Because of the high concentration of piece of acryl-glass, which leaves a narrow space oH-, conversion of co is fast by reaction 5. With of a few tenths of a millimetre between the lime- 2 increasing ca-concentration pH drops, and con- stone surface and its lower plane face (see Figure sequently slow conversion of co by reaction (4) 3, below). Distilled water in equilibrium with the 2 takes over in controlling the rates. as a conclusion p in the atmosphere by use of a peristaltic pump CO we state that for low concentrations c the rates are 2 is introduced into the upper compartment, and a given by the relation film of constant thickness moves down in laminar α (0.3 c − c < c < c (13, 14) flow at ambient temperature of 20°c. This film is eq ); 0 0.3 1 eq R = . α ( c − c c > < established by drawing down the water along the  c  eq ); 0.36 0.9 2 eq limestone surface by use of a wet paper strip as wide as the film is desired to be. The water film does not touch the acryl walls but is kept by sur- Experimental determination of face tension. it does not change its shape, even dissolution rates in region 1 when its depth varies by a factor of three. The sur- face of the film is absolutely plain as can be seen by When a thin water layer of width W flows down a a mirror like reflection of light. The flow rate Q is smooth, plane limestone surface with inclination measured by collecting 10 ml of water at the out- angle γ it dissolves calcite and the concentration let hole at the end of the channel, and measuring c(x) of calcium along its flow path increases. The the time needed. The calcium concentration c of end amount of calcite dissolved during one second this sample is then measured for various values of between positions x and x dx is given by α ( c Q. Furthermore, water in equilibrium with atmo- 1 app – c( x))· dx·W. Due to mass conservation this must spheric p and calcite is used to measure c . The CO2 eq 28 KRF•1 • OK.indd 28 15.12.2009 10:41:56 Wolfgang Dreybrodt and Georg Kaufmann, Physics and chemistry of dissolution on subaerialy exposed soluble rocks by flowing water films Figure 4: Calcium concentration versus inverse of flow rate for experimental data (squares). Q is the total total flow rate of the film. The straight line is a least square fit to the data. Figure 3: Experimental set up to measure limestone dis- rate downstream increases (see Figure 1). if at x’= solution rates in region 1 (top and side view). Length l of channel 120 cm, width of channel 5 cm, average 0 the flow rate is Q ; then at a later position x’ it is 0 width W of water film 4 cm. given by Q = Q q x’ cosγ = Q q’x’. (18) 0 0 Mass conservation demands that calcium concentrations are determined by mea- W ( Q + q' x') c + W α ⋅( c − c dx = (19) eq ) ' suring electrical conductivity, which for such low 0 concentrations is linear with calcium concentra- = W ( Q + q' x'+ q' dx')( c + dc), 0 tion. The experiment was performed at 25°c. eqn. 16 can be rewritten to where c is the average concentration at position x’,  c  α ⋅ l ⋅ W (17) and W is the width of the film. ᾶ ≈ α· f , where f end 1 −ln1−  = . α α  c  is a correction factor considering the roughness of  app Q  total the rock surface. if one assumes that the rock sur- Figure 4 shows the plot of the experimental face consists of small half spheres densely packed, data in terms of ln(1- c / c ) versus 1/ Q . This instead of a smooth plane, the surface area avail- end app total can be fitted with a straight line by using c = 0.3 able for dissolution will increase by f = 2. This app α c = 0.17 mmol/cm3. From the slope 0.129 of the gives an estimation on the order of magnitude of eq line one finds α = 2.6·10-4 cm s-1, which is in rea- f . equation 19 states that the outflow of calcium 1 α sonable agreement to the theoretical predictions at position x’ dx’ is given by the inflow at posi- of α th = 5·10-4 cm s-1 and c th = 0.36· c . tion x’ plus the amount of calcium ions dissolved 1 app eq per time between x’ and x’ dx’. neglecting terms with dx’· dc one finds a differential equation Dissolution on bare rock surfaces dx' dc (20) = , Q + q' x'  α( c − c − q c eq ) ' When rain falls onto an inclined surface the flow 0 29 KRF•1 • OK.indd 29 15.12.2009 10:41:56 Karst Rock Features • Karren Sculpturing with solution q'  α c   (21) ρ g (24) eq q' x' 3 − q' c( x') = 1  − (1+ ) + α . l = h tanγ . q'+ α  Q 3 c η q  0  Therefore, by plotting l versus tan γ one should For large values of x’ the concentration ap- find a straight line. This indeed is the case for the proaches the value glew and Ford᾽s (1980) data, as shown by Figure α c =  (22) 5. The slope of this line is 14 cm, from which one ∞ c . q' eq +  α finds a critical thickness h = 7.7·10-3 cm if one as- c sumes a smooth surface. For a rough surface with 90% of this value is reached at a distance k = h one finds a value of 10-2 cm. glew and Ford c q'   (23) measured a value below (1.5 ± 0.5)·10-2 cm, which ' Q 0 q' x = 10 +   α −1. is in good agreement. They also measured disso- 0.9 q'   lution rates of 4·10-3 cm/h. For their experimental data one finds c ≈ 0.66 c from eqn. 17. ∞ eq For Q = 0 the concentration c is established The amount of flow leaving a rock of width W at 0 ∞ immediately. Therefore dissolution rates are uni- x’ is equal to the amount of rainfall which falls to form downstream if one assumes that ᾶ is inde- the area W·x. it carries away the mass of rock q· c ∞ pendent of the thickness of the water sheet. This is · x which is dissolved from the rock’s surface area not true for gypsum. a reasonable approximation Wx’ = Wx/cos γ. converting the mass of dissolved is to use average values. For gypsum α is maximal material to its volume one finds the retreat of rock 7.1·10-3cm s-1 if the rates are controlled by surface R = c ⋅ ⋅ γ ρ (25) ∞ q D cos / g , reactions and at a sheet thickness of 0.1 mm it is 1.56·10-3cm s-1 (see eqn. 10). at a sheet thickness of where ρ = 2.3 g/cm3 is the density of gypsum. g 0.5 mm one finds α = 3.8·10-4 cm s-1. With the experimental conditions of glew and Ford one finds R = 2.7·10-3·cos γ (cm/h). This fits D reasonably well into their data set. However, it rep- Rillenkarren Experiments on formation of rillenkarren on gypsum glew and Ford (1980) experimentally simulated the formation of rillenkarren on gypsum by ex- posing inclined surfaces of plaster of paris to a rainfall intensity of 38 mm/hour, which lasted for 500 h. They obtained well developed rillenkarren. Their average length from the crest to the “aus- gleichsfläche” was dependent on the angle of in- clination, as shown in Figure 5. Ford and glew ar- gued that the “ausgleichsfläche” could form only when the water film exceeds a critical thickness h , c Figure 5: Length of experimental rillenkarren versus which should be higher than the roughness k of slope, tan γ. The squares are experimental data from the rock. With this assumption by use of eqns. 1 Glew and Ford (1980). The line represents eqn. 24 with and 2 one finds h =7.7.10-3 cm. c 30 KRF•1 • OK.indd 30 15.12.2009 10:41:56 Wolfgang Dreybrodt and Georg Kaufmann, Physics and chemistry of dissolution on subaerialy exposed soluble rocks by flowing water films   ­€  ‚ ƒ„€   Figure 6: Length of natural rillenkarren on limestone versus slope tan γ. From J. Lundberg and A. Ginés, personal communication. The straight lines are fits to l = A . tan γ. resents a lower limit because one assumes laminar tilted by 30°. Water samples collected from the flow. splashing raindrops may disturb this flow karren at various distances from the crest were and cause mixing of the solution by which the ef- used to measure the calcium concentration pro- fective diffusion constants increase. a factor of 10 file along the karren. petterson found an almost is sufficient to rise c to 0.9 c . linear increase from 75 mg/l of calcium at 5 cm ∞ eq in a recent work petterson (2001) has exposed to a value of 105 mg/l at 40 cm. The average value rillenkarren channels modelled from real lime- was 90 mg/l ± 15 mg/l. stone rillenkarren by plaster of paris, to artificial With an average film thickness of 0.5 mm one rain of 115 mm/hour intensity. By using an optical finds ᾶ = 7.6·10-4 cm s-1. With a rainfall intensity technique he measured the thickness of the lami- of 115 mm/h = 3.2·10-3 cm s-1 by use of eqn. 22 one nar flowing water films along the karren rills. The obtains a value c = 118 mg/l. in view of the ap- ∞ thickness of these films, measured at a distance of proximations this can be regarded as good agree- 5 cm to 40 cm from the upper edge, range from ment to experiment and proves our theoretical 0.2 mm up to 0.8 mm, when the karren model was considerations. 31 KRF•1 • OK.indd 31 15.12.2009 10:41:57 Karst Rock Features • Karren Sculpturing Interpretation of field data of rillenkarren 35 a large body of data has been collected, which re- 30 lates the lengths of rillenkarren to the slope of the 25 rock surface where they grow. From eqn. 24 one 20 expects a linear relation of length and slope. 15 l = A⋅ tan γ length L (cm) (26) 10 with L = 12.6 + 0.5 T ρ 5 g 3 A = h 3 c η q . 0 -10 -5 0 5 10 15 20 25 30 mean annual temperature T °C Figure 6 shows average lengths of rillenkarren Figure 7: Length of natural rillenkarren on limestone ver- versus slope (tan γ) for several areas (ginés and sus mean annual temperature. From J. Lundberg and A. lundberg, personal communication, 2006). The Ginés, personal communication. straight line represents a least square fit by the re- lation l = const·tan γ to the data points with γ ≤ 64° (tan γ ≤ 2). rillenkarren in serra de tramuntana as a function although the scatter of points, which could be of altitude above sea level, taken from lundberg caused by differing values of precipitation q at dif- and ginés (personal communication, 2006). ferent sites and times is significant one finds a = There is a clear decrease of length with altitude 12.6 ± 3 cm for all plots. From this by use of eqn. h, which can be caused by two reasons. First there (26) one obtains h3/ q = (3.9 ± 0.2) ·10-4 cm2s. is a linear relation between altitude and tempera- c Figure 7 shows the relationship of length with ture. The up most abszissa shows the correspond- mean annual temperature as reported by lund- ing temperature given by berg and ginés (personal communication, 2006). The data can be fitted by a relation l = 0.5 T + 12.6 T = 17 – 0.0065 h (°c), (29) (cm), where T is in °c. The variation of l in tem- perature could result from the temperature de- where the altitude h is in m. pendence of η which can be presented with an ac- Furthermore mean annual precipitation q is av curacy within 2% by the empirical relation related to altitude h by   (27) 1/η = 53.8 + 2.76 cm s T   ,  g  q = 461 + 0.4 h [mm/year]. (30) av that is valid between 0°c and 25°c. see upper abscissa in Figure 8. introducing this into eqn. 26 one finds using We now assume that the actual rainfall to the h3/ q = 6.11 ·10-4 cm2s and tan γ = 1 rock is related to q by q = f ·q , where f is a con- c av q av q stant. l =12 + 0.52⋅ T [ cm] . (28) Both q and viscosity η depend on altitude. using eqns. 26, 27, 29 and 30 one can calculate the The value of h3/ q is close to that found from the length as a function of altitude. With h3/ q as a fit-c c dependence of length on slope in the previous ex- ting parameter one obtains the curve in Figure 8. ample. The curve underestimates the large lengths, but as a final example we discuss the data present- shows the general trend. Whether it is a reason- ed in Figure 8 which relates the average lengths of able estimation must be judged from the value 32 KRF•1 • OK.indd 32 15.12.2009 10:41:57 Wolfgang Dreybrodt and Georg Kaufmann, Physics and chemistry of dissolution on subaerialy exposed soluble rocks by flowing water films of h3/ q( h). if one assumes that 1000 mm/year c correspond to an average actual precipitation of ­ 10 mm/hour one obtains h = 0.005 cm and cor- c respondingly h ­ € 3/ q( h) = 6.7·10-4 cm2s. This value is c also close to those found in the previous examples. assuming an average actual precipitation of 10 mm/h dominant in the formation of karren one   finds h = 0.0059 cm from the length-slope rela- c tion and h = 0.0065 cm from the length-temper- c ature relation. in all three examples we have assumed an av- erage precipitation of about 10 mm/hour dur- ing the formation of rillenkarren. This is a value, which seems possible. For higher precipitation the length would be smaller and would be overprint- Figure 8: Length of natural rillenkarren on limestone ed by lower precipitation yielding longer karren. (Mallorca) versus altitude above sea level. From J. Lund- at low precipitation rates (1mm/hour) the karren berg and A. Ginés, personal communication. The curve become very long (2 m) and will form very slowly, represents the fit discussed in the text. such that they may not be detected. in summary, glew’s and Ford’s idea that karren tion running off the rock has a concentration of length is determined by a critical thickness h of 0.5 c . at lower rainfall intensities of 4 mm/hour c eq the down flowing water film can be used to ex- one finds c = 0.9 c . Therefore it is reasonable to eq plain field data. one should keep in mind that at a take an average value c = 0.75 c for all the water eq precipitation rate of 10 mm/h a film thickness of during one year’s rainfall. From this one finds a 0.006 cm is attained after 27 cm on a smooth rock denudation rate of 1 mm/year. surface inclined by 45°. We do not know at present the physical reason, why this critical thickness avoids further growth Limestone of rillenkarren. This requires experimental obser- vations of flow rates and chemical composition of For dissolution under linear kinetics with a rate the water flowing on natural karren on limestone law during rain storms of various intensities. R = α( c − c (31) eq ), the time t, which is needed until a volume ele- Denudation rates in the field ment with initial concentration zero attains con- centration of 0.63 c is given by eq Gypsum T = h /α . (32) Denudation rates on subaerial exposed gyp- For limestone with a film thickness of 0.2 mm sum samples have been reported by cucchi et one finds T = 10-2/ᾶ = 20 s to attain c = 0.64 c . in 1 1 app al. (1996). in an observation station close to tri- the slower region 2, ᾶ = 2·10-5 cm s-1 and the time 2 est (italy) with a yearly rainfall of 1,350 mm they to reach c = 0.63 c is T = 500 s. under natural eq 2 found 0.9 mm/year as an average during an obser- rainfall flow velocities are on the order of 1 cm vation time of eight years. s-1. Therefore dissolution will be effective only in at rainfall intensities of 40 mm/hour the solu- region 1. even when the water dissolved limestone 33 KRF•1 • OK.indd 33 15.12.2009 10:41:57 Karst Rock Features • Karren Sculpturing in region 2, the dissolution rates were about two ed. The specific conductivities were measured in orders of magnitude lower. in other words, all the the field. The conductivity of rain water was 6 μs/ water, which falls to the rock surface, will leave it cm, whereas the water from the karren exhibited with concentration c derived from dissolution in 57 μs/cm. analysis for calcium in the lab yielded ∞ region 1. a value of 0.25 mmol/liter, 38% of the saturation With ᾶ = 10-3 cm s-1 one finds value of 65 mmol/liter at 10°c, the temperature 1 3 10− (33) during collection of the sample. This result is in c = ⋅ ∞ c , 3 − 5 10 + 2.8⋅10− ⋅ p ⋅cos app γ good agreement to what one expects from eqn. 33. where p is the rainfall intensity in mm/h. at low slope angles (cos γ ≈ 1) and for rainfall Discussion and conclusion intensities of 1 mm/h, c = 0.97 c = 0.29 mmol/ l. ∞ app at 10 mm/h, c = 0.24 mmol/ l, and for extreme We have presented some basic principles of flow ∞ intensities of 40 mm/h c = 0.14 mmol/ l. dynamics of thin water films that can approxi- ∞ cucchi et al. (1996), by using micrometers, mate flow on natural rock surfaces under rainfall measured surface denudation rates on a huge conditions. although these approximations are number of limestone samples with slope angles of crude they can be used for realistic estimations. about 15 degrees in the karst of triest. They found to understand the formation of geomorpho- average dissolution rates sampled over eight years logic features on rock surfaces basic knowledge of 0.015 ± 0.01 mm/year. at an average rainfall of of the dissolution rates by flowing water sheets is 1,350 mm/year in this region one needs an average needed. Water in equilibrium with the p of the CO2 run-off concentration c = 0.97 c to explain this atmosphere dissolves limestone quickly up to a ∞ eq number. a closer inspection of the distribution of concentration of c ≈ 0.3 c . For higher concen- app eq rainfall-depth distribution is therefore necessary trations the dissolution rates drop rapidly. The to verify this number. anyway, our findings sup- time to reach the concentration c under natural app port that denudation on bare rock by the dissolu- rainfall conditions is on the order of 10 seconds, tional action of rainwater is caused by fast dissolu- sufficiently short, that all dissolution will be af- tion in region 1 of Figure 2. fected in this regime of concentrations. even if the We have performed a first attempt to measure solution would reach concentrations higher than concentrations of rainwater flowing from the sur- c then dissolution rates drop to such low values app, face of a karren formation of limestone from li- that they become insignificant. We have presented pica, slovenia, exhibited in front of the postojna experimental data, which confirm this behaviour. cave. after two days of heavy rainfalls, cleaning it is also possible to understand from these kinet- the rock from dust, water was collected during a ics denudation rates of limestone measured in the medium strong rainfall of a few millimeters/h by field. use of an aluminum foil attached to the rock. Fig- For gypsum dissolution rates are controlled by ure 9 shows the experimental situation. The water mixed kinetics of surface reactions and molecular had flown on top of the formation, which exhib- diffusion. Therefore, the rates become dependent its only a slight inclination of about 10° degree for on the thickness of the flowing water sheet. it is about one meter, then down one half meter, almost possible, however, to predict denudation rates on vertically, where it was channelled by the foil and gypsum, as obtained from field data. Furthermore collected into a beaker. This flowpath is depicted experimental findings on rillenkarren can be ex- by the grey line. Measures were taken to prevent plained. dilution of the sample by rainwater dripping into it should be noted that we have neglected tem- it. in parallel a sample of rainwater was collect- perature dependence and have used 20°c as stan- 34 KRF•1 • OK.indd 34 15.12.2009 10:41:57 Wolfgang Dreybrodt and Georg Kaufmann, Physics and chemistry of dissolution on subaerialy exposed soluble rocks by flowing water films Figure 9: Karren formation, from which water was collected. The grey line marks the flow path. The water was collected at the end of this line. dard. since many of the constants used depend on geneous environment channelling can occur and temperature, however, some temperature depen- parallel flow paths can arise, where the flow rates dence on the denudation rates is expected. in view are higher. For limestone then the concentration of the many approximations this is not of high c decreases and dissolution rates correspond- ∞ significance. ingly increase. in gypsum the solution is close to We have not addressed the issue of rillenkar- saturation and therefore the amount of dissolved ren formation. at present one may only speculate. rock is proportional to the volume of the flowing The surface of the rocks acts to flow like a two- water. one therefore could imagine that rillenkar- dimensional porous medium. in such an inhomo- ren could only originate at rough rocks. This issue 35 KRF•1 • OK.indd 35 15.12.2009 10:42:00 Karst Rock Features • Karren Sculpturing can be handled experimentally by simulating kar- be of utmost use for a better understanding. one ren formation experimentally on polished and of the purposes of this work is to stimulate such rough samples of plaster of paris. research. an object of further research should be to mea- sure flow velocities on limestone surfaces under natural conditions in dependence of rainfall in- Acknowledgement tensity, and also to take samples of the water at various locations on that surface to obtain calci- We thank Joyce lundberg and angel ginés for um concentrations. such experimental data could providing their field data in Figures 6, 7 and 8. 36 KRF•1 • OK.indd 36 15.12.2009 10:42:00 BioKarstic processes associateD 3 WitH Karren DevelopMent Heather VILES Karren features in many environments are cov- colonized by biofilms. on terrestrial limestone ered with a variety of organisms, and many au- surfaces on aldabra atoll, indian ocean, they thors have suggested that microorganisms, plants found that a cyanobacteria-dominated biofilm and animals may contribute to surface weather- recolonized cleared squares on the surface within ing of limestone and other soluble rocks and the 16 years at many sites. colonization was found to development of karst features. in this chapter i re- be rather patchy, however, with some particularly view the evidence for biological contributions to dry or hard sites experiencing very little recolo- karren development, starting with a consideration nization even after over a decade. Many biofilms of the types of organisms found on soluble rock growing on carbonate rocks or building stones are surfaces. highly biodiverse, as indicated by the findings of tomaselli et al. (2000) from a survey of european buildings. Organisms found on limestone Higher plants can also be key components of surfaces rock surface flora, although their growth is in many cases limited by the absence of soils. lime- Most bare rock surfaces are discoloured by a layer stone pavements, for example, often contain a of microorganisms, which together make up a bi- wide range of plants, many endemic to karst areas ofilm. such biofilms contain a mixture of micro- and of great conservation value (Ward and evans, organisms, including fungi, cyanobacteria (also 1976; Webb, 1995). animals also commonly range known as blue-green algae) and lichens as well over and inhabit bare limestone surfaces, espe- as associated extracellular products which create cially along coastal exposures where a whole suite a slimy surface. Biofilms have been observed on of sessile and motile animals have been found to many limestone surfaces, creating a grey to black occur. some animals are capable of extracting patina where they are dominated by cyanobac- nutrients directly from the rock surface, whilst teria and fungi and a multicoloured patchwork others make use of the rock surface biofilm as a where dominated by lichens. source of food. Figure 1 shows a characteristic biofilm on a organisms inhabit a range of rock surface limestone surface. such biofilms can be several niches, as shown in Figure 2. The terminology millimetres thick. research by Viles et al. (2000) presented in Figure 2, which derives from the illustrates the speed at which surfaces can become work of golubić and others, has been applied 37 KRF•1 • OK.indd 37 15.12.2009 10:42:00 Karst Rock Features • Karren Sculpturing Figure 1: Mixed cyanobacterial biofilm on Quaternary age limestones from Aldabra Atol , Indian Ocean. Width of view is 1 m. Figure 2: Rock surface niche terminology (from Golubić et al., 1981). mainly to microorganisms but can also be ap- several different types of endolithic niche. organ- plied to higher plants and animals. at the sim- isms which actively penetrate into the surface are plest level organisms growing on a rock surface called “euendolithic” whereas those that inhabit are termed “epilithic”. Those that live under the preformed fractures and cavities connected to the surface are referred to as “endolithic”. There are surface are called “chasmoendolithic” and those 38 KRF•1 • OK.indd 38 15.12.2009 10:42:01 Heather Viles, Biokarstic processes associated with karren development Figure 3: Lichens growing on marble in the central Namib desert. The west-facing slopes (left side of image) are characterized by luxurious lichen growths, and the east facing ones with thin crustose forms. that live in a subsurface layer within pores etc. and often species-poor assemblages. coastal are called “cryptoendolithic”. Finally, organisms limestone exposures support a diverse suite of which live under stones on the surface are called organisms tolerant of marine conditions includ- “hypolithic”. ing gastropods and algae, but with few lichens or The community of microorganisms, plants and higher plants. Finally, rock surfaces in cave en- animals found on bare limestone surfaces and vironments are populated by highly specialized the niches that they inhabit, varies according to communities – especially in areas well away from climate and the nature of the environment. in light where non-phototropic organisms domi- particular cases carbonate island terrains support nate. Viles (1995) proposed that at the macroscale unusual plant and animal communities, such as across a gradient of decreasing rainfall, biofilms those on aldabra atoll dominated by the giant would change from being dominated by epilithic tortoise Geochelone gigantea. in general the major forms to endolithic ones. controls are light and rainfall levels. on terrestrial at the more local level, rock surface communi- karst surfaces, for example, there are gross differ- ties are influenced by smaller scale variations in ences between temperate, semi-arid and humid light and water availability. For example, on many tropical communities. studies from around the inland limestone scarp slopes water flow paths be- world indicate fairly similar biofilm communi- come concentrated in some areas, producing fer- ties across most environments, although the arid tile grounds for thick, rich biofilms (often called and cold extremes are characterized by unusual Tintenstriche, because they look like a streak of ink 39 KRF•1 • OK.indd 39 15.12.2009 10:42:02 Karst Rock Features • Karren Sculpturing running down the scarp face). in other drier areas duce quite complicated effects. For example, fungi biofilms are thinner and less species rich. simi- have been found to produce spicily etched calcite lar microenvironmental differences are found in as well as sparmicritization in experiments by arid terrains. on marble exposures in the central Jones and pemberton (1987). some lichens, for ex- namib desert, in southern africa, for example, ample, produce different patterns of etching and well developed lichen communities with foliose boreholes under different parts of the thallus as and fruticose forms common are found on the evidenced by studies from Jerusalem which indi- west-facing slopes which receive high amounts of cate pinhead pits caused by apothecia or perithe- fog, whereas on the east-facing slopes which expe- cia of endolithic lichens and microgrooves formed rience the full force of the desiccating east winds by dissolution at the meeting point of lichen thalli only a thin cover of crustose lichens is found (see (Danin et al., 1983). Figure 4 shows fungal hyphae Figure 3). endolithic forms are found more com- from the base of crustose lichens growing on mar- monly on the harsher, east-facing surfaces. Both ble in the central namib desert boring their way plants and rock-dwelling animals will tend to in- into the rock. lichens can also have a biophysical habit depressions and fractures within the rock weathering effect on limestone surfaces, as dem- surface, where there is more moisture, shade and onstrated in the pioneering experiments of e. Jen- soil development. nie Fry in the 1920s (e.g. Fry, 1927) and more re- cently by Moses and smith (1993). lichens are ca- pable of absorbing an enormous amount of water Weathering and erosive action of relative to their size and weight, and on wetting organisms on limestone surfaces and drying of lichens partially attached to lime- stone through hyphae at the base of the thallus it has commonly been found that the microorgan- considerable stresses are put onto the limestone, isms, plants and animals inhabiting karst terrain causing flaking and granular disintegration. sim- play active roles in denuding the surface. Both bi- ilar biophysical effects have been ascribed to chas- ophysical and biochemical processes are involved, moendolithic cyanobacteria which can expand and organisms inhabiting both epilithic and en- by 300% on wetting and dislodge calcite crys- dolithic niches can cause weathering and erosion. tals from the sides of fissures (Danin and cane- looking firstly at biofilms, many studies have va, 1990). Many organisms can produce both bio- shown the biochemical weathering effect of epi- chemical and biophysical effects as evidenced by lithic biofilms that produce acid exudations which studies on lichens from the genus Xanthoparmelia can contribute to calcite dissolution. When bio- growing on sandstone in arizona (paradise, 1997). films are removed from limestone surfaces the un- paradise found that biophysical weathering domi- derlying surface is often found to be etched and nated under the centre of the lichen cortex, with pitted (Viles, 1987). euendolithic microorganisms biochemical weathering predominant towards produce similar biochemical weathering, but in the edge of the thallus. synergistic associations of this case focused on producing the holes and tun- biological weathering processes with dissolution nels which the organisms then inhabit. in some and physical weathering may also occur. papida et cases a networks of euendolithic boreholes is pro- al. (2000), for example, found enhanced physical duced stretching several millimetres into the rock. weathering of limestones to occur under experi- This produces an altered near surface layer with mental conditions when inoculated with mixed high porosity. cryptoendolithic biofilms may microbial populations, in comparison with fresh also contribute to exfoliation of limestone surfac- limestone samples. es, as suggested recently for the Hirao-dai karst plant roots can have endolithic growth forms, in Japan (Darabos, 2003). Many organisms pro- producing tunnels in limestone surfaces through 40 KRF•1 • OK.indd 40 15.12.2009 10:42:02 Heather Viles, Biokarstic processes associated with karren development Figure 4: Fungal hyphae from crustose lichens boring into marble, central Namib desert. Scale bar = 50 microns. biochemical and biophysical processes. pioneer- remove particulate limestone along with the bio- ing experiments carried out in the 19th century by film. Detailed experimental studies by andrews Julius sachs indicated the efficacy of roots from and Williams (2000) on the chalk shore platforms plants such as Phaseolus multiflora (bean) etching along the south coast of england found a consid- into polished marble surfaces. recent attempts to erable amount of calcium carbonate within limpet reproduce such experiments have met with mixed ( Patel a vulgata) faecal pellets, which appeared to success, however (Mottershead and Viles, 2004). have come from limpets grazing on algae. Vis- several studies have shown that plants are capa- ible limpet grazing trails were found on the chalk ble of taking up calcium and magnesium from surface, emanating from the “home scar”. The limestone and other rock surfaces and soil miner- “home scar” also appears to be excavated by a als through biochemical activity in the roots and combination of physical and chemical processes rhizosphere and thus may play a role in sculpting by the limpets themselves, perhaps enlarging a both bare and subsoil limestone surfaces (Hins- pre-existing depression within the chalk surface. inger et al., 2001). limpets here are thought to be responsible for animals actively denude limestone surfaces 12% of downwearing of the platform in areas they through a range of biochemical and biophysical frequent, and 35% or more where population den- processes. Many gastropods, for example, can sities are very high. similar processes are carried graze effectively on rock surface biofilms and out by a range of rock-dwelling organisms, largely 41 KRF•1 • OK.indd 41 15.12.2009 10:42:03 Karst Rock Features • Karren Sculpturing on coastal limestone exposures, although stanton organic acids which may be produced in soils and (1984) and others have found distinctive hollows by decaying organic material (see the early work that appear to be produced by terrestrial gastro- of Murray and love, 1930). trudgill (1985), for pods. example, illustrates the highly corrosive nature Biochemical attack on limestone surfaces can of water acidified as it flows over tree bark pro- occur without direct involvement of organisms. ducing characteristically polished and weathered For example, many reports have been made of limestone surfaces under trees. animal urine the chelating and dissolving effect of a range of may also dissolve calcite, producing runnels. it Figure 5: The impacts of rock wal aby urine on limestone surface biofilms, Napier Range, NW Australia. Width of view is 7 m, on the upper edge. 42 KRF•1 • OK.indd 42 15.12.2009 10:42:05 Heather Viles, Biokarstic processes associated with karren development can also have an indirect effect on the weather- tion of such landforms are often called biokarstic. ing of limestone surfaces, as evidenced in Figure 5 Biokarst features can be erosional or depositional, which shows the defoliating impact of rock wal- or involve a combination of the two processes, and laby urine on limestone surface biofilms from the are commonly found on exposed limestone sur- napier range in nW australia. faces in a range of environmental settings (Viles, although a wide range of biochemical and bio- 1984). an early paper by Jones (1965) describes physical processes capable of effecting weathering many of the erosional features found on limestone and erosion can be identified, it must also be re- pavements as being at least partly biokarstic in or- membered that organisms can protect limestone igin. organic-rich soils cause accelerated dissolu- from other agents of denudation as has been found tion here, and endolithic lichens on clint surfac- for other rocks (e.g. Kurtz and netoff, 2001; Miku- es produce roughened surfaces from dissolution. las, 1999). Biofilms, for example, act to bind sur- More recent work has identified suites of small- faces together and absorb incoming rainfall and scale karren along coastal areas as being biokarst runoff thereby reducing the inorganic dissolution. (e.g. schneider and torunski, 1983). until the community decays and dies the under- if organisms commonly inhabit limestone sur- lying surface is protected. lichens act in a similar faces on which karren features are developed, can fashion, acting as a net agent of protection even an explicit link be made between some of the bio- whilst producing fungal boreholes at the base of chemical and biophysical processes reviewed in the thallus, but contributing to dramatic episodic the previous section and the production of kar- surface removal as they decay. some lichen spe- ren? if so, then karren may be at least partly bi- cies decay from the centre, with large sections of okarstic in nature. at the simplest level, the near the thallus peeling away bringing with it portions ubiquity of biofilms on limestone surfaces (in of the underlying rock. recent studies by carter non-soil covered terrain) and the importance of and Viles (2003) indicate the general protective organic acids in most soils suggest that both bare role played by black lichen-dominated biofilms on and soil-covered karren are influenced by organic limestone used as a building material. in field and processes. However, it has so far been found to be laboratory experiments a cover of epilithic lichens difficult to prove a strong link between organic ( Verrucaria nigrescens) was found to retain mois- processes and the development of karren features. ture and dampen thermal stresses at the limestone some progress has been made with experimental surface thus reducing the potential for weathering. studies, for example the work of Fiol et al. (1996) rock surface dwelling biofilms may also contrib- on the influence of rock surface microorganism ute to surface protection through biomineraliza- communities in ril enkarren development. Their tion processes, as found by rodríguez-navarro investigations showed that mechanical detach- et al. (2003) in an experimental study. The bacte- ment of small particles is a key process in the ria Myxococcus xanthus was found to be able to development of rillenkarren, caused by raindrop produce a protective and consolidating carbonate impact which is enhanced in areas where endo- matrix on stone samples under laboratory condi- lithic cyanobacteria have previously corroded the tions. surface. in some cases highly unusual karren features have been identified which appear to owe their Biokarst and karren origin dominantly to biological processes, such as phytokarst. The classic phytokarst landscape is Biokarst refers to karst landforms created, or influ- that described by Folk et al. (1973) at Hell, grand enced to a significant degree, by biological proc- cayman island. Here, a series of limestone pinna- esses. in turn, the processes involved in the forma- cles in a low-lying swampy environment have been 43 KRF•1 • OK.indd 43 15.12.2009 10:42:05 Karst Rock Features • Karren Sculpturing Figure 6: Phytokarst pinnacle from Aldabra Atol . blackened and dissected in a random spongework tified type of phytokarst are the light-orientated pattern which Folk et al. ascribe to the action of erosional pinnacles found in the lit zone of many cyanobacteria (blue-green algae). Figure 6 shows cave entrances (as reported by Bull and laverty similar randomly sculpted forms from aldabra in 1982 in Mulu, Borneo, for example, and some- atoll. Jones (1989) has provided further detailed times given the alternative name of photokarren). microscopic observations of this phytokarst and However, making a convincing process/form link illustrated the variety of weathering roles played between biofilm processes and phytokarst has by the cyanobacteria and fungi dominated bio- proved to be difficult and there may be a range of films. on hard dolostones, epilithic microflora controls operating at different scales (Viles, 2001). dominates, whereas on softer limestones a diverse indeed, taboroši et al. (2004) propose that much endolithic flora is found. another commonly iden- karren on young island karst is better called “eo- 44 KRF•1 • OK.indd 44 15.12.2009 10:42:06 Heather Viles, Biokarstic processes associated with karren development genetic karren” rather than “phytokarst” as it is and bioerosion occur, especially as they are often probably polygenetic in origin and the heterogene- highly spatially and temporally patchy. Danin ous nature of the young limestones exerts a major (1983), however, used information on the depth of control on the resulting karren forms. pits occupied by cyanobacteria on dated walls in in other cases, process/form links may be easier Jerusalem to estimate an annual biological weath- to prove. some interesting work has been done by ering rate of 0.005 mm per year, which probably simms (1990) on coastal phytokarst or photokar- exceeds the rate of dissolution in this semi-arid ren in ireland. He finds ample evidence of light area. More recent work in Jerusalem and rome orientated pinnacles adjacent in ancient cave pas- indicates pitting of marble by cyanobacteria from sages. The photokarren develop near unroofed the genus Myxosarcina to be occurring at the rate sections where light can enter. simms notes that of 0.025 mm per year (Danin and caneva, 1990). scallops, formed when the caves were active, are one key issue in assessing the role of organ- well preserved in other parts of the caves despite isms in karren formation is that of temporal scale. the ingress of seawater today. This implies that are organisms permanent enough features of the seawater is not having a direct dissolutional effect, rock surface environment over the timescales of and thus the creation of the photokarren is the karren formation? another issue is spatial scale. only form of weathering occurring today. Many of the biological processes take place at the sub-millimetre scale, whereas karren features de- velop over the centimetre to metre scale. even if Issues for further work biological processes play a role in karren develop- ment, other factors undoubtedly exert larger scale Most surfaces on which karren features are devel- controls (such as jointing). Thus, it is probably oped are exposed to biological influences. How- more accurate to say that biological processes con- ever, it is difficult to prove whether such biological tribute to the formation of many karren and other influences are a necessary part of karren forma- karst landforms. Whether they can be seen to play tion, or whether they have little real impact. in- the dominant or decisive role is perhaps much less deed, some authors have argued in the past that important. biological processes act to degrade karren features The debate over biokarst draws attention to the produced by inorganic dissolution processes. Fur- many, diverse ways in which biological processes ther information on rates of biological weather- influence karst features and furthermore the large ing may help resolve some of the questions about role that organisms play in the global carbon- the role of biokarstic processes. it is difficult to ate cycle (schneider and le campion alsumard, assess the speed at which biological weathering 1999). 45 KRF•1 • OK.indd 45 15.12.2009 10:42:06 KRF•1 • OK.indd 46 15.12.2009 10:42:06 Karren siMulation WitH plaster 4 oF paris MoDels Tadej SLABE The experimental modelling of rock features in The chapter presents the latest findings of ex- plaster helps reveal the manner of their forma- periments in the formation of subsoil rock relief tion, the development of individual rock features and plaster blocks exposed to rain. in nature, and their connections in rock relief. it also helps us distinguish the proportion and significance of the legacy of various factors that Previous experiments described in the participated in the formation of rock relief and literature indirectly therefore the various periods of its de- velopment. pluhar and Ford (1970) studied the formation of limitations do exist regarding either the size of flutes (rillenkarren) using hydrochloric acid on a the models or the more rapid solubility of plaster dolomite block. However, when they covered the compared to carbonate rock. primarily, we can dolomite with a layer of quartz sand, only micro follow the manner of the shaping of soluble stone pits developed. in various conditions and the development of the glew (1976) and glew and Ford (1980) studied rock relief on it; however, the direct comparison flutes with experiments exposing plaster surfaces of individual rock features on plaster and on rock inclined from 22.5° to 60° to artificial rain. They is more difficult, especially due to their size. as a determined that flutes develop where the layer of rule, features on plaster are smaller (slabe, 1995b). water flowing off an inclined surface is thin and The rapid solubility of plaster influences their fre- does not prevent the direct impact of raindrops quently jagged form and rough surface. experi- on the rock. under a thicker layer of water, how- ments on the formation of subsoil rock features ever, a smooth rock surface develops. The length must be interrupted and models must be taken of the flutes is related to the inclination of the rock apart since this is the only way to observe their surface and their cross-sections have a parabola continuous development. However, the experi- shape, the shape that most efficiently directs the ments continue to confirm that their use is help- erosive action of raindrops along the axis of the ful. flute. industrial plaster (caso x 2 water) is used in Dzulynski et al. (1988) used modeling to study 4 the experiments. in one litre of water, 2.5 grams the formation of karren, exposing a fissured piece of gypsum are dissolved at 20°c (Klimchouk, of plaster to artificial rain. They studied karren 2000a). shaped by rainwater as well as subsoil karren. 47 KRF•1 • OK.indd 47 15.12.2009 10:42:06 Karst Rock Features • Karren Sculpturing They were interested in the dissection of the plas- Recent experiments ter into channels that developed along the fissures, the formation of the protuberances between them, Experimental formation of subsoil karren and the influence of the level of water surround- ing them on their development. Thicker columns some of the findings from the experiments de- formed when the plaster was crisscrossed by only scribed below are presented in detail in Zeitschrift a few fissures, and when the network of fissures für geomorphologie (slabe, 2005). was dense, the columns were thinner. They de- i attempted to verify the descriptions of various termined that the amount of rain influences the subsoil features (see chapter 11) with experiments speed of the dissolution of the plaster but not the on the formation of subsoil karren. form of the artificial karren. Karren that devel- We sliced a plaster cube into small columns oped on plaster covered with sand was similar to with six-centimetre square cross-sections and that which developed on bare plaster, and only the heights of 30 centimetres (Figure 1). The separated columns were less sharply dissected. on the walls columns were placed tightly side by side in a large of a small model of uncovered karren, it was possi- bucket and covered with soil. We drilled small ble to discern vertical channels at the end of lapies holes in the bottom of the bucket and then filled wells (karrenröhren) and funnel-shaped recesses it with water. We provided a continuous supply of on their tops with channels underneath them. water to keep the surface of water five centime- experiments with plaster have been employed tres above the surface of the soil. slowly, the water in researching cave rock features as well. scal- started to percolate through the soil and then flow lops were studied by rudnicki (1960), curl (1966), through the holes in the bottom of the bucket. goodchild and Ford (1971), and allen (1972). after the experiment, which with breaks lasted such experiments helped Quinif (1973) explain almost 400 hours, the columns were eighteen to the formation of ceiling pockets, ewers (1966, twenty-seven centimetres tall (Figure 2). 1972, 1982) study the development of the net- The upper two thirds of the walls of the col- work of original watercourses through rock, and umns are eroded with tiny features, the traces of lauritzen (1981) study above-sediment channels. the percolation of water through the alluvium and tućan (1911) exposed limestone and dolomite to its flow along the contact point between plaster hydrochloric and nitric acid and established that and alluvium. their surfaces were similar as on the karst surface, The entire surface of the peak section of col- of course with characteristic differences between umns is minutely eroded with micro recesses (Fig- limestone and dolomite. trudgill (1985) describes ure 3a). The micro recesses are up to two centime- the smoothing of rock surfaces due to exposure tres in diameter, although the majority is smaller. to acidic waters. He proved his findings through it appears they are the consequence of the con- laboratory experiments using acids and scanning stant percolating of water through the most per- rock surfaces with an electron microscope. meable layer of soil. as a rule, epikarst rock fea- experimental research (slabe, 1995a, b) on cave tures (see chapter 11) have this sort of roughness above-sediment ceiling channels and anastomoses, on rock covered with soil where the contact with below-sediment flutes, various types of scallops the rock is loose. and the influence of rock and hydraulic conditions The larger recesses found as a rule in the middle on their size and shape, and ceiling pockets that of columns (Figure 3b) have relatively smooth sur- occurred due to the percolation of water through faces. i called recesses of this type ‒ incomparably fissures helped a great deal in conceptualizing larger, of course ‒ subsoil scallops (slabe, 1998). their development and the diverse formation of along fissures or other weak spots in the rock, karst caves. individual deeper semicircular or channel-shaped 48 KRF•1 • OK.indd 48 15.12.2009 10:42:06 Tadej Slabe, Karren simulation with plaster of Paris models ranged from one to three centimetres formed on the walls of the lower sections of the columns (Figure 3c). The locally flooded zone, which devel- oped because the quantity of water flowing along the contact was greater than the quantity that could flow out of the perforated bottom of the bucket, reached the upper level of the channels. This margin is often marked by notches (see chap- ter 11). This subsoil formation of karren is charac- teristic of the alluvium-covered and periodically flooded valley systems of the lower karst regions in southern slovenia and the estavelle mouths on cerkniško polje. on the lower surfaces of the columns (Fig- ure 3c), including those cut in half horizontally, Figure 1: Plaster block from which stone forest devel- oped. there are distinctive networks of above-sediment anastomoses (slabe, 1995a). The channels compos- ing them have omega-shaped cross-sections with recesses often occur in many cases that can in diameters measuring up to three centimetres. The time grow into subsoil tubes. networks have several stories. Channels (see chapter 11) whose diameters angular and square cross-sections were only Figure 2: Subsoil stone forest in plaster. 49 KRF•1 • OK.indd 49 15.12.2009 10:42:08 Karst Rock Features • Karren Sculpturing preserved in the lower sections of the columns. sections of the columns is the consequence of the The characteristically pointed shape of the upper dispersed percolation of water through the soil that covered the plaster. in the final experiment (Figure 4) of this series, we covered larger plaster pillars (the cross-section of the larger pillars was 20 centimetres, and they were 25 centimetres tall) with fine-grained clay, which is poorly permeable to water as therefore was the contact between the clay and the pillars. subsoil shafts were the first feature to develop be- tween the columns. on the upper sections of the columns (Figure 4, left), vertical subsoil channels a formed as parts of the shafts with funnel-shaped mouths at the top. Their cross-sections reached 3 centimetres. it appears that water finds the most conductive path along the poorly permeable con- tact between plaster and clay and forms streams. This hypothesis was confirmed by the bubbles that appeared on the water surface, which revealed numerous distinctive ponors. special channels formed in the locally flooded zone, like those de- scribed in the previous experiment, developed in the lower section of the columns. anastomoses developed on the lower planes of the columns. The peaks of the columns gradually sharpened, nearing a pointed shape. We interrupted the ex- periment several times to observe all the stages of development of subsoil karren. after 800 hours, when the experiment was con- cluded, the two taller columns were 20 centimetres high, and the width of the thicker column was 20 centimetres. From the smaller columns remained two 5-centimetres tall and up to 1.5 centimetres thick pieces of plaster. b The columns are quite pointed (Figure 4, right), which is most evident in the two larger columns, c especially the thickest. The sharp side edges were preserved. over time, the rock relief became increasingly similar to that in the first experiment, especially on individual faces although as a rule the surface of the columns was flatter. new characteristics are Figure 3: Column with characteristic rock relief: a. re- preserved, such as dissection with more or less cesses and protuberances; b. recesses; c. channels and vertical and, depending on the varying degrees anastomoses. of permeability at the contact between the plaster 50 KRF•1 • OK.indd 50 15.12.2009 10:42:12 Tadej Slabe, Karren simulation with plaster of Paris models Figure 4: Development of gypsum subsoil pil ar; left: early stage, right: final stage of experiment. and clay, more or less meandering subsoil chan- rock peaks exposed to rain and their rock relief, nels. Higher on the walls of the columns than in which often formed on top of the legacy of older, the first experiments, horizontal or variously in- especially subsoil relief. i therefore decided to ex- clined notches are preserved, reflecting the more pose small plaster blocks with sides measuring completely filled cracks and tighter contact be- forty centimetres square to rain. tween the plaster and clay and consequentially the i exposed three blocks of various shapes to nat- lower permeability of the model. The surface of ural and artificial rain. i leaned an uncut cube at the plaster was weathered in more places, mean- a 27° angle and cut a second cube diagonally into ing the solution was not carried away everywhere two halves. i placed the first resulting prism on a at the same time. shorter face and wedged it so its longest face had a 36° incline. i placed the second one on an edge so that one of its smaller square faces was horizon- Experimental modeling of “rock peaks” in tal, the second was vertical like the triangular side plaster exposed to rain faces, and the longest face was overhanging with a 45° incline. i thus obtained flat faces inclined at in studying the rock relief of karren and stone for- 27°, 36°, 63°, and 90°, overhanging faces with in- ests, questions arose regarding the development of clination of 9°, 27°, and 45°, and edges of various 51 KRF•1 • OK.indd 51 15.12.2009 10:42:14 Karst Rock Features • Karren Sculpturing created (Figure 5). The upper sections were cov- ered by flutes (Figure 5a), the central sections were relatively smooth, and channels developed on the bottom sections. This characteristic shaping of the faces is also shown in a drawing in a book by Ford and Williams (1989). The flutes grew slowly from the upper edge downwards and deepened. at first, their form was indistinct and it was impossible to measure their size in detail because the ridges between them were rounded. However, i believe that their width did not differ substantially from the size of “mature” flutes. on the face inclined at 27°, the flutes average 3.6 centimetres long, 0.7 centimetres wide, and 0.5 centimetres deep in the upper part; lower down they are more shallow and end in a wedge-shape. on the face inclined at 36° they average 4 centimetres long, 0.8 centimetres wide, and 0.5 centimetres deep in the upper part. on the steepest surface (63°), their most distinct parts average 7 centimetres long, 0.5 centimetres wide, and 0.3 centimetres deep in the upper sec- tion. in the last case, it is difficult to determine the length because individual flutes extend all the way to the channels in the bottom section of the face. Figure 5: Surface of plaster exposed to rain with charac- on the steepest surface, the flutes are therefore teristic rock relief: a. flutes; b. smooth surface; c. chan- longer and also somewhat narrower. They also ap- nels. peared on the overhanging surface inclined at 9°. individual channels (Hortonian-type runnels; lengths along them. i cut a channel similar to a Ford and Williams, 1989) formed on the bottom subsoil channel, one centimetre in diameter, into section of the inclined faces (27°, 36°, 63°). at the face with an inclination of 63°. The surface of first, they were relatively straight, 0.5 centimetres the faces with different inclinations was smooth. wide and 10 centimetres long. Between them were We can distinguish two kinds of development larger areas without channels. initially, the chan- on the plaster blocks. The variously inclined faces nels mainly deepened. Their cross-sections took exposed directly to rain and the vertical faces the shape of the letter omega turned upside down. were formed in a characteristic way. The over- The ridges between them were gradually sharp- hanging faces, however, were formed uniquely in ened by rainwater. The channels widened to 2 cen- both cases. timetres. in the bottom section thus forms a web after ten hours, straight, narrow, long, and of channels that initially develop as collectors of shallow scallop-like recesses first appeared in the water from the upper section of the inclined face middle of the face with a 36° inclination, and nar- and when they become deeper, rainwater starts to row channels appeared on their lower sections. shape them as well as carving flutes on areas be- The upper edges were dissected into semicircu- tween the channels (Figure 5c). lar notches, below which flutes began to appear. a special type of channel, which Ford and Wil- gradually, three distinct sections of the faces were liams (1989) call a “Hortonian-type dissolution 52 KRF•1 • OK.indd 52 15.12.2009 10:42:15 Tadej Slabe, Karren simulation with plaster of Paris models Figure 6: Channels on a vertical wal . channel”, forms on vertical faces on which water grew by one centimetre, and a smaller meander- flows from the horizontal top (Figure 6). at first, ing channel cut into the bottom, particularly dis- channels with three-millimetre diameters formed tinct in their lower part. The mouths of the chan- that meandered slightly. They were most distinct nels became funnel-shaped, 3.5 centimetres wide, on the upper section of the faces, below edges and at first dissected by flutes. This is a frequent that were jagged with semicircular notches. These and characteristic form for the peaks of the lunan channels extended to the bottom of the vertical stone forests (Knez and slabe, 2002). faces. initially they deepened and their cross-sec- The entire plaster blocks are shaped in a char- tions acquired the form of the greek letter omega. acteristic way. The exposed edges remain flat and Between the channels were larger undissected sharp but become jagged in accordance with areas, but channels later covered the majority of their dissection by flutes. The outer corners of the face. The channels gradually begin to merge the horizontal and vertical faces are dissected by and the largest reach 3 centimetres in diameter notches that are the mouths of vertical channels with smaller channels remaining on the ridges be- into which water flows from the upper face. all tween them. The rock relief of the pillars in spain’s the plaster blocks get sharper, and the upper sec- el torcal stone forest was formed in this fashion. tions of originally vertical faces incline inwards The upper sections of the vertical faces of the plas- and are increasingly exposed to the direct action ter blocks that are directly exposed to rain gradu- of the rain. on the plaster block with a flat upper ally become less steep; so far, they have diverged face, the vertical side faces bow concavely and the by three centimetres from the vertical. The edges corners project outward three or four centimetres. between the upper parts of channels, which are increasingly exposed directly to rain, therefore become sharp and the channels open semicircu- Conclusion larly. slowly they transform into flutes (Figure 7). i carved two channels with diameters of one in spite of the limitations of experiments of this centimetre into the surfaces inclined at 63° and type described in the introduction, particular- 36° to represent subsoil channels. Their diameters ly the rapid solubility of plaster due to which the 53 KRF•1 • OK.indd 53 15.12.2009 10:42:19 Karst Rock Features • Karren Sculpturing Figure 7: Development of rain flutes on channels. surface is often more minutely dissected and the known at first and are only later discovered in na- limited possibilities for monitoring its formation, ture with the help of or because of our previous- these experiments offer numerous advantages. ly acquired knowledge. in any case, these experi- They complement the knowledge acquired in na- ments should continue, supported of course with ture and frequently open new directions for fur- a comprehensive and interdisciplinary foundation ther thought and research. often the forms that and upgraded knowledge. develop during the shaping of models are un- 54 KRF•1 • OK.indd 54 15.12.2009 10:42:21 tHe proBleM oF rillenKarren 5 DevelopMent: a MoDelling perspective Matija PERNE and Franci GABROVŠEK a few chapters discuss morphological aspects of mation. The rain is continuously distributed to rillenkarren and observation of their formation. the surface of the water film. consequences of However, a satisfactory model of rillenkarren for- raindrop impacts are not investigated; mation based on the first principles has not been • we use thin film approximation for the navier- presented so far. For the time being, it would be stokes equation. additionally we neglect sur- of a benefit if we could prove or disprove some as- face tension; sumptions used for the rillenkarren development. • the development of rillenkarren has to go experimental and field evidences lead to the through a stage in which the surface is only following conclusions for the rillenkarren devel- gently undulated, covered with shallow rills opment: which are becoming deeper. • rillenkarren form on initially flat inclined sur- a satisfactory model of rillenkarren formation faces of soluble rock under constant rain; thus has to predict that a rill which is not as deep • they form on all soluble surfaces provided that as mature rills will deepen. these are exposed to the rain long enough; • the formation of rillenkarren is dissolutional; they form if dissolution rates are different on Methods and results different positions on the rock surface. to convert these observations into a mathemat- Protorill ical model, we assume a flow of thin water film with uniform (continuous) recharge over an in- all the models are tested on the same rock surface clined surface of soluble rock. The question which form, dubbed protorill. it has parabolic shape be- we want to answer is whether such film alone is cause every gentle curve is parabolic in the first sufficient for the rillenkarren development. our approximation. its slope is 40° and it is 2 cm wide, model includes some other assumptions which which are typical values for rillenkarren, while are making the calculations easier: it is only 0.5 mm deep, much less than real ril- • concentration of dissolved rock is small enough lenkarren. The upper 16 cm of such a rill were that it does not influence hydrodynamics, i.e. studied. The density of water is taken to be 103 kg/ the flow is calculated in advance; m3 and its viscosity 10-3 pas (Figure 1). • the flow is calculated in a steady-state approxi- 55 KRF•1 • OK.indd 55 15.12.2009 10:42:21 Karst Rock Features • Karren Sculpturing Figure 1: The protoril . Rain is uniformly distributed over the whole surface. Blue arrows indicate flow lines. The flow Fluid flow is described by navier-stokes equation the water film thickness and to the surface slope. which is generally difficult to solve. Because of That is that some reasonable approximations are used in 3 ρ g h j = − ∇( z + m , (2) 0 ) order to simplify it. η 3 For our needs water can be considered incom- pressible. in this case the navier-stokes equation where g is acceleration of gravity and ∇ stands for becomes derivation in x and y directions only. conserva- ∂v ρ + (v ⋅∇)  2 v = f − ∇ p +η∇   v, (1) tion of water gives another equation: b  t ∂  where ρ is density, v is velocity, t is time, f are b body forces per unit volume, p is pressure and η is viscosity. We neglect inertial forces to get the so-called lu- brication approximation (Kondic, 2003). in other words, we neglect the whole left side of equation 1, and second derivatives of velocity in direction parallel to the surface. additionally we neglect surface tension. a local cartesian coordinate system in which the rock surface lies in xy plane is introduced (Fig- ure 2). Here α is the surface slope, h is thickness of the water film, m = h / cosα is water depth and z is elevation of rock surface above a reference 0 level. Within the mentioned approximations the density of water flow j is proportional to the third Figure 2: Coordinate systems. The local coordinates are power of the water film thickness and to the sur- introduced, so that the rock surface lies in xy plane. See face slope. j is proportional to the third power of the text for notations. 56 KRF•1 • OK.indd 56 15.12.2009 10:42:22 Matija Perne and Franci Gabrovšek, The problem of ril enkarren development: a model ing perspective Figure 3: The left graph shows the rock surface and the right one the steady-state water surface when the rock is exposed to rain. h ∂ ∇ ⋅ j = v (4) ∇ ⋅ j = v d  r − , t ∂ (3) where v d is time average of rain intensity. The flow where v r is rain intensity in the z direction. density j is parallel to the water surface slope. The The equations 2 and 3 can be solved numeri- equation is nonlinear because j depends both on cally for arbitrary surface shape. We applied the the third power of water film thickness, or m, and method of time propagation which is efficient on gradient of m itself. But if the local water depth enough to find an approximate steady-state solu- is much smaller than typical height differences tion. The algorithm flows as follows: an initial ap- between points on the rock surface, the rock sur- proximation for m is taken, j is calculated from face slope is approximately equal to the water sur- equation 2 and the new m after a short time step is face slope and can be used in its place. Thus the calculated from equation 3. The procedure is then equation becomes linear and easy to solve analyti- repeated with the new m as initial approximation cally. We used the method of characteristics to until m converges to a steady state and does not solve it. change anymore. results of numerical and analytical solutions Figure 3 shows a test result of the algorithm on are shown on Figure 4. The results are almost iden- a slope with a depression. a pool of water fills the tical for the upper part of the rill. on its lower part depression the same way as in reality. The method the numerical solution becomes smoother while performs as good as expected, or even better. it the analytical one stays sharp. The numerical al- should be noted that in this case an assumption gorithm suffers from numerical diffusion which of lubrication approximation that water and rock smoothnes the solution. Differences could also surfaces are nearly parallel is not fulfilled but the result from the additional approximation used result is realistic anyway. in analytical calculation. it turns out that the dif- to test the numerical solution, which is in- ference between both solutions is important only herently only approximate, the water flow over when the water is deep in comparison to the rill the protorill can be calculated analytically. The and the presumptions of the analytical method steady-state form of the equation 3 is are no longer fulfilled. Thus, both solutions are in 57 KRF•1 • OK.indd 57 15.12.2009 10:42:22 Karst Rock Features • Karren Sculpturing Figure 4: The depth of water film as a function of position on the protoril . Left graph presents the results of analyti- cal calculation and the right graph the numerical one. The ril dips from left to right. The results are for rain inten- sity of 10 mm/h. Al coordinates are in mm. agreement when the analytical one is correct. That the film thickness (Dreybrodt, 1988; Kaufmann means the difference results from the additional and Dreybrodt, 2007). Therefore, we use equation approximation in analytical method, so the nu- ∂ρ (5) merical solution was used in further calculations. S = β ( c − c eq ) ∂ results are also in good agreement with experi- t mental data given by pettersson (2001). to describe the dissolution rates. β is a rate con- stant, ρ S is surface density of dissolved matter, and ceq is the equilibrium concentration of calci- Dissolution kinetics um with respect to calcite and c the concentration of calcium in the solution. The dissolution rates of calcite are determined by in the cases of gypsum or salt CO conversion 2 three rate-controlling processes (Kaufman and plays no role in dissolution. even more, we assume Dreybrodt, 2007): that the surface reaction is fast and only diffusion • the kinetics of dissolution at the mineral sur- is rate limiting. For more details on gypsum dis- face, which depends on the chemical composi- solution a reader is referred to Jeschke et al. (2001). tion of the solution at the mineral surface; Diffusion is described by diffusion equation • the diffusion of ionic species and co towards (6) 2 Dc 2 and away from the calcite surface; = D∇ c, Dt • the conversion of co into H+ and Hco-. 2 3 Depending on particular conditions any proc- where D is diffusion coefficient and Dc/Dt denotes ess can be rate limiting (see Kaufman and Drey- substantive (or material) derivative of concentra- brodt, 2007). in the case of a few tenths of a mil- tion. The dissolution rate at the rock surface be- limetre thick water film, molecular diffusion and comes: co conversion are rate limiting. it turns out that (7) 2 ∂ρ c ∂ the dissolution rate is only weakly dependent on S = D , t ∂ z ∂ z=0 58 KRF•1 • OK.indd 58 15.12.2009 10:42:22 Matija Perne and Franci Gabrovšek, The problem of ril enkarren development: a model ing perspective where z is the coordinate normal to the rock sur- slower; as rain is falling vertically, there is less rain face. for a given surface on steeper slopes so concentra- tion of dissolved limestone is higher and dissolu- tion rate is slower according to equation 5. Models of rillenkarren formation on the other hand, for a given vertically pro- jected surface there is more rock surface if the Both modes of dissolution, for limestone and for slope is steeper. Therefore, steeper surfaces lower gypsum and salt, are first applied on inclined flat faster than gentler ones. surface and then on the protorill. For limestone, the equation 5 is used. Therefore dependence of Limestone, protorill concentration on z is not taken into account. Water flow is always parallel to the slope direc- For gypsum and salt, equation 7 is used with tion of local water surface. The shape of the water the value of − 2 9 m D =10 s which is close to real surface is calculated as described in section be- diffusion constants for these substances at normal fore. We use it to calculate flow lines, paths along temperatures. which water flows downward. The protorill is uni- formly covered with flow lines such as schemati- Limestone, flat surface cally shown on Figure 1 and dissolution rates are The surface is oriented so that its upper edge is calculated along every one of them. each flow line horizontal. rainwater then flows in the direction can be treated independently, because water in the of the slope. it turns out that the dissolution rate film neither enters nor leaves it. The lateral gra- is the same everywhere on the surface (Kaufmann dients of concentration are assumed to be small and Dreybrodt, 2007; see chapter 2). enough to neglect the diffusion of solutes into or on a steeper surface, the dissolution rate is out of the flow line. Dissolution rates on the points Figure 5: Values of (c – c )/(c cos α) eq eq for β = 10-7 m/s. According to equa- tion 5, the values are proportional to the rate of lowering of the surface of a limestone protoril , i.e. higher points mean faster lowering. The top of the ril is on the top left side. Values on horizontal axes are in mm. 59 KRF•1 • OK.indd 59 15.12.2009 10:42:23 Karst Rock Features • Karren Sculpturing Figure 6: Values of (1/c ) (∂c/∂z)z = 0 in mm-1. These are eq proportional to the rate of lowering of the surface of a protoril made of gypsum or salt according to equation 7, i.e. higher points mean faster lowering. The top of the ril is on the top left side. Values are calculated with resolution of 51 nodes along z direction in finite differ- ence scheme. Values on horizontal axes are in mm. of rock surface that do not lie directly on a calcu- Gypsum or salt, protorill lated flow line are obtained by interpolating. The The model for the flat surface has to be only slight- results are shown on Figure 5. ly modified to handle dissolution on a flow line along a curved surface. The same flow lines as for Gypsum or salt, flat surface the limestone protorill are used, dissolution rates on these rocks, the concentration of dissolved all over the rill are calculated and are presented rock is dependent on z while at the rock surface on Figure 6. it is assumed to be at c (see previous section), so eq mass transport in both x and z direction is taken into account. From the lubrication approximation Discussion and conclusion water velocity field is calculated and advection in both x and z directions is accounted for. Diffusion The presented models do not predict rillenkarren in x direction is neglected because of small con- formation although it is known from nature and centration gradients while in z it is the main force experiments that they do form under the mod- driving the mass transport and so has to be taken elled circumstances. This means that something into account. We end with advection diffusion crucial for rillenkarren formation was not ac- problem, which can be solved with finite differ- counted for correctly. ence scheme applying suitable coordinate trans- some of the approximations, simplifications formation. Dissolution rates directly follow from and inaccuracies common to all models are: the resulting concentration field. in this case, dis- • instead of navier-stokes equation an approx- solution rates on different points on the flat sur- imation is used as described before. The ap- face are different. 60 KRF•1 • OK.indd 60 15.12.2009 10:42:23 Matija Perne and Franci Gabrovšek, The problem of ril enkarren development: a model ing perspective proximation works well on average but on the The raindrop does not stay at the surface of the upper edge of the rock it does not; water film, as presumed for the calculations, but • steady-state approximation of the water flow is pushes off some of the old film. it is also possible used; that it does not stay at the site of impact, maybe it • the phenomena at raindrop impacts are not ac- bounces toward the centre of the rill, effectively counted for. Fresh water is added only at the increasing rainfall rate at the centre. This effect is water surface, eventual penetration of drops not taken into account either. so it would make into the film is neglected. sense to include raindrop impact and non-steady conditions during rillenkarren formation are state situation into future models of rillenkarren certainly not steady-state. The steady-state shape formation. of water film on 2 cm by 16 cm is calculated using experiments have shown that the size of drop- time propagation. it turns out that after three lets is not crucial for rillenkarren formation (see seconds of simulated water flow the film shape is chapter 4) but we know of no experiments on ril- very near the steady state, even if initial state is lenkarren formation without drops, that is under far from the steady one. We assume that state in fog or condensation. The importance of raindrop reality is nearly steady if a lot of drops fall on the impacts and non-steady-state effects might even rill in less than 3 s. if we take 1 mm3 as an average be easier to check experimentally than numeri- raindrop volume (elert, 2001), in the simulated cally. rainfall rate of 10 mm/h only 27 drops impact the rill in three seconds. 61 KRF•1 • OK.indd 61 15.12.2009 10:42:23 KRF•1 • OK.indd 62 15.12.2009 10:42:23 soMe MetHoDologies 6 on Karren researcH Gábor TÓTH The recognition of karren forms and research on of karstification, but which also depict the kar- them commenced at the end of the 19th century ren forms in detail (Bögli, 1980; trudgill, 1985; when karren did not represent an independent Jennings, 1985; White, 1988; Ford and Williams, group among the karst landforms. a. Favre was 1989). the first to mention the karren forms, calling them lapiés (Favre, 1867). This term became naturalized in many languages and is still used as the german Mapping methods expression Karren. a few years before Favre, sachs made experiments to produce karren forms in lab- as the karren forms are so extremely complex and oratory conditions by root corrosion, although he such small-sized features, the traditional mapping did not nominate them yet (sachs, 1865). at the end methods could not be adopted for their documen- of the century eckert investigated the evolution of tation. karren forms and the effect of vegetation (eckert, The first problem is to determine the best scale 1898). For a long period classification of karren ratios, because scales used on geomorphologic forms accounts for the mainstream of researches, maps are not suitable for the correct delineation of an outstanding researcher being J. cvijić (cvijić, karren forms. When measuring the single forms, 1924). The most comprehensive and recently used maps at the scale 1:5 and 1:10 can be chosen in system classifying karren forms was established by order to make a correct record, but in the case Bögli, who is considered to be the most significant of larger karren surfaces maps at scales 1:20, 1:50, authority on karren researches (Bögli, 1951, 1960a, and 1:100 values are preferred. These scales refer 1976, 1980). His studies on karren morphology are to the karren features of the temperate climate the most cited and his terminology was adopted in zone; for tropical karren forms different scales several languages. may be appropriate. two conference volumes are significant in the another important problem is the application present morphological literature: those edited by of traditional map-making instruments. regard- K. paterson and M. M. sweeting (1986), and by J. ing the tiny features, the method of contour line Fornós and a. ginés (1996), which contains the mapping can only be utilized with good results newest and most modern methods and informa- in very special situations. The extensively applied tion about karren forms. it is worth mentioning “grid method” is outlined in detail later in this five other studies dealing with the total process study. These difficulties mean that mapping of 63 KRF•1 • OK.indd 63 15.12.2009 10:42:23 Karst Rock Features • Karren Sculpturing karren forms cannot be done on traditional geo- squared paper which is calibrated in advance. By morphologic maps that depict larger geomorpho- linking up the resulting points the karren features logic areas. certainly, these maps are suitable for of the surface become easy to sketch. The distance the location of karren forms, e.g. the morphologi- of the net grid is basically determined by the size cal map of lapiés de tsanfleuron, which depicts of the mapped area as well as by the election of a the location of the karrenfields (schoeneich et al., useful scale. 1998). By using this procedure, maps on the scales of only the larger forms are suitable for mark- 1:10, 1:20 and 1:100 can be made applying 10, 20 ing on contour maps, so the application of the and 50 centimetre-distanced nets. These methods method is very limited. Veress and Barna tried to were applied by szunyogh, Veress and tóth espe- survey several well developed features when they cially in order to study the mountain karren fea- were mapping solution runnels (rinnenkarren) in tures (szunyogh et al., 1998; Veress et al., 1995). the totes gebirge in austria (Veress et al., 1995). The largest area mapped (20 x 25 metres) was sur- These features are 10‒15 metres long and their veyed near the Widerkar peak of totes gebirge widths are 0.5‒1 metre. using this method in the (Figure 2). delineation, the terrace grooves and microforms Karren morphological maps can be drawn by in the karren forms were shown precisely. square-net mapping of smaller surfaces, and the it seems more useful to choose mapping meth- history of dissolution of the area can be deduced ods that adapt better to the sizes of karren forms, from these maps (Veress and tóth, 2001). so making measure as precise as possible. one of these methods is square-grid mapping, which gives suitable recording for middle-sized Morphometrical methods (2 x 2 metres) and larger (20 x 25 metres) karren surfaces. The basic principle of this method is to Morphometry is widely used being easily adapt- cover the chosen area with a horizontal, suitably able to particular circumstances and conditions, meshed net, and then to determine the distance of and suitable for the examination both of single points of the forms in relation to the points of the forms and karren terrains. a further advantage of net (Figure 1). in the field these data are drawn on morphometrical methods is that they can be used Figure 1: Grid mapping in the Julian Alps. 64 KRF•1 • OK.indd 64 15.12.2009 10:42:25 Gábor Tóth, Some methodologies on karren research Figure 2: The largest mapped karrenfield area in Totes Gebirge. 65 KRF•1 • OK.indd 65 15.12.2009 10:42:41 Karst Rock Features • Karren Sculpturing Figure 3: Parts of a heel-print karren (trittkarren). 1. slope; 2. riser; 3. tread; 4. foreground; h : smallest height of riser; h : 1 2 greatest height of riser; s : greatest width of tread; s : greatest length of tread; f : width of foreground. w 1 w X maximum width horizontal width X M cross-sectional cross-sectional AREA large – ‘Al’ AREA small – ‘Ar’ D max Hor D mid Hor T Figure 4: Different parameters for meandering karren morphometry (after Hutchinson, 1996). X: field indentified channel rims; D mid Hor: horizontal mid-depth; Shape: Al/Ar; M: mid-point of channel width; D max Hor: horizontal maximum depth; form ratio: width/depth; T: channel lowest point/thalweg. in places which are difficult to approach, since it p. J. Vincent investigated the different param- requires only easily movable equipment. For data eters of heel-print karren (trittkarren) and re- collection it can be used for manual or instrumen- searched how they were interlinked. He discov- tal measures, which then are analysed by statisti- ered that the relationship between the character- cal methods. istic parameters of heel-print karren (e.g. the heel 66 KRF•1 • OK.indd 66 15.12.2009 10:42:41 Gábor Tóth, Some methodologies on karren research height, the bottom, etc.) was not accidental (Vin- cent, 1983a). Z. Balogh also studied trittkarren with a simi- lar method and measured several parameters (the height of the riser, the angle of the tread, the length of the tread, the width of the riser, the lenght of the arch of the riser and the width of the foreground (Figure 3) on slopes with different angles (10°, 20°, 25°, and 40°). analysing the data he revealed that different sized trittkarren represented the differ- ent phases of an evolution process (Balogh, 1998). l. rose and p. J. Vincent measured the widths of grikes (kluftkarren) in three different areas, and Figure 5: Estimation of specific karren-solution: it ex- then they depicted the results on bar charts based presses how many centimetres of karren landform are on frequency. They found that grikes of the three present on 1 metre of exposure of karrenfield. areas had evolved in two different periods, before and after the period of ice cover (rose and Vin- cent, 1986a). also in ril enkarren morphometry, it is feasible andering karren, applying morphometric meth- to collect data to define the diagnostic width for ods; they differentiated the types of meanders by such a microform (ginés, 1996b). Measurements detailed measuring of both sides of the bend (Ver- of the length of longer rillenkarren and altitude ess and tóth, 2004). above sea level correlate negatively at sites receiv- another basic investigation method for the ing more than 800 mm of rainfall in Mediterrane- morphogenetics of karren landforms is to make a an-climate karrenfields (a. ginés, 1990). transect on significant karren outcrops (Veress et in the case of meandering karren some pa- al., 2001a). The type and occurrence of the forms rameters originating from the meandering of the placed along a spread band-chain is determined, groove are worth defining, because they provide measuring its width, depth and direction and the most important information for the classifica- the slope angle and direction of the terrain. The tion and make possible the comparison of these lengths of transects are 15‒25 metres, depending microforms with other types of meanders. J. Zel- on the size of the karren outcrop. By analysing the ler was the first to measure the sinuosity of mean- resulting data several specific and global param- dering karren, the lengths of inflexed arches and eters can be estimated. in this manner the value the width of the belt. comparing these data with of specific karren solution can be determined by other data of fluvial meanders and glaciers he adding the width of the karren forms occurring found out that the sinuosity of meandering karren along the section and dividing this total karren is the largest. analysing the results he also dem- width by the length of the whole transect. so the onstrated that the wider the groove is, the larger specific karren solution shows how many centi- the size of the belt results (Zeller, 1967). metres of forms have significantly dissolved in 1 D. W. Hutchinson emphasized the shape of metre (Figures 5, 6). This value characterizes the the meandering grooves in his research and chose grade of development of the karrenfield. This cal- the parameters to be measured for this purpose culation can be applied to each distinct karren (Hutchinson, 1996). His survey criteria are shown form. an additional characteristic value is the in Figure 4. density of landforms. The density of forms can be M. Veress and g. tóth also examined the me- counted by dividing the number of occurrences 67 KRF•1 • OK.indd 67 15.12.2009 10:42:42 Karst Rock Features • Karren Sculpturing scribed in a circular diagram divided into 20° spacing with weighted average (Figure 7). The radius of the sector shows the occurrence of each form. For a better depiction of the data sets, the forms are grouped by their genetics. The first group is constituted by the runnels or grooves with different genetics, which evolved along the slope direction. The grikes were put into the sec- ond group, because they occur perpendicular to slope direction and are formed by the coalescence of vertical karren hol ows or are preformed by the direction of fissures. The third group contains the heel-print karren (trittkarren), solution pans (ka- menitzas) and karren wel s (Karrenröhren), which are basically circular forms. Their evolution is strongly influenced by the slope angle. By applying the data of different transects some conclusions can be drawn regarding how the top- ographic and stratigraphic position of the karst terrain (mainly the slope angle and the direction) influences the evolution of forms and which types of forms are preferred. another widespread and successfully applied method is the comparison of different param- eters of the same form. The basic principle of the method is to choose one or more characteristic Figure 6: Specific karren-solution in the various vegeta- parameters (width, depth, length) and to ana- tion zones per karren features. Showing the specific lyse their measurements under different circum- solutions, one close to each other, and measuring spe- stances (slope angle, exposure, precipitation). it is cific solutions in different vegetation zones, the rela- worth using this method to differentiate a karren tionship between the vegetation and karren forms can form into subtypes. in this case parameters must be deduced. 1. pine zone; 2. mountain pine zone; 3. without vegetation (1–3 Julian Alps, Totes Gebirge and be chosen according to the characteristics of the Dachstein); 4. mountain pine zone (Asiago plateau). forms, and the proportions are used to draw con- clusions. of each karren feature by the length of the whole Examination of dissolution process transect (table 1). in this way we will see how many individuals of the different forms are found These methods refer to the process of dissolu- in 1 metre. From this data we can conclude the tion, the rate of dissolution and the denudation frequency in which forms are increased by certain of the surface caused by the development of kar- factors (slope angle, precipitation, vegetation, ex- ren forms. The methods are classified into three posure). For instance, the increase of slope angle groups: is favourable for the development of wall karren. The first group determines the beginning of the The directions of karren features can be de- forms evolution and evaluates the rate of solution 68 KRF•1 • OK.indd 68 15.12.2009 10:42:42 Gábor Tóth, Some methodologies on karren research y 1.64 1.48 2.19 3.28 2.27 2.20 1.76 1.78 1.84 1.93 1.59 1.55 1.3 0.89 1.38 1.6 1.48 1.8 1.48 1.9 2.64 1.48 1.58 2.52 1.92 1.36 1.72 : Ju- densit (no/m) and Total ion etre; H 43.28 27.56 40.56 37.52 32.93 43.96 33.52 27.11 26.56 25.57 29.63 25.35 21.58 26.89 21.52 18.85 19.24 37.96 34.24 34.36 54.12 29.48 40.83 47.28 42.84 36.72 42.32 1 m s.s. (cm/m) - - - - - - - - - - - - - - - - - - - - - y en 0.29 0.08 0.38 0.18 0.41 0.05 arr en) densit (no/m) specific solut t k arr - - - - - - - - - - - - - - - - - - - - - 1.0 (Trittk 5.27 27.0 3.68 0.81 6.62 Heel-prin s.s. (cm/m) - - - - - - - - - - - - - y each karren feature on 0.12 0.25 0.07 0.12 0.11 0.16 0.04 0.04 0.09 0.04 0.25 0.24 0.24 0.12 es the values of densit (no/m) - - - - - - - - - - - - - 4.5 2.4 1 3.2 7.4 Kamenitzas 3.56 0.73 8.00 1.29 0.32 0.86 11.16 4.64 5.12 s.s. (cm/m) - 1 - y 0.44 0.4 0.25 2.4 0.86 1.55 1.64 0.67 1.04 1.0 1.09 1.15 0.77 0.97 1.4 0.08 0.4 0.08 0.22 0.12 0.48 0.2 0.4 0.72 en) densit (no/m) illen- and arr - - 5.4 5.2 1.8 r different terrains and giv Runnels and es (R 5.32 11.44 25.52 16.47 25.47 31.64 13.22 15.47 15.64 21.5 20.75 15.85 14.90 13.50 1.88 2.77 3.12 5.96 16.25 2.24 9.32 15.32 etre; density: occurrences of flut Rinnenk fou s.s. (cm/m) 1 m en - - - y 0.4 0.2 0.2 s on arr 0.36 0.25 0.32 0.47 0.08 0.04 0.24 0.29 0.04 0.06 0.15 0.04 0.36 0.31 0.24 0.04 0.04 0.16 0.28 0.36 0.08 densit (no/m) ells and k pipes - - - 6.88 8.88 4.5 6.00 2.78 0.8 5.33 5.0 0.68 2.75 7.22 4.35 0.8 4.44 2.67 3.18 2.12 0.8 0.37 1.48 6.8 12.6 1.08 en w 10.13 tains the data of Karr s.s. (cm/m) - - - - - - - - - - - - - - - 1 - - con the karren form y 0.18 0.72 0.64 1.04 1.56 0.68 0.08 0.28 0.4 en h of arr densit (no/m) dt - - - - - - - - - - - - - - - - - work k 0.81 11 30.6 1.4 3.96 4.28 s. The table Net 10.52 13.12 20.18 14.36 , total wi s.s. (cm/m) - - - ion y 0.2 0.4 0.2 0.2 0.4 en) 0.84 0.52 0.87 0.56 0.33 0.16 1.67 0.19 0.64 0.13 0.83 0.28 0.22 0.68 0.28 0.29 2.08 0.72 0.76 arr karren form tk densit (no/m) siago plateau. luf - - - 7.3 18 19 : A es (K 29.76 9.72 21.94 6.00 4.07 2.45 13.89 2.08 4.93 5.81 1.85 4.44 19.67 3.16 9.48 15.56 15.08 8.36 13.04 42.16 14.24 Grik s.s. (cm/m) 41 36 34 44 32 27 44 15 39 27 35 31 35 16 20 32 37 45 37 37 66 37 37 63 48 34 43 . S. s.: specific solut ms and density of achstein; A . of the : D no for ion o e 24 o 26 o o o o o o o o o o o o o o o o o o o o o o o 28 28 10 15 20 30 31 17 21 4 8 25 0 4 4 5 6 10 15 0 0 0 4 rge; D facSur ebi slope angle 1695 1715 1695 1776 1776 1800- 1900* 1800- 1900* 1900- 2000* 1900- 2000* 1900- 2000* 1630 1820 2051 2090 2098 1900- 2100* 2055 2055 2055 2055 2055 2055 2055 2061 2061 2061 2061 pa each karren landform altitude (m) lps; T: Totes G t Table 1: Specific karren-solut Number of the transec H I/2 H I//1 H I//3 H II//1 H II//2 T/ 4 T 5 T 3 T 2 T 1 D I/1 density of lian A D II/1 D III/1 H III/2 H III/1 H IV/1 A I/6 A I/2 A I/7 A I/1 A I/3 A I/5 A I/4 A II/2 A II/3 A II/4 A II/1 *altitude on m 69 KRF•1 • OK.indd 69 15.12.2009 10:42:42 Karst Rock Features • Karren Sculpturing Figure 7: Directional dispersion of the karren features along T 1 section (mountain pine zone, Totes Gebirge). The circular diagram shows the relationship between the formation of karren features, the slope angle and the strike directions. 1. runnels; 2. grikes (kluftkarren); 3. slope direction with angle of gradient; 4. strike direction; 5. density of fractures (occurrences/10 centimetres). by measuring the sizes of the forms. This process researchers consider the beginning of evolution gives average values, and does not take into con- of forms started at the regression of the ice cover, sideration that intensity of dissolution might have supposing that evolution began immediately. The changed. in the case of alpine karrenfields most first researcher to be mentioned is a. Bögli, who 70 KRF•1 • OK.indd 70 15.12.2009 10:42:42 Gábor Tóth, Some methodologies on karren research determined the rate of solutional erosion by meas- ence of the soil on the processes of karstification. uring the heights of karren tables (Karrentische; Many researchers examine the chemical compo- Bögli, 1961). The heights of karren tables were sition of the soil. in the case of a karstic surface 10‒15 centimetres, so ‒ if the recession of the ice covered by soil the most frequent method is to put cover is considered to have happened ten thou- limestone tablets of known masses into the soil; sand years ago ‒ the velocity of erosion was 10‒15 then after a period of time their masses are again millimetres per one thousand years. measured. The reduction of the tablet mass shows D. sellier used a similar method on granite sur- the measure of dissolution (gams, 1985; trudgill, faces measuring the sizes of forms evolved on the 1975, 1985; Kashima and urushibara-Yoshino, megaliths in Brittany supposing that the megaliths 1996). trudgill found out that in case of high con- had been erected 5,000 years ago (sellier, 1997). tent of heavy metals and small value of limestone The second group is the collection of applica- content the intensity of solution is larger. tions which measures directly the velocity of solu- to summarize the researches, the most doubt- tion and erosion by measuring the speed of ero- ful fields of karren research are the velocity of so- sion compared to a fixed height given in a certain lution and the growth rate of karren features. The point of time. F. cucchi put metallic points in the main reason for this is that the process of solution rock and measured the extent of the surface’s deg- and its velocity are influenced by several environ- radation compared to the original surface (cuc- mental conditions. The amount of precipitation, chi et al., 1996). The same method was applied the quality of rock, the vegetation, the soil and by a French expedition on the island of Diego de the temperature could strongly modify the value almagro in chile. They had painted signs using of local velocity of the dissolution process. after waterproof painting 50 years before and the sur- a survey on the available data, most of the values face protected with the painting proved to be 3 are 0.4‒10 millimetres per a thousand years, with millimetres higher than the surrounding surface 1,500‒2,500 millimetres of annual precipitation. (Hobléa et al., 2001). The third group represents other chemical methods for investigations on dissolution proc- Examination of topographic and esses. several scientists have tried to determine lithological conditions the solute content of karstwater at different lo- cations for measuring of the erosion (sweeting, as several features of limestone rocks could in- 1966; newson, 1970; Thomas, 1970). in order to fluence the process of dissolution, examination apply this method it is essential that the catch- of lithological condition is widely represented in ment area of a certain spring should be correctly the literature. This brief summary describes the delimited. F. Hobléa et al. measured the solute methods linked directly to karst morphology. content of solution pans on the island of Diego roughness is a factor which can strongly in- de almagro and determined the beginning of dis- fluence the spatial expansion of solution process, solution process at ten thousand years Bp which and it has been investigated by several scientists. was the time of the ice regression. in this way the J. Moses et al. applied the electron microscope rate of erosion was measured as 95 millimetres per to examine roughness (Moses and Viles, 1996). J. thousand years (Hobléa et al., 2001). considering crowther applied a manual method: he put a car- the precipitation of 8,000 millimetres per year on penter gauge vertically on the surface of the rock, the island, the survey data are in good agreement then took photos of the profiles obtained; the pho- with the dissolution of the temperate climate zone. tos were then digitalized and the height difference examinations of solutional erosion are often between consecutive points was applied to deter- applied on covered surfaces to analyse the influ- mine the measure of roughness (crowther, 1996). 71 KRF•1 • OK.indd 71 15.12.2009 10:42:42 Karst Rock Features • Karren Sculpturing The relation between the development of karren grain size of the rock substrate by analysing the forms and the slope angle is essential as longitu- evolution of ril enkarren and found that high cal- dinal forms evolve down slope; the slope angle of cite concentration and fine granulation helped to the karren outcrops determines the drift speed of evolve the rillenkarren (Vincent, 1996). the solvent. one of these investigations have been done by J. r. glew and D. c. Ford, who dripped solvent on gypsum surfaces with different angles Theoretical method of karren of slope (35°, 45°, 55°) and found that the increase development of slope angle increased the lengths of rillenkar- ren. Then they dripped solvent on different rock a theoretical-physical study of the process of kar- surfaces and realized that wider rillenkarren had ren development has been developed recently. it evolved on limestone surfaces rather than on salt is based on the equation system of the karstifica- or gypsum surfaces (glew and Ford, 1980). D. n. tion of a sloping limestone terrain, considering Mottershead divided the slope into 10° intervals the hydrodynamic, chemical and morphological and measured the frequency of rillenkarren. His rules of the karstification processes (szunyogh, results show that their number is the largest on 2000a). The main magnitude was a three-variable the slope of 60‒70° (Mottershead, 1996b). function, which can describe the form of rock sur- The types, sizes and frequency of karren fea- face in time. The determination of this form also tures are greatly influenced by the quality of the postulates the calculation of the speed of flowing rock. on porous or large grained limestone the water over limestone surfaces, the concentration dissolution is less. The evolution of kluftkarren is of solution caco in water and the thickness of 3 enhanced by a strong network of fissures on the liquid film. it is done as a computer algorithm for rock. p. Vincent examined the calcite content and the solution of equations (szunyogh, 2000b). 72 KRF•1 • OK.indd 72 15.12.2009 10:42:42 Microrills 7 Lluís GÓMEZ-PUJOL and Joan J. FORNÓS Microril s are the smallest form of the exokarstic ments (ginés, 1999a), although they are present linear features on limestone and gypsum rocks. on localities with rain precipitation greater than They are typically described as about one millime- 800 mm per year (Ford and lundberg, 1987). The tre wide rills, round bottomed and packed togeth- microrills distribution could be broadly related to er with characteristic tightly sinuous to anastomo- lithological control and to genetic agents such as sing plan view patterns on gentle slopes, becom- dew and sea spray water inputs. capillary flow is ing more parallel and straighter with increasing believed to explain much of their characteristic slope (Ford and lundberg, 1987). sinuosity (Ford and Williams, 2007). They appear widespread, from mountain to su- The aim of the present paper is to contribute to pralittoral domain in rock coasts, and from arid the understanding of the origin and evolution of to temperate environments; and rarely in cold microrills through an examination and descrip- environments (Davis, 1957). Microrills are found tion of the form of features, rock texture and lithol- in coastal environments, both macro- and mic- ogy properties from samples collected in Balearic rotidal, such as Vancouver island, canada, or the islands both in coastal and mountain sites. Welsh coast, great Britain (Ford and lundberg, 1987), on the northern palawan coast, philippines (longman and Brownlee, 1980), Dalmatian coast, Study sites and methods croatia (perica et al., 2004), sicilia (Macaluso and sauro, 1996a), and also at the Balearic islands The island of Menorca together with Mallorca (the (ginés, 1993; gómez-pujol and Fornós, 2001, greatest of the Balearic archipelago) are charac- 2004a). They also have been reported in arid and terized by diversified and abundant karstic geo- semi-arid continental environments such as the morphology. The karstic features, both the subter- chillagoe and gregory karsts, australia (Jennings, ranean (ginés J., 1995; ginés and Fornós, 2004) 1981; Dunkerley, 1983; see chapter 31, tropical and the surficial ones (ginés and ginés, 1995), monsoon karren in australia), the Mohave and constitute one of the most characteristic aspects colorado deserts, southern california (lauder- of these island's geographical context. in particu- milk and Woodford, 1932), pakistan (cílek, 1989), lar, extensive areas of Mallorca, especially in the and from southeast Morocco (smith, 1986, 1988). main range (serra de tramuntana), and of both, These micromorphological features are believed Mallorca and Menorca coastal sites, present an ex- to be representative of arid or semi-arid environ- ceptional range of exokarstic morphologies. Kar- 73 KRF•1 • OK.indd 73 15.12.2009 10:42:42 Karst Rock Features • Karren Sculpturing renfields cover large expanses of limestone expo- along each microrill-path were obtained when sures lacking a soil covering. in the case of moun- the feature limits were clear enough. results of tain environments the microrills analysed in this the measurements are presented as mean width paper appear in combination with ril enkarren values for microrills in each study location; the forms that evolve to meandering decantation mean width of the thinnest individual microrill flutes, as well as with kluftkarren and some char- and the mean width of the widest microrill were acteristic smooth surfaces. at coastal sites mi- also reported. a qualitative seM and resin cast crorills appear between a complex assemblage of study (Moses et al., 1995; taylor and Viles, 2000; mesokarren formed by the combination of pinna- Viles, 2001) was done on pieces of 10×10×10 mm cles and basin pools. The rocks involved in these in order to, first, identify the abundance of na- karrenfields are mainly formed of lower liassic nomorphologies (Viles and Moses, 1998), secondly massive limestones and breccias, as well as con- to conclude which kind of processes (chemical, glomerates and calcareous breccias from the Bur- physical or biological) are operating on microrill digalian. reefal limestones, calcarenites and asso- formation, and thirdly, to elucidate the relation of ciated sediments, tortonian and Messinian in age, the microrill plan view orientation pattern with are the characteristic rocks in both Mallorca and the joint and rock structure alignments. Menorca coastal sites. samples of limestone with well-developed mi- crorills were hand-cut using a chisel at different Microrills rock type, plan view and coastal environments of Mallorca and Menorca morphometry as well as in mountain landscapes of the Mallorca northern range (table 1). From each limestone systematic data on microrill width, length and rock, fragments have been prepared for thin sec- plan view pattern are scarce in literature. since tion study, optical microscope observation and laudermilk and Woodford (1932) there is an seM exploration. Thin sections allowed analysis agreement in distinguishing four types of micro- of rock texture properties, sorting and grain-size rills according to plan view and cross section. This mean diameter. photos of optical microscope classification (see discussion in grimes, 2007) samples were used to assess microrill width and separates four main types that range from par- geometry. Microrill width was obtained by means allel and moderately sinuous and rather shallow of a digital image processing standard software rills, to less tightly packed and shallower rills with measuring the distance between the crests that smooth crests and a polished appearance. Ford delimits laterally the microrill. several measures and lundberg (1987) suggest that the develop- Table 1: Sampling sites and main environmental parameters (climate data from Guijarro, 1986). Ma. Mallorca; Me. Menorca. Island Locations Altitude Precipitation Temperature Environment Rock type (m) (mm) (°C) Ma Puig Major (PM) 1,300 m 1,246.6 9.1 mountain Lower Jurassic massive limestone Ma Cala Sant Vicenç (CB) > 2 m 731.1 16.5 coastal Lower Miocene carbonate breccia Ma Punta de Tacàritx (PT) > 2 m 580.8 16.2 coastal Quaternary carbonate breccia Upper Miocene Ma Cala Màrmols (CM) > 3 m 321.3 17.6 coastal biocalcarenite Upper Miocene Me Punta Prima (PP) >2 m 444.9 17.1 coastal biocalcarenite Upper Miocene Me Cap d’en Font (CF) > 4 m 545.4 16.6 coastal biocalcarenite 74 KRF•1 • OK.indd 74 15.12.2009 10:42:43 Lluís Gómez-Pujol and Joan J. Fornós, Microril s ment and the plan pattern of microrills are conse- rock types described, where microrills have been quences of rock type and texture and the differing identified at Balearic islands, are fine-grained and amount of effective liquid. homogenous in grain size. texture control on in the Balearic islands microrills appear patch- microrill development is shown by their different ily on mountain and coastal limestone outcrops. presence according to textural variation between on the mountains, where the Jurassic material conglomerate clasts and matrix. in thin section forms most of the structured terrains, microrills analysis and seM observations the grain size are related to lower Jurassic massive micrite lime- ranges from maximum mean diameters of 8.65 stones with stratified levels with bioclastic and μm to minimum mean diameters of 3.23 μm and oolites components (Fornós and gelabert, 1995). all the samples are well sorted. The absolute values at coastal sites microrills sculpture a widespread of grain size ranges from 2.305 μm to 17.43 μm. variety of rock exposures that comprise from This fact shows that microrills are developed only neogene to Quaternary limestones and calcaren- on a mudstone texture. ites. This kind of feature has also been identified Microrills appear on rock surfaces where veg- in lower Miocene conglomerates, namely in the etation and biofilm colonization are not present sant elm fm (rodríguez-perea, 1984). Most of the and on rocks cropping out with ancient rounded pebbles which compose these carbonate Miocene subsoil karren, which have been recently exposed. conglomerates are lower Jurassic. Microrills also The orientation of these surfaces is generally hori- appear at coastal sites in upper Miocene rocks. zontal or near horizontal. although some of the The upper Miocene is composed by massive reefal rocks are rough in a mm or lower scale, at cm- limestones, calcarenites and calcisiltites that form scale microrills do not develop on vertical or respectively, the reefal unit and the terminal rough surfaces. complex (Fornós and pomar, 1983). Finally, mi- There are no differences in microrills plan view crorills also are developed on Quaternary brec- pattern between mountain and coastal study sites cias. These Quaternary deposits correspond to despite the differences in rain supply and litholo- aeolianite matrix-supported clasts related to the gy. straight and sinuous or meander-like rills can mixing of dune episodes with alluvial and collu- be seen on different rocks (Figure 1). at cala sant vial events. The clasts resulted from Jurassic, cre- Vicenç (Figure 1g, h) microrills appear developed taceous and lower Miocene denudation materials on Miocene breccia with Miocene and Jurassic and are composed of micritic limestones. all the matrix-supported clasts. These fine-grained rocks Table 2: Microril s morphometrical assessment. Sample Site Mean width SD Min Max N (mm) (mm) (mm) CM a Cala Màrmols, Ma 0.869 0.267 0.453 1.375 57 CM b Cala Màrmols, Ma 0.728 0.207 0.412 1.050 59 CM c Cala Màrmols, Ma 0.758 0.181 0.509 1.163 63 CB a Cala Sant Vicenç, Ma 0.747 0.138 0.547 1.016 55 CB b Cala Sant Vicenç, Ma 0.539 0.115 0.269 0.781 48 CB c Cala Sant Vicenç, Ma 0.760 0.198 0.514 1.128 49 PT a Punta de Tacàritx, Ma 0.440 0.106 0.256 0.606 61 PT b Punta de Tacàritx, Ma 0.611 0.141 0.446 0.926 52 PT c Punta de Tacàrtix, Ma 0.997 0.233 0.632 1.434 46 PM a Puig Major, Ma 0.648 0.112 0.391 0.918 32 PP a Punta Prima, Me 0.662 0.152 0.503 1.033 48 CF a Cap d’en Font, Me 0.470 0.096 0.307 0.644 57 75 KRF•1 • OK.indd 75 15.12.2009 10:42:43 Karst Rock Features • Karren Sculpturing a, b c, d e, f g, h Figure 1: Microril photographs from mountain and coastal sites at Balearic islands study sites: a. microril linear-like forms on Lower Miocene limestones at Cala Sant Vicenç, NE-Mal orca; b. plan view of microril features showing meandering-like forms on Lower Miocene limestones at Cala Sant Vicenç, NE-Mal orca; c. Cap d’en Font, Menor- ca. Note the blue-green algae and fungi colonization at microril bottoms; d. microril s developed on micritic Jurassic limestones at Puig Major, Mal orca. The microril s, path is normal to the main microfractures and  76 KRF•1 • OK.indd 76 15.12.2009 10:42:45 Lluís Gómez-Pujol and Joan J. Fornós, Microril s      ­ Figure 2a: Microril textural and morphological properties. Relationship between microril width and rock grain size. present a wide array of orientations, outcrop expo- outcrops (Figure 1d) it seems that meandering sition surfaces and slopes and do not show a clear forms correspond to fine-grained textures and relation with the plan pattern or the length of the straight forms to coarser textures. nevertheless, microrills. The same type of clast can have mean- there are many cases where sinuous and straight dering microrills and straight microrills in several microrills share wall partitions or where the same cases. The same situation can be identified at the microrills follow both configurations changing punta de tacàritx (northern Mallorca) coastal site, from sinuous to straight path. The plan view of where rock outcrop is built up by Quaternary car- microrills uses to be parallel or moderately sinu- bonate aeolianites that support Jurassic limestone ous but there is not any evidence at any study lo- clasts. There are many cases where on the same cality of a dendritic or truly branched organiza- clast microrills follow both meander and straight tion of the microrills. paths (Figure 1e). in the case of the lower and The dimensions of the microrills are quite vari- upper Miocene limestone outcrops (Figure 1a, b, able in length and more or less regular in width. c, f) and also in lower Jurassic massive limestone it has not been possible to measure accurately mi-  joint directions in the rock; e. meandering and sinuous microril s, plan view on a Jurassic clast at Punta de Tacàritx, Mallorca; f. microril s developed in a rough surface at Cala Màrmols, Mallorca; g. sinuous and rectilinear mi- croril s at Cala Sant Vicenç developed on clasts from a carbonate matrix supported breccia Lower Miocene in age. Note that there are some pits that begin to modify the microril s plan view; h. detail of microril distribution at Cala Sant Vicenç rock outcrops where, differences in breccia clast and matrix texture, demonstrate the textural control on microril development. 77 KRF•1 • OK.indd 77 15.12.2009 10:42:45 Karst Rock Features • Karren Sculpturing       Figure 2b: Histograms of microril width. crorill depth, although this seems to be close to clear relationship between the age of rocks and the the width values or slightly lower. Microrill width microrills width. Figure 2 illustrates the microrills measured on different rock textures in mountain mean width values and width distribution against and coastal environments from Mallorca and the mean rock grain size. The characteristic micro- Menorca ranges from a minimum width of 0.26 to rill width frequency distribution shows a slightly a maximum of 1.43 mm (table 2). The mean width skewed typical peak-pointed shape (Figure 2b). value for the whole microrills present at the dif- The scatter graph (Figure 2a) points up two main ferent sampled surfaces ranges from 0.44 to 1.00 relationships. The first one is that microrills are de- mm. it is important to note that the mean width veloped only on mudstones or fine-grained rocks. standard deviation is not bigger than 0.10 to 0.27 The second is that microrill width is not depend- mm, which shows that the homogeneity in width ent on the rock grain size (r2 < 0.01; p < 0.001). if is the essential parameter of this feature regard- the data on microrills developed on upper Mi- less of environments and rock types. There is not a ocene rock outcrops are assessed alone, then there 78 KRF•1 • OK.indd 78 15.12.2009 10:42:45 Lluís Gómez-Pujol and Joan J. Fornós, Microril s Figure 3: Microril s plan view. They are not true linear forms and can be considered as an alignment of micro depres- sions. The graph shows the ridge crests (dotted), the direction of slopes (arrows) and the concave bottoms (circles). is a weak negative correlation between microrills 2007). sample observations show that there are width and rock grain size (r2 = 0.62; p < 0.001). many features closed by well-defined partitions, and other ones that connect with neighbouring features although there are secondary wall parti- Optical microscope and sEM observations tions, or soft chains that separate opposite slopes. seM observations contribute to previous informa- Binocular microscope observations show that tion. Microrills are not true channels but a set of conventional features that we use to describe micro depressions joined or sharing walls aligned microrills are not real lineal elements (Figure 3). according to a rock texture pattern. Thus, in some slope orientations are not organized as a logical or cases chains and depression walls correspond to classical flow feature. Microrills that to the naked coarser grains whereas microrills bottoms over- eye seem one individual lineal feature are, in fact, lie finer grains (Figure 4). seM micrographs and an array of elongated or near circular depres- resin casts reveal that microrills are not aligned sions that share wall partitions. They can be com- according to joints and micro fractures present in pared, just from a morphological point of view, to the rock, some of them clearly cross or have their an elongated polygonal karst (Ford and Williams, 79 KRF•1 • OK.indd 79 15.12.2009 10:42:46 Karst Rock Features • Karren Sculpturing a b c d Figure 4: Microril s SEM photographs: a. general overview of microril s and fungal biopitting. Note the microril s crossing the joints filled by calcite crystals; b. detail of the array of elongated or near circular depressions that share wal s inside the gross morphology of microril s; c. and d. in some cases microril s chains and depressions wal s cor- respond to coarser rock grains. major axis orientation perpendicular to structural ure 5). There are few samples in which evidences alignments (Figures 4, 7a, b). of biological weathering nanomorphologies such scanning electron micrographs also show that as circular etched pits, etched tunnels and filament grain boundary widening, V-in-V etching and shaped trenches are present, and these are mainly blocky etching are very abundant nanomorpholo- at microrill bottoms. nevertheless, lichen fructif-gies. all these nanomorphologies, according to erous bodies, fungi and blue-green algae pits ap- Viles and Moses (1998), are thought to be related pear in those samples that have been exposed for to dissolution and crystallographic control on dis- longer times and in more humid locations (Fig- solution in its morphological expression. in this ure 6); although they have no relation with the de- way, the general overview is that rock grains, be- velopment of the microrills. all these features are cause of solutional attack, seem to be floating (Fig- clearly superimposed over the solution and crys- 80 KRF•1 • OK.indd 80 15.12.2009 10:42:47 Lluís Gómez-Pujol and Joan J. Fornós, Microril s a b c d Figure 5: SEM micrographs showing the grain size and good sorting at different rock exposures where microril s are developed. Note that rock grains seem to be floating, they are poorly cemented. tallographic controlled nanomorphologies, and are cases in which depressions are isolated. in the gross morphology of the microrills. some cases a second order of small pits appears resin cast observations support the informa- in the bottom of microrills and those relates to in- tion from the previous techniques. similarities cipient colonization of fungi or blue-green algae are found from different samples and study loca- superimposed on the gross morphology of micro- tions. in the inverse resin cast image of microrills rills (Figure 7a, c). the walls appear as a set of depressions and they resemble the geometry of intestines (Figure 7) in which bottom surfaces are smooth and crests Some genetic considerations slightly sharp. The plan view of resin casts rein- forces the seM observations of the non-linear From our observations in the Balearic islands, nature of microrills. it could be seen how depres- it would seem that microrills do not show truly sions are attached one to the other, although there linear microkarren morphologies. among other 81 KRF•1 • OK.indd 81 15.12.2009 10:42:48 Karst Rock Features • Karren Sculpturing a b c d Figure 6: Nanomorphologies identified in microril wal s and bottoms: a. and c. detailed observations of rock grains where V in V and crystal limits widening a crystal growth controlled nanomorphologies are the most abundant features. It can be appreciated a second set of nanomorphologies, although less important, as micropits and circu- lar etchings related to blue-green or fungal biological action; b. and d. lichen and fungi fructiferous bodies and as- sociated pits appear in those samples related to longer time exposed surfaces and more humid positions. factors, rock texture is the key element for their variables can better explain the disposition and development. rock texture must be fine enough organization of microrills. Detailed observations (mudstone) to develop microrills. The mean grain show that the channels are not simple channels, size that we have measured ranges from 3.23 μm but rather chains of depressions, and that mean- to 8.65 μm (Figure 2a). The fact that the set of der-like forms are related to the size of rock grains aligned or joined depressions do not follow frac- and to the plan view arrangement of depressions. tures and joint directions enhances the idea that Microrills appear in a wide set of environments, microrill genesis is related to the textural nature with mean annual precipitation ranging from of the rock rather than its structure (Vacher and greater than 800 mm down to less than 400 mm. Mylroie, 2002; taboroši et al., 2004). all these nevertheless, solution, according to seM observa- 82 KRF•1 • OK.indd 82 15.12.2009 10:42:49 Lluís Gómez-Pujol and Joan J. Fornós, Microril s a b c d Figure 7: Resin cast observations show that microril s do not follow the microstructural and joint alignments of rock (a and b) and that the gross morphology of microril s is of rounded bottoms (a and c) and sharp wal s (b and d), as wel as the different depressions that appear inside a microril (d) In some cases it can be appreciated smal depres- sions superimposed to microril paths that seem related to biological weathering agents (a and c). tions, is the key process operating on the develop- these features. in some way the evolution in two ment of this feature. The absence of real channel stages of this micromorphology has resemblances forms exclude runoff as the gross agent for micro- to those described for rillenkarren by Fiol et al. rill enlargement. For this reason thin water films (1996). The first mechanism by which the rock derived from dew or sea spray and driven by cap- grains are weakened corresponds to dew solution, illary forces seem to be the mechanism by which and in a second stage rain or water film runoff microrills are formed and then enlarged. grains detaches the grain and mechanically enhances detached by solution are pushed out of the de- the morphology. The biological contribution is a pressions in a mechanical way by runoff and rain, secondary agent that can affect the microrill de- and this fact is enhanced by the shallow nature of velopment, helping us to understand the temporal 83 KRF•1 • OK.indd 83 15.12.2009 10:42:50 Karst Rock Features • Karren Sculpturing framework of the microrill's genetic model. Mi- addressed on the Balearic islands microrill char- crorills never appear on highly-dissected relief acterization it is highlighted that microrills are a zones (i.e. pinnacle zone at coastal karren, or well more complex form than was initially thought. a developed rillenkarren surfaces). However, it is worldwide field study exploring all the features de- easy to find them landward from coastal sites and scribed would be advisable to get a better knowl- on round surfaces recently exposed in mountain edge of the specific role of rock texture and nature, or inland environments. according to observa- as well as how dew or water films are operating tion from longer exposed samples, as biological jointly with runoff in the shaping of microrills. action increases then microrill formations disap- pear and evolve to biokarstic features. Thus, ac- cording to the model presented by a. ginés (1995, Acknowledgements 1996a), karren assemblages can be understood or observed from an ecological point of view in The authors are indebted to staff of the electron which they are interpreted as a succession of mor- Microscope unit and to the laboratory staff of the phologies. in this way, microrills would be the Department of earth sciences at Balearic islands first step in this karren forms colonization or de- university. We are grateful to Dr. angel ginés and velopment of recent exposed limestone rocks to Dr. Ken g. grimes for their helpful comments the weathering agents. and suggestions. This study is a contribution to The literature on microrills is scarce, and most the investigation project cgl2006‒11242‒c03‒01 of the works are concerned with descriptive ap- Bte from the Ministerio de educación y ciencia- proaches. There is not much work that links FeDer. lgp is indebted to the “consejo superior quantitative data on microrills dimensions and de investigaciones científicas” (csic) for the observations at different scales to rock properties funding provided in the Jae-Doc program. and the active weathering processes. From data 84 KRF•1 • OK.indd 84 15.12.2009 10:42:50 cavernous WeatHering 8 Andrew S. GOUDIE in common with sandstones, granites and many ever, as smith and Mcalister (1986) have rightly other rock types, the surfaces of limestones and pointed out: dolomites may frequently be pitted with a range of “There has been a tendency in field studies of cavernous weathering forms of different sizes. The salt weathering for either dangerous circular ar- smaller features (a few cm in size) are commonly guments or a form of ‘guilt by association’. in the called alveoles or honeycombs, whereas the larger first case, the presence of certain landforms in an features (which may be some metres in size) tend area is used to infer the activity of salt weather- to be known as tafoni. Both alveoles and tafoni are ing mechanisms, without incontrovertible evi- common in coastal environments, and tafoni ap- dence that the landforms derive solely from these pear to be especially common in drylands. tafo- mechanisms. in particular, tafoni, which are com- ni developed on limestones have been described monly attributed to salt weathering, may…be con- from Malta (Hunt, 1996), Bahrain (Doornkamp vergent forms produced by different mechanisms, et al., 1980), and north africa (smith, 1978, 1986), of which salt weathering is only one possibility. and are widely developed on the limestones of the Whereas, in the second case it is assumed that be- oman mountains. They are also reported on lime- cause certain salts are abundant in an area, and stone buildings. However, most detailed studies of because rocks are being weathered, then those cavernous weathering forms and the processes salts must be responsible for the weathering.” that form them, have been conducted on non-car- The role of salt may cause chemical weather- bonate rocks (goudie and Viles, 1997). ing (Young, 1987), volume changes in clay min- The apparent preferential development of tafoni erals (pye and Mottershead, 1995) as well as the and alveoles in coastal and dryland environments physical mechanisms of crystal growth, crystal gives some prima facie support to the idea that hydration and volumetric expansion. a wide salt and/or wind abrasion may be involved in their range of salt minerals have been found associated development. indeed, salt weathering has often with tafoni, including halite, gypsum and vari- been seen as associated with the development of ous forms of magnesium sulphate (rögner, 1986; cavernous weathering forms (Martini, 1978) and Mustoe, 1983; Bradley et al., 1978). experimental simulations have shown that lime- preferential development of alveoles has been stones are very susceptible to salt attack (goudie, noted in windy situations (cardell et al., 2003), 1974, 1999) or to attack by a combination of salt and the simulation study of rodríguez-navarro and wind (rodríguez-navarro et al., 1999). How- et al. (1999) suggested that wind and salt could 85 KRF•1 • OK.indd 85 15.12.2009 10:42:50 Karst Rock Features • Karren Sculpturing work in tandem to produce honeycomb weather- (1996) has argued that Maltese tafoni are dissolu- ing forms. as they wrote: “incipient honeycomb tion phenomena that were initiated underground weathering in a homogeneous limestone has been (e.g. under a mantle of soil), subsequently ex- experimentally reproduced by wind exposure and humed, and then enlarged by subaerial weather- salt crystallisation. our experiments show that ing processes. heterogeneous wind flow over a stone surface is some scientists have argued that the develop- important in the development of small, randomly ment of cavernous weathering forms requires distributed cavities. a reduction in air pressure the presence of rocks with case-hardened exteri- within the cavities results in increased wind speed ors and weakened interiors. once the outer skin and rapid evaporation. a high evaporation rate is breached, it is argued, weathering and erosive and evaporative cooling of the saline solution in mechanisms can then hollow out the rock’s interi- the cavity leads to more rapid and greater granu- or (Winkler, 1979). However, whether case-hard- lar disintegration than in the surrounding areas. ening is a requisite has been challenged by Brad- it seems that this local supersaturation and sub- ley et al. (1978), conca and rossman (1985) and sequent build-up of salt crystallisation pressure smith and Mcalister (1986). on the other hand it ultimately result in the formation of honeycomb is remarkable how many tafoni do appear to occur features.” in rocks that have been case-hardened (goudie et some cavernous weathering forms may be the al., 2002). result of seepage erosion and arise from concen- another possible reason, apart from breach- trated weathering related to greater moisture ing of a case-hardened carapace, why cavernous availability at ground level (smith, 1978). Hunt forms may develop, is a positive feedback effect (smith and Mcalister, 1986): “… once a hollow is initiated it creates an environment in which weathering is favoured, weathering in turn ex- tends the hollow to produce an optimum form in which weathering is further enhanced, and so on. such re-enforcement cannot continue indefinitely and a condition must eventually be reached where the weathering rate is reduced. This could occur, for example, where a cavern becomes so deep that it either prevents the ingress of moist air, or rock temperature variations are reduced to the point where precipitation and evaporation no longer occur.” They argue that on exposed cliff faces salts would be deposited by the outward migration of salts derived from within the rock, but that they would be removed by subsequent rainwash, whereas salts precipitated in hollows would be protected from such leaching and so could cause salt weathering to occur, which would progres- sively expand the hollow. it is probable that cavernous weathering forms Figure 1: Tafoni. such as tafoni are the result of a combination of 86 KRF•1 • OK.indd 86 15.12.2009 10:42:50 Andrew S. Goudie, Cavernous weathering processes, as indicated in the model presented larger cavernous weathering forms are more (goudie and Viles, 1997), Figure 1. than just morphological curiosities. in time they There is very little hard and fast data on the lead to the undercutting of slopes, so that slope rates at which tafoni may develop in limestones. failure occurs. Hume (1925) argued that the lime- an experiment undertaken in Bahrain in a silt- stone cliffs of the eocene Ma’aza limestone plateau stone member of the al Buhayr carbonate forma- in upper egypt has been so oversteepened by ex- tion indicated that back wall retreat and rock flour treme cavernous weathering that they had become production were rapid (goudie and Viles, 1997). “absolutely unscaleable”. 87 KRF•1 • OK.indd 87 15.12.2009 10:42:50 KRF•1 • OK.indd 88 15.12.2009 10:42:50 KluFtKarren or griKes as FunDaMental 9 Karstic pHenoMena Helen S. GOLDIE Kluftkarren are fissures in karstifiable rocks, espe- as they are also known, is their distinction from cially limestone or dolomite, that have been wid- many, solutional etching, karren features since ened by weathering and erosion processes, largely they are linear features determined by rock struc- corrosional. They are fundamental to karst, as a tures such as joints and veins. Kluftkarren are limestone without fissures of some sort is karstifi- thus not so exclusively a solutional feature as are able to only a limited extent. Definitions of karst some karren types, and fissure formation can in- testify to and support this notion. For example, a deed produce a visible topographic feature with- glacially scoured limestone surface without any out any solution having yet taken place, for exam- remaining fissures has essentially had all of its ple, where pressure release caused by quarrying karstic characteristics removed. limestones that has occurred. are unfissured may develop surface solution pools Ford and Williams (2007) define kluftkarren for example, but without openings into the rock as “the master karren features in most karren as- there will be no notion of drainage going under- semblages. They are the principal drains, either to ground, so the fundamentally important third di- the deeper epikarst, or to dolines or to surface dis- mension of karst will not develop. charge such as river channels.” They identify their relationships to major joint sets or systems and to bedding planes. some kluftkarren or grikes cut Literature: definitions, dimensions through only one rock layer, others may extend and distribution through two or more. These latter will receive drainage from shallower fissures and therefore de- Kluftkarren can be discussed in various terms in- velop dominance of the drainage systems. goldie cluding scale, degree of development, and their re- (1976) defined grikes as features cutting through lationships to other karst features, including other at least one bed of rock. karren ( lapiés) features, particularly clints, or it can be seen that grikes are absolutely funda- Flachkarren. Development can be examined from mental to karst landform and landscape develop- both ends of the spectrum, that is, from where the ment, and without them karst would be restricted. features begin, to where they are evolving. There indeed, karstifiable rocks must be mechanically has been much work on karren definition and de- competent and not with too few fissures, nor too scription (Fornós and ginés, 1996) and an impor- many, nor a dominance of high porosity, since tant point as regards kluftkarren, grikes, or cutters, these properties prevent true karstic landform 89 KRF•1 • OK.indd 89 15.12.2009 10:42:51 Karst Rock Features • Karren Sculpturing suites from developing. The chalk of southern with the form of the intervening rock, composing england is one limestone where the nature of the flachkarren, clints, pinnacles or other remnant fissures and the porosity restrict karstification, un- solid forms. like in the harder, more mechanically competent Much research has material on or reference to and less-fissured limestones of the carboniferous grikes (the term to be used here), and it would in northern england and elsewhere in the British be impossible to summarize all but the main isles, or the cretaceous limestones of the alps, for ideas that more specifically focus on these fea- example. some of the difference in karstifiability tures. Factors affecting grike development must is because the joint or fissure spacing in the chalk be outlined and it is worth commenting on their is quite small, reflecting bed thickness. Bed thick- distribution, which is world wide in any fissured nesses of 10 to 30 cm are typical in chalk and joint strong limestone exposed at the surface. although spacing is similar (ameen, 1995; gunn, 2004). grikes are most clearly seen in plan on glacially Making comparison with other karren features, scoured surfaces with some time to solutionally White (1988) describes kluftkarren thus: “in the open, and have been thought therefore to be best middle of the size scale, and perhaps most widely displayed in glaciated upland karsts, the fissures distributed, are the linear slots cut in the bedrock along which solution has dissected mature karsts along a guiding structural element.” in reality the such as pinnacled karst, even the stone Forest of size range is from incipient cracks only millime- china (gunn, 2004), are essentially the same basic tres wide to several metres in width and depth and feature, however hard they might be to observe in tens of metres length. some sources define these plan view. indeed, definitions and discussion of features as beginning with larger lower width di- grikes in the literature often move on to consider mensions, but in a theoretical sense it is important enlarged features and complex areas. This is not to consider absolutely minimal dimensions and a the place for an extended description of such fea- scarcely visible hairline crack in the rock is the tures but it is relevant to remark on various ma- beginning of the feature. as soon as such a crack ture griked landscapes in different environments exists solution processes can begin to operate on to set the grike networks of the British isles in a the opening, there is thus a grike or cleft from mm wider context. Feeney (Fornós and ginés, 1996) widths. since grikes are the features that cut up compared limestone pavement in new York state the landscape, and along which processes oper- with the nahanni karst, noting similarities inde- ate, White referred to their identification as one pendent of their very different scales. However, of the three principal landforms of karst along his diagrams do not incorporate the type of ma- with sinkholes and caves. in fact, as already im- ture surface roundedness which is observable in plied, it can be argued that their presence is even mature sites in the British isles, southern spain, or more fundamental, that the existence of fissures the Giant Grikeland of australia (gale et al., 1997). is a necessary condition for the development of The giant grikeland is an established example karst, along with a strong and soluble rock, and of a mature griked area (Jennings and sweeting, the presence of water. sinkholes develop in asso- 1963), now with a semiarid climate, although not ciation with fissures and so do caves. so fissures always, the chillagoe area is another example are a condition for the development of these fun- (Marker, 1976b). The more maturely weathered damental forms. Kluftkarren or grikes are merely areas of northern england that are thought to the visible surface expression of the fissures of a have survived the Devensian glaciation have in karstifiable rock. as such, they are a significant common with these other maturely griked areas component of the epikarst, or the outer skin of the very massive limestones and considerable time for whole karst system. The variation in their scale the landforms to evolve without the radical hiatus and form is enormous and relates in partnership of erosional stripping of the well-weathered upper 90 KRF•1 • OK.indd 90 15.12.2009 10:42:51 Helen S. Goldie, Kluftkarren or grikes as fundamental karstic phenomena blocks by, for example, glacial scour or other proc- section angles of 60, 90 and 120 degrees in plan ess. There are more similarities than differences view. Ford (gunn, 2004) refers to jointing densi- between the basic forms of many of these different ty being inversely proportional to bed thickness landscapes around the globe and an apparent dif- and its considerable variation between carbonate ference can often be ascribed to scale. formations. Thin to medium beds, < 10 to ca 30 cm, are regarded as unfavourable to good karst development due to the wide dispersal of solution. Factors affecting grikes Thicker beds, > 30 cm, even > 100 cm, with joint spacing of 100 cm or more, are the most favoura- Factors affecting grike characteristics can be ble for karstification, allowing as they do some fo- grouped as geological, process and time-related. cussing of solution processes. explanation of vari- of necessity, some detailed description of grike ations from this generalization has been discussed characteristics and examples must be incorporat- by geologists (e.g. narr and suppe, 1991; gross et ed into this discussion. rock characteristics influ- al., 1995), with considerable emphasis on the role ence development of structural lines that are the of macroflaws in the rock material in influencing starting point for grikes. processes produce the joint occurrence and pattern. features, still constrained by the material prop- The distribution of fissures in depth through erties of the rock, and perform their action over rock layers influences how extensive and large varying time periods. Historical considerations, grikes can become, and the propagation of joints involving sequences of development, and possi- across beds is of considerable importance here bly changes in conditions, are also very important. (Helgeson and aydin, 1991). an important char- grikes can be examined from two main aspects: acteristic influencing grike depth development from the plan view or map of their spacing and are the presence, scale and pattern of veins, which pattern which is clearly important for its influence may or may not be contained in joints. gillespie on the pattern development of landforms; and et al. (2001) demonstrate that veins have different from the cross-sectional view, that is their side- influences on the grike network from jointing, in ways shaping and their depth. particular the greater likelihood of persistence by veins through a greater number of rock beds than the joints (Figure 1). This can be affected by the Geological factors affecting grike presence or thickness of shale layers, or by other development changing lithologies. This has material impact in explaining varying plan patterns of grikes and also it is clear that the existence of fissures in a rock influences development in the vertical plane. Ma- mass is the prior condition for grike development. turely griked areas which have not been scoured The prior condition producing the fissures are the by glaciation, for example el torcal de antequera, tectonic forces that put strain on these rock mass- spain (Figure 2), show grikes extending through es, the release of which results in brittle fracture. several beds of rock to attain great depths and influences on fissure patterns and density thus ul- leaving tower-like features between. limited ex- timately explain the potential for grike develop- amples of grike persistence through more than ment. Basic geological characteristics include bed about 2 or 3 beds do exist in northern england, in thickness, macroflaws within the rock material spite of glacial erosion. They are best observed on and tectonic forces affecting the rock mass. The the edges of outcrops or cliffs, in sheltered, higher fissures that develop from fracture patterns result- altitude locations, such as High sleets, above lit- ing from stress fields in the rock masses, namely tondale in Yorkshire, or by the Monk’s path (Fig- the regional joint systems, tend to produce inter- ure 3) above cowside Beck (goldie, 2006). 91 KRF•1 • OK.indd 91 15.12.2009 10:42:51 Karst Rock Features • Karren Sculpturing a angles of jointing in the vertical direction can significantly influence a grike. Most joints are ap- proximately perpendicular to bedding planes, but some are at lower angles, and very massive beds may have curiously curved grikes running at quite low angles across outcrops. Bedding planes in such rocks may also be uneven or curved (Fig- ure 4). angled joints produce sloping grikes, af- fecting grike wall solutional features. in plan most joints are also more or less straight with some minor irregularities, but curved joints are not un- known, for instance on limestone pavement on the west coast of the Burren near Murrough, ire- b land (Figure 5), where the beds are the favourable massive type. gillespie et al.’s work on the Burren distin- guishes vein networks from joint networks, an idea applicable elsewhere. Veins run north-south in the area of sheshymore on the Burren and the joint patterns run east-west with enough join- ing up to produce a grike network with plenty of low angles (Figure 1). This low angled joining produces long and narrow clints, not a rectan- gular pattern. a separate set of cross joints, or c veins, is needed for rectangularity. it is suggested that other areas of long narrow clints, identified by morphometric work (goldie and cox, 2000), reflect the same influences but at different scales according to local tectonic stresses. The similar- ity between the patterns displayed at The clouds (Figure 6) and sheshymore (joints only, not veins) is interesting although the clints at The clouds are smaller; whereas at scar close in Yorkshire there are similarities with the total overall pattern at sheshymore (with veins) with both strong vein- ing and cross jointing, which is seen particularly in the maturely developed area (Figure 7) of scar Figure 1: a. Block diagram, demonstrating possible re- close. lationships between bed thickness, joint spacings and veins (red lines) based on Gillespie et al. (2001); b. one fascinating pavement site on arainn, grike pattern, derived from photograph of pavement county galway, ireland, has very straight joints at Sheshymore, Burren (Gillespie et al., 2001), with one at almost perfect right angles with virtually no main joint set, and veins crossing (red lines). Notional oblique angles (Figure 8). This is in a narrow tec- scale: diagram is 30 m across; c. hooks and “en echelon” tonically stressed zone (langridge, 1971) and its features, plans of minor surface fracture and veining simple network contrasts with complex patterns features (after Ameen, 1995). Notional scale: diagrams are each 1 m across. involving several joint sets and minor plan fea- 92 KRF•1 • OK.indd 92 15.12.2009 10:42:51 Helen S. Goldie, Kluftkarren or grikes as fundamental karstic phenomena Figure 2: El Torcal de Antequera, Spain. General view showing varied massive beds surviving to form towers, wel - rounded upper edges and good runnel ing (photo by L. Hook). tures stemming from the joint forma- tion process. in plan view these latter include hooking patterns and en eche- lon features (Figure 1); and, on the face of joints and therefore on the sides of grikes, many small surfaces mark- ings, such as hackles, steps and ridges (ameen, 1995). These small marks may influence how these surfaces erode by solution; certainly hooking and en echelon features are seen on less-dissected pavement areas such as gaitbarrows, cumbria. grike patterns may also be affected by rotation of joint directions across areas, reflecting changes in stress fields (gillespie et al., 2001). Further generalizations about grikes need comment and amplifica- tion. For example, White (1988) refers to their orientation in folded areas as along the strike. This is not exclusively Figure 3: High Sleets, Littondale, UK. Tower-like outcrop, with grikes through several beds. so, however. grikes may favourably 93 KRF•1 • OK.indd 93 15.12.2009 10:42:53 Karst Rock Features • Karren Sculpturing Figure 4: Dowkabottom, Littondale, UK. Massive limestone outcrops with curved grikes. develop along this strongly stressed direction but another is not exclusive to folded limestones, it there may still be a rectilinear pattern if cross is observable in many horizontal pavements, the joints or veins are involved. an example is at Hut- Burren, ireland (Figure 9), for example. effects of ton roof crags in nW england where strongly stretch and compression in folded terrain can be dipping limestones have two equally strong grike important, for example it is very relevant to un- directions. What is most striking here is that re- derstanding gently folded outcrops at great asby lations between topographic slope, dip alignment scar, as well as more pronounced folding such as and grike orientations produce diamond patterns at The clouds (Fornós and ginés, 1996). (Figure 1). at The clouds, strike direction appears Ford and Williams (2007) observed that grike to have the stronger influence. on the other hand length is inversely proportional to the density of a difference in strength between one joint set and major joints. This is an understandable observa- 94 KRF•1 • OK.indd 94 15.12.2009 10:42:54 Helen S. Goldie, Kluftkarren or grikes as fundamental karstic phenomena Figure 5: Murroogh, west coast of the Burren, Ireland. Curved grikes in very massive limestones. Figure 6: The Clouds, Cumbria, UK. Lower limestone beds at centre of denuded anticline with joint set parallel to strike of fold with similar pattern to that from one set of joints at Sheshymore on the Burren (Gillespie et al., 2001). 95 KRF•1 • OK.indd 95 15.12.2009 10:42:55 Karst Rock Features • Karren Sculpturing Figure 7: Scar Close, Yorkshire, UK. Aerial photograph of Scar Close demonstrating different grike networks, dendritic (Dr) networks seen to the east where acidic waters come off drift and peat, rectilinear (R) on older, longer-exposed surfaces at outer, west edge of outcrop, and in the foreground between modern wal s (long and straight) are signs of old settlement/wal s and damaged surfaces where grikes are less clearly visible (Da). Long wide griked lines pocked with large holes cross from northwest to southeast. tion in some ways, implying that a major joint will identify the most important grikes. scar close has persist for great distances, but not all major joints major grikes spaced tens of metres apart, although will do this. such a pattern may partly depend on in between there are many other grikes, some of the degree of joint opening; if this is well devel- which are as wide as the very long features (Fig- oped it may be virtually impossible to determine ure 7). The major wide grikes persist across coun- individual grike length on the ground, although try both in the well-evolved rectangular area, and aerial photography may resolve the question and in less continuous lines they visibly influence the 96 KRF•1 • OK.indd 96 15.12.2009 10:42:57 Helen S. Goldie, Kluftkarren or grikes as fundamental karstic phenomena dendritic patterned less evolved area. sites with closely spaced fractures still have the grikes pro- ceeding over long distances, for example in the Burren, ireland (Figure 9). The effect is reflected in clint-grike dimensional work which finds that square clints are very rare (goldie and cox, 2000), with the ratio of clint width to clint length gen- erally being well below 1 (indicating squareness). There are many complex patterns with different levels of development, so a tight relationship be- tween grike length and joint spacing is unlikely. Figure 8: Arainn, Ireland. South of Cil Ronain with very length also depends on time available for fissure straight, rectangular patterned closely spaced grikes. widening and linking up of small features to form true grikes. The influence on joint spacing by bed thick- ness (Ford and Williams, 1989; gross et al., 1995) needs further discussion. Widely spaced grikes in thick strata may be able to become deep, long and possibly very wide too, although this depends on propagation of joints between beds. el torcal de antequera in southern spain well illustrates (Figure 2) a site with plenty of time for grikes to develop through several beds, including some very massive ones nearly two metres thick, not having been interrupted by glacial scour in the Quaternary. Wider spaced joints in thick beds usually link across well-bedded limestones into further thick beds permitting grike depths to be- come many metres deep. in other sites stronger beds beneath weak ones may restrict grike depth. This appears to be the case in parts of great asby scars (see chapter 22). plan patterns of grikes at any locality certainly need very careful analysis to be fully understood. High alpine sites such as at sanetsch, glattalp, and the Julian alps, are useful due to their scanty vegetation allowing features to be seen clearly in both plan and section views (Figure 10). Figure 9: Burren, Ireland. Pavement showing differences between strength of two joint sets. Process influences on grike development particular it must be remembered that the many possible solutional processes are not the only ero- processes affecting grike evolution are wider rang- sional processes affecting grikes. The other main ing than may seem obvious in a karstic context. in processes are mechanical. in numerous circum- 97 KRF•1 • OK.indd 97 15.12.2009 10:42:58 Karst Rock Features • Karren Sculpturing Figure 10: Triglav, Slovenia. Stepped limestone with grike persistence through several beds. stances mechanical processes may complete- very like a long, wide grike of 10-fold size across ly overshadow solutional action. There are two scar close, both have plan variation in their line main situations where this occurs: firstly where pattern due to this linking up of many features. the limestone involved is well fractured, in the although solution processes will not be dis- horizontal or vertical plane or both, and secondly cussed in detail here, being adequately treated where some major mechanical process is able to re- elsewhere, it must be emphasized that the solu- move blocks of limestone. in the latter case human tional environment of grikes includes infiltration, activities must not be forgotten for their ability direct rainfall, stream flow, and soil percolation to have major landform effects (goldie, 1986), al- water. Biological influences are very important, though cliffs clearly provide a location where natu- as the many initial small features preparatory to ral large-scale mechanical effects are likely. fully developed grikes are the pans, and slits in to understand how grikes develop it is useful which vegetation can grow for as long as moisture to consider a clear, ungriked surface, after varying is available. These small features initiate in dips lengths of time and in differing solution conditions. and flaws in the rock surface. some may develop small slits or kamenitzas develop into larger and in plan form until moisture drains out as the slit more complex features, which merge at varying opens up to drain under the relevant bed (Fig- size stages. long grikes are invariably associated ure 11). grikes covered in acid soil will develop with many minor features along their line linking rounded smooth features on their sides and be up over time (Figure 10) which occurs on different protected from mechanical processes until soil scales. in plan a simple grike at gaitbarrows looks erosion exposes them. 98 KRF•1 • OK.indd 98 15.12.2009 10:42:59 Helen S. Goldie, Kluftkarren or grikes as fundamental karstic phenomena Figure 11: Gaitbarrows, Lancashire, UK. Complex area of radiating runnels and grikes around what was original y a kamenitza. physical influences have already been implied impose a pattern due to its own properties. such in discussion of geological factors. erosion of may be the case where the rock surface is very im- karstic features itself influences the geological maturely karstified, possibly even devoid of pre- properties, i.e. the fissures, by pressure release for existing widened fissures, and there is a strongly instance. pressure release cracks can be gener- acidic source of water directed across that surface. ated, or existing cracks widen, due to removal of Dendritic solutional runnels result that will grad- lateral confining material from human quarrying, ually cut through the rock, and become grikes. glaciation, or long-term geomorphic processes. With time the rock properties influence locations glaciation as a specific influence has already been for solution and the rectilinear pattern reflecting referred to concerning mature grike development. these rock characteristics develops after the den- also grikes cannot be dissociated from what is dritic one. scar close in Yorkshire, uK, shows this affecting the solid forms of their sides, the clints, sequence (Figure 7). flachkarren, pinnacles and so on, which includes the formation of other karren features such as ril- lenkaren and rundkarren. Differences between Time vertical or horizontal attack affect grikes. over- all, processes affecting grikes and their patterns The whole issue of time is embedded in the dis- will be strongly guided by the geological factors cussions of geological and process influences on already outlined. However, there are situations grike development. if the grike evolution process where the solutional environment may initially involves the total breakdown of the intervening 99 KRF•1 • OK.indd 99 15.12.2009 10:43:00 Karst Rock Features • Karren Sculpturing Figure 12: Trowbarrow Quarry, Arnside-Silverdale, UK. Inter- stratal palaeokarst from Late Carboniferous showing grike and kamenitza-like features on what is now a vertical quarry face, exposed in 20th century by quarrying. clints, involving mechanical as well as solutional by mechanical processes of clear simple grike fea- processes, then it permits much erosion and the tures is inevitable in well-fractured and well-bed- development of very deep and wide, open gaps ded rocks. in less-fissured thick beds mechani- through the limestone. Widened features are cal processes may also eventually become more given various local terms, including bogaz and important as solution dissects the rock. in many strugas. giant grikeland has already been consid- places however, the decay of one type of feature ered. as time progresses what may be regarded simply produces another feature, perhaps at a as a perfect or typical karst landscape will decay scale order larger. due to a combination of processes. The question The model from goldie and cox (2000) (see of when a particular landscape type is perfect or chapter 22) (Figure 13) summarizes many possi- best expressed is a very challenging one. ‘Decay’ bilities in the field of grike development. an im- 100 KRF•1 • OK.indd 100 15.12.2009 10:43:01 Helen S. Goldie, Kluftkarren or grikes as fundamental karstic phenomena Figure 13: Goldie-Cox model. Sketch showing pavement-scar sequence indicating topography before glaciation. Factors affecting variety of topography after glaciation include depth of glacial scour and plucking, lithological variation including palaeokarst bedding plane features, interacting with the pre-glacial topography. Notional clint size ~1 m in columns. VM. very massive; M. massive; L. laminated; A-E. limestone bedding planes. portant further point is the possibility that grikes Morphometric work and related features are not merely mature but relict from earlier periods altogether and re-ac- it is valuable to discuss general morphometric tivated at the present day. a palaeokarstic expla- work as it establishes a basis, or structure on which nation of the large grike holes at great asby scar, to fit specific cases. goldie and cox (2000) present for instance, dated at the late carboniferous, is data on clint width and length, and grike width given by Vincent (1995). These palaeokarstic ex- and depth (table 1). clint width to clint length ra- planations accord with and relate to survival from tios have already been referred to regarding grike glacial scour that must help account for some of patterns (Figure 14). only a moderate correlation the better-developed grikes in northern england. between grike widths and depths was found, sug- This is especially so since it is now thought that gesting that vertical and horizontal developments limestone solution rates in the Holocene in many of these features are not yoked. Variations be- parts of northern england have been much lower tween and within field sites were related to depth than assumed until very recently (goldie, 2005). and date of glacial scour, rates of post-glacial so- Thus many wide and deep grikes and related lution, tectonic disturbance, and lithology, not features could be pre-Devensian, possibly much necessarily in that order of importance. Human earlier, in origin. an example of probable late- impact was identified as influencing present land- carboniferous features is supplied by vertically forms both directly and indirectly. surveyed grike bedded limestones at trowbarrow Quarry, arn- widths vary over two orders of magnitude, from 1 side-silverdale, uK. Here grikes and other karstic cm to over 1 m. outcrops sampled in switzerland features are perpendicular to bedding (Figure 12) (glattalp and sanetsch) and nW england (cum- demonstrating that they must have been formed bria and ingleborough) had relatively wide grikes, before tilting to their present position, which oc- in spite of some of these areas being the most re- curred in the late carboniferous. cently or severely ice-scoured. it is suggested that 101 KRF•1 • OK.indd 101 15.12.2009 10:43:01 Karst Rock Features • Karren Sculpturing 70 Table 1: Grike widths and depths: summary statistics for each sample area (from Goldie and Cox, 2000). Meas- 60 urements rounded to nearest cm. Columns show: n 50 (sample size); p 25 (lower quartile); med: median; p 75 ns of 0.025 40 (upper quartile); max: maximum; range = max–min; iqr n bi = p 75–p 25. 30 Grike widths n min p 25 med p 75 max range iqr 20 Sanetsch 100 5 12 15 20 140 135 9 frequency i 10 Glattalp 29 10 20 28 30 80 70 10 0 Ingleborough 138 1 13 18 25 56 55 12 0 0.2 0.4 0.6 0.8 1 Malham 35 2 8 13 23 61 59 15 clint width/clint length Wharfedale 70 1 10 15 23 76 75 13 Figure 14: Frequency distribution of clint width/clint Cumbria 250 2 10 16 20 140 138 10 length for al areas. Wales 85 1 7 10 17 46 45 10 Burren 510 1 10 13 17 71 70 7 Arainn 200 1 6 9 11 100 99 5 Grike depths wider grikes could have existed before glacial ero- Sanetsch 100 20 43 55 73 150 130 31 sion and survived to develop greater width in the Glattalp 29 50 60 73 100 140 90 40 Holocene (Figure 13). The data for the whole pop- Ingleborough 138 5 71 99 132 244 239 61 ulation of over 1,400 showed 95 % of grike widths Malham 35 38 66 84 96 244 206 30 are below 30 cm, whilst the median is 13 cm. Wharfedale 70 15 53 71 91 168 153 38 Cumbria 250 12 54 83 116 274 262 62 grike depth data also showed a great range, vary- Wales 85 4 27 42 56 95 91 29 ing 70-fold from 4 cm to 274 cm, with a median Burren 510 8 57 84 114 260 252 57 of 74 cm. These grike measures can be ambiguous. Arainn 200 9 38 58 81 157 148 43 The four areas of the largest clints, glattalp, Bur- ren, sanetsch, and arainn, have very varied grike characteristics. Bed thickness needs examination, as an important factor influencing joint spacing cussion on geological influences earlier suggested and, if linked to strength, possible survival from that even where there is only one major joint di- glacial scour. clint morphometry helps in under- rection there is meandering and oblique joining standing grike patterns, particularly the ratio of which limits clint elongation (e.g. The clouds). clint width to length which varies from just above it is concluded that any discussion of grikes can 0 to exactly 1, which would be square. photogenic extend very widely, involving many global karst- clints that are squarish are often depicted in texts lands, and a continuum with smaller and larger as typical and these would have a ratio close to 1. karst features. so definitions and dimensions are The data in goldie and cox actually indicates that crucial. grikes really are truly fundamental karst this pattern is fairly uncommon, and that ratios landscape features, a pre-condition for karst. il- of about 0.4 are the most frequent (Figure 14) al- lustrative examples in the case study chapter (see though the least common clints are the extremely chapter 22) more fully demonstrate some of the elongated, with ratios approaching zero. The dis- generalizations here. 102 KRF•1 • OK.indd 102 15.12.2009 10:43:01 suBsoil sHaping 10 Anikó ZSENI The first classification of karren forms, which undercut solution runnels (Hohlkarren), solution served as a basis for most of the later researchers notches (Korrosionskehlen), and he also described and which discussed subsoil karren, was by Bögli. swamp slots (an extreme form of Korrosionskeh- Bögli (1960a) classified karren types according to len). covered karren include: rounded solution their genesis: free karren develop where the rock runnels (rundkarren), cavernous subsoil weather- is bare, half free karren where the rock is partly ing (kavernöse Karren) and solution pipes (geolo- covered and covered karren where the rock is cov- gische orgeln). ered by soil or dense vegetation. according to his gams (1976) classified subsoil karren as fol- scheme, free-karren are Ril en-, Tritt-, Rinnen-, lows: 1. subsoil rundkarren, 2. subsoil niches, 3. Mäander-, Wand- and Kluftkarren, half-free covered bogaz (subsoil bogaz), 4. subsoil cavernous karren are Kamenitzas, Korrosionskehlen and karren or subsoil tubes (rock holes), 5. covered so- Hohlkarren, while covered karren are kavernösen lution pans (subsoil kamenitzas, covered kamen- Karren, geologische Orgeln and Rundkarren. Bögli itzas), 6. subsoil (covered) wel s, 7. covered dolines (1951) had previously debated the genesis of rund- and 8. fil ed pits. He used the term subcutaneous karren. Bögli’s classification and the german karst as the synonym of subsoil karst. The term names he gave them have been adopted by most subkutane Karren was used in the speleological of the later researchers. unfortunately, many ad- dictionary edited by trimmel (1965) and means ditional terms have been added by other authors karren which develop or are being transformed and some authors have used or translated Bögli’s under soil or vegetation layer. see also Williams terms in a different sense to the original. (1983). Bauer (1962) gave a detailed explanation of how sweeting (1972) describes the chief types of the forms under soil develop. He claimed that the karren. covered karren comprise rundkarren, water penetrating through soil produces a basical- karren formed by roots and soil (which she in- ly different form than subaerial outflow does. He correctly termed Deckenkarren) and hohlkarren, named forms under soil cover soil-karren (Bod- while solution basins (kamenitzas) and kluftkar- enkarren). ren ( grikes) can be both free and covered. Jennings (1971, 1985) also differentiated be- Ford, lundberg and Williams reviewed the tween partly or totally soil-covered conditions. nature and genesis of karren developed under He translated Bögli’s terms as follows. partly-cov- soil cover and in open-air conditions (Ford and ered karren include: solution pans (Kamenitzas), lundberg, 1987; Ford and Williams, 1989). The 103 KRF•1 • OK.indd 103 15.12.2009 10:43:01 Karst Rock Features • Karren Sculpturing karren classification of Ford and Williams (1989) The role of soil cover in the evolution is based on morphology with subdivisions that of karren features incorporate genetic factors. Their main groups are circular plan forms, fracture controlled lin- There has been a lot of research in the internation- ear forms, hydrodynamically controlled linear al literature not only on the classification of sub- forms and polygenetic forms. in some cases they soil forms but also on the soils on karst. Much of gave the same name for the similar forms which this deals with the role of soil cover in the evo- can be found or can be developed both beneath lution of limestone forms. sweeting (1966, 1972) soil and on bare rock (e.g.: solution pits, karren in her study about the solution process of lime- shafts or wells). in those cases i put a “subsoil” at- stone, assigned a great role to the soils and plants tribute before the name that they gave. Thus, they in the development and evolution of limestone recognized the following subsoil forms: subsoil pavements. pigott (1962, 1970) also studied the pits, subsoil karren shafts or wel s, cutters (subsoil connection between soil and evolution of forms grikes), subsoil pinnacles, rundkarren ( subsoil so- occurring on limestone. trudgill (1975, 1985) de- lution runnels), pinnacle karst, ruiniform. soil also scribed the connection between morphology of has an important effect in the genesis and forma- karren and the different dissolution processes of tion of decantation or overspill forms: these are subaerial and subsoil conditions. He emphasized decantation runnels (“Wandkarren”; Bögli, 1960a) the importance of soil reaction in the subsoil so- and decantation flutings. lution processes: cuspate, arcuate and runnelled a. ginés (1995, 1996a) named karren features forms are only found under acid soils and if these that developed on the bedrock surface beneath forms are observed on bare rocks then soil loss has soil cover subcutaneous karren or cryptolapiaz. occurred (trudgill, 1975, 1985, 1986; trudgill and He listed smooth and rounded rock surfaces, hol- inkpen, 1993). experiments by urushibara-Yoshi- lows, pits, tubes and pinnacles. no et al. (1999a) proved that the solution rates of slabe (1999) distinguished rock forms (which limestone tablets in soils are several times higher are due to various factors acting on the surface of than those in the air. the rock) from karst forms (which are due to water Jakucs p. (1956) considered the biogenic effects percolation through fissured or porous rock). For very important in the development of surface rock forms that developed on karst surfaces cov- karst forms. He thought that the acid-secretion ered by soil or sediment he used the term subcu- of plants and microorganisms in soil was the pri- taneous. according to their origin, subcutaneous mary factor in the development of karren. Jakucs rock forms are grouped as follows (slabe, 1999): 1. l. (1977) emphasized the common role of soil, cli- a consequence of water flowing along the contact mate and vegetation in the evolution of karst fea- between rock and soil, in the form of subcutaneous tures. He proved that the rate of limestone solu- smal and large channels (rundkarren), subcutane- tion mainly depends on the biological aggressivity ous scal ops; 2. a consequence of the percolation of soil covering the limestone. so karst-corrosion of water through the soil, in the form of subcuta- is determined by the physical and chemical prop- neous smal and large channels (rundkarren and erties of soils, closely connected with microcli- kavernösen karren), subcutaneous smal and large mate (Jakucs, 1977). recesses; 3. a consequence of an inflow of water to soil has an important role in the evolution of the surface of the soil surrounding the rock, in the different karst features, such as types of karren form of subcutaneous half-bel s and subcutaneous (Figure 1). Briefly, the difference between forms notches. Forms in group 1 and 2 develop under which occur under a soil and those which devel- soil; forms in group 3 develop at and just below oped in open-air conditions is that angular forms the ground surface. are produced directly by dissolution by rain and 104 KRF•1 • OK.indd 104 15.12.2009 10:43:01 Anikó Zseni, Subsoil shaping Figure 1: Subaerial and subsoil bedrock forms (after Trudgil , 1986, modified). surface runoff, while rounded forms have formed soil and the denudation rate of limestone is higher under a soil or peat cover (Bögli, 1960a; Jennings, in the case of subsoil/subcutaneous karren than 1971; sweeting, 1972; trudgill, 1986). of subaerial, bare karren. in the case of subsoil so- Karren which develop in open-air conditions lution, the co dissolved by rainwater from the 2 are fretted and sharp, having been subjected to air is not the only factor. Billions of microorgan- much more selective corrosion. on bare rock isms live in the soils which cover the limestone. surfaces rapid runoff occurs on sloping surfaces, These microorganisms break down the organic whereas on flat surfaces or in basins water will be matter (decomposition of organic waste, fallen resident on the surface much longer. it can be said foliage, animal remains, etc.) and produce co 2 that a soil-free surface suffers only episodic disso- and other acids from it. The macroflora produces lution by rainwater. However, in addition, rainwa- co directly through respiration of roots. The 2 ter has considerably less dissolution potential than regularly renewed co produced by micro- and 2 acid soil water. Thus, under conditions of rapid, macroflora changes the composition of soil-air, so transient flow, where water depths are limited, the flora becomes, indirectly, the main factor of only the rapidly soluble constituents of the rock karst corrosion. in the case of temperate, mediter- will be removed. Forms will be related to disso- ranean and tropical karst corrosion the most cru- lution kinetics and the overall chemical reactions cial source of the dissolving power of water is the will be rate limited (trudgill, 1985; Ford and lun- co originated from biogenic process in the soil 2 dberg, 1987). The main types of karren that form (Jakucs, 1977). although the main role is played on bare surfaces are discussed in other chapters by co , various organic acids (fulvic, humic acids, 2 of this book. in addition, selective solution pro- formic acid, acetic acid, oxalic acid, lactic acid, duces deep grikes and runnels and a rough pitted propionic acid, humus, etc.) ‒ originating from surface. the bioactivity of soils ‒ also take part in the dis- Due to the higher co concentration and the solution of limestone. Jakucs (1977) declared that 2 longer time period of direct water-rock contact, the dissolution of limestone or karstification is the solution of limestone is more intense under essentially a response by the bedrock to the phe- 105 KRF•1 • OK.indd 105 15.12.2009 10:43:02 Karst Rock Features • Karren Sculpturing Calcareous soil Acid soil rain rain zone of solution acid soil calcareous soil acid water calcareous water active solution at soil/rock interface limestone calcareous soil: no production of large grikes limestone acid water percolates down in times of heavy rainfall and dissolve large grykes calcareous soil 0 40 cm soil washed down into large opened zone of active solution joints and possibly to caves Figure 2: Weathering under acid and calcareous soils (after Trudgil , 1985, modified). nomena of biological and chemical development or has been a covered karst, we can assume that of the paedosphere covering the rock. the bulk of karren was originally generated, de- The chemical production of acids in the soil is veloped and shaped when the bedrock was still extremely climate-sensitive, as the biological ac- buried below soil cover (a. ginés, 1995) (see later tivity of soil microorganisms is very sensitive to section in this chapter on “remnant subsoil forms variations in temperature, soil-moisture, etc. so, inherited and transformed after soil removal”). in the case of soil-covered karst areas the intensity under drift and soils, rock surfaces can be at- of limestone corrosion is determined indirectly tacked from many directions, so smooth, rounded, by the climatic conditions of the area, through its runnelled and pinnacled forms may be found fre- effect on the paedo- and biosphere. That is why quently with etched surfaces and comparatively striking differences (in the order of a magnitude) shallow incision of the bedrock. subsoil attack on and the very characteristic regional morphologic tabular blocks from three sides may give rise to differences of the karst forms can be seen in dif- a pinnacle form. arcuate and cuspate forms also ferent climate zones of the earth. in the tropics, occur under acid soils. where both the rich vegetation cover and the mi- according to trudgill (1985), the role of a soil croorganisms in the soil exhibit much more vig- cover in karst evolution depends on the pH of soil. orous activity, the effect of the biogenic, subsoil in general, calcareous soils with high pH (pH of formation is much higher (Jakucs, 1977). 7 to 9, where calcium carbonate content is high- so the surface karstic phenomenon and forms er than 10%) protect the underlying limestone are mostly the results of processes initiated by almost completely from erosion, because water subsoil biogenic corrosion and only in a smaller becomes saturated with bicarbonate on passing part by processes initiated by tectonics or erosion through the soil profile. so the soil water arriving (Jakucs et al., 1983). and as almost every karst is at soil-bedrock interface is incapable of dissolving 106 KRF•1 • OK.indd 106 15.12.2009 10:43:02 Anikó Zseni, Subsoil shaping Figure 3: Tafoni-like features promoted under soil covering at the foot of a karren pinnacle. Lluc karrenfield, Mallorca, Spain (photo by A. Ginés). the bedrock (Williams, 1966; trudgill, 1985). The lowering of the soil surface down the developing existence of preserved glacial surfaces under drift cleft, possibly also with soil loss into near surface and soils must be related to the calcareous cover. cave systems. But surface wash induced by the so there is a relationship between soil characteris- felling of trees also causes erosion. so, for example, tics, especially pH and calcium carbonate content many limestone pavements could have been post- and the rate of bedrock erosion (sweeting, 1966; glacially covered more extensively with drifts and trudgill, 1986). soils (trudgill, 1985). The presence of smoothly if the percolating water is not saturated, then eroded surfaces in many pavement areas supports solution will take place. under acid soil limestone this hypothesis (see chapter 9). is extensively weathered. erosion of limestone is soil moisture distribution, drainage rates, soil most severe beneath deposits supporting an acid texture, soil depth, water-flow rates, slope, vegeta- vegetation and with a pH between 4 and 7 and a tion, and the nature of the limestone will all have calcium carbonate content less than 1% (trudgill, a strong influence on the distribution of subsoil 1985). The extensive weathering under acid soils solutional erosion and resulting landforms. leads to the formation of solutionally opened The degree of erosion is commonly greater joints (grikes), and soil is often washed down into under deeper soils, which may reflect the increase the larger grikes (Figure 2). This can result in the in soil air carbon dioxide values with greater 107 KRF•1 • OK.indd 107 15.12.2009 10:43:03 Karst Rock Features • Karren Sculpturing Figure 4: Irregular subsoil sculpturing on ex- posed rundkarren. Hutton Roof Crags lime- stone pavement area, England, UK (photo by H. S. Goldie). depth. There seems to be less erosion beneath soils incorporated in the soil, making it less acid. softer on steep slopes than under those on flat or gently rocks typically support a calcareous flora, while sloping ground. soils on the steeper upslope sites hard limestones may support a calcifuge vegeta- tend to be thinner than those on flat ground; the tion, which has the effect of increasing the erosion top of the soil profile is often truncated and there rate (trudgill, 1985). is a higher rate of downslope removal of mate- time is a factor that has great importance in rial. Hypothetically, a light textured soil should determining whether or not soil characteristics enhance leaching with a concomitant decrease will tend to encourage weathering or protect the in pH and increase in erosion, but this effect is rock beneath it (trudgill, 1985). masked by other factors (trudgill, 1985). The na- ture of the limestone is important in determining the rate of its erosion beneath a soil. Hard, mas- Subsoil forms sive limestones do not break down easily and the soils above them tend to be easily leached. More grouping of subsoil karren forms is not easy, as friable rocks fragment more easily and become similar forms have been given different names by 108 KRF•1 • OK.indd 108 15.12.2009 10:43:06 Anikó Zseni, Subsoil shaping Figure 5: Cavernous zone at the foot of a karren pinnacle, Australia (photo by K. G. Grimes). Width of view is 5 m. different people. as we can see at the beginning of Irregular subsoil sculpturing – the this chapter, the researchers’ approach to classifi- kavernösen Karren of Bögli cation is also different. some authors have used a variety of genetic classifications (e.g. Bögli, 1960a, Cavernous karren 1978) while others have used descriptive or mixed cavernous features are openings generally of a classifications, and this has influenced their ter- small size bored into the rocks. They are complex minology. There is a tendency for gradations to of irregular, randomly shaped, inter-connecting occur both in the nature of the forms and in their cavities intricately perforating the rock (look like sizes and orientations. Different people have put a “spongework”) (Bögli, 1960a, 1978). The size of the boundaries at different points along those the cavities is from a few centimetres to more than continua. a metre across. cavernous weathering is not al- also, it has not helped that the original terminol- ways the result of subsoil corrosion, but frequent- ogy evolved in europe before the tropical karren ly some tafoni-like features are clearly promoted were well described. The different languages (and under soil covering (Figure 3). translations) also increase the confusion. Despite There is a lot of overlap in the usage, by differ- the above mentioned difficulties, a classification of ent authors, of the terms honeycomb, boneyard, subsoil karren has been created. table 1 gives us cavernous weathering, cavernous karren and root- an overview of the different subsoil karren forms, karren. They are all deep, irregular sculpturing of including the mostly used synonyms as well. the rock (Figure 4), caused by a soggy material in 109 KRF•1 • OK.indd 109 15.12.2009 10:43:07 Karst Rock Features • Karren Sculpturing Table 1: Classification of subsoil karren. 1. Irregular subsoil sculpturing 2. Horizontal notches at or just 3. Pits and vertical pipes/shafts 4. Linear subsoil channels 5. Pinnacles and cutters, (the kavernösen Karren of Bögli) below the ground surface (the geologischen Orgeln of (the Rundkarren and Hohlkar- covered bogaz Bögli) ren of Bögli) 1.1. Cavernous karren 2.1. Subsoil notches 3.1. Solution pits 4.1. Rundkarren variations: variations/synonyms: variations/synonyms: variations/synonyms: variations/synonyms: cutters, honeycomb, small subsoil notches, subsoil pits, rounded runnels, subsoil pinnacles, boneyard, smaller subcutaneous notches (= subcutaneous recesses, Hohlkarren (= undercut solution covered bogaz, cavernous weathering, subsoil niches), tapering hollows, runnels), ruiniforms subsoil cavernous karren, large subsoil notches, covered solution pans large subcutaneous channels, rock holes, larger sub-cutaneous notches small subcutaneous channels subsoil hollows, 3.2. Karren wells subsoil tubes, 2.2. Subsoil half-bells synonym: 4.2. Decantation or overspill root-karren subcutaneous half-bells subsoil (= covered) wells forms decantation runnels 1.2. Subsoil scallops 3.3. Karren shafts and solution variations: variations/synonyms: pipes soil-fed wandkarren, subcutaneous scallops, synonyms: filled pits, soil-fed meanderkarren smooth undulating surfaces deep subcutaneous recesses contact with limestone, with irregular concentra- into the initially fine cracks in the rock eventually tions of activity related to roots, or other localized enlarge these into a spongework of wider, mean- flow and owing to its biological or chemical effects dering and branching dissolution channels, which upon the rock surfaces. are usually round or oval in cross section. These on a scale greater than the tiny cavernous grade to cavernous karren, and the term “root weathering, subsoil cavernous karren (subsoil lapiés” has been used in a broader sense to include tubes and rock holes included) have round cross- those forms (Jakucs, 1977). sections (gams, 1976) and they run in all direc- tions (Figure 5). Their development usually fol- subsoil scallops lows a joint or fissure. Subsoil scal ops (subcutaneous scallops, after slabe, slabe (1999) used the name subcutaneous tubes 1999) are scallop- or ripple-like forms, 15‒50 cm for small irregular karst cavities, which appear in wavelength and shallow, but mostly deeper on filled with sediment or soil. They develop mainly the upper side (Figure 6). Because of subsoil or- in distinctly fissured rock: the initially small tubes igin they have smooth walls. They are found on filled with soil are widened by dissolution. The steep to overhanging rock surfaces surrounded vegetation often has an important effect in their by soil or sediment and observed following dis- formation. smaller ones are 1‒10 cm in diameter, tinct fissures along which soil-filled cracks devel- larger ones can be up to 1 m. op. They are usually linked in a network. accord- root-karren ( root grooves) in the strict sense of ing to slabe (1999) subcutaneous scallops devel- etchings that show the shape of the causative root op on a relatively permeable contact between the are not often recognizable as such (e.g. see plate rock and the sediment when there is simultaneous 3.4 of Bögli, 1978). These are formed under com- water flow along both the contact and through the pact soil where roots etch into the limestone sur- soil and sediment. face, or penetrate along joints (Jakucs, 1977). The Smooth undulating surfaces are also common evolution of these subsoil corrosion channels is features under soil, showing a larger and irregular helped by the root-acids of greater plants and the wavelength than subcutaneous scallops and giv- intense co production of microorganisms sur- ing a characteristic rounded pattern to the walls 2 rounding the root system. The roots penetrating of clefts, cutters and solution pipes. 110 KRF•1 • OK.indd 110 15.12.2009 10:43:07 Anikó Zseni, Subsoil shaping Horizontal notches at or just below the bottom in a semicircular fashion and usually is re- ground surface – the Korrosionskehlen of shaped by the dissolving effect of rainwater. Bögli The term “subsoil niche” (gams, 1971, 1976) may refer to slabe’s small subcutaneous notches. subsoil notches a subsoil niche is a horizontal recess, which is Smal subsoil notches (smaller subcutaneous usually many times wider than deep and its depth notches, after slabe, 1999) are shaped as semicir- into the rock face is usually less than 10 cm. They cular horizontal channels. They are 10‒20 cm in can be found in unbedded or thick bedded lime- diameter and their upper edge is sharper while stone. Their cross section is irregular. on vertical the lower is rounded. Large subsoil notches (larger rock walls they are more open and semicylindri- subcutaneous notches, after slabe, 1999) are in- cal, while on inclined wall they are rather like a dented 1 m or more into the limestone and can furrow. similar forms can be found in caves filled be up to 1 m high. They develop due to the corro- or formerly filled with loam or clay. sion of limestone in contact with the soil at or near note that the forms described here are quite a stable ground surface (Figure 7). smaller semi- distinct from notches that form by the preferential circular notches appear first and then they grow solution of rock along horizontal bedding planes: larger with the slow lowering of the level of soil these latter are narrower and relatively deep in re- or sediment. They occur when more water flows lation to their height. to the soil-limestone contact area than can imme- diately be conducted away and therefore the so- subsoil half-bells lution of limestone is more rapid and stronger in Subsoil half-bel s (subcutaneous half-bells, after this area. slabe, 1999) develop under vertical subaerial The lower part of subsoil notches is undercut, channels from which large amounts of water enter smooth and rounded because here the solution is the sediments or soils at a localized point (Fig- stronger and lasts longer. The bottom of the notch ure 8). Just as in the case of subsoil notches, sub- is horizontal. The upper part descends toward the soil half-bells occur when more water flows into Figure 6: Subsoil scallops exposed in an excava- tion. Chil agoe, northeast Queensland, Australia (photo by K. G. Grimes). Width of view is 100 cm. 111 KRF•1 • OK.indd 111 15.12.2009 10:43:09 Karst Rock Features • Karren Sculpturing Figure 7: Subsoil notches. Solid arrows indicate water flow, open arrow indicates a dropping soil surface (drawing by K. G. Grimes). the limestone contact zone than can be conducted away, therefore intensive solution can take place. Their shape and size is controlled by the amount and solution power of the water that reaches the soil, the permeability of the soil-limestone contact zone and the rate of the soil lowering. The walls of the subsoil half-bells from which soil has recently removed can be criss-crossed by subsoil recesses (i.e. small rundkarren). Figure 8: Subsoil half-bel (drawing by K. G. Grimes). Pits and vertical wells, shafts and pipes – the geologischen Orgeln of Bögli to make a distinction between pits, wells, pipes and shafts we use the ratio of depth to width (D/W) rather than the absolute depth (Figure 9). “pits” are shallow holes with a D/W ≤ 1 (they are wider than their depth). “Wells” are vertical holes with a D/W between 1 and 2. “shafts” or “pipes” (we use them as synonyms) are deeper vertical holes with a D/W > 2. in that D/W scheme a pan has very low D/W (< 0.2). For inclined and hori- zontal pipes we suggest using “tube”. solution pits Subsoil solution pits are shallow circular, semicir- Figure 9: Distinction between pan, pit, wel and shaft or cular, oval, rectangular or irregular plan forms pipe, based on depth to width ratio (drawing by K. G. with rounded edges and rounded or tapering Grimes). floors (in contrast to solution pans ‒ which are sub- 112 KRF•1 • OK.indd 112 15.12.2009 10:43:09 Anikó Zseni, Subsoil shaping aerial forms and have a flat bottom). smaller ones are 1‒5 cm in diameter, larger ones are up to 1 m (with a D/W ≤ 1). according to Ford and Williams (1989) they occur both on bare rocks and under soil cover (see chapter 15). slabe (1999) included solution pits in his “subcutaneous recesses”, along with deeper solution pipes. He suggested that they developed under a thin soil layer due to the perco- lation of water through the soil to the rock. Their formation begins on little spots on the rock under soil and the solution power of the water percolating through the soil enlarges them. a lot of them, located along small joints, taper down into them; others have developed at a cluster of primary pores or a vug. They are often side by side or already connected. rundkarren (subcutaneous channels) often lead from the larger solution pits. Many subcu- taneous forms can develop from the initial pits. on flat surfaces solution pans can develop from the more rounded pits when the rock is denuded, so there are several intermediate stages of forms. gams (1976) described covered solution pans (subsoil pans, subsoil kamenitzas, covered ka- menitzas), which hold stagnant water after the soil Figure 10: Sections of karren shafts, tubes and solution has been stripped off. subsoil pans have more oval pipes exposed in a marble quarry at Wombeyan, New South Wales, Australia (photo by K. G. Grimes). Width of and elongated form than surface kamenitzas and view is 5 m. lack the flat bottom. Wider solution pits can also evolve into either vertical soil-filled wells, shafts or into an irregular solution is intense, because the dominant condi- spongework (cavernous karren). tion is epikarstic. They can grade to a “boneyard” or “spongework” morphology. gams (1976) called Karren wells them subsoil (covered) wells. Karren wel s (Ford and Williams, 1989) are short (usually few centimetres to 2‒3 m) vertical or in- Karren shafts and/or solution pipes clined caves draining into the epikarst, with a Karren shafts are vertical, cylindrical holes devel- D/W = 1–2. longer versions are called solution oped by vertical dissolution processes (they are pipes or shafts. Their cross section is often circu- similar to karren wells, but karren shafts and solu- lar, elliptical or funnel-shaped. Their formation is tion pipes are deeper, with a D/W > 2). Downward initiated by joints, calcite veins, fissures or bed- water flow can occur along the intersection of ding planes: they can develop from proto-caves or joints or where the bedding plane is dipping steep- from pits and pans where the floor of the former ly, or water moving vertically down through po- features intersects bedding planes. They can be rous limestone can be concentrated into localised very complex and variable if developed under paths that form a field of solution pipes (lundberg deep, periodically saturated soil cover where dis- and taggart, 1995). They can be as much as 20 m 113 KRF•1 • OK.indd 113 15.12.2009 10:43:11 Karst Rock Features • Karren Sculpturing Figure 11: Exposed rundkarren. Great Asby Scar limestone pavement area, England, UK (photo by H. S. Goldie). in depth, filled with soil, loam, clay, sand, gravel, lundberg, 1987; Ford and Williams, 1989). They rubble, etc. Where open pipes are seen, this repre- have rounded cross sections and smooth surfac- sents loss of an earlier fill. They are often excavat- es (Figure 11). They are similar to rinnenkarren ed on walls of quarries (Figure 10). if the filling is (see chapter 17), as both are Hortonian chan- allochtonous, the pipe (as a rocky form) seems to nels, heading where sheetflow or wash on a slope be fossil. They are most common in porous lime- breaks down into linear threads and, as they gain stone (see chapter 42). gams (1976) called them discharge downstream, they normally widen and filled pits. slabe (1999) called them deep subcuta- deepen downstream. Both of them develop on neous recesses. These features constitute the tran- most carbonate rocks, but are best formed on ho- sition between the deeper karren and the epikarst. mogeneous and medium- to fine-grained carbon- ate rocks. They also develop well on gypsum, ba- salt, granite and sandstones but are not known on Linear subsoil channels – the Rundkarren salt. The main difference is that rinnenkarren de- and Hohlkarren of Bögli velop on bare slopes and have sharp rims and flat or rounded bottoms. Rundkarren The width of rundkarren can vary from 2‒50 rundkarren are solution runnels which are cm, their depth is up to 50 cm or more and their formed under soil or acid till, moss, vegetation, length is from a few centimetres to over 10 m. The humus or litter cover (sweeting, 1972; Ford and size of rundkarren is usually related to time: the 114 KRF•1 • OK.indd 114 15.12.2009 10:43:11 Anikó Zseni, Subsoil shaping longer the period of soil-covering the deeper the the soil and the quantity of water flowing over the runnel. But it also depends on the power of solu- contact, along with the composition of the rock. tion processes under the soil. slabe (1999) distinguished between large and They are basically drainage features and are re- small subcutaneous channels. large channels de- lated to the slope and dip of the rock. on steeper velop when larger quantities of water flow contin- slopes they tend to run parallel (slabe, 1999). on uously along the permeable contact with the rock. gentle slopes or flat surfaces their development is They have diameters from 20‒100 cm or more. usually orientated towards pre-existing solutional Their width can fluctuate along their length. openings along major joints: they often have den- small channels are 5‒20 cm in diameter and dritic confluence or centripetal orientation into a they are equally wide along the entire length or karren shaft or grike (Figure 12). wider at the contact with other channels. These are slabe (1999) used the term “subcutaneous the more typical rundkarren. They criss-cross the channels” for vertical (or sometimes horizontal or vertical or sloping wall at various angles, can be inclined) channels of various sizes, usually with linked in a network and can have winding shapes. semicircular cross-section, which develop where less permeable the contact is the more winding water flow is concentrated along the contact be- channels develop. The most windings channels tween the wall and the sediment that covers the are 1‒5 cm in diameter. small channels can occur rock and fills the cracks along vertical fissures. on the walls of larger channels. Vertical subcuta- The size and shape of the channels are dictated neous channels are common in stone forests. primarily by the permeability of the contact with Hohlkarren furrows are an accentuated vari- Figure 12: Subsoil rundkarren exposed by soil stripping adjacent to a cliff. Border Rivers karst region, southeast Queensland, Australia (photo by K. G. Grimes). Width of view is 1 m. 115 KRF•1 • OK.indd 115 15.12.2009 10:43:12 Karst Rock Features • Karren Sculpturing ety of rundkarren formed under an acidic peat or ren on gentle slopes) (Ford and Williams, 1989). humus cover (Bögli, 1960a; sweeting, 1972). They Because they do not collect additional acidic water are rounded and smooth ‒ similar to rundkarren downslope, away from the source, the size of the ‒ but tend to have broader troughs and more un- channels reduce downslope. Decantation runnels dercut sides. They are 0.6‒1 m in depth, about 50 occur where overspill is at a point. cm wide and can be several metres long. Jennings The dimensions of runnels are in proportion to (1985) called them “undercut solution runnels”. the volume and acidity of water. generally run- nels have 1‒10 cm in widths and depths and up to Decantation or overspill forms 10 m in length. They are usually sharp-edged as Decantation or overspill forms are classes pro- they form subaerially. posed by Ford and lundberg (1987) and include “halbfrei” or “partly covered” karren of previous classifications (Bögli, 1960a; sweeting, 1972; Jen- Pinnacles and cutters, covered bogaz and nings, 1971). ruiniforms Mostly they occur upon bare and partly cov- ered surfaces but can also form beneath soil. The pinnacles are residual forms. Dissolution beneath common feature of these forms is that the solvent deep and acid soil, or over a long time enlarges is supplied as an overspill from an upslope store. grikes (kluftkarren; see chapter 9). These become This store can be soil, moss, humus, snow bank, much widened at the top and taper with depth to etc., retained in a depression and overspilling to form cutters (a north american term) (Figure 13). form decantation runnels on the adjoining bare simultaneously, intervening clint tops are cut surface (wandkarren on steep slopes, meanderkar- back by subsoil runnels to form subsoil pinnacles: Figure 13: Cutters (subsoil klufkarren) exposed in a quarry. Wombeyan, New South Wales, Australia (photo by K. G. Grimes). Width of view is 7 m. 116 KRF•1 • OK.indd 116 15.12.2009 10:43:13 Anikó Zseni, Subsoil shaping peaks with rounded runnel incuts (Ford and lun- Bogaz are also a type of kluftkarren: a more dberg, 1987; Ford and Williams, 1989) (Figures 14, complex and larger scale corridor landform than 15). They appear to form best in very thick bed- grikes. covered bogaz (or subsoil bogaz) (gams, ded to massive strata with well-spaced joints. The 1976) are corridor-like features, which can de- pyramidal forms cannot easily retain soil and be- velop by widening of joints, fissures, fault zones cause of soil erosion their tops often become ex- and zones of fractured limestone. They are 2‒4 m posed. Then they become sharpened by subaerial in width and up to 5 m or more in depth. They karren forms (see next section on “remnant sub- are more frequent in thick bedded or massive soil forms inherited and transformed after soil re- limestone. after soil erosion, covered bogaz can moval”). pinnacles can be several metres in height. be transformed into surface bogaz. Most surface as their development needs a long time they are bogaz occur on inclined slopes where soil erosion common in the humid tropics and rare in recently is faster. glaciated areas. The best known pinnacle karst is ruiniforms were first described by perna and the “stone Forest” in Yunnan, china (see chapters sauro (1978). They are terrains where soil has 35, 36). been removed from very deep and wide grikes but Figure 14: Cutters and pinnacles form by progressive subsoil enlargement of vertical joints (drawing by K. G. Grimes). Figure 15: Subsoil pinnacles being exposed by soil erosion at the edge of an old quarry. Face is about 5 m high. Quarry north of Railton, Tasmania, Australia (photo by K. G. Grimes). 117 KRF•1 • OK.indd 117 15.12.2009 10:43:15 Karst Rock Features • Karren Sculpturing Figure 16: Rillenkarren growing over subsoil-inherited forms. Rillenkar- ren with steps and horizontal so- lution ripples exposed by soil ero- sion. Border Rivers karst region, southeast Queensland, Australia (photo by K. G. Grimes). where the clints are not sharpened into pinnacles comes sharper, rough and serrated (Figure 17). but stand out like miniature city blocks in a ru- Many karst sites of the world have gone through ined townscape (see chapter 38). removal or qualitative/quantitative changes of soil cover and vegetation. Deforestation also enhances soil erosion. conversely, bare surfaces can be colo- Remnant subsoil forms inherited and nized by vegetation and covered by soil or till de- transformed after soil removal posits in glaciated areas. The alternation of buried and bare stages during their life causes the mixed as we see, subsoil forms are characterized by origin of many karren forms. sometimes a clear smooth and rounded surfaces. after soil remov- distinction between bare and covered karst forms al they are exposed to subaerial solution and me- is very difficult to make or even impossible. chanical and biological weathering (rain-water, a. ginés (1995) in his two-stages-model of kar- algae, fungi, etc.), which reshapes their forms ren development assumed that beneath natural (Figure 16). The originally smooth surface be- vegetation and soil cover an intense growth of 118 KRF•1 • OK.indd 118 15.12.2009 10:43:16 Anikó Zseni, Subsoil shaping Figure 17: Sharp break between smooth subsoil surface and sharp rillenkarren above. Molong, New South Wales, Australia (photo by K. G. Grimes). subsoil karren is produced. His model, based in ren and other subaerial karren forms appear. Jakucs (1977), attributes great importance to the Morphometric analyses (e.g. width, depth, length loss of soil. after soil removal the former sub- of rillenkarren) can be used for dating the age of soil forms begin to be reshaped and transformed soil erosion or deforestation and for reconstruc- under subaerial conditions (rainfall, rapid runoff tion of a palaeoenvironment (ginés, 1996b). water, etc.). The soil loss usually results from eco- if a limestone was covered by acid soil then logical crisis (e.g. climate change) and anthropo- more and many kinds of subsoil forms are ex- genic effects: deforestation, change in cultivation, posed by soil removal. These begin to reshape. as etc. a. ginés’ (1995) two-stages-model can be ap- an example, according to sweeting (1972), the plied to most karren landscapes in warm and tem- runnels of Hutton roof limestone pavement in perate climate. Britain, which are narrow, have rounded edges if a limestone was covered by a calcareous soil and deepen downslope just like a closely spaced, and the dissolution power of the percolating water overdeepened rinnenkarren, were originally was weak, then after soil removal the remaining rundkarren which have been modified after soil surface is smooth, plain with no or only a few sub- removal. soil karren forms (Figure 18). on the bare surfaces another good and spectacular example are the microkarren, rillenkarren, solution pans, trittkar- high columns (pinnacles) of stone forests. on the 119 KRF•1 • OK.indd 119 15.12.2009 10:43:17 Karst Rock Features • Karren Sculpturing Figure 18: Soil cover protects underlying limestone from solution at Great Asby Scar limestone pavement area, Eng- land, UK (photo by H. S. Goldie). lower part of the columns, rounded subsoil sur- Recent soil studies in England and faces and formations are predominant and these Hungary grade up to the more strongly reshaped, rougher and sharper upper part of the columns, which to understand the evolution of different karstic have been corroded by rainwater for a longer time features, especially types of karren, the investi- (see chapters 35, 11). gation of soils on karst landforms is very impor- 120 KRF•1 • OK.indd 120 15.12.2009 10:43:19 Anikó Zseni, Subsoil shaping tant. There are some interesting questions, e.g.: ety of karst features within short distances, the are there any connections between solutional results show only minor differences in soil pH. power of soil (e.g. pH and carbonate content) and soils on both the Hungarian karrenfields seem the depth, smoothness, and roundness of lime- to be rather homogeneous, while on the english stone forms? to answer this question soil samples limestone pavements not only the appearance were collected from limestone pavement areas of of the limestone surface but even the soil pH north england and on karrenfields of aggtelek can vary considerably within short distances; Karst and Villányi Mountain, Hungary (Zseni, • the results of the english measurements veri- 1999, 2002a, b; Zseni and Keveiné-Bárány, 2000; fy that the soils with lower pH are associated Zseni et al., 2003). samples were from rundkar- with deeper solution features, not surprisingly ren, grikes and cavernous karren, representing as their solvent power is greater than that of the different solutional and tectonic features of lime- soils with higher pH; stone. During the examination, the pH and car- • the neutral and weakly basic pH-values of Hun- bonate content of the soils were determined and garian soil samples came from deep and well- the connection between the soil characteristics developed solutional forms. This suggests the and features of limestone was examined. Various almost complete protection of the underly- solutional conditions, which result from soil vari- ing limestone from dissolution (although the ations, have produced a variety of forms. carbonate contents are low). By reason of the Human influences have played an important present soil pH, these solutional forms in the role in the evolution of both Hungarian karren- Hungarian karrenfields either had to develop fields. previously the slopes were covered by sedi- formerly (during the period of dense vegetation ment but intensive cultivation (viniculture) caused cover and more acid soil) or there are processes soil erosion, and increasing exposure of the lime- which can lower the pH occasionally or season- stone blocks. on the bare and thin-covered sur- ally, so that solution can take place. faces of limestone different solutional forms can be seen. The comparison of the Hungarian and english Acknowledgements soils results in the following: • great differences can be found in soil pH asso- The author thanks Helen s. goldie and Ken g. ciated with the different karst features in the grimes for the photos and Ken g. grimes for his english samples. However, in the Hungarian very valuable suggestions and for making im- karrenfields, in spite of the similarly rich vari- provements to the english text. 121 KRF•1 • OK.indd 121 15.12.2009 10:43:19 KRF•1 • OK.indd 122 15.12.2009 10:43:19 signiFicant suBsoil rocK ForMs 11 Tadej SLABE and Hong LIU shapes created on karst surfaces covered by soil or subsoil rock forms deviate from these character- sediment are called subsoil rock forms. istics. under great magnification, the subsoil rock soil or various types of sediment that complete- surface as a rule is distinctly finely rough due to ly or partially cover carbonate rock influence the the even corrosion of the grained rock (slabe, shaping of the rock. Water flowing along the con- 1994). tact between the cover and the rock creates sub- subsoil rock forms were described by Bögli soil channels (table 1) and subsoil scallops. We (1960a) as part of karren. subsoil channels were can distinguish tiny forms at the most permeable introduced by Williams (1966). The importance contact between rock and soil. These include fine- and development of this type of shaping of rock ly dissected channels, small cups, steps, and small was presented by gams (1971) and sweeting pendants. Water that percolates through the soil (1972). Jennings (1973) divides subsoil rock forms forms subsoil cups and solution pipes. When so into those formed by the dissolving of partly much water flows down the rock to the soil or covered or completely covered limestone. in the sediment that it can not all flow away rapidly be- first group, he includes solution notches, and in tween the rock face and the soil or sediment, it the second, deep subsoil solution pipes. nicod carves out half-bells and notches. unique rock (1976) presents the subsoil shaping of rock in the forms also occur due to the oscillation of the level Mediterranean. subsoil scallops are described by of the water table. sauro (1976b). subsoil rock forms, mostly chan- in this chapter we distinguish the subsoil dis- nels, are graphically presented in the atlas of rock section of the rock that is mostly the consequence forms prepared by perna and sauro (1978). Bögli of the rock structure and its fissuring and stratifi- (1981) describes rock forms that occurred under cation, that is, spots of weakness in the rock, from the ground and calls them rundkarren. Fabre and the subsoil rock forms created by the factors men- nicod (1982) amalgamate the knowledge on the tioned above. subsoil shaping of rock. The importance of the Due to the relatively even dissolving of rock shaping of rock under soil is defined by trudgill under soil and sediment, the rock is rounded as (1985), who also describes subsoil cups (trudgill, are the subsoil rock forms, and the surface of the 1986). in his classification of karren, White (1988) rock is relatively smooth to the naked eye or char- singles out those which were covered by soil. Ford acteristically rough on diversely-structured or and Williams (1989) also describe karren shafts recrystalized carbonate rock. only the smallest and wel s, solution pits, and channels that occur 123 KRF•1 • OK.indd 123 15.12.2009 10:43:19 Karst Rock Features • Karren Sculpturing Table 1: Subsoil rock forms. Subsoil rock forms Under soil and sediment At the level of soil Due to the percolation of water Due to the flowing of water along Due to the water flow to the contact Due to the periodically flooded rock through soil the contact between rock and soil channels flutes half-bells scallops cups minute subsoil dissections of the rock notches under weathered debris. Karren formed by water a rule enlarging them outwards when the rock is flowing through the soil is also described by a. covered by fine-grained sediment or soil. Their ginés (1990). He names the karren formed under cross-sections are therefore circular or elliptical the ground subsoil karren (ginés, 1996a) and also along fissures. subsoil cups are often found side- specifies some subsoil forms: tubes and cups. The by-side or already connected. subsoil tubes can importance of the subsoil shaping of rock is also develop from subsoil cups, especially on fissured emphasized by those studying the lunan stone or porous rock. forest, who describe it as a form of covered karst: if the rock becomes exposed, solution pans can sub-jacent after chen et al. (1986), crypto karst also develop from subsoil cups found on horizon- after Maire et al. (1991) and sweeting (1995) and tal surfaces. This development is illustrated by subcutaneous after slabe (1999) and Knez and gams (1971), who calls subsoil cups of this kind slabe (2001a, b, 2002). The rock is often over- covered kamenitzas. grown or criss-crossed by roots (Jakucs, 1977; special subsoil cups form under newly occur- ollier, 1984). traces of this kind are not directly ring weathered debris. as the exposed surface be- classified as subsoil rock forms, but of course veg- comes overgrown, disintegrating vegetation piles etation influences the formation of the described up on the rock, retaining moisture and accelerat- subsoil rock relief. ing the corrosion of the rock. The cups are ini- tially shallow and have gently sloping walls. Their diameters measure from a few centimetres up to Rock forms occurring due to the many decimetres. some have channels through percolation of water through soil and which excess water drains away. on inclined sur- sediment faces, their upper parts are semicircular and wide, and they taper downwards. The cups also form subsoil cups under moss, lichen, or algae that cover the rock in places. under a thinner layer of porous soil that partly or This section could also include subsoil chan- entirely covers the rock, smaller and larger sub- nels formed due to the converging and flowing of soil cups form on horizontal surfaces (Figure 1). the water that percolates through the soil. How- The former are one to five centimetres in diam- ever, due to their characteristic formation, their eter and the latter are larger. They occur due to description matches that of subsoil channels oc- the percolation of water through the soil to the curring due to the flowing of water at the contact rock. as a rule, they form on weak spots in the of the rock and the soil. rock. The water saturates the soil in the cups, as 124 KRF•1 • OK.indd 124 15.12.2009 10:43:19 Tadej Slabe and Hong Liu, Significant subsoil rock forms Figure 1: Above subsoil cups in Lunan stone forest, and below a subsoil cup in south Atlas (Morocco), width of view is 4 m. 125 KRF•1 • OK.indd 125 15.12.2009 10:43:21 Karst Rock Features • Karren Sculpturing Subsoil rock forms occurring due to tion of soil and sediment. Distinctive and diverse the flowing of water along the contact channels are found in the lunan stone forest. song between rock and soil or sediment (1986) attributes the widening of subsoil channels to the mixing of water sliding along the contact subsoil channels and water penetrating through the soil. smaller subsoil channels with diameters of five to twenty subsoil channels form due to the concentrated centimetres criss-cross the wall at various angles flow of water along the contact with the soil. as and are often sinuous. They are equally wide over a rule, the largest channels form when the water their entire length or wider where they intersect runs down vertical or steep contact points. These other channels. They can be joined in a network. large (Figure 2), usually vertical and separate as a rule, the smallest channels, whose diameters channels have diameters from twenty centimetres reach only five centimetres, are the most sinuous. to one metre and more. along fissures, where the Their occurrence is most distinctly influenced as channels are most frequent, they are also deeper, well by the structure and fissuring of the rock on and along the most pronounced, a subsoil shaft which they form. (well) can develop. The diameter of a channel can subsoil channels form primarily through the vary along its length. Deeper under soil and sedi- moistening of the soil and sediment at a permea- ment, large channels frequently narrow. The rock ble contact with the rock and less often by distinc- therefore dissolves most rapidly in the upper sec- tive smaller flows. This is indicated by their shape Figure 2: Subsoil channels. 126 KRF•1 • OK.indd 126 15.12.2009 10:43:23 Tadej Slabe and Hong Liu, Significant subsoil rock forms Figure 3: Subsoil channels. and frequent dissection by horizontal notches. The type of contact between the wall and the soil smaller tubes with a diameter of up to one centi- can differ in places or can change. Meandering metre through which water flows often form at the small channels can therefore occur on the walls of bottom of channels between rock and clay. along larger channels (Figure 3). along less permeable the contact with moist soil, the dissolving of rock contacts, subsoil channels are larger at the level of is more distinct and of longer duration. along the sediment and soil, while below it they quickly with the composition and disintegration of the narrow. gams (1997) identifies the link between rock, the permeability of the contact with the soil the growth of subsoil hollows and the permeabil- and the quantity of water flowing along the con- ity of their filling. in the upper part immediately tact largely dictate the size and shape of the chan- below the surface, such channels most often have nels. it appears that smaller and more meander- funnel-shaped mouths whose diameter can ex- ing channels form along less permeable contacts. ceed one metre. 127 KRF•1 • OK.indd 127 15.12.2009 10:43:24 Karst Rock Features • Karren Sculpturing Figure 4: Subsoil channels. subsoil channels also occur on limestone when they are joined in a branched network. Deeper the latter is in contact with flysch and on the walls channels can have smaller subsoil flutes on their of old roofless caves filled with sediment. When walls. in the lunan stone forest, unique channels water runs on the contact through a narrow crack with semicircular or upside-down omega-shaped in the rock, the channel widens several dozen cen- cross-sections form at the bottom of cracks be- timetres below the surface. Here and there, sub- tween rock pillars or teeth where fissures wedge soil scallops also occur below narrow mouths. out (Figure 5). on more or less gently sloping rock covered by special channels (Figure 6), called hohlkarren soil, channels (Figure 4) with semicircular bot- by sweeting (1972), form when they are filled with toms develop, being frequently described in the soil or when their bottoms are covered and the literature as rundkarren (sweeting, 1972; perna rock around them is bare. in most cases, they have and sauro, 1978) and subsoil runnels (trudgill, the characteristic shape of an upside-down greek 1985). These are the consequence of the joining of letter omega. They can have several stories. as the water percolating through the soil. on steep sur- level of the soil dropped, it remained only on the faces, they can be parallel (Williams, 1966), and bottom of the channels and thus deepened and we can talk about subsoil flutes since the water widened them. channels also lead from subsoil percolating through the soil flows evenly over the cups formed where water streams flow together. entire surface. on gently sloping rock surfaces, They often developed from subsoil tubes that were 128 KRF•1 • OK.indd 128 15.12.2009 10:43:26 Tadej Slabe and Hong Liu, Significant subsoil rock forms Figure 5: Subsoil channels. Width of view is 1.5 m. uncovered when the upper strata of rock disinte- formerly uncovered karst becomes overgrown grated. These become subsoil channels from the again (Jennings, 1973). This is also characteris- moment the level of the soil surrounding the pil- tic of the classic Karst region of slovenia where lars drops lower than the channels. We can often weathered debris or a thick layer of moss is ever trace the transition from channels formed on rock more distinctly covering the rock. completely covered with soil to those where only in the areas of locally flooded zones, above-sed- the channels are covered. after uncovering, when iment channels and anastomoses (slabe, 1992) can they contain no more soil, they are transformed occur. These are also characteristic of the lower by rainwater. subsoil channels start to reshape the planes of the basal, carbonate conglomerates in rock, including the rain-created rock forms, when flysch (Figure 7). 129 KRF•1 • OK.indd 129 15.12.2009 10:43:28 Karst Rock Features • Karren Sculpturing Figure 6: Subsoil channels. Width of view is 3 m, in the middle. Figure 7: Subsoil anasto- moses. Width of view is 1.5 m. 130 KRF•1 • OK.indd 130 15.12.2009 10:43:32 Tadej Slabe and Hong Liu, Significant subsoil rock forms Figure 8: Subsoil scallops. subsoil scallops as a rule, subsoil scallops do not form along distinct fissures that criss-cross walls or along subsoil scallops (Figure 8) occur due to the flow bedding planes where semicircular channels de- of water along the entire-permeable contact of the velop instead. rock with the soil. These are subsoil cups with di- along-sediment cups are also frequent on the ameters between fifteen and fifty centimetres con- walls of caves filled with fine-grain sediment. nected in a network. They are shallow and most often a little deeper in their upper part. as a rule, they are found on overhanging rock surfaces. on Rock relief of subsoil karren that is distinctly overhanging surfaces, they can be ar- periodically flooded ranged one above the other like fish scales (Fig- ure 9). Their narrower lower parts protrude from The peaks of subsoil karren (Figure 10) that are the rock wall. periodically reached by the water table and are en- subsoil scallops can be observed along distinct tirely formed below the ground are sharp. rela- fissures where cracks filled with soil occur. They tively smooth rock characteristic of formation be- are also characteristic of overhanging walls of pil- neath soil and fine-grained sediment dominates lars in the stone forests. The circumference of sub- the upper part. subsoil notches are most pro- soil notches is often dissected below the ground nounced in the lower part of the karren. larger by subsoil scallops. in places, channels lead to the horizontal notches reach up to one metre in di- higher subsoil scallops that are larger in size. ameter, and smaller ones are found one above the 131 KRF•1 • OK.indd 131 15.12.2009 10:43:33 Karst Rock Features • Karren Sculpturing Figure 9: Subsoil scallops. Width of view is 2.5 m. other. semi-panned notches are the conclusions the rock, most often minute fissures, subsoil cups of vertical subsoil channels that formed along the form that in time can grow into tubes. Between most conductive paths. The individual peaks of the cups and channels, there are subsoil tubes subsoil teeth above the most distinct notches are criss-crossing the rock at various angles. spongy. subsoil channels on this karren can be This type of formation of subsoil karren is il- divided into vertical and horizontal. The former lustrated by an experiment using plaster pillars are conductors of the oscillating water table along that we covered with soil and then exposed to ar- the most conductive paths. The latter, which criss- tificial rain. The water drained from the model at cross more sloping rock and larger rock surfaces, the bottom. The upper section of the pillars was are further formed by the moisture that remains shaped by the water that percolated through the in them the longest after the lowering of the level soil in a dispersed fashion, while the lower part of the water table. similarly, along weaknesses in was shaped in a locally flooded zone. The outflow 132 KRF•1 • OK.indd 132 15.12.2009 10:43:35 Tadej Slabe and Hong Liu, Significant subsoil rock forms Figure 10: Subsoil karren occasional y flooded. of the water was too slow and the water therefore soil but does not fill completely and dolines that accumulated at the bottom part of the model. are thinly covered with soil are often dissected by to sum up, two dominant processes for the unique subsoil cups, while overhangs are dissect- formation of this type of subsoil karren can be ed with ceiling pendants. on gently sloping sec- deduced from the shape of the karren and its tions in such conditions, the rock is dissected by rock relief. The rock forms that are the traces of steps. subsoil cups (Figure 11) are semi-circular frequent oscillation of the level of the water table or oblong and arranged in steps. Their diameters that floods the karren from below give it a special measure between 0.5 and two centimetres. The stamp. When the water table is low, the karren is largest are combinations of smaller ones. in all shaped by the water that periodically and dispers- cases, the rock forms are connected in a network. edly percolates from the surface through the soil Their formation and shape are primarily the con- and slides evenly down the rock. it remains longer sequence of the characteristics of the structure of in the subsoil cups and gently sloping channels the grained rock and the inclination of the wall and along the less permeable contact between the along which the water slides. in such conditions rock and the sediment surrounding it. even the less distinct channels that often form in cracks are minutely dissected with subsoil cups. it appears that the water sliding down the rock Minute subsoil dissection of rock carries soil and deposits it at individual points in subsoil cups on vertical surfaces, on steps on The walls of cracks through which water carries gently sloping surfaces, and on pendants on over- 133 KRF•1 • OK.indd 133 15.12.2009 10:43:37 Karst Rock Features • Karren Sculpturing Figure 11: Subsoil cups. Width of view is 0.5 m. hangs. Moist sediment in subsoil cups can cor- are dissected in the rock relief by above-sediment rode the rock more effectively since the water ac- (slabe, 1995a) and under-sediment channels. cumulates in the sediment and remains in it for smaller tubes with diameters between one cen- a longer period. on inclined surfaces, the water timetre and one decimetre pierce the rock in vari- deposits sediment on the most gently sloping sec- ous directions and are often connected into a sys- tions. old sediment protects them from corrosion, tem. Most form on distinctly fissured or porous while the rock lying next to them corrodes more rock, and vegetation often plays an important role quickly and is therefore dissected into steps. on in their formation. overhanging surfaces, the sediment collects on the pendants and protects them from dissolving. Rock forms that occur at the level of the soil or sediment subsoil tubes subsoil notches The rock below the ground is often criss-crossed by tubes of various sizes, karst hollows that dur- subsoil notches form due to the corrosion of the ing their formation are filled with sediment or soil. rock along a long-lasting level of sediment or soil They are of various size and shape. The larger ones surrounding it. The water flows to the contact over 134 KRF•1 • OK.indd 134 15.12.2009 10:43:39 Tadej Slabe and Hong Liu, Significant subsoil rock forms Figure 12: Subsoil notch. a larger surface, more or less distinctly corrodes it, lar fashion. The upper part of the notch is reshaped and then flows away between the rock and the soil. due to water sliding down the rock. smaller semi- smaller subsoil notches with diameters between circular notches develop first and can then grow ten and twenty centimetres have the shape of semi- increasingly larger with the slow lowering of the circular horizontal channels, only with their upper level of sediment. The notches can be seen at vari- edges most often being sharper and the lower ous heights on rock that was rapidly and sporadi- edges more rounded. larger (Figure 12) subsoil cally exposed. smaller exposed notches are more notches: undercut notches (Waltham, 1984; Ford et reshaped by rainwater, and larger ones less dis- al. 1997), solution notches (Jennings, 1973), swamp tinctly so. The water that shapes them often flows undercut (ollier, 1984) are corroded one metre or between the rock and soil below them and creates more into the rock, and in the lunan stone forest subsoil channels or scallops. they are frequently up to one metre high. The lower The forms described above are distinguished part of the notches is undercut. The rock was sub- from notches formed due to the often more rapid ject to faster, long-term dissolving below the moist dissolving of rock along horizontal bedding ground and is therefore rounded and smooth. The planes. as a rule, the latter are narrower and very lower section of the notch is horizontal, while the frequently relatively deep relative to the diameter upper tapers out toward the lower in a semicircu- of the opening. 135 KRF•1 • OK.indd 135 15.12.2009 10:43:40 Karst Rock Features • Karren Sculpturing Figure 13: Subsoil half-bel . Width of view is 5 m. subsoil half-bells half-bell shapes whose shape and size are related to the quantity of water flowing to the soil, the Half-bells (Figure 13) form below channels that permeability of the contact between the rock and continuously bring larger quantities of water to soil, and the speed at which the level of the soil the sediment or soil surrounding the rock. The or sediment lowered. The upper part of a channel contact is not conductive enough for all of the can also be a tube where it formed along a distinct water that reaches it. There are large and distinct fissure, while the wall is only corroded along the half-bells in the lunan stone forests. above the widenings. immediately below the ground, the soil or sediment there are characteristic bell or walls of large bells can be dissected by oblong sub- 136 KRF•1 • OK.indd 136 15.12.2009 10:43:42 Tadej Slabe and Hong Liu, Significant subsoil rock forms soil scallops that reach up to one metre in diame- surfaces and on the basis of their porosity and the ter. Deeper under the ground, bell-like widenings height of fossils and chert protruding from them often narrow gradually into subsoil channels. determined the degree of the corrosion of the rock. Surface of subsoil rock forms Conclusion if the rock is relatively evenly structured, the sur- The sediments and soils covering rock in layers of face of subsoil rock forms is usually smooth. With varying thickness have different structures. This the great magnification of a scanning electronic influences their permeability and the permeabil- microscope, however, we can see that it is minute- ity of the contact between them and the rock and ly dissected, the consequence of the faster dissolv- the manner in which water flows through them ing of the rock at weak spots, that is, at the contact and along the contact. The rock and its compo- between the various particles that compose it. The sition, stratification, and fracturedness determine smoothness or roughness of the rock is of course the development of subsoil rock forms and their influenced by its composition and fracturedness. appearance and surface. slowly dissolving particles, which can protrude subsoil rock forms are a distinct and indicative distinctly from the rock surface, remain on it sign of the formation of rock sculpturing under (slabe, 1994) even though they are subject to fast- the ground and often an important trace of the er dissolving due to their larger exposed surfaces. development of the karst surface and its use. trudgill (1985) measured the roughness of rock 137 KRF•1 • OK.indd 137 15.12.2009 10:43:42 KRF•1 • OK.indd 138 15.12.2009 10:43:42 KaMenitzas 12 Franco CUCCHI Kamenitzas are closed depressions that develop depth is always a fraction of their width. They are on karst surfaces exposed to the atmosphere. The usually rounded in shape, with decimetre-scale presence of static water produces small, round, plan dimension and centimetre-scale depth. Most closed pans (Figure 1) that are shallow as com- have a diameter ranging between 4‒5 cm and 1‒2 pared with their depth. m, even though some up to 6 m have been docu- Kamenitzas form on horizontal, or slightly in- mented (Bryan, 1920). in a palaeokarst surface clined, undulating surfaces where water does not in the upper pennington formation limestone flow but collects into the small depressions. Their (Mississippian-latest chesterian), Humbert and genesis is usually more controlled by micro-relief Driese (2001) found a “flat-floored kamenitza with patterns rather than by discontinuity of surfaces. a width of 7 m and a depth of only 0.5 m”. in other The floors of kamenitzas are usually planar and palaeokarst deposits, the aran limestones in scot- horizontal but sometimes show minute irregu- land, Vincent found some “palaeopits”, remains of larities and rough protuberances. Their walls are karst depressions (personal communication). sub-vertical or slightly inward leaning and their Their genesis has been widely discussed. ac- Figure 1: Perfectly round-shaped kamenitzas in Cretaceous micrites of the Classical Karst (left), width of view is 85 cm, and (right), width of view is 50 cm, in Cretaceous breccia of Kornat island (Croatia). In one of them the overflow channel favours the formation of notches. 139 KRF•1 • OK.indd 139 15.12.2009 10:43:45 Karst Rock Features • Karren Sculpturing cording to the major theories, they could be ei- 1972), true solution pan (gams, 1974), solution ther essentially chemical (gavrilović, 1968; Forti, pan (Ford and Williams, 1989); 1972) or essentially biochemical, and thus linked • French: lapiés à nid de poules (gèze, 1973); to the action of endolithic algae (perna and sauro, • german: Napfkarren (Bögli, 1960a); 1978). some researchers are in favour of an ex- • italian: vaschette di corrosione; clusively corrosive origin and therefore take into • polish: miseczka krasowa or kamenica; consideration increases in growth or changes in • slovene: škavnica (also kamenica); shape, which are linked to phytokarst phenomena • croatian term see cvijić, 1924; (Belloni, 1969; Belloni and orombelli, 1970) or to • spanish (also caribbean): tinajitas (udden, the presence of dissolved deposits (Bögli, 1960a). 1925). other researchers (gams, 1974) differentiate two types of kamenitzas: those formed by means of subcutaneous corrosion on primary depressions, Evolution, lithological and structural and those formed on bare carbonate surface, control which might be related or not related to pre-exist- ing subcutaneous depressions. When the surface depression fills up during pe- some consider them semi-covered landforms, riods of rain, the rainwater lasts there until total assuming their formation has already started evaporation occurs. The dimension of the pans while they are still covered by clods of earth or (probably also linked to climatic conditions and rock fragments. Kamenitzas, in fact, are often to rainfall regime) almost always allows for their present on fairly smooth surfaces, whose present- nearly total filling and sometimes for marginal day undulations are likely to have been inherited overflow that form effluent runnels (Figure 2). in from covered karst conditions. soil erosion and time, the water collecting inside the pan, becomes exhumation then leads to the development of ka- more aggressive on the pan walls rather than on menitzas and landforms of dynamic corrosion its floor. The solute, in fact, accumulates on the origin such as rinnenkarren and ril enkarren. pan floor due to its density and insoluble impuri- even though the term kamenitza or (plural) ties and other material (mainly silt) caught in the kamenitzas – kamenica, kamenice and also ka- depression “protect” the floor, thus decelerating minitza (the term seems to come from the wrong the solution process. Moreover, carbon dioxide hypothesis according to which a small pebble ‒ diffusion from the atmosphere rapidly decreases kamen in serbo-croat ‒ left on a limestone sur- with depth, thus leading quickly to over-satura- face led to the formation of a depression. cvijić, tion. The depression walls are more subject to the 1924; gavrilović, 1968) is now widely accepted, water action because superficial water, still inter- below are some corresponding terms used in dif- acting with the atmosphere, replaces over-satu- ferent countries and languages (rose and Vincent, rated and dense water, which seeps downwards, 1986b): mixing with silt on the floor of the pan. This ex- • Brazilian, portuguese: marmite; plains why kamenitzas tend to widen rather than • czech and slovak: kamenice; to deepen (Figure 3). • english: solution cup or solution pan (Zotov, organic material or fine sediments that either 1941), rock tank (Bryan, 1920), lapiés pot- float on the water or are transported by the wind hole, etched pothole (udden, 1925), solution can accumulate in the depression, inducing pro- pit (Wenthworth, 1944), solution pan (Fry and longed humidity and therefore corrosion on the swineford, 1947), corrosion basin, clint pool bottom thus explaining the evolution of the typi- (sweeting, 1966), rock pool (Williams, 1966), cal undulating floor and other minute concave- corrosion cup and solution basins (sweeting, cup depressions. 140 KRF•1 • OK.indd 140 15.12.2009 10:43:45 Franco Cucchi, Kamenitzas Figure 2: A kamenitza and pits of Gait Barrow, UK (photo P. Vincent). The same result, often accompanied by limited from rock fragments to twigs, from seeds and pol- furrows of parietal corrosion origin, can be linked len, to decaying birds and small mammals. to the phytokarst, and the biological activity of The presence of such material on a damp sur- fungi, lichens, algae, etc. crowther (1997) shows face can favour the formation of a kamenitza. that individual karren possess distinctive rough- clayey deposits and organic accumulations (such ness characteristics that seem to be attributable to as leaves and twigs, for instance) restrain water differences in the (assumed) nature of water flow, and induce some prolonged humidity in contact i.e. greater turbulence inflow along rillenkarren with the rock, thus leading to static corrosion phe- and unstepped rinnenkarren produces rougher nomena. surfaces than those found on steps and stepped spillage water usually overflows and creates a flats, and to the presence or absence of litter/ solution runnel – also decantation runnel or over- humic soil fill within kamenitzas. flow channel. at the outlet point, karst phenom- in this respect, it ought to be noted that kamen- ena become dynamic, thus forming a furrow that, itzas act as “traps”, which can accumulate sedi- often, lowers rapidly, to the extent that it reaches ments either transported by the wind, or runoff the level of the kamenitza bottom. in this case, ka- water “captured” by the depression (Figures 4, menitzas do not widen any longer, thus becoming 5, 6). The sediments can vary in origin, ranging accumulation cups, which feed karst furrows. as 141 KRF•1 • OK.indd 141 15.12.2009 10:43:46 Karst Rock Features • Karren Sculpturing 1 solute accumulation 2 silt and impurites solute accumulation 3 max water level solution runnel clast organic material, micro-notche soil max water level 4 max water level solution runnel clast organic material, micro-notche soil organic material, soil 5 Figure 3: In the kamenitza the wal s evolve more than the bottom because of solute and other materials accumu- late on the floor. A. organic material, soil, clasts and biological activity induce a sort of subcutaneous corrosion on the bottom and cause undulating or minute depressions; B. micro-notches along the pan wal s form when the so- lution runnel – the over flow channel – progressively deepens. 142 KRF•1 • OK.indd 142 15.12.2009 10:43:47 Franco Cucchi, Kamenitzas a b c d e f Figure 4: Kamenitzas of Classical Karst: a. filled by debris and soil; b. by “terra rossa”; c. by a smal Juniperus plant; d. by foliage; e. by mosses (width of view is 65 cm); f. kamenitza of Fanes Natural Park (Dolomites, Italy) at 2,288 m a.s.l. partial y filled by soil and polished by snow during winter (width of view is 70 cm). 143 KRF•1 • OK.indd 143 15.12.2009 10:43:54 Karst Rock Features • Karren Sculpturing the corrosion processes no longer affect the whole wall but only its wet parts. sometimes real micro- notches form (Figures 2, 4b). This term is bor- rowed from a term used to describe the peculiar landforms that originate along the coast, at the sea level, due to bio-corrosion or bio-alteration of rocky walls. a particular development characterizes the kamenitzas resting along the shoreline (Figure 7) as, beyond the “standard” ones, there are other factors coming into play, such as those linked to water swash, spray zone, mixing corrosion, the Figure 5: Kamenitzas of Viñales Park, Cuba. mix of salted waters, endo- or exolithic organ- isms and the intense biological activity (perica et al., 2004). They may also be remarkable in size and are characterized by irregular and indented floors and rims, due to corrosion (dynamic, static and point corrosion) caused by water mixing, bio- karst phenomena and erosion caused by wave mo- tion. it is possible to recognize several phases in the development of kamenitzas: formation, develop- ment, degradation and disappearance (Figure 8). The first phase consists in the corrosion caused by waters in the depressions following denudation. corrosion continues and, during the second phase, lateral widening occurs, sometimes accompanied by outflow and/or runnel channels at the same level as the bottom of kamenitzas (Figures 3, 4a, 8). Therefore degradation occurs due to the widening of the outflow groove, and the kamenitza margin gradually lowers, thus becoming rounded. Finally, the kamenitza disappears due to the progressive overhanging of the walls and to the widening of the outflow channel (Figure 3). The development of kamenitzas can also stop when an open frac- ture is encountered during the development phase. Figure 6: Miseczka krasowa filled by ice during winter oc- The water, in fact, can widen it and then disappear curred in Upper Jurassic limestone of Cracow-Czesto- in the infiltration point (Figure 2). chowa Upland, south Poland (photo A. Tyc). experimental studies on the evolution rate of kamenitzas are scarce. sweeting (1966) observed on an experimental site in northern england an the solution runnel deepens, the level of the water increase in depth up to 3‒5 cm within less than present in the depression progressively decreases ten years. it should be noted that such rates are ex- and the walls tend therefore to overhang, because ceptional. Zotov (1941) hypothesized that solution 144 KRF•1 • OK.indd 144 15.12.2009 10:43:57 Franco Cucchi, Kamenitzas Figure 7: Kamenitzas along the shorelines of the La Habana, Cuba, harbour (left) and of the Mali Lošinj island, Croatia (right). Figure 8: Elongated kamenitza characterized by complex evolution. The original elongated form first evolved into a form with a drainage channel that originated lateral notches and karren runnels. It deepened afterwards at its cen- tre. At present, it also features two round sub-depressions (similar to corrosion cups) favoured by sediments and or- ganic material. cups would be completely destroyed in a few thou- Measurements in the classical Karst (cucchi et sand years. rose and Vincent (1986b) estimated al., 1987, 1990) proved that the floor of a kamenit- that the time required by a kamenitza to form and za has a lowering rate of approximately 0.02–0.03 reach a depth of 10 cm and a radius of 20 cm was mm/year: a 4‒5 cm deep kamenitza would need at approximately 3,260 years. least 2,500 years to be completely formed. 145 KRF•1 • OK.indd 145 15.12.2009 10:44:01 Karst Rock Features • Karren Sculpturing ready-formed kamenitzas can be found on surfaces that are currently exposed to weathering: they may belong to two or even three periods of formation because of their size range. one might think that they form and evolve rapidly until they reach a “standard” size determined by climatic and geographic conditions, and develop more slowly afterwards. in this sense, they are equilib- rium landforms. as with all landforms linked to chemical disso- lution (but also for those stemming from erosion), the morphology of kamenitzas is conditioned by Figure 9: Kamenitzas conditioned by discontinuity of the the lithological and structural characteristics of Classical Karst. Note the secondary central depressions corroded by sediments accumulation. Width of view is the host rock, i.e. by that complex set of factors 55 cm. Figure 10: Elongated kamenitzas reach- ing a depth of up to 1.5 m on the lime- stone pavements near Borgo Grotta Gigante (Classical Karst) partly condi- tioned by discontinuity planes perpen- dicular to sub-horizontal stratification. 146 KRF•1 • OK.indd 146 15.12.2009 10:44:05 Franco Cucchi, Kamenitzas that leads to the so-called selective corrosion (Fig- ures 4e, 8). in case of fine-grained homogeneous rocks (such as micritic limestone), landforms are smooth, regular and symmetrical; if the rock is coarse-grained and heterogeneous (as in the case of breccias), the forms originated by selec- tive corrosion are less regular (Figure 1). also the presence of organic residues within the rock may cause roughness and irregularities (Figure 8). it is in fact known that the dissolution also depends on the caco crystal size, as its rate inversely pro- 3 portional to grain size. experiments with micro- erosion meter on classical Karst have been made Figure 11: A “corrosion cup” of Perolas cave (Iporanga, São since 1970, showing lowering rates ranging be- Paolo, Brazil), with diameter amounting to approximate- tween 0.04 and 0.01 mm/year, with annual rainfall ly 30 cm and depth to nearly 20 cm (photo P. Forti). amounting to approximately 1,350 mm/year in Mediterranean climate (cucchi et al., 1987, 1996). The lowering rate can vary: it is higher in case of micritic limestone (mudstones), lower in case of which traverse the pavements in low-altitude sparitic limestone (rudstones or grainstones) and limestone pavements in the British isles of gait even lower in case of dolomitic limestone and Barrows national nature reserve. Dissolution limestone dolomite. selective corrosion phenom- and fracture of the calcite veins at the bases of a ena that highlight the petrography and structure kamenitza eventually leads to water leakage and of affected limestones can therefore manifest. the termination of kamenitza development. The shape, or plan, of kamenitzas is also condi- tioned by the discontinuities present in the rock mass. if fractures lead to higher porosity and per- Other similar forms meability along a plan intersecting a kamenitza, the intersection becomes the preferential direc- Corrosion cups are peculiar formations, similar to tion of development, thus originating elongated kamenitzas, recently described in slovenian (Mi- forms, also called linear or elongated kamenitzas hevc, 2001: korozijske kotlice) and Brazilian (Forti (Figures 9, 10). Joints can simply condition the et al., 2001: marmitta da corrosione) caves. They alignment of a kamenitza when it forms on them. are similar in geometry to some peculiar “pot- if this happens, the fracture is persistent and in- hole-like” forms (Figure 11), described in non- volves the whole stratum. When fractures are less carbonate rocks (White, 1988; Opferkessel or solu- persistent, they can either condition only one of tion basins in granites). in the depressions formed the kamenitza sides, or originate complex-plan on microcrystalline lightly dolomitic limestone, kamenitzas if two or more fractures families are sand deposits with silt, essentially composed of present. Fracture sets can therefore differently quartz or silicates saturated with water rich in affect the size and morphology of kamenitzas; organic substances, favour static corrosion proc- only non-persistent fractures cannot lead to the esses that preferentially affect the bottom: the dis- evolution of karst grikes or holes. rose and Vin- solved calcite diffuses in the solution (Figure 12). cent (1986b) describe active and fossil kamenitzas Depressions whose shape evolution is heavily con- strongly controlled by the presence of calcite veins ditioned by the sediment accumulations on their 147 KRF•1 • OK.indd 147 15.12.2009 10:44:06 Karst Rock Features • Karren Sculpturing   Figure 12: The genetic model of a “corrosion cup” according to Forti et al. (2001). Solution cup evolution is controlled both by the degree of under-saturation and the time during which the corrosion process is active. If these parame- ters remain constant in time, the process leads to a progressive tapering of the equilibrium cup floor with the evo- lution of a cylindrical section. If the degree of super-saturation and/or the time of activity increase, the equilibrium diameter of the cup is forced to increase and, consequently, the cup starts evolving into a cone shaped depres- sion. During a flood, part or whole of the sediment fil ing the depression is washed away. At the end of the flood, sand and organic material are deposited on the bottom of the depression (pre-existing ril flutes, scallops and the already developed corrosion cups), which is filled with water. The oxidation process of the organic material releases carbon dioxide, which in turn causes the dissolution of the limestone located at the bottom of the depression. This process lasts until water floods the depression again. The dissolved calcite migrates through diffusion layer and at the end of the process, accumulates on top of the sandy sediment. Genetical y, corrosion cups may be subdivided into three zones: in the lower one, corrosion is induced by the oxidation of the organic material; in the intermedi- ate section, the transfer of CO and Ca2+ towards the air-water interface and the sinking of calcite rafts towards the 2 top of the sand fil ings are dominant; in the upper zone, CO diffusion in the atmosphere, and evaporation lead cal- 2 cium carbonate to deposit mainly as calcite rafts but sometimes also as thin calcite crusts. floors and by environmental characteristics are, in Venezuela, and ethiopia), and in metamorphic fact, not uncommon. This is basically the case in rocks. tropical and sub-tropical environments, where or- on granites, for instance, flat-bottomed, sub- ganic substances and co are particularly impor- circular depressions similar in genesis (but not in 2 tant for the evolution of karst phenomena. shape) to the so-called “tafoni”, can originate on landforms similar to those typical of carbonate the rocky walls, owing to attack by solution weath- rocks (White, 1988) have been observed on non- ering. sub-horizontal landforms comparable to carbonate rocks: in granites (isola d’elba, italy, kamenitzas could be the result of a concomitance and antarctic), in basalts, quartz-feldspar sand- of factors, such as selective corrosion in femic dif- stones with calcite or quartzite cement (tepuys, ferentiates (involving the contact with the host 148 KRF•1 • OK.indd 148 15.12.2009 10:44:06 Franco Cucchi, Kamenitzas Figure 13: A cup in the pink granite of Capo d’Orso, Sardinia, Italy. Figure 14: Femic nodules in riolite, Atacama desert, Chile, because alteration generate cups: pseudo-kamenitzas? 149 KRF•1 • OK.indd 149 15.12.2009 10:44:09 Karst Rock Features • Karren Sculpturing granites), in areas with different crystallization sands “playa basins” (small, roughly circular to (both regarding crystal size and regarding adher- oval, internally drained depressions), whose gen- ence and cohesion). Furthermore, they can be in- esis is conditioned by a concurrence of geomor- duced by progressive alteration and disintegration phic, pedogenic, hydro-chemical and biological due to temperature fragmenting, to the swelling processes, state that they are economically “im- of hydro-sensitive crystals, to silicate feldspar ar- portant because they collect runoff and recharge gillification, as well as to bio- and phyto-alteration the aquifer”. (Figure 13). interesting pseudo kamenitzas origi- nate from femic nodules alteration in riolites or in ignimbrites (Figure 14). Acknowledgements to conclude, it is undeniable that, the aggres- sion mechanisms can vary and interact in various i would like to thank paolo Forti, ugo sauro and ways with the different rocks involved, neverthe- peter Vincent for having critically and construc- less they can generate very similar landforms. tively contributed to the drawing up of my paper, gustavson et al. (1995), recognizing tens of thou- and silvia Mancaleoni for the english translation. 150 KRF•1 • OK.indd 150 15.12.2009 10:44:09 trittKarren 13 Márton VERESS Trittkarren are steps that develop on bare slopes from about 3 cm to 5 cm high, while the tread may (Figure 1). This karren form is also called step kar- be from about 20 cm to 100 cm wide. according to ren (Werner, 1975) and heelprint karren (Bögli, Haserodt (1965), trittkarren occur at altitudes be- 1980). trittkarren can be from 2 cm to 25 cm wide. tween 1,900 and 2,200 metres in the alps. trittkar- according to sweeting (1973), the riser may be ren occur mostly on surfaces with a small dip angle. according to sweeting (1973), they can de- velop in the initial phases of surface karren forma- tion. trittkarren occur on marble (Vincent, 1983a), on gypsum (calaforra, 1996; Macaluso and sauro, 1996a), and on sandstone (Veress, 2003). trittkar- ren have a riser, a tread, and a foreground (Vincent, 1983a; Veress and lakotár, 1995). The riser is usu- ally curved and surrounds the tread. The tread is the flat, more or less horizontal section, the riser is the vertical surface at the back, and the foreground is the sloping section at the front of the tread that is not surrounded by a riser. We can distinguish the following trittkarren types: • embrionic trittkarren differ from developed trittkarren. Their shape is varied, they are only partly like developed trittkarren, and they are smaller (Figure 2); • nischenkarren (Figure 3) are a type of trittkar- ren whose wide tread is surrounded by a small riser-system that is interconnected (Haserodt, 1965); • trichterkarren (Figure 4) are a type of trittkar- Figure 1: Trittkarren and rinnenkarren on bedding plane ren whose tread is missing (Bögli, 1951); surface (Dachstein). • uvala trittkarren (Figure 5) are a type of 151 KRF•1 • OK.indd 151 15.12.2009 10:44:09 Karst Rock Features • Karren Sculpturing Figure 2: Embryonic trittkarren; the risers of several are more arched and their treads wider; below a grike on “Ausgleichsfläche” (Totes Gebirge). Width of view is 2.5 m. Figure 4: Trichterkarren (Dachstein). Width of view is 0.5 m. 1. trichterkarren. Figure 3: Almost closed “Nischenkarren” (Totes Gebirge). trittkarren whose risers are coalescing (Veress, 2000a); • Veress et al. (2006) described ripplekarren on marble (from Diego de almagro island). The risers of this type are straight and are not curved. although the risers of this form are straight and not curved, these forms are indeed small steps and “ripplekarren” can therefore be considered as trittkarren; • step-trittkarren develop around or on the mar- gins of grikes and shafts. Their form is semi- circular. The treads and risers occur in a series, and the risers of separate “step-trittkarren” can develop parallel to each other (Figure 6); • there are karren forms that have treads but do Figure 5: Uvala trittkarren (Dachstein). Width of view is 0.5 m. 1. uvala trittkarren. not have risers (Figure 7). They occur on very 152 KRF•1 • OK.indd 152 15.12.2009 10:44:10 Márton Veress, Trittkarren Figure 6: “Step-trittkar- ren” with narrow steps (Totes Gebirge). Figure 7: Trittkarren- like forms (Dachstein). Width of view is 0.5 m. 1. trittkarren like forms; 2. wandkarren. steep slopes. These forms are like trittkarren Methods of trittkarren research but they may also be a special variety of Spitz karren. choppy (1996) calls these forms tetra- near the svartisen glacier, Vincent (1983a) inves- hedron karren. tigated the relationships between the values of trittkarren occur in rinnenkarren, in kameni- various parameters of trittkarren (Figure 8), an- tzas, on Ausgleichsfläche, and on ridges between alyzing the distribution frequency of six param- rinnenkarren. They occur on bare rock as well. eters (Figure 9). He grouped the data into inter- 153 KRF•1 • OK.indd 153 15.12.2009 10:44:11 Karst Rock Features • Karren Sculpturing 40° slopes: the height of the riser, the angle of the tread, the length of the tread, the width of the riser, the length of the arch of the riser, and the width of the foreground. assuming that trittkarren of dif- ferent sizes in the same place represent different phases of development, Balogh (1998) describes the development of trittkarren as follows: the riser of the trittkarren develops evenly on slopes with a small inclination and the tread of the trittkarren lowers evenly as well. The treads of the trittkar- Figure 8: Morphological parameters describing the ren do not lower on slopes with a medium incli- shape of trittkarren (according to Vincent, 1983a). MBA. maximum angle of riser; SA. average angle of side wal s; nation. The riser develops mainly in its central VRC. radius of curvature of riser tread junction at center; section. The lower part of the riser dissolves more, HRC. maximum horizontal radius of curvature of riser; and therefore the riser becomes steeper during its MW. maximum horizontal width; VH. vertical height of development. The upper part of the riser dissolves riser. if the angle of the slope is about 40°. The riser be- comes gently sloped during its development. vals. The maximum values are the following with the boundary of the interval shown in paren- Morphology of trittkarren thesis: VH 5‒6 cm (2‒10 cm), Vrc 2‒8 cm (2‒18 cm), Hrc 8‒10 cm (4‒18 cm), MW 12.5‒15 cm trittkarren occur in groups. according to Bögli (4‒21 cm), MBa 50°–60° (30°–80°), and sa 40°– (1951) and Haserodt (1965), the tread of a trittkar- 50° (10°–80°). The meanings of the abbreviations ren is an “ausgleichsfläche”. according to Bauer are shown in Figure 8. Vincent was able to show (1962), there is a close relationship between the a linear function relationship using regressional angle of the bearing slope and the morphology of calculation. For example, the form of the function trittkarren. if the angle of the slope is small (less between the MW and the VH parameters is VH than 10°), the height of the riser will be one or two = –2.27 + 0.53 MW, where r = 0.575 is the value centimetres, and the width of the tread a few deci- of the correlation coefficient. He created a matrix metres (Figure 3). if the angle of the slope is be- by using the correlation coefficients, and we pres- tween 10° and 30°, the riser and the tread will be ent the components of the matrix in table 1. The larger (the height of the riser will be several centi- asterix beside a number indicates a 95% reliability metres, and the width of the tread one or two deci- level of the correlated components in the matrix. metres, Figure 5). if the inclination of the slope Working in the totes gebirge mountain range, is large (above 30°), the height of the riser will be Balogh (1998) measured the following parameters several decimetres while the width of the tread of 240 trittkarren occuring on 10°–20°, 25°, and will be one or two centimetres. The riser can be gently sloping (Figure 10a) or steep (Figure 10b). Viewed from above, the curve of the riser can be Table 1: Matrix of correlation coefficients (Vincent, 1983a). short (Figures 4, 10a) or long (Figures 5, 10f). in VH – the first case, the length of the riser is smaller than MBA .50* – a semi-circle; in the latter case, its length is longer VRC .71* .17 – MW .58 .41* .50* – than a semi-circle. The riser may also be almost HRC .32* .26* .33* .46* – closed (Figure 10c). The curve of the riser can SA .24 .30* .06 .01 –.24 VH MBA VRC MW HRC be almost straight (Figure 10d) or angular like a 154 KRF•1 • OK.indd 154 15.12.2009 10:44:11 Márton Veress, Trittkarren            ­€‚ƒ „ ˆ‰  Š ­ƒˆ      „ †‡†ˆ‰  ‹ ‰ˆƒ Figure 9: Histograms of the six morphological variables, n = 45 (from Vincent, 1983a). 155 KRF•1 • OK.indd 155 15.12.2009 10:44:11 Karst Rock Features • Karren Sculpturing Figure 10: Model trittkarren (from Veress and Tóth, 2002). a. trittkarren with gentle riser; b. trittkarren with steep riser; c. trittkarren with almost circular riser arch; d. trittkarren with straight riser; e. trittkarren without tread with uvala trittkarren below it; f. trittkarren with ril s on riser; g. trittkarren with ril s on tread; h. trittkarren with sharply angled riser and gentle tread; j. trittkarren with undulating tread; k. trittkarren with elongated tread; l. trittkarren with ridge and trittkarren on its tread; m. trichterkarren. 156 KRF•1 • OK.indd 156 15.12.2009 10:44:12 Márton Veress, Trittkarren Figure 11: Trittkarren with riser similar to a right angle; below it are trittkarren with peaks on their treads (Dachstein). Width of view is 0.5 m. 1. trittkarren similar to a right angle. right angle (Figures 10h, 11). Ril enkarren (Fig- Development of trittkarren ure 10f) or secondary (small sized) trittkarren can occur on a riser. The tread may be horizontal and according to Bögli (1960a), trittkarren devel- planar, or complex. it can be composed of surfac- op where the intensity of the dissolution is great. es with various slope angles. smaller forms can This phenomenon occurs where the flow of water occur on the surface of a tread, for example, waves is thin. in one of his later papers, Bögli (1976) (Figure 10j), steps (Figure 11), ridges (Figures 10k, claimed that these forms developed by dissolution l), peaks, secondary trittkarren (Figure 10l), rillen- under snow when small drops of melting snow fall karren (Figure 10g), and kamenitzas. into already existing depressions on the limestone trittkarren and rinnenkarren can occur on surface. Haserodt (1965) thought that trittkarren the tread of “nischenkarren” (Figure 12). a small occur under micro snowdrifts. according to Hase- rinnenkarren can occur on the interior of trich- rodt (1965), the development of trittkarren occurs terkarren. due to the melting of snow patches. The melting The characteristics of trichterkarren are the fol- dissolves limestone through surface corrosion. lowing: This process needs the presence of snow patches • they are large; for a long time, a characteristic of northern slopes. • their morphology is not varied; The amount of the dissolution (and thus the size • they do not occur in groups; of the forms) depends on the value of the satura- • they develop together with other karren forms tion. according to sweeting (1973), trittkarren de- (such as trittkarren); velop due to large rain drops (during intense rain- • mostly they can develop together with rinnen- fall). she thinks there is horizontal dissolution karren; on their areas primarily because two water layers • small rinnenkarren can occur inside them. merge on the surface. according to Ford and Wil- liams (1989), trittkarren develop on homogenous 157 KRF•1 • OK.indd 157 15.12.2009 10:44:12 Karst Rock Features • Karren Sculpturing Figure 12: An area of “Nischenkarren” treads (Dachstein). 1. “Ausgleichsfläche”; 2. rinnenkarren that developed on the “Ausgleichsfläche”; 3. remnant surface (“peak”) of the “Ausgleichsfläche” with rillenkarren; 4. “Nischenkarren” treads; 5. young trittkarren on “Ausgleichsfläche”. fine-grained rock if micro-steps developed earli- water fed by snow. The following facts prove this er on its surface. These micro-steps could develop, process: for example, by erosion. according to Ford and • trittkarren develop under rillenkarren on a Williams (1989), these forms can develop from slope. The “ausgleichsfläche” often develop kamenitzas. according to several other authors below the trittkarren down the slope. Therefore, (Vincent, 1983a; trudgill, 1985; Veress and lako- the water saturation belt does not occur at the tár, 1995), the development of trittkarren can be trittkarren but on the “ausgleichsfläche”; caused by turbulent flooding. carbon dioxide en- • there are trittkarren that occur on ridges be- ters the water due to the turbulence. as the tur- tween rinnenkarren that can only develop bulence grows, the surface of the rock becomes under snow. if they are not covered with snow, uneven, which further increases the turbulence of they can only get water from the rain that falls the flow. according to Jennings (1985), the devel- on their surface, but this water drains quick- opment of rillenkarren is rapid. The surface of the ly into the adjacent troughs. if the adjacent “ausgleichsfläche” contributes to the development rinnenkarren are filled with snow, the melt of trittkarren. water can flow along the ridges; The morphology of trittkarren suggests they • sheet water develops on areas of trittkarren. The develop through surface dissolution under sheet presence of rillenkarren on the riser and on the 158 KRF•1 • OK.indd 158 15.12.2009 10:44:12 Márton Veress, Trittkarren tread proves the presence of sheet water. The sheet water so it can become turbulent and help presence of rillenkarren also proves that sheet the development of rillenkarren. We think that water dissolution can occur without snow; trittkarren can develop if the dissolution happens • secondary trittkarren can develop on the per- locally and periodically, which is possible if sheet pendicular riser of the primary trittkarren. water flows under snow. secondary trittkarren cannot be created by The development of trittkarren can happen in dropping rainwater since the drops can not touch three phases: the surface of the riser: • in the first phase, a planar surface develops on • the density of trittkarren is great; a slope. This surface part becomes the tread of • the width of the tread of “nischenkarren” the trittkarren; proves the large surface extent of dissolution; • if the inclination of the slope is small, in the sec- • secondary trittkarren develop on the tread of ond phase the tread and the riser form simul- uvala trittkarren. taneously, probably because there is no or little Dissolution occurs on the tread. The following turbulence on the slope. Because of the laminar facts prove this process: flow, the dissolution is similar over a larger area. • the rills of a riser can continue on the tread; These trittkarren are stable forms, and the tread • the uneven surface of a tread including “peaks” increases more and more. on a slope with a me- and ridges is a sign of dissolution on the tread; dium inclination, the dissolution is greatest at • young trittkarren exist on the ice-free bottom the junction of the riser and the tread. The dip of the valley of the Hallstatt glacier because of the tread decreases while the riser becomes the environment of the trittkarren was cov- steeper. on slopes with a steeper inclination, the ered with ice only a few decades ago. The ris- riser subsequently becomes less steep because ers of the trittkarren are short and the dip of dissolution is more rapid. as a result, trittkar- their tread is large. The dip of the tread of older ren cannot exist on such slopes for a long time; trittkarren is small since the inclination of the however, we think that trittkarren can redevel- tread decreases during its development. This op since their density is also high on such slopes; can only be explained by the dissolution of the • in the third phase, the intensity of the disso- tread. lution can vary along the length of the riser trittkarren often occur below one another along when the curve of the riser is longer. The shape a slope. This fact shows that dissolution happens of the riser does not change if the dissolution repeatedly down a slope. given the presence of the is similar all along the riser, and the riser be- tread, the dissolution must be local. We think that comes longer as it develops. The development trittkarren develop on slopes where the stream of trittkarren with small risers is rapid, which of sheet water is turbulent, and at such places the is why the width of the tread is great. When local dissolution is significant. That local turbu- the risers coalesce, the trittkarren change into lence develops under snow, is proven by the fact “nischenkarren.” in the case of trittkarren that trittkarren occur on ridges between rinnen- with higher risers (if the riser is not destroyed), karren. Furthermore, there are no rillenkarren the speed of the development is slower. uvala on most trittkarren and therefore the turbulent trittkarren (with small width of tread) develop sheet water cannot be caused by rain. according if the trittkarren are close to each other. to glew and Ford (1980), raindrops must strike 159 KRF•1 • OK.indd 159 15.12.2009 10:44:12 KRF•1 • OK.indd 160 15.12.2009 10:44:13 corrosion terraces, a 14 MegaausgleicHsFläcHe or a speciFic lanDForM oF Bare glacioKarst Jurij KUNAVER The article deals with corrosion terraces, a specific of Kanin Mts, i.e. on the northern (italian) side landform of bare glaciokarst. They are a secondary (Figure 2), and especially when similar forms were stage of development of different corrosion land- also discovered elsewhere. The same landform forms located on limestone rock surface. They re- was found in the Dinaric mountains as well: i.e. in semble to Ausgleichsflächen from Bögli (1960a), the south Velebit, in the central part of prokletije but they are of greater dimensions and more com- Mts, and in the north limestone alps of austria, plex development. Furthermore they are also ge- e.g. in steinernes Meer. so we have come to the oecologically important. conclusion that these cases are not isolated, but until recently the term corrosion terraces has that they represent relatively infrequent surface not been used for the surface landforms of bare forms of the bare mountains or high mountain karst. They are delimited by a few tens of centi- karst which are regularly more or less associated metres high walls cut into bedrock and have been to these forms. Just were in particular the corro- linked so far with ausgleichsflächen and described sion terraces found in the prokletije Mts (caf Bor as a subclass of them in the case of Kanin Mts, area at the altitude of 1,800 m a.s.l.) that raised slovenia (Kunaver, 1983). The basic characteristic some basic questions about their origin and de- of these surface landforms is the minimum of 10 velopment. owing to their particular position cm high wall dividing two flat surfaces levelled and geological characteristics of the area, as well by corrosion. Figure 1 shows, that in some cases as because of the proximity of vegetation and deep kluftkarren or even minor Schachtdolinen or its possible intense change in the past, the ques- kotlich ( kotlič in slovene; snow kettle in english; tion was also raised, especially about the impact puit à neige in French) can be cut into these flat of vegetation and soil cover on the development surfaces; however, they can be claimed to be of of corrosion terraces; and thus also the question younger origin. in other places walls, measuring about direct and indirect impact of man. up to 30 cm or even more, seem to be remnants of some extremely large ausgleichsfläche surface which mainly do not develop nowadays any more. Morphological characteristics of They are supposed to be remnants of certain forms corrosion terraces and terminology which occurred and developed in the past. a question about these forms arose when at first we linked the terraces, as far as their mor- similar ones were also discovered in other parts phology and origin are concerned, with aus- 161 KRF•1 • OK.indd 161 15.12.2009 10:44:13 Karst Rock Features • Karren Sculpturing Figure 1: Remains of an older corrosion terrace bellow Veliki Babanski Skedenj, 2,000 m a.s.l., western Kaninski podi plateau. Figure 2: Corrosion terraces below Col delle Erbe, 2,100 m a.s.l., northern Kanin mountains, Italy. 162 KRF•1 • OK.indd 162 15.12.2009 10:44:16 Jurij Kunaver, Corrosion terraces, a megaausgleichsfläche or a specific landform of bare glaciokarst Figure 3: Shallow ausgleichsfläche on Bjelić, 2,000 m a.s.l., Caf Bor, Prokletije mountains, Montene- gro. Width of view is 2 m, in the middle. Figure 4: A great kamenitza on Kačarjeva glava, 2,000 m a.s.l., southern Kaninski podi plateau. gleichsflächen forms, i.e. regarding the process sion terraces, since several cases have been found of their genesis (Figure 3). But after similar forms where at least two or three flat shelves are laid one have also been found in considerably diverse envi- above the other and separated by a steep or verti- ronments and at different altitudes, the question cal slope or wall. about their origin claims to be solved in another let us look now at some concrete examples. way. considerations about the Holocene climatic Formerly described corrossion terraces (Kunaver, and vegetational fluctuations and the impact of 1983, 1991) on the Kaninski podi plateau under man have also must be included here. at the same Veliki Babanski skedenj occur at the altitude of time, the need for giving to the phenomenon an 2,000 m a.s.l. They represent a first type of cor- adequate name is becoming evident: to differen- rosion terraces (Figure 1). The height of wall is tiate between ausgleichsfläche, which have much about 30 cm. The characteristics of the shelves lower walls and are of recent origin, and such consist in only slightly dissected rock surfaces, much bigger terraces we suggest to introduce a the lowest one of them, surrounded by the wall on new technical term. They could be called corro- three sides, resembles a large ausgleichsfläche. its 163 KRF•1 • OK.indd 163 15.12.2009 10:44:20 Karst Rock Features • Karren Sculpturing bottom is perforated by a widened corrosion fis- teau at the altitude ca. 2,100 m a.s.l., under col sure caused by the progressive corrosion process. delle erbe. such corrosion terraces fully resemble above this, the next rock shelf follows being very the above described ones, and thus they belong to poorly dissected and ends again into a steep wall, the first described type (Figure 2). of approximately equal height as the lower one. another type of ausgleichsfläche has been found But this one is flat on the one part and inclined in the central part of the Kaninski podi at the alti- on the other. tude of 2,100 m a.s.l., as well as on steinernes Meer an alternative way of explanation for the devel- in the north austrian limestone alps. The latter opment of some corrosion terraces is offered by location lies on a limestone pavement, appr. 850 m the unique case of a great kamenitza on the lower north of riemannhaus, in close vicinity of a loca- edge of Kaninski podi plateau, 2,000 m a.s.l. (Fig- tion of trittkarren. in both cases, these are a kind ure 4). With exceptionally big dimensions, 5 x 2 of more or less distinct linear bends in otherwise meters, this kamenitza seems to be a quite old sur- evenly inclined surface of limestone pavement. face karst feature. exceptional is also its location The next area to be mentioned in this context near the top of extreme roches moutonnées, which is caf Bor (1,810 m a.s.l.) which belongs to the prove the connection of this and similar land- glaciokarst of Bjelić in the central prokletije (plav, forms on limestone rock, at least in its early de- Montenegro). it is a location of corrosion terraces velopment. This kamenitza is now not more active developed on the roches moutonnées, near the al- because its bottom is perforated by a karren well pine pasture of caf Bor below the saddle of the and filled with turf. although it would be quite same name, representing a border line between interesting to establish the time of the start and Montenegro and albania. This region of typi- the end of its development, it is at this moment cal high mountain glaciokarst has been poorly difficult to say anything about this. But it is evi- known and described so far. By now, it has been dent, that the original kamenitza form, with the studied by J. cvijić and recently by s. Belij. These undercutting walls and flat rock bottom, will be corrosion terraces have a slightly lower position preserved still for a long time. when compared with the above mentioned ones, on the northern italian side of Kanin Mts, and they also differ from them by their typical there is an area of distinctive corrosion terraces, location. in this area the vast limestone region which lie in a smooth, glacially eroded karst pla- of Bjelić tectonically meets with non carbonate Figure 5: Successive corrosion ter- races at Caf Bor, possibly originat- ing from subcutaneous corrosion, 1,800 m a.s.l., Prokletije moun- tains, Montenegro. 164 KRF•1 • OK.indd 164 15.12.2009 10:44:22 Jurij Kunaver, Corrosion terraces, a megaausgleichsfläche or a specific landform of bare glaciokarst flysch-like region located to the north. From the amounts to the average of 15–20 cm. in their latter, superficial denudation and erosion brought close vicinity there are certain shelves which are vast deposition of fans of glacial and periglacial partly filled with turf, but such cases only occur origin which accumulated on the contact with sporadically. There is an open question about this limestone and on the limestone itself (Figure 5). turf, whether it is of primary or secondary origin glacial reshaping of roches moutonnées is in (Figure 5). the same area also proved by certain karrentische corrosion terraces of caf Bor reveal closer ge- which have a maximum of 10 cm pedestal. The netic and formal connection with the subterrane- prolonged roches moutonnées, running parallel an karst forms. There might have been some more to the valley, have well preserved forms of glacial vegetation and soil in the past on this bare ridge, erosion, to which the positioning of rock bedding but probably they disappeared due to the impact has contributed in the first place. But intense al- of man. namely, this is a pasture area of the near- teration of the surface caused by Holocene corro- by alpine pasture of caf Bor and because of the sion activity of soil and vegetation is also evident. vicinity of a karrentische it is difficult to claim that There are also some deep schachtdoline-like de- vegetation and soil had substantially changed the pressions on the ridge, most probably of older ori- surface of the former glacially rounded ridge. gin. Thus, traces of Würm glacial erosion can be let us mention in the end, as an illustration and observed on schachtdoline-like depressions and comparison, the extraordinary large ausgleichs- corrosion terraces. close proximity of these phe- fläche and kamenitzas of recent origin in the re- nomena, which might genetically even exclude gion of Bojin Kuk (1,100 m a.s.l.) at Veliko rujno one another, raises some questions about the de- in the southern Velebit. Bojin Kuk is known by its velopment of this bare karst surface in Holocene. steep towers, mainly formed by exfoliation in the Here, corrosion terraces mostly occur in the form Jelar-promina limestones of eocene age. The walls of narrow rounded shelves, located for the most of the shelves are up to 15 cm high, but the walls of part on the top of the ridge. They can follow one old abandoned shelves are even up to 40 cm high, another or they can occur as independent for- or even more (Figures 6, 7, 8). mations. The height of the slope above the shelf Figure 6: A great ka- menitza and cor- rosion terrace in development. Bo- jinac, 1,100 m a.s.l., southern Velebit, Croatia. Width of view is 10 m, in the middle. 165 KRF•1 • OK.indd 165 15.12.2009 10:44:24 Karst Rock Features • Karren Sculpturing Figure 7: Fresh corrosion terrace. Bojinac, 1,100 m a.s.l., southern Velebit, Croatia. Figure 8: Two phases and lev- els in the development of cor- rosion terraces: the higher and older, and the lower and younger. Bojinac, 1,100 m a.s.l., southern Velebit, Croatia. Problems of development and probably be also found in other places. The conclusions above mentioned examples prove the regular- ity of occurrence of such forms in the regions Because of morphological characteristics and of bare, non-dissected rock, glaciokarstic sur- similarities of locations, we assume that the above faces. a great majority of the examples come described examples are also genetically alike. The from the altitudes between 2,000 m and 2,100 differences occur, above all, in the dimensions m a.s.l., one comes from 1,800 m a.s.l., and an- and directions of corrosion terraces, which can other from 1,100 m a.s.l. except for the lowest be the result of local conditions and/or the result location, all corrosion terraces belong to areas of unequal durations of development. We can un- of typical glaciokarst; doubtedly state the following facts: • morphologically these forms are, to the great • locations of these forms at the above described extent, independent formations, which can be sites are not isolated or casual but can most proved by the following: 166 KRF•1 • OK.indd 166 15.12.2009 10:44:28 Jurij Kunaver, Corrosion terraces, a megaausgleichsfläche or a specific landform of bare glaciokarst - it is the cross-section of the terrace which is trudgill, 1985; Ford and Williams, 1989), the important for the morphological identifica- above described terraces can be called corro- tion; sion terraces or megaausgleichsfläche. since the - transition from a more or less flat surface into main accent of this form is its wall, where the wall is sharp. it is usually composed of two morphology and the dimensions of the upper parts: the upper, vertical one, which pass- and the lower levelled parts are less important, es into its lower evenly formed concave end. although still being its components, this form This transition into flat surface is not in the might be terminologically defined as corrosion form of corrosion notch as in the case of ka- terrace; menitzas and very often also in the case of • as for genesis, it might be said that ausgleichs- ausgleichsflächen; fläche as well as corrosion terrace, are both the - the surface of the vertical wall and its concave result of a similar or even the same process, i.e. end is not all evenly smooth but can be com- horizontal levelling of a bare limestone surface posed of individual more or less shelved seg- which also comprises formation and regression ments sculptered by rillenkarren; of walls. although the formative features are - somehow the profile of the wall resembles the similar to those of trittkarren (and, to a lesser wall of trittkarren; extent, to kamenitzas), it is necessary to distin- - the next element is the levelling, which is usu- guish between them because of the differences ally slightly wavy, non-dissected and not very in dimensions; wide shelf that can pass at its bottom side into • corrosion terraces are the result of development the next terrace; in longer periods of Holocene. in interpreting - water is drained in the form of film and rarely terraces geomorphologically, it is necessary accumulates; to take into account the climatic, vegetation- - the above described characteristics point to al and soil changes for each individual region. differences between these forms and the typ- The fundamental question is whether the cor- ical ausgleichsflächen which are still being rosion terraces have developed independently formed; or are they a successive formation, originating • as for terminology (sweeting, 1973; gams et al., formerly in some other way. Judging from our 1973; Kunaver, 1973a; perna and sauro, 1978; observations, we incline to support the second Figure 9: An example of ausgleichs- fläche on Velika vrata, 1,900 m a.s.l., Komna, Julian Alps. Width of view is 10 m, in the middle. 167 KRF•1 • OK.indd 167 15.12.2009 10:44:30 Karst Rock Features • Karren Sculpturing Figure 10: Even in a dolomitic limestone rock of Altipiano di San Martino, Dolomites, Italia, 2,300 m a.s.l., a kind of large ausgleichsfläche could devel- op. Width of view is 4.5 m, in the middle. possibility. in fact, it is a matter of development another indicator of their age is also the fact of a corrosion terrace which had formerly been that some terraces have remained untouched smaller. by vertical karst dissecting, while in other cases We assume that the initial corrosion wall, out this process has already begun and can com- of which a larger one developed later, had been pletely ruin them. eventually, karst terraces are a form of an ausgleichsfläche or some other not forms to be searched for in strongly dissect- horizontal landform of the bare karst (e.g. a de- ed kluftkarren surfaces. pression, excavated either by the subcutaneous if a supposition is made that corrosion terraces corrosion or by the glacial erosion). Therefore originate from depressions in rock surfaces gen- not only is the original form important, but erated by continuous or discontinuous vegetation also the fact, that the bedrock and, in proper and soil cover (which is very likely at these alti- climatic conditions, rock surface can be sculp- tudes in the past), then it can be concluded that tered and developed, for a longer period, under at some time in the past vegetation limit was at special circumstances which result in spe- higher altitude and that the arrangement of alti- cial forms. This is the process of levelling of a tude belts in high mountain regions was differ- rock surface (Figures 9, 10). corrosion terraces ent from the present one. This fact is more or less could be therefore defined as zonally reformed proved by means of the presence of hohlkarren older climozonal karst forms; features, the frequency of which is higher at the • a tendency of augmenting the height of walls is altitude between 1,900 m and 2,000 m a.s.l. in our evident as a result of their retreat. a conclusion regions. There can be no doubt today about shift- about their age can also be made from this fact. ing of altitudes of climatic and vegetational zones according to this thesis, lower terraces are of in the past, caused both by climatic oscillation younger date, and vice versa – the higher ones and the impact of man. are of older date, i.e. their genesis was longer. 168 KRF•1 • OK.indd 168 15.12.2009 10:44:32 rainpits: an outline oF tHeir 15 cHaracteristics anD genesis Angel GINÉS and Joyce LUNDBERG rainpits are small karren depressions not great- explore the rather limited literature, present some er than a few centimetres in size, circular in plan of our unpublished field studies, attempt to out- view, semi-spherical to parabolic in cross section, line their defining characteristics, and speculate with very sharp edges and very regular morpholo- on their genesis. gies. They typically occur in suites on the flat tops of rock bosses, packing all the available space so that the sides of neighbouring pits meet to form Terminology and classification a knife-edge, with no inter-pit flat surfaces (Fig- ure 1). They develop on any soluble rock but are Because these features are not often reported in most often reported on carbonates. They have the literature, there is no clear consensus on termi- rarely been studied in detail. in this chapter we nology. some authors use the general term pitting, Figure 1: Closely packed rainpits in an extremely rough and jagged lime- stone surface from El Co- lomer (Serra de Tramun- tana mountain range, Mal- lorca). Width of view is 40 cm. 169 KRF•1 • OK.indd 169 15.12.2009 10:44:34 Karst Rock Features • Karren Sculpturing but this designation is more frequently utilized (1989) include them under the group of “circu- (together with boring) for nano- to small-scale lar plan forms”. They fit into “small dissolutional karren-forms that lack any clear size limitations forms” as defined by Macaluso and sauro (1996a). and morphologies. in his detailed “lexique des ginés (2004) places them in the “free karren, sin- termes français de spéléologie physique et de gle” forms developed in a solutional environment Karstologie”, gèze (1973) lists the following syno- of storm showers and within a scale limit of 1 to nyms for rainpits: lapiés à cupules in French and 10 cm. although White (1988) classifies rainpits Grübchenkarren or Napfkarren in german. in separately from rillenkarren under “etched forms” some publications, mainly in those devoted to rather than “hydraulic forms”, we believe that the the early discussion on rillenkarren genesis (e.g. rainpits and rillenkarren are genetically related Dunkerley, 1979, 1983), they are called solution (see discussion below). pits. in their papers on gypsum karren, Macaluso We define them here as small-scale, free, single, and sauro (1996a) suggest the neologism minute circular, hydrodynamically-controlled by droplet craters, changing to mini-rain craters in a publi- impact, produced in a solutional environment of cation issued a few months later (Macaluso and direct rainfall and within a scale limit of 1 to 4 cm. sauro, 1996b). none of these alternative terms have achieved widespread use in the literature. taking into account the bulk of the short and Description of rainpits and scattered publications on this type of karren, es- rainpit-related features pecially those associated with semi-arid karst environments, it seems that the best option is to Description assume the term introduced in the international literature by the australian researcher, Jennings in spite of being very distinctive and recogniz- (1971). Thus we suggest that rainpit be the pre- able features, a review of the literature reveals ferred term. note that some authors, for example that there is little consensus on what constitutes White (1988), separate it into two words, as in rain a rainpit. Jennings and sweeting (1963) describe pit. rainpits as “tiny hemispherical hollows, about 5 rainpits are not always included in classifica- mm to 20 mm deep”. in the definitive citation of tion schemes, perhaps because they do not occur Jennings (1971), rainpits are merely described as in all environments. They are quite common in “being usually less than 3 cm across and 2 cm deep” environments characterized by semi-arid and and are considered to be “the simplest effect of even arid climates. However, they seem to be rath- rain falling on bare rock”. no other description is er infrequent in alpine karrenfields. it is probably included regarding the shape or profile of the pits. because of this that no specific mention of this fea- White (1988) offers a broad description of rain- ture appears in the well-known karren classifica- pits as “the smallest of the features resulting from tions developed by Bögli (1960a, 1980), although etching of the bedrock” and indicates that they they would fit into his “free karren” category be- are “few millimetres to a few centimetres in di- cause they usually form on bare rock surfaces. The ameter, roughly circular, symmetrical pits etched small size of these karren features may account for into the bare limestone surface”. in their classi- the disregard perceived by Jennings (1983) with re- fication of small scale solution sculpturing, Ford spect to the scant knowledge on arid and semiarid and Williams (1989) describe rainpits as “circular, karst landforms in specialized literature. White oval or irregular forms in plan view, with rounded (1988) classifies rainpits under “etched forms” or tapering floors and greater than 1.0 cm in di- (the etching being of massive bedrock rather than ameter”. according to Macaluso and sauro (1996a, of structural weaknesses). Ford and Williams b), rainpits “are crater-like depressions; …their 170 KRF•1 • OK.indd 170 15.12.2009 10:44:34 Angel Ginés and Joyce Lundberg, Rainpits: an outline of their characteristics and genesis borders are nearly elliptical or polygonal, with a tops of rock outcrops: on the steeper flanks of the diameter of 12–30 mm and a depth of 1–30 mm, rocks, developing radially outwards from the rain- and their cross profiles are parabolic, with round- pits, are groups of rills. The same pattern was rec- ed bottoms, steep sides and sharp crests”. smith ognized by Macaluso and sauro (1996b) on nar- et al. (1996) speak about “rounded solution pits row gypsum crests from Verzino (calabria, italy). with V-shaped cross profiles”. ginés (1996a, 2004) in the karren assemblages where rainpits and ril- defines rainpits as “small hollowed cup-like kar- lenkarren coexist, the rills are frequently distrib- ren features, sub-circular in plan, nearly parabolic uted downslope over the sides of the protruding in cross section, whose diameter ranges from 0.5 rocks whilst the tops and upper parts of the ridges cm to 5 cm, and exceptionally exceeding 2 cm in are occupied by rainpits (Figure 2). This is an alter- depth; frequently they appear clustered in groups”. native to the common herringbone pattern usual- The most common properties appear to be ly produced where rillenkarren ribs meet back to that they form on bare rock, have roughly circu- back at the crests of pinnacles. in some rather ex- lar plans and parabolic profiles, and are approxi- ceptional cases rainpits are reported to be aligned, mately 1–3 cm in diameter and depth. However, interrupting the flow of small runnels (Figure 3) in view of the variety of descriptions and defini- or even producing distortions inside rillenkarren tions quoted above, some obvious needs arise. The flutes (ginés, 1999a). However, special attention is first is to choose size limits for rainpits, in order required in order to distinguish rainpit features to permit a useful distinction from the larger pits from other small concave features caused by the (e.g. some kamenitzas or solution basins) as well as from the smaller ones (e.g. some biokarst bor- ings or solutional micropits). a second need is to define the shape of the bottom or the cross section, in order to differentiate rainpits from the charac- teristic flat bottom of many kamenitzas and also from the irregular profiles that are typical of solu- tional etching and biokarstic boring pits. a third need is to document the pattern in space in order to distinguish rainpits (that invariably occur in suites and frequently in non-random patterns) from isolated solutional or biokarstic pits and also from randomly-spaced or structurally-governed etchings. Distribution as a general rule, in many descriptions about the karren assemblages in which rainpits occur, it is stated that they develop mainly on gentle slopes or on the summits of the rocks rather than on their steep flanks. a repeated pattern, observed and de- scribed by Dunkerley (1979) from limestone out- Figure 2: Typical pattern of rainpits and rillenkarren on a gently rounded clint in the Gregory Karst, Northern Ter- crops, is of a well-packed group of nearly circular ritory, Australia (photo by K. G. Grimes). Width of view is depressions (i.e. rainpits) that appear on the flat 55 cm. 171 KRF•1 • OK.indd 171 15.12.2009 10:44:36 Karst Rock Features • Karren Sculpturing Figure 3: Sloped lime- stone surface in El Co- lomer (Mallorca), show- ing rainpits in associa- tion with rillenkarren and greater rinnen- karren grooves. Figure 4: Rainpits from El Colomer (Mallorca) in a ribbed or lineated pattern. According to Dunkerley (1979), the conjunction of rainpit- like depressions like these could be involved in the early stages of rillenkarren formation. Width of view is 50 cm. action of thin water films, like the so-called cock- the coalescence of rainpits becomes extensive, the ling patterns described by sweeting (1972). circular or elliptical contour of these small basins on bare exposed rock surfaces rainpits usually is substituted by polygonal shapes, and the sharp occur clustered in closely-packed groups (Figure 1) rims between them produce a characteristic hon- and sometimes coalesced (Figures 4, 5), although eycombed pattern over the surface of the rocks they may occur singly (Jennings, 1985). Where (Figures 1, 5). Macaluso and sauro (1996b) com- 172 KRF•1 • OK.indd 172 15.12.2009 10:44:41 Angel Ginés and Joyce Lundberg, Rainpits: an outline of their characteristics and genesis Figure 5: Zenithal view of clustered rainpits on a 40° sloping rock in El Colomer (Mallorca). Note the con- spicuous biokarstic overdeepen- ings on the bottoms of many rain- pits. Width of view is 55 cm. pare the micro-topography resulting from exten- uous thin water sheets over the rock. in normal sive rainpit coalescence with “miniature versions rainpits suites, as well as coalesced pits, drainage of a polygonal doline karst”. The whole surface patterns can not be recognized; no rivulets are then shows a chaotic appearance divided into ir- seen to form. splash is obvious and presumably is regular and often jagged micro-watersheds. a major way (along with overspilling or decanta- tion) that water is transferred from pit to pit and then down the slope. Hydrology The prevalence of rainpits in arid and semi-arid environments may be a consequence of the scar- rainpits form from the direct impact of raindrops city of sheet or channel flow. limited amounts onto bare rock surfaces but only at the crests or of rainfall cannot generate enough flow of water summits where deep water films do not occur. over karst surfaces to produce sustained sheet The mechanism for drainage of rainpits is not or channel flow. Thus channelled karren forms, obvious. They are clearly associated with surface apart from rillenkarren, are not common in as- roughness which hinders the formation of contin- sociation with rainpits. in this manner, the group 173 KRF•1 • OK.indd 173 15.12.2009 10:44:43 Karst Rock Features • Karren Sculpturing Figure 6: Tiny pitting and rainpits of different sizes in the Gregory Karst, Northern Territory, Aus- tralia (photo by K. G. Grimes). Width of view is 35 cm. of karren features classified by White (1988) as Variations in pit form etched forms develop from nanoscale weathering and biokarstic processes in the absence of signifi- Wherever water drips onto bare rock, a pit will cant flow action. instead, dissolution occurs in form, but for many it is apparent that they are not water films and pools that do not readily drain ef- simple rainpits. simple rainpits form where rain- ficiently. rainpits are one expression of this kind fall hits bare rock of low slope angle. The rain falls of rough etched surface characterized by tiny ir- with no impedance and no other mechanism for regularities, borings and pittings (Figure 5). in the erosion is apparent (at least macroscopically). semi-arid localities of the Mallorcan mountains Where pits develop from drips that are not sim- (ginés, 1996a), the geographical distribution of ple direct rainfall, the drip size is typically larger rainpits is strongly correlated with that of micro- than raindrops and the distribution is not neces- rills ( ril ensteine), another karren form that seems sarily random. in addition, the chemistry is likely to be related to limited water supply. to differ from that of simple rainfall. in all these 174 KRF•1 • OK.indd 174 15.12.2009 10:44:45 Angel Ginés and Joyce Lundberg, Rainpits: an outline of their characteristics and genesis cases the pits are not the product of simple rainfall of subsoil pitting towards polygonal rainpit-fields: and are not part of the genuine rainpit category. is this merely an occasional occurrence or are the instead, they should be called drip pits. examples majority of well-packed rainpits simply the final include pits developed where leaf interception de- result of this process? livers large drops that may have enhanced aggres- sivity. Decantation drips from overhanging rocks, mats of vegetation, snow melt or drips inside caves Morphometric data also fit into this category of “drip pits”. some drip pits are distributed in groups but many are sin- no protocol has ever been established for the meas- gle. The size of drip pits will vary with size of drip; urement of rainpits. as with the sampling of any they are often larger than rainpits. population, care must be taken that all samples in- simple rainpits form on low angle surfaces that cluded are within the population, in this case of are approximately normal to the incoming rain. simple rainpits. Morphometric study of rainpits is another distinct type of a pit, of roughly the same similar to that of rillenkarren (except for length). dimensions as rainpits but found on steeply dip- Width (the distance between edges or cusps) may ping, vertical, or even overhanging faces, does not be measured along the slope and across the slope appear to be formed by either flowing or dripping (as a measure of roundness of plan). Depth (the water. These pits are ~1–3 cm both in diameter maximum depth from the cusp-to-cusp line) is and depth, bored horizontally into steep faces, measured at the deepest point. some measures very similar to borings by large gastropods such of the nature of the edge (steepness of side slope, as whelks. While the mechanism of formation of asymmetry of sides) should be included. profiles these horizontal pits is not at all obvious (they do are recorded by means of a carpenter’s profile occur in supra-littoral coastal locations but they gauge (Figure 7). This is not always ideal: unless also occur in many inland situations where there the pits are arranged in a line, any linear profile has been no possibility of gastropod boring), it is that includes several pits may not sample the cen- clear that these pits are not caused by direct rain- tre point and thus the deepest part of the profile fall and should not be called “rainpits”. for each pit. (This proved not to be a problem for Finally, careful observation of nearly flat lime- lundberg, in 1976, profiling rainpits in chillagoe, stone surfaces where a thin cover of soil has been australia. The rainpits tended to occur in lines.) removed by recent erosion, demonstrates that ex- slope angle of the rock face on which the pits are tensive pitting can also be produced as a particu- developed may vary across the top of a boss and lar case of subsoil karren. additional research and thus should be recorded for each pit. The spatial statistical data are required in this respect, but distribution patterns might be recorded by over- some examples of subsoil pitting from Dragonera head photography and perhaps measured by near- islet and Mallorca (Balearic islands) show some est neighbour analysis. significant differences regarding dimensions, dis- actual data based on careful observations in the tribution and sharpness when compared with field and real measures of rainpits are very scarce genuine rainpits. subsoil pitting produces pits of in the literature. our bibliographical research, very different size (many of them smaller than presumably representative of the current know- normal rainpits), that are shallower and smoother, ledge on these features, produced only a few ref- with blunted rims and generally less coalescence. erences. For instance, Dunkerley (1983) describes casual observation suggests that "subsoil pitting" the micro-topography of several sampling sites in seems to be quickly transformed into true rainpits the chillagoe karst (australia) and comments on when it becomes exposed to direct rainfall. an in- the presence of “solution pits of normal dimen- teresting question then arises about the evolution sions (i.e. about 1–2 cm in diameter)”. smith et 175 KRF•1 • OK.indd 175 15.12.2009 10:44:45 Karst Rock Features • Karren Sculpturing Figure 7: Profile of two rainpits recorded by means of a car- penter’s profile gauge. al. (1996), speaking about surface weathering fea- Data on rainpits from Chillagoe (Lundberg, tures from tatahouine (southern tunisia), found 1977a) a variety of solutional features, including “…solu- tion pans, extensive pitting (1–3 cm in width and Width, depth, radius of circle that fits into the depth) and some poorly developed rillenkarren”. base (the more parabolic the profile, the smaller statistical data based on a significant number the circle) and sharpness (average angle of top 0.5 of measurements of individual rainpits are scant. cm of pit walls to the vertical) were measured on one of the few rigorous studies is that by lund- 1,710 rainpit individuals. results are summarized berg (1977a) in chillagoe (australia) for two well- in table 1. The pits are on average 1.5–2 cm wide differentiated kinds of rocks: namely fossiliferous and 0.5–1 cm deep. The pits on reefal limestone reefal limestone and coarsely-crystalline marble. are wider, deeper and sharper than those on mar- The majority of the rainpits are simple in form, oc- ble. a pertinent observation is that the rock sur- curring in clusters on the tops of rounded bosses. faces on marble are regularly destroyed by exfolia- rainpits occur on slopes from 0ůp to 20°, but the tion or disintegration and thus the morphometric majority are in the range 0°–10°. on the steeper indices for the short-lived rainpits on marble may slopes pits are noticeably asymmetric where the not be of fully mature forms. For this reason, fur- upslope side is taller. complex forms observed ther analysis is confined to pits on the reefal lime- include rainpits grading into cups, and into ril- stone (n = 1,029). lenkarren. only the clearest, simple forms were Depth (54% variation) is more variable than measured. The analyses presented here include width (36% variation). This suggests that the char- some from the 1976 unpublished theses and some acteristic width is reached while the depth contin- additional analyses. ues to increase. The simplest measure of shape, av- erage width to depth ratio is 2.48 (s = 1.12). ratio of depth to radius of fitted circle, average 1.60 (s 176 KRF•1 • OK.indd 176 15.12.2009 10:44:48 Angel Ginés and Joyce Lundberg, Rainpits: an outline of their characteristics and genesis Table 1: Morphometric data from rainpits in Chil agoe, temperatures and precipitation values below 500 northern Queensland (Lundberg, 1977a). mm per year. The number of measurements made Lithology Reef limestone Marble on each site varied between 36 and 139. These pits Width (SD) 20.9 mm (7.6) 16.3 mm (8.2) are rather small: the means of the rainpit diam- Depth (SD) 9.3 mm (5.0) 5.4 mm (3.2) eters are 7.5 mm in the mountains of Dingri, 8.1 Radius 7.8 mm (4.6) 7.9 mm (7.1) Sharpness 33° (13) 44° (13) mm in amdo and 7.9 mm in lhasa valley; and the Slope angle of rock 6.6° (6.2) 10.1° (9.5) mean depth values are 3.1, 3.0 and 4.2 mm respec- N 1,029 681 tively. even the maximum diameter found in each sampling site appears to be small: between 11 and 12 mm. = 1.30), indicates that the pit profiles are typically The rainpits in Mallorca island appear to be shallow parabolas. simple in form and origin. ginés (1996a) meas- Further analysis shows that the profile changes ured 50 rainpits from a karren site 200 meters with size. if depth is used as a measure of size, and above sea level, in the mountain range called serra W/D ratio as a measure of shape, a plot of W/D de tramuntana (Mallorca island). The climate is against D shows that small pits are shallow and, as characterized by an annual precipitation between they enlarge, approach a characteristic W/D ratio of 1.5 (Figure 8, above: exponential relationship, r2 = 0.48, p = 5e–108). if depth to radius of fitted circle (D/r) is used as another measure of shape (the more parabolic the profile, the higher the D/r value), a strong relationship is seen between log D/r and log D (Figure 8, below: r2 = 0.53, p = 6e–166). This shows that small rainpits are more rounded and larger ones more parabolic. Width, however, does not show a significant correlation. if width reaches equilibrium early on in the devel- opment, then the form can get deeper and more parabolic without any change in width. Data on rainpits from other sources  smith (1986) describing some arid karstic sites from  southeast Morocco reports a survey made by Kerr  (1983) in Wadi akerboûss, which identified signifi- cant differences between rainpits developed in ver-  tical surfaces (average diameter 14.5 mm and depth  10.9 mm) and rainpits developed on horizontal sur- faces (average diameter 9.4 mm and depth 7.2 mm);  but no indications about the number of individuals     measured is given. it is not clear if the two popula- tions are both simple rainpits. Figure 8: Rainpits from Chil agoe, Queensland, Australia: abowe is relationship between width to depth ratio Zhang (1994) measured rainpits in three karst (W/D) and depth (D), and below relationship between areas of tibet characterized by remarkably low depth to radius ratio (log D/R) and depth (log D). 177 KRF•1 • OK.indd 177 15.12.2009 10:44:48 Karst Rock Features • Karren Sculpturing Figure 9: A general view at the meso-scale of the karren assemblag- es in El Colomer (Mal- lorca). At this level of observation, rinnen- karren grooves are outstanding as wel as the grey biokarstic patina coating al the rock surfaces. Evidence of subsoil karren rem- nants can also be no- ticed. 800 and 900 mm that falls in the form of short and progress in the understanding of rainpit gen- intense showers, following the typical Mediterra- esis is hindered by these extremely scarce mor- nean pattern of rain distribution that implies a phometric data (table 2), but at least the defini- very dry summer. The sampling shows that in this tion of the feature is becoming substantiated by karren outcrop (Figure 9) the mean of the rainpit several sets of measurements. it is very clear that diameters is 20.1 mm; with a maximum value of many more studies are required before a defini- 61 mm, a modal value around 14 mm, and a mini- tion of rainpits in terms of morphometric crite- mum value of 9 mm. additional measurements ria and diagnostic shape characteristics can be obtained in the same karstic area show congruent presented. it is not clear if all these features are diameter mean values of 20.5 mm (n = 100) and in fact simple rainpits: perhaps the very small ti- depth mean values 10.4 mm (n = 50). betan rainpits documented by Zhang (1994) call Table 2: Summary table of morphometric data on rainpits. Location Notes Width (mm) Depth (mm) No Reference Chillagoe, Australia reef limestone. 20.9 9.3 1,029 Lundberg, 1977a Chillagoe, Australia marble 16.3 5.4 681 Lundberg, 1977a Tramuntana, Mallorca 200 m a.s.l. 20.1 50 Ginés, 1998 Tramuntana, Mallorca 200 m a.s.l. 20.5 10.4 100 Ginés, unpublished data 50 Dingri, Tibet low temperature and precipitation 7.5 3.1 36–139 Zhang, 1994 Lhasa, Tibet low temperature and precipitation 7.9 4.2 36–139 Zhang, 1994 Amdo, Tibet low temperature and precipitation 8.1 3.0 36–139 Zhang, 1994 Morocco vertical surfaces 14.5 10.9 Smith, 1987 after Kerr, 1983 Morocco horizontal surfaces 9.4 7.2 Smith, 1987 after Kerr, 1983 178 KRF•1 • OK.indd 178 15.12.2009 10:44:51 Angel Ginés and Joyce Lundberg, Rainpits: an outline of their characteristics and genesis Table 3: Comparison of morphometric measures on rainpits and rillenkarren of reefal limestones in Chil agoe, Aus- tralia (Lundberg, 1977a). Significant differences are shown by Z-scores of > 3.3 (either positive or negative) and probability values < 0.001. The morphometric measures are explained in the text. Ril s (no = 1,174) Pits (no = 1,029) Z-score p Width,SD 19.47 (7.33) 20.95 (7.64) –4.4 < 5e–5 Depth, SD 8.39 (4.38) 9.34 (5.00) –4.5 < 1e–5 Sharpness, SD 35.49 (13.07) 33.19 (12.94) 4.2 < 5e–4 Radius 7.30 (3.84) 7.78 (4.60) –2.6 0.01 W/D ratio 2.72 (1.30) 2.48 (1.12) 4.3 < 5e–4 D/R ratio 1.43 (1.03) 1.60 (1.30) 3.25 0.0015 for a more precise definition of the feature. in the A preliminary discussion on the absence of clear diagnostic criteria, some degree genesis of rainpits of confusion is likely because of the natural con- tinuum that can be expected between nanoscale The relationship between rainpits and pitting, rainpits and kamenitzas, as suggested im- rillenkarren plicitly by Ford and Williams (1989). in fact, the minimum diameters of pits reported from tibet rainpits and rillenkarren have so many resem- are as small as 5 mm and in the same paper ba- blances (except for the obvious difference that sin-shaped features ranging from 2 cm to 15 cm rillenkarren are elongate down slope) that genet- are classified as solution pans (kamenitzas). Ma- ic similarity is suspected. They have a strikingly caluso and sauro (1996a) are probably thinking similar trough-width and cross-sectional profile, of this problem when they state that “minute cra- both being very close to a parabolic shape. Both ters [called rainpits in this chapter] must not be develop on bare rock that is open to unmodified mistaken for micro-pits, described by Ford and raindrop impact. They are often associated in the Williams (1989), which correspond to the micro- same karren outcrop as an integrated karren as- honeycombs (or micro-alveoli) of biological cor- semblage: the rainpits form on summits of rocky rosion”. blocks where raindrops impinge directly onto an a short note on the effect of lithological varia- approximately flat surface while the rillenkarren tion is relevant here. Factors controlling the pres- develop from the crest downwards where rain ence or absence of rainpits are probably similar to drops fall directly onto a sloping surface. those for rillenkarren (see chapter 16): both need lundberg (1977a) measured both rills and pits rocks of high purity, high hardness, and low po- on limestones in chillagoe (Queensland, austral- rosity and generally high homogeneity. in chil- ia). table 3 shows that the distributions have con- lagoe (australia), lundberg (1977a) found that, siderable overlap at the 1 standard deviation range. apart from radius, all the morphometric measures However, there is a consistent slight difference: on rainpits showed a statistically significant dif- rainpits are marginally wider and deeper than ril- ference for two lithologies, the pits on reefal lime- lenkarren (by ~7% for width, 10–20% for depth). stone being wider, deeper and sharper than those While the mean values are very close, the Z-test of on the more coarsely crystalline marble (table 1). significance (Dunkerley, 1983) shows that rainpits grain size may prove to be an important variable are significantly wider and deeper than rillenkar- once further studies have been done. ren on both rock types at 0.01 probability level. (to 179 KRF•1 • OK.indd 179 15.12.2009 10:44:51 Karst Rock Features • Karren Sculpturing place this in perspective, the significance of the difference between rock types is much higher than between rillenkarren and rainpits within one rock type.) in all locations rainpits conformed to the same parabolic profile as rillenkarren (radius of fitted circle showed no significant difference), but W/D ratios and D/r ratios indicate that pits are a slightly deeper form than rills. Development of profile over time is almost identical for both, rainpits and rillenkarren. For Figure 10: Frequency distribution of rainpit widths in El W/D against D, rillenkarren yield y = 7.6x-0.56, r2 Colomer (Mallorca) compared to rillenkarren widths in = 0.40, p = 4e–125, while rainpits yield y = 7.4x-0.55, the same locality. r2 = 0.44, p = 2e–129. Both reach a characteristic W/D ratio of ~1.5. log D/r against log D for ril- lenkarren yields y = 0.95x + 0.8, r2 = 0.53, and p = 2e–188, and for rainpits yields y = 0.99x + 0.8, r2 ical point of view), the available documentation = 0.53, p = 6e–166. These suggest that both, rain- intimates that environmental conditions may en- pits and rillenkarren, reach a characteristic width hance or inhibit the development of rainpits. The early on in their development, become deeper majority of occurrences (smith, 1986; goudie et over time and approximate to an equilibrium al., 1989; Veni, 1994; Zhang, 1994) correspond para bolic profile. to karst areas characterized by small amounts of The morphometric observations made by rainfall, such as the Hamada de Meski (southeast ginés (1998b and unpublished data) in el co- Morocco), the limestone ranges (western aus- lomer (Mallorca) are in good agreement with the tralia), the edwards plateau (texas, usa) or the observations from chillagoe: cross sections are lhasa valley (tibet, china) respectively. The pres- predominantly parabolic and rainpit diameters ence of rainpits in places like chillagoe (Queens- appear to be slightly, but significantly, wider than land) and serra de tramuntana (Mallorca island) rillenkarren flutes. Distribution frequency of both, extends the distribution to tropical monsoonal rillenkarren and rainpits from el colomer, is pre- and mediterranean climates, but in all the cases sented in Figure 10, showing a clear skewness in receiving less than 900 mm of precipitation per the curve toward greater values, which may sug- year. gest an additional solutional enlargement caused White (1988) argues that rainpits “occur where by the stagnation of water in the pits after rainfall the bare rock is exposed in climates in which showers. Many of the wider pits reported in this minor etching can be preserved” and indicates site show non-parabolic cross sections and prob- that “rainpits are found on many limestones of the ably could be the result of modifications produced american southwest”. Veni (1994) agrees in this by some kind of “kamenitza-effect”. Further stud- respect when he affirms that “…like cleft karren ies are obviously needed to address this question. and tinajas (the texan word for kamenitzas), rain- pits are common in arid climates… but are often overlooked as just rough rock”. The presence of Environmental conditions and rainpit rainpits in desert karst terrains is also document- occurrence ed in north africa (Morocco and tunisia), with annual precipitation less than 200 mm (smith, in spite of the sparse and unevenly-distributed 1986; smith et al., 1996; nelhans and svensson, studies on rainpits (especially from the geograph- 1997). 180 KRF•1 • OK.indd 180 15.12.2009 10:44:51 Angel Ginés and Joyce Lundberg, Rainpits: an outline of their characteristics and genesis several references confirm that rainpits are the case of Mallorca, the altitudinal “extinction associated, in arid environments, with poorly limit” observed for rainpits corresponds approxi- developed rillenkarren (Jennings, 1983; smith, mately to the annual rainfall isohyet of 900 mm 1986; smith et al., 1996; Veni, 1994; nelhans and and mean annual temperatures around 15° to svensson, 1997; ginés, 1999a), but with increasing 16°c. These results, especially concerning precipi- amounts of precipitation it appears that both kind tation amounts, are in good agreement with the of karren types are able to grow together in op- data available from the australian karrenfields of timum conditions. The coexistence of well-devel- chillagoe (Queensland) and napier range (west- oped rainpits and rillenkarren was reported early ern australia), where rainpits and rillenkarren from some well-known australian locations, such constitute characteristic karren assemblages. as napier range and chillagoe, as well as more This study of rainpits exemplifies the usefulness recently from different Mallorcan locations at the of the environmental approach to the understand- periphery of the serra de tramuntana limestone ing of karren forms (the environmental approach range. However, the striking absence of rainpits is obviously complementary to the physico-chem- in many Mediterranean karrenfields where con- ical approach). it demonstrates that aridity is an spicuous rillenkarren are the dominant feature, as important factor in the development of this spe- occurs in the main mountains of Mallorca, calls cific karren feature. it seems that the recurrence for some additional specific explanation. of strong but short showers enhances the develop- owing to the existence of good karren outcrops ment of rainpits but the aridity is just enough to in serra de tramuntana, Mallorca, that are dis- avoid the competitive substitution of the rainpit tributed along the climatic gradient of decreasing form by other forms that grow more efficiently temperatures and increasing precipitation, over with higher rainfall, as seem to be the case of ril- the whole range of altitudes from 0 to 1,450 me- lenkarren. tres a.s.l., these mountains can be regarded as an excellent natural laboratory for the study of karren assemblages. For this reason, within the frame- some provisional ideas about the genesis of work of a wider research project, ginés (1996a) rainpits developed a statistical sampling strategy in order to verify the observed extinction of rainpits with as early as 1979 Dunkerley pondered the idea that increasing altitude and eventually to elucidate the rillenkarren develop from conjunction of rainpits main factors involved in their development or in- into downslope sequences: solutional attack was hibition. The sampling was carried out at 100 sites presumed to begin at scattered points on a rock (each delimited by approximately 50 m2 of karren- face, chains of depressions becoming connect- field surface) where the presence of both, rainpits ed and smoothed downslope. Dunkerley (1979) and rillenkarren, were recorded using the follow- states clearly that in “those sites where solution ing semi-quantitative scale: abundant (10), com- pits (rainpits) are developed on nearly horizon- mon (8), Frequent (6), occasional (4), rare (2) and tal surfaces and grade directly into solution flutes Zero or absent (0). The data show a clear overlap- (rillenkarren) as the slope of the rock increases. . ping of their distribution in the altitude range the diameters of the solution pits are equivalent from 50 to 400 metres a.s.l., as well as extinction to the width of the flutes”. Macaluso and sauro of the rainpits above 500 metres a.s.l. where con- (1996b), again noting the similarities of widths versely rillenkarren become dominant. over this and depths, discuss the possible genetic relation- “extinction limit” the rillenkarren crests exhibit ship of rainpits and rillenkarren. if this is true, the typical herringbone pattern on the tops of then one might expect a good morphological cor- the projecting rocks but rainpits are lacking. in relation in statistical terms. as discussed above, 181 KRF•1 • OK.indd 181 15.12.2009 10:44:52 Karst Rock Features • Karren Sculpturing the morphometric data from Queensland and explain the similarities in size and cross section Mallorca do lend support to this concept. between rillenkarren and rainpits (namely, their Dunkerley (1979) suggests, in the caption of a parabolic shapes), stagnation of water after rain- picture taken from a karren outcrop in Huon pe- fall may explain some of the differences observed ninsula, that solution flutes [or rillenkarren] are between them. if pits fill with water, then their “evolving from an irregularly pitted rock face” and form might be more rounded than tapering, and remarks, in the text of the paper, on the presence eventually also larger. rounded or flat bottoms of “lineations”, considered to be “suggestive of in- could indicate that water stands temporarily in- completely developed solution flutes”. These kinds side the pit producing further dissolution and of supposed early stages of rillenkarren that result over-widening of the former hollow. The recurrent from the “conjunction of the depressions giving presence of small amounts of water and moisture the appearance of ill-formed flutes” (Dunkerley, in the hollow can also favour biokarst processes. 1979) are also a common feature in many sites of Different kinds of biokarst borings are able to Mallorca (Figure 4). The joint occurrence of both cause over-deepening of rainpits (Figure 11), as karren types, in chillagoe, Mallorca, and, as re- demonstrated in several Mallorcan sites. counted by Dunkerley (1983), in Wee Jasper (new south Wales, australia) and Huon (papua, new guinea), lends further support to the proposal of summary of processes a genetic relationship. However, the work of ginés (1996a) on the environmental limitations for both if we presume that the geneses of rainpits and ril- forms (above), indicates that they are not inter-de- lenkarren are related, then the processes of ril- pendent: both forms do develop in isolation from lenkarren formation can be assumed to operate, at one another. least to some extent, for rainpits (see chapter 16), Macaluso and sauro (1996b) suggest that rain- although rainpits are likely to involve a slightly pit erosion appears to be result of some kind of more complex suite of processes. “splash erosion” focused on the centres of depres- The chemistry of dissolution is probably re- sions, while rills are also the result of similar stricted for most of the time during active rainfall “splash erosion” but not point-centered. it seems to BÖgli’s (1960a) phase 1 and 2, but after a rain- that the impact of the raindrop forces water out of fall event stored water may move into phase 3. ob- the pits such that it does not usually flow over the viously, sampling of karst waters from rainpits is rims. Thus the parabolic shape is maintained, but necessary in order to understand the sequence of it requires that the effect of direct interception of solutional events that can occur during and after raindrops is not significantly overshadowed either short and recurrent storm showers. by the effects of overspilling flow down the sides Biological processes similar to those docu- of the pit or by the effects of water stagnation on mented for rillenkarren by Fiol et al. (1996) are the bottom of the pit. By this reasoning, splash- likely to be important, but have not yet been in- ing is probably an even more important factor for vestigated. seM studies of nano-scale features of rainpits than for rillenkarren, owing to the differ- rainpit cusps, walls and bottoms are suggested. ent mechanisms of removal of spent solvent that physical processes are probably similar to ril- are associated with each one. nevertheless, rain- lenkarren but with variations. Mechanical impact pits can retain water and thus may have increased of rainbeat is important, probably involving the aggressivity from further dissolution of atmos- removal of protrusions produced by preliminary pheric co or biological action (both biochemical biological weathering. Hydrodynamic controls 2 and biophysical). are expressed through raindrop kinetic energy assuming that interception of raindrops can impinging on the rock surface through a thin film 182 KRF•1 • OK.indd 182 15.12.2009 10:44:52 Angel Ginés and Joyce Lundberg, Rainpits: an outline of their characteristics and genesis Figure 11: Close-up of several rainpits from El Colomer (Mallorca), showing numerous biokarstic borings over the concave surfaces. of water. considerations of critical thickness of fall intensity. We also presume that controls on water film in relation to raindrop size and energy reaction rate such as rock solubility, temperature, are presumed to be important. rainpits form in turbulence of flow, and thickness of laminar layer a solutional environment of direct rainfall. The should control rainpit dimensions and rate of for- kinetic energy of raindrops cuts through the thin mation. rates of formation are likely to be similar film of water keeping flow turbulent, causing to those of rillenkarren – of the order of 103 years rapid direct dissolution of the rock surface, rather for carbonates (Mottershead and lucas, 2001). than delayed dissolution after diffusion through a laminar sublayer. rain splash, and sometimes also overspilling, continually renews the solvent. The Acknowledgements discussion of the boundary layer model for ril- lenkarren (see chapter 16) is pertinent for rainpits. The authors thank Ken g. grimes for photos (Fig- We feel that much of this is operative for rainpits ures 2, 6) and Ken g. grimes and Joaquín ginés and that several of the implications for rillenkar- for their very valuable comments and suggestions. ren are likely to be true also for rainpits. For ex- This paper is a contribution to the investigation ample, kinetic energy of raindrops should control project cgl2006-11242-c03-01/Bte from the the thickness of film that can be penetrated and Ministerio de educación y ciencia-FeDer. thus rainpit dimensions should vary with rain- 183 KRF•1 • OK.indd 183 15.12.2009 10:44:53 KRF•1 • OK.indd 184 15.12.2009 10:44:53 rillenKarren 16 Joyce LUNDBERG and Angel GINÉS rillenkarren – often referred to by the somewhat sauro (1996b) note that mini-spikes may develop ambiguous solution flutes (ambiguous because in the nodal points between the borders of con- the same term is used for quite different forms tiguous rills or rainpits; these are especially appar- akin to scal ops that develop under stream cur- ent in the deep relief of salt, and where the rilled rent activity on cave walls; curl, 1966) – are small surface is convex (Figure 2c). scale, straight, narrow, closely packed, parallel so- in long profile, rillenkarren have a relatively lutional channels that head at the crest of a bare constant width and depth for much of their length rock slope, and are extinguished downslope (Fig- but decrease in depth before their downslope ter- ure 1a, b, c). They develop on any soluble rock but mination in the smooth planar surface with no are most commonly documented on carbonates. channeled erosion, termed the ausgleichsfläche Their dimensions in limestone outcrops are typi- (Bögli, 1960a). it is often, but not always, a con- cally 12–25 mm in width, 2–6 mm in depth and tinuation of the rilled slope profile at the same 100–300 mm in length. individual flutes are para- slope angle. in certain places it appears to develop bolic in cross-section and are separated by sharp- into a horizontal surface that is not related to bed- ly pronounced cusp lines described by glew and ding planes (Macaluso and sauro, 1996b). all ril- Ford (1980) as “razor-sharp edges” (Figure 1b). lenkarren, given an adequate expanse of homoge- in plan view, they may form a simple suite of neous rock, will reach a characteristic length and parallel flutes on a planar surface with “remark- give way to the ausgleichsfläche; however, in natu- able regularity of form and dimension” (glew and ral situations it is rare for a rock surface to be un- Ford, 1980). Their development to either side of a interrupted by inhomogeneities such as less solu- crest often produces a herringbone pattern, where ble laminae, sedimentary structures, or fractures. the original straight crest line is modified to a Thus the ausgleichsfläche is not always evident in wiggly line between alternating rills (Figure 1d). the field (Figure 2a, b). in discussing rills on gyp- on planar slope facets they rarely bifurcate. How- sum, stenson and Ford (1993) point out that the ever, on convex slopes they will diverge, and on observed length is dependent on the down-slope concave slopes they converge and are transformed extent of exposed bedrock. into runnels (Figure 2c). around a summit they in terms of worldwide distribution, rillenkar- may form a complex of rills splaying out in all ren were first reported from alpine environments directions from the crest centred around a suite of europe where they are widespread (e.g. glat- of rainpits on top of the summit. Macaluso and talp, switzerland). They have since been reported 185 KRF•1 • OK.indd 185 15.12.2009 10:44:53 Karst Rock Features • Karren Sculpturing a, b c d, e Figure 1: Typical simple rillenkarren: a. shallow rillenkarren on dolomitic limestones in Cortina D’Ampezzo, Italian Alps. These show al the typical features: they head at the crest of the block and extinguish downslope at the belt of non-channelled erosion, the ausgleichsfläche. The crustose lichens colonizing the rocks have no apparent ef- fect on the ril profile. Width of view is 20 cm; b. diagram of rillenkarren showing typical features; c. sharp, deep rillenkarren on limestone of Wee Jasper, New South Wales, Australia. These do not natural y extinguish in the aus- gleichsfläche; rather they are terminated by a fracture. Width of view is 45 cm; d. rillenkarren on a landslide block of limestone in Surprise valley, Rocky Mts, Canada. The herringbone pattern is created where ril s form to either side of the original y straight crest line. Width of view is 90 cm; e. using the profile gauge to document ril s on gypsum in Svalbard. These ril s extinguish in the soil and vegetation mat. Width of view is 45 cm. 186 KRF•1 • OK.indd 186 15.12.2009 10:44:55 Joyce Lundberg and Angel Ginés, Ril enkarren from many climates ranging from tropical mon- Rillenkarren morphometry soon (e.g. chillagoe, australia), to mediterranean (e.g. Mallorca island, spain), and humid temper- Many morphometric studies of rillenkarren on ate (e.g. trentino, italy). rillenkarren formation carbonates have been reported: (e.g. Belloni and is inhibited in situations with cover of sediments orombelli, 1970; Heinemann et al., 1977; lund- or vegetation (e.g. many equatorial regions), berg, 1977a; Dunkerley, 1979, 1983; goudie et al., where rainfall is limited (e.g. desert regions), 1989; Bordoy and ginés, 1990; Zhang, 1994; Vin- where rainfall is of low intensity (e.g. British isles), cent, 1996; Mottershead, 1996a; crowther, 1998; where the rate of non-dissolutional weathering or gil, 1992; a. ginés, 1990, 1996a, 1998a). similar erosion exceeds the rate of rill formation (e.g. arc- studies on other soluble rocks are few (e.g. sten- tic regions), and where lithological characteristics son and Ford, 1993, measured flutes on gypsum). of the rock are unsuitable (e.g. highly inhomoge- Mottershead et al. (2000) tried to address this neous texture). paucity of comparative data on non-carbonates by measuring a small number of rillenkarren on a great variety of sites on limestone, gypsum and Classification and terminology salt. Most of the discussion below relates to data from carbonates only. rillenkarren are “small dissolutional forms” as defined by Macaluso and sauro (1996b) where at least two of the three dimensions are measured Protocol for rill morphometric in centimetres, but in general less than one metre measurements long. Bögli (1960a, 1980) classifies them as a type of free Karren – they develop only on rock surfac- typical measures reported for rillenkarren mor- es that are free of soil/vegetation cover. Ford and phometry are width, depth, length, and slope lundberg (1987) and Ford and Williams (1989) angle. often a profile is taken of the rill cross sec- introduce further constraints on genesis: ril- tion with a carpenter’s profile gauge whose pins lenkarren are sited under “linear forms – hydro- conform to the shape of the rills (Figure 1e). How- dynamical y control ed”, and gravitomorphic solu- ever, the measures are not necessarily taken in the tion channels. ginés (2004) places them with “free same way. karren, single” forms in a solutional environment Width is relatively clear cut: it is the distance of direct rainfall and within a scale limit of 1 cm between cusps. Thus width data in the literature to 1 m. are probably reliable and consistent. rillenkarren forms were depicted in old de- Depth is not as clear. Depth is a measure of scriptions of alpine karsts, and eckert (1902) used surface denudation and thus ought to be from a the term Kannelierungen. in the early german presumed original surface. in most cases this may literature they were called Firstkarren and more be the plane of the profile gauge but this is correct usually “rillenkarren” (Bögli, 1951, 1960a; Bauer, only if this plane conforms to the tops of most of 1962; trimmel, 1965). Then, during the 1970s, the cusps. if the cusp top is regarded as the origi- Monroe (1970), Jennings (1971), sweeting (1972), nal rock surface, then depth has to be measured and gèze (1973) spread the term “rillenkarren” normal to the cusp-to-cusp line. Figure 2d shows into the international literature. in this way ril- how this may yield a smaller depth. Most publi- lenkarren become a synonym of solution flutes cations do not clarify their depth measurement (after the translation suggested by Jennings, 1971) protocol and thus may not be directly comparable. as well as equivalent to the French word canne- length data in the literature are also sometimes lures. problematic. length is often constrained by rock 187 KRF•1 • OK.indd 187 15.12.2009 10:44:56 Karst Rock Features • Karren Sculpturing a, b c, d Figure 2: Complex ril enkarren, and some of the difficulties encountered in field documentation: a. ril enkarren develop- ing on the tops of formerly rounded sub-soil pinnacles, Wee Jasper, New South Wales, Australia. This il ustrates one of the difficulties in measuring true ril enkarren length – the smooth rock just above the soil may not be the ausgleichs- fläche; rather it may simply be an indication of recent soil retreat; b. ril enkarren in natural rocksalt, Cardona, Spain. The ril form is continual y interrupted by detrital bands in the rock; the exact position of the crest of each ril is not clear; and the ausgleichsfläche is not clear. The ril s in the foreground appear to extinguish, but those in the background do not; c. very sharp, deep ril s in rock salt of Cardona, Spain. Width of view is 65 cm. These show the mini-spikes be- tween the borders of contiguous ril s mentioned by Macaluso and Sauro (1996b). The forms in the foreground are a mixture of deep, asymmetrical pits, mini-spikes, and ril s. This also shows two channels that are not ril enkarren: both are fluvial channels (i.e. rinnenkarren) from concentration of drainage in a concavity. The one is the wide channel of rectangular cross section, the second is the wide channel on the left-hand side where the original ril cusps are being destroyed; d. measurement protocol for ril enkarren width and depth. 188 KRF•1 • OK.indd 188 15.12.2009 10:44:57 Joyce Lundberg and Angel Ginés, Ril enkarren properties (joints, variations in texture, length of rillenkarren rills below). random sampling is rock face available, and length of rock face avail- typically assumed (e.g. lundberg, 1977a; ginés, able at an appropriate angle; Figure 2a, b) rather 1996b) although rarely specified. selective sam- than hydrodynamic properties. Mottershead et al. pling may be required in order to limit complex- (2000) observe that flute length is defined by the ity of controls (e.g. Mottershead, 1996b). in places length of the divides, the flute ending where the with few rillenkarren, sampling strategy may be divides die out downslope. The only measure of reduced to taking all samples that can be found. true rill length has to be from the crest to the aus- gleichsfläche (e.g. ginés, 1996b; crowther, 1998). if this is not possible, length must be reported Rillenkarren morphometry on carbonate as a minimum and not included in the general rocks calculations of averages. ginés (1996b, 1999a) measured every tenth longest rill in a set on the Width, depth and cross-sectional profile assumption that the longest rills represent the po- table 1 summarizes data for rills in carbonates tential length. For all length measurements, care from many regions of the world. Figure 3 shows must be taken to correctly identify any abnor- width (W) to depth (D) graphs for our data from mally long rills caused by glew and Ford’s (1980) various parts of the world. several general pat- edge effect where water drains over the side edge terns emerge from these data. rills on carbon- of the block. ates have width limits of ~5 to ~50 mm and depth although rill depth is relatively constant down- limits of ~0.5 to ~20 mm. all regions show a rill, a further detail that ought to be specified is strong relationship of width and depth (by sim- the position down-rill at which the profile was ple power functions). Both crowther (1998) and measured. ginés (1996b) took profiles 5 cm below Mottershead (1996a, b) report a significant rela- the crest, Mottershead (1996a, b) at 10 cm from tionship of width and depth for the rills at lluc, the crest, while lundberg (1977a) and crowther Mallorca. in general depth is more variable than (1998) chose a distance proportional to the rill width, both within rills and between rills. Mot- set under study, midway down-rill. Mottershead tershead (1996b) reports a small increase in rill (1996a, b) limited sources of variation by sampling width downslope, while depth increases from the only the straight slope facets; this eliminated the head to a maximum at 0.25 mm to 0.35 mm of the convergent or divergent rills that occur on curved down-rill length. Flutes are shallower on steeper slopes. slopes (Mottershead, 1996b). ginés (1996b, 1999a) slope angle is unambiguous to measure. slope shows a very slight positive skew for rillenkar- of the ausgleichsfläche is rarely included, but ren width frequency distributions but mean and ought to be. modal values are almost the same; the depth fre- several studies have included a measure of the quency distributions do show a distinct positive parabolic profile. Dunkerley (1983) and Motter- skew. Because of a moderate positive skew, Mot- shead (1996b) fitted a parabolic function to the tershead et al. (2000) report only median values rill profile. crowther (1998) provided an asym- but all other publications quote mean values. metry index in addition to fitting a curve to each not all these data are equal in reliability. The side of the rill profile. Both lundberg (1977a) and median values from Mottershead et al. (2000) are crowther (1998) provide a measure of the sharp- included at the end of the table. These data are ness of the inter-rill cusp. from a small number of samples over a great vari- The general problems of field sampling apply ety of climates and some of the individual values here. The definition of the population to be sam- are rather strange. For roughly normal distribu- pled is not always clear (see discussion of non- tions, means and medians should be compara- 189 KRF•1 • OK.indd 189 15.12.2009 10:44:57 Karst Rock Features • Karren Sculpturing , 1990 s. The dge- inés ple y and G ta ta ta sam e g, 1987 ordo , 1996a ., 1977 , 1996b , 1999a enc ord, 1980 , 1998 , 1979 , 1979 , 1983 , 1983 ., 1989 er undber , 1996b , 1996b , 1999a , 1996b g, 1977a g, 1977a t, 1996 t, 1996 t, 1996 t, 1996 , 1994 , 1994 , 1994 on, 1980 ber of erley erley erley erley er, 1985 er, 1985 er, 1985 en en en en Ref ershead ershead a and Simón, 1979 Ginés , 1996b , 1999a; B Ginés Ginés Ginés Mark Mark Mark Zhang Zhang Zhang unpublished da unpublished da Crowther Gil, 1989, 1992 uev Dunk Dunk Dunk Dunk Vinc Vinc Vinc Vinc Osmast Glew and F unpublished da Mott Mott Ginés Lundber Lundber Goudie et al Ford and L Heinemann et al Pérez-C and num ean values, in acknowle , 1990, 1996binésA. G 8 53 36 20 49 74 74 n 150 100 200 480 100 277 293 555 428 204 505 14 18 14 21 SD 7.3 6.7 4.7 6.1 6.9 8.3 8.4 6.1 3.6 7.2 cm) 12.7 16 19 edian rather than as m 18.36 16.3 24.2 19.5 22.7 25.3 17.9 12.5 11.6 34.8 22.5 33.5 11.3 20.4 19.3 12.7 ean, standard deviation Length ( ) as m en as m 6.63 4.54 3.93 3.82 4.17 5.23 3.91 2.73 W/D n 29 113 150 30 108 100 200 49 74 93 1,200 100 680 1,174 204 ere available SD 0.1 0.2 0.1 0.1 0.2 0.1 0.3 0.2 0.2 0.3 0.3 0.4 cm) 0.19 0.35 0.46 0.44 0.42 0.29 0.49 0.5 0.27 0.4 0.44 0.84 0.71 0.42 0.39 en (wh Depth ( n 60 113 38 72 150 30 20 100 200 49 50 32 50 74 93 2,590 100 100 300 100 277 293 680 1,174 555 428 626 100 204 easure is giv 0.5 0.3 0.2 0.4 0.3 0.3 0.6 0.6 0.5 0.7 0.5 0.5 0.5 0.9 0.5 0.6 0.7 0.7 0.5 0.8 0.8 0.6 th al the others since they are giv SD wi cm) 1.37 1.4 1.7 1.8 1.6 1.6 1.9 1.9 1.6 1.132 1.54 1.72 1.59 1.82 1.76 1.84 1.64 1.69 1.43 2.52 1.52 1.95 1.74 2.08 1.95 1.36 1.76 1.85 1.81 1.72 2.09 1.68 1.08 1.74 idth ( parable W ate rocks. Each m one one one one e e e one one one one one one one one one one one one one one one one one one one one one one one one one Lithology marble limest limest limest limest limest limest limest limest limest limest limest limest limest limest limest limest limest marble limest marble limest limest dolomit dolomit dolomit limest limest limest limest limest limest limest limest limest . .l. ion es a.s etric data for carbon alia) alia) e## alia) frica) frica) frica) orphom ation Austr Austr e## e## ver 4,500 metr the distribut Loc Austr oastal sit ia (Spain) ales (W alia) alia) alia) alia) alia ( outh A outh A outh A of Canada) Canada)* ca (Spain)# oastal sit tabr ales ( ysia) outh W Austr Austr Austr Austr ustr aal (S aal (S aal (S .K.) oastal sit China) / o way) ca island (Spain)** an eland) / c ottershead et al. (2000) are not quite com or ies ( ies ( ca island (Spain) ern A .K.) / c nkarren m ock ock , Mallor ew S outh est .K.) / c M , R , R ana, C , Transv , Transv , Transv ia) ange Valencia (Spain) ndalusia (Spain) ew S , W est est est ngland (U urren (Ir China) China) n Tibet ( awak (Mala the skew tia) tia) ca island (Spain) ca island (Spain) ca island (Spain) , Mallor c, Mallor , N ueensland ( ueensland ( ueensland ( ueensland ( W W W , E ngland (U ales (U , B ther Austr d island (N Valley Valley , Q , Q , Q , Q ange Croa Croa tana r , Valencia (Spain) asper arrows , E e, W ead Tibet ( , nor ala, Sar ein ( ent of prise prise ia ( ia ( , Mallor , Mallor , Mallor arades tero de Astr ri, Tibet ( Table 1: Rille data from m Svalbar Sur Sur Istr Istr Lluc Lluc Lluc Ses P Vall den Mar Tramun ee J Mor La Safor Barx, Valencia (Spain) La Madalena, Grazalema, A Cooleman Plain, N W Chillagoe Chillagoe Chillagoe Chillagoe Napier R Griqualand Griqualand Griqualand Gait B Arnside Moelfr Black H Lhasa, Ding Amdo Batu P Dachst 190 KRF•1 • OK.indd 190 15.12.2009 10:44:58 Joyce Lundberg and Angel Ginés, Ril enkarren e ., 1977 ., 2000 ., 2000 ., 2000 ., 2000 ., 2000 ., 2000 ., 2000 ., 2000 encer rombelli, 1970 rombelli, 1970 , 1998a , 1998a , 1998a , 1998a Ref Ginés Ginés Ginés Ginés ershead et al ershead et al ershead et al ershead et al ershead et al ershead et al ershead et al ershead et al Heinemann et al Belloni and O Belloni and O Mott Mott Mott Mott Mott Mott Mott Mott 49 88 n 410 n SD cm) 14.0 15.6 20.0 44 35 35 20 32 22 28 30 30 19.23 6.62 12.61 25.85 Median Length ( 2.63 2.94 4.37 1.15 3.22 5.52 2.45 2.33 4.83 6.5 4.07 3.63 3.55 4.99 3.77 W/D Median n 49 88 n ~600 SD cm) 0.7 0.8 0.44 0.16 0.26 0.60 1.66 1.29 0.39 0.25 0.72 0.44 0.53 0.35 0.47 Median Depth ( n 49 88 100 200 100 50 n ~600 0.4 0.4 0.4 0.3 SD cm) 2.0 2.4 1.67 1.5 1.54 1.44 1.69 0.28 1.41 1.97 4.09 2.27 2.25 1.7 2.06 1.7 1.51 1.6 1.81 idth ( Median W one one one one one one one one one one one one one one one one plingam Lithology limest limest limest limest limest limest limest limest limest limest limest limest limest limest limest limest .l. elective sy s es a.s ated b y) ) restimve one (Ital Figure 5 y) erland) witz erland) / 2,475 metr 980 ( alia) Table XII) lightly os s ation e (Ital erland) ns witz ord, 1 999 ( ea Loc Triest witz Austr arso di Monfalc nd F ength i n m berlandes (S erland) g (S ales ( alia) alia) alia) es inés, 1 es, C arso di , Valais (S W lew a y) witz ribour Austr Austr e) eland) et sit te, C erner O , F Austr ouilles outh rom G ltitudes, l rom G ot included i abar igan e, B wyz (S teys w Greec ch alues ew S ca island (Spain) Victoria ( urren (Ir any a rand' G , N Victoria ( rete ( ca island (Spain) , B y of these 8 sit ites n Veronesi (Ital rotta G engst en, S W, Victoria ( alculated f ib and Z ange: lo ange: high asper ead rom m , Mallor /D c ata f oastal s Prealpi Podcr ee J Borgo G Sieben H Bödmer Vallon des Mor Lapis de G Mean of all v SD 1 SD r 1 SD r W Buchan Buchan MR, Buchan PH, Lassithi, C Lluc Es Molí, Mallor Black H Summar econstructed f* r ** W # d ## c 191 KRF•1 • OK.indd 191 15.12.2009 10:44:58 Karst Rock Features • Karren Sculpturing   ­€ ­€„ „­€„  ­€ ‚ƒ ‚ƒ ‚ƒ ‚ƒ     Figure 3: Width/depth graphs for rillenkarren: marbles in Svalbard (arctic climate), unpublished data; limestones in Rocky Mts, Canada (continental/mountain climate), unpublished data; limestones in Mallorca, Spain (mediterrane- an climate), from data in Ginés (1999b); limestones in Chil agoe, Queensland (tropical monsoon climate), from data in Lundberg (1977a). For a parabolic profile, W = √(4D/a), shown by the red line. ble: although they claim that values of flute size depth. For our chillagoe data (fitting a parabolic are comparable, the data from Mottershead et profile to the width and depth data) the best fit al. (2000) for Wee Jasper (median width of 4.09 value for the constant is 0.073, so the equation of cm) seems to be very strongly out of keeping with the relationship is W = √(4D/0.073). However, if Dunkerley’s (1979) data from the same region the anomalously large, shallow rills are omitted (mean width of 1.9 cm) and out of line with other (< 2% of the data), the constant that best fits is rill width measurements. 0.077 mm-1. The data from the rockies fit well to a Most rills are reported as having a parabolic constant of 0.065 mm-1, and, less well, Mallorca to profile (e.g. glew and Ford, 1980) but it is not often 0.051 mm-1, and svalbard to 0.053 mm-1. quantified by means of curve fitting. crowther The profile of the developing rill changes as the (1998) reports that 80% of the rills studied in Mal- forms enlarge: our data indicate that the parabolic lorca are parabolic, but asymmetrical sides are profile becomes more closed as the rills are in- common. He also reports some rectilinear rills cised until the form reaches equilibrium. if depth considered to be truncated or immature forms. is used as an indicator of development, and W/D For a simple parabolic function of the form y = ratio as an indicator of profile, a plot of W/D ratio a x2, lower values for a indicate a more open pa- against D shows the changes (Figure 4). if all rills rabola, and higher values a more closed parabo- from the same population conform to the same la. When depth D and width W are measured in parabolic profile, then the relationship is simply mm, a has the unit mm-1. Thus D = a(W2/4) and W/D = √(4/ a D) or W/D = const.Dn and the expo- W = √(4D/ a). Mottershead (1996a, b) reports an nent should be –0.5. However, as Figure 4 shows, a value of 0.075, and crowther (1998) 0.073, for the exponents for our data are slightly higher be- lluc, Mallorca. Figure 3 shows fits of width versus cause the values for a that best fit the majority of 192 KRF•1 • OK.indd 192 15.12.2009 10:45:00 Joyce Lundberg and Angel Ginés, Ril enkarren ­€        ‚ ƒ       Figure 4: Change in width/depth ratios with rillenkarren development: marbles in Svalbard, unpublished data; lime- stones in Rocky Mts, Canada, unpublished data; limestones in Mallorca, Spain, data from Ginés (1999b); limestones in Chil agoe, Queensland, data from Lundberg (1977a). the data somewhat underestimate W/D for the Relationship of rillenkarren width/depth smallest rills. For rockies rills < ~3 mm in depth and climate/altitude a = 0.04 mm-1, the value of a rises to 0.065 mm-1 From Figures 3 and 4 it is apparent that all regions for the majority of the population (as shown in have the smallest rills but the cooler the region the Figure 3). For Mallorca rills < ~3 mm in depth a smaller the maximum size. Thus rill size appears = 0.03 mm-1, rising to the standard value of 0.051 to be related to climate. However, width is rela- mm-1. chillagoe is closest to the ideal parabolic tively invariant (1б = 14%), so the majority of the population with a = 0.06 mm-1 for the small rills, relationship with climatic variables is of depth (1б increasing to 0.073 mm-1 for the majority of the = 37%). in the absence of climate data for most of population. This shift in form with development the published sites we have estimated mean annual is also apparent if the relationship between depth temperature and precipitation values (www.world- and the size of circle that fits into the base of the climate.com, means from 1861–1989). Width (lin- rill is used as an indicator of profile (e.g. for chil- ear or log scale) proves to have no significant re- lagoe data; lundberg, 1976). lationship with precipitation or with temperature. 193 KRF•1 • OK.indd 193 15.12.2009 10:45:01 Karst Rock Features • Karren Sculpturing        Figure 5: Rillenkarren from sites in Table 1: left. relationship of depth (log) with mean annual temperature; right. rela- tionship of length with mean annual temperature. Temperatures are from the nearest climate station, rather than the actual site. Depth (log scale) shows no relationship with pre- nual temperatures and rainfall for each site yields cipitation, but a relatively strong one with tempera- no relationship with precipitation but a weak one ture: r2 = 0.60, p = 0.002 (Figure 5, left). with temperature: r2 = 0.40, p = 0.006 (Figure 5, ginés’ (1996b) study of the relationship between right – omitting the exceptionally long rills from rillenkarren and altitude shows very clear results. Wee Jasper – with the Wee Jasper point included He measured rillenkarren from nearly 100 sites in r2 is only 0.30, p = 0.02). The relationship is pre- Mallorca, at altitudes ranging from ~150 to ~1,150 sumably governed by viscosity. Dreybrodt and metres a.s.l. no relationship of width and altitude Kaufmann (see chapter 2) test these data against is apparent but depth (log) shows a very clear their theoretical models and find a relatively close negative relationship with altitude (re-plotted in relationship. of course, empirical data include a Figure 6, left: r2 = 0.91, p = 1.5e–6). rill depth var- variety of rainfall intensities and slope angles so ies from 0.63 cm down to 0.28 cm over the 1,000 the data do not fit perfectly. m range (Bordoy and ginés, 1990; ginés, 1999a). ginés’ (1996b) study of rills in Mallorca shows temperature decreases and rainfall increases with a much more clear relationship of length with altitude. Thus the relationship is with climate. so climate using altitude as a proxy for increased we can say that rill depth is lowest at high rainfall rainfall and decreased temperatures (re-plotted and low temperatures, but, from these data, the in Figure 6, right). Here the r2 value is a robust two effects can not be separated out. 0.75 (p = 1e–11). rather than using the mean length of all rills, ginés plots the mean of every Rillenkarren Length tenth longest rill in each set (assuming these to length data are summarized in table 1. The pub- represent the “optimal” rill length for each sam- lished values that are only estimates of maximum pled location). This relationship is clearest if the and minimum sizes, have not been included here. few sites receiving less than 800 mm per annum Mean lengths range from ~10 cm to ~35 cm. of precipitation are eliminated (a. ginés, 1990, 1999a). rill length varies from ~50 cm at 100 m Relationship of rillenkarren length and a.s.l. to ~10 cm at 1,200 m a.s.l. Dreybrodt and climate/altitude Kaufmann (see chapter 2) use these data in their plotting length against the estimated mean an- model, assuming that rainfall intensity and water 194 KRF•1 • OK.indd 194 15.12.2009 10:45:01 Joyce Lundberg and Angel Ginés, Ril enkarren    ­      ­        €    ­          Figure 6: Rillenkarren of Serra de Tramuntana, Mallorca island, Spain: left. relationship of ril depth with altitude; right. relationship of ril length with altitude. Data from Ginés (1996b). Temperature and precipitation values are estima- ted from Palma de Mallorca mean annual values (17°C, 461mm, 4 m a.s.l.), using the climatic gradients for Serra de Tramuntana of – 0.65°C per 100 m, and 80 mm of rainfal per 100 m. viscosity (temperature) vary with altitude (and clarified the issue. Figure 7 shows rill length in re- ignoring other variables such as ranges in slope lation to slope for various sites. all regions show and rainfall intensities) to calculate the expected optimum development within a slope range of lengths. While not mirroring the empirical data ~20° to ~80°, although there is some indication precisely, the predicted lengths are relatively close. that the cooler regions have rills on more gentle That empirical data can be approximated by equa- slopes. Three of them show a positive relationship tions based on hydrodynamics alone, is important of length and slope, but the data from Mallorca for consideration of rillenkarren formation (see (ginés, 1999a) do not clearly show this. discussion below). it may be that the data of ginés (1999a) and those of Dunkerley (1979) include many slope types and Relationship of rill length and slope complex rills. in order to limit complexity, Mot- The simulations of glew and Ford (1980) yield a tershead (1996a, b) chose only those rills with very strong direct relationship of rill length and unconstrained length on simple planform slopes slope (r2 = 0.92). natural rills are not so simple, of constant gradient with no apparent joints and but they do seem to show a relationship; e.g. Ford approximately horizontal crest lines. These rills, and lundberg (1987) find a weak relationship for from lluc, Mallorca, demonstrate strongly that data from various field sites (r2 = 0.31, p = 1e–6). length does vary with slope: cosine of slope angle it is of interest that Dunkerley (1979), working in and log of flute length show r2 values of 0.31–0.62, nsW (australia), reports a very weak to non-ex- p = < 0.01. crowther (1998), also working at lluc, istent relationship with no significant change in confirms this relationship. Following the theoreti- flute length over the 30–80° slope range, although cal considerations of Dreybrodt and Kaufmann slopes of < 30° have shorter flutes. ginés (1999a) (see chapter 2), it is more correct to consider the also reports no relationship of rill length and slope relationship of tangent of slope and length (as for several sites in mainland spain and Mallorca. shown in Figure 7 – the fitted lines are empirical; Further studies from field sites have partly see chapter 2 for a discussion of the theoretical 195 KRF•1 • OK.indd 195 15.12.2009 10:45:02 Karst Rock Features • Karren Sculpturing       ­   €   ‚ Figure 7: Rillenkarren length in relation to slope for marbles in Svalbard (unpublished data), limestones in Istria, Croatia (unpublished data), limestones in Rocky Mts, Canada (unpublished data), and limestones in Mallorca (data from Ginés, 1999a). relationships using these data). The relationships affect presence or absence of rillenkarren. High displayed by field data are of course complicated purity, high hardness, and low porosity seem by variations in temperature, slope angle, rainfall to be important. sweeting (1972) observes that intensity, and surface roughness. nevertheless, rills are best developed on dense, massive, fine- they conform reasonably well to the model. The grained, strong/hard rocks. Marker (1985) reports implications of this are further discussed below in that rillenkarren on hard precambrian dolomitic the section on rillenkarren formation. limestones in south africa are restricted to beds that are dense with regular grain size and low im- Relationship of rill morphometry and purities, and that the best rills develop on rocks lithology of carbonates with closely packed, homogeneous, small crys- The role of lithological variations has been of in- tals. goudie et al. (1989) note that hardness and terest since the earliest morphometric studies on low porosity are probably necessary pre-disposing rillenkarren (e.g. lundberg, 1977a, b; Dunker- conditions but are not the sole explanation for rill ley, 1979, 1983; Marker, 1985; goudie et al., 1989). development, at least in western australia. The different lithological characteristics that have Homogeneity is normally required: rillenkar- been examined, include homogeneity, grain size, ren are very irregular in form where fossils domi- and surface roughness. nate rock texture. goudie et al. (1989) report that The first question is how lithological variations only the more pure and homogeneous beds in 196 KRF•1 • OK.indd 196 15.12.2009 10:45:03 Joyce Lundberg and Angel Ginés, Ril enkarren the limestones of western australia develop rills. Rillenkarren morphometry on evaporite However, Mottershead (1996b) notes that the rocks limestones of the classic karren site at lluc, Mal- lorca, are fine-grained, hard, pure, with low dolo- rillenkarren on other soluble rocks have been mite content, but they are not very homogeneous. studied very much less often than on carbon- Vincent (1996) notes that rillenkarren-bearing ates. nevertheless there is enough photograph- rocks in the British isles are all very hard, ex- ic (e.g. Figures 1e, 2b, c) and morphometric data tremely pure and dolomite-poor. Both for his data (table 2) to confirm that rillenkarren on evapo- from uK and for data from western australia rite rocks are fundamentally similar. rigor- from goudie et al. (1989), he shows that % calcite ous statistical data have been published for only in fabric and % calcite in cement is the key to ex- three locations: nova scotia, canada (stenson plaining rill presence or absence. and Ford, 1993), spain and uK (Mottershead et grain size/shape may be important where it al., 2000). unpublished data from svalbard have affects friability. sweeting (1972) indicates that been added to table 2. The rills in svalbard are rills do not form where rock disintegrates, such forming on anhydrite. The process of hydration is as some marbles and dolomites. Ford (1996) at- clearly destroying rills, but their size – compara- tributes the scarcity of rillenkarren on dolostones ble with those from nova scotia and uK – sug- of canada to the medium- to coarse-grained gests that they have reached maturity. The rills in character of the rock. spain are slightly larger than the others. Macaluso The second major question is how lithological and sauro (1996b) give estimates of < 2 cm for rill variations affect the morphology of rillenkarren. width and 0.4 to 1.5 cm for depth in gypsum of again, no clear pattern emerges from the litera- another Mediterranean region, south-west sicily. ture. grain size is generally presumed to be impor- These are considerably higher values than the oth- tant. glew and Ford (1980) argue, from evidence ers, but without a rigorous treatment of data, they of simulations, that flute width increases with in- have not been included in the table. creasing grain size. However, this is not true for These few data do not allow any relationships chillagoe, australia. Both lundberg (1977a, b) with slope or climate to be tested. Mottershead et and Dunkerley (1983) find significantly larger rills al. (2000) have measured lengths and slope angles on fine-grained but less homogeneous fossilifer- but present only the median values for each. Thus ous reef limestone compared to rills on coarsely we cannot assess the relationship of slope and crystalline marble. lundberg suggests that the length. average length for rills on gypsum in sval- rills on the marbles are destroyed by exfoliation bard are 12.7 cm, in spain 12.0 cm (median value, before they reach maximum size. Dunkerley, also Mottershead et al. 2000), and in glew and Ford’s finding significant differences between groups (1980) simulation, 14.2 cm. within the reef limestones, suggests that grain- size is only one of the variables. Marker (1985) also Relationship of rill morphometry and finds smaller rills on the coarser-grained rocks of lithology of evaporites Kimberley, south africa, and larger rills on the The data available do not allow a rigorous assess- dense, fine-grained rocks. However, it is apparent ment of the relationship of rock properties and rill from her data table that several of the dense, fine- morphology. general observations indicate that grained beds do not show rillenkarren at all. Mot- lithology is significant, perhaps more to the oc- tershead et al. (2000), in a study of “lithological currence of rillenkarren than to the morphology. control of solution flute form”, observe significant For example, stenson and Ford (1993) note that variations in flute size between limestone sites but rills are somewhat more likely to form on anhy- fail to give details of lithological variations. drite than on the more friable gypsum. calaforra 197 KRF•1 • OK.indd 197 15.12.2009 10:45:03 Karst Rock Features • Karren Sculpturing Table 2: Rillenkarren morphometric data for evaporite rocks. Each measure is given (where available) as mean, standard deviation and number of samples. The data from Mottershead et al. (2000) are not quite comparable with al the others since they are given as median rather than as mean values, in acknowledgement of the skew of the distribution. Location Lithology Width (cm) SD n Depth (cm) SD n W/D Length (cm) SD n Reference Svalbard anhydrite 0.92 0.25 114 0.35 0.2 50 2.94 12.68 7.33 8 unpublished data Nova Scotia, Canada gypsum 0.8 0.15 226 0.26 0.18 226 4.64 Stenson and Ford, 1993 UK gypsum 0.89# ~100 0.27# ~100 3.82# 13* 1 Mottershead et al., 2000 6 sites in Spain# gypsum 1.09# ~750 0.37# ~750 3.64# 10* 3 Mottershead et al., 2000 Mean 0.92 0.12 0.31 0.05 3.76 11.9 1.6 1 SD range 0.80–1.05 0.25–0.36 3.06–8.22 10.3–13.5 gypsum** 0.45 * 14.54 9.13 7 Glew and Ford, 1980 Cardona, Spain rocksalt 2.16 0.36 96 unpublished data Cardona, Spain rocksalt 2.47* ~60 1.39* ~60 1.82* 14.9 Mottershead et al., 2000 5 sites in Cardona, Spain salt 1.73* ~250 0.86* ~250 2.32* 21 Mottershead et al., 2000 salt** 1.75* Glew and Ford, 1980 * median values ** from simulation # values averaged from medians for all sites, n estimated from 125 profiles for Spain and 20 profiles for UK (1996) observes that rillenkarren are best devel- on rock salt, with intermediate-sized crystals of 5 oped on massive, finely micro-crystalline gypsum to 10 mm, are the largest at 2.5 cm wide and 1.4 (e.g. in sicily, italy); they are not found on macro- cm deep; the rills on carnalite, with the smallest crystalline gypsum or on selenites (e.g. in sorbas, crystals of 1 mm (but the more porous rock), are spain). Forti (1996) suggests that crystal size needs intermediate in size. certainly their data indicate to be < 0.5 mm for normal karren to develop on that the relationship of rill size and crystal size is gypsum. However, these observations are entirely not simple. contradicted by the findings of Mottershead et al. (2000). They observe that the widest flutes in their gypsum study sites in spain are from the coarse- Effect of rock type on rillenkarren, grained facies. They give an example of rills at the comparing evaporites and carbonates sorbas site (0.93 cm wide) that are at a smaller scale than the crystals. They note that particularly it has been tacitly assumed since early studies that deep flutes (0.65 cm deep, 1.07 cm wide) are devel- rock solubility must play a large role in the proc- oped in the macro-crystalline gypsum of crystal esses in action and thus be apparent in the mor- size 5 to 10 mm. smaller flutes (0.42 cm deep, 0.91 phometry. it has seemed intuitively obvious that cm wide) develop on the microcrystalline gypsum the greatly enhanced solubility of gypsum and salt of crystal size 0.1 to 0.5 mm. compared to limestone would contribute to larger reports of variations in rillenkarren on salt are rills that form faster. even more rare. Mottershead et al. (2000) claim table 3 summarizes the available data on gyp- that coarsely granular rock bears larger rills; sum, on salt, and on limestone. if enough data are however, their data from cardona, spain, seem assembled for each rock type, then it may be feasi- to contradict this. The rills on halite, the purest ble to assess differences in mean values. However, rock, with the biggest crystals up to 30 mm wide, at present a comparison of means and ranges has are only 1.33 cm wide and 0.75 cm deep; the rills no real value in view of the complexity represent- 198 KRF•1 • OK.indd 198 15.12.2009 10:45:03 Joyce Lundberg and Angel Ginés, Ril enkarren ed by the mean values and the huge differences in dramatic with salt rills, ~80% deeper than lime- rigour of studies. in order to explore effect of rock stone rills, but, again, with such limited data we type alone, sites should be sought where the differ- cannot offer any generalizations. ent rock types co-exist, within the same climatic The data do not allow a test of the relationship of and altitudinal zone. it may be impossible to iso- length to rock type. Few data are available for rill late chemical variations from physical variations lengths on salt or gypsum because the full length since it is unlikely that any one field site would is rarely expressed. Mottershead et al. (2000) note have several rock types, all of the same texture. that “the longest flutes are associated with the With so very little data for rills on evaporites, it deepest strata, representing the longest available is premature to offer a firm conclusion about dif- rock slope” rather than the natural length of the ferences in width. glew and Ford (1980) in their rill. simulations report the widest rills on salt, but their comparison is with field data from limestones of canada, which are amongst the narrowest rills on Non-rillenkarren or complex rills limestone. Mottershead et al. (2000) find that rills on salt are significantly narrower than those on it is important that all comparisons are of the limestone. However, they acknowledge that their same features. rillenkarren are simple forms samples on limestone are biased towards the large from raindrops falling on bare rock where flow is rills in australia and their data on salt are from governed by gravity alone. The term cannot nec- only one location in spain. Their median value for essarily be applied to all suites of small dissolu- limestone is rather higher than the mean shown tional channels. narrow channels often develop in in table 3, but both, 1.73 cm for salt and 1.81 cm soluble rocks simply as a result of channelled cor- for limestone, are well within the 1б range of val- rosion. These are more correctly termed rinnen- ues for limestone worldwide. glew and Ford’s karren (Bögli, 1960a) rather than “rillenkarren” (1980) simulated rills on salt are actually very sim- and are usually, but not necessarily, at a larger ilar in width to the natural rills of Mottershead et scale. The presence of scal ops inside a trough is al. (2000) and the average rill width on limestone. usually a clear indication of current flow and in- Viewed in this light, the widths of limestone and dicative of rinnenkarren. salt rills are not significantly different. another mechanism for channel formation is We also suggest that the limited data do not yet decantation (Ford and lundberg, 1987). Decanta- allow a valid comparison of widths for gypsum tion from a joint or bedding plane, or a vegeta- rills and limestone rills and that it may be diffi- tion mat, or a bank of snow often produces par- cult to isolate chemical influences from textural allel flutes that extinguish downslope. although influences. gypsum rill average widths are clear- typically larger and shallower, in all other respects ly outside the 2б range for limestone worldwide. they resemble rillenkarren but they are formed by However, the limited data presented above from a completely different mechanism. in cases where studies on marbles and gypsum in the same re- the origin of the decantation water is no longer gion, svalbard, show that, while the average width apparent (e.g. snow or no-longer-extant vegeta- of rills on gypsum (0.92 cm) is lower than for mar- tion), the flutes may be mis-identified. ble (1.37 cm), they overlap at the 1б range and thus it is apparent that some forms described in the cannot be considered to be significantly different. literature as ‘rillenkarren’ may actually be more neither rock has been studied for textural proper- complex forms. For example, the source of water ties. is not simply rainfall for the features described in Differences in average depths from Motters- Mazari (1988): the caption for photo 1 – “solution head et al. (2000) for salt and limestone are more is augmented by soil cap at the top” – suggests 199 KRF•1 • OK.indd 199 15.12.2009 10:45:03 Karst Rock Features • Karren Sculpturing that these are decantation flutes. That they are not from a survey of over 100 sites along the 90 km simple rillenkarren, must also be suspected from long serra de tramuntana, reports only two in- the dimensions: at 3–7 cm wide and 3–5 cm deep stances of rillenkarren wider than 2.1 cm. These these are significantly bigger than all rillenkarren are from locations over 800 m a.s.l., which experi- reported in the literature. ence snow several times a year, and from locations another example of a non-rainfall water source under a forest cover where rills are presumably af- are rills in seashore locations that are open to salt fected by tree canopy interception. in both cases spray. This is true for the rills reported by Vincent the source of water is not simply direct rainfall. (1996) for three of the four sites in the British isles Based on the data available at present, we sug- where the rills are affected by wave spray during gest that features in limestone that have a mean “most spring tides and storm events”. width outside of the range ~1.3 to ~2.1 cm should every field situation should be carefully studied. be considered suspect, and the formative mecha- rills that develop down the sides of clint blocks nisms carefully studied. are particularly difficult because they may change function. at the beginning of rainfall they act as simple rillenkarren, but as rainfall continues Rillenkarren formation they often act as decantation routeways for water draining from the clint top. The features of rillenkarren that must be ex- simple rillenkarren develop where the water plained in any model of formation are the separa- falls from above. sweeting and lancaster (1982) tion of foci of erosion into almost evenly spaced describe rills on marbles in the namib desert components, the parabolic cross profile of each that form only on the sides of boulders facing rill, the formation at a crest and extinguishing the incoming advection fog. These rill dimen- downstream, and the continuation of erosion in sions (2.2–3.0 cm wide, 0.15–0.32 cm deep) are the ausgleichsfläche as sheet flow. also on the high end of published rillenkarren Modes of formation may not be apparent from dimensions. true rillenkarren are purely gravito- field evidence because of the complexities of na- morphic features: sweeting and lancaster’s (1982) ture. glew and Ford, in 1980, were the first to rills are further complicated by the action of wind simulate rillenkarren production in the labora- in controlling flow dynamics, as are the impres- tory and thereby limit the variables. They created sive wind-enhanced dissolution rills in patagonia rillenkarren on gypsum blocks in the laboratory (Maire et al., 1999). and documented their development. rill width The rills described by Migon and Dach (1995) emerged early on; the first narrow and variable rills on porphyritic granite in poland have also to be coalesced in stages until the stable, characteristic, assigned to the non-simple category. These are and relatively constant width was reached. Width probably pliocene in age and relict, with a poorly then remained stable as rills deepened and length- known history. ened. They developed at the crest and propa gated it seems that rillenkarren width is invariant downslope, lengthening and deepening steadily at enough to be able to offer a definition of true ril- the beginning, then at a reduced rate until a stable lenkarren based on their width. ginés (1996b) length was reached. The rill cross section devel- demonstrates that width is the most stable charac- oped into a parabolic form as they deepened. The teristic of rillenkarren. table 1 demonstrates that rill trough and ausgleichsfläche surfaces showed rills from widely different locations have rather parallel retreat. close mean widths: the frequency distribution These observations are confirmed in other plas- centres on 1.70 cm, the majority of the population ter simulations by slabe (2005) and are gener- lying between 1.50 cm and 1.95 cm. ginés (1999a), ally supported by field evidence. Mottershead and 200 KRF•1 • OK.indd 200 15.12.2009 10:45:03 Joyce Lundberg and Angel Ginés, Ril enkarren lucas (2001), examining rill development in natu- um to be reached in a water film of 0.2 mm thick- ral gypsum on surfaces of known age, confirm that ness for the fast phase of dissolution, but 500 seco- width stabilizes early whereas depth continues to nds for the slower. increase. Field evidence that flute length and slope if dissociation is the only process, then karren angle is maintained as divides are lowered, sug- waters should have very low hardness values. em- gests that the whole system is in dynamic equilib- pirical data for caco contents of karren waters on 3 rium; rills maintain their overall form over time, bare rock are few but do indicate very low values. but change in detail as cusp lines shift because Dunkerley (1983) quotes total hardness values of existing flutes are captured and new ones are initi- only 22 to 28 ppm for flow distances of 120 to 170 ated (Mottershead, 1996a; crowther, 1998). cm – much longer than typical rillenkarren. Fiol studies of morphometry (both at the macro- et al. (1992) report values between 9 and 45 ppm and nano-scales) and of chemistry help to eluci- for natural rainfall waters collected at the end of date process. rillenkarren formation, as with al- rillenkarren flutes. Fiol et al. (1996) find values of most every karst feature, appears to be the prod- around 10 ppm, and Mottershead (1996a) around uct of chemical, biological, and physical processes. 20 ppm, from rillenkarren irrigated with distilled water. Chemical processes Biological processes The chemical processes that produce rillenkarren must take place in a very thin film of water (see rillenkarren are limited to surfaces that are free relevant discussion in Dreybrodt and Kaufman, of macroscopic cover. However, Fiol et al. (1996) chapter 2) and in the few seconds required for convincingly demonstrate that biological action flow from crest to ausgleichsfläche. on evaporite at the microscopic scale is an important part of rocks, simple dissociation is realistically the only rillenkarren erosion in carbonates (although not possible chemical process (it is unlikely that bio- in gypsum and halite – Mottershead and lucas, logical action will be important). Bögli (1960a) 2004). Fiol et al. show, by scanning electron mi- asserts that rills on carbonates must also form croscopy (seM), that colonies of cyanobacterial by simple, rapid dissolution, and cannot include cells inhabit the rock surfaces, and that bio-ero- the long-term and slow complexing of co . ril- sion weakens the crystalline structure of the lime- 2 lenkarren must thus be formed from simple dis- stone. The impact of raindrops is then presumed sociation of caco in water. This is Bögli’s phase to cause the detachment of particles contribut- 3 1, yielding water of around 14 ppm caco content. ing to the considerable particulate load in run-off 3 phase 2 of the chemical reaction involves the co water. 2 that is chemically dissolved in rain, governed by The effect of colonization by lichen is not clear. atmospheric co (relatively constant worldwide) Macaluso and sauro (1996b) and Mottershead and 2 and temperature (less dissolution of co at high- lucas (2000) found that lichen hindered, but did 2 er temperatures). This yields another ppm or so. not completely inhibit, dissolution of gypsum sur- These first two phases take only seconds to com- faces. Yet sharp rills on gypsum in svalbard show plete. Further phases of carbonate dissolution re- no apparent effect of colonization by crustose li- quire longer time (an order of magnitude longer chen (unpublished data). The effect of lichens on and much slower than the drainage rate of rills) carbonates is similarly unclear. Moses and Viles and are thus not relevant for rillenkarren forma- (1996) offer direct seM evidence of bio-erosion tion. Dreybrodt and Kaufmann calculate that 20 under lichen cover on rillenkarren from carbon- seconds would be required for apparent equilibri- ates in eastern australia: they show that the whole 201 KRF•1 • OK.indd 201 15.12.2009 10:45:03 Karst Rock Features • Karren Sculpturing rill surface, both ridge and base, is dissected with While the mechanics of removal may not be circular etch pits and tunnels and that the lichen clear, it is apparent that loss of particulate matter thallus colonizes the top 0.1 to 0.15 mm with fun- is a significant part of rillenkarren erosion, espe- gal hyphae penetrating to 1.4 mm. on the other cially in carbonates. hand, in their experiments of particulate matter removal by raindrop impact on different lime- Hydrodynamic stone surfaces, Fiol et al. (1996) found that lichen- Bögli (1960a), observing that rillenkarren form covered rock released significantly fewer particles only where bare rock is exposed to uniform rain- than “bare” rock (which actually is colonized by fall coverage, suggested the probable importance endolithic cyanobacteria). of physical characteristics of the water-precipita- tion, mode of flow, thickness of water layer, and predicted that rillenkarren should increase in Physical processes length with increasing slope and with intensity of rainfall. since then, discussions about rill forma- physical processes may be expressed as the me- tion have focussed mainly on hydrodynamic con- chanical removal of particles and/or in the influ- trols. (Dunkerley, in 1979, finding no evidence in ence of hydrodynamic action on the process of field studies for the predicted increase in length dissolution. The two potential sources of physical with slope, questioned hydrodynamics as the control in the rill environment are raindrop im- principal control. However, later field studies did pact and fluid flow. raindrop impact is generally find the predicted increase in length with slope, considered to be the more significant of the two, and thus the focus of studies returned to hydro- but discussion continues. dynamic controls.) glew and Ford, in 1980, postulated that hy- Mechanical drodynamic controls – rain drop kinetic energy Fiol et al. (1996) show that mechanical removal and thickness of water film – are dominant. rain- of small limestone protrusions (albeit produced drops, striking randomly over the whole surface, as a direct result of biological activity) is one of keep the upper layer of water in constant motion, the principal processes involved in the growth of renewing the solvent and creating mixing. Where rillenkarren in carbonates (accounting for nearly the water film is thin enough, and raindrop en- half of the erosion). Mottershead (1996a) suggests ergy great enough, the drops penetrate the lami- that fluid flow down the rill trough may cause me- nar sublayer to cause dissolution directly on the chanical erosion but Fiol et al. (1996) argue that surface rather than by diffusion through the lam- it is caused by raindrop impact. Mottershead and inar sublayer. rillenkarren are the result of dis- lucas (2004) offer further evidence of the impor- solution from direct raindrop impact on the rock tance of mechanical removal of particulate mat- surface; the ausgleichsfläche is the expression of ter, this time from gypsum and, to a much lesser dissolution under a deeper water film. While new extent, salt surfaces. The nano-morphologies in- research has highlighted additional processes dicate delicate promontories and etchings as well such as biological and mechanical action, the es- as considerable loosened particulate matter (pre- sence of these interpretations provided the basis of sumably caused by physical or chemical weather- studi es for the next 15 years or so. ing because there is no evidence of any biological action). They suggest that the mechanical force of raindrop impact will detach these and contribute Development of the ausgleichsfläche to the sediment load of run-off water from gyp- sum surfaces in the field. The development of the ausgleichsfläche is not 202 KRF•1 • OK.indd 202 15.12.2009 10:45:03 Joyce Lundberg and Angel Ginés, Ril enkarren well understood. Field evidence is quite vari- The relationship of rillenkarren and rainpits able; in some cases the ausgleichsfläche contin- ues to develop at the same angle of slope as the rillenkarren and rainpits (Jennings, 1985) have rilled belt (as happened in glew and Ford’s sim- so many similarities that a genetic relationship is ulations, 1980). in other cases the rillenkarren suspected. This idea is further developed in chap- will terminate in an obvious decrease of gradi- ter 15 (rainpits: an outline of their characteristics ent into the smooth face of a step karren (Motter- and genesis). shead, 1996b). Macaluso and sauro (1996b) sug- gest that the ausgleichsfläche develops where ero- sion slackens. This is probably not caused by an Modelling rillenkarren formation approach to saturation. rather it is likely that the dissolution rate will vary for the two regions. it any model for rillenkarren formation must offer a may be that the evidence from glew and Ford’s reasonable explanation for the salient features and (1980) simulation of parallel retreat of slope does embrace all the available evidence of chemical, bi- not necessarily apply to carbonates because of the ological, mechanical and hydrodynamic process- high solubility of gypsum. For carbonates, where es. The simulation studies of glew and Ford (1980) direct raindrop impact is inhibited, and the reac- provide the basis for the theoretical model of ril- tion becomes diffusion-limited, dissolution may lenkarren development discussed below, which be correspondingly reduced. However, velocity of we have called the “raindrop impact and bound- flow also governs dissolution rate. Discharge and ary layer model”. velocity increase downslope, so reaction rate may increase down the ausgleichsfläche. Raindrop impact and boundary layer model if dissolution proceeds at different rates for the glew and Ford (1980) offer a hypothesis of ril- rilled section and the ausgleichsfläche, then the lenkarren formation. it includes the following profile of a single face should thus become modi- concepts: fied into a double face, with a break of slope at the • rillenkarren are a rim effect induced by direct junction of rill and ausgleichsfläche, a situation rainfall impact on the rock surface; that is often observed in the field but not consist- • dissolution is fast and at the point of impact – ently. no further dissolution occurs from splash or from water draining down the rill; • the differentiation into two distinct morpho- Rates of formation logical zones (rilled and planar) is a conse- quence of a critical change in the thickness of rates of formation are estimated from simulation the film of water flowing over the surface and studies (e.g. glew and Ford, 1980), from water the degree of turbulence at the base caused by chemistry studies (e.g. Dunkerley, 1983), and impacting raindrops; from field studies of dated surfaces (e.g. gams, • raindrop impact produces a constant dissolu- 1990; Mottershead and lucas, 2001). table 4 sum- tion rate over the rilled section, where the water marizes these estimates for carbonates, gypsum, film becomes deeper than the critical thickness, and salt. rates of formation are in the order of 103 a laminar boundary layer protects the sur- years for carbonates, 102 years for gypsum, and 101 face from direct raindrop impact and the aus- years for salt (Mottershead and lucas, 2001). This gleichsfläche continues to dissolve in a planar information can, in turn, be used as a rough dat- form (Figure 8, left); ing tool: e.g. gams (1990) used depth of rills as • the troughs erode only, or principally, through measure of time since deforestation. direct raindrop impact – channel flow is not 203 KRF•1 • OK.indd 203 15.12.2009 10:45:04 Karst Rock Features • Karren Sculpturing important; thus the rillenkarren channel is therefore raindrop kinetic energy relates to rain- radically different from the conventional Hor- fall intensity, all by power laws (Kinnell, 1987; tonian stream channel that propagates downs- uijlenhoet and stricker, 1999; salles et al., 2002). lope of the belt of sheet overland flow where For example, light rain, heavy rain, and thunder- converging flow creates channels. storms have a mean droplet diameter and kinetic The implications of this model are: a) that ki- energy of 1 mm and 1 kJ/(m2 h), 2.1 mm and 103 netic energy of raindrops should control the thick- kJ/(m2 h), and 3 mm and 104 kJ/(m2 h) respectively ness of film that can be penetrated and thus ril- (auerswald, 1998). obviously, the higher the ki- lenkarren length (the relationship should be with netic energy of the drop the thicker the laminar rainfall intensity rather than total rainfall, and layer that can be penetrated (Figure 8, left). rain should relate to water viscosity); b) that controls splash erosion on soils is especially effective where on depth of flow downstream should control ril- the water film is less than ~0.1 to 0.3 of the drop lenkarren length (discharge, slope angle, surface diameter; and water films thicker than 3 drop di- roughness); and c) that controls on reaction rate ameters protect the surface from raindrop impacts should control rillenkarren dimensions (e.g. rock (auerswald, 1998). in glew and Ford’s (1980) sim- solubility, temperature, turbulence of flow, thick- ulations drop size was centred on 1.0 to 1.5 mm. ness of laminar layer). rock solubility will interact They found experimentally that the limiting film with hydrodynamic controls to affect rill depth, thickness was 0.15 mm on a 45° slope. This is 0.1 width and speed of formation. of the drop diameter – at the lower end of the ef- a) effect of raindrop kinetic energy: raindrop fective water film thickness for raindrop erosion size and energy vary with the type and intensity of soil. Dreybrodt and Kaufmann (see chapter 2) of rain events and thus will vary geographically. modelled film thickness using only hydrodynam- Mean drop diameter relates to rainfall intensity, ic controls (density, viscosity, slope angle, rainfall and terminal fall speed relates to diameter, and intensity) in the region of 0.075 to 0.1 mm. petter- Figure 8: Structure of water film and controls on rillenkarren length: left. critical thickness of boundary layer depends on raindrop size/energy level; more intense rainfal should produce longer ril s; right. critical thickness of boundary layer relates simply to slope angle; steeper slopes should produce longer ril s. 204 KRF•1 • OK.indd 204 15.12.2009 10:45:04 Joyce Lundberg and Angel Ginés, Ril enkarren son (2001) measured thicknesses of 0.2 to 0.8 mm The slightly concave long profile of the rill on a 30° slope. trough observed by Mottershead (1996a, b) can although rillenkarren morphology appears to also be explained by raindrop kinetic energy. The have no relationship to total (mean annual) pre- position of the critical depth will vary with varia- cipitation, variations in raindrop properties may tion in rainfall intensity. For any region, raindrop explain much of the variation found in rillenkar- energy varies temporally. Thus there will be a con- ren properties. tropical rainfall is typically con- stant slight shifting of the position of the critical vective, and characterized by larger raindrops depth throughout a single storm, throughout the than the lighter rains of temperate regions (cal- year, and over climatic cycles. The gentle transi- der, 1996). raindrops in mediterranean regions tion from rill to ausgleichsfläche is a response to are variable but the majority are small (e.g. in the natural variations in rainfall intensities and cerdá’s 1997 study in the Western Mediterranean, the associated variations in the position of criti- the majority of raindrops were less than 1 mm). cal depth. Flow in the upper part of the rill is al- Therefore rillenkarren ought to be longer (and ways shallower than the critical depth. Flow in the presumably wider and deeper) in subtropical re- lower part will vary. a region with more constant gions compared to temperate and Mediterranean rainfall intensities should have a sharper transi- regions. The only substantial data sets relevant to tion. this question are from subtropical, monsoonal The impact of raindrop kinetic energy should chillagoe (Dunkerley, 1983) and Mallorca (ginés, also vary with water viscosity (see chapter 2), 1999a). The difference in average rill length for which in turn varies inversely with temperature. chillagoe (33 ± 20 cm, n = 428, with an average all other things being equal, rill length should be slope of 65° ± 12) and for Mallorca (23 ± 8 cm, n greater at lower viscosities and higher tempera- = 200, with an average slope of 58° ± 12) is highly tures, but this effect will be complicated by the significant (z-score of 9). if we account for the dif- impact of temperature on reaction rate (below). ference in slope by using the relationship of length b) Depth of flow controls (discharge, slope and slope shown in the data of Mottershead angle, surface roughness, porosity): Discharge is (1996a) from Mallorca to calculate the average directly related to rainfall intensity (q). Depth in- length in Mallorca for an average slope of 65°, this creases downslope with length l (as 3√(l/q). The would be 26 cm. testing this against the length higher discharge caused by higher rainfall in- for chillagoe is still highly significant (z-score of tensity should therefore reduce the length of ril- 6). similarly rill width in chillagoe is significantly lenkarren, but is counteracted to some extent by larger than in Mallorca (z-score of 4.6) and rill the higher kinetic energy of the drops. depth (z-score of 14). However, these comparisons increasing slope angle reduces the effective are complicated by temperature differences, chil- rainfall per unit area, increases the velocity of lagoe’s mean annual temperature being ~26°c flow and thus decreases the thickness of the water and Mallorca’s ~17°c. film (Figure 8, right). The positive relationship of a survey of a population of rills forming under rillenkarren length and slope support this argu- trees reported in ginés (1999a) is relevant to this ment. The position downslope at which the criti- discussion. The mean width, at ~2.5 cm, is clearly cal depth is reached is a function of slope and outside the normal range of rillenkarren widths length. logically, if hydrodynamic controls are for all the non-forested regions in the rest of the the only or principal control, then the best rela- study. This is explained as the result of intercep- tionship should be produced by length against tan tion of rain by the forest canopy and the subse- slope (see chapter 2). However, for our empirical quent formation of larger raindrops (Brandt, 1989, data, the relationship is not very strong. For is- 1990; Hall and calder, 1993). tria and rocky Mts data, a slight improvement is 205 KRF•1 • OK.indd 205 15.12.2009 10:45:04 Karst Rock Features • Karren Sculpturing shown when the data are limited to slopes smaller that the mechanical strength of the rock can sup- than ~60°: e.g. for istria limestones data l = 1.9 port the steep sides; Dunkerley, 1979). However, if tan slope + 14, r2 = 0.16 for all data, but for slopes it were so simple, then the depths of rillenkarren less than ~60° the relationship is l = 9.4 tan slope on gypsum ought to be larger than on limestone + 9, r2 = 0.63. and smaller than on salt. This does not appear to surface roughness provides additional fric- be so (table 3), although the impact of biological tional retardation and thus decreases velocity of erosion on limestone has not been taken into ac- flow, increasing the depth of the water film and count in this comparison. reducing the length of the rillenkarren. if surface There is no logical reason to expect length to roughness causes significant retardation of flow, vary with reaction rate. it does show a relation- then the slope of the relationship of length by ship with mean annual temperature (Figure 5, slope angle would be lower. Data are not yet avail- left), and with altitude (Figure 6, right), but this is able to test this effect. most probably explained by the effect of tempera- rock porosity (within limits – very porous ture on fluid viscosity and thus on the depth of rocks will not show any rillenkarren development the critical layer. if reaction rate is an important at all) should also affect flow depth and thus rill control on rillenkarren length, then length should length. a slightly porous rock will have a slight- vary directly with solubility for different rock ly less deep film, so slightly more porous rocks types in the same field situation. The very limited should have longer rills. data from svalbard show average rill length on c) reaction rate (rock solubility, temperature, marble to be 12.7 cm (n = 8) and on gypsum to turbulence of flow, thickness of laminar layer): be only 9.0 (n = 5). putting together length data There is a simple positive relationship of rate of dis- from spain from various sources, limestone rills sociation and temperature (reaction rate doubles are 20 cm (ginés, 1996b), gypsum rills are 10 cm every 10°c, a logarithmic function). if rillenkar- (the spanish sites taken from Mottershead et al., ren are produced by simple dissociation of calcite 2000), and salt rills are 21 cm (Mottershead et al., (Bögli, 1960a; also see chapter 2), and if depth is a 2000). until further data are collected, it appears measure of reaction rate, then depth should show that rock type does have a significant effect on rill a relationship to temperature. The demonstrated length, but it is not governed simply by reaction log-normal relationship of depth and altitude in kinetics because length is not in order of reaction Mallorca (Figure 6, left), combined with the ob- rate. servations of very deep rills on salt (Mottershead The relationship of rillenkarren width and re- et al., 2000) do appear to confirm that depth is action rate is not so clear. Width shows no rela- strongly controlled by reaction rate (on condition tionship to mean annual temperature and ginés Table 3: Summary of rillenkarren morphometry. Averages and 1 standard deviation ranges for rillenkarren on car- bonates, gypsum, and salt from Tables 1 and 2, and median values from Mottershead et al. (2000). Rock type Width (cm) Depth (cm) W/D Length (cm) Source ± 1 SD ± 1 SD ± 1 SD limestone 1.70 0.44 4.37 19.23 Table 1 1.43–1.98 0.26–0.60 3.22–5.52 12.61–25.85 limestone 1.81* 0.47* 3.77* 30* Mottershead et al., 2000 gypsum 0.92 0.31 3.76 11.9 Table 2 0.80–1.05 0.25–0.36 3.06–8.22 10.3–13.5 gypsum 1.07* 0.37* 3.27* 12.0* Mottershead et al., 2000 salt 1.73* 0.86* 2.32* 21.0* Mottershead et al., 2000 * median values 206 KRF•1 • OK.indd 206 15.12.2009 10:45:04 Joyce Lundberg and Angel Ginés, Ril enkarren Table 4: Estimates of rates of formation for rillenkarren. Rock Time (yrs) Place Method Reference limestone 724–1,159 Lluc, Mallorca chemistry Mottershead and Lucas (2001); Fiol et al. (1996) 400–600 Hercegnovi, Montenegro field Jakucs (1977) 2,000–2,500 Lluc, Mallorca chemistry Mottershead (1996a) 2,100 Dinaric karst field Gams (1990) 1,330–2,600 Queensland, Australia chemistry Dunkerley (1983); Mottershead and Lucas (2001) gypsum < 26–100 United Kingdom field Mottershead and Lucas (2001) < 20 Mallorca, Spain field Mottershead and Lucas (2001) 4–38 Crete, Greece field Zezza (1994); Mottershead and Lucas (2001) 40–100 Nova Scotia, Canada field Stenson and Ford (1993) salt 8–20 Catalunya, Spain field Mottershead and Lucas (2001) (1996b) demonstrates that width does not vary soluble than fine (as discussed with reference to with altitude in serra de tramuntana, Mallorca. rillenkarren by Dunkerley, 1983). Ford and Wil- as discussed above, width is the most invariant of liams (1989) give many examples of finer grained rillenkarren characteristics, mean widths show- rocks being more soluble. This might explain why ing only minor differences for widely differing lo- rills on coarser texture are smaller than those on cations. These observations suggest that rillenkar- finer texture. However, it does not explain why ren width is not controlled by reaction rate. rills on gypsum seem to be so small. We must if the supply of fresh solvent is not inhibited by conclude that at present this question cannot be a laminar layer, then the reactions are reaction- answered. rate limited rather than diffusion-rate limited. grain size should also affect rill length (again Thus, surface roughness should be greater in the in association with reaction rate). if coarse grain belt of rillenkarren compared to the belt of non- is associated with rougher surface texture (and channelled erosion of the ausgleichsfläche – ex- with reaction-rate-limited dissolution within the actly as crowther (1997) observes. rills, that is to be expected), then flow down the reaction rate should also vary with grain size. rill is inhibited so that length to the critical depth While the empirical data are few, there seems to would be shorter. in the absence of adequate data be slightly more evidence that small rills are asso- on length we cannot assess this. ciated with coarse grain for all three rock types. in summary, the data show smaller rills for coarser Challenges to this model grain in carbonates of chillagoe and south africa The only real challenges to this model come from (lundberg, 1977b; Dunkerley, 1983; Marker, 1985), Mottershead. in 1996, while claiming that his best developed rills on microcrystalline gypsum model is “entirely consistent with the conclusions of spain and italy (calaforra, 1996), and smaller of glew and Ford (1980)”, he suggests three fun- rills on coarsely crystalline halite of spain (Mot- damentally different processes. arguing that rain- tershead et al., 2000). publications do not always drop impact will lead to divide reduction and thus give quantitative details of grain size, so as yet it is raindrop action alone would destroy the rills, he not possible to make a table of grain size against suggests that rills can be maintained only because flute dimensions. troughs are deepened by channel flow. He further rill cross section is presumably related to both, claims that downslope channel deepening be- rain splash and reaction rate. For a constant rain comes limited by decreasing aggressiveness rather splash, a slower reaction rate would produce than critical depth, and thus the rill gives way to smaller rills. coarse texture is often quoted as less the ausgleichsfläche (Mottershead, 1996b). 207 KRF•1 • OK.indd 207 15.12.2009 10:45:04 Karst Rock Features • Karren Sculpturing We offer several counters to these arguments. rill magnitude. Their final point is that the rill First, simple modelling of the dissolution from long profile is produced by differences in channel raindrops falling vertically onto a parabolic pro- downcutting rates (rather than rill divide erosion file demonstrates that the form is maintained and rate – in apparent contradiction to the designa- does not require any additional trough deepening. tion of channels as subordinate forms): at the second, if the troughs require focussed erosion by top of the rill both, mechanical shear forces and channel flow, then they require an upstream area chemical aggressivity, will be most effective; mid- of sheetflow (with no channelled erosion) for col- way down both forces will be balanced; towards lection of discharge, in the standard Hortonian the base both will be weakest (because increased fashion. Yet rillenkarren are usually close to their depth of flow reduces the chance of direct rain- maximum depth right at the crests. Third, glew beat on the rock surface). and Ford in 1980 had already rejected the argu- obviously, we have no argument with some ments of changing aggressivity. For the highly parts of the model, but we suggest that the focuss- soluble evaporite rocks, water certainly does not ing of any process (mechanical or chemical) on approach saturation, yet the rills still naturally ex- the inter-rill divides will necessarily cause them tinguish downslope and give way to an ausgleichs- to be lowered while the troughs remain relatively fläche at the same slope angle. unaffected. The form must therefore lose its integ- By 2004 Mottershead has shifted focus, offering rity over time. Modelling of dissolution at point instead a kind of omni-inclusive model incorpo- of raindrop impact (dissolution, related to cosine rating raindrop impact and critical depth but with of slope, occurs by surface reduction normal to a focus on mechanical stresses (imposed by rain slope) shows that the maximum effect is in the drop impact rather than channel flow) and a focus troughs and that the form is maintained over on the processes on inter-rill divides rather than time (Figure 9, above). if we add mechanical ero- troughs (Mottershead and lucas, 2004). The new sion (shear, related to sine of slope, acts vertically), (rather complex and confusing) model suggests: then the trough-to-divide distance decreases and 1. that rill troughs are “subordinate forms” to rill the rills eventually extinguish (Figure 9, below). divides; 2. that rainbeat is a significant force caus- ing mechanical damage on the steep slopes of the Limitations of models rill divides but not on the flatter trough (because it is likely that there is some truth in every model, the proportion of shear force versus compression but at present, the available data offer the most will increase with slope); 3. that the steep slope support to the simple rain-drop impact and criti- of the rill divide is also the region of maximum cal depth model. There certainly are a great many chemical action, with the maximum surface ex- variations in nature, but on the whole, the ma- posed to dissolution – the solvent is spread more jority of evidence from rill morphometry studies thinly, bringing more of the solvent into contact supports glew and Ford’s (1980) model. Howev- with the rock – but they also say that the maxi- er, Dunkerley’s (1983) arguments, that many ob- mum dissolution is at the point of rainbeat impact served characteristics of rillenkarren are still not (up to a critical depth); 4. that the troughs have explained by this model, should not be forgotten. the deepest water film, maximizing the solvent While it may be the best explanation we can volume and increasing the potential for dissolu- come up with, there are some problems apparent tion; 5. that the variation of dissolution kinetics with the raindrop impact theory. Dreybrodt (per- with lithology and the concentration of solvent in sonal comment, 2006) has pointed out that rain- channels explains why the “relatively deeper flutes drop impact on any one point is actually rather are present on the more soluble rocks” (sic) – but rare: for rainfall intensity of 10 mm/hr (heavy they also say that mechanical strength governs rainfall) and drop size of 1 mm radius (volume 208 KRF•1 • OK.indd 208 15.12.2009 10:45:05 Joyce Lundberg and Angel Ginés, Ril enkarren 0.004 cm3), only 4 drops fall onto a 1 cm2 area every minute. Furthermore, the work of petterson (2001) indicates that laminar flow is the norm and turbulent flow rare. another consideration that is not fully under- stood is the effect of reaction kinetics on rill for- mation. Flow velocities are typically ~2 cm per second but it takes about 20–30 seconds to reach apparent equilibrium (see chapter 2). logically, rills should not reach their maximum size until the flow has reached ~40–60 cm. Thus short rills and rills that head at the crest should not be possi- ble. enhanced dissolution under raindrop impact, and within only a second of impact, seems to be the only explanation that applies to all situations. as far as we understand, no model has fully explained why the region of randomly impinging raindrops should develop into such neat parallel channels rather than random surface lowering. glew and Ford’s model explains why the form  develops at the crest, why it extinguishes, and how it is maintained. once a rill has attained the parabolic profile, it is easy to maintain its equilib- Figure 9: Profiles produced by model ing: above. result rium profile solely by raindrop impact. raindrop of dissolution alone, where solvent from vertical y fal - impact is delivered uniformly. The surface area ing raindrops is distributed according to surface area over which the impact is dissipated increases with and removal of molecules is parallel to slope. The form slope angle of each facet. Depth of surface lower- is maintained over time; below. result of dissolution in ing is thus a direct function of surface area ex- combination with mechanical erosion as envisioned by Mottershead and Lucas (2004). The form is destroyed posed to rain. a curved form, once established, is over time. maintained. This simple concept works if dissolu- tion is the only process. it also works for mechani- cal action if kinetic energy is dissipated according dissolution alone. Biological and mechanical ac- to surface area, ignoring potential differences in tion may then be added for susceptible lithologies, shear versus compressional forces. What is diffi- as modifiers of rill form and dimension. cult to model, is the initial, highly ordered pattern of differential erosion that sets up the profile. We feel that simple dissolution from raindrop Conclusions impact explains the basic formation of rillenkar- ren on all rock types. rills in carbonates are prob- rillenkarren have been reported from a varie- ably the result of a combination of dissolutional, ty of geographic locations worldwide. We have biological, and mechanical action. However, be- summarized the available data for around 20 lo- cause biological action is not observed on gypsum cations ranging from arctic to tropical climates. or halite, and because mechanical action is prob- rillenkarren width is remarkably constant world- ably not active on the smooth surfaces of halite, wide, while depth and length vary with slope/ any model of rill formation must focus first on temperature/altitude. none of the morphometric 209 KRF•1 • OK.indd 209 15.12.2009 10:45:05 Karst Rock Features • Karren Sculpturing properties varies with mean annual precipitation. locality. The relationship of altitude and climatic Because width is so constant, rillenkarren may be variables needs to be clarified. The complexity of defined by their width: any suite of channels of the relationship of slope and rillenkarren length mean width outside the range ~1.3 to ~2.1 cm are needs further study. While the focus on detail and probably not simple rillenkarren. consistency of studies recommended by crowther The best explanation of rillenkarren formation (1998) is important, the value of the research lies is that it is by rapid dissociation where raindrops mainly in the sampling strategy, which should be impact directly onto the rock surface, uninhib- clearly stated. sampling efforts should be organ- ited by the laminar boundary layer. The kinetic ized around ecological principles, to elucidate the energy of the raindrop dictates the critical depth impact of environmental variables. experiments of flow that can be penetrated. The belt of non- on the chemistry of natural and simulated runoff channelled erosion, the ausgleichsfläche, develops waters are vital. Future research efforts could also where the critical depth is exceeded. rill dimen- profitably be focussed into more simulations in- sions are controlled largely by raindrop kinetic troducing additional controls, such as grain size, energy and water viscosity. The effect of lithology raindrop size, temperature, initial slope charac- is not clear but there is a suggestion of smaller rills teristics. on coarser-grained rocks. Future research efforts should focus on the areas of weakness. Morphometric data on ril- Acknowledgements lenkarren are reasonably comprehensive (al- though length data are not often available). How- We thank Wolfgang Dreybrodt for very help- ever, data on the ausgleichsfläche are rare. We ful comments on an earlier version of this chap- need considerably more data from regions other ter. This work was partially supported by the re- than the two favourites, australia and Mallorca. search fund of Ministerio de educación y ciencia ideally a search should be made for rillenkarren – FeDer, cgl2006–11242–c03–01/Bte. on limestone, gypsum, and salt within the same 210 KRF•1 • OK.indd 210 15.12.2009 10:45:05 rinnenKarren 17 Márton VERESS Rinnenkarren are solution channels ( runnels, flu- ded solution channels. several researchers explain tes) that occur parallel with each other and whose the development of this form by dissolution under direction coincides with the line of the dip of the soil (eckert, 1902; Bögli, 1976; Jennings, 1985; slope. according to eckert (1898), Bögli (1976), sweeting, 1955). other researchers think that and Ford and Williams (1989), rinnenkarren are rundkarren could have developed when rinnen- several decimeters wide and deep, and they can be several dozen meters long. They are large forms that cover extensive areas and do not taper out at the ends (Figure 1). according to Haserodt (1965), rinnenkarren occur between the altitudes of 480 and 2,300 meters in the alps, and according to Kunaver (1984), between 1,650 and 1,700 meters in the Julian alps. They can develop parallely on steeper slopes (as mentioned above), but they can also be separated as main channels and subsidiary channels on gentler slopes where they create inter- locking systems. large channels can have depths and widths of about one meter (Veress, 1995). ac- cording to Wagner (1950), rinnenkarren occur on slopes between 30° and 90°, but we believe rather that wandkarren develop on the steeper slopes (ca. 60°–90°). The types of rinnenkarren rinnenkarren can be divided into several types. according to Bögli (1976), the surface remnant be- tween channels can be flat or round. Bögli (1976) Figure 1: Rinnenkarren (Triglav Lakes valley, Slovenia). called the latter karren form Rundkarren or roun- Width of view is 3.5 m. 211 KRF•1 • OK.indd 211 15.12.2009 10:45:06 Karst Rock Features • Karren Sculpturing Figure 2: Semi-ex- humated rounded rinnenkarren (Totes Gebirge, Austria). karren were transformed during their develop- along the slope, taking a branched form. They are ment (Bögli, 1960a; Haserodt, 1965; louis, 1968; fed from the upper part of the slope (for example, Wagner, 1950). Those surfaces that are covered from an Ausgleichsfläche area), but in our opinion with soil today were bare during the glacial peri- they receive further water from their margins on od, and therefore rinnenkarren could have devel- the lower part of the slope. according to gladysz oped on such surfaces. These rinnenkarren were (1987), they are complex forms ( complex chan- covered with soil after the retreat of the ice in the nels) from about three to five meters below their Holocene, and due to dissolution under the soil, upper ends and their development is also complex. the ridges between the channels were rounded off. There can be pits ( karren sinkholes) and grikes on The side walls of the channels dissolved steeply (re- their bottoms. according to Ford and Williams sulting in a u profile), depressions could develop (1989), there are single rinnenkarren that receive at the bottom of channels ( Korrosionshohlkehlen), their water directly from rain whose width and and the slope angle of the bottom of the channels depth decrease farther down the slope. There are became smaller (Bögli, 1976; White, 1988). The also decantation channels fed by kamenitzas. The side walls of the channels can also be overhang- water supply of decantation channels is therefore ing due to dissolution under the soil. Bögli (1976) local, but they can also receive water dropping called this type Hohlkarren. The soil can erode from leaves and the trunks of trees. according to and the rounded ridges partly or totally protrude sauro (1976b), all rinnenkarren are decantation (Figure 2). channels. according to Ford and Williams (1989), according to Jennings (1985), flutes and small- “decantation flutes” are forms of large density with er channels ( rain solution channels) form on the narrow combs between them. These forms are fed ridges between channels. We consider the latter to by sheet water from the upper part of a slope. be ril enkarren. Bögli (1960a) described Regenrinnenkarren, Ford and Williams (1989) describe one rinnen- forms that develop on a steep slope due to rainwa- karren type as Horton-type channels whose forms ter. However, this form type is probably a decanta- become larger and more complex downward tion channel or a special wandkarren. according 212 KRF•1 • OK.indd 212 15.12.2009 10:45:07 Márton Veress, Rinnenkarren to Jennings (1985), this type can develop with the trum and Mason, 1948), on quartzite (White et al., coalescing of neighbouring rills. crowther (1997) 1966; Marker, 1976a; White, 1988), on halite (Ma- distinguishes different types of rinnenkarren ac- caluso and sauro, 1996a), and on calcareous green- cording to their vertical section: those with a flat schist (Veress and szabó, 1996; Veress et al., 1996). bottom, with a stepped bottom ( step rinnenkar- large channels can develop in particular on gran- ren), and with a changing bottom slope angle ite and halite. The rinnenkarren on halite (parajd, ( bevel rinnenkarren). romania) begin on covered sediments, and they rinnenkarren can develop on a variety of rock. can be several meters wide. rills and channels can For example, they can occur on marble (Veress et develop on their bottoms. rinnenkarren on mar- al., 2006), on granite (rassmusson, 1959), on cal- ble (Diego de almagro island, chile) are also large. careous sandstone, on calcareous conglomerate, Here, the bottoms of the rinnenkarren are stepped on amorphous silica sandstone (Veress and Kocsis, and they join dissolution basins. 1996), on gypsum (calaforra, 1996), on basalt (Bar- combinations of various karren forms occur on      Figure 3: The morphological complexes of different parts of a channel. a, b. upper part of the channel; c, d. middle parts of the channel; e. lower part of the channel; 1. dip of the surface. 213 KRF•1 • OK.indd 213 15.12.2009 10:45:08 Karst Rock Features • Karren Sculpturing slopes. The most common combinations of forms type iii. Complex rinnenkarren have “simple are the following: ril enkarren–Ausgleichsfläche complex” and “manifold complex” forms. in the ( solution level)–rinnenkarren; Ausgleichsfläche– case of simple complex rinnenkarren, only type rinnenkarren; and ril enkarren–rinnenkarren. ril- iii rinnenkarren develop in a type i rinnenkarren lenkarren can develop under sheet water where the (although occasionally two type iii rinnenkarren current of the water is turbulent from time to time. can develop at the bottom of a type i rinnen- Dissolution is small and superficial on the aus- karren). Scal ops can occur at the bottom of type gleichsfläche. rinnenkarren develop on the slope iii rinnenkarren. The rate of the size increase of if the water separates into streams (Bögli, 1960a, the bearing rinnenkarren and the rate of the size 1976; trudgill, 1985; Ford and Williams, 1989). increase of the inner rinnenkarren can be congru- ent or incongruent. in the case of manifold com- plex rinnenkarren, type ii rinnenkarren occur Morphology of rinnenkarren inside type i rinnenkarren, and type iii rinnen- karren inside type ii rinnenkarren (Figures 5, 4a. to describe rinnenkarren, we must consider their e; Veress, 1995, 2000b, 2002). environment and their morphology. Trittkarren, Terraces can develop at the bottom of the chan- leafkarren (szunyogh et al., 1998), and kamenitzas nels (Veress, 1993) in wide complex rinnenkarren. can occur near rinnenkarren and feed the rinnen- terraces are the remnants of former rinnenkarren karren. Hanging kamenitzas are often connected bottoms. The surface of terraces slopes toward the to a main channel by a small subsidiary channel. middle of the channels and to the end of the chan- The morphology of rinnenkarren varies accord- nels. The terraces drop in gently concave stages ing to sections of the channels (Figure 3). Forms from the edges of the main type i rinnenkarren with a water-feeding function such as kamenitzas but have sharp edges where they meet the steeper are characteristic of the upper sections of rinnen- walls of a lower channel. karren, while forms created by water flowing on according to Veress (1995, 2000c, 2002), steps, the channel bottoms (steps, bottom basins) char- bottom basins (Figure 6), and bottom kamenitzas acterize the middle section of the channels. Vari- occur at the bottom of channels. Potholes, which ous types of pits increasingly dominate the mid- can coalesce into one another, occur at the bot- dle and lower sections of the channels. toms of the basins. The cross-section of a channel can take sim- The pits or karren sinkholes in rinnenkarren ple (Figure 4a.a–d) or complex forms (Figures 5, are called bottom channel pits and there are sever- 4a.e). The shape of the ridges between the chan- al types. Chimney pits can develop where rinnen- nels depends on the density and shape of the karren are crossed by crevices or grikes (Figures channels (Figure 4B). The cross-sections of simple 3e, 7). soil and plants can accumulate in the pits, rinnenkarren can be V, u, , or (Figure 4a.a– and the diameter of occupied pits is greater than d; Veress, 2000b). those without soil and plants. The so-called chan- We group simple rinnenkarren according to nel-end pits occur in the lower sections of rinnen- the size of the cross-section (Veress, 1995, 2000b, karren (Figures 3e, 7). Their upper edges coincide 2002). The widths and depths of Type I rinnenkar- with the upper edges of the channels. ren are several decimeters, while the widths and in general, the development of karren caves is depths of Type III rinnenkarren are only several similar to the development of rinnenkarren but centimetres. The smaller type iii rinnenkarren occurs below the surface. Karren caves develop can develop on surfaces without other karren along joints or bedding planes and can be sink- formations. The widths and depths of Type II hole karren caves, spring karren caves, or through rinnenkarren lie between the sizes of type i and karren caves (Figures 3d, 8, 9). 214 KRF•1 • OK.indd 214 15.12.2009 10:45:08 Márton Veress, Rinnenkarren Figure 4: Types of channels (A) and ridges between channels (B) in profile. A: a–d. simple channels; e. a complex channel; B: a. flat and narrow ridges between channels; b. flat and wide ridges between channels; c. rounded ridges between channels (round karren); d. ridges between channels that have become sharp; e. flat ridges between channels that become narrow downwards; f. rounded ridges between channels that become narrow downwards; g. wide ridges between channels with smal channels; h. windows through ridges between channels; 1. limestone; 2. former channel. 215 KRF•1 • OK.indd 215 15.12.2009 10:45:08 Karst Rock Features • Karren Sculpturing Figure 5: A complex channel (Totes Gebirge). Width of Figure 6: A channel bottom with basins and steps view is 50 cm. 1. Type I channel; 2. Type I channel; 3. (Triglav Lakes valley). Width of view is 35 cm. 1. step; 2. Type I I channel. bottom basin; 3. pothole; 4. debris. The development of rinnenkarren and was only 60 mg/l. according to trudgill (1986), their forms rinnenkarren can also develop from subsoil dis- solutional forms. groups of karren forms such as rillenkarren, We believe the following factors (in combi- rinnenkarren, wandkarren, and meänderkar- nation or separately) cause the development of ren are created by flowing water on bare slopes rinnenkarren: (White, 1988; Ford and Williams, 1989). rinnen- • the development of rinnenkarren occurs under karren can be formed by rivulets as the limestone turbulently flowing rivulets. The laminar flow dissolves under them (Ford and Williams, 1989). becomes turbulent if the value of the reynolds according to parry (1960), they could have de- number is higher than 1,500–6,000. The value veloped during the ice ages due to abundant melt of the reynolds number depends on the angle water. He theorized that the quantity of co in the of the slope (and furthermore on the change 2 atmosphere was higher than it is today, although of the slope angle), on the depth of the flowing this theory has not been confirmed by measure- water, and on the roughness of the surface (em- ments. according to smith (1969), the quantity mett, 1970). The roughness of the bottoms of of caco in the melt water on somerset island the channels is considerable (crowther, 1997), 3 216 KRF•1 • OK.indd 216 15.12.2009 10:45:09 Márton Veress, Rinnenkarren and therefore the development and the main- tenance of the turbulent flow can be caused by the roughness of the bottom of the channel as well. Because of the turbulent flow, the bound- ary layer is broken (curl, 1966; Ford, 1980; trudgill, 1985). a new boundary layer develops repeatedly and because it is unsaturated, ca2+ ions can enter it from the limestone; • plant and soil patches since they produce more co (Jennings, 1985; Ford and Williams, 1989); 2 • mixing corrosion: Zentai (2000) showed that the area of the diameter of the main channel below two coalesced rinnenkarren is larger than the diameter lengths of the two subsidi- ary rinnenkarren added together. Therefore the solubility of the water increases where the wa- ters of two subsidiary rinnenkarren join each other; • the difference in the ion concentration between the boundary layer and flowing water is great- er if the velocity of the current is higher; there- fore, the quantity of the ion transport increas- es outside the boundary layer (Dubljanszkij, Figure 7: Pits (Totes Gebirge). Width of view is 3.5 m. 1. 1987). Therefore, more ions are able to enter the bedding plane; 2. head of the bed; 3. channel bottom boundary layer from the limestone; pit; 4. channel end pit; 5. channel; 6. grike. • co enters the flowing water from the atmos- 2 phere (Jennings, 1985); • according to Mariko et al. (1994) and Körner (1999), the quantity of co is high if the plants 2 are covered with snow (and the snow is solid). This phenomena can be explained by the fact that the plants are unable to photosynthesize but are able to dissimilate. For this reason the average diameter of channels found in a meter distance on bare slopes is 3.65 dm2/m, accord- ing to our measurements, while this value is 9.35 dm2/m on slopes covered with dwarf pine. We measured the density of karren forms (in- cluding rinnenkarren) and their specific dissolu- tion (the overall width of the karren forms to a distance of one meter) in several high mountain areas (Dachstein, totes gebirge, Julian alps). The Figure 8: Karren channel cave system (observation, from decreasing of the values of this data is only small Veress, 1995). 1. bedding plane; 2. through karren chan- nel cave; 3. karren channel swallet; 4. karren cave swal- in the course of the increase of the altitude. The let; 5. debouchure; 6. swallet type karren cave; 7. spring density of rinnenkarren is 0.87 piece/meter at karren channel cave; 8. retreat of bathycapture; 9. grike. 217 KRF•1 • OK.indd 217 15.12.2009 10:45:09 Karst Rock Features • Karren Sculpturing lets can be sheet water or local sources. The lat- ter can include soil patches, kamenitzas, leaf- karren, trittenkarren, and cave karren; • the further growth of the channels is due to percolating or flowing water. When the perco- lating water originates from the snow filling the channel, the dissolution occurs everywhere in the channel and its quantity is the same. How- ever, the amount of dissolution is small since the co content of snow is also small. in the 2 case of channel development generated by a riv- ulet, the channel becomes deeper and wider if the rivulet fills the channel completely. if this does not happen and the water flow decreases quickly, dissolution occurs mainly at the bot- tom of the channel. The growth of the chan- nels on bare slopes is mainly due to percolat- ing water as the water flow and co content of 2 the rivulets are small. since the co content 2 of such snow is small, the growth of the chan- nels is negligible. The growth of the channels is significant on slopes with Pinus mugo. The greater growth of the channels is due to flow- Figure 9: Sinkhole karren cave and through karren cave (Totes Gebirge). Width of view is 65 cm. 1. Type I chan- ing water since melt water originating from the nel; 2. sinkhole karren cave; 3. through karren cave; 4. snow covering the dwarf pines has a high co 2 unroofed karren cave. content. The different channel shapes can develop in two different ways. either the sheet water originating 1,700 meters (in the pine belt), 1.18 piece/meter at from the boundary surface or the water (rivulet) 1,800–2,000 meters (in the Pinus mugo belt), and flowing in the channel dissolves the side slope of 0.79 piece/meter at 2,100 meters (on bare surface). the rinnenkarren. However, sheet water can not The specific dissolution of the rinnenkarren is dissolve the channels under the following condi- 12.83 cm/m, 20.29 cm/m, and 9.47 cm/m at these tions: altitudes respectively (Veress et al., 2001a). • the margins of channels (except of roundkar- We can explain this with two factors: ren) are sharp; • the quantity of co can be high in solid snow • the side walls of some rinnenkarren are unbro- 2 even on surfaces with no soil cover because the ken and free from smaller karren forms (Figure co cannot escape the snow; 5). rillenkarren occur in the neighbourhood of 2 • the higher the altitude, the longer the melting some rinnenkarren. These rillenkarren cannot of snow takes. The water is therefore in contact continue at the margins of the channels. Hence, with the rock for a long time. the water is saturated before it enters the chan- We distinguish two phases in the development nel; of rinnenkarren: • the channel wall slopes similarly (except with • the development of channels begins under riv- meanderkarren) along its whole length; ulets (embryonal phase). The origin of the rivu- • the line of dip of the bearing surface and the 218 KRF•1 • OK.indd 218 15.12.2009 10:45:10 Márton Veress, Rinnenkarren direction of the channel are equal. Therefore water from the boundary surface cannot flow into the channel. sometimes we observe that channels can develop on ridges and water can not flow into the channel because of the ridges. The shape of a channel is determined by the quantity of the water present and its changing quantity over a period of time. These character- istics depend on several factors including: the ve- locity of the current (which depends, for example, on the angle of the bearing slope) and the quan- tity of the recharge (which depends, for example, on the thickness of the snow and the intensity of the melting). in an active period, the dissolution works in a narrower width because the quantity of water decreases. in this case, the shape of the rinnenkarren takes a V-form in cross-section. The wall of a channel is perpendicular if the quantity of water does not change over a longer period. internal channels develop (type ii or type iii rinnenkarren) in a channel (which could develop earlier) if the quantity of water decreases consid- Figure 10: Characteristic cross-sections of an evolved erably and subsequently does not change further channel in various stages of development (Totes Ge- birge, from Veress, 2000b). 1. roof destroyed by col- for a long period. lapse; 2. roof destroyed by merging of a channel and a Karren terraces develop if the older channel karren channel cave; 3. Type I I channel; 4. upper chan- can not grow. The internal (younger) and smaller nel section developed by surface solution independ- channel can partly consume the bottom of the ently from bathycapture; lower channel section inten- older channel. terraces can also develop because sively developing due to beheading; 5. channel section of bottom channel pits since internal channels de- developed by subsurface solution. In the A-A' and B-B' cross-sections, the evolution occurred with the col- velop by corroding backwards from bottom chan- lapse of the roof of the karren channel cave; in the case nel pits. of the C-C’ cross-section, it occurred through coalesc- The deepening of a channel can take the follow- ing; bathycapture occurred between the C-C' and D-D' ing forms: sections. • the channel deepens in the direction of the channel head and the channel becomes longer during its development. The head of the chan- nel extends up the slope; • the depth of the channel decreases below the • the depth and length of the channel develop lower end of the channel as it extends down the uniformly as the head of the channel extends line of dip of the slope; up the slope; • some sections of the bottom of the channel are • the depth of the channel increases uniformly dissolved to varying degrees. while its length does not change; in the first two cases, the development of the • the depth of the channel increases downward channel is regressional. The causes for this phe- along the dip of the slope while its length does nomena are the following. rivulets develop along not change; the line of dip of the slope, and such currents have 219 KRF•1 • OK.indd 219 15.12.2009 10:45:10 Karst Rock Features • Karren Sculpturing is characteristic of decantation channels. “Bevels” develop in the sixth and seventh cases. We can also distinguish other channel develop- ment processes: • from “leafkarren” (szunyogh et al., 1998); • from trittkarren (Veress and tóth, 2002); • from scallops (curl, 1966); • from ripple karren on marble (Diego de alma- gro island; Veress et al., 2006); • from the opening-up of karren caves (Veress, 2000b, 2002). We observed eddies (Veress, 2000c; Veress et al., 2006) in the bottom basins of rinnenkarren and at the bottom steps of rinnenkarren in the Julian alps and on Diego de almagro island. We there- fore believe the development of these forms occurs at eddies. capture pits and channel pits can de- velop at bathycaptures. The sites of bathycaptures develop back toward the ends of the rinnenkarren. The development of karren caves can largely be linked to rinnenkarren (Veress, 1995, 2000b, 2002). Karren sinkholes are capture pits and channel pits. it can sometimes happen that a karren cavity swal- let develops first through dissolution at a bedding plane, and then a channel develops from the swal- Figure 11: Channel and karren cave which are partly coa- lescing (Totes Gebirge). Width of view is 80 cm. 1. chan- let ( blind channel). Karren caves can develop like nel; 2. former karren cave; 3. evolved channel; 4. roof deltas if there are bathycaptures at the bottom of a remnant. channel. spring karren caves develop below each other if bathycaptures occur in a karren cave. it is also possible that a karren cave network of several levels develops. an evolved channel develops when the highest velocity. in the first case, the develop- a channel coalesces with a karren cave under it. ment of the channel is exclusively regressional; in This can occur through dissolution and collapse. the second case, it can be partly rain-furrowed at the beginning of the process karren windows (regressional rain-furrowed channel development). and karren bridges develop. These rinnenkarren in the third and fourth cases, the length of the channels have a double cross-section (Figure 10). channel does not change and therefore their de- These double cross-sections occur at the karren velopment is rain-furrowed. The rate of deepen- swallet of earlier karren caves. The upper section ing can increase toward the lower section of the of the channel grows narrow in its cross-section channel due to the increase of the current veloc- (a former channel), while the lower section of the ity. The deepening of the channel results in a hori- channel is basically circular in cross-section (a zontal movement. The dissolving of the bottom of former karren cave). roof remnants can remain the channel is more significant in the lower parts in the wall. than in the upper section of the form. in the fifth according to Veress and tóth (2001), channels case, the development is antiregressional, which can coalesce during regression. The process oc- 220 KRF•1 • OK.indd 220 15.12.2009 10:45:10 Márton Veress, Rinnenkarren Figure 12: Coalescing of channel ends (a) and false beheading (b, c, d) (from Veress and Nacsa, 1998). 1. Type I chan- nel; 2. slope direction of Type I channel bottom; 3. channel bottom watershed divide; 4. step; 5. limestone; 6. ground surface before channel entrenchment; 7. master channel; 8. regressing tributary channel; 9. step; 10. chan- nel bottom watershed divide; Ieft. view from above; right. cross-section; a. the regressing tributary channel heads join; b. the beheading channel deepens constantly; c. the beheading channel retreats; d. the beheading channel entrenches to the bottom of the main channel. 221 KRF•1 • OK.indd 221 15.12.2009 10:45:11 Karst Rock Features • Karren Sculpturing curs when the ends of the channels coalesce into heading, either a channel end reaches the wall of each other (Figure 12a). it can also happen that another channel or several channel ends coalesce the end of a regressing channel reaches or cuts into each other and the direction of the flow of through the side wall of another channel. This is water along the bottom of the beheaded channels called “false beheading”. in the event of false be- will change partly or totally. The water of the be- heading, the original direction of the flow of water headed channel will run partly or totally into the along the bottom of the “beheaded” channel does beheading channel. not change (Figure 12b-d). in the case of “real” be- 222 KRF•1 • OK.indd 222 15.12.2009 10:45:11 MeanDerKarren 18 Márton VERESS Meanderkarren are described by Bögli (1976) and on limestone but also on other rock, for example, others (Jennings, 1985; Ford and Williams, 1989) on evaporites (Macaluso and sauro, 1996a; cala- as a special type of rinnenkarren. according to forra, 1996) and marble (Veress et al., 2006). Bögli (1960a), the cross-section of meanderkar- Hutchinson (1996) published a new way of ren decreases along the direction of the slope. We classifying meanderkarren as young and ma- present pictures in this chapter that show the typi- ture types. He distinguishes gutters that have a cal asymmetrical cross-section and the under-de- V cross-section, gorges that have steep sides, and tailed morphology of meanderkarren. a meandering type of the mature type. This latter small micro-meanderkarren ( decantation mi- type is characterized when a smaller meandering cro-meander) are described by Macaluso and channel occurs inside a large straight channel. ac- sauro (1996a), who claim these forms are found cording to Hutchinson (1996), meanderkarren on halite but not on limestone. However, we dis- have two main characteristics: sinuosity and an covered micro-meanderkarren on limestone as asymmetrical cross-section. according to Hutch- well, for example, a few on the sides of an abra- inson (1996), meandering occurs when rinnen- sional solution spike karren on lokrum island karren become old and when the inclination of (croatia). We should also mention that various the surface is between 7° and 14°. He notes that authors describe the morphology of meanderkar- young channels flatten downwards along the slope ren differently. sauro (1973a), for example, regards while older ones do not and that older channels channels that have non-symmetrical cross-sec- have less sinuosity than young ones. tions and changing directions as meanderkarren according to Ford and Williams (1989), mean- (for this type we use the term false meander). in derkarren develop where the flow of the water is his later publication, Bögli (1976) describes these slow, while according to Zeller (1967), they may forms emphasizing their small size and that they be formed where the velocity of the water current are developed by a solutional flow percolating is high (Froud-number 1.8–20) and the velocity from the soil. according to Ford and lundberg of the water current is higher than the velocity of (1987), the sinuous Horton-type rinnenkarren are the currents of river and melt water. Hutchinson associated with meanderkarren, while according (1996) describes the development of meanderkar- to sweeting (1972) meanderkarren are internal ren as a characteristic process when rinnenkarren channels of larger channels. We have already ob- become older, while according to Davies and served that meanderkarren can develop not only sutherland (1980) they are forms that adapt to the 223 KRF•1 • OK.indd 223 15.12.2009 10:45:11 Karst Rock Features • Karren Sculpturing flow (along the line of least resistance). according the parameters applied to rivers. The morphology to Zeller (1967), meanderkarren develop during of the loops of the meanderkarren and the mean- turbulent flow, related as well to when the flow dering rivers may be different, but in both cases changes from turbulent to laminar. However, Zel- the cause of the meandering is the swinging of the ler (1967) also suggests that secondary currents channel line. The differences are as follows: can cause the development of meanderkarren as • the profile of meanderkarren is asymmetrical. well. The bed of an alluvial meandering river may be slightly asymmetrical due to lateral erosion. on the concave side of a meanderkarren the bank Morphology and development of will be steep but it will never be overhanging. meanderkarren on the convex side of the meanderkarren the bank will be gentle, where shoals can develop. Veress (1998, 2000d) distinguishes two groups of if a valley with V cross-section is created by lin- meanderkarren: false and true meanderkarren. eal erosion, its cross-section will be symmetri- False meanderkarren change direction in such a cal; way that the cross-section of the channel is sym- • the morphology of the loops of meanderkarren metrical when true meanderkarren do not develop is largely like that of the valleys of forced mean- during the deepening of the channel (Figures 1a, 2). dering rivers. The cross-sections of the loops of The cross-sections of the meanders of true mean- forced meandering rivers are also asymmetri- derkarren become asymmetrical (Figures 1b, 3). cal. With this valley type the bank of the river like other authors, Veress (1998, 2000d) clas- can overhang somewhat over the concave side sified the morphometry of meanderkarren using of the loop if the river bed developed in rock. Figure 1: False (a) and true meanderkarren (b). 1. vertical channel side; 2. stream line; 3. edge of concave channel side; 4. moderately sloping side of channel (skirt); 5. cross-section; 6. bedrock; 7. meander scour grooves (from Veress, 2000b). 224 KRF•1 • OK.indd 224 15.12.2009 10:45:11 Márton Veress, Meanderkarren The side wall of a meanderkarren overhangs on the concave side of a meandering channel while the slope of the wall is gentle on the convex side of the meandering channel. Veress calls this gen- tle side wall a skirt. The form of the skirt is very varied both in profile and seen from above. ac- cording to the data measured by Veress (1998, 2000d), at the meanderkarren of a channel the degree of inclination of the opposite walls is very different. He established that the size of the over- hang is less than the lateral extension of the skirt. skirts sometimes can extend beyond the edge of the overhanging wal . The shape of a skirt is half- conical or half-pyramid, seen from above (Figure 3). if the meandering is complex, the skirt will be complex as well. skirts frequently become de- tached and form oxbows and islands ( karren in- selbergs) for many reasons (Veress, 2000b; tóth and Balogh, 2000), for example, when the flowing water at the bottom of the channel hits a skirt and dissolves its way through it. The roof of the tun- nel that develops this way will eventually collapse. The process can also occur when some of the water flowing at the bottom of the channel overflows at Figure 2: False meander (Julian Alps, Slovenia). a karren neck and dissolves the rock, resulting in a detached form (Figure 4). Figure 3: True meander (Julian Alps, Slovenia). Width of view is 35 cm. 1. meander scour grooves. 225 KRF•1 • OK.indd 225 15.12.2009 10:45:12 Karst Rock Features • Karren Sculpturing Figure 4: Detached loop–neck (Totes Ge- birge). Width of view is 1.5m. 1. detached kar- ren oxbow; 2. karren neck; 3. karren recess (karren “inselberg”). Meanderkarren scour grooves can develop on a new boundary layer wil develop repeatedly and the overhanging side walls of meanderkarren. because it wil be unsaturated, ca2+-ions can enter They occur one below another at several levels it from the limestone. Veress (2000d) therefore ex- (Figure 3) and are like small horizontal channels plains the development of the asymmetrical cross- on the side walls. Meanderkarren terraces are sections by the asymmetrical relationships of the embayments of the bottom of the channel on the currents. The channel line wil not stay in the mid- side wall that is below the concave edge. Hanging dle of the channel (in the middle of the rivulet flow- terraces develop when the bottom of the channel ing down the slope); instead, it wil move lateral y is able to deepen more intensively in the middle. and at some points wil even reach the wal of the Meanderkarren terraces can also develop on parts channel ( oscil ation of the channel line). The dis- of the surface of a skirt where the inclination is solving of the channel wal wil be more intensive small. here than on the opposite side of the channel where according to Veress (1998, 2000d), the cross-sec- the velocity of the flowing water is smal er because tions of channels are asymmetrical because the rate the channel line has moved away. The development of solution is different on opposite sides of the me- of meanderkarren as well as the development of anderkarren. He explains this by the fact that the meanderkarren types may be explained by the velocity of the current is different on the opposite oscillation of the channel line (see Types of mean- sides of a channel. The difference in the ion con- derkarre n in the following). centration between the boundary layer and flow- according to Veress (1998, 2000d), the oscilla- ing water is larger if the velocity of the current is tion of the channel line can be explained by exter- higher, and therefore the quantity of the ion trans- nal and internal causes. He regards morphology port out of the boundary layer increases (Dubljan- as an external cause. He observed this in several szkij, 1987). Therefore, more ions pass from the cases, for example, when water from a tributary limestone to the boundary layer. a fast current channel flows into the main channel (in this case, causes turbulence and therefore the boundary layer the main channel meanders locally below the is broken (curl, 1966; Ford, 1980; trudgil , 1985). mouth of the tributary channel). Meandering 226 KRF•1 • OK.indd 226 15.12.2009 10:45:12 Márton Veress, Meanderkarren can also be caused by a false meanderkarren, by where b is the width of the channel. a change in the inclination of the bottom of the The functions show that the wider the river or channel, by calcareous spar lenses, and by an older meanderkarren is, the bigger the curves and the skirt. The channel line can also often oscillate in zone of the meanderkarren are. The above param- a homogeneous environment. in this case, we ex- eters of meandering forms are probably associated plain the process by the changing conditions in with the width of the water flow. the current of the water (internal cause). probably Hutchinson (1996) examined the proportions of due to the current of the water, a wave motion de- the shapes of meanderkarren and their forms on velops that can change the oscillation of the chan- Mal orca. He defined the ratio of the shape (“form nel line of the rivulet downstream. ratio”) as the quotient of the width and the depth of the forms, defining width as the distance be- tween identifiable channel rims. He measured Methods two different width data: “Maximum width” and “Horizontal width”. presumably he calculated the Meanderkarren were first quantitatively described mean of this width data and used it as the width by Zel er (1967) on Mount silbern (switzerland) in his calculations. He measured the depth of along with other different types of meandering in- the forms at two points: “Horizontal mid-Depth” cluding beds of al uvial meandering rivers, forced where the depth is measured in the middle of the meandering rivers, and channels of meandering horizontal width (Dmid Hor), and “Horizontal melt water from a glacier. He used the following maximum-Depth” where the maximum depth is features to describe meandering: sinuosity ( w), measured perpendicularly to the horizontal width length of curve (λ), inclination of the bottom (Dmax Hor). We also think he used a mean depth of the channel ( J ), and amplitude ( A). The last calculated from these two measurements. He s equals the width of the meanderkarren zone. The could calculate the shape of the forms from the sinuosity is calculated as the ratio of the chan- quotient “Al” and “Ar” where “Al” is the cross-sec- nel line and the axis length of the meanderkarren tional “area large” – calculated using the Dmax zone. He measured the sinuosity of meanderkarren, Hor data – and “Ar” is the cross-sectional “area the length of the meanderkarren curve, the width small” – calculated using the Dmid Hor data. He of the meanderkarren zone, and the inclination of found that the average form ratio is larger for young the bottom of the channel. He found the sinuos- channels (9.1) than for mature channels (2.54). The ity of the meanderkarren in the silbern area to be change of the shape is probably less significant; 2.84, the length of the curve of the meanderkarren however, the asymmetry of mature meanderkar- 5.8 cm, the inclination of the bottom of the chan- ren is greater. Therefore, the maximum depth of a nels 60–500o/oo, and the amplitude 6.6 cm. He also meanderkarren channel shifts away from the mid- found that meanderkarren can wander lateral y dle of the channel in the course of its development. and that their side wal s overhang. He describes the Hutchinson assumed there was a direct relation- length of the meanderkarren curve, the width of ship between the shape and the ratio of the forms. the meanderkarren zone which he cal ed amplitude Hutchinson (1996) also investigated the slope ( A), and the width of the meanderkarren channels inclination data relative to the sinuosity of mean- in a function. one linear function can be adapted derkarren. He measured the sinuosity on the axis to the above-mentioned parameters of each of the of abscissa (x-axis) and the inclination on the axis four different types of meandering. These func- of ordinate (y-axis) of a coordinate system. How- tions are the following: ever, he did not do a regression analysis but merely λ=10.0·b1.025, used a line to connect the points. He noticed that A=4.5·b1.0, the amount of sinuosity increases with a decrease of 227 KRF•1 • OK.indd 227 15.12.2009 10:45:12 Karst Rock Features • Karren Sculpturing Figure 5: Constructed channel lines of channel No. 7 (number in parenthe- ses identifies the loop; Totes Gebirge, Austria, from Veress, 2000d). 1. edge of Type I channel; 2. lower edge of skirt; 3. end of gently sloping chan- nel; 4. bottom of overhanging wal at the plane of the channel bottom; 5. inflection point; 6. present channel line; 7. previous channel line; 8. ac- cessory straight along which the Sk k and Sk values can be measured. j 228 KRF•1 • OK.indd 228 15.12.2009 10:45:13 Márton Veress, Meanderkarren inclination and concluded that sinuosity depends of the meanderkarren ( β) as follows using lac- on the degree of inclination. according to Zel er zay’s (1982) method (table 1): (1967), sinuosity (meandering) develops if the angle β mi = of the slope is between 3.44° and 36°, while accord- λ ing to Hutchinson (1996), it occurs if the angle of the slope is between 25° and 40° (although in ma- Table 1: River bend types specified by their develop- ture channels it can occur between 7° and 14°). ment (Laczay, 1982). Veress (1998, 2000d) used different parameters Type of river bend Value of β (described below) to construct maps of seven me- Undeveloped bend <1.1 anderkarren in totes gebirge (austria). The data Developed bend 1.1–1.4 from four of the meanderkarren appears in table Wel developed bend 1.4–3.5 2. He wished to identify the factors that can deter- Ful y developed bend >3.5 mine the oscillation of the channel line in order to describe typical meanderkarren development. The parameters used by Veress (2000d) to de- a constructed channel line of a meanderkarren scribe meanderkarren are shown in Figure 6: wave drawn by Veress (2000d) is shown in Figure 5. length ( λ), the length of the curve ( m), the width i Veress used the channel line as the basis for the of the meanderkarren zone ( m ), etc. He acquires sz various parameters of meandering. He defines these parameters by drawing the oscil ating chan- wave length ( λ) as the smallest distance between nel line and identifying the inflection points. The adjacent inflection points, which mark where the channel line moves to the bottoms of the side wal s channel line is in the middle of the channel. The at the concave channel edges, and by connecting length of the curve ( m) is the distance measured these points we can identify the channel line. i between the neighbouring inflection points along The channel line will become lower and lower the channel line. The width of the meanderkarren due to the deepening of the channel. The chan- zone is the shortest distance between the bound- nel line moves laterally and obliquely downward ary lines of the meanderkarren. The boundary as well during its deepening. Veress (2000b) called lines touch the edges of a channel on its concave this process the “slippage” of the channel line. sides. Veress (2000d) calculated the development if the slippage of a deepening channel is ex- Figure 6: Parameters of a meander (from Veress and Tóth, 2004). 1. channel margin; 2. channel line; 3. inflection point; 4. boundary line of meanders; λ. wave length of meander; m . length of arch of meander, m . width of mean- i sz der zone; b. width of channel. 229 KRF•1 • OK.indd 229 15.12.2009 10:45:13 Karst Rock Features • Karren Sculpturing pressed in units, we can determine the intensity also depend on the width of the channel (Zeller of the slippage of the channel line. The intensity of 1967). Therefore Veress and tóth (2004) employed slippage ( L) can be calculated using the following the idea of the width of the specific meanderkar- i formula: ren zone. They could calculate it from the equa- Sk − j Sk L k = tion as follows: i M m msz = where Sk is the difference between the present and f b j original channel lines measured on the map, Sk is where m is the width of the specific meanderkar- k f the difference between the initial slippage and the ren zone, m is the width of the meanderkarren sz original channel lines measured on the map, and M zone, and b is the width of the channel. is the depth of the channel in the bend. The slippage of the channel line will be greater where the oscillation of the channel line occurs Types of meanderkarren due to an external cause. according to Veress (2000d), the stream characteristics of the surface using the morphology and the measured parame- water flowing downward have little to do with the ters of meanderkarren, Veress (2000b) and Veress oscillation of the channel line. and tóth (2004) classified them into the following The loops can not be ranked using the width of types: “looping,” “remnant,” “developing,” and the meanderkarren zone alone since their values “perishing” (Figure 7). Figure 7: Meanderkarren types (from Veress, 2000d). I. planimetric representation: 1. limestone; 2. Type I channel; 3. skirt starting at channel rim; 4. skirt on lower part of channel side; 5. overhanging wal ; 6. site of section; I . on cross- section: 7. overhanging side wal ; 8. skirt; 9. karren recess; 10. symmetrical cross-section channel and part of chan- nel; 11. asymmetrical cross-section channel; a. false meander; b. “remnant” meander; c. “looping” meander; d. “de- veloping” meander. 230 KRF•1 • OK.indd 230 15.12.2009 10:45:13 Márton Veress, Meanderkarren Figure 8: Morphological map of channel 4 (Totes Gebirge; from Veress and Barna, 1998). Planimetric representation: 1. vertical side wal of Type I channel; 2. gentle side wal of Type I channel; 3. vertical side wal of Type I I channel; 4. plane channel bottom; 5. depth of channel (in centimetres); 6. slope direction of channel bottom; 7. number of meander loop; 8. developing skirt; 9. asymmetrical skirt; 10. half skirt; 11. overhanging wal ; 12. meander terrace on overhanging wal ; 13. meander scour groove and major meander scour groove (position and size of the smal meander scour  231 KRF•1 • OK.indd 231 15.12.2009 10:45:14 Karst Rock Features • Karren Sculpturing it is clear that looping meanderkarren may be their curves is over 30 cm and the average of the distinguished from the other types by using the meanderkarren development value is above 2.5. value of curve-length and the value of the devel- according to Veress (2000d), the slippage of the opment of the meanderkarren. The average length channel line can best characterize the morphol- of the curve of “looping” meanderkarren is 45.75 ogy of a meanderkarren because it can explain its cm, and the average value for the development of a asymmetry. The average intensity of slippage of “looping” meanderkarren is about 3.0. The average the channel line is 0.33 for “looping” meanderkar- length of the curve of “developing” and “remnant ren, 0.17 for developing meanderkarren, and 0.12 meanderkarren” is 21.06 cm and 30.54 cm respec- for remnant meanderkarren. tively, and the development value of these mean- The sinuosity and the width of the meanderkar- derkarren is 1.96 and 2.17 respectively (table 2). ren zone can be different for different mean- When the loops that belong to the different types derkarren types. For example, the sinuosity is 2.1 are compared, the difference is especially great. for the number 11 and 12 loops of the “looping” Thus, for example, in the case of the meanderkar- meanderkarren seen in Figure 5 while this value ren seen in Figure 4 the development value is 4.27. is 1.7 for “developing” meanderkarren and “rem- The value of the development is 2.76 in the case of nant” meanderkarren sections (Figure 8, number loop number 11 of the “looping” meanderkarren 11, 12, 13 loops). The sinuosity values calculated seen in Figure 5, but this value is only 1.66 in the by Zeller (1967) are between 1.1 and 1.7. accord- case of the number 3 “remnant” meanderkarren ing to Hutchinson (1996), the value of sinuosity is in the same meanderkarren. between 1.0 and 1.6 if the slope inclination is be- Meanderkarren can be ranked in the “looping” tween 25° and 40°. meanderkarren type when the average length of The width of the specific meanderkarren zone is Table 2: Averages of parameters (using data measured in the No. 3, 4, 6, and 7 channels; from Veress, 2000d). Meander type Looping Developing Remnant All types cause of swinging external internal external all internal external all internal external n = 3 n = 7 n = 13 n = 20 n = 2 n = 6 n = 8 n = 10 21 length of bend 45.75 23.5 18.6153 21.0576 27.5 33.5833 30.54 25.5 32.6495 (66.71)* wave length 15.125 12.0 10.38 11.19 16.0 13.8333 14.9166 14.0 13.1128 (16.87)* stage of development 3.0233 2.0210 1.8992 1.9601 1.7239 2.6181 2.171 1.8724 2.5135 (4.08)* intensity of slippage 0.3336 0.1056 0.2300 0.1678 0.08 0.1608 0.1204 0.0928 0.2439 (-3.32)* 0.2966** * the number in parenthesis belongs to channel No. 6 (n=7) ** with the data of channel No. 3 (n=3)  groove in the bend are not drawn to scale); 14. estimated channel line, with stream direction (number in brackets indicates type of channel in which channel line developed); 15. estimated inflection point; 16. bifurcation of channel line connected to function; 17. bifurcation of channel line connected to slippage; 18. uniting channel lines; 19. site of cross-section; 20. karren “inselberg” at bottom of channel (with altitude data); 21. slope direction and degree of slope of boundary surface; a. straight channel segment; b. false meander channel segment; c. true meander segment (c 1 “remnant” meander, c “looping” meander, c “developing” meander), the swinging of the channel line can occur due 2 3 to an internal (α) or external (β) cause (β false meandering of channel, β bend and its skirt). Cross-section: I: Type I 1 2 channel; 1. overhanging side wal of concave channel margin; 2. meander terrace on side wal of concave channel margin; 3. meander scour groove on side wal of concave channel margin; 4. ridge between meander scour grooves; 5. skirt; 6. vertical wal of skirt that developed due to solution; 7. bottom of channel. 232 KRF•1 • OK.indd 232 15.12.2009 10:45:14 Márton Veress, Meanderkarren Figure 9: “Looping” meander (Julian Alps). Width of view Figure 11: “Perishing” meander (Julian Alps). Width of is 55 cm. view is 2.5 m, in the middle; 1. asymmetrical cross-sec- tion, 2. symmetrical cross-section. Figure 10: “Developing” meander (Julian Alps). Width of view is 50 cm. 1. complex skirt; 2. partly detached skirt (karren “inselberg”). 233 KRF•1 • OK.indd 233 15.12.2009 10:45:15 Karst Rock Features • Karren Sculpturing Figure 12: Inherited meanders of composite karren channels (from Veress, 1995). I, I , I I. channel types; a. from the be- ginning forced meandering, similar to strained meander; b. from the beginning forced meandering, shifted mean- der; c-d. strained meander; e. free meander. 2.41 for “looping” meanderkarren (Figure 5, loops ren is small. The oscillation of the channel line oc- 11, 12) while it is 1.56 for “remnant” meanderkar- curs in part of a false meanderkarren during the ren (loops 2 to 10 in Figure 5). This parameter is deepening of the channel. considered small for remnant meanderkarren. “remnant” meanderkarren may develop in a “looping” meanderkarren will develop if the straight channel as the sides of the channel are dis- channel line swings due to a direction change in sected by curved sections. The curved sections be- a false meanderkarren at its curvature (Figures 4, come the concave sides of the channel, and where 7c, 9). The length of the curve of a “looping” me- the curved sections meet we can see the peaks that anderkarren is large as is the width of the mean- become the convex sides of the channels where derkarren zone, the sinuosity of the meanderkar- skirts develop. The peaks and skirts occur oppo- ren, and the intensity of slippage. on the other site the center of the concave sections of “remnant” hand, the wave length of “looping” meanderkar- meanderkarren (Figures 3, 7b, 8). The length of 234 KRF•1 • OK.indd 234 15.12.2009 10:45:15 Márton Veress, Meanderkarren the curve of “remnant” meanderkarren is small Meanderkarren can often be complex. in this as is the width of the meanderkarren zone, the case, a smal er younger channel wil develop at the intensity of slippage, and the development of the bottom of a larger older one. When the meander- meanderkarren. For this type of meanderkarren, ing of the internal channel is similar to the mean- the oscillation of the channel line begins in the dering of the bearing channel, the process is cal ed rivulet on the surface before the development of forced meandering because the meandering of the the channel begins. internal channel is “forced” by the bearing channel “Developing” meanderkarren can originate (Figures 12a, b). it is also possible for the meander- if the oscillation of the channel line occurs later ing of an internal channel to be independent of the than when the channel started deepening. These meandering of the bearing channel (it can also be meanderkarren have symmetrical cross-sections that the bearing channel does not meander, Figure in the upper part and asymmetrical cross-sections 8d). When the bearing channel is able to limit the in the lower part. in this type of meanderkarren, meandering of the internal channel, the meander- the vertical side walls will start to overhang at a ing of the latter channel wil be strained (Figure certain depth below the concave channel rim and 12c, d). one type of complex meanderkarren is the begin to slope gently at a similar depth below the “developing-looping” meanderkarren seen on the convex channel rim (Figures 7d, 10). marble of Diego de almagro island. This type de- Perishing meanderkarren occur rarely. This velops from “remnant” meanderkarren. The loops type will develop if the oscillation of the chan- border the large skirts that developed from the side nel line ends during the deepening of the channel. wal s of the “remnant” meanderkarren of the bear- Therefore at the lower part of the channel a sym- ing channels. metrical cross-section develops (Figure 11). on the concave side the wall of the channel changes from overhanging to vertical. 235 KRF•1 • OK.indd 235 15.12.2009 10:45:15 KRF•1 • OK.indd 236 15.12.2009 10:45:15 WanDKarren 19 Márton VERESS Ril enkarren, rinnenkarren, and wandkarren (wall liams, 1989). Jennings (1985), for example, calls karren) are created by water flowing on slopes wandkarren wal solutional runnels. (Ford and Williams, 1989). Wandkarren devel- Wandkarren occur mostly on cuestas (Figure 1), op on vertical slopes (for example on the walls of in karst forms, on the walls of shafts, on the sides shafts). They are parallel to each other and have of pits, on the walls of poljes, and on the sides of a semi-circular cross section (Bögli, 1960a). ac- dolines. They can also develop on steep coasts, on cording to german researchers, wandkarren can the sides of boulders, in caves (choppy, 1996), on be independent karren forms (Bögli, 1960a) but in slopes shaped by glaciers on cuestas, erosion steps the english research literature, these forms are de- on the sides of roche moutonnnée rocks, on the scribed as a type of rinnenkarren (Ford and Wil- slopes of horns, and on the slopes of glacier val- Figure 1: Wandkarren on cuesta (Totes Gebirge, Austria). 1. schichtfu- genkarren; 2. wandkar- ren that wedge out at schichtfugenkarren; 3. wandkarren that cross schichtfugenkarren; 4. complex wandkarren. 237 KRF•1 • OK.indd 237 15.12.2009 10:45:16 Karst Rock Features • Karren Sculpturing leys. Their altitude distribution is varied as they width is between 4 and 12 centimetres. on Diego can occur near sea level (Diego de almagro island) de almagro island (chile), the size of wandkar- and above the snow line in the alps. Wandkarren ren is great, and their width can be several me- can develop on limestone, on marble (Diego de tres. taking all the studied forms into account, almagro), on granite (corsica, Mongolia), and on we can establish that the most forms which be- halite (atacama desert). long to a certain width interval can be found at lines marked D-5/2000 and D-13/2000. at the line marked D-5/2000, 43.75% of the forms belong to Methods the width interval of 4 to 6 centimetres and in the case of D-13/2000 line 33% of the forms belong to Veress et al. (2003) investigated wandkarren with the width interval of 8‒10 centimetres. The shape the following methods. They measured the width, of the wandkarren in Dachstein is between 0.14 depth, and direction of wandkarren, and the in- and 28. The greatest frequency of the shape is be- clination of the bearing slopes along seven lines. tween 2 and 4 (line marked D-13/2000, 71.42%) using the measured data, they calculated the spe- and 0 and 2 (line marked D-4/1999, 80%). The cific dissolution of the wandkarren (ratio of the density of wandkarren can range between 1.49 total width of the forms and the length of the line), and 14.55 pieces/meter, while the value of the spe- their density (ratio of the number of forms and cific dissolution is 19.37‒39.31 cm/m (table 1). the length of the line), their cross-section shapes Wandkarren develop mostly along a down- (ratio of the width and depth), and the differences dip, a characteristic shown for the lines marked between the direction of the wandkarren and the D-13/2000 and D-5/2000. 61.98% of the wandkar- direction of the inclination of their bearing slopes ren of the line marked D-13/2000 differs less than (direction differences). This data is presented in 20° from the slope direction of the bearing slope, tables 1 and 2 grouped into intervals. The slope while this value for the line marked D-5/2000 is angles of the bearing surface along the lines vary 100%. it can also happen that the difference be- (table 1). We think that wandkarren can also de- tween the direction of the form and that of the velop on gentler inclinations. bearing slope is more than 20°. This is the case for 42.3% of the wandkarren of the line marked D-19/2000 and for 47.36% of the wandkarren of Size and morphology of wandkarren the line marked D-16/2000. The specific dissolution is independent of the The width of wandkarren is between 2.5 and 34 altitude or the dip of the bearing surface (table 1), centimetres in Dachstein. Their most common but connections appear between width, shape, den- Table 1: Characteristics of selected wandkarren. D. Dachstein; Ch. Chile; ö.sz. total of measured widths along a line; f.sz. specific width; s. density; last figure of line name indicates year measurements taken. Angle slope Characteristics of wandkarren Name of a line Altitude [m] Length of a line [m] of bearing surface ö.sz. [cm] number [pc] f.sz. [cm/m] s. [pc/m] D-4/1999 1700 10.2 51 401 32 39.31 3.14 D-5/2000 1990 5.5 55 450.5 80 81.91 14.55 D-13/2000 2180 4.5 75 226 21 50.22 4.67 D-14/2000 2157 7 48 163 18 23.29 2.57 D-16/2000 2115 17.5 90 339 26 19.37 1.49 D-19/2000 2106 9 75 279 17 31.00 1.89 D-20/2000 2078 12.5 73 371.3 28 29.70 2.24 average - - - 408 31.71 39.26 4.36 Ch-2/2002 500* 20 90 804 6 40.2 0.3 * estimated 238 KRF•1 • OK.indd 238 15.12.2009 10:45:16 Márton Veress, Wandkarren Figure 2: Shapes of wandkarren (cross-section). 1. grike-type wandkarren: 1/1 slanting, 1/2 straight, 1/3 wedging-out, 1/4 flat-bottomed; 2. half-pipe wandkarren: 2/1 planar surfaces between wandkarren, 2/2 ridges between wand- karren, 2/3 scallops in wandkarren, 2/4 rounded surfaces between wandkarren; 3. cavernous wandkarren: 3/1 wid- ening with flat bottom, 3/2 widening with curved bottom; 4. complex wandkarren: 4/1 with one internal channel, 4/2 with internal ridge; 5. large wandkarren on marble (Diego de Almagro island): 5/1 with smaller wandkarren, 5/2 with scallops; 6. limestone, marble; 7. debris. 239 KRF•1 • OK.indd 239 15.12.2009 10:45:17 Karst Rock Features • Karren Sculpturing 9 9 10 10 9 9 10 10 a b 9 9 10 10 9 9 10 10 a b 9 ‘Schichtfugen 0 20 cm karren cave’ cave 10 1 2 3 4 9 5 6 10 7 8 Figure 3: Genetic types of wandkarren. I.a. half-pipe wandkarren; I.b. grike-like wandkarren; I .a. wandkarren devel- oping under water sheet originating from surface; I .b. wandkarren developing under rivulets originating from surface; I .c. wandkarren developing under water flowing from hollows in the bed head; 1. limestone; 2. bedding plane in profile; 3. soil; 4. wandkarren; 5. water sheet; 6. rivulet; 7. water current; 8. inclination; 9. bedding plane; 10. head of the bed. 240 KRF•1 • OK.indd 240 15.12.2009 10:45:17 Márton Veress, Wandkarren sity, and specific dissolution. For lines where the • complex wandkarren (Figures 1.4, 2.4) can be value of shapes of the wandkarren is high, the den- separated into at least two parts. For example, sity and specific dissolution are large as well. such the form can be the coalescing of two smaller wandkarren occur on the lines marked D-5/2000 wandkarren. More frequently, however, small and D-13/2000. The wandkarren of these lines wandkarren divide the side walls or the bottom have a half-pipe shape. on lines where the value of a larger wandkarren (Figure 2.5/1). of shapes of the wandkarren is small, the density Wandkarren can extend over the complete and the specific dissolution are also small. such length or only on a part of the bearing slope. in wandkarren occur on the lines marked D-16/2000, the first case, wandkarren can have a continuous D-19/2000, D-20/2000, and D-4/1999. along these development or the form can be interrupted by lines, grike-like wandkarren developed. schichtfugenkarren or by variously shaped hollows The width of wandkarren varies. according to (Figures 7, 8). schichtfugenkarren are grooves the data, the width of half-pipe wandkarren can with a depth of one or two decimeters that deepen spread less, while the width of grike-like forms can into the rock for several meters along bedding spread to a greater degree. We hypothesize that planes. The width of the wandkarren can be un- where wandkarren occur in great density, the ad- changed or smaller below the schichtfugenkarren, jacent forms do not let each other widen. since the or the wandkarren can also be transformed into grike-like forms are far from each other, they can several smaller forms below a schichtfugenkar- widen freely and can therefore have great width. ren. The upper end of the wandkarren can be at or below the top of the bearing slope. The lower end of the wandkarren can wedge out or end at a Types of wandkarren schichtfugenkarren (Figure 9). When they begin at a schichtfugenkarren (Figure 10), they can according to cross-section, wandkarren can be extend to the lower edge of the bearing slope or the following (Figure 2): to another schichtfugenkarren , or they can also • grike-like wandkarren (Figures 2.1, 3.i.b, 4) wedge out. Wandkarren that end at the lower part have sharp edges and flat side walls. We can of the slope, can wedge out or join a soil-covered distinguish different types as well. Their side surface or another karren form. walls can take different “V” shapes. if the side slopes do not cross each other, the forms end in a flat surface and the end of the wandkar- Development of wandkarren ren can have debris or not. The side wall can be vertical to or slanting relative to the bear- We distinguish two development types of wand- ing slope. soil can fill the ends of nearly vertical karren according to their characteristics: side-walled forms (Figure 5); • half-pipe or ril -type wandkarren (Figure 3.i.a) • half-pipe wandkarren (Figures 2.2, 3.i.a, 6) have develop under sheet water that flows down the curved side walls. This type also varies. The bearing slope. Their cross-section, the value half-pipe cross-section can be similar to a sem- characterizing their shape, and their great den- icircle or to an ellipse. The bearing slope be- sity prove this fact. sheet water can originate tween these forms is rarely wide (it can be flat or from the bedding plane above the slope (Fig- rounded), and its shape is more often ridge-like. ure 3.ii.a) or from a schichtfugenkarren (Fig- Scal ops can occur on the walls of these forms; ure 3.ii.c). When the bedding plane is covered • cavernous wandkarren can widen at the bottom with soil, there is less probability they will de- of the form (Figure 2.3), and the bottoms can velop since soil can retain rainwater. Melt water be flat or curved; 241 KRF•1 • OK.indd 241 15.12.2009 10:45:17 Karst Rock Features • Karren Sculpturing Figure 4: Grike-like wandkarren (Totes Ge- birge). Width of view is 4 m. Figure 5: Wandkar- ren formed below soil from a distance (Asiago plateau, Italy). Width of view is 2.5 m. also plays an important role in the development merous source points), or soil patches. Wand- of this type of wandkarren; karren can help the development of a soil patch. • grike-like or rinnen-type wandkarren (Fig- rivulets can also flow from a hollow (Figure ure 3.i.b) develop under rivulets (Figure 3.ii.b), 3.ii.c). Because the distance between rivulets is which their small density proves. rivulets are random, their discharge, the period of their ex- fed from bare bedding planes, soil (with nu- istence, the date of their development, and the 242 KRF•1 • OK.indd 242 15.12.2009 10:45:18 Márton Veress, Wandkarren co saturation level of their water can differ to 2 a great extent. neighbouring forms can there- fore have different sizes. Forms that are close to each other can have different ages and activity levels. We believe, cavernous wandkarren can be included in this type since we believe the cause of their development is an increasing dis- charge or an increasing dissolution effect. The directions wandkarren take illustrate that they can be created by two different kinds of water flow. The direction of those forms that develop under sheet water differs less from the dip line of the bearing slope. on one hand, the flow of water can not be diverted by the irregu- larity of the surface, and on the other, the forms do not have space to develop in different direc- tions. Those forms that develop under rivulets have a different dip direction relative to the bearing slope since the direction of the flow is changed by the irregularity of the slope and the development of wandkarren occurs in various directions; • complex wandkarren develop if the rivulet can- not fill the form. if there is only one rivulet, we Figure 6: Half-pipe wandkarren (Totes Gebirge). Width of can see only one internal wandkarren while if view is 25 cm. there are several rivulets we can observe several internal wandkarren; • covered wandkarren develop if soil fills the several factors contribute to the antiregres- forms. This type of development occurs where sional development of wandkarren, including the the quantity of plants and soil is significant (for following: example, the southern slopes of the alps or on • the water flow is faster in already-developed the lower part of a mountain). Because of the wandkarren and therefore saturation of the soil accumulation, water cannot flow in the water occurs over a longer path; forms and dissolution occurs under the soil. • because of plants, the co saturation value of 2 The development of wandkarren is antiregres- the water increases. sional, which the following points demonstrate: The number of rillenkarren and channels de- • if they do not begin at schichtfugenkarren , creases as the angle of the slope increases, but as their upper end is always at the beginning of the angle becomes greater wandkarren develop. the upper margin of a slope; The development of a turbulent flow, which causes • their lower ends can wedge out at different alti- an increase of rock dissolution, depends on the tudes; thickness of the water current and its velocity • their width usually decreases in the direction of (emmett, 1970; trudgill, 1985). The thickness of their lower ends; the water becomes smaller because the velocity of • there are no rillenkarren and rinnenkarren on the water increases as the angle of the slope in- slopes with steep inclinations (70°–90°). creases and therefore a turbulent flow can not de- 243 KRF•1 • OK.indd 243 15.12.2009 10:45:18 Karst Rock Features • Karren Sculpturing Table 2: Distribution of wandkarren along selected lines from Dachstein relative to width. Greatest frequency Greatest frequency of Most common interval Most common width Name of a line Width interval [cm] Shape interval of shapes relative to shapes in interval width [%] interval [pieces] [%] D–4/1999 3–50 4–6; 8–10 20; 20 0.14–4 0–2 80 D–5/2000 2.5–11 4–6 43.75 0.75–18 2–4 37.5 D–13/2000 6–16 8–10 33 1.22–5.5 2–4 71.42 D–14/2000 4–16 8–10 29.4 1–10 4–6 35.3 D–16/2000 2–27 8–10 24 0.25–28 0–2 48 D–19/2000 2–61 10–12 31.5 0.19–12 0–2 55.55 D–20/2000 4.3–34 10–12 29.6 0.35–13 0–2 46.42 Notes: - intervals are in 2 centimetres; - width interval: the width of the smal est and biggest wandkarren that occur along a line; - most common width: the width of the wandkarren grouped in an interval (along a line); - shape: ratio of the width and depth of wandkarren; - shape interval: ratio of the smal est and greatest width and depth along a line; - the greatest frequency of shapes: the most common shape along a line from the shapes grouped in an interval; - last figure of the line name indicates year measurements taken; - line D-4/1999 is on side of a polje (area 2); - line D-5/2000 is near Däummel lake on the edge of a doline (area 3); - line D-13/2000 is on the side of a boulder which is on a moraine; - the sites of lines D-14/2000, D-16/2000, D-19/2000, and D-20/2000 are on bed-heads of cuestas (area 1). Figure 7: Development types of wandkarren where the bearing surface is crossed by schichtfugenkarren. 1. depth of wandkarren (cm); 2. schichtfugenkarren that developed along bedding plane; 3. inclination and angle of bear- ing surface. 244 KRF•1 • OK.indd 244 15.12.2009 10:45:19 Márton Veress, Wandkarren Figure 8: Wandkarren that cross schichtfugenkarren Figure 9: Wandkarren that end at schichtfugenkarren (Dachstein). 1. schichtfugenkarren. (Dachstein). 1. schichtfugenkarren. Figure 10: Wandkar- ren that start from schichtfugenkar- ren (Totes Gebirge). 1. schichtfugenkar- ren; 2. wandkarren; 3. karren forms whose direction differs from inclination of a bear- ing slope. 245 KRF•1 • OK.indd 245 15.12.2009 10:45:20 Karst Rock Features • Karren Sculpturing Figure 11: Genetic types of schichtfugenkarren developed on heads of beds. a. inclinations of the bedding plane and the head of the bed are similar; b. inclinations of the bedding plane and the head of the bed are opposing; c. inclinations of the joint and the head of the bed can be opposite (see top and bottom parts of the Figure) or simi- lar (see middle of the Figure); 1. limestone; 2. bedding plane; 3. joint; 4. schichtfugenkarren; 5. infiltration of unsatu- rated water; 6. snow; 7. head of the bed; 8. surface of the bedding plane. Figure 12: Development of wandkarren with different lengths. 1. limestone; 2. bedding plane (in cross-section); 3. joint; 4. schichtfugenkarren; 5. unsaturated water; 6. saturated water; 7. site of complete saturation of water flowing on the head of the bed; 8. dissolution increases because water mixes on the head of the bed with unsaturated water flowing from schichtfugenkarren; 9. head of the bed; 10. bedding plane; a. unsaturated water can dissolve rock anywhere (width of wandkarren is similar above and below schichtfugenkarren); b. water flowing down the head of the bed becomes saturated at the level of the highest schichtfugenkarren, but water flowing from schichtfugenkarren is unsaturated (width of wandkarren is smal er below schictfugenkarren); c. water from the head of the bed above schichtfugenkarren can no longer dissolve rock but mixes with unsaturated water flowing from schichtfugenkarren (wandkarren crosses several schichtfugenkarren but its width does not change); d. water is saturated (wandkarren can only develop under schichtfugenkarren if unsaturated water flows from schichtfugenkarren); e. water is unsaturated but flows into schichtfugenkarren (wandkarren do not develop between schichtfugenkarren).  246 KRF•1 • OK.indd 246 15.12.2009 10:45:20 Márton Veress, Wandkarren 247 KRF•1 • OK.indd 247 15.12.2009 10:45:20 Karst Rock Features • Karren Sculpturing velop. Thus the intensity of dissolution on surfaces will have started at the wandkarren, so its inner with a small dip (5°–50°) decreases as the slope height will be the largest there. angle increases. Wandkarren with different lengths and widths a turbulent flow can develop although the develop as follows (Figure 12): thickness of the sheet water is small if the angle of • a wandkarren cuts a schichtfugenkarren with- the bearing slope is large (e.g. greater than 70°). if out changing its width because only saturated the velocity of the flow is high on slopes with large water leaves the schichtfugenkarren (Figure inclinations, a turbulent flow develops regardless 12a); of the small thickness of the water. Therefore, if • we can observe that the width of a wandkar- the inclination of the slope is great, the intensity ren decreases after it cuts a schichtfugenkarren. of dissolution increases and wandkarren develop. in this case, the water flowing down the wall grike-like wandkarren generally cut across becomes more or less saturated at the height of schichtfugenkarren or develop from the site of the schichtfugenkarren, but the water leaving the schichtfugenkarren (depending on the time of the schichtfugenkarren is unsaturated because origin of the forms). according to Weber (1967), there is a wandkarren below it (Figure 12b). schichtfugenkarren develop due to dissolution oc- The smaller width of the wandkarren below curring along bedding planes. schichtfugenkar- the schichtfugenkarren indicates that while the ren can develop in two ways (Figure 11). They water that leaves the schichtfugenkarren is un- can develop when water percolating through the saturated because the width of the wandkarren rock creates caves along the bedding planes in the is small, the water can only dissolve the wand- direction of the face of the bed heads. schichtfu- karren to a smaller extent; genkarren can also develop as water flows down • wandkarren cut some schichtfugenkarren the walls and infiltrates the rock along bedding without a decrease or even an increase in their planes to dissolve cavities. However, this latter width. in these cases, the solubility of the water case happens rarely since usually only a small flowing down decreases minimally since the quantity of the water flowing on the walls of a bed water leaving the schichtfugenkarren is un- head can infiltrate the rock. saturated and is able to maintain solubility as We can formulate a genetic model of wandkar- it mixes with the water flowing down the wall ren crossing schichtfugenkarren in the case where (Figure 12c); the schichtfugenkarren is older than the wand- • wandkarren begin at schichtfugenkarren be- karren. such cases have the following character- cause the water flowing down the bed heads is istics: already saturated. Wandkarren can only devel- • the width of a wandkarren abruptly be- op below schichtfugenkarren when unsaturat- comes smaller when the wandkarren crosses ed water leaves the schichtfugenkarren (Figure a schicht fugenkarren. This can only happen if 12d); the schichtfugenkarren developed earlier than • wandkarren do not develop between two the section of wandkarren below it; schichtfugenkarren when the water flowing • the inner height of schichtfugenkarren does not down the bed head flows into the upper schicht- change. The inner height of schichtfugenkarren fugenkarren or becomes saturated at this alti- can decrease (away from a wandkarren) if the tude. at the same time, unsaturated water flows schichtfugenkarren is younger than the wand- out of the lower schichtfugenkarren (Figure karren. in this case, the schichtfugenkarren 12e). 248 KRF•1 • OK.indd 248 15.12.2009 10:45:20 coastal Karren 20 Joyce LUNDBERG coastal karren are the small-scale dissolutional ess that is not confined to coastal features. since and/or bio-erosional features that develop on rock coastal karren are impacted by a great variety of surfaces in the narrow zone of exposed rock be- processes (including abrasion, hydraulic action, side a body of water, and within the direct range wave action, wetting and drying, corrosion, disso- of the action of that water. The erosional processes lution, bio-erosion, bio-construction and, in cold are directly related to the water surface and op- regions, frost action), it is preferable to avoid ge- erate in (a) the inundated zone immediately be- netic terms. Many of the features described below side the shore, (b) the zone of alternating water also occur on non-soluble rocks but are usually levels, (c) the zone of splash, and (d) the zone of less clearly developed. spray. The processes are most commonly observed coastal karren do not include features related in association with salt water, but they also occur to groundwater action in the coastal environ- in brackish water and fresh water environments. ment, usually of salt-fresh mixing (e.g. smart et unlike most terrestrial settings, tectonic joint- al., 1988), often described as coastal karst. in ad- ing is rarely the focus of activity. The formation is dition, coastal karren do not strictly include fea- controlled by very small scale, very local hydrol- tures such as alveoli, tafoni honeycombing, and ogy and chemistry of rock surfaces; large-scale certain shore platforms that are produced by salt environmental settings are less significant than weathering (e.g. Matsukura and Matsuoka, 1991), proximity to organisms, to water, and to vegeta- although these may be associated with dissolu- tion or sediment cover. The dominant controls tional features. are wetting regime, biological activity, and energy levels. in many settings, bio-erosion dominates. if abrasion overrides, or if cliff retreat rate exceeds Morphology that of karren formation, then karren are absent. a variety of other terms has been used to de- one of the most striking features of coastal kar- scribe coastal karren, such as intertidal karren, lit- ren is that they can be very delicately etched into toral karren, biokarst (Viles, 1984; schneider and surprisingly jagged surfaces. They range in scale torunski, 1983), phytokarren, phytokarst (Folk et from sub-millimetre to several metres. The forms al., 1973; Jones, 1989). some of these obviously in- are usually distributed in zones relative to wetting clude only coastal forms; others, such as biokarst regimes and biological zonation. Zonation is most or phyto karst/karren indicate a general proc- obvious in meso- to macro-tidal salt-water coasts. 249 KRF•1 • OK.indd 249 15.12.2009 10:45:20 Karst Rock Features • Karren Sculpturing The variety of forms reported in the literature tomed; Figure 1a, b, d), simple/complex pits (W/D appears to be legion. However, variations can be ≤ 1), and pinnacle and basin spitzkarren (trudgill, explained if coastal karren are viewed as a series 1979) (Figure 1c, e). all erosional forms show var- of modules made up of basic building blocks: the ying degrees of isolation or interconnectedness. modules are then assembled in different propor- simple pans/basins and simple pits are normally tions in different geographic settings. This will isolated. pinnacles and basins are usually well vary with: (a) organisms present (which may cause, inter-connected so that water flows freely between or protect from, erosion); (b) exposure levels basins. in warm, salt waters of high energy level, (wave energy, wind energy, wave and wind orien- bio-constructors such as coralline algae or ser- tation); (c) tidal range and regime; (d) the balance pulid worms may protect surfaces from erosion of fresh and salt water; (e) lithological variations and build up small ramparts of encrustations; (chemistry, crystallography, depositional fabric); pans and basins may also become dammed or (f) structural variations (sedimentary structure, rimmed (Figure 1f, g). bed thickness, orientation, dip, joint frequency The blocks are assembled into modules. These and orientation). include: the erosional ramp, the erosional notch, The basic units may develop at different scales and the shore platform. some places have all three (for example, a rounded basin may range from a of these: e.g. a reasonably exposed tropical sea few millimetres to a metre in width). They may coast will have a notch, a shore platform, and a also display a pseudo-fractal nature (torunski, short erosional ramp. some places have only one, 1979), of forms within forms (so a 30 cm wide e.g. a long ramp is characteristic of exposed re- basin may enclose a surface made up of 3 cm wide gions anywhere. The module can further be de- basins, which may in turn enclose a surface of 0.3 scribed by its position in relation to the shore, e.g. cm wide basins). a notch may be supra-tidal, inter-tidal, or sub-tid- al. in view of the complexity of genesis, and the incidences of convergent evolution, descriptors Building blocks and modules using genetic terms such as “surf”, “bio-erosion- al”, etc., should be avoided unless the genesis is The basic building blocks include negative, rem- evident. These modules are not unique to karst nant and (occasionally) positive forms. The nega- rocks, although the dominance of dissolution, the tive forms are produced by varying proportions detailed forms of the units, and the particular as- of dissolution through fresh-salt water mixing, semblage of modules often is. dissolution through biogenically-induced aggres- The erosional ramp, which may stretch from siveness, mechanical quarrying by grazing organ- wave base to the supra-littoral spray zone, is the isms, abrasion, hydraulic action of water, etc. The most common form worldwide, ranging from remnant forms, often the most manifest, are sim- tropical areas, to temperate areas, to cold regions ply the places where erosion has not yet operated; (Figure 2a, b, c), and usually associated with ex- in places this occurs where colonies of organisms posed environments. it is an obvious consequence provide protection. positive forms are produced of the diminution of erosion with distance from where encrustations build forms up. a common wave and splash energy. in many tropical regions characteristic of karst shorelines is the dynamic the deposits from the last interglacial high sea interaction of both constructional and destruc- level that originally formed a raised platform are tional processes in the one location (rust and Ker- now being modified into a ramp (e.g. Moses, 2003). shaw, 2000). The erosional ramp comprises zones of pits, The negative and remnant basic units include: pans, basins, and pinnacles, varying with water pans/basins (W/D > 1, flat-bottomed/round-bot- regime, biological action, and exposure. typically 250 KRF•1 • OK.indd 250 15.12.2009 10:45:20 Joyce Lundberg, Coastal karren the zones furthest from wave action have small sects a sloped cliff or an erosional ramp, then the pits and flat-bottomed basins (but see discus- top of the notch takes on the form of a visor, which sion of supra-littoral basins below). Biological ac- in the splash zone bears pinnacles and basins. tion is more pronounced closer to the water. The Thus a commonly-described duo is the notch and negative forms become bigger closer to the water visor (Figure 2g). Where a stack stands separate surface and the pinnacles more pronounced. Ba- from the coast, notching will occur on all sides sins usually become more rounded, perhaps with producing a mushroom rock (Figure 2e). overhanging rims, towards the water. The form and elevation of the notch relates to Jaggedness increases with increasing exposure exposure level, and to tidal range. in sheltered (Figure 1e, f). With strong splash and high en- and micro-tidal environments, processes (mainly ergy levels the projections dominate and basins biological activity) are focused on a small verti- are poorly developed. Runnels may appear where cal expanse of cliff: the notch digs deeply into the gravitational draining dominates (Figure 1d). Me- rock at mid tide level but is not very tall (Figure chanical fracture and dissolution are active. Few 2d). as tidal range or exposure increase, processes macroscopic organisms are apparent and direct are spread out over a larger expanse of cliff face: bio-erosion is minor. Endoliths bore into surfaces the notch will be less incised but taller. Where bio- but epiliths are rare. on the reef rock of Bonaire, genic activity is dominant then the surface will be netherlands antilles, the pinnacles are oriented inhabited by boring and grazing organisms and towards the incoming splash. may be made up of many sharp-edged pits. Where The forms vary with lithology: e.g. in puerto abrasion and hydraulic action are dominant (only rico reef rock is much more jagged and irregu- in high energy environments) the notch will be lar than beach rock, and more deeply etched but less deeply incised with smooth, rounded surfaces not as sharply fretted as eolianites. Mylroie and devoid of organisms. The sub-tidal environment Mylroie (see chapter 39) describe karren from san below the breaking waves is not subject to much salvador island, Bahamas, that is very strongly mechanical or dissolutional action, so this part of influenced by lithology, developed on eogenetic the notch is probably purely bio-erosional. clearly, rocks that show little post-depositional diagenesis. notches may result from a variety of processes: bi- The erosional notch, a horizontal groove paral- ological (e.g. abensperg-traun et al., 1990; Hodg- lel to sea level cut into bedrock but not structural- kin, 1970), chemical (e.g. Higgins, 1980; rust and ly controlled, is the next most common form. it is Kershaw, 2000), and physical (trudgill, 1976a; very well developed in tropical regions where bio- Kershaw and guo, 2001). convergent evolution eroders abound, especially in sheltered regions. may produce similar forms on non-karst rocks: in temperate regions, with fewer bio-eroders, the for example, erosional notches just below mid- notch is not well developed. in addition, most of tide level develop very rapidly in conglomerates of the temperate regions displaying coastal karren new Brunswick, canada (trenhaile et al., 1998). are in exposed settings, so that the erosion ramp notch elevation is not always simply and clearly dominates over the intertidal notch. a third factor related to sea level: in more turbulent conditions is that tropical coastal karren are almost invari- the notch may be as much as 2 m above sea level ably developed in young carbonates of relatively (rust and Kershaw, 2000). low diagenetic maturity, whereas temperate coast- lithology has some impact on notch form. al karren are in substantially older rocks. notches in reef rock are usually more complexly The intertidal zone is the focus of intense bio- fretted than those on eolianites. it is of interest genic activity, abrasion, hydraulic action, and per- that Holocene eolianites of san salvador island haps dissolutional activity. This produces a notch, (see chapter 39) do not show much notch develop- incised typically from 1 to 5 m. if the notch inter- ment, while adjacent pleistocene eolianites, pre- 251 KRF•1 • OK.indd 251 15.12.2009 10:45:20 Karst Rock Features • Karren Sculpturing a b e f c d g Figure 1: The basic building blocks: a. basins of the upper littoral of a sheltered bay, north-east Vancouver island. The rock is very hard, fine-grained limestone. The pools are isolated, relatively flat-bottomed, and smooth bottomed. The microscopic green algae lining the pools have been grazed in patches. The inter-basin rock is roughly fretted; b. smal basins in the supra-tidal swash/splash zone of Port-au-Port Peninsula, Newfoundland (photo by C. Malis, used with permission), width of view is 1.5 m; c. pinnacles and basins close to low water level on the west coast of Ireland, Burren District. The pinnacles are protected by barnacle colonies (white). Mussels (black) occupy the slight- ly more protected pinnacle wal s. The pinnacle wal s and the basin floors are themselves made up of smaller ba- sins, many created by, and occupied by, the spiny sea urchin Paracentrotus lividus; d. basins modified by slope in the mid-tide level of Gower Peninsula, south Wales. The basins are cut into a dip slope and al tend towards an elon- gate form; e. pinnacles modified by high energy splash from Burren District, west coast of Ireland. These pinnacles show a form very similar to the sharp pinnacles of Puerto Rico (right); f. rimmed pools developed in eolianites and exposed at very low tide, Punta Maracayo, Playa de Sardinera, Puerto Rico. In the foreground are the basins and pinnacles of the erosional ramp. Below this the rimmed pools, making up the shore platform, are normal y covered with surf; g. constructional rampart exposed at low tide, Boca Kokolishi, northern coast of Bonaire, Netherlands An- tilles. Although this is in a smal bay it directly faces the dominant wave direction and thus receives high energy. The intertidal shore platform and associated protective rampart are clearly developed. The erosional, rather than constructional nature of the platform is apparent from the remnant isolated sea stacks. 252 KRF•1 • OK.indd 252 15.12.2009 10:45:26 Joyce Lundberg, Coastal karren sumably more diagenetically altered, do. Bird et al. rock surface at the shore line. it may develop at dif- (1979) report notches incised 30 m deep into the ferent elevations in relation to dominant processes very easily eroded calcareous marls of Barbados. such as waves, tides, surf (e.g. sub-tide, mid-tide, Dip or slope also relates to notch form: torunski high-tide). in general, a shore platform develops (1979) observed that steeper slopes result in deep- wherever differential lateral planation causes cliff er notches. retreat above a datum but not below it. The most The shore platform (Figure 2c) is any horizontal common shore platforms are simply the wave-cut 253 KRF•1 • OK.indd 253 15.12.2009 10:45:30 Karst Rock Features • Karren Sculpturing platforms that develop at wave base or at the base of steps separating wide, shallow, flat-bottomed of subaerial weathering – top of permanent satu- pans, called rimmed pools or vasques (guilcher, ration (trenhaile, 2002). However, in limestones 1953). each riser is a narrow, sinuous, lobed ridge, a particular form of shore platform develops in protected by encrustations. The encrusters thrive the mid-tide to surf zone where bio-constructors on the rim edges because water continually flows flourish (in warm water regions such as mediterra- over the edge as it drains to lower levels. The form nean and tropical areas). Various terms have been is very similar to that of tufa dams or rimstone used in the literature such as tidal platform, plate- pools and gours. This is the classic “plate-forme à forme à vasques, solution bench, or surf platform. vasques” (Figure 1f) of guilcher (1953). in warm waters with high energy levels, en- crusting organisms abound (e.g. calcareous algae, vermitid gastropods, serpulid worms). These coat Transition between zones the rock surface with carbonate, protecting it from erosion. The crust may be lithified by the pumping The nature of the transitions varies with tropical action of seawater through the accretions. Many and temperate karren. some of the tropical kar- organisms take advantage of the protection offered ren show a distinct division between zones, most and hide in crevices between encrustations. The obviously in the sheltered regions. The edge of the encrusters are not very tolerant of emersion and splash pinnacles and basin zone is clearly demar- prefer a continually renewed water supply of high cated at the edge of the visor giving way to the energy level. Thus they concentrate below mean notch. similarly the upper and lower edges of the tide level or in the surf zone of very exposed coasts. intertidal constructional platform are relative- The rock above (and below to a lesser extent), with- ly clear. in more exposed tropical regions and in out this protection, is open to normal notching temperate regions, most zones do not have a de- processes. Thus, over time, the notched part re- finitive edge: all show some degree of gradation treats while the encrusted zone remains relatively between forms. as denudation continues each intact. The remnant form is the shore platform. zone moves landward (even those zones semi-pro- The outer edge may have an accumulation built up tected by encrustations). Thus denudation in any into a rampart (the armoured rim, or trottoir) (see one spot lowers that area into the influence of the Figure 1g). This can create a landward moat, where water regime of the next zone. so, each upper zone the rampart bears the brunt of the waves and the gets transformed into the zone below it. water flows downhill towards the land. sometimes shore platforms are built entirely of bio-constructors. in less exposed localities (e.g. Regional variations mediterranean regions) a small and rather deli- cate bench is built out from a steep rock surface, or salt water coastal karren from the end of the shore platform, by encrusta- tion alone: this is then called a corniche (trenhaile, salt-water coastal karren display some or all of the 1987). naylor and Viles’ (2002) site in crete con- modules described above, depending on condi- sists of bio-constructional boiler reefs (the term tions. Four general situations are discussed below: boiler referring to the action of the water) on a tropical, temperate, mediterranean, and cold re- gently sloping surface. shore platforms with clear gions (Figure 3). constructional morphology are reported from the Mariana islands (Mylroie, pers. comm. 2004). Tropical regions The shore platform usually has a gently slop- tropical coastal karren are widespread because ing surface form. The slope is made up of a series carbonate coasts are so common in the tropics. 254 KRF•1 • OK.indd 254 15.12.2009 10:45:30 Joyce Lundberg, Coastal karren These rocks are typically diagenetically immature, a ramped platform with rimmed pools, the pools rarely of greater age than the late pleistocene. The becoming smaller and the rims becoming higher karren are more complex than temperate and cold and more irregular upslope. The rims grade into region forms, and have more often been studied. projections and the shallow pools grade into ba- They have been reported from a variety of locations sins. The upper part of the ramp is then of well- including Hawaii (Wentworth, 1939; guilcher, developed splash zone basins and pinnacles. 1953), Florida Keys (ginsburg, 1953), Bikini (re- a dynamic equilibrium develops as the cliff velle and emery, 1957), puerto rico (Kaye, 1959), edge, notches and platform retreat together. it is of Bahamas (newell, 1961), guam (emery, 1962), interest that this retreat sometimes exposes flank Bermuda (neumann, 1966), Barbados (tricart, margin caves that have developed in the landmass. 1972), aldabra atoll (trudgill, 1976b), the nether- These may then be mistaken for notches. lands antilles (Focke, 1977, 1978a, b), Kenya (Bird Variations can be produced through diverse and guilcher, 1982), Madagascar (Battistini, 1981), controls. Most of these significantly modify the india (Bedi and rao, 1984), and the cayman is- forms only in moderately exposed situations: very lands (Woodruffe et al., 1983; spencer, 1985a, b). sheltered locations will still show the notch, and The obvious zones include: splash/spray zone very exposed locations will still show the erosion- basins and pinnacles on the erosional ramp, visor al ramp. tidal range is important in sheltered and and notch, inter-tidal to surf zone shore platform, moderately exposed locations; at higher exposure and low tidal or subtidal notch. not all coasts will levels wave height is more important. The height have all of these units and the proportions and of the cliff is important (Hodgkin, 1970): for high forms will vary. The model shown in Figure 3 is cliffs, splash energy impinges only on the vertical based on that by Focke (1977, 1978a, b) taking into cliff face producing fretting and fluting on the face, account observations by neumann (1966), safriel and the splash zone ramp does not develop even (1966), Battistini and guilcher (1982), Dalon- in exposed conditions. in contrast, a low cliff will geville and guilcher (1982), and modified accord- develop a splash zone ramp even in moderate con- ing to field observation. ditions because the thin visor will easily collapse. in this model, form is clearly related to expo- Very thin beds may be affected by only one part of sure level. The very sheltered areas have only the the shore processes: e.g. beachrock often emerges intertidal notch grading into the subtidal notch from sandy beaches and is abraded, dissolved and without any obvious break in form. spray zone bio-eroded into shallow grooves and basins which fretting is usually minor. The more exposed plac- become more dissected with longer exposure time es have a taller, less incised notch and splash will (revelle and emery, 1957; Hopley and Mackay, extend beyond the notch to cause splash zone pit- 1978). sediment accumulation may prevent the ting. as conditions become less suitable for borers development of notches even in sheltered areas and more suitable for encrusters, part of the notch and promote very smooth notch development in becomes protected so that a “double notch” form more exposed areas (Hodgkin, 1970; tjia, 1985). develops: i.e. the intertidal notch forms above, the Fossil notches from former sea levels may compli- tidal platform interrupts it, and the low- or subti- cate the shore profile (e.g. red sea; guilcher, 1952). dal notch forms below (Figure 2f). usually such lithology may be significant. calcareous eoli- areas are of high enough energy levels to have a anites usually are of relatively low strength. This well-developed splash zone. in the more exposed contributes to the development of large basins (e.g. areas the platform can be at higher elevation. the vast rimmed pools in Madagascar up to 500 m Where energy levels are very high then coastal long; Battistini, 1981), small but sharp spray zone retreat is fast and notches are poorly developed. pinnacles as projections break off easily, and wide instead the low tidal terrace gives way upslope to shore platforms as visors collapse easily. The me- 255 KRF•1 • OK.indd 255 15.12.2009 10:45:30 Karst Rock Features • Karren Sculpturing a b e f c d g Figure 2: The modules: a. erosional ramp in Paleozoic limestones, Gower Peninsula, south Wales, showing zona- tion in relation to water levels. The basins are cut into a dip slope and al tend towards an elongate form; b. looking from the supra-littoral zone at the erosional ramp developed in marbles, west coast of northern Norway at Gås- bakken. Supra-littoral basins are clear in the foreground; closer to high tide mark the surface is not pitted; the inter- tidal algal mat providing bio-protection is visible in the background; c. erosional ramp in Pleistocene reef rock, east coast of Bonaire, Netherlands Antilles. The ramp is made up of sharply fretted spitzkarren and basins. It gives way seaward to the shore platform exposed between waves at low tide. Under the platform is a sub-tidal notch (not visible in the photograph); d. a deeply incised intertidal notch developed in Pleistocene limestones at Boca Slag- baai, north western coast of Bonaire, Netherlands Antilles. This notch developed in a very sheltered environment that is now cut off from the sea by a beach barrier. The absence of a pinnacled visor and the slightly pitted nature of the cliff above is further indication of the sheltered conditions; e. notching on al sides of a sea stack in Miocene carbonates from Isla de Mona, Puerto Rico, creates a mushroom rock, which wil eventual y fal over; f. a fossil inter- tidal notch from last interglacial high sea level, in Pleistocene reef deposits, north eastern coast, Bonaire, Nether- lands Antilles. This shows a smal double notch separated by a smal intertidal platform, characteristic of a mode- rately exposed environment. Height of the right edge is 4 m; g. a smal visor and intertidal notch cut into Pleis- tocene reef rock on the south coast of Dominican Republic. The cliff is high enough to avoid splash on its top, so the visor displays only a smal erosional ramp of pinnacles and basins. The host rock cannot support a wide visor and col apses easily. The details of form reflect the highly heterogenous nature of the reef rock, which has been minimal y diagenetical y altered. Height of the right edge is 2 m. 256 KRF•1 • OK.indd 256 15.12.2009 10:45:35 Joyce Lundberg, Coastal karren chanically stronger beach and lagoonal calcaren- ites of lord Howe island, australia (Moses, 2003), have a deeply dissected visor in the splash zone cut into highly fretted and jagged pinnacles and basins up to 1 m wide and deep. rock strength may not always affect form as expected: e.g. soft coral marl in Barbados has not collapsed although an intertidal notch has incised more than 30 m (tricart, 1972). lithological differences may not affect gross morphology but may affect details: e.g. in Malaysia both Quaternary reef rock and crys- talline palaeozoic limestones are deeply notched but the reef rock is sharp and jagged while the crystalline rock is smooth (Hodgkin, 1970). Finally, catastrophic events such as hurricanes or tsunamis may rip up parts of the shore, leaving guilcher (1953, 1958) shows the littoral zone exposed flat bedding planes (e.g. grand cayman; of cool temperate regions such as the British isles Jones and Hunter, 1992). as pitted to varying depths: the highest zone that gets sea spray has small pits; the upper part of the Temperate regions intertidal zone shows flat-bottomed and larger There is relatively little literature on coastal kar- pools; the lower intertidal is more dissected by ren from temperate areas (see review in trenhaile, sharp pinnacles or lapiés between pools. trudgill 1987). essentially temperate region coastal karren (1987) describes the forms as “scoriaceous” cock- are all variations on the erosional ramp (Figure ling and f retting in the spray zone, giving way 2a), with some indications of small-scale sub-tidal downslope to a pinnacled zone in mid-tidal re- notches. it is important to note that most of the gion, with pools taking over seaward. The great- carbonates from temperate regions are diageneti- est relief is usually in the mid-intertidal zone (ley, cally mature and thus the karren are not strictly 1979; lundberg, 1977c), with more subdued relief comparable with tropical karren on diagenetically landward and seaward. The form is related to dis- immature rocks. tribution of bio-erosive organisms, in some cases 257 KRF•1 • OK.indd 257 15.12.2009 10:45:39 Karst Rock Features • Karren Sculpturing the form being directly attributable to particular notches around sicily are generally shallow with organisms (trudgill, 1987). The clearest example very small constructional rims of coralline algae of this is the deepening of permanently wet ba- (rust and Kershaw, 2000). sins (mid-low tide levels) by echinoderms (Figure catastrophic events in the form of earthquakes 1c). notches do occur in cool temperate regions, and sea level change are apparent in some re- e.g. in chalk and limestone in Britain (trenhaile, gions. antecedent conditions thus lead to inherit- 1987), although generally poorly defined. trudgill ance and modification of forms: e.g. Dalongeville (1987) shows many profiles from ireland with (1977) described a fossil platform above the mod- cliffed edges and sometimes slightly overhang- ern platform in lebanon where the basins in the ing edges at low tide level. ley (1977, 1979) in a fossil platform are being modified into crater- study of coastal karren of south Wales, found that shapes by spray and rain. in sicily the complicated the degree of development of pinnacles increased notch profiles reflecting variations in recent uplift with the purity of the limestones. increased poros- rates are preserved only in the more sheltered lo- ity and permeability gave increased complexity of cations (rust and Kershaw, 2000). microrelief features, caused by rock heterogeneity. The coastal karren of eastern Mallorca is de- The Burren District, county clare, ireland, pos- scribed in detail by gómez-pujol and Fornós (see sibly the best developed temperate coastal karren chapter 40). with clear zonation of forms in relation to organ- isms (lundberg, 1974, 1977c, 2004; trudgill, 1987; Cold regions trudgill and crabtree, 1987; trudgill et al., 1987), The literature on coastal karren from cold regions is described in detail by Drew (see chapter 41). is severely restricted. The west coast of newfound- land (Malis, 1997; Malis and Ford, 1995) is a mi- Mediterranean regions cro-tidal region with a late Holocene history of These are essentially intermediate between tropical rising sea level (~0.7 mm) where sea ice remains and temperate forms. The cooler parts show a sim- fast to the shore for 5 months of the year. The lim- ple erosional ramp. The warmer parts show varia- ited karren that develop are micro-pits and simple tions on the notch and shore platform. again, ex- basins, generally < 5 cm deep and < 10 cm wide, posure, although never so impressive as the more many with poorly defined edges (Figure 1b). There open ocean coasts, plays a part. trenhaile (1987) is little karren development in the intertidal zone notes that the low tidal range of the Mediterranean proper; basin size then increases with increasing allows for deep notches and protruding visors. height above mean low water mark to a maximum guilcher (1953, 1958) observed the warmer of ~14 cm wide and ~6 cm deep in the backshore parts of the northern spanish coast showing zone. The most abundant karren develop where evidence of a shore platform sloping steeply sea- wave and splash energy, sub-aerial and sub-aque- ward, ending in a low cliff, but without the classic ous exposure are optimally balanced, in the supra- rimmed pools. near Marseilles, on the Mediter- tidal swash zone (~12 cm wide, ~5 cm deep). oth- ranean sea coast of France, pinnacles and basins erwise no obvious zonation of form is apparent. form in the spray zone, sometimes with a high tide unlike almost every other region of coastal karren notch, and an intertidal constructional corniche development, bio-erosion is reported to play only (guilcher, 1953) while nearby, near nice, there is a very minor role here. geological properties and no corniche. Dalongeville (1977) described from exposure levels are the dominant controls. lebanon a platform a few metres wide with wide, lundberg and lauritzen (2002) tried to develop deep basins (i.e. deeper than rimmed pools but a model of karren development on cold coasts not as jagged as the pinnacled pools of the tem- from studies in northern norway and svalbard. perate regions) just above mean tide level. The The basic form is an erosional ramp and the kar- 258 KRF•1 • OK.indd 258 15.12.2009 10:45:39 Joyce Lundberg, Coastal karren ren are simple basins (Figure 2b). They discovered in lord Howe island, australia. However, where that the greatest development is in the supra-litto- vegetation is absent, the supra-littoral may show ral (see discussion below) because isostatic uplift is surprisingly large isolated basins where water is the dominant process. The forms are also often an pooled and aggressivity can be enhanced (Figure overprint on the legacy of glacial erosional forms, 4a): e.g. Kaye (1959) observed spray zone pitting in with rudimentary pitting where water pools. Thus puerto rico eolianites up to 6 m across. the karren displayed today are not in equilibrium in humid temperate regions such as western with the coastal processes in action. ireland, evaporation is minimal; salt spray and a rudimentary supra-littoral erosional notch rain water mixing is dominant. The supra-littoral resulting from frost action was observed in some has flat-bottomed isolated pans, with bare inter- jointed marbles of northern norway and svalbard vening rock surfaces. Where a fresh ground water (unpublished field observation). source emerges and mixes with seawater, very big, deep pools (filled with the green alga characteris- tic of brackish water, Enteromorpha) develop. Brackish water coastal karren The supra-littoral of the cold but semi-arid re- gions of northern norway shows various bowl- Brackish water is produced either through mixing shaped depressions (Moe and Johannessen, 1980; of salt and fresh water (e.g. in the zone of sea spray Holbye, 1989; lundberg and lauritzen, 2002). and rainfall) or by evaporation of fresh water. This These bowls are developed only in exposed situa- develops in the supra-littoral zone of seacoasts, or tions where basins are typically ~20 cm deep. The in inland sites with high evaporation. The proc- bowls are made up of hierarchies of pits; many esses include salt-fresh water mixing dissolution, show considerable salt weathering with large salt evaporation, salt weathering (Mottershead and crystals in remnant pools (Figure 4b). This results, pye, 1994), and removal of loosened materials by particularly in the more coarsely-grained marbles, wind. The karren develop into varieties of basin in grusification and the breakdown of the bowl and pits, usually lined by salt encrustations, with edges. some spal ation and grusification caused by salt action. The limit between the upper spray zone of intertidal karren and the lowermost region of ter- Fresh water coastal karren restrial karren that gets some salt spray is some- times obscure. in the north east of Bonaire, neth- Fresh water coastal karren are very rarely de- erlands antilles, the coast facing the prevailing scribed in the literature. one example comes from winds grades from the splash zone basins and pin- the silurian dolostones making up the lakeshore nacles characteristic of the tops of the visors up to of lake Huron, Bruce peninsula, ontario, canada a complex and very sharp pit karren oriented to- (Vajoczki and Ford, 2000). Here the form is a sub- wards the prevailing wind. lithology is important dued erosional ramp with simple to complex pits in that aphanitic rocks in all climates show micro- (Figure 4d). The pits reached up to ~2 cm in width rills (rillensteine) in the zone of salt-fresh mixing and ~5 cm in depth. pit depth has a strong posi- (Figure 4c). tive correlation with water depth. Vajoczki and The supra-littoral seacoasts of tropical regions Ford (2000) consider that dissolution, possibly as- often simply grade into dense vegetation with no sociated with biofilms on the rock surface, is like- obvious karren features: e.g. Moses (2003) ob- ly to be the dominant control on their formation. served a gradual diminution of basin size and The lake has had a complex history of level chan- rock micro-relief in the transition zone between ges since glacial retreat; the pits are developed in the supra-tidal splash zone and the vegetation line the zone of changing water levels. 259 KRF•1 • OK.indd 259 15.12.2009 10:45:39 Karst Rock Features • Karren Sculpturing a second example comes from the carbonifer- corrosion in intertidal and splash-spray zones but ous limestones of ireland. again, the pits are as- there is a theoretical problem because seawater is sociated with changing water levels on lakeshores. normally saturated or supersaturated, especially Here tubular pits develop on the underside of in the tropics. boulders and bedding planes and develops up- explanations involving the interaction of fresh wards. simms (2002) called them röhrenkarren and salt water can immediately be contradicted (tube karren). These forms are described in detail by the presence of excellent corrosional forms by Drew (see chapter 41). on small isolated stacks with no fresh water stor- age, and in the red sea where fresh groundwater does not flow (Macfayden, 1930). in a few cases Rates of development the development of the intertidal notch is clearly enhanced by salt-fresh water mixing corrosion: rates of erosion are measured directly with mi- e.g, Waltham and Hamilton-smith (2004) note cro-erosion meters, and/or relative to some dated that sea-level notches in the islands of Ha long event (trenhaile, 1987). relatively sheltered areas Bay, Vietnam, seem to develop only on the larger typically retreat by about 1 mm per year (Hodg- islands with fresh water catchments. kin, 1964; schneider, 1976). rates vary with lo- it has been suggested that biochemical proc- cation: e.g. 0.09‒2.7 mm per year, aldabra atoll esses play an important role by modifying water (trudgill, 1976b; Viles and trudgill, 1984); chemistry. This is probably applicable only where 0.2‒3.8 mm per year, great Barrier reef, aus- seawater is trapped in enclosed bodies. During tralia (trudgill, 1983); at least 1 mm/yr, eastern the night excess respirational production of co 2 Mediterranean (Kelletat, 1991). rates from a sin- over photosynthetic removal increases the ag- gle location vary with exposure level: e.g. open gressivity of pool waters (emery, 1946; schneider, coasts on grand cayman 2.7 mm/yr, reef-pro- 1976; trudgill, 1976a). obviously such an effect is tected coasts 0.45 mm/yr, but where bio-con- less in open waters where pH changes of, for ex- struction dominates over bio-erosion only 0.17 ample, only 0.15 units can be detected (schmalz mm per year (spencer, 1985a, b). surprisingly, and swanson, 1969). indeed, it has been found erosion rates for calcareous sandstones in West- that biological activity may actually inhibit disso- ern australia (abensperg-traun et al., 1990) are lution in that organic coatings may prevent con- comparable with those for pure limestones, rang- tact of rock with water (schneider, 1976). ing from 0.2 mm/yr to 0.8 mm/yr. Many workers have thus concluded that dis- solution, even biochemically mediated, can only be of minor importance (cooke, 1977; trudgill, Processes 1976b; torunski, 1979) and probably explains only some smaller forms, although it may be more sig- The environment is extremely complex and there nificant in cooler waters (alexandersson, 1976). remains much controversy about the processes in none of the workers has investigated the impact operation. of potential salt-fresh water mixing from meteoric rather than groundwater input. Dissolution Biological action Discussions about the potential for, and possible importance of, dissolution are summarized by The focus of explanations has moved towards con- tre nhaile (1987). Field evidence indicates efficient sidering effects of direct bio-erosion, of boring in- 260 KRF•1 • OK.indd 260 15.12.2009 10:45:40 Joyce Lundberg, Coastal karren                 ­ Figure 3: Diagrammatic shore profiles to show regional variations. The tropical model shows a clear relationship of form and exposure level (a). Local variations on the basic model (b and c) are usual y expressed most clearly in the moderately exposed situation, where variations typical y produce an erosional ramp rather than a double notch and tidal shore platform. The temperate model shows a simple erosional ramp with basins and pinnacles. The cold region diagram cannot be called a “model” since it reflects the impact of isostatic uplift more than intertidal ero- sion. 261 KRF•1 • OK.indd 261 15.12.2009 10:45:40 Karst Rock Features • Karren Sculpturing vertebrates and microflora. Viles (1984) has gone within the photic zone. These are probably the so far as to call coastal karren “biokarst”. There most important borers in the spray and inter- is a very large literature on bio-erosion, not all of tidal zones (Hodgkin, 1970; golubić et al., 1975; it relevant to coastal karren development: Brom- schneider, 1976; Dalongeville, 1977), but fungi ley (1978), Viles (1984), trudgill (1985), trenhaile (probably mainly in bottoms of pools) and lichen (1987), spencer (1988), and spencer and Viles (mainly in drier parts of splash zone; see schnei- (2002) give reviews of this topic. der, 1976) can also bore. Filaments of cyanobac- Bio-erosion is the removal of lithic substrate by teria may penetrate 500 to 900 microns, and of direct organic activities (neumann, 1966). Most lichen several millimetres. up to one third of the exposed rock surfaces both in the marine and ter- rock may be occupied by borehole (schneider and restrial environment acquire a complex “bio-film” torunski, 1983). dominated by cyanobacteria, but including algae, Boring by invertebrates is also common. Mol- fungi and lichens (Viles et al., 2000) and most of luscs often carve a home scar to shelter in. gastro- these biofilms show a distinct relationship with pods, such as small whelks, are important over most rock surface weathering. Bio-erosion is particu- of the coastal karren range, contributing much of larly important in the tropics because of the richly the fretted nature of the rock surface (e.g. Hodgkin, diverse flora and fauna, susceptible rock, and 1970). chitons are also common, and more vora- not very vigorous wave action (trenhaile, 1987). cious, inter-tidal eroders (e.g. abensperg-traun et Duane et al. (2003) found active bio-erosion in al., 1990). echinoderms colonize close to low tide the arid coastal terraces of Morocco. scanning level and are important in deepening intertidal electron microscope studies in ireland (trudgill, pools. Boring clionid sponges are more significant 1987) showed bio-erosion to be the most important in subtidal locations and are thus important for agent of formation of medium-scale and micro- subtidal notch development (Yonge, 1963). morphology. Both epiliths and endoliths are preyed upon by levels of activity and species distribution are grazing invertebrates such as chitons, gastropods, controlled mainly by moisture, and therefore tidal echinoids, crabs, parrot fish. grazers mechanically regime and wave energy. generally the wetter the rasp the rock surfaces, which have been weakened environment the greater the bio-erosion (schnei- by biochemical weathering or by boring, with der, 1976), but it may be reduced where energy lev- some form of hardened teeth or radulae. a large els are too high for organisms to survive or where gastropod can carve a groove 0.5 mm deep and 1 abrasion or sediments prevent colonization. rock mm wide in a single traverse (newell and imbrie, type is important in that soft, fine-grained rocks 1955). The rate of rock removal is determined by are easily bored, especially by mechanical borers a kind of homeostatic balance between the rate such as invertebrates, and the greater the carbon- of grazing and the rate of boring (golubić et al., ate content the easier it is to chemically bore (e.g. 1975). by algae, fungi). Moses (2003) demonstrates that The action of microbes is not always simple bio- algal boring on calcarenite avoids the grains and erosion: in the marine terraces of Morocco, Duane operates selectively on the calcite cement. et al. (2003) found a complex suite of organisms Bio-erosion is effected by a combination of sur- responsible for both disintegration of matrix and face dwellers, direct borers and by grazers. epi- deposition of micro-stromatolites to depths of 50 liths, such as algae and cyanobacteria, live on the cm within the limestone. rock surface and contribute aggressive organic Biological action often results in protection of chemicals. endoliths actively remove rock by bor- the rock surface: this can be effected simply by a ing into it. The depth of penetration depends on covering of organisms resistant to wave energy or light levels: algae and cyanobacteria must stay by the presence of calcareous encrusting organ- 262 KRF•1 • OK.indd 262 15.12.2009 10:45:40 Joyce Lundberg, Coastal karren a c b d Figure 4: Brackish and fresh water forms: a. very wide pans of the supralittoral, north-eastern coast of Bonaire, Neth- erlands Antilles. This is open to rain water input, salt water input only during storms, and often subject to intense evaporation. Although the pools are occupied by brightly-coloured algae that can tolerate hyper-saline condi- tions, it is likely that the principal actions are salt-fresh water mixing corrosion and salt weathering; b. supra-litto- ral basins from the west coast of northern Norway at Gåsbakken. Slope causes the basin asymmetry (shallow lips, deep backwal s). Surfaces within basins have secondary 1–2 cm deep semi-spherical pits. Basin rims are rounded from salt action and grusification. Width of view is 75 cm; c. rillensteine (microril s) developed on aphanitic lime- stone in the supra-littoral zone of Gower Peninsula, south Wales. Width of view is 7 cm; d. freshwater pitting in do- lostones of Bruce Peninsula, around Tobermory, Lake Huron, Canada. This sample is from the splash zone above water level. The below-water level pits are deeper. Width of view is 90 cm. 263 KRF•1 • OK.indd 263 15.12.2009 10:45:46 Karst Rock Features • Karren Sculpturing isms. For example, naylor and Viles (2002) found ered to be true coastal karren. However, in reality, that colonizing macro-algae afford bio-protection processes often are not segregated. The influence and that this is most apparent under exposed con- of salt action on coastal karren is variously report- ditions. The types of organisms performing a pro- ed. emery (1946) described karren in calcareous tective function vary with temperature. in cool sandstones of southern california as dominated temperate regions barnacles colonize the drier by shallow flat-bottomed tidal pools with raised parts of the upper littoral and dense colonies of rims where evaporation has deposited salt. in mussels the upstanding parts of the lower littoral, contrast, medium to coarse-grained limestones in thus facilitating differential downcutting of the coastal zones often with strong drying winds may basins. in tropical regions calcareous algae, ver- erode in the form of tafoni (cavernous weathering metid gastropods and serpulids worms both pro- forms with spherical elliptical hollows; trenhaile, tect and encrust. 1987; Kelletat, 1991). Moses (2003) found evidence Biological processes work to varying degrees for salt weathering producing granular disintegra- in all limestone coasts of the world. The resultant tion of calcarenites in the splash zone, but in ob- karren forms relate to the interplay of environ- vious association with bio-erosion. The supra-lit- mental factors (such as tidal regimes, water tem- toral bowls in the marbles of norway (lundberg perature) and inter-specific interactions (such as and lauritzen, 2002) clearly show granular disin- competition, grazing and predation) (spencer and tegration from salt action. Viles, 2002). For example, colonization by echi- noderms may have little effect in a sheltered en- vironment such as the Mediterranean (torunski, Abrasion 1979) or in the sublittoral because there is no need for a substantial home scar for protection, but The presence of tools for abrasion normally results in the lower eulittoral echinoderms modify the in a less diverse biota and the smoothing of forms, environment considerably with distinctive semi- but trudgill (1979) examined the role of abrasion spheroidal cups. There is a positive feedback effect in relation to other processes in aldabra atoll and once the first colony is established and the rock is concluded that even in areas far from beaches rapidly removed to create a deep pool (trudgill, abrasion can be important. 1987). However, if the water temperature is high then there is greater competition, the lower eulit- toral may also be colonized by encrusting organ- Acknowledgements isms which inhibit echinoderm colonization and home scar formation and thus deep pools may not Many thanks to John Mylroie and Joan Fornós form. a factor that is becoming more apparent is and lluís gómez-pujol for excellent comments on that the biological environment of a rock surface an earlier version of this manuscript, and to craig is dynamic and may show considerable temporal/ Malis for the photograph in Figure 1b. spatial fluctuation (spencer and Viles, 2002). salt action as noted above in the introduction, features that are strictly produced by salt action are not consid- 264 KRF•1 • OK.indd 264 15.12.2009 10:45:46 case studies KRF•1 • OK.indd 265 15.12.2009 10:45:47 KRF•1 • OK.indd 266 15.12.2009 10:45:47 liMestone paveMents in tHe BritisH isles 21 Peter VINCENT Limestone pavements are among the most distinc- By now it should be self-evident that conven- tive landforms in the British isles, and for the pur- tional definitions of limestone pavement fail to poses of this chapter i have adopted the following accommodate the polygenetic nature of these definition, slightly modified, from goudie (1990): surfaces as presently identified in the British karst “exposed areas of bare limestone bedding planes, landscape. if all this is a little disturbing then i both flat and sloping, which are often, but not al- will have achieved my purpose. to regard these ways, fretted by microforms produced dominant- features as a simple glaciokarst and then place the ly by solution activity.” research emphasis on the presence and measure- it is noteworthy that above definition says ment of karren is to miss the point entirely. For nothing about the processes which may have a proper understanding of the genesis and vari- eroded the limestone to give rise to the pavement, ety of these intriguing landforms it is necessary nor does it indicate that the presence of karren is to examine denudation process in relation to the a necessary condition. as far as the latter point detailed geology (Vincent, 1995). is concerned, there are, for example, several fine limestone pavements have attracted the atten- examples of pavement in northern england that tion of British scientists for more than a century are almost totally free of karren. and while it is (sweeting, 1972; Williams, 1966) but somehow true that much pavement in the British isles has developments in karst geomorphology seem not been produced by glacial abrasion, it is also true to have engaged much with developments in our that some pavements have been exhumed, mostly understanding of the lower carboniferous rock by glacial plucking ‒ but not always. to nuance succession. indeed, as far as the author is aware, this point further there are magnificent exam- there have been no major studies in this regard ples on the coasts of western ireland, north Wales since the work of sweeting and sweeting (1969) and north-west england of limestone pavements who investigated the relationships between litho- which may have been exhumed by marine action logy and pavement form in north-west Yorkshire rather than by ice. These are not marine abrasion (england) and the Burren, county clare (irish platforms but demonstrably limestone pavements, republic). as we shall see. Yet a further type of glacially Both the rise of process geomorphology and scoured pavement is found in northern england the quantitative revolution in geomorphology and forms part of a truncated relict palaeokarst further deflected research away from the geology landscape. and two research themes have prevailed for much 267 KRF•1 • OK.indd 267 15.12.2009 10:45:47 Karst Rock Features • Karren Sculpturing of the thirty years or so. some researchers have Geological control investigated the influence of plants and the soil cover on the corrosion of the pavement surfaces, since the 1970s there have been a number of de- for example trudgill (1985), while others have in- velopments in carbonate and carboniferous geol- vestigated karren types and pavement morphom- ogy that have a direct bearing on the genesis and etry (goldie, 1996; Vincent, 1983b; rose and Vin- morphology of British limestone pavements. With cent, 1986b). The time is now ripe for a paradigm the exception of small areas of pavements devel- shift back to an examination of the essential rela- oped on the cambro-silurian Durness limestones tionship between pavement form and genesis in in northern scotland, limestone pavements in the relation to the lower carboniferous stratigraphy. British isles are developed in lower carbonifer- ous (Dinantian) carbonate successions, and in particular the asbian and early Brigantian stages. Pavement distribution These sediments were deposited on extensive shal- low-water, flat-topped carbonate platforms (Walk- although the presence of cyclic limestone se- den, 1972) which show marked small-scale rhyth- quences of asbian age are the main requirement mic cyclicity, which is absent from the underlying for the formation of limestone pavements in the Holkerian and earlier strata. The cyclicity is due to British isles their distribution does not always co- changes in relative sea level but its exact cause is incide with a suitable process domain. Thus, for still a matter of debate. example, there are asbian stage limestones in the asbian limestones display a variety of litholo- Mendips, in Derbyshire and in south Wales but gies which are an expression of water depth and limestone pavements are for the most part ab- environmental energy in which they were laid sent. in the case of Derbyshire and the Mendips, down. individual cycles comprise a sequence of pavement-forming ice scour was absent as both carbonate lithofacies arranged in a shallowing areas lie south of the ice limits of the last glacia- upward succession. each cycle was initiated by a tion Maximum (lgM). lgM ice did pass across rise in sea level that inundated the carbonate plat- south Wales but the limestone outcrops are nar- form. intervening marine regressions culminated row and the bedding planes upturned (Waltham in subaerial exposure and karstification. et al., 1997). The average thickness of the cycles varies from The largest area of asbian stage pavement in a few metres in the early asbian to around ten the world is found on the Burren, in county clare, metres in the overlying Brigantian. some 25 to irish republic (Figure 1). Here, some 290 km2 of 35 cycles have now been identified in the asbian pavement have been formed as ice passed south- and in the Brigantian there are probably no more wards across galway Bay from the mountains of than a dozen or so. if it is assumed that the as- connemarra and the corrib Basin. bian lasted for about 9 Ma and the Brigantian for 6 Within mainland Britain there are only 29 km2 Ma (george et al., 1976), individual asbian cycles of limestone pavement and as much as 97 percent lasted between 260,000 and 360,000 years, and has been damaged. The most extensive upland Brigantian 500,000. palaeokarstic surfaces, clay pavement occurs in the ingleborough and great palaeosols, and calcretization of the host rock at asby scar regions of the northern pennines (Fig- the top of each cycle are diagnostic features of the ure 1). There are also small areas of pavement subaerial exposure of Dinantian carbonate plat- found on north Wales and in the far north-west forms during periods of low sea level. of scotland. in the asbian and early Brigantian rocks of the Derbyshire Dome Walkden (1974) showed conclusively that the major bedding planes were 268 KRF•1 • OK.indd 268 15.12.2009 10:45:47 Peter Vincent, Limestone pavements in the British Isles Caledonian Hinterland Asby Scar The Burren St. Georges Land shelf and platforms carbonates Figure 1: Distribution of Asbian shelf sediments in the British Isles. produced by contemporaneous carbonifer- accumulations of volcanic dust (Walkden, 1972). ous subaerial weathering and the formation of These clays are often known in the literature as karstic surfaces. according to Walkden, the pal- wayboard clays ‒ an old quarrying term. The cor- aeokarstic surfaces underlie thick bentonitic clays roded, karstified surfaces can be shown to have which are interpreted by Walkden as subaerial formed before the next deposition of limestone 269 KRF•1 • OK.indd 269 15.12.2009 10:45:48 Karst Rock Features • Karren Sculpturing Figure 2: Clay wayboard in Asbian limestone. Crosby Ravensworth, Cumbria, England. Figure 3: Mul ach Mór (Burren). Schichttreppenkarst pavement sequence developed on calcreted horizons in the Asbian limestones. 270 KRF•1 • OK.indd 270 15.12.2009 10:45:52 Peter Vincent, Limestone pavements in the British Isles because the lower surface of the overlying lime- and later, by farming activities, and partly as a re- stone is always sub-horizontal and does not follow sult of the development of the grike systems which the irregular palaeokarstic surface (Figure 2). This drained the pavement of its silty cover (raistrick, general model of palaeokarstic development holds 1947). looked at in another light, British lime- for all asbian shelf limestones in the British isles stone pavements are magnificent examples of a and thus includes all the major limestone pave- soil-eroded landscape. ment regions. at norber Brow, a few miles south of ingle- of particular relevance to the development of borough, glacial erratics, eroded from an inlier limestone pavements is the diagenetic alteration of silu rian greywacke in the valley bottom, have of the limestone immediately under the wayboard been lifted by ice onto pavements on the valley clay. Horbury (1987) showed that these limestones sides. These erratics are now perched on lime- are mottled with dark-grey circular markings sur- stone pedestals mostly 400‒500 mm above the rounded by pale-grey matrix. The darker patches general pavement level, the erratics having pro- are interpreted as a diagenetic alteration of the tected limestone surface from postglacial solution. limestone caused by organic acids and associated There is nothing extreme about the norber site with vegetation growing in the subaerial wayboard. and it is probably reasonable to assume this figure in addition many palaeosurfaces have crusts of for pavement lowering in northern england. dark grey laminar calcrete. This sub-wayboard The coastal pavements at Fanore (irish grid diagenesis results in poorly jointed, massive lime- M 13 08), in the south west of the Burren, are in- stones which are relatively resistant to erosion as teresting since open grikes probably existed be- compared with those limestones above the way- fore dunes buried the pavements. erosion of the board. This differential jointing explains the devel- dunes has now revealed a polished pavement sur- opment of the schichttreppenkarst (Bögli, 1964) face with the grikes having been filled by walls of by differential glacial erosion, both in the Burren proud beachrock formed from the dissolution of and north-west england (Figure 3). a particularly the shells in the dune sand. good exposure of a clay wayboard can be found in ireland at the entrance of the ailwee caves (irish grid M 23 05) and the associated calcreted lime- Pavement types stone bench can be traced all round the Bally- vaughan embayment (irish grid M 08 23). The intersection of geological control and proc- ess domains naturally gives rise to five pavement types, namely: glacially plucked joint-dominant Post-glacial pavement development pavements; glacially abraided calcrete-dominant pavements without palaeokarst; glacially exhumed although there are exceptions, as we shall see calcrete-dominant pavements with palaeokarst; later, a general model for post-glacial pavement glacially exhumed calcrete-dominant pavements exposure is as follows. There is now abundant evi- with palaeokarst; glacially truncated palaeokarst; dence that most limestone pavements in the Brit- and marine exhumed paleocarst pavements. ish isles were at one time covered by late gla- cial loess (Vincent and lee, 1981; Vincent, 2004) which, in wetter situations, may have developed a Glacially plucked joint-dominant pavements peaty cover. it was under this blanket of loess that widespread rundkarren developed. Much of this This type of pavement is formed where glacial loess cover has now been lost partly through dis- scour coincided with non-calcreted horizons turbance by Mesolithic and iron age clearances above clay wayboards. These zones have high joint 271 KRF•1 • OK.indd 271 15.12.2009 10:45:52 Karst Rock Features • Karren Sculpturing Figure 4: Aerial view of Asby Scar showing a mosaic of glacial y trun- cated palaeokarst, ex- humed calcrete-dom- inated pavement, and glacial y eroded joint- dominant pavement. densities which made them less resistant to gla- stones pavement north of castle Folds is but one cial plucking processes such as hydraulic jacks of many in the area, and at gait Barrows national and heat pump effects. The net effect has been to nature reserve (sD 482 775) ‒ especially the al- produce a pavement with abundant surface clitter most grikeless central pavement. in both cases, and relatively small clints. Many of the pavements schmidt hammer results confirm the high com- on the flanks of ingleborough are of this type and pressive strength brought about by case hardening good examples can be found at the head of crum- due to the formation of calcrete. mack Dale. it is interesting to speculate why glacial scour did not penetrate deeper into the carbonate cyclic Glacially exhumed calcrete-dominant succession in these areas. one possibility is that pavements with palaeokarst the ice was not particularly erosive and may have been thin and also possibly sluggish due to a cold sub-wayboard surfaces are laterally quite variable base. another possibility is that the ice was ero- (Vanstone, 1998) and where wayboards were thick, sive but scour was time-limited. karstified, calcreted surfaces developed. in places, these surfaces have been exhumed by glacial ac- tion. The evidence for exhumation is unequivocal Glacially abraided calcrete-dominant in that the palaeokarst can be seen passing lateral- pavements without palaeokarst ly underneath wayboard clays. This type of pave- ment is also massive, with few joints, but contains Where glacial scour coincided with the hardened, saucer-shaped pits and some runnels and a non- joint-poor calcreted horizons beneath the way- planar topography. good examples can be seen at board clays the pavements were abraded but not gait Barrows national nature reserve in north- plucked. The net result was to produce a smooth ern england, on the Morecambe Bay coast near pavement almost a totally without joints. There grange-over-sands and silverdale, and also at the are excellent examples of such pavements at great entrance to the ailwee caves in the northern Bur- asby scar (nY 6409 – nY 6809) ‒ the shining ren. 272 KRF•1 • OK.indd 272 15.12.2009 10:45:55 Peter Vincent, Limestone pavements in the British Isles Figure 5: Palaeokarstic pavement exhumed from under wayboard and block beach fal . Inishmore, Aran islands, Ireland. Glacially truncated palaeokarst sion of palaeokarstic surfaces exhumed by marine action. These are spectacularly developed at many at great asby scar in cumbria it is clear that ice sites on the atlantic facing coasts of the aran is- moving eastward away from an ice dome over the lands, western ireland and on the north coast of lake District was not effective in completely re- anglesey, north Wales. moving a pre-glacial palaeokarst (Figure 4). at at first glance it might seem that these pave- this site, a glacially truncated pre-glacial karst is ments are just marine abrasion platforms with fretted with almost completely circular solution biokarstic pits but this is demonstrably wrong. pipes which now form part of the general pave- The evidence is remarkably simple and merely ment mosaic. similar sites can be found on the requires the investigator to look cliffward rather fells above appleby in the eden Valley, and possi- than seaward where it will be seen that the karsti- bly in the ingleborough region (Jones, 1965). fied surface continues into the cliff section and excavations of the solution pipes at asby reveal under well-exposed wayboard clays (Figure 5). that they are filled with a dense, mottled, reddish Furthermore, it is clear from topographic profiles clay. XrD analysis shows the clay to be kaolinite measured immediately under the wayboards and rich and completely unlike a wayboard clay. The also out on the exposed pavement that there has pit clays may originally have been a saprolite de- been little surface lowering or pit deepening/wid- veloped on the shap granites to the west which ening since the pavement was exposed. The exact was carried onto great asby scar by glaciers and cause of the exposure has not been examined in plastered onto the surface of a pre-Quaternary detail but one possibility is that these pavements karst. have been exhumed as a result of the development of so-called block beaches. During major storms the heavily jointed limestones on the cliff face be- Marine exhumed palaeokarst pavements come loaded with water, fail and collapse. These fallen blocks are then eventually removed by wave The definition of limestone pavement suggested at action, and the poorly jointed palaeokarstic pave- the beginning of this chapter allows for the inclu- ment is then exposed. an alternative explanation 273 KRF•1 • OK.indd 273 15.12.2009 10:45:55 Karst Rock Features • Karren Sculpturing might relate the exhumation to glacial plucking limestones. in general, differential glacial pluck- of the cliff faces as ice passed from connemarra ing and abrasion seems to have been controlled southward across the islands and on out into the by the calcretization and consequent joint densi- atlantic. ty variation within each cycle. British limestone pavements are thus in part postglacial, in part exhumed lower carboniferous surfaces, and in Conclusions places part of an older pre-Quaternary, possibly tertiary, karst. Viewed in this way, these fascinat- it now seems clear that the morphology of Brit- ing landforms are clearly important lines of evi- ish and irish limestone pavements can mostly be dence when attempting to decipher the wider geo- accounted for by interaction between erosion/ex- morphology of the regions in which they are lo- posure processes and cyclic events in the asbian cated. 274 KRF•1 • OK.indd 274 15.12.2009 10:45:55 case stuDies oF griKes 22 in tHe BritisH isles Helen S. GOLDIE an initial discussion of general morphometry is The site with the greatest median depth was appropriate to establish a basis on which to fit ingleborough, 99 cm, and the shallowest was 42 specific local cases. as Day says (gunn, 2004): “… cm in Wales but ambiguity of grike measures has quantification of form has made inter-regional been commented on with great range and local comparison more rigorous. it has also helped in variety. in order to understand the variety of re- the development of meaningful indices of land- lationships between measures, pre-glacial condi- scape morphology and it has clarified the role of tions, as well as depth of glacial scour, must be lithological, structural and other factors influ- invoked. reference can be made to variants of the encing karst landform development.” reference morphological model, based on the goldie–cox is made to goldie and cox’s work on limestone model (2000) (see Figure 13 in chapter 9). The bed pavement morphometry at British, irish and swiss thickness factor must be carefully examined lo- sites, presenting data on clint and grike measures. cally, as an important influence on joint spacing Variations between and within field sites relate to and thus grikes. Bed thickness will also, through depth and timing of glacial scour, rates of post- being linked to strength, influence survival of gla- glacial solution, tectonic disturbance, and lithol- cial scour, since massive beds, with widely spaced ogy, whilst human impact is also identified as in- joints, produce very large blocks of rock, propor- fluencing present-day landforms directly and in- tionately harder for ice to remove. rose and Vin- directly. an important possibility is that some of cent (1986a), examining grike data at gaitbarrows, the wider grikes existed before glacial erosion and identified a bimodal distribution and considered survived to be re-activated and to develop greater that this demonstrates two ages of grikes. How- width and depth during the Holocene. The irish ever, a bimodal distribution is not necessary to sites in the Burren and on arainn are in areas support pre-glacial survival. pre-glacial survival heavily scoured by the irish midland ice-sheet will merely widen population range as surviving (Mccabe, 1987), and in sympathy with this the scoured fissures will range in size from minimal data here show relatively narrow grikes (see chap- sizes up to some large survivors, and the minimal ter 9), although it is still theoretically possible for ones will merge with the new Holocene grikes. there also to be larger, pre-Devensian survivors. Varied morphology is consistent with sugges- The arainn pavements have the grike population tions about ice scour, given varied pre-glacial with the lowest median width (9 cm), although conditions as well as differing degrees of scour. this site has maximum width of 1 metre. arainn’s grikes, consistent with deep recent ice 275 KRF•1 • OK.indd 275 15.12.2009 10:45:56 Karst Rock Features • Karren Sculpturing scour, are indicated by level D on the develop- barrows (lancashire, uK) is also an excellent ment model (see Figure 13 in chapter 9). The site to examine for the evolution of grikes from glattalp (switzerland) grikes may reflect lesser immature features with which perna’s 1996 ob- scour resulting from divergent ice streams on a servations of the trentino area make interesting broad col, a similar situation to sanetsch (swit- comparison. zerland). at sanetsch at present very recently de- pavements with smal er clints stil show enor- glaciated limestone surfaces have visible fissures mous variety in grike data. The ingleborough and filled with debris now, which will be the wider cumbria sample areas have both deep and wide grikes of the future. sanetsch’s limestone is also grikes but high variability particularly over depth. strikingly veined, and these veins affect how the High grike depth variability is also observed in the grikes develop. gillespie et al. (2001) discuss both Yorkshire Dales areas further east, in Malham and jointing and veining and their separate influenc- Wharfedale. There has also been a complex rela- es on fissure patterns in the Burren pavements. tionship here between topography and glaciation, The Burren’s moderately wide and deep grikes with recent field observations supporting the view include several locations where glacial scour has that features have partly survived glacial scour. produced situations B or c in the model. gait- other Dales sites have experienced considerable ˆ ‰ ‡ˆ ‚  ƒ  €„   € †      ­     € ‰Š‹ƒ ‰Š‹ƒ Ž‹ Œ‹ Figure 1: Location map of main field sites in north-west England. 276 KRF•1 • OK.indd 276 15.12.2009 10:46:44 Helen S. Goldie, Case studies of grikes in the British Isles human impact from direct removal of upper clints dividual grikes and networks (rose and Vincent, especial y in Wharfedale. Human removal of clints 1986b). Figure 3 shows a short sequence of round has a significant effect on grike depth, though holes resulting from boring into the limestone by a somewhat more limited effect on grike width plants along such a veined ‘weakness’. Figure 4 (goldie, 1986). clearly grike depth reflects the lay- shows a more developed line of small solution fea- ers of rock through which fissures exist although tures, which will eventually merge and become a fracture propagation from one bed to another must true grike. The plan outline of that grike reflects not be assumed. Joints may exist only through one its evolution, and the wavy plan outline of more rock layer, whilst others propagate across several developed grikes betray similar origins. thus crucial y influencing grike depth. Human re- a further development of vein influence is that moval of clints may thus mean very immediate ef- their clustered alignments at staggered angles as- fects on clint sizing, but not always. sociated with structural influences (en echelon) Major factors influencing grike development can best be illustrated with the case studies (Fig- ure 1) which include: gaitbarrows, The clouds, great asby scar, Farleton Knott (including Hut- ton roof crags and Farleton Fell), Whernside and ingleborough, Malham Moor (including cowside Beck, High sleets and Dowkabottom), and Whar- fedale and Hampsfield Fell, and some compari- son with Derbyshire. The sites will be discussed according to locally important characteristics (Waltham et al., 1997; goldie, 1995, 2006), for ex- ample, early stage features, structural influences, stream flow, palaeokarst characteristics and evi- dence of mature development including glacial survival, or human influence. Immature grike features: Gaitbarrows clearly one major influence on grike characteris- tics is the time available since the grike began to evolve. However, when a feature becomes a grike from an unfissured stripped surface is a question that is hard to answer, as there is no minimal defi- nition. Figure 2 shows early stages of grike forma- tion at this site in nW england, which has been well scoured by ice in the late Devensian (rose and Vincent, 1986a) but retains a great variety of features, including probable survivors of ice scour that developed into larger features in the Holocene. Figure 2: Line of bio-corrosion holes developing along a The carboniferous limestone here has calcite veins vein in wel scoured limestone pavement. Gaitbarrows, that influence alignment and features of both in- Lancashire, UK. 277 KRF•1 • OK.indd 277 15.12.2009 10:46:45 Karst Rock Features • Karren Sculpturing result in linear zones of complex grike patterns. at gaitbarrows such zones are lower than less griked areas (Figure 5). This may be because greater frac- ture density here favours increased solutional lowering. another possibility is that glacial scour gouged down into these zones preferentially due to their vein-related weakening. Thus after glacia- tion, erosion favoured these areas for more than one reason, the rock weaknesses, and also their lowering by glacial scour, both resulting in drain- age waters focussing on these lower zones. Structural influences Figure 3: Kamenitza which has been drained downwards by the development of a slit which breaks through the The importance of folds has been mentioned (see top bed. Gaitbarrows, Lancashire, UK. Width of view is chapter 9). great asby scar in cumbria demon- 1.8 m. strates this with a gently folded structure (Figure 6). The clouds, cumbria, are carboniferous lime- stone outcrops very close to a major shear stress fault (Dent Fault) that strongly influences the to- pography (underhill et al., 1988; goldie, in Fornós and ginés, 1996), particularly causing close rock fissuring. typical clint and grike arrays here have strikingly closely spaced grikes and narrow clints. However, there is also now a greater appreciation of the age of some of these features since, despite the closeness of jointing, it is obvious from Figure 7 that some beds are extremely thick, ca. 1 to 1.5 m, similar to those observed in the Malham area of Yorkshire, and similarly well-rounded in out- line. This roundedness does not result from gla- cial scour as it extends all round the blocks; it is a shaping resulting from pre-glacial karstification (Figure 8). Thus, in spite of glaciation and lengthy periods of periglacial conditions in closely-jointed rock, The clouds retains remnants of ‘maturely’ weathered karstic features. The usual relationship between bed thickness and joint spacing is over- ridden where major tectonic stresses have meant joints are closer together than at outcrops of sim- Figure 4: View along a prominent vein showing a series ilar thickness but distant from such influences, of features at different stages of development along such as in the Malham area. this line. Gaitbarrows, Lancashire, UK. 278 KRF•1 • OK.indd 278 15.12.2009 10:46:47 Helen S. Goldie, Case studies of grikes in the British Isles Stream flow influences scar close on ingleborough has some of the most massive pavements in Yorkshire with a mixture of evolving grike network patterns. The inner edge of the outcrop near peaty glacial deposits (gosden, 1968) has acidic drainage off these ‘islands’ and the marginal shale cover (see annotated diagram Fig- ure 9). This drainage has caused dendritic runnel- ling patterns, forming grikes when the top lime- stone layer is cut through. towards the outer valley side of the outcrop, where the limestone has been longer exposed, more rectilinear patterns reflect increasing influence by rock fractures on grike de- velopment rather than the superimposed dendritic patterns. There is also evidence that the outer scar edge has pressure release fissures, since numerous grikes run sub-parallel to it. lastly, at a scale order higher than individual grikes, large griked areas run across scar close from south-east to north- west, which have a varied plan outline comparable to that referred to earlier for gaitbarrows, but in which the holes are essentially large grike holes or dolines often with significant soil and vegetation. The spacing of these wider griked lines is of the order of tens of metres, and a similar scale of fis- Figure 5: Linear complex of grikes developed along sures is responsible for major indentations in cliff veins in association with a major fissure, al of which outlines, and for small valleys in this area. is at a lower level than the general pavement surface. ingleborough also has sinkholes on its west Gaitbarrows, Lancashire, UK. side, worth discussion here as they are in pave- ment-like outcrops and are fundamentally wid- smooth-sided, about 50 cm to one metre wide, ened fissures (Figure 10). These examples are very caused by solution from significant streams flow- Figure 6: Sketch profile across Great Asby Scar, showing relations between geological features and landforms. C. mesa (Castle Folds); S. main syncline; A. denuded anticline; M. mushroom features; H. large holes on steep slope; LH. large holes and extended holes. Not to scale, notional distance across section is ca. 400 m. 279 KRF•1 • OK.indd 279 15.12.2009 10:46:48 Karst Rock Features • Karren Sculpturing Figure 7: The Clouds. Massive beds on upper layer at amphitheatre edge (centre of anticline). ing off the ingleborough massif, which they swal- low. Dry versions are found further out on the outcrop, abandoned as streams that formed them disappeared underground down newly opening upstream fissures. similar exceptional ‘grikes’ in- volving stream swallowing include Hunt pot, near pen-y-ghent. Thus grikes are also features known by other terms; cavities, or major drainage points, also termed potholes. on arainn runnels of simi- lar dimensions contain peaty streams. Many such dry features, now on limestone pavement out- crops abandoned from past drainage points, pro- Figure 8: Plan diagram (not to scale) of rounding dem- vide puzzles for geomorphologists. onstrates that roundedness is found at the sides (C-D) and backs (E-F and G-H) of outcrops edges to the same degree as at the facing edge (A-B). This indicates that Palaeokarstic features and mature grikes the rounding is caused not by glacial scour along the outer edge (X-Y) but is probably the remain of karstic processes. evidence from grike dimensions, shapes, and re- lationships to other features, helps address the 280 KRF•1 • OK.indd 280 15.12.2009 10:46:49 Helen S. Goldie, Case studies of grikes in the British Isles Figure 9: Sketch cross section looking east across stepped boulder – pavement sequence, Newbiggin Crags, Farle- ton Knott, Cumbria, UK, showing sequence of maturely weathered and scoured limestone beds. I. upper pave- ment layer, wel dissected; I . “normal” pavement, merging on into same bed; I I. extremely wel rounded pavement edge with boulder shapes; IV. “normal” pavement; V. same bed as IV, but with grike holes and wel rounded clints and boulders. Not to scale: notional distance across section is ca. 50 m. Figure 10: Ingleborough sink demonstrating stream dis- appearance down a wid- ened grike. 281 KRF•1 • OK.indd 281 15.12.2009 10:46:51 Karst Rock Features • Karren Sculpturing question of landform age. This can be illustrated and newbiggin crags (Figure 12). general grike at many carboniferous limestone outcrops in nW patterns have been discussed by Moseley (1972). england and elsewhere in the British isles. rose There are many distinctive mature karst features and Vincent (1986a) discussed the significance of along outcrop edges (Figure 13), including higher the erratics in grikes at gaitbarrows concerning beds, and the sides of major geological weakness- grike maturity and similar evidence is observable es cutting through the area. These weaknesses at scales Moor, scar close and souther scales in produce narrow low-lying zones, which must have the Whernside-ingleborough area of Yorkshire, been sheltered from ice scour in order for the well- and at great asby scar, among many examples. an important grike characteristic is the flared shaping of their upper edges and it is clear that the concept of a grike cannot be separated from clint definition (Figure 11). The diagram summarizes the range of possibilities, demonstrating develop- ment of curvature with time. Bed thickness and fracture spacing influence the outcome. The de- velopment model (see Figure 13 in chapter 9) ex- pands on the various likely relationships between the main influencing factors. numerous sites have particular aspects of maturity to exemplify. Farleton Knott: The area known as Farleton Knott includes the well-known pavement sites, Farleton Fell, Holme park Fell, Hutton roof crags Figure 12: Sketch map of Farleton Knott, Cumbria, UK, northern half, to show location of mature karst fea- tures cited in text, and relations to direction of Lake District ice flow (double arrow). Solid line. main faults and major fractures; numbered arrows. angles of dip; dashed lines. outline of bedding plane edges; FF. Far- leton Fel ; NBC. Newbiggin Crags; HPF. Holme Park Fel ; R. rounded cliff tops; A. tower-like features and em- bayments; m. massive rounded edges; OXO. NBC se- Figure 11: Grike evolution, defining and grading round- quence of boulders and pavement; Q. Holme Quarry; edness. G, G’. original grike centre lines; W. original dots. glacial y-moved boulders-approximate distribu- block top width; H. original block top height; a–e. stag- tion; 1. ice spreading towards sheltered east side; 2. ice es of upper roundedness of the limestone; f–g. stages diffluent over central ’hump’. Notional distance across of under-rounding of the limestone. map = 1.3 km. 282 KRF•1 • OK.indd 282 15.12.2009 10:46:51 Helen S. Goldie, Case studies of grikes in the British Isles Figure 13: Massive rounded bed edge clints at the north end of Farleton Knott (Farleton Fel ). rounded massive features found there now to sur- vive. The Hutton roof crags dolines are mostly located along these fissure lines and their edges have rounded boulders and clints, and wide flared grikes that cannot have evolved entirely during the Holocene. The famous sloping diamonds are shown in Figure 14. The north-east end of newbiggin crags on Far- leton Knott has some very unusual features not found in the same sequence anywhere else known to the author (Figure 15). in general they are simi- lar to the boulder features along the Hutton roof dolines and differ slightly from mature features at great asby scar where grike holes dominate. a very massive upper limestone bed at newbiggin crags has permitted remnant clints between wid- ening grikes to become boulder-like. Below these features is conventional pavement, and below this is another massive bed with widely flared grike holes, more akin to the great asby scar holes. The relationships between the various layers point strongly to the more rounded features surviving Figure 14: Diamond-shaped clints on the steep limb of glacial scour, whereas the middle layer must have the Hutton Roof Monocline (Hutton Roof Crags). 283 KRF•1 • OK.indd 283 15.12.2009 10:46:53 Karst Rock Features • Karren Sculpturing Figure 15: Panorama looking east at the northern end, Newbiggin Crags. To the far left are wel rounded clints with flared grike holes leading on to the relatively regular pavemented clints in the same bed, to the right are very wel rounded clints and boulders of the more massive upper limestone layer (Newbiggin Crags). Figure 16: Panorama looking east from the north end of Castle Folds showing varied grike and clint shapes. To the left in the distance are the relatively rectangular shaped clints in the lower massive bed, to the right are the wel - rounded clints around the grike hole of the more massive uppermost limestone layer (Great Asby Scar). Figure 17: Bed edge features around the surviving upper limestone layer north of Castle Folds and in its ‘shadow’ in terms of glacial scour from the south (Great Asby Scar). Figure 18: Close-up of holes to the south end of Castle Folds demonstrating the large number of fissures radi- ating around them (Great Asby Scar). 284 KRF•1 • OK.indd 284 15.12.2009 10:47:03 Helen S. Goldie, Case studies of grikes in the British Isles had features small enough to have been plucked and origin in the late carboniferous is one pos- or scoured away as it now forms a flat runnelled sible explanation. surface. Measurements here suggest ca. 8 cm of at great asby scar the grike holes also vary solution since bedding plane stripping. considerably. Those in the very thick upper beds Further variety of grike features on Farleton have a few joints crossing aiding formation, and Knott includes regular rectangular clints, to the are vertically oriented to topography not bedding south along newbiggin crags (nBc), and on the (Figure 6). others on horizontal beds areas are south-east corner the famous diamond-patterned smaller and shallower, but in thinner beds and features at Hutton roof (Figure 14). grike patterns of more complex form having numerous joints at both sites suggest that joint sets are very nearly and other solutional features draining towards equally strong in two directions at right angles to the centre. in a study of 37 holes, an average of each other. at Hutton roof the clints are diamond- 5.6 runnels or grikes focussed towards holes, the shaped because of the relationship between orien- average diameter of the top of the flare was 1.45 tation of topographic slope, itself related to incli- m and there was a positive relationship between nation of this limb of the Hutton roof Monocline this and the number of features draining towards (Moseley, 1972), and orientations of the joints sets. the hole (Figure 19). When the features formed squareness as indicated by a clint width to length in their totality is debatable. greater soil cover ratio of 1 (see chapter 9) is not attained, but the in the Holocene is evidenced by smooth surface ratio at these nBc sites is 0.65, at the higher end solutional features and archaeological evidence of the whole population (see chapter 9) indicating of human occupation, unlikely without more soil that these clints are squarer than many. than now (richmond, 1933). Holocene sub-soil great asby scar is part of the extensive car- solution would operate at variable rates depend- boniferous limestone escarpment west of Kirkby ent on conditions of acidity and moisture. evolu- stephen (Figure 1) (goldie, 1996; Waltham et al., tion in the Holocene would have occurred, but to 1997). it has very varied grike patterns, the most what extent this compares with evolution occur- rectangular being where upper beds have been ring much earlier, including the late carbonifer- stripped away (Figure 16). The upper beds are very ous, now needs careful appraisal in view of new massive, often ca. 1 m thick, and where they sur- vive (Figure 17) they are pocked by grike holes of 2‒3 m diameter, usually deep and soil-filled (Fig- ure 18). Vincent (1995) has analysed these features and their fill and explained them as dating from the late carboniferous, which accords nicely with recent determinations of lower solution rates (goldie, 2005). lower solution rates allow for some grikes and grike holes in northern england hav- ing developed over a longer time, possibly since well before the Quaternary, not merely in long interglacials before the Devensian glaciation, or in the Holocene. it is probable that rates of surface solutional lowering in many dry upland limestone areas are 15 to 20 cm in 15 ka at the most, possibly even lower, for example, 5 to 10 cm, as assessed Figure 19: Scatter graph of grike hole upper flare diam- from pedestal evidence. Thus grike holes of 1 to 2 eter (y axis) against number of joints and runnels drain- metres wide or even more, must be palaeofeatures, ing towards each hole (x axis). 285 KRF•1 • OK.indd 285 15.12.2009 10:47:03 Karst Rock Features • Karren Sculpturing evidence concerning palaeokarst features and so- lution rates in these upland limestone areas. structural influence on the holes is clearest in the upper bed, which persists around parts of the synclinal outcrops (Figure 6), and this may have affected how grikes evolved and whether limestones were removed easily by glacial strip- ping because of effects involving joint widening over anticlines. grikes have been strongly influ- enced by an underlying thinly bedded limestone layer, which will have suffered much mechanical weathering, undercut overlying limestone and hence allowed mechanical erosion of the upper layer. in addition a shadow effect is demonstrated where the mesa (castle Folds) has sheltered areas in the lea of ice-flow from the south, leaving the upper bed. The edge of this demonstrates many evolutionary features of upper grike edges as the limestone erodes (Figures 16, 17, 18). Few clints at great asby scar have, however, developed as far as those in upper beds at newbiggin crags which Figure 20: Monk’s Path, above Arncliffe, Littondale; are boulder-like, possibly because the limestone rounded outcrops at ca. 480 m a.s.l. bed at newbiggin is thicker. However, differences between sites could also be due to variations in corrosion conditions or time available. great asby Figure 21: Lee Gate High Mark; rounded outcrops at ca. 460 m a.s.l. 286 KRF•1 • OK.indd 286 15.12.2009 10:47:05 Helen S. Goldie, Case studies of grikes in the British Isles Figure 22: Scoured pave- ment at 330 m a.s.l. show- ing lack of rounded fea- tures, but narrow grikes and flat clint tops (Litton- dale, Yorkshire, UK). scar’s holes could have been modified to become to support this thesis (Murphy and lord, 2003). more vertical by copious melt-water and there is Further north and east, towards cowside Beck, the possibility of biologically enhanced corrosion cote gill and littondale, near Dowkabottom and from vegetation growing in the deep soils in the High sleets, there are many maturely weathered holes. some grike holes still retain shrubs. parts of outcrops, including almost pinnacled clints by the great asby (near castle Folds) may also have their Monk’s path (Figure 20) over 2 m in height on grike features widened by enhanced corrosion their downslope side, and simple tower-like forms. from human and animal occupation, particularly it is suggested that the tower features (2 to 3 m the ‘entrance track’ to castle Folds. high) at Dowkabottom, at about 380 m a.s.l., are Malham-littondale: north of Malham, many located above the main littondale ice-flow. Fur- well-rounded limestone outcrops are found at al- ther evidence for this is provided by pavement on titudes of about 400 m a.s.l. (Figures 20, 21, 22). the main valley side with narrow, shallow, rubble- There is general agreement that this area is a ma- filled grikes at a lower level, about 330 m a.s.l., ture karst with sweeting (1966) regarding it as the which appears to have been ice-scoured (Figure most evolved area of the Yorkshire Dales karst. 22), lacking rounded clints, towers, wide grikes or in addition to being at higher altitudes the most grike flares. it is worth bearing in mind linton’s rounded features, with widely flaring grikes are in views on nearby pennine tors in gritstone: “… the massive, thick limestones and usually in sheltered present landscape still owes not only its main out- locations. lines but also much of its small scale relief to proc- locations include around edges of large sur- esses that were operating before and between the face karst depressions east of clapham High Mark, glacial episodes” (linton, 1964). The observation near Back pasture, and around lea gate High applies to neighbouring limestone features as well. Mark (Figures 21, 23), a situation similar to Far- another simple comparison, of nW england leton Knott. Many factors favour these forms hav- and Derbyshire, also highlights the idea that ing survived severe ice scour in the last glaciation. many features in glaciated areas might be pre- There is also evidence from cave studies in the area Devensian or earlier in origin. Derbyshire was not 287 KRF•1 • OK.indd 287 15.12.2009 10:47:06 Karst Rock Features • Karren Sculpturing ƒ ƒ  ­    ‚ € Figure 23: Malham-Littondale high country, indicating locations of mature surface outcrops (based on a map by M. E. Marker, 2003). Large, bold m. mature rounded features; CG. Cote Gil ; H1. Clapham High Mark; H2. Malham Lea Gate; J. Back Pasture; K. High Mark; L. Monk’s Path; O. Wharfedale; HS. High Sleets; DB. Dowkabottom. Highlighted contour is 450 m; NCF and MCF. North and Mid Craven Faults. glaciated in the late Devensian (Waltham et al., their surface landforms by human activities. This 1997), yet outcrops of pavement exist, regarded is discussed at length for various sites (goldie, as palaeokarstic in origin (Walkden, 1972), with 1986, 1993), particularly in nW england, but also characteristics similar to those on outcrops in in Wales, scotland and ireland. two areas consid- many glaciated parts of nW england (Figure 24). ered here demonstrate the effects on grike charac- Their similarity supports the idea that nW eng- teristics and patterns: Wharfedale, Yorkshire, and land has small glacial survival landforms in lime- Hampsfield Fell, cumbria. stone outcrops. Wharfedale is a well-settled valley, where stone has been taken off the land for building and other purposes for millenia. limestone has been burnt Human activities for lime to add to soil and this process has seen much surface limestone removal in past centuries Many limestone outcrop areas of the British isles to be burnt in kilns whose remains dot the land- have suffered direct damage and alteration of scape. More recently (20th century mainly) remov- 288 KRF•1 • OK.indd 288 15.12.2009 10:47:07 Helen S. Goldie, Case studies of grikes in the British Isles Figure 24: Rounded pavement edge near Blah (Derby- Figure 25: Rounded pavement edge north of Grassing- shire). Width of view is 5.5 m, below. ton (Wharfedale). al of solutionally shaped clints for decorative use placed clints in heaps, much reduced grikes, and has had a damaging effect on the appearance of larger clints without runnelling are among the ef- many valley side pavement outcrops. also, stone fects of relatively recent 20th century removal. removal for walling has occurred, most intensely in the late 18th and early 19th century enclosure period. all these activities have left limestone Conclusions outcrops with various ages of artificially affected forms. The morphometric data for pavements all these cases involve the landform assemblage sampled here show moderately shallow grikes, known as limestone pavement, but their features most likely due to these activities (see chapter vary enormously for many reasons. The grikes 9, table 1). runnelling also seems immature on range from immature slits in surfaces almost as apparently suitable outcrops, for example, near freshly-scoured as those of surfaces near new- grass Wood, which is consistent with older clint ly-retreated glaciers, to well-rounded limestone removal. However, there is also very massive ma- blocks left between widening grikes more akin to ture outcrop that has survived damage (Figure 25) features observed in semi-arid or mediterranean and probably also glacial scour. areas not affected by any Quaternary glacial ero- sion. The fact that many clints in these areas are so well-eroded that the grike flares merge at the Table 1: Clint data at Newbiggin Crags (metres). top of the clint begs the question of where the Length Width Height/GD grike finishes. The easy answer is when grikes are Upper bed (I I) 1.6 1.2 1.5 truncated by glacial scour or human action, and Lower bed (IV) 2.8 1.6 1.5 have an approximately right-angled and or abrupt change of slope at their top, but when there is a gradual grike flare it is not obvious where the Hampsfield Fell west of Morecambe Bay is dis- turning point between features comes. Degree of tinguished for being subject to the first limestone roundedness is one measure that could be charac- pavement protection order (lpo) (goldie, 1986, terized in this situation to put some sort of quanti- 1993) placed on it after extensive removal of clint tative index on the features in order to distinguish tops for garden rockery stone. a freshly rough- them from site to site. ened surface with much small sugary debris, dis- obvious visual weathering differences can be 289 KRF•1 • OK.indd 289 15.12.2009 10:47:09 Karst Rock Features • Karren Sculpturing seen between the top limestone layers of different plain detailed differences in the landforms. litho- sites. at the shilin stone Forest in china these are logical variation can bring in the importance of very sharply runnelled (Waltham, in gunn, 2004). non-solutional processes, and tectonic influences el torcal in spain is moderately runnelled. How- introduce such factors as uplift with some areas ever, many of the cases here are smooth, rounded experiencing uplift that helps to sustain lengthy and less runnelled than in the aforementioned periods of karstification by raising base-level. areas. These different sites have basic features in Fundamentally, however, these varied landforms common in that they have developed their nega- would not exist the way they are without a suit- tive landforms, the grikes, along weaknesses in able level of fissuring in the respective limestones, the limestones, and local differences in erosional fissuring which is a basic condition for karst de- conditions, both now and in the past, help to ex- velopment. 290 KRF•1 • OK.indd 290 15.12.2009 10:47:09 The Karrenfields of The MuoTa Valley: 23 Type localiTies of The Main Karren Types afTer The noMenclaTure by alfred bögli Michel MONBARON and Andres WILDBERGER The famous karst specialist Alfred Bögli studied 1961). The northern part of the valley consists intensively the area of Silberen-Charetalp-Mären of the Drusberg nappe (northern Helvetic). The (canton of Schwyz, Switzerland) (Figure 1), his fa- more interesting southern part is shaped within vourite terrain for surface geomorphological re- the strata of the Axen nappe s.l. (middle Helvetic). search. Working on these carbonate rocks of the The Jurassic strata of the latter are left in the south, Helvetic nappes, he classified and defined a no- the Cretaceous and Tertiary formations are accu- menclature, still used nowadays (Ginés, 2004), in mulated by thrusting movements in the northern order to describe the different karrenforms of the part of the Axen nappe s.l. In this pile of nappes “haute montagne calcaire” (a term used by Maire, and slices, the karstified limestones of the different 1990). Between 1951 and 1987 Bögli published ex- nappes are in contact with each other and enlarge tensively on this topic, with some of his publica- the thickness of the karstified series. Therefore, a tions still being very influential. We chose to illus- good prerequisite is given for the existence of long trate some of the most characteristic forms of this and deep caves (Hölloch: 194.5 km long, 939 m karstic landscape, paying tribute to Alfred Bögli deep; Silberen system: 37.8 km long, 888 m deep; and remembering his very useful contributions to state 2008). the field of karst research. The main karstified limestone formations in the Axen nappe are the Quinten limestone (Malm, Upper Jurassic), the Urgonian limestone (Bar- Geological and climatic introduction remian-Aptian, Lower Cretaceous) and the See- wen limestone (Cenomanian-Coniacian, Upper Some of the most extended bare karrenfields Cretaceous). In addition, other carbonate forma- (“Karrenfeld”, a german word meaning ground tions of lesser importance in the context of karsti- with karst tracks) of Switzerland (about 50 km2) fication occur in this area. The entire stratigraphic (Bögli, 1987) are located in the southern part of column covers strata of Triassic to Palaeogene the Muota valley. Within this large surface area (besides more or less pure limestones, sandy and several types of karren forms can be found and siliceous limestones, marls, silty marls, dolo- and many of Bögli’s published examples originate sandstones are present). from this region (Bögli, 1951‒1976). The mean annual precipitation lies between The whole valley of the Muota river (Figure 2) 2,000 mm in valleys and 3,000 mm on summits tectonically belongs to the Helvetic realm (Hantke, (Kirchhofer and Sevruk, 1991). The correspond- 291 KRF•2 • OK.indd 291 15.12.2009 10:56:54 Karst Rock Features • Karren Sculpturing 705000 710000 715000 Figure 1: Geographical lo- cation of the study area. Switzerland Fulberg 000012 cross section S i l b e r n 000502 Pfannenstock M 00 u 000 o 2 ta C h a r e t a l p rive cross section r 00 Mär en 0 591 Kilometers 0 1 2 3 4 5km 0 0.5 1 2 3 4 ing mean annual temperatures range between 8°C scribed and illustrated in turn below. They are (valley) and -1°C (summits) (Kirchhofer, 1982). A classified by their size and the lithology of the significant part of the precipitations is snow which host rock. Most of these forms are contained in melts again during spring and summer. the Urgonian limestone formation which is 180 m thick and covers the Silberen plateau (Figure 3). The Quinten limestone formation (400 m thick) Nomenclature of the main forms is well represented in the Mären area and has se- veral karstic features, too, which occur less fre- The karren are the most developed forms of the quently in other carbonate formations. Some of exokarst in this part of the Helvetic Alps. Table the most spectacular forms of these two areas will 1 lists the main karrenforms which will be de- be shown below. 292 KRF•2 • OK.indd 292 15.12.2009 10:56:55 Michel Monbaron and Andres Wildberger, The karrenfields of the Muota val ey: type localities of the main karren types … The big karrenfields of the Muotatal mit with an altitude of 2,319 m a.s.l. During the area Pleistocene glaciations this top was covered by a glacier which flowed off on all sides of the plateau. Silberen karrenfield The glacier removed loose rock material and left the ground almost bare with a humocky relief (in The Silberen (the bare, bright grey rocks give the French “roches moutonnées”) occurring occasion- “silvery mountain” its name) is a flat-topped sum- ally. It is now sculpted by Holocene pavements. €ƒ   ƒ † ­  ­ „ € ­ ­ ‚ Figure 2: Geological cross-section. Figure 3: Karrenfield on Urgonian limestone, Silberen plateau. The intensive faulting causes a jig-saw puzzle of karstic (Urgonian and Seewen limestones), semi-karstic (Garschel a-fm. p.p.) and non- karstic rocks (Garschel a-fm. p.p., Tertiary). 293 KRF•2 • OK.indd 293 15.12.2009 10:56:56 Karst Rock Features • Karren Sculpturing Table 1: Different types of exokarst features in the Muotatal area. Macro-scale Meso- to micro-scale types m] unnels lutes arren rikes unnels its hickness [ chichtrippenkarst valas arrentables ndering r eelsteps all k Formation Lithology outonnées Solution f ea ative t Dolines / U Corrosion p Roches m Kluftkarren / G Rinnenkarren / R Trittkarren / H andkarren / W Approxim Karrentische / K Rilenkarren / W Schichttreppenkarst / S Mäanderkarren / M Fossiliferous micrite with some argil ous Seewen limestone* 50 x x x – x   x x x – flakes, strata often indistinctly developed Brisi limestone* (Member of the Garschel a forma- Biosparite with quartz–sand (up to 20 %) 5 – – x – x   – – – – tion) Mainly biosparites and biomicrites, mas- Urgonian limestone* sive limestone or strata often indistinctly 180   x x        developed Mainly fossiliferous micrite, strata distinctly Quinten limestone** developed (in the range of decimetres up to 400        x x x  1 metre) * Silberen’s karrenfield (1,800 to 2,300 m a.s.l.) ** Mären’s karrenfield (2,200 to 2,400 m a.s.l.) Relative frequency: – type not known  intermediate x rare / indistinctly developed  frequent The conspicuous karrenfields are developed main- rock material and climate patterns. For the karst ly in Urgonian and Seewen limestone at an alti- of the Muota valley this means, for example, that tude between 1,800 and 2,300 m a.s.l. rinnenkarren on Urgonian limestone are rare The karren types are determined by the parent above 2,000 m a.s.l. but rather widespread below this altitude; just as ril enkarren are rare on See- wen limestone but frequent on Urgonian lime- stone (Table 1). Below, about 1,800 m a.s.l., rocks are incom- pletely covered by soil and vegetation, mainly for- est. Within this “green karst” the bare limestone is visible only in patches and its karren types are transformed or even absent. Mäander- and wandkarren are subtypes of rinnenkarren: mäanderkarren (Figure 4) are present on flat rocks (slope angles less than about 20°) whereas wandkarren can be found on very steep slopes (more than about 70°). Figure 4: Mäanderkarren (meandering karren) on a small A most typical and frequent microform of this slope surface, Urgonian limestone, Silberen plateau. region, the rillenkarren, can be observed on the Width of view is 80 cm. sparitic Urgonian limestones of the Silberen kar- 294 KRF•2 • OK.indd 294 15.12.2009 10:56:57 Michel Monbaron and Andres Wildberger, The karrenfields of the Muota val ey: type localities of the main karren types … Figure 5: Smal -scale rillenkarren on crests between medium-scale kluftkarren features (Silberen plateau). Width of view is 90 cm, in the middle. renfield. These karren decorate the crests separat- ing the grikes and solution runnels (Figure 5). The rillenkarren are more or less perpendicular to the runnels. Figure 6: Corrosion pits and rillenkarren on Urgonian Corrosion pits are small features spreading on limestone, Silberen. In some pits the red-brown Urgonian limestone of the Silberen region, and al- minerals limonite / goethite / haematite are still though not mentioned by Bögli, they are present- present; in other pits they are removed by rain water ed here due to their rather frequent occurrence. (compare empty pits in the right side of the crest). Width of view is 35 cm, below. The process causing these pits is the oxidation of pyrite in contact with air and water from precipi- tation. This chemical reaction results in sulphu- Pleistocene glaciers. The polished glacial marks ric acid and minerals like limonite, goethite and (Figure 7) of the last glaciation (Würm) are still haematite. The acid dissolves the limestone locally recognizable and soil formations are thin and and leaves behind a hemispherical pit with a di- patchy. Various forms of karren which have been ameter of several centimetres. The pits are filled influenced by the main tectonic structures, bank- with iron oxides or stay empty if washed out (Fig- ing of rocks and the distinct ruggedness of the ure 6). In cave environments a reaction between massif can be found on the Mären plateau. Two sulphuric acid and limestone acts as the source for typical examples are described below. gypsum crystals (Bögli, 1972). The frontal backward erosion of limestone lay- ers due to frost attack (Aubert, 1969) or glacier erosion have resulted in a staircase-like arrange- The Mären karrenfield ment ( schichttreppenkarst) (Figures 7 and 8). The Mären plateau is mainly characterized by this The karren of the Mären plateau are located be- particular karstic morphology. A similar form tween 2,200 m and 2,400 m a.s.l. defined by Bögli (1964), the schichtrippenkarst, is Essentially, karren landforms are developed in less frequent. the Quinten limestone formation (Malm). As in Smaller typical forms common in the Mären the Silberen plateau, the area was once covered by area illustrate the powerful dissolution of the 295 KRF•2 • OK.indd 295 15.12.2009 10:56:58 Karst Rock Features • Karren Sculpturing Figure 7: Witnesses of gla- cier-polished rocks in the Mären area. Figure 8: Schichttrep- penkarst near the small valley of Bockalpeli, at the western border of the Mären plateau. run-off water on the pure micritic limestones of constantly being enlarged by solution processes Quinten. resulting in kluftkarren or grikes which underline Tectonic fractures can be well observed on the even more the diversity of different fracture fami- strata surfaces and, due to the present network lies (Figure 9). of these fractures, conclusions about the tecto- The second most frequent forms are the rinnen- nic stress directions can be made. The cracks are karren, created by runoff on the top and slopes of the 296 KRF•2 • OK.indd 296 15.12.2009 10:56:59 Michel Monbaron and Andres Wildberger, The karrenfields of the Muota val ey: type localities of the main karren types … Figure 9: Kluftkarren on Quinten limestone, Mären pla- Figure 10: Rinnenkarren features, at the northern border teau. of the Mären plateau. limestone beds (Figure 10). The result is a network of treat of the local glacier – the Tardiglacial, which channels running parallel at the steepest slope. happened nearly 12,000 years BP. Boulders, left Both solution forms kluftkarren and rinnen- behind over the limestone layers during the last karren are very common and occur at different spa- glacial retreat, have preserved the rock surface tial scales, that is approximately 10-1 to 30 metres. against solution by rain and snow melt water since By means of analysing the karren tables ( Kar- then. The height of the limestone pedestal under rentische) (Figure 11) it is possible to determine the boulders represents the average thickness of the dissolution rates of the carbonates during the dissolved limestone removed during the last the whole post-glacial period since the last re- twelve thousand years. Figure 11: Karren tables or Karrentische with pedes- tal, Mären plateau. Width of view is 15 m in the middle. 297 KRF•2 • OK.indd 297 15.12.2009 10:57:00 Karst Rock Features • Karren Sculpturing Figure 12: Boulder chaos resulting from the dis- location of the staircase front. Finally, effects of cryoclasty on the limestone can serve as an indicator for unresolved problems could be observed on the bare plateau. Gelifrac- as well as a driver for future researches. tion has dislocated the blocks, which are piled up Bögli’s natural laboratory was situated not far in a rocky chaos (Figure 12). from his home, near Lucerne in Central Switzer- Dolines and uvalas are less frequent in the land. He was also interested in the phenomenon Muota region but occur mainly in zones with of endokarst, and hence it is not surprising that loose material on the top of karstified rocks (suf- he was also engaged in the exploration of the Höl- fosion dolines). loch Cave in near Silberen, one of the longest cave systems in the world (Bögli, 1970, 1980). Conclusions Acknowledgements Alfred Bögli studied and defined typical and wide- spread karren forms of the alpine karst between the We would like to thank Beat Niederberger for upper forest line and the periglacial zone. He tried contributing to the overview map and geological to explain the different karren types as the result cross-section, and Ingo Heinrich for translating of the different stages of the corrosion strength of the manuscript. water. This attempt, although not fully successful, 298 KRF•2 • OK.indd 298 15.12.2009 10:57:01 The naTure of liMesTone paVeMenTs 24 in The cenTral parT of The souThern Kanin plaTeau (KaninsKi podi), WesTern Julian alps Jurij KUNAVER The aim of this paper is to introduce one of the kettle, Schacht-doline, Kessel-, Karrendoline, puit most significant characteristics of the Kanin Mo- à neige), normally associated with them, belong untains, the limestone pavements, found in a va- therefore to the most important characteristics riety of forms, according to their origin and evo- of the mountain karst surface (Haserodt, 1965; lution. They reflect the relations between the ge- Maire, 1990; Ford and Williams, 1989). For this ological structures and the surface as well as the reason this phenomenon seems to be interesting evolution of the mountain karst surface in late and justified. Pleistocene and Holocene epoch. Strong traces of The wider area was partly affected by human glaciation can be found there, as well as an abun- impact, e.g. by grazing on the mountain pastures, dance of corrosional forms as the consequences which is now in strong decline. A great number of large amounts of rain and snow (3,400 mm per of human traces date from World War I because year); not to speak of Holocene evolution of ve- of the Austro-Italian front (1916‒1918) and, from getation and soil that also influenced the deve- 1975 on, due to the development of alpine skiing. lopment of pavements, at least in lower altitudes. Human impacts may have also caused the lower- The pavement character of the Kanin plateau and ing of the forest line in last centuries. the very common kotlich ( kotlič in Slovene, snow The Kanin Mountains in NW Slovenia (Figure   Figure 1: Geographical location of the Kanin Mts. 299 KRF•2 • OK.indd 299 15.12.2009 10:57:04 Karst Rock Features • Karren Sculpturing 1) have become one of the karstologically most ern slopes of the Kanin Mts have had a normal investigated high karst mountain range of the Ju- stratigraphic evolution from the Upper Trias to lian Alps in the last ten to fifteen years, not only the Cretaceous period, following the former view because of a rich variety of surface karst phenom- of Kossmat. The bedrock of the mountain range ena but also because of 12 deep shafts with depths is composed of massive dolomite (Hauptdolomit) from 500 to over 1,500 m and about 100 km of of light grey to white colour, which changes up- underground channels explored so far. One of the wards into a stratified coarse-grained to micrite shafts, Vrtiglavica (–643 m), has the world record dolomite. Such dolomite is found at the bottom of of a continuous vertical shaft. The investigations the Možnica and the Krnica valleys. The visible of karst geomorphology and hydrology of the thickness of the dolomite in the Možnica valley is mountains are in full progress (Audra, 2000; Ku- 400 m, and the thickness of Dachstein limestone naver, 1998; Komac, 2001). 1,000‒1,200 m, 200 metres of which are made of Strong karst springs at the foot of the mountain massive limestone (Buser, 1976). Some areas along range are the result of abundant precipitation and the fault lines, in a width of some metres, are dolo- asymmetry in the underground water discharge mitized too. between the northern and southern sides of the The dolomite gradually changes upwards into range. The hypothesis that this mountain range is Dachstein limestone. The limestone is stratified one of the most characteristic areas of high moun- and has a typical Loferitic development with a tain karst in the Southern Limestone Alps with layer᾽s width of 0.2–2 m, rarely up to 10 m. The extremely interesting surface and subterranean higher lying strata of Dachstein limestone are karst phenomena was verified by geomorphologi- formed largely of micrite limestone, while the cal and speleological research results (Kunaver, highest parts of this sequence are composed of 1973a, b, 1983, 1984, 1991; Pirnat, 2002). massive micrite limestone. Characteristic of this limestone are very clearly visible stromatolitic lay- ers on the contacts of strata, which contain a high General geology and geomorphology proportion of dolomite, being the result of an of the area early diagenetic dolomitization (Ogorelec, 1996). Stromatolitic layers contribute a lot to the strong The area is extremely rich in pavement surfac- mechanical disintegration and development of es also, compared to some other areas of moun- undercuttings and half caves in the subnival zone tain karst in the Alps, due to the congruence of in the Kanin Mts. The area of dipped slopes in the the bedding planes and the inclination of the southern side of the border ridge between Pre- surface (a dip slope), both facing south or south- streljenik and Visoki Kanin, just along the moun- east. This is the basic, distinctive characteristic of tain path, is typical for excellent examples of this the Kanin Mts. The southern slopes of the Kanin sort of mechanical disintegration along the stro- Mts are therefore one of the best examples of re- matolitic layers. lief conformity and of structural landforms on a An important characteristic of Dachstein lime- large scale in the Julian Alps. The central part of stone is also the infillings of syngenetic corro- the mountains of Dachstein in the Northern Aus- sion hollows and joints with younger, presumably trian limestone Alps, for example, is for the most Jurassic material, which additionally makes this part completely different due to the total discrep- limestone similar to that in the Northern Aus- ancy between the surface, that slopes towards the trian Limestone Alps. In the southern slopes of north, and the prevailing incidence of strata that the Kanin Mts, Buser confirmed and determined is facing south. the extent of Jurassic and Cretaceous limestones According to Buser (1976, 1986a, b), the south- and of other rocks of that age, that are of consider- 300 KRF•2 • OK.indd 300 15.12.2009 10:57:04 Jurij Kunaver, The nature of limestone pavements in the central part of the southern Kanin plateau (Kaninski podi), Western Julian Alps able importance for the verification of a synclinal m structure of this area. The evolution of these lay- kluftkarren 1750 karrenbrunnen ers varies highly, but they are not very extensive potholes rinnenkarren if compared with the Upper Triassic strata (Buser, 1700 kettle kamenitzas 1976). In Buser᾽s view, the Bovec basin is the bot- glacial till tom of a large syncline, filled with Cretaceous flysch and hence the Kanin Mts are the northern 1650 limb of it. The dip of strata is south-westwards in the western part of the range, eastwards in the 1600 eastern part and southwards on Mt. Rombon. This is linked with a bowl-like shape of the syn- 1550 cline. The anticline was later often discontinued 20 60 100 200 300 m as a result of subsequent faulting and thus its axis Figure 2: Vertical distribution of karst landforms on in- (east–west) was horizontally shifted. clined glacial y abraded limestone pavement on a Besides lithology, fault lines, running either lower edge of the plateau of Kanin, according the re- parallel or transversal to the range, also affected cession of glacial till in the upper part and concentra- tion of the precipitation water and snow in the lower the evolution and the present-day shape of land- part, 1,800 metres a.s.l. (after J. Kunaver, 1983). forms and are therefore of great importance. Most often they run in the north, northeast and north- west direction. To the south of the main W–E ridge with the depend on the fault line system, but are connected highest peak of High Kanin (2,585 metres a.s.l.), with dipped slope strata and are of erosional ori- which divides the northern Italian side from the gin, like other similar features. Their orientation southern, Slovenian side, there is the main part of is largely in agreement with the dip of Dachstein the plateau of Kaninski podi, lying about 200‒300 limestone strata on the slopes. Evidently the flow m below the ridge, with an area of nearly 10 km2 of ice had the greatest modifying effect on slopes, that represents the most massive part of the moun- especially in their upper parts, also because of tain range. From here the plateau sweeps down in reduced space in comparison with a broad accu- steps from the more gentle upper parts below the mulation area on the plateau. Thus we can sup- ridge (between about 2,300 and 2,000 m a.s.l.) to port Linton’s claim that the present-day land sur- the lower edge of the plateau (between 2,000 and face was reduced to a lower elevation than that of 1,800 m a.s.l.). From there downwards, the incli- the former weathered surface owing to preglacial nation of slopes increases to about 22°. The slopes weathering of bedrock during the ice ages (Linton, are discontinued by smaller structural or erosion- 1963). Skednji could also be called divides, which al scarps, about six in all. On some of them glacial corresponds with the term, invented by the same till was deposited and preserved as patches, which author (Figure 3). influenced the distribution of surface karst forms The extensive central part of the southern (Figure 2). Kanin high karst plateau or Kaninski podi (in The slopes are enclosed by narrow side ridges, Slovene) is not only characterized by extremely called skedenj ( sing.), skednji (plur.) (a local Slo- well developed limestone pavements with many vene term for the barn or better, for the hay-rack), high alpine glaciokarst landforms, but also with extending far down the slopes. According to our many caves and shafts. The plateau Kaninski podi investigations, these distinct and slim ridges, re- is characterized also by some long dry valleys sembling giant rock walls, with up to 300 m of rel- of polygenetic origin, which start on the highest ative height and often less than 100 m wide, do not part of the plateau and end at an altitude of about 301 KRF•2 • OK.indd 301 15.12.2009 10:57:05 Karst Rock Features • Karren Sculpturing Figure 3: Mali Skedenj, the erosional remnant on the southern Kanin dip slopes of presumably preglacial relief (1,999 m a.s.l.). 1,800 m a.s.l., being more modified by glacial ero- the higher places, and so are limestone pave- sion in the lower than in the higher parts. In the ments that occur in different combinations (Fig- central part of the plateau they are over-deepened ure 4), one above the other like roofing tiles or fish by greater karst depressions like Veliki Dol, which scales. The dip of strata of the plateau Kaninski exceed 700 x 450 m in size and is up to 80 m deep. podi is mostly between 15° and 22°. In some plac- The original valley depressions could have de- es the dip increases up to 28° or even 30°. The low- veloped as a result of the older fluvial processes est dips were measured in the western part of the but their present feature is more likely to be only plateau, between Mali Skedenj and Veliki Skedenj, a morphological inheritance, as they were influ- particularly below Veliki Skedenj, where the dip enced by many phases of non-fluvial development was as low as 10°, and in some places even lower. It in Quaternary. appears that the strata inclination slowly decreas- es in the direction towards the border ridge be- cause of the vicinity of an anticline crest. Hori- The Kanin pattern of limestone zontal strata and corresponding limestone pave- pavement areas – the interrelation of ments are rarely to be found in the area, except on limestone beds, topographic surface the plateau Goričica. and glacial abrasion In the Julian Alps, the structural relief of lime- stone pavements primarily occurs in bedrock As pointed out earlier, the geological strata of the made up of Dachstein limestone. Evidence of this Kanin Mts are not only more or less steeply in- claim is provided by place names that contain the clined towards the Bovec basin but also all over word “lašt” in different combinations, the Slovene 302 KRF•2 • OK.indd 302 15.12.2009 10:57:08 Jurij Kunaver, The nature of limestone pavements in the central part of the southern Kanin plateau (Kaninski podi), Western Julian Alps Figure 4: The system of limestone pavements of Kaninski podi: I. the even flat and sloping limestone pavements; II. the flat and sloping cuesta-like limestone pavements; II . the flat and sloping stepped limestone pavements; IV. the flat and sloping inverse limestone pavements. term for limestone pavement (Gams et al., 1973). It ments are found both at the bottom of dry valleys is too early to say anything about the resemblance and some larger depressions, and at the top of ris- between the Slovene “lašt, lašti” (lasht, plur. lash- ing ground. The first outcrop from the top is the ti) and “lastra” , a name used for the limestone upper plateau just below the main mountain crest, pavements in the area of Dolomites in north Italy. or the highest peak Visoki Kanin, 2,587 m a.s.l. It is followed by the ridge between Visoki and Nizki Talir and both shallow valley depressions Zadnji The location and the system of Dol and Dol Za Mostmi. Extreme pavement areas limestone pavements are Zgornja and Spodnja Osojnica near the moun- tain hut. South from Veliki Dol there is the next In the central area of the Kanin plateau the pave- vast area between Gnila Glava and Mali Dol on 303 KRF•2 • OK.indd 303 15.12.2009 10:57:09 Karst Rock Features • Karren Sculpturing Figure 5: The even sloping pavements in the upper part of Kaninski podi (2,300 m a.s.l.). Figure 6: The even sloping pavements in the central part of Kaninski podi with kotliči (2,050 m a.s.l.). 304 KRF•2 • OK.indd 304 15.12.2009 10:57:12 Jurij Kunaver, The nature of limestone pavements in the central part of the southern Kanin plateau (Kaninski podi), Western Julian Alps one side, and the ridge Konjc on the other. Small- tributions have become a good starting point for er and secluded areas of sloping or inclined pave- further classifications of pavements. Taking the ments are situated to the east of the ridge Veliki Kanin Mts as an example, we have defined eleven and Mali Babanski Skedenj, for instance Hudi Lašt, kinds of pavements, divided into four, respective- in the upper part of Razor and in the area of Skripi ly in two groups, the ones predominantly found under the upper station of the Kanin cable rail- on flat surfaces and those found on sloping or in- way. There are relatively few pavements in the west clined surfaces (Kunaver, 1983). of Veliki and Mali Babanski Skedenj. Prestreljenik For better differentiation of pavements we sug- plateau has a fairly pavement-like surface, but it gest the following classification, which uses the is anthropogenically quite changed. There are degree of congruence between the inclination of pavements in the pass side of Prestreljenik sad- the surface and the position of the layers as a crite- dle. Pavements on Goričica are not mentioned in rion. The first group includes pavements with the this paper, although they also express a very good highest degree of congruence between the surface development. In general, it may be claimed that and geological structure or flat layers of lime- such a landform characterizes more than half of stone, among others flat even pavements, which the plateau area. could also be called stratified plates, and have The pavements on the Kanin plateau and in the the simplest form, like Williams’ idealized pave- Kanin Mountains in general differ according to: ments. They can be either entirely flat or more or 1) different relation between the incidence of the less inclined (Figure 4.I.1, I.2, I.3). It is difficult to layers and the inclination of the surface, 2) dif- define the border between the more and less in- ferent assemblages of surface karst forms, 3) dif- clined pavements, but it has been decided that the ferent height of pavements above sea level, and 4) boundary-line should be the inclination lower or different size of the pavements. The height above higher than 10°. Williams believes that the upper sea level has been crucial for the intensity of ap- limit to find pavements is 45°, but it is hard to pearing of karst phenomena. The tendency to- agree with the idea that plates with steeper incli- wards the formation of limestone pavements is nation are not possible or they are not considered found in all places where the strata are orientated pavements any more. It is true, however, that there more or less in the same way as the movement of are fewer corrosion forms if the inclination of the glacial masses. The strata in the Kanin Mts, how- surface is steeper (Figures 5, 6). ever, only rarely lie in the opposite direction, as do Another form is the so-called cuesta-like pave- those on the lower sides of larger depressions. ment ( Schichtrippenkars t, according to Bögli), which appeared as a result of discordance be- tween the inclinations of the surface and the lay- Basic classification of pavements ers. An average flat surface can be very configured with inclined conformable pavements of different The starting point for a limestone pavement sys- length and shorter unconformable surface, if the tem in the Kanin Mts was provided by Bögli layers are inclined for over 10° and more (Figure (1964) and Williams (1966). The latter distin- 4.II.1). Kotliči or Schachtdolines are very common guished between the three most frequent kinds phenomena on such a surface that can be often of pavement, the idealized or flat even pavement, found in the area of the upper Kanin plateau. They the inclined pavement and the stepped pavement can be found almost under every short inclined (after Sweeting, 1973). At the same time, A. Bögli pavement, since there is usually plenty of snow distinguished only between Schichttreppen- and that stays on the surface, while snow-water flows Schicht rippenkarst, which means stepped and towards the foot of the pavement. Theoretically cuesta-like pavements (Bögli, 1978). Both con- both surfaces become the same if the inclination 305 KRF•2 • OK.indd 305 15.12.2009 10:57:12 Karst Rock Features • Karren Sculpturing Figure 7: Short cuesta-like pavements on the flat surface of uppermost Kaninski podi plateau with the entrances of kotliči and shafts (2,250 m a.s.l.). of the layer is below 45°, which usually does not side of the layer inverted upwards. Holocene cor- happen on Kanin (Figures 7, 8). rosion has already lowered the original surface, The cuesta-like pavements are very common but the glacial round shaping is still clearly visible on slopes, normally when the inclination of the (Figure 9). surface is smaller than the inclination of the lay- The third form of pavements is the stepped pave- ers. This is typical of the lower and middle parts ment, found in places where the surface is steeper of the Kanin plateau. Pavements of this kind can than the inclination of the layers. The first exam- also be found on steeper slopes, but only in case of ple is a combination of more or less flat layers and the above-mentioned relation between the surface an inclined surface, which is relatively rare in the and the layers. Beside the cuesta-like pavements Kanin Mts (Figure 4.III.1). More frequent forms on flat areas we distinguish more gently sloping are the stepped pavements on steeper bends and and steeper cuesta-like pavements (Figure 4.II.2, lower steeper parts of the plateaus, where the sur- II.3). A very strong and clear glacial grinding is face is steeper than the layers. Beside the above- typical of the cuesta-like pavements with the front mentioned stepped pavements we know some 306 KRF•2 • OK.indd 306 15.12.2009 10:57:14 Jurij Kunaver, The nature of limestone pavements in the central part of the southern Kanin plateau (Kaninski podi), Western Julian Alps Figure 8: The irregularly shaped cuesta-like pavements in the central part of Kaninski podi plateau (2,000 m a.s.l.). Figure 9: Glacial y abraded upper scarp edge on a limestone pavement (2,300 m a.s.l.). Width of view is 6 m, in the middle. 307 KRF•2 • OK.indd 307 15.12.2009 10:57:18 Karst Rock Features • Karren Sculpturing Figure 10: The stepped limestone pavement on Gorenja Osojnica (2,250 m a.s.l.). Figure 11: The bottom of initial valley depression of rectangular shape with pavements and kluftkarren, Veliki Graben (2,000 m a.s.l.). 308 KRF•2 • OK.indd 308 15.12.2009 10:57:21 Jurij Kunaver, The nature of limestone pavements in the central part of the southern Kanin plateau (Kaninski podi), Western Julian Alps gentle as well as steep stepped pavements (Figures 4.III.2, III.3; 10, 11). Special attention should be paid to cases, which are quite common, where the orientation of prev- alent gradient of relief and that of the dip of strata differs up to 90°. The maximum difference of up to 180° leads to the formation of limestone pave- ments which are inclined towards the slope or are unconformable. This fourth form is the inverse or unconformable pavement, which are not so fre- quent on the Kanin plateau (Figure 4.IV.1, IV.2). Its presence depends on local topography. Dimensions and character of the pavements Figure 12: The lower part of a limestone pavement with The height and frequency of the steps of strata and Trümmerkarren (2,100 m a.s.l.). the size of limestone pavements are an external sign of the extent of relief and structure conform- ity. On the flat land surface ice moved along the strike, plucking away less resistant upper strata from the surface and the cuesta-like limestone pavements were formed. The direction and the dip of strata in relationship to the direction of ice movement and to the general tendency of relief inclination determine whether limestone pave- ments are less extensive and steep with high steps of strata or they are wider and longer and steps of the strata between them are lower or even smooth and flat. The relationship between local relief, surface inclination, and strata orientation is shown by the shape of limestone pavements. Quite com- mon are those with a straight line on the lower side, i.e. along the step of strata, while the upper Figure 13: The lower part of a limestone pavement with or exterior side is due to glacial scour of irregu- Hohlkarren (the northern Italian side of Kanin Mts, 1,850 lar shape. Limestone pavements are often wider m a.s.l.). Width of view is 1.5 m, below. and glaciated particularly on the upper side, while they narrow downwards in the shape of a triangle (Figure 8). phenomenon. The nature of karren depends most The surface corrosion forms on pavements does often either on the nature of the rock or on the not differ much from the cases described in the lit- altitude of the location (i.e. the influence of veg- erature. They are mostly of younger, Holocene age. etation and soil) (Figures 12, 13). Also older land- The napfkarren or karrenfussnäpfe are a common forms are present, e.g. the rounded karren and tu- 309 KRF•2 • OK.indd 309 15.12.2009 10:57:25 Karst Rock Features • Karren Sculpturing bular pipes, which could only be interpreted as the deeper parts of former karren which were eroded by ice (Figure 14). Also many kotliči, one of predominant surface karst forms of medium size, are mostly of younger origin, although their development (of the biggest ones) has presumably begun even earlier. In par- ticular the fossil kotliči, which are not so rare espe- cially in the area of limestone pavements in con- sideration, and which contain morainic material, can be reliably claimed as older than the last ice age. The morphogenetic independence and fun- damental characteristics of the process and con- ditions in which kotliči are formed were reported by previous studies of this karst form. Their link with thickly stratified Triassic limestones and with weak lines in rocks (Kunaver, 1976) is their most typical feature (Figure 15). In terms of intensity of glacial erosion it is pos- sible to recognize a rather strong influence on the surface morphology of the Kanin Mts, as already mentioned. In general, the smaller abrasion land- Figure 14: Rounded opening of a glacial y eroded karren; forms were denuded a great deal by the Holocene central Kaninski podi plateau (2,050 m a.s.l.). Width of karst erosion lowering. Therefore the bare rock view is 75 cm. surface, which clearly looks glacially abraded, e.g. on the scarp edges of limestone strata, looking towards the direction of glacial flow, was in fact Figure 15: View along the even sloping pavements in the central part of Kaninski podi plateau, with entrances of kotliči (2,050 m a.s.l.). Width of view is 7 m, in the middle. 310 KRF•2 • OK.indd 310 15.12.2009 10:57:29 Jurij Kunaver, The nature of limestone pavements in the central part of the southern Kanin plateau (Kaninski podi), Western Julian Alps Figure 16: The fossil erosional groove or notch as the effect of the former subglacial water erosion. Lower part of the Kanin- ski podi plateau (1,900 m a.s.l.). Figure 17: Karrenfield in the northern upper part of Kaninski podi plateau, presumably the remain- der of an older karsti- fication because of a weaker glacial erosion in that part (2,250 m a.s.l.). Width of view is 15 m, in the middle. intensively lowered (at least 20‒30 cm), but still ated mostly on the edges between less and more shows a typical glacial morphology (Figure 9). steep slopes. Here the frequency of limestone The lowest south-eastern parts of the plateau pavements is much smaller. Kaninski podi are noticeably glacially abraded, es- Smaller forms of subglacial abrasion like pecially their lower margin with many rock drum- grooves, with overhanging lips, too, which have lins and steep rock bars or steps, which widen in different gradients, even opposite gradients, are some places into a smaller rock amphitheatres. often found at the bottom and also on the steep They are accompanied by some typical glacial sides of some dry valleys, along with some typical grooves, a sort of shorter rounded furrows, situ- roches moutonées areas. The formation of these 311 KRF•2 • OK.indd 311 15.12.2009 10:57:33 Karst Rock Features • Karren Sculpturing erosional grooves or notches can be attributed Conclusion to the subglacial erosion action of glacial water. In many places these subglacial forms remained In those sections of land surface that were most completely unaffected by corrosion under mo- affected by glacial erosion, the process would start raine cover (Figure 16). On such a bedrock sur- nearly from a virgin ice scoured surface, while face there shows striation in many places, which in others the process would be resumed, though has disappeared from the surface exposed to cor- under slightly altered conditions. In addition, the rosion for a longer time. Glacial mills, which were till immediately after the last ice age covered a also searched for by Desio, when he commented much larger area of the bare rock surface, com- on the report by Brazzo about such phenomena pared to the present situation, with varying thick- north of the border ridge, were not to be found ness and to an unknown extent. But even larger anywhere (Kunaver, 1983). On the other hand, sections of the plateau remained more or less bare. some of the uppermost areas of Kaninski podi Thus the rock surface was affected by a process plateau are less affected by the glaciation (Figure of corrosion gradually and in different periods of 17). the Holocene. This process may be referred to as a phenomenon of successive inclusion of rock sur- face into the karstification process, which is to be considered for pavements of all sorts. 312 KRF•2 • OK.indd 312 15.12.2009 10:57:33 Karren feaTures 25 in The dachsTein MounTain Gábor TÓTH We have carried out annual karren morphologi- Dachstein mountain only. The most extensive of cal research Austria since 1995. The main pur- these formations is the Upper Triassic Dachstein pose of investigations in Dachstein was the study limestone that achieves 1,500 m thickness in the of karren landform assemblages and karrenfields. Dachstein mountains. It is made up of two types: Considering the many factors influencing karren the bedded Dachstein limestone and the unbed- development (slope angle, exposition, vegetation ded fracture-traversed limestone. and soil cover, flow characteristics, solvent quan- The Triassic rocks are covered partially only tity, precipitation), it is easy to understand that by small-thickness Jurassic limestone, the age of the landforms of karren terrains show several idi- which is Upper Malm. Small amounts of Creta- osyncrasies in various locations. Accordingly, we ceous rocks are also present, while Lower Creta- examined forms as follow: surfaces of the same ceous is totally absent. From the Upper Cretaceous slope as well as surfaces characterized by a differ- are found sandstone, conglomerate and marls. ent kind of exposure, namely, small dipping sur- According to tectonical history, several main faces bordered by cracks, steep wall karren out- units can be distinguished: the Dachstein massif crops, surfaces water-supplied from vegetation, which bends the Hallstatt nappe, and the thrust covered terrains, and karren forms near a glacier. sheet bordering the Dachstein mountain from the Another part of our measurements comprised south. The karstic forms are mainly represented the analysis of single forms. We especially empha- by the Dachstein massif, which is broken up along sized meandering karren, trittkarren (heel-print fractures. The dislocated blocks induce significant karren) and the different types of ril enkarren (so- relief energy. lution flutes) and rinnenkarren (solution runnels). Typical karren terrains in the Geological survey Northern Calcareous Alps The Northern Calcareous Alps consist of Triassic, Karren morphological research was accomplished Jurassic and Cretaceous limestones, Upper Trias- in several terrains of the Dachstein and Totes Ge- sic Dachstein limestone being the most remark- birge mountains. able rock regarding karren landforms. The Middle Complex examinations were performed with Triassic dolomite appears in the southern wall of morphometrical methods on the Dachstein pla- 313 KRF•2 • OK.indd 313 15.12.2009 10:57:33 Karst Rock Features • Karren Sculpturing teau near the Krippenstein peak at the intersec- • the most high altitude forms are the elevations tion of 661 and 662 pathways (Figure 1). The range at the edge of the plateau (Krippenstein); of altitude on the adjacent peaks (about 2,000 m • the fourth outstanding forms are the bedding- in height) encloses the main karstified internal plane terrains of various size. terrains of the plateau. In the surrounding envi- Young karren landforms can be found in the ronment four kinds of forms can be recognized: vicinity of the Hallstatt-glacier where the exact • older and large-sized paleodolines that are con- date of the ice recession is known. Not far from siderably transformed by the ice and in which this place, under the Simony house, we measured lakes also may occur (Lake Däumel); the different intensity of karren denudation on the • the whole area is characterized by young shafts bed-escarpments (cuesta) of several stepped pave- produced by active karstification; ment karst (schichttreppenkarst) locations.    Figure 1: Research areas in Dachstein. 1. schichttreppenkarst under the Simony house; 2. foreground of the Hal statt- glacier; 3. zonal and local karren assemblages near Krippenstein peak. 314 KRF•2 • OK.indd 314 15.12.2009 10:57:34 Gábor Tóth, Karren features in the Dachstein mountain The terrains selected for the mapping in Totes 2.20 1.76 1.78 1.84 1.93 1.59 1.55 1.30 Gebirge are aligned near the Scheibling peak along Density (No/m) the 230 Path. Here the chosen sites have karren Total developed on bedding-plane surfaces correspond- m) 43.96 33.52 27.11 26.56 25.57 29.63 25.35 21.58 ing to former cirques or slightly dipping terrace ccurrences of S.s. (cm/ plateaus in the side of glacial valleys. Their sur- – – – – 0.29 0.08 0.38 0.18 faces are almost impassable due to vertical karst ensity. o Density (No/m) forms and appear dissected into smaller parts by grikes (kluftkarren). etre; d – – – – Trittkarren m) 5.27 27.0 3.68 0.81 Several microforms of a karren location simi- S.s. (cm/ lar to the above mentioned area were examined. s on 1 m – – – – – – 0.12 0.11 The researched area can be found under Widerkar orm Density (No/m) peak, at an altitude of 1,800 metres, on the incline enitzas of a glacier valley, which is without any outlet, – – – – – – arren f Kam m) 8.00 1.29 draining as an independent karstic entity, and S.s. he k (cm/ separated into smaller parts along fissures and cracks (Veress and Tóth, 2001). h of t 1.55 1.64 0.67 1.04 1.00 1.09 1.15 0.77 dt Density (No/m) otal wi Runnels m) Research methods 15.47 , t S.s. 25.47 31.64 13.22 15.64 21.50 20.75 15.85 (cm/ ion – – A useful method was mapping through a square olut 0.08 0.04 0.24 0.29 0.04 0.20 ipes grid, which provides suitable documentation for els Density (No/m) middle-sized and larger karren surfaces. The basic pecific s arren p – – Karren w m) 2.78 0.80 5.33 5.00 0.68 2.75 principle of this method is to cover the chosen . s. s and k S.s. (cm/ area with a horizontal, suitably meshed net, and ap; S – – – – – – – then determine the distance of points of forms 0.18 compared to the points of the net. By using this arren Density (No/m) procedure, maps on the scales of 1:10, 1:20 and – – – – – – – m) 0.81 1:100 have been made, applying 10, 20 and 50 cen- Network k S.s. altitude on m achstein. (cm/ timetre-distanced nets. The largest mapped area s. * . D – (20 x 25 metres = 500 m2) was surveyed next to the 0.16 1.67 0.19 0.64 0.13 0.20 0.40 orm Widerkar peak of Totes Gebirge in Austria. rge; D Density (No/m) Using surveying instruments, contoured topo- ebi arren f Grikes – m) 2.45 13.89 2.08 4.93 5.81 1.85 4.44 graphical maps of larger karren forms have also S.s. otes-G (cm/ been made (Veress et al., 1995). This method . T 27 44 15 39 27 35 31 35 seems to be the most suitable for mapping single, ensity of k No. several metres long and wide troughs. etre; T nd d The basic investigation method referring to the 10˚ 15˚ 20˚ 30˚ 31˚ 17˚ 21˚ morphogenetics of karren landforms was to make ion a Surface Slope angle a transect on selected profiles (Veress et al., 2001a). olut We determined on the transects the type of each eature on 1 m 1630 1820 2051 karren form occurring along a spread band- Height pecific s arren f 1800–1900* 1800–1900* 1900–2000* 1900–2000* 1900–2000* chain, measuring its width, depth and direction : S and the slope angle and direction of the terrain. Site /1 I/1 I/1 Table 1 code each k T 4 T 5 T 3 T 2 T 1 D I D I D I 315 KRF•2 • OK.indd 315 15.12.2009 10:57:34 Karst Rock Features • Karren Sculpturing Figure 2: Directional dis- persion of the karren features along D IV/1 transect (mountain pine zone, Dachstein). 1. wall karren; 2. karren wel s and karren pipes; 3. slope direction with the angle of gradient; 4. strike direction; 5. densi- ty of fractures (ocurrence per 10 centimetres). The lengths of each transect were 15‒25 metres, into two levels. At the lower level between 1,400 depending on the size of the measured outcrop. and 1,800 metres a.s.l. cavernous forms are typi- Analysing the data, several specific and global pa- cal, whereas at the upper level between 1,800 and rameters can be determined. In this manner the 2,000 metres a.s.l. karrenfields are dominant value of specific karren dissolution can be esti- (Hartlieb, 1999). According to our own examina- mated by dividing the whole widths of the karren tions, the different types of karren forms occur in features along the section by the length of the sec- a wider altitude interval, modifying the two kar- tion (Table 1, Figures 2, 3). ren zones mentioned above. As the formations of the two zones extend beyond their limits, it is more useful to classify the forms by their genet- Morphologic outline ics. The whole karstic landforms of Dachstein can be gathered into three general groups: subsur- The karren landforms of Dachstein are surveyed face forms, large superficial karst landforms and principally from the view point of karren land- smaller karstic forms, namely the karren features. form assemblages. As our examinations have been focussed on the Glacial influence and intensive karstification small-scale morphology of the mountain, hence determine the actual landforms of Dachstein. the karren landforms of the mountain are being After M. Hartlieb the karstic phenomena fall presented (Figure 4). 316 KRF•2 • OK.indd 316 15.12.2009 10:57:35 Gábor Tóth, Karren features in the Dachstein mountain Figure 3: Directional dis- persion of the karren features along H II /2 transect (mountain pine zone, Dachstein). 1. grikes (kluftkarren); 2. karren wel s and karren pipes; 3. slope direction with the angle of gradi- ent; 4. strike direction; 5. density of fractures (oc- currence per 10 centi- metres). Karren features are basically divided into three is characterized by a bedding-plane catchment at groups, especially in the case of gentle dip and the top, a water transmission in the middle and a suitable slope length. In the upper zone of the water drainage towards the bottom. karrenfields the elementary, embryonic stage is Karren forms ‒ integrated as karren assemblages shown, represented by variable sized microforms. ‒ are classified into two separable groups based on Solution flutes (rillenkarren) start at the upper the stratigraphic position of the surfaces. The first edge of the slope and are characterized by U- group consists of the less steeply dipping bedding shaped cross sections. Below this area solution lev- planes and is characterized by longitudinal and els (ausgleichsflächen) can be found, which serve circular forms. The second group contains the wall as the feeding area for the rinnenkarren (solution karren (only longitudinal forms) developed on the runnels) located in the middle zone. The most steeply dipping beds. If structural benches (cues- general forms of the middle zone are single solu- tas) have been created by glacial erosion, the whole tion runnels, as well as meandering karren, ka- landform complex of the area is called stepped (or menitzas and heel-print karren. On the lower part staircase) pavement karst (schichttreppenkarst). of the slope karren wel s (karrenröhren), karren By investigating the surfaces near the peak of shafts and fissures drain off the solvent water. In Krippenstein we found out coalescent circular this way, it can be demonstrated that the typical forms could develop into longitudinal ones. On high mountain karren in the Dachstein outcrops the other hand, the different karren features (run- 317 KRF•2 • OK.indd 317 15.12.2009 10:57:36 Karst Rock Features • Karren Sculpturing Figure 4: Karrenfields in Dachstein. nels, trittkarren, grikes and karren wells) strongly and trittkarren. Trittkarren evolve on the oppo- influence each other’s evolution. Several trittkar- site sides of the roches moutonées, where glacial ren (heel-print karren), being placed above each striae do not influence the water flow. other, could frequently develop into downward The measurements at the ice-free base of the runnels and could serve as their catchment area. Hallstatt-glacier produced surprising results. Similarly, circular forms may change into longi- Knowing the recession rate of ice cover it was pos- tudinal forms when small karren wells and pipes sible to define the speed of evolution of the karren. join to develop grikes. Our examinations considered the elementary kar- Investigations performed at several levels show ren forms evolved during the recess of ice cover that the degree of the surface dissolution depends (Figure 5). The common characteristic feature on the altitude and slope angle of the terrain. The of these forms is that their sizes and frequencies amount of dissolution is strikingly high on the gradually increase with increasing distance from gently dipping terrains with southern exposure the glacier (Veress et al., 2001b). and on the karren surfaces above 2,000 metres a.s.l. According to the dates marked at the valley bot- In the latter case one possible reason for intensive tom, some small-sized solution runnels (Figure 6) corrosion is the thickly accumulated melting snow, evolve after some years, because their evolution which guarantees long-lasting dissolution. process is strongly promoted by the glacier striae. The evolution period of primitive, extremely shal- low kamenitzas is seven years, the first trittkarren Juvenile karrenfields in Dachstein can be established after 23 years. Juvenile karren development is characteristic of the areas where the ice receded not long ago. In Schichttreppenkarst in Dachstein the glacier foregrounds elementary karren forms appear, most of them being preformed by glacial Stepped pavement karsts (schichttreppenkarst) striae. The first forms to appear are rillenkarren stand out among the most spectacular terrains 318 KRF•2 • OK.indd 318 15.12.2009 10:57:37 Gábor Tóth, Karren features in the Dachstein mountain Figure 5: Juvenile karren- fields surrounding the ice-free bottom of the Hal statt-glacier. Figure 6: Juvenile karren forms at the ice-free bottom of the Hal statt- glacier. Width of view is 2 m, below. with karren surfaces; their characteristics were the shrinking of the ice, began the formation of described first by Bögli (1964). This kind of gla- karren features on the ice-prepared gently in- ciokarstic complex form evolves on horizontal clined bedding planes and vertical scarps promot- or gently inclined well-bedded limestone dur- ed by glacial striae and fissures beneath the glacier. ing glacial and later solutional denudation. The Such schichttreppenkarst terrain can be found in two parts of the schichttreppenkarst are the ice- Dachstein near Moderstein and under the Simony rounded scarp (the more abrupt and tilted side of house. It is worth mentioning that the temporal the bed) and the horizontal bedding plane. After relationship between the dissolution and the gla- 319 KRF•2 • OK.indd 319 15.12.2009 10:57:38 Karst Rock Features • Karren Sculpturing Figure 7: Local form complex with a small karren pipe. Figure 8: Local form complex with trittkarren (heel-print karren). Width of view is 50 cm, in the middle. cial erosion is not yet explained. It is well known ed into independent karren surfaces ( units) along that dissolution may occur also on ice-covered grikes which are not interconnected hydrologi- terrain where the ice does not continuously touch cally. Also snow plays an important role in the de- the surface of the rock (the air renders the water velopment of the schichttreppenkarst. The snow aggressive). However, these are isolated terrains, accumulates at the vertical and horizontal adjust- some parts of which are denuded by glacial ero- ments of the structural bench (cuesta) causing a sion in their early stage. The wall karren are the slow and long-lasting dissolution, especially when main forms over the scarps. The diameter of their the flat surface under the structural bench dips U-shaped cross-section is 1–5 cm and their length towards the wall above it. In this case a zone will ranges from 1 to 5 m. Advanced forms can be develop at the border of the horizontal and verti- found densely packed within some centimetre cal part characterized with karren wells or shafts. distance in the foreground of the Simony house. The 5°‒15° slope angle makes various landforms of bedding planes possible, the most characteristic Karren complex forms in Dachstein of which are trittkarren, kamenitzas and mean- dering karren. The important peculiarity of the The karren landscapes constitute karren complex bedding planes is that they are frequently separat- forms (or karren assemblages) that one can clas- 320 KRF•2 • OK.indd 320 15.12.2009 10:57:40 Gábor Tóth, Karren features in the Dachstein mountain sify by several aspects (lithological, biogeograph- ical, climatical). By our investigations, in the Dachstein we established three morphogenetic types of terrains: moderately dipping karren areas, karren zones found on terrains of steeper dipping and karren complex forms developed on structur- al bench terrains. Each of them is associated to a well-distinguishable kind of surface development. Gently dipping local complex forms These karren complex forms usually develop on inclines of less than 10° and vary morphologically and hydrologically. Other important phenomenon is that their wa- ter-course does not cross over their margins. The water is drained into the depth by shafts and kar- ren wells while their boundaries are ridges and grikes. Their size is fairly variable, ranging from a few m2 to 50‒60 m2. The most frequent features of these local karren assemblages are rillenkarren, rinnenkarren, kar- ren pipes (Figure 7), karren wells and trittkarren (Figures 8, 9). In the course of their development, Figure 9: Local form complex with concave trittkarren the karren areas may merge into each other and (heel-print karren). expand, respectively (Tóth, 2003). Zonal karren complex forms in the formation of flat surfaces (Ford and Wil- liams, 1989). The solution of small quantities is ex- These result from normal karren evolution on in- plained by the falling precipitation and the lami- clines of 10°‒60° dipping. Their size ranges be- nar flow. The formation of the flat surface is also tween 50 and 300‒400 m2 (Figure 10). The locali- caused by the long-lasting solvent process of small zation of the several karren forms on the incline extent because of the thaw of the accumulated varies with the flow features. The rillenkarren snow. The water arriving from the solution level are located at the top and their evolution down- feeds the zone characterized by solution runnels wards is related to sheet water flow when the solu- (rinnenkarren), meandering karren and trittkar- tion level (ausgleichsfläche) appears at their bot- ren underneath. Here the flow divides into water tom, the solutional development of which is not branches and is always turbulent. yet clarified. By one of the theories, the solvent is The zone located at the lower part of the slope saturated here and no more solvent action is pos- has the most varied and well developed form com- sible. In contrast, however, the solution level is de- plexes. Mixture corrosion may result in intensive graded with the surface. By another hypothesis, dissolution at the confluence of the water branch- the degrading of the terrains takes place resulting es, explaining the frequent gradual change of the 321 KRF•2 • OK.indd 321 15.12.2009 10:57:41 Karst Rock Features • Karren Sculpturing Figure 10: Zonal karren surface near the Krippenstein peak. forms into each other. The main forms of this are Different dipping non-local complex forms the mature rinnenkarren (solution runnels). Finally, the infiltration zone close to the surface These arise in karren terrains consisting of steeper consists of grikes, frequently composed by align- scarps and gently sloping bedding planes. These ments of small pipes, karren wells and shafts. The karren structural bench terrains correspond to water flow is strongly delimited by the rinnen- the before-mentioned stepped pavement karst karren systems placed above this zone. The solu- (schichttreppenkarst). tion shafts and karren wells drain the water away at the terminations of the runnels, while their ac- cretion results in the formation of grikes perpen- dicular to the slope direction. 322 KRF•2 • OK.indd 322 15.12.2009 10:57:42 glacioKarsT landforMs 26 of The loWer adige and sarca Valleys Ugo SAURO In the lower Adige valley and in the area of the tongue of Adige, has complex assemblages of kar- Garda Lake (Italian Southern Alps) many typical ren (Figure 1). Among the most common forms glaciokarst landscapes are recognisable. The fin- are kamenitzas, distinguishable in three main est is probably that of Canale, where a rock bench, subtypes: solution cups, solution pans, and solu- abraded during the late Pleistocene by the glacial tion pans nested inside runnels. The karren devel- Figure 1: Detail of the glaciokarst rocky landscape of Canale in the southern Adige valley. 323 KRF•2 • OK.indd 323 15.12.2009 10:57:43 Karst Rock Features • Karren Sculpturing Figure 2: A relative- ly deep solution cup that holds water al- most permanently. There are two outlets: the one to the left is fed by the overflow- ing water, while that to the right is more recent and is fed by water through a fis- sure. Figure 3: A solution pan with a well developed outlet Figure 4: A large runnel with solution pans nested inside. with small meanders nested inside. The sides of the pans are characterized by different de- grees of edge overhang. 324 KRF•2 • OK.indd 324 15.12.2009 10:57:49 Ugo Sauro, Glaciokarst landforms of the lower Adige and Sarca val eys oped inside the rocky mass like the minute-shafts depths between a few millimetres and some deci- and the grikes seem to evolve by a speleogenetical metres. It is possible to distinguish at least three process from an inner network of cavities towards main morphological types: solution cups, single outside. solution pans, and chains of solution pans nested inside large runnels. Solution cups (Figure 2) are basins with surface Description of the karren features openings smaller than their floors, therefore hav- ing overhanging sides; their depths are relatively In the lower Adige valley and Sarca valley of Ital- large, and their floors are flat and horizontal. ian Southern Alps, some typical glaciokarst ter- Solution pans (Figure 3) are shallower kameni- rains are recognisable. They are worthy of note tzas, with greater width/depth ratios; the larger both for their geographical position and for the forms are also more elongated. Each solution cup variety of karren landforms developed during and solution pan has a solution runnel as its outlet. the late Pleistocene and Holocene. These land- The solution pans organised in chains and forms are situated at very low altitudes (between nested inside large runnels are deeper forms, 100 and 300 m above sea level) and relatively low with varying degrees of overhanging rims (Fig- latitudes (between 45°30’ and 46°00’N), and are ure 4) within their profiles that are an expression characterized by a sub-mediterranean vegetation of cyclic evolution (Sauro, 1973b). that is influenced by the extensive rock outcrops The evolution of each different type of kame- exposed to solar radiation. nitza is probably controlled by the relative size Probably the finest karst landscape is that of the hydrographic basin that feeds the hollow. situated on the northern side of the village of The cups have very small basins, while the solu- Canale in the lower right slope of the Adige val- tion pans have relatively larger basins. Evapora- ley, about 150‒200 m a.s.l., developed on a mas- tion of accumulated rainwater run-off may cause sive lens of an oolitic limestone of Lower Jurassic the deposition of thin layers of silt on the floors age, and shaped by the glacial tongue to form a of the hollows. If the catchment is small, the rock bench (Corrà, 1972; Sauro, 1973a; Perna and layer of silt is very thin, and probably does not Sauro, 1978). Abrasion by the glacial tongue, and match the solutional deepening of the kamenitza, in particular by the lodgement debris at the base which therefore evolves into a deep cup profile. If of the glacier, shaped the bedrock into round- the catchment is larger, the layer of silt becomes ed knobs elongated in the direction of ice flow thicker and the solutional effort develops lateral ( Rundhöckerkarst, in the German literature). Into widening rather than deepening. Within the so- these glacially moulded surfaces, solution has lution pans that are nested as chains inside a large sculptured a large variety of karren, which are runnel, the turbulence of the run-off flow during expressions of the different microenvironments rainstorms is able to remove the silt layer, and and hydrological processes. therefore the landform may deepen. Analysis of these landforms reveals the style Water is present for most of the time in the so- and the progression of karst morphogenesis in lution cups, while the solution pans are normally a massive limestone. Kamenitzas (solution ba- empty of water except during and immediately sins) are very well developed. These basins were after rainfall events. So the environment in the the first karst features to develop after the retreat two types of solution basins also evolves differ- of the glacier. Most of the basins are cut into the ently with respect to biological colonization of higher parts of elongated rock ridges. The vari- the rock surface. ability of these features is substantial: diameters After periods of rain, some circular patches of range from a few centimetres to some metres, and the bare rock surface remain wet for longer pe- 325 KRF•2 • OK.indd 325 15.12.2009 10:57:49 Karst Rock Features • Karren Sculpturing Figure 5: A large, inactive kamenitza developed before the last episode of glacial abrasion when the original form was partly eroded, especial y on its right side. riods than do the surrounding surfaces. These vance into the Adige valley. They have been partly patches appear to be on rock of a higher porosity, abraded by glacial erosion. probably a consequence of some type of biological Many of the solution runnels, that originated colonization. This biological invasion predisposes as outlets from the solution cups and pans, are the rock to the development of kamenitzas, which characterized by sequences of small depressions seem to begin to develop like virtual forms within similar to heelprint karren; these forms are also the rock before becoming real depressions. similar to the runnels starting from soil patches Some of the largest solution pans (the largest (Humusrinnenkarren, in the German literature). is 4.5 m long), which are now completely inactive On the relatively smoothed and gently rounded (Figure 5), seem to have evolved in two phases glaciokarst knobs, rillenkarren or solution rills of karstic dissolution separated by a glacier ad- are not well developed. In particular, they are 326 KRF•2 • OK.indd 326 15.12.2009 10:57:51 Ugo Sauro, Glaciokarst landforms of the lower Adige and Sarca val eys Figure 6: Grikes and minute-shafts that are progressively subdivid- ing the rock surface into isolated blocks. absent or scarce where extensive and continu- bare, fissured limestone pavements within the ous, gently sloping surfaces exist. On the contrary, karst terrains (Figure 6). In the Canale area, the they are abundant where the rock surface is par- oolitic limestone is massive and lacks visible bed- tially dissected by grikes (kluftkarren), where ex- ding planes and fracture discontinuities. However, tensive sheet flow of water cannot develop during discontinuities do exist, and are revealed by de- rainstorms. velopment of the structurally controlled karren. Minute-shafts, kluftkarren grikes and bedding These karren forms develop in the more porous plane fissures are structurally controlled features parts of the rock, and not as a consequence of that have developed within the rock mass. In some rainfall run-off flowing directly into holes. It is ev- areas, they are noticeably abundant and subdivide ident that most of these cavities are not open from the rock into isolated blocks, creating distinctive, the outer surface towards the inside, but from an 327 KRF•2 • OK.indd 327 15.12.2009 10:57:53 Karst Rock Features • Karren Sculpturing internal network of cavities towards outside. An other interesting glaciokarst landscapes. Of par- important role in their development is probably ticular note, for both their karren assemblages played by the air circulation inside the cavities and their scenic positions, are those situated that is induced by the diurnal cycle of solar radia- north of the Garda lake, including the limestone tion acting as a heat pump, and by the water con- pavements developed on a flatiron of Eocene lime- densation associated with this circulation. stone near the village of Nago, and the Calodri In areas near the Canale karst, there are many ridge north of the town of Arco. 328 KRF•2 • OK.indd 328 15.12.2009 10:57:53 Karren in paTagonia, 27 a naTural laboraTory for hydroaeolian dissoluTion Richard MAIRE, Stéphane JAILLET and Fabien HOBLÉA The unusually well developed hydroaeolian kar- These karst areas are located between 50 and 52°S ren landforms of the Chilean Patagonia archipela- in Madre de Dios and Diego de Almagro islands, go are a paramount natural heritage which we in the province of Última Esperanza (XIIth region have begun to study recently (Maire, 1999; Jail- of Chile: Magallanes y Antártica Chilena) (Figure let et al., 2000; Hobléa et al., 2001; Maire, 2004). 1). They are the most southerly and the most in- 12 th region Madre de Dios Diego de Puerta Natales Almagro Estrecho de €‚€ Evangelistas Magallanes Punta Arenas Tierra  del Fuego ­ Cape Horn        Nm 2000  Figure 1: A. Location map of Madre de Dios and Diego de Almagro islands; B. detail of the studied area in Madre de Dios (after an aerial photograph, Servicio Geographico Militar del Chile); C. map of the studied experimental catch- ment constituted by step-like karren. 329 KRF•2 • OK.indd 329 15.12.2009 10:58:04 Karst Rock Features • Karren Sculpturing hospitable on Earth due to the subpolar climate ed and layers are subvertical (Avenir peak, 815 m characterized by the extreme rainfall and strong a.s.l.). The Huemul formation which combines the winds (“roaring fifties”). The hydroaeolian karren Duc d’York and Denaro formations is constituted are a specificity of Patagonian karsts and several by dark bands of volcanic sandstones and grau- forms have never been described before. Because wackes intruded by dikes of lamprophyre. The pure limestones are rare in Chile, these islands metamorphic complex, located in all the SW half, were recognized in 1930‒1950 as within an inven- is formed by a thick sequence of pelitic schists, tory of mineral resources (Biese, 1956, 1957; Ce- green shales and amphibolites (Escobar, 1980). cioni, 1982). This folded metamorphic formation does not exist in Madre de Dios; it represents a deeper structural level of the same accretion prism according For- The geoclimatic context sythe and Mpodozis (1983). During Upper Carboniferous and Lower Per- Litho-stratigraphy mian, coral reefs formed on underwater volcanic intraoceanic mounts, forming atolls surrounded The carbonate terrains of the archipelagos form by bioclastic limestone formations which have part of the pre-Jurassic basement of the Andean been locally dolomitized (Cerro Pelantaro in Cordillera which was the former Pacific margin Diego de Almagro). Between Upper Paleozoic of Gondwana. Limestone and marble outcrops and Lower Mesozoic, this ancient active margin constitute a band a few kilometres wide bounded was a accretionary prism, that is to say a thrust to the west by the Pacific Ocean and to the east sedimentary arc with a Pacific vergence built by the Patagonian granodiorite batholiths dating from the subduction zone. The metamorphism of from the early Cretaceous (Halpern, 1973; For- limestones into marbles is extensive for the Ploma sythe, 1981; Forsythe and Mpodozis, 1983). They formation and variable for Tarlton limestones. In are interbedded with volcano-sedimentary rocks the two islands, limestones and marbles show nu- and dikes. In Madre de Dios, Forsythe and Mpo- merous dikes of plutonic and subvolcanic rocks. dozis (1983) distinguish three sedimentary for- A contact metamorphism is visible on the edge of mations. The Tarlton limestones are massive lime- the largest dikes: brecciation of limestone, recrys- stones and marbles more than 500 m thick and tallization and fragilization of limestone (“sugar” date from the Upper Carboniferous and Lower cryptocrystalline facies), limestone fragments in- Permian. The Denaro formation is a sequence cluded into subvolcanic rocks. By differential dis- showing from bottom to top thick submarine ba- solution, these dikes and veins can be used as a salts (pillow lavas), a level of biogenic cherts and tool to measure postglacial surficial dissolution of red shale (30‒60 m), grey calcarenites (50 m) inter- limestone and marble. bedded with argilite and siliceous nodules. The Duc d´York formation is a very thick volcano- sedi mentary sequence of flysch type. A hyperhumid subpolar climate In Diego de Almagro, situated 150 km south of Madre de Dios, the geological formations are The cold oceanic climate of Patagonia is related to similar (Escobar, 1980; Cecioni, 1982; Forsythe the interaction between the tropical anticyclones and Mpodozis, 1983; Maire, 1999). The Ploma (subtropical convergence of South Pacific) and the formation correlates with Tarlton limestones. It is southern low pressures (polar front) which ac- formed by massive and sparitic white marbles and counts for the climate of the “roaring fifties”. These sometimes grey dolomites. These carbonate ter- rotational cells coming from west generate huge rains of several hundreds metres thick are thrust- precipitation at the contact of the first mountains 330 KRF•2 • OK.indd 330 15.12.2009 10:58:04 Richard Maire, Stéphane Jail et and Fabien Hobléa, Karren in Patagonia, a natural laboratory for hydroaeolian dissolution and very strong winds (NW, SW, W). Below 50°S, istence of the magellanic forest of Nothofagus, one Última Esperanza archipelago has an isothermic of the last primary forests in the world, is a spe- subpolar climate called Tundra Isotérmico (Zamo- cificity of Chilean Patagonia and Tierra del Fuego. ra and Santana, 1979). To have a subpolar climate, Even if the climate is subpolar, this forest can be the temperature of the warmest month must not developed in the shelter of rock dolines and at the be greater than +10°C; this temperature is normal- bottom of cliffs up to an altitude of 400 m. Because ly the thermic limit for trees. Nevertherless, the ex- of a strong wind, the very low growth of some Not- Figure 2: Giant rinnen- karren and canyon-like wandkarren in Madre de Dios in the Carbonifer- ous and Permian lime- stones (Tarlton lime- stones). Width of view is 10 m, below. 331 KRF•2 • OK.indd 331 15.12.2009 10:58:06 Karst Rock Features • Karren Sculpturing Figure 3: The floor of a large flachkarren showing the laminar flow by wind deflection during a shower, Madre de Dios (scale = 0.8 m in the foreground), photo Ultima Patagonia. hofagus gives birth to bonsai with sometimes hori- in Guarello station. The mean speed of wind is 70 zontal trunks of 5 to 10 m long. km/h at Guarello station with a north-west domi- Precipitation reaches 7,330 mm per year (80% nant direction. The absolute maximum at Madre rain, 20% snow) at the Guarello station (Madre de Dios is unknown, but in February 2001 (Diego de Dios) for the 1950‒1970 period (Zamora and de Almagro) we have measured frequent speeds Santana, 1979). The whole of the external archi- between 120 and 140 km/h. In Evangelistas sta- pelagos of Última Esperanza between 49 and 52°S tion situated 52°30’S, south of Diego de Almagro are included into the isohyet 6,000 mm per year as (Figure 1.A), the absolute maximum is 183 km per the Patagonia precipitation map (Toledo and Za- hour in July (NW). pater, 1991) shows. This very rainy climate is dif- ferent from the humid alpine climate which has a snow precipitation of more than 65 to 70% (Maire, Types of karren and hydroaeolian 1990). The subpolar climate of Patagonia shows karren a regularity of precipitation with an average of 611 mm per month and 802 mm for the wettest The karren of Patagonia show remarkable and month (November) and 441 mm for the driest huge solution features due to the extreme rainfall month (June). The wettest year (1960) reached and strong winds upon ice-smoothed rock karst 8,495 mm (Zamora and Santana, 1979). But re- (Figure 2). The laminar flow, the concentrated cently, February 2008 recorded almost 1000 mm flow and the strong wind are three parameters 332 KRF•2 • OK.indd 332 15.12.2009 10:58:08 Richard Maire, Stéphane Jail et and Fabien Hobléa, Karren in Patagonia, a natural laboratory for hydroaeolian dissolution which combine more or less in relation with slope, topography, fracturation and exposure. Karren due to a mixing of a laminar flow and microturbulent flow There are three types of karren due to a mixing of a laminar flow and microturbulence flow related to wind deflection (frictional effect and gravity). Aeolian flachkarren (Figure 3): The flachkarren are known in the alpine karsts (Maire, 1990), but in Patagonia they are remarkably well formed on null or very weak slope (< 5°). Here the horizon- tal dissolution by the wind deflection is due to a quick laminar flow pushed and spread by strong winds in a dominant direction. The laminar flow can reach a high velocity despite the frictional ef- fect on the rock. The streamlines probably divide into two layers: a water film with microturbulence at the rock surface and, just above, a layer of few millimetres with the laminar flow. The water film responsible for the corrosion action is renewed Figure 4: Wave-like ripples on a limestone scarpment in continuously. This very specific process generates Madre de Dios formed by a laminar flow and wind de- flection. flat and very smooth surfaces reaching more than 1,000‒2,000 m2 never observed before at this scale. Wave-like ripples (Figure 4): These are a typical deflection and solutional form. They have been flow) due to the wind shocks. The process goes on described in New Zealand (Owen Range), in an by positive retroaction. alpine and windy context, by J. Jennings (1985). In Trittkarren (terrace-like ripples) (Figure 5): In Patagonia, they are widespread all over the karst Madre de Dios, trittkarren cover large surfaces of when the slope is steep (30‒70°), on walls and on roches moutonnées and are different enough from the edge of solution runnels and karren shafts. the classic heel-print karren described by Bögli These small ripples constitute a dense network of (1960a, 1980). When the slope is above 25‒30° semi-circular microstairs, each measuring 1 to 3 we have wave-like ripples, below 20‒25° the step cm long, 0.5 to 1 cm wide and 1 cm high. On ver- process by laminar flow and regressive dissolution tical walls, step-like rims become small rounded form small staged terraces of 2 to 10 cm wide and scallops like on a wall of a cave tube. So these several tens of centimetres long. These terraces micro-karren give a regular micro-crenellated follow the topographic contours and evoke per- surface. The genetic process of the regressive dis- fectly at a small scale the terraced cultivation of solution by small steps is connected with micro- south-east Asia. In a small “talweg”, the terrace- turbulences flow (water film) generated by the gra- like ripples transform in large concave steps called vity (slope), the frictional effect at the contact of step-like karren (infra). As in the Burren (Ireland), the micro-irregularities of rock and above all the trittkarren are also due to heavy rain in an oce- rythmic renewing by the pulsed flow (unsteady anic context (Sweeting, 1973). 333 KRF•2 • OK.indd 333 15.12.2009 10:58:10 Karst Rock Features • Karren Sculpturing Figure 5: Trittkarren (terrace-like ripples) covering the roches moutonnées karst in Madre de Dios, altitude 450 m (photo Ultima Patagonia). Step-like karren: a mixing form of laminar, tremity of steps because of a strong wind (water microturbulent and turbulent flows seems to go up) is a specific process for which an unsteady flow is responsible. If the distance is only The step-like karren constitute a specific and a few metres between a flachkarren and a karren widespread form which combine steps and run- shaft, there is a succession of steps more and more nels due to a laminar flow in upstream and a tur- high, a few cm to 0.5‒1 m high, on the convex wall bulent flow in downstream because of an increase of the karren shaft. During a rainfall the laminar of water concentration (Figure 6). The process flow concentrates and begins to become turbulent of step-like karren occurs especially on col and in the axis of the upper “talweg”. At the end of the rounded eminences due to original glacial topog- rainfall, the laminar flow disappears completely raphy in the vicinity of the flachkarren. As for and a small concentrated flow continues to run in the terrace-like ripples, the step regressive disso- the axis of the step-like karren. lution occurs when there is a beginning of water Depending on the topography and slope, concentration. First concave steps of 0.3 to 0.6 m step-like karren can continue by steps, 20 to 40 wide and 1 to 3 cm high occur in small “talwegs” cm wide and 5 to 25 cm high, in the bottom of (5‒25°), often over a distance of several tens of me- rinnenkarren 0.5 to 1.5 m deep with a slope be- tres. The bottom of each step shows a corrosion tween 25° to more than 40°. The height of steps depression which fills with water during the flow. increases with the slope. When the slope is steep, The recurrent interruption of water flow at the ex- between 40° and 90°, the steps are more numer- 334 KRF•2 • OK.indd 334 15.12.2009 10:58:11 Richard Maire, Stéphane Jail et and Fabien Hobléa, Karren in Patagonia, a natural laboratory for hydroaeolian dissolution Figure 6: Step-like karren above “Col du Bélier”, al- titude 400 m, in Madre de Dios. Width of view is 5 m, in the middle. ous and higher, up to several metres, before dis- high, like on the western flank of Cerro Pelantaro appearing. (Diego de Almagro) or in the west part of Madre de Dios (Figure 2). At the base of cliffs, each solu- tion runnel reaches several metres wide and deep Rinnenkarren, canyon-like wandkarren and as parallel small canyons (Figure 2). Rinnenkaren meanderkarren and canyon-like wandkarren with steps generate small gorges reaching until 4 to 8 m deep and 1 The large marble and limestone cliffs and glacial m wide in the bottom. These young canyons are valley sides are striated by huge rinnenkarren and generally 10 to 30 m long and show steps reach- wandkarren reaching sometimes 100 to 300 m ing sometimes several metres high just before the 335 KRF•2 • OK.indd 335 15.12.2009 10:58:12 Karst Rock Features • Karren Sculpturing Figure 7: Giant meanderkarren on the marble dome, al- Figure 8: Circular karren shafts with step-like karren. All titude 650 m, in Diego de Almagro. Width of view is 15 the limestone surface is covered by wave-like ripples m, in the middle. (photo Ultima Patagonia). Width of view is 8 m, in the middle. swallow hole. They occur by symmetric lower- for the process of the meander deepening related ing dissolution of each step. This morphology is a to the slope in a massive soluble rock as marble. good model to study the process of the regressive erosion. Meanderkarren can reach exceptional dimen- Pinnacles, kluftkarren and karren shafts sions even in catchments of only 1,000 to 1,500 m2. A remarkable example has been observed on Pinnacles of 10 to 15 m height can exist on the the top of a marble dome, at 600 m high, in Diego edge of rock dolines where a magellanic forest de Almagro, south-west of Avenir peak. During a grows. These forms join by thin and sharp arêtes rainfall the water in a flachkarren basin of about which exhibit wave-like ripples and sometimes 1,500 m2 moves NNE down a gentle slope, on rillenkarren. Klufkarren occur especially in the which forms a series of meanders with an ampli- fractured Tarlton limestones of Madre de Dios on tude of several metres (Figure 7). With the slope cols and domes. They are 2 to 5 m wide and 10 to increase, the meander deepening can reach sever- 50 m long. We often observe an asymmetry of the al metres deep near the edge of the valley side and edges related to the influence of the wind direc- 0.6 m wide at the bottom. This type of meander tion (W to NW). The side exposed to wind forms a canyon is also a natural model at the middle scale sharp angle less than 90°. The opposite side shows 336 KRF•2 • OK.indd 336 15.12.2009 10:58:15 Richard Maire, Stéphane Jail et and Fabien Hobléa, Karren in Patagonia, a natural laboratory for hydroaeolian dissolution Figure 9: A new family of karren, “the rock comets”, hydroaeolian wedges formed behind errat- ic rocks (“Col du Bélier”, Madre de Dios). Width of view is 9 m, in the mid- dle. Figure 10: An asymmetrical shaft in Madre de Dios caused by wind deflec- tion. Width of view is 10 m, in the middle. a convex edge often corroded by step-like karren portant caves of Patagonia, which are former gla- (Figure 10). Circular karren shafts of 10 to 20 m cial sinkholes (location in Figure 1.B). In Diego wide are formed by centripetal sinks of several de Almagro, the Avenir swallow hole, a 50 m wa- rinnenkarren and step-like karren from an ini- terfall at the contact of marble and volcanic sand- tial klufkarren (Figure 8). This is a typical post- stones, is a typical Holocene glacial lake sinkhole glacial morphology. But other karren shafts are (Maire, 1999, 2004). larger and seem older, for example, the upper en- Located downstream of runnels, the karren trance 100 m karren shaft of “Perte du Temps” (2.6 shafts are young solution swallow holes mainly km long) and the upper entrance swallow hole of of postglacial age. They play a main role for quick “Perte du Futur” (–385 m deep), the two most im- absorption of water into the endokarst. For ex- 337 KRF•2 • OK.indd 337 15.12.2009 10:58:18 Karst Rock Features • Karren Sculpturing Figure 11: Other hydroaeo- lian wedges showing the residual and horizontal form by differential dis- solution (“Col du Bélier”, Madre de Dios). Width of view is 5 m, in the mid- dle. ample, on the side of the Abraham fjord, in Diego posite direction of the dominant wind. One can de Almagro, an active shaft with small horizon- observe several remarkable types never described tal passages is directly fed by a large wandkarren. before: the hydroaeolian wedges behind erratic During a rainfall, a flood of 5 litres per second has blocks, the hydroaeolian triangles and oblique been observed at 20 m deep. This example shows pillars behind dikes and the keel-shaped wedges that one large solution runnel (or a small group) (without a shield). can provoke a sudden flood in the entrance of the Hydroaeolian wedges or “rock comets” (behind karren shafts a few minutes after the beginning erratic blocks): These exceptional residual forms of the rain. By comparison, an instantaneous 100 are easy to interpret. The reference site is the “Col l/s flood has been observed at 150 m deep in an du Bélier” near altitude 400 m (Seno Soplador, inclined tube of “Perte du Temps” cave (Madre Madre de Dios, Figure 1.B). A group of two dozen de Dios) 15 minutes after the beginning of the erratic blocks of plutonic rocks exhibit horizontal rain. Several floods can occur in the cave every and parallel aeolian wedges located on the leeward day. This is a main problem for cave exploration sides of each glacial block with NW–SE direction in Patagonia. (Figure 9). The length of the ridges is 0.5 to 2.5 m and the height is 10 to 25 cm behind the block (Figure 11). Elsewhere, for bigger blocks, they can Hydroaeolian wedges (“rock comets”) and reach 40 to 60 cm high and 3 to 4 m long (“Col the role of horizontal solution by deflection de la Baie”) (Figure 12a.h). The transverse sec- The strong and permanent winds facilitate the horizontal laminar flow as for the flachkarren (supra). Also they generate specific hydroaeolian karren as wedges behind erratic blocks and resid- Figure 12a: Morphological indicators of postglacial and ual small dikes of lamprophyre and subvolcanic actual differential dissolution in Madre de Dios and Diego de Almagro. Mean postglacial dissolution is rocks. These aeolian ridges are elongated forms about 1,000 mm in Madre de Dios (limestones) and 750 due to a differential dissolution, located in the op- mm in Diego de Almagro.  338 KRF•2 • OK.indd 338 15.12.2009 10:58:20 Richard Maire, Stéphane Jail et and Fabien Hobléa, Karren in Patagonia, a natural laboratory for hydroaeolian dissolution   €     ­ ­ †  ­ ‚ ƒ„  Š  ‰ ­  ‡  ­ ‡ ‹€ ­ ˆ  ŒŽ ‘ ’ ­ ‹€“ ­ ­  ‹‹† 339 KRF•2 • OK.indd 339 15.12.2009 10:58:20 Karst Rock Features • Karren Sculpturing Figure 12b: A pedestal of 170 cm on Monte Roberto near altitude 400 m, Madre de Dios. Width of view is 6 m. tion is triangular. During the rainfall, the strong replaced by a fragment of a small dike which is wind spreads the water horizontally and obliquely. raised by the differential dissolution. The refer- Upon a flachkarren, each erratic block constitutes ence site is also the “Col du Bélier”. We can ob- a shield which protects against dissolution a lime- serve some large triangular aeolian ridges behind stone elongated surface. There is a differential dis- small dikes of lamprophyres protected by the solution between the non-protected flachkarren wind. They can be generally 0.6 m to 1.1 m high, and the protected part. Aeolian wedges behind 1 m wide and 1 to 2 m long. There exists also a blocks are the horizontal equivalent of vertical unique dramatic example looking like a batter- pedestals situated under an erratic block called ing-ram (bélier). This is an oblique pedestal in- karrentische or tables of the corrosion (Bögli, 1961). clined to 35° protected behind a thin fragment of These last residual forms are frequent in the alpine the dike which is now isolated from the bedrock karst where the wind influence is not dominant. surface because of the erosion (Figures 12a.e, 13). But in Madre de Dios, in some protected places, This inclined pedestal is 3 m long, 0.5 m wide and the highest pedestals measure between 80 to 170 1.4 m in its highest point and indicates a huge cm like west of Monte Roberto near altitude 400 surficial corrosion. m (Figure 12b). Hydroaeolian ship’s bottom-like wedges: These Hydroaeolian triangles and pillars (behind remarkable karren are also a positive form of dikes): These residual solution forms are similar differential dissolution but without a protection to horizontal wedges. Here the erratic block is block or a dike. The reference sites are all in Madre 340 KRF•2 • OK.indd 340 15.12.2009 10:58:21 Richard Maire, Stéphane Jail et and Fabien Hobléa, Karren in Patagonia, a natural laboratory for hydroaeolian dissolution Figure 13: A hydroaeolian oblique pil ar formed behind a small dyke of lamprophyre (Col du Bélier, Madre de Dios). Width of view is 4 m, in the middle. de Dios: above Seno Soplador and also the “Col du the Venturi effect. In comparison to hydroaeolian Bélier” (Figures 1.B, 14). These ridges can reach 5 wedges situated behind erratic blocks, these resid- m in lenght and 70 cm in height. The upstream ual forms are poorly understood. They are prob- side, facing the wind, is steep and the downstream ably due to an original process of differential evap- part finishes as a sharp triangle or can be digitated. oration. After each shower, we observe the ridges The whole shape looks like a reversed ship bottom. dry more quickly than flachkarren because of the The arête is more or less crenellated and the sides wind. This process repeats several hundred times show step-like ripples. A statistical study of 129 per year and accounts for why the ridges corrode aeolian ridges indicates they are 0.2 to 5.3 m long more slowly than flachkarren. Once the process (mean 1.25 m) (Jaillet et. al., 2000). The direction has begun, there is a positive retroaction effect as is between N 123° and N 180°, with a dominant for the wave-like ripples. direction NNW-SSE (330‒345°/150‒165°). The wedges are grouped and occur preferentially on cols and eminences showing a former glacial mor- Sword-shaped karren and asymmetrical phology in roches moutonnées. Every group of ae- entrance karren shafts olian ridges has a similar direction and length of the same order. The morphogenesis is controlled The wind influence also explains a lot of residu- by the exposure to wind, especially in the glacial al sharp forms such as knife-shaped limestone valleys which channel and accelerate the wind by edges and asymetrical blocks. The most spec- 341 KRF•2 • OK.indd 341 15.12.2009 10:58:22 Karst Rock Features • Karren Sculpturing Figure 14: Hydroaeolian wedges like the bottoms of ships, without a shield. We explain this new kind of karren by a differen- tial evaporation because of the wind after every shower. Width of view is 7 m, in the middle. tacular is a sword-shaped karren. This extraor- thin separation walls are perforated by circular dinary morphology has been observed on Cerro holes. This sharp coastal morphology inherited Pelantaro in Diego de Almagro. This karren is a by the uplift of intertidal pools is accentuated by a marble sword, 0.80 m long, 5 cm wide, 2 to 3 cm strong mixing with heavy rain and spume spread thick, with very sharp arêtes. Precipitation, wind, by wind. purity of the sparitic marble and a specific site ef- In the carbonate islands of Patagonia, the fect (exposure) explain this residual karren. An- coastal staged notches of corrosion are amongst other typical morphology are asymmetrical en- the most remarkable tide-mark in the world (Fig- trances of the karren shafts. We have already seen ure 15). The reference site is in the Abraham fjord the asymmetrical klufkarren in Madre de Dios (Diego de Almagro), at the bottom of the marble (Figure 10). In Diego de Almagro, on the top of dome crossed by the “Perte de l’Avenir” cave. Five the marble dome (alt. 600‒700 m), between the notches are located at elevations up to 10.5 m (0 Huemul fjord and Abraham fjord, entrance ob- = high water). From bottom to top, we observe a lique karren shafts and klufkarren are eroded by notch 1 situated at high water level (0) where there the strong wind with asymmetrical holes and is a mixing corrosion between marine water and sharp edges (Maire, 1999). overland flow from rain coming from the wall. The notches 2 and 3 are located at +4.5 m and 5 m. The notches 4 and 5 are located at +7 m and Coastal karren and coastal erosion notches +8 m. There is a sharp contact at +10.5 m between the inclined karren wall and the top of the step 5 Coastal karren show large and sharp solution of 2.5 m high. This remarkable step, 12 m high in forms on the margin uplift during postglacial the low water, 10.5 m in the high water, indicates times because of the glacio-isostatic response, es- a strong uplift of more than 10 m since the end pecially between 0 and +6 m. The intertidal zone of the last glaciation. In Madre de Dios (Guarel- is only 1 to 1.5 m high and exhibits solution pans lo, Seno Soplador), marine notches are also well and pools several metres wide as in the north of formed, between 0 and +6 m. Guarello island. Between +1,5 m and +6 m, some 342 KRF•2 • OK.indd 342 15.12.2009 10:58:24 Richard Maire, Stéphane Jail et and Fabien Hobléa, Karren in Patagonia, a natural laboratory for hydroaeolian dissolution Figure 15: Marine staged notches, 10.5 m high, in Abraham fjord (Diego de Almagro) because of the glacio-isostasic response during the Holocene. A study of an experimental catchment pools located at the foot of each step into the step- with step-like karren like karren. These small basins form a water re- serve on the catchment. Between two showers, the The conditions of surficial flow have been studied water accumulated in pools dissolves more CaCO3 in a small experimental basin situated in Madre de than the running water and contributes to enlarg- Dios. It measures 972 m2 (perimeter 185 m, height ing the depression. At the beginning of the rain- 26 m) and drains a sinkhole. The catchment area fall, this water is pushed downstream by the pis- of the drainage network is constituted by 310 m of ton-flow. step-like karren (Figure 1.C). A study of a flood series At the top of the sinkhole we have studied, over a period of 3.5 hours each, two floods caused by two rainfalls. Precipitations, discharges, conduc- tivities and temperatures were measured between every 1 and 10 minutes according to the flow in- tensity. The short response time, from 5 to 10 minutes, is related to the small size of the catch- ment. The longest flow distance is 75 m. For each increase of a discharge, there is a small peak of conductivity of 10 µS/cm which is interpreted as   a piston-flow (Figures 16, 17). It is followed by a decrease of conductivity related to the dilution of Figure 16: Precipitations, discharges and measurements water during the flood peak. The process of a pis- of conductivity on the experimental catchment in ton-flow is probably due to the existence of small Madre de Dios (cf. Figures 1.A, B). 343 KRF•2 • OK.indd 343 15.12.2009 10:58:27 Karst Rock Features • Karren Sculpturing ­  €  €   €      ­ ‚  Figure 17: Measures of conductivity during the rainfall (12/02/2001) in wandkarren and in the outlet situated at the base of the marble cliff. “Perte de l’Avenir”, Diego de Almagro. The volume and the role of step pools ly good. With 1,250 steps, the total volume of the pools is about 410 litres (= 0.4 mm). To understand the hydrologic regime of a flood, The combined volume of the two rainfalls is we have measured the length, the width and the 2,430 litres (2.5 mm) on the experimental catch- depth of the 124 pools situated along the axis of ment. The flow is 2,220 litres (2.3 mm), with a rate the main step-like karren drain. The multipli- flow of 90 % and an evaporation of 10 %, on the cation of the three values divided by two gives a same order as the rate calculated by the equations good approximation of the volume. The curve of of Turc and Thornwhaite. The volume of the pools the pool volume from upstream to downstream (410 litres = 0.4 mm) has no influence in the hy- shows an acute limit at the level of the step n 39. drologic regime because the pools are full at the On the two sides of this limit, the curves are rela- beginning and at the end of the experiment. Dur- tively right and it is possible to determine a mean ing the 3.5 hour experiment, the water volume in unit volume per step which can be extrapolated for the other step-like karren. To characterize Table 1: Step-like karren; evaluation of the volume of the different drainage parts, we apply the Horton water in the step pools by the method of mean vol- classification. The pools of drains of the orders 1 ume extrapolated on two areas of the experimental and 2 have a unit volume of 0.14 litres (Table 1). catchment. The pools of drains of the orders 3 and 4 have a Horton classification Orders 1–2 Orders 3–4 Total unit volume of 3.96 litres. In the parts of the or- (upstream) (downstream) ders 1 and 2, the steps have a mean length of 0.25 Length of drains (karren) 253 m 57 m 310 m m. Downstream, in the parts of the orders 3 and Number of steps 1187 63 1250 Number of steps per metre 4.7 1.1 4.0 4 the steps are 1 m long. Moreover, downstream Volume of basin per step 0.14 l 3.96 l - of the main drain (orders 3 and 4) nearly all the (main drain) pools have been measured, so the result is relative- Total volume of basins 161 l 250 l 410 l 344 KRF•2 • OK.indd 344 15.12.2009 10:58:28 Richard Maire, Stéphane Jail et and Fabien Hobléa, Karren in Patagonia, a natural laboratory for hydroaeolian dissolution the pools has been renewed five times. In one year, hours) due to a permanent sinkhole which drains for a mean annual precipitation of 8,000 mm, the an impervious catchment in volcanic sandstones water of the pools is regenerated fifty times per day with lakes and peat-bogs. according our experience. Despite the renewing of the water in the pools, the electrical conductivity increases except during flood peaks, when there is Quantification of surficial dissolution a slight decrease. The two showers studied were not able to dilute strongly the water storage inside the Present and sub-present surficial dissolution pools. This is due to the persistence of a water film on the limestone surface between the two rainfalls. Present surficial dissolution: The conductivity The carbonate concentration of this film grows and of the karren water varies from 52 to 107 µS/cm contributes to the surficial dissolution. and the total hardness (TH) from 14 to 35 mg/l of equivalent CaCO . These variations are related 3 to the flow distance (about 20 to 60 m) and the Relations between a surficial flow and contact time with the rock. The karren water with underground floods mosses and small epikarstic springs reaches 135 to 161 µS/cm and 57 to 80 mg/l (pH = 7.30 to 7.65). The transit time of the floods in the caves is related Measurements of conductivity have been collect- to the transit time of a concentration surficial flow. ed during a rainfall at the bottom of a 60° cliff dug The response time for a small catchment (1,000 m2) by wandkarren 40 to 80 m long and 0.4 to 0.7 m is quick, between 5 and 10 minutes. But this re- wide near the “Perte de l’Avenir” (Diego de Alma- sponse can be reached in less than 5 minutes if the gro) (Figure 17). The discharge varies from 0.1 to rainfall intensity increases. Several transit times 0.3 litre per second and the conductivity is 59.8 to in the endokarst have been measured at differ- 106.9 µS/cm connected with the karren length. At ent depths. The first example is related to a flood the level of the outlet which sinks in the “Perte de generated by a few wandkarren and observed in a l’Avenir” the mean conductivity is 98.5 µS/cm. young cave of Diego de Almagro (Abraham fjord). Present specific dissolution: The conductivity The flood of about 5 litres per second occurs at 15 and the total hardness have been measured on the m deep 5 minutes after the beginning of the rain- experimental catchment in several samples. A lin- fall. Another underground flood, which was more ear relationship (R = 0,99) allows the conductivity severe, has been observed in an inclined tube to be converted to dissolved CaCO . It shows that 3 of the “Perte du Temps” (Madre de Dios) at 150 100 grams of dissolved CaCO have been exported 3 m deep. The rainfall began at 12 o’clock and the during the experience and represent a limestone flood wave of 100 litres per second occurs at 12h15΄. volume of 37,037 mm3. For a catchment of 972 m2, This short transit time supposes a surficial con- this dissolved volume is similar to a limestone centration of 5 to 10 minutes and an underground slice of 3.81.10-5 millimetre thick. The high purity transit time also of 5 to 10 minutes. These instan- of Tarlton limestone permits to neglect the weak taneous subterranean floods are connected direct- rate of insoluble minerals. Despite its punctual ly with the rainfall intensity and the rapidity of character, this dissolved surficial slice can be ex- flow concentration on karren. For the “Perte du trapolated if we compare it with other indicators. Temps”, we distinguish two kind of underground For the postglacial time estimated to 10,000 years, drains: 1) temporary drain with a great discharge the surficial ablation would be 0.95 m. This ex- variability (500 to 0.1 l/s in 24 hours) connected trapolated value is of the same order as the post- with karren sinkholes; 2) perennial drain with a glacial dissolution measured with morphological middle discharge variability (1,000 to 100 l/s in 24 indicators (infra, Figure 12a). 345 KRF•2 • OK.indd 345 15.12.2009 10:58:28 Karst Rock Features • Karren Sculpturing Figure 18: Traces of paint around the Quarry of Guarello (concession limits) dating back to 1948 showing a raised relief of 3 mm in about 50 years. Surficial dissolution in 50 years: A precise mor- Table 2: Evaluation of the surface dissolution in 50 years phological indicator was measured in 1997 near and during postglacial times according to the paint the quarry of Guarello exploited for lime. Traces traces of Guarello and the small dikes, cherts and ped- of paint for limits of concession dating back to estals highlighted by differential dissolution. 1948 show a raised relief of 3 mm (Figures 18, 12 Value of dissolution Dissolution Postglacial dissolution Morphometric indicators in 50 years in 10,000 years a.j). So the rate of surface dissolution is 3 mm in Experimental basin 4.5 mm 950 mm 50 years, that is to say a specific dissolution of 60 (Madre de Dios) (extrapolated) (extrapolated) Anthropic indicator 3 mm 600 mm m3km-2a-1. For 10,000 years, the surficial ablation (paint traces, 1948) (measured ) (extrapolated) would be 0.60 m. Morphological indicators 3 to 4.5 mm 600–900 mm (Diego de Almagro) (extrapolated) (measured) Morphological indicators 3 to 7 mm 600–1,400 mm (Madre de Dios) (extrapolated) (measured) Estimate of postglacial surficial dissolution by morphological indicators macrosparite marble (Figure 12a.d). These dif- The best morphological indicators are small dikes ferences are related with altitude, exposure to of basalt and lamprophyre and aeolian wedges wind and petrographic facies. These values are protected by erratic blocks highlighted by dif- the highest known in the world of the surficial ferential dissolution (Figure 12a, Table 2). They dissolution on naked karren and are four to six stand proud by 600‒1,400 mm with a mean value times more important than in classic alpine and of 1,000 mm in Madre de Dios. Amongst the pyrenean karst (Maire, 1990). In Madre de Dios, highest values, we have measured 1,100 mm for a broad dike projected by 3 to 4 m, is due to the a dike (Figures 19, 12a) and 1,400 mm for the differential dissolution of the surrounding lime- inclined limestone pillar near the “Col du Béli- stone which was weakened by the contact meta- er” (Figures 20, 12a.e). In Diego de Almagro, the morphism (Figures 20, 12a). Nevertheless, this postglacial dissolution is 600 to 900 mm (mean exceptional dissolution cannot be extrapolated 750 mm) probably because of the more resistant for the hard limestone and marble. 346 KRF•2 • OK.indd 346 15.12.2009 10:58:30 Richard Maire, Stéphane Jail et and Fabien Hobléa, Karren in Patagonia, a natural laboratory for hydroaeolian dissolution Figure 19: A raised dyke of subvolcanic rock with a differential dissolution of about 1.10 m (cf. Figure 12a.b). Width of view is 5 m, in the middle. Conclusion certain fundamental morphogenetic processes in the karstology and general geomorphology. The The subpolar Patagonian karst is unique on Earth high speed of a surface dissolution, about 1,000 and potentially merits the distinction of a World mm for Holocene period, permits direct analy- Heritage site to allow it to be protected and stud- sis of the relation between morphology (effect) ied in the near future (Peter, 2001). They consti- and process (cause) which is normally difficult in tute a natural laboratory showing in real time earth scien ces because of inherited phenomena. 347 KRF•2 • OK.indd 347 15.12.2009 10:58:31 Karst Rock Features • Karren Sculpturing Figure 20: A broad dike with a differential dissolution of 3–4 m because of the limestone contact metamorphism. Width of view is 16 m, in the middle. Heavy rain and strong wind are responsible for a continue the research is the very inhospitable en- quick laminar flow which generates hydroaeolian vironment and the financing of the speleo-karsto- karren as flachkarren, wave-like ripples, trittkar- logical expeditions. ren (steplike ripples) and wedges behind erratic blocks. The small catchments represented by the karren arborescence are functional models which Acknowledgements follow the laws of the flow and regressive erosion- dissolution in the connection with specific crite- We thank the French Federation of Speleology ria (Veress and TÓth, 2001). The rythmic forms and the members of the speleological expeditions as wave-like ripples, trittkarren and meanderkar- in Patagonia. We hereby express our thanks to ren have a fractal dimension that can be studied, Guilaine Reaud-Thomas for drawing the Figures as in subterranean scallops, by the small wavelet and Arthur Palmer for editorial suggestions. method (Horoi, 2001). Now the biggest problem to 348 KRF•2 • OK.indd 348 15.12.2009 10:58:33 cuTTers and pinnacles in The saleM 28 liMesTone of indiana Arthur N. PALMER The Salem limestone of Indiana, USA, of Missis- stone. From the standpoint of karst, the Salem is sippian (early Carboniferous) age, is well known as noted for its well-developed epikarst exposed in a source of building-stone (Figures 1, 2, 3). Many the sawed faces of quarries and in roadcuts. As architectural landmarks, such as the Empire State shown in Figures 2 and 3, the epikarst consists Building in New York City, are faced with this mainly of high-relief joint-guided fissures ( cutters) with intervening fin-shaped pinnacles. These fea- tures are generally covered by soil and represent a subsoil version of kluftkarren. The quarry faces  provide excellent cross sections of the epikarst, and where solutional features or insoluble content have interfered with the economic quality of the rock, numerous abandoned blocks provide multi-  dimensional views. Previous reports give general descriptions of the cutters and pinnacles in the Salem limestone, but with few details (e.g. Malott, 1945; Powell, 1961). Although this present paper expands on  these early papers, it is based on limited field work and many questions remain for future detailed study. Regrettably, access to quarries is increas- ingly difficult because of liability concerns, and abandoned quarries are becoming overgrown with vegetation. Figure 1: Location map and cross section showing the distribution of the Salem limestone in southern Indi- ana. Cross section is drawn east–west at a location be- Geologic setting tween Bloomington and Bedford. 1. Borden Group (silt- stone); 2. Harrodsburg limestone; 3. Salem limestone; 4. The Salem consists mainly of a massive, granu- St. Louis limestone; 5. Ste. Genevieve and Paoli lime- stones; 6. younger strata, mainly quartz sandstones and lar limestone (skeletal grainstone) deposited in shales. All strata shown are of Lower Carboniferous age. shoals and inter-shoal environments (Brown, 349 KRF•2 • OK.indd 349 15.12.2009 10:58:34 Karst Rock Features • Karren Sculpturing Figure 2: Cutters and pinnacles exposed in the sawed face of a dimension-stone quarry in the Salem limestone, Law- rence County, Indiana. This is part of face 3 used in the data analysis. The measured sample includes all visible cut- ters in this photo, plus their downward extensions. The sample includes what appear to be isolated pockets, be- cause they connect with cutters at higher elevations. 1990). Its total thickness averages 20‒30 m, al- breaks in the building-stone facies, even though though the building-stone facies rarely includes they involve little or no change in depositional more than half of this thickness. The building- texture. The conspicuous ledges in the quarry stone facies contains less than 1% insoluble ma- faces are determined mainly by successive stages terial. It is traditionally considered oolitic, but in of quarry deepening, rather than by stratification. fact it is composed almost entirely of fossil frag- The Salem, along with its underlying and over- ments dominated by foraminifera. The Salem has lying carbonate strata, is exposed in the low-relief gradational contacts with the underlying Har- karst surface known locally as the Mitchell Plain rodsburg limestone, a massive bryozoan-rich (Figure 1). This is correlative with the more ex- limestone, and with the overlying St. Louis lime- tensive Pennyroyal plateau of Kentucky. The local stone, a thin-bedded limestone containing inter- structural dip is very gentle and averages 0.25‒0.5° bedded dolomite, shale, and chert. The building- to the west-southwest. The Salem contains promi- stone facies is best developed between the cities nent vertical joints that form two major sets, with of Bloomington and Bedford (Figure 1). To the spacings of about 2‒10 m. The strike directions of north the Salem is covered by Pleistocene gla- the dominant joint sets are nearly perpendicular cial deposits, and to the south it becomes thinner to each other, at 70‒95° and 170‒185°. The former bedded and impure. set, which is dominant, is nearly parallel to the dip. Temporary pauses in Salem deposition allowed Joints in both sets have been enlarged by epikarst the development of lithified surfaces known as dissolution, and they also determine the trends of hardgrounds. They provide the only significant many underlying caves. 350 KRF•2 • OK.indd 350 15.12.2009 10:58:34 Arthur N. Palmer, Cutters and pinnacles in the Salem limestone of Indiana Figure 3: Parts of faces 1 (top) and 2 (bottom) used in the data analy- sis. Cutters are narrow- er and more closely spaced than in Figure 2. In face 1 (top), the meas- ured sample includes the six main cutters but excludes the thin fissure at right. Width of view is 9 m. In face 2 (bottom), the sample includes only the two largest cutters. The narrow fis- sures are simply sawed grooves. 351 KRF•2 • OK.indd 351 15.12.2009 10:58:36 Karst Rock Features • Karren Sculpturing Field and laboratory observations gap between the soil plug and the bedrock walls. Genetically, these networks and enlarged pores Cutters in the Salem limestone are solutionally resemble maze caves formed by floodwater (Pal- widened fissures that decrease in width down- mer, 1975). ward. Some have smaller fissures extending out- One might expect the cutters to be best devel- ward from them at various dip angles, and in local oped beneath topographic lows, where infiltration areas some cutters are surrounded by zones of so- can concentrate. But this seems not to be the case, lutional y enlarged primary pores that diminish in because exposures show no consistent relation to size away from the cutters. The networks of sub- the configuration of the land surface. Some of the sidiary fissures and enlarged pores are apparently most conspicuous cutters are located beneath con- formed by periodic flooding of epikarst fissures vex topography. This suggests that cutter growth during high flow (Figure 4). This tendency is en- depends mainly on local infiltration through the hanced where shrinkage of drying soil produces a overlying soil. Figure 4: Enlarged pores in the bedrock wal s of a cutter, the result of periodic floodwater re- charge. Width of view is 1.2 m. 352 KRF•2 • OK.indd 352 15.12.2009 10:58:37 Arthur N. Palmer, Cutters and pinnacles in the Salem limestone of Indiana Many cutters have an erratic variation in width conduits perched on sparse bedding-plane part- with depth (Figure 2). Lithologic control is very ings and relict hardgrounds. Soil in the cutters small in the massive Salem because of the nearly varies from yellow quartz silt to red, blocky clay uniform texture and low insoluble content, and composed of illite, kaolinite, quartz, and goethite. most of the width variations are caused instead The red clay appears to be residual from carbon- by perching of vadose water on poorly permeable ate weathering, because it has the most intimate soil plugs. However, in the relatively impure tran- contact with the rock surfaces. In composition the sition beds at and around the St. Louis contact, red clay is typical of insoluble materials from the petrographic analysis shows that cutters are con- St. Louis. The residuum cannot come exclusively spicuously narrower where the rock is dolomitic from the relatively pure Salem, because too lit- or has a high insoluble content. In places the basal tle of that rock has been dissolved in the epikarst St. Louis is both dolomitic and high in insoluble to supply more than a tiny percentage of the soil. content (up to 60%). Epikarst is greatly subdued Furthermore, the insoluble content of the Salem is in these beds. Instead the rock tends to weather limited almost entirely to quartz silt. There were into small fragments, whose combined surface probably several sources for the yellow silt fill, be- areas are large enough to consume nearly all the cause the Mitchell Plain was once covered by more aggressiveness of infiltrating water within short than 10 m of Pliocene alluvial, lacustrine, collu- flow distances. Where the impure St. Louis cap vial, and residual deposits (Powell, 1964; Palmer approaches a thickness of roughly one metre, fis- and Palmer, 1975; Granger et al., 2001). Quater- sures in the underlying Salem diminish greatly in nary loess was later added to the mix. number and width (Figure 5). Where the St. Louis The effect of these varied soil materials on infil- exceeds 2 m, fissures in the Salem are virtually ab- tration has not been quantified, but comparison sent. with similar materials suggests that the blocky clay Figure 5: Variation in cutter development in the Salem limestone vs. thickness of overlying impure St. Louis lime- stone. Widely spaced cutters, such as those in Figure 2, has an inherent hydraulic conductivity less than tend to be wide and very irregular in their upper about 10-8cm/sec. The conductivity of the sandier sections. In comparison, closely spaced fissures, soil is orders of magnitude greater. However, de- as well as the lower sections of widely spaced ones, termining the effective hydraulic conductivity of taper downward in a more uniform manner (Fig- the epikarst soil is greatly complicated by macro- ure 3). A few fissures lead downward to sinuous pores and shrinkage cracks, especially within the 353 KRF•2 • OK.indd 353 15.12.2009 10:58:38 Karst Rock Features • Karren Sculpturing Salem epikarst become narrower downward at a roughly exponential rate, which is in accord with the diminishing dissolution rate as infiltrat- ing water approaches saturation. Also, the over- all development of epikarst diminishes sharply with depth, partly because of the decreasing width and spatial frequency of initial fractures and partings in response to erosional unloading (Williams, 1983; Klimchouk, 2000b). At shallow depths, dissolution is fairly uniform along many alternate flow paths, regardless of their discharge, so a network of widened fissures and pores is typi- Figure 6: Irregular porosity in the wal s of a cutter along a cal. The origin of epikarst resembles that of maze former sulphate zone. Width of view is 70 cm. caves, where aggressive water follows short paths through carbonate rock (Palmer, 1991). Farther below the surface, as water approaches calcite saturation, the enlargement is highly selective clay. The conductivity is also a function of soil because the most rapid dissolution takes place moisture content, which varies with time. Macro- along paths of greatest discharge. Thus, below pores transmit water readily if there is ample sup- the epikarst, only the most favourable paths are ply from the surface, but they serve as flow barri- enlarged significantly. During much of the year ers during slow, diffuse infiltration, because their there is no net infiltration, owing to evapotranspi- capillary potential is much higher than that of the ration of soil moisture. However, capillary mois- surrounding silt and clay. Water is drawn to areas ture can still continue to dissolve the bedrock at of low capillary potential (small pores, low mois- these times. ture content), while the larger pores conduct water only during the wettest of conditions. In places the Salem epikarst contains irregular Geochemical modelling of cutters porosity and textures that are clearly related to now-vanished sulphates (Figure 6). Calcite and A finite-difference model was devised to simu- quartz pseudomorphs of evaporites are common, late cutter growth. The real system is highly com- as are moldic pores from sulphate dissolution. plex, with large temporal and spatial variations Gypsum is common in these beds farther down- in recharge, soil character, moisture content, CO2 dip, where they are protected by thick overlying production, capillary potential, etc. A model that rocks. Local acid sources, such as pyrite oxidation, includes this degree of complexity would be dif- are demonstrated by abundant iron oxides, many ficult to interpret and would lead to questions as of which are pseudomorphic after pyrite. The deep to whether the proper boundary conditions had red and yellow colouring of the soil (terra rossa) is been chosen. For this reason, the simplest possible derived from the pyrite oxidation. The pyrite was model was used, so that discrepancies between it almost certainly produced by reduction of the and real fissures could be interpreted in a broad original sulphates. Thus these colours, so typical qualitative manner. of the karst of the east-central U.S., may be de- The chosen computerized model is similar to rived in part from now-vanished sulphates by way that used to simulate the enlargement of water- of intermediate iron sulphides. filled fissures (see Dreybrodt, 1990; Palmer, 1991; As viewed in vertical exposures, cutters in the Gabrovšek, 2000; and relevant chapters in Klim- 354 KRF•2 • OK.indd 354 15.12.2009 10:58:39 Arthur N. Palmer, Cutters and pinnacles in the Salem limestone of Indiana chouk et al., 2000), except that the epikarst system more directly on discharge, which differs greatly is open, rather than closed to CO exchange, be- among the various flow paths. 2 cause its water is in continuous contact with soil One of the major factors governing the disso- CO . An idealized vertical fissure 20 m from top lution of limestone is the specific discharge (q) 2 to bottom was divided into 0.1 cm depth incre- through the fissure. This is equivalent to volumet- ments. A uniform amount of vertical flow was ric discharge (volume/time) per unit fissure width. introduced at the upper end. In each depth incre- It is also equivalent to velocity x soil porosity x fis- ment, the rate of solutional widening and result- sure width. In the model, pCO = 0.03 atm and 2 ing percentage of calcite saturation were calcu- temperature = 15°C, both of which are typical of lated. The resulting water chemistry was passed soils in the Mitchell Plain (Miotke, 1974). Specific from each increment to the underlying one, and discharges (q) were varied between 0.1 and 10 cm2/ the calculations were repeated until the bottom of sec but were held constant during each simula- the fissure was reached. The entire sequence was tion. Model results are shown in Figure 7 for sev- repeated in one-year time steps for a total of 1,000 eral values of specific discharge. On the semi-log years of simulated time. It was assumed that mass plot, the slope of fissure depth in relation to log transfer was rapid enough through the soil that of fissure width should remain roughly constant dissolved species were uniform across the entire as the fissure enlarges, as long as q remains con- fissure width at any given elevation. This is highly stant. There is little tendency for q to increase as unlikely, but it provides a standard to which the the fissures widen, because their catchment areas actual conditions can be compared. remain nearly constant. The kinetic variables of Plummer et al. (1978) Several quarry faces of Salem limestone were were used, because their measurements provide chosen for analysis on the basis of accessibility the maximum dissolution rates typical of open systems. They used turbulent flow in their experi- ments, which is not appropriate for infiltration through soil. But according to Plummer and Wig- ley (1976), turbulence has little effect on dissolu- tion rate as pH rises above about 7, a value that is rapidly reached by seepage water in karst. As water approaches equilibrium with dissolved calcite, the reaction order changes from about 1‒2 to about 4, at roughly 70% saturation (Plummer et  al., 1978; see Palmer, 1991, for a summary of their experimental results, and Dreybrodt, 1990, 1996, for an alternative approach). At low-order kinetics (saturation ratio less than about 70%), the dissolu- tion rate is relatively rapid. The onset of high-or- Figure 7: Idealized curves of fissure enlargement from der kinetics farther downflow sharply reduces the finite-difference model ing, showing rate of solutional dissolution rate, because the base number (1 – C/ widening of epikarst fissures vs. depth and specific Cs, where C/Cs = saturation ratio) is less than 1.0. discharge. pCO = 0.03 atm, T = 15˚C, and saturation 2 Dissolutional enlargement of openings is fairly concentration = 290 mg/litre CaCO equivalents. The 3 uniform at low saturation ratios, because it is not inflection points in the curves indicate the transition so dependent on discharge rates. At high satura- from low-order to high-order kinetics with depth (high and low dissolution rates respectively). In this over- tion ratios the enlargement rate varies much more simplified model the profiles are the same, regardless between alternate flow paths because it depends of whether the fissures are soil-filled or open. 355 KRF•2 • OK.indd 355 15.12.2009 10:58:40 Karst Rock Features • Karren Sculpturing 0 files were then compared with the idealized pat- terns predicted by the model. 5 face 1 face (m) Model interpretation face 2 face 3 The model shows the shapes of ideal cutters after 10 1,000 years of continuous development (Figure 7). depth below sur These were compared to the actual cutter cross sections viewed in the quarry walls (Figure 8). Be- 15 cause the actual cutters are of unknown age, their –1 0 1 2 3 proper positions on the graph are uncertain. Fur- log fissure width (cm) thermore, the water that enters the cutters has Figure 8: Cumulative data for epikarst fissures in three already acquired a certain dissolved load before quarry faces (shown partly in Figures 2 and 3). it starts its downward journey. On the semi-log plots, the curves for real cutter geometries retain their shape when they are shifted vertically. In ad- dition, they retain their relative sizes if they are  shifted horizontally (i.e. if a fissure is 10 times wider at one elevation than at another, this ratio does not change). The best fit was obtained when the field data were shifted downward and to the left, to match the modelled curves of depth vs. en- largement rate. Data for the cutters in quarry faces 1 and 2 fit the curve for the modelled plot for q = 0.1 cm2/sec. Those for face 3 fit best to the plot for q = 3 cm2/sec. This seems appropriate, because the fissures in face 3 are wider than those in faces 1 and 2, and they are also more widely spaced, so they accommodate a greater amount of recharge than those in faces 1 and 2. The measured profiles Figure 9: Matches between field data and model curves. represent the actual shape after an unknown time Note the lateral and vertical shifts in data from their original positions in Figure 8, which were required to fit of development. The amount of horizontal shift the ideal graphs. should be a crude indicator of their age (relative to the 1,000-year standard). The amount of down- ward shift should be proportional to the amount of dissolution that has taken place before the water and quality of cutter-and-pinnacle development. enters the fissures. The 4 m downward shift re- Others were disqualified because their cutter quired for the cutters in face 2 may reflect the fact profiles were distorted by differential dissolution that the uppermost bed (about 3 m thick) has been in the vicinity of local dolomites and shaly beds. stripped away by quarrying. To smooth out non-systematic irregularities, the The resulting fits, shown in Figure 9, are sur- mean width of all fissures exposed in each face prisingly good. This is unexpected, because the was calculated as a function of depth below the idealized model is obviously wrong. As expected, surface (Figure 8). The resulting composite pro- the mean taper of the cutters is strongly controlled 356 KRF•2 • OK.indd 356 15.12.2009 10:58:42 Arthur N. Palmer, Cutters and pinnacles in the Salem limestone of Indiana by dissolution kinetics. But the attempt to under- gentler taper, by high-order kinetics. Where cut- stand the nature of cutter growth by contrasting ters are closely spaced, each receives less recharge, the field data to a purposely over-simplified model and most of their water has reached high-order seems to have failed. kinetics by the time it enters them, owing to dis- There are some obvious differences in behavi- solution at the soil/bedrock contact and with car- our between the model and the field conditions bonate fragments within the soil. The fact that the that are not illustrated in the graphs. In the field, data for the narrow cutters needed to be shifted aggressive water is able to penetrate more deeply downward to fit the ideal curves of the model than in the model, because water in contact with demonstrates that the water has lost much of its the real limestone walls does not efficiently trans- aggressiveness before it enters the narrow cutters. fer its dissolved load to the interior of the soil plugs. The apparent close fit to the model may indicate that much of the infiltrating water moves along Age of cutters and pinnacles the walls of the fissures along the soil/bedrock contact. This topic needs further investigation. In Figure 9 the amount of lateral shift required to It appears from Figure 9 that the taper of the fit the field data to the modelled curves should be cutters depends partly on whether the solvent proportional to the age of the cutters. This is only a water experiences low- or high-order kinetics. The crude approximation. The amount of lateral shift data from face 3 fits best to the low-order upper suggests that the cutters in faces 1 and 2 required part of the graph, whereas the other faces fit a between 75,000 and 500,000 years of cumulative higher-order dissolution graph (i.e. slower disso- infiltration at small q values and with slow (high- lution rates). At high recharge rates (wide cutters) order) reaction kinetics. The data for face 3 re- the water enters the cutters while it still retains quired little horizontal shift, which suggests that much of its aggressiveness. Where the reaction its cutters are younger than those in faces 1 and order increases, the cutters become narrower. The 2, but with greater q and rapid (low-order) kinet- result is a bulbous upper part leading downward ics. These assumptions need closer scrutiny before to a narrower and more evenly tapered fissure they can be taken seriously. (Figure 2). In narrow cutters, by contrast, the re- The low-relief Mitchell Plain surface, in which charge rates are low, and by the time the water en- the Salem epikarst is developed, dates mainly ters the cutters most of the dissolutional capacity from the late Tertiary period (Powell, 1964; Palm- has already been consumed in lowering the soil/ er and Palmer, 1975). In Kentucky, passage levels bedrock surface. Therefore, in the narrow cutters in Mammoth Cave that correlate with the Pen- the water follows high-order dissolution kinet- nyroyal plateau (a continuation of the Mitchell ics over their entire length. Still it is unlikely that Plain) contain quartz-rich sediments that show cutter shape is an accurate indicator of low-order cosmogenic radionuclide ages of several mil- vs. high-order kinetics, because the depth of the lion years (Granger et al., 2001). These passages transition varies with the q value, which in turn were filled during a major period of aggradation varies greatly with time. about 2.6 million years ago (estimate revised from Although the quantitative details of the model Granger et al., 2001). Sediments accumulated to are not realistic, the general relationships seem thicknesses of at least 10 m, both in the cave and valid. Wider cutter spacing provides greater dis- on the karst surface. These sediments are well pre- charge rates because the catchment areas are served in many upper-level cave passages, and in- larger. The upper parts of widely spaced cutters place remnants of the sediments can still be iden- enlarge rapidly by low-order kinetics, while the tified at the surface (Ray, 1996). lower parts are enlarged more gradually, with a Where does the Salem epikarst fit into the 357 KRF•2 • OK.indd 357 15.12.2009 10:58:42 Karst Rock Features • Karren Sculpturing picture? Because of the low structural dip of the dissolution kinetics. Because of the small size of Salem limestone in Indiana, small amounts of the sample and the idealized nature of the model, denudation can cause great lateral displacements the quantitative results must be considered tenta- of geologic contacts. As shown in Figure 5, the tive. Force-fitting of data to idealized curves does St. Louis limestone had to be removed by erosion not prove a functional relationship. Nevertheless, before the Salem epikarst could begin to develop. this study opens a line of inquiry that may help And yet, where local land surfaces cut discordantly to explain cutter development and shapes. Fu- across both the Salem and St. Louis, cutters in the ture field work should include measurements of Salem extend very close to the St. Louis contact. soil character and infiltration patterns. The role of This shows either that the contact has not shifted now-vanished sulphates should also be considered. greatly since the epikarst began to form, or that the cutters develop quite rapidly. A maximum age of half a million years for the cutters and pinna- Acknowledgements cles (suggested by the model) is consistent with the antiquity of the overall Mitchell Plain surface Many thanks to Margaret Palmer, Oneonta, NY, as determined from cave-sediment dating. for performing the petrographic and X-ray dif- fraction analyses, and for help with the field work. She tallied the cutter cross sections in the quarry Conclusions faces independently, to avoid personal bias in ob- taining fits to the finite-difference model. Richard Cutters and pinnacles in the Salem limestone shed Powell and Todd Thompson of the Indiana Geo- light on the evolution of the entire Mitchell Plain. logical Survey, Andrew Cheney of C & H Corpo- The simplified modelling approach used here has ration, and Daniel Chase of Indianapolis were es- shown that cutter morphology is closely tied to pecially helpful in directing us to field sites. 358 KRF•2 • OK.indd 358 15.12.2009 10:58:42 Types of Karren and Their genesis 29 on The VelebiT MounTain Dražen PERICA and Tihomir MARJANAC Numerous karst forms are found on carbon- Velebit Mt has attracted geologists and geogra- ate rocks of the Velebit mountain ridge. There phers to study its karst features. occur numerous varieties of karren (“grižine” in Karren grikes of the Velebit Mt were studied Croatian), whose genesis and number are con- by Simeonović (1921), Cvijić (1926, 1927), Poljak ditioned by interaction of geology, climate, soils, (1929a, b), Rogić (1958), and more recently Bognar flora and geomorphology, as well as by anthro- and Blazek (1986), Perica (1998), and Perica et al. pogenic influences. The term grižine encompass- (1995, 1999). es various types of small-scale corrosional forms such as various types of karren, grikes, solution pans ‒ kamenitzas (“kamenice” in Croatian), root Geology of the Velebit Mt karren, karren wells, karst tables, pot-like karren, and the karren locally referred to as “sige” (tufa- The Velebit Mt is a part of the extensive Dinaric like karren). Several types of karren display dif- karst region. It consists of a wide variety of strati- ferences as a consequence of their genesis; directly graphic units, ranging in age from Carboniferous under the atmospheric water, or owing to subcu- to Quaternary (Sokač, 1973). The oldest strati- taneous corrosion by the water which was second- graphic unit is Middle Permian, which is exposed ary enriched by soil-derived biogenic CO . only in the central mountain area, and forms a 2 tectonically reduced anticline core. The units of Paleozoic age are conformably overlain by Meso- Previous research zoic units. Triassic deposits are predominantly represented by dolomites, whereas the Jurassic The first true naturalist to study the Velebit Mt deposits comprise various types of shallow ma- karst was apparently Hacquet (1785). Fras (1835) rine limestones. The deposits of Cretaceous age provided first descriptions of some speleological are represented by carbonate platform carbon- features on the Velebit Mt, and the first geologi- ates. These units are transgressively overlain by cal research were conducted by Vienna Geological Tertiary shallow marine carbonates and deep ma- Institute in 1862. The first compilation of geologi- rine clastics. The youngest Tertiary unit, though cal data was prepared by Hauer (1867‒1871) who of controversial age and origin, are the so-called published a geological map of Austro-Hungarian Jelar-beds, commonly referred to also as the Jelar- Monarchy at a scale of 1:576,000. Ever since, the breccia. The youngest deposits of the Velebit Mt 359 KRF•2 • OK.indd 359 15.12.2009 10:58:42 Karst Rock Features • Karren Sculpturing are Pleistocene glacial and glacio-fluvial deposits, tioned by prevalence of carbonate rocks on the preserved on the mountain’s highest parts. Velebit Mt, namely by the high CaCO content. 3 The Jelar-breccia (first described by Bahun, The size of exokarstic and endokarstic features in- 1974) is the lithostratigraphic unit bearing the dicates that they were most commonly formed in best developed karst features. The Jelar-breccia is rock successions dominated by limestones, such a massive, or thick-bedded carbonate rock, and as the Middle and Upper Jurassic limestones with comprises predominantly angular, poorly sorted a high percentage of pure CaCO . 3 debris. It is commonly grain-supported, although Structural predisposition implies abundance of matrix-supported varieties occur locally. The bre- primary and secondary fissures and voids, as well ccia has a carbonate matrix, which is gray to red- as inclination of beds. “Classical” carbonates (pri- dish coloured. The stratigraphic composition of marily Jelar-beds) are characterized by numerous debris is varied. Clasts of Cretaceous limestones primary fissures. Bedding commonly affects the and dolomites are the most common, and sub- size and shape of karst features in layered rocks ordinately there occur Triassic carbonates and (e.g. asymmetrical dolines), where steepened bed- Paleogene limestones. The debris grain sizes are ding promoted more intensive corrosion primari- very variable, and clasts range from a few mm up ly along diastromes. Tectonics created fault planes, to several decimetres in size, but also up to several fractures and fissures, which initiated formation metres or more across. of numerous karst features. One of the prominent The area covered by Jelar-beds (breccia) reaches characteristics of these karst features (created ca. 690 km2 (Figure 1), but its thickness is poorly along secondary fissures) is their elongation along known. The only available direct account on its the strike of the fissure. thickness was acquired by drill-holes during con- Formation of karst features and their frequency struction of the St. Rok road tunnel (Matičec et also depend on slope inclination. The karst fea- al., 1999) on the southern part of the Velebit Mt, tures are more varied and larger on horizontal which documented the thickness of Jelar breccia and gently inclined slopes (up to 12°). On steeper of 300 m. Not only is the thickness of Jelar-breccia slopes, karst features are much rarer, primarily poorly known, but also its age remains a contro- because of faster superficial drainage, and conse- versy. It was treated as an Eocene-Oligocene unit quently decreased corrosion intensity. by authors of the General Geological Maps of Croatia (Ivanović et al., 1973; Mamužić et al., 1969; Šušnjar et al., 1970), which means that it represents Climatic conditions a post-Eocene flysch sedimentary unit. However, Tari Kovačić and Mrinjek (1994) interpreted the Climatic factors (rainfall and temperature) are breccia as a post-Cretaceous but pre-flysch sedi- important because of their influence on dura- mentary unit of the Early Eocene age. The young- tion and intensity of corrosion. Rainfall increases est debris found in Jelar-breccia are clasts of Early with altitude, but it is unevenly distributed. The Eocene Alveolina limestones, which constrain its whole Velebit Mt is characterised by a mediterra- age to Upper Lutetian-Bartonian span (Sakač et nean pluviometric regime, with more abundant al., 1993; Vlahović et al., 1999). rainfall in the cold than in the warm season. The lowest rainfall is in the coastal part of SW moun- tain slope (ca. 1,200 mm annually). The rainfall Superficial corrosion features and non-proportionally increases with altitude. The karren of Velebit Mt rainfall increase is lowest on the north Velebit Mt where the 2,000 mm isohyet attains altitude The development of karst relief has been condi- of about 1,400 m. Gradually, the 2,000 mm iso- 360 KRF•2 • OK.indd 360 15.12.2009 10:58:42 Dražen Perica and Tihomir Marjanac, Types of karren and their genesis on the Velebit mountain ‡ †‡‚       ­€ ‚ƒ ƒ„ ƒ  Figure 1: Simplified geological map of the Velebit Mt and neighbouring areas, after Oluić et al. (1972). Circle marks the Paklenica National Park. hyet lowers to an altitude of 900 m on the south (from mist, fog and clouds) in the crestal part of Velebit Mt, where rainfall reaches ca. 3,500 mm the mountain, which accumulates orographically. at the mountain crest (Bunovac at the leeward The contribution of this additional precipitation side of the mountain crest: 3,419 mm, Perica and on Zavižan meteorological post (1,594 m a.s.l.) in Orešić, 1997). However, the intensity of corrosion period 1955‒1965 was 249% (Kirigin, 1967), with is significantly affected by additional precipitation significant differences between winter (343%) and 361 KRF•2 • OK.indd 361 15.12.2009 10:58:44 Karst Rock Features • Karren Sculpturing summer (171%) seasons. On the NE flank (Lika a major role in shaping of the relief. A short veg- side) of the mountain rainfall gradually falls etative period and a large number of cold, icy and below the isohyet of 2,000 mm. freezing days, favour soil dryness and conditions The thermal effect of the sea is restricted to the curtailing the corrosion (Perica, 1998). coastal zone and the lower SW mountain slopes, The formation of the karst relief was signifi- but it is also weakened by the presence of islands cantly affected by the climate of the last Ice Age. and the Ravni Kotari plateau which stretch paral- On the one hand karren were destroyed by ero- lel to the coast-line. The average annual air tem- sion, or reshaped, and on the other the melting of perature on the lowest part of the SW mountain snow and ice formed new karst features such as flank reaches about 15°C (Senj 14.5°C, Karlobag speleological features having drainage function. 15.6°C), and about 3°C in the crestal part (Zavižan Finally, the anthropogenic influence in the past at 1,594 m a.s.l., 3.5°C) of the Velebit mountain. was related to deforestation, when enhanced den- With an increase in altitude, the air temperature udation and soil erosion exposed rock at the sur- rapidly, but unevenly, decreases. The annual ver- face. This is particularly well seen on the Velebit tical gradient between Karlobag (30 m a.s.l.) and Mt crest and its SW slopes. In the long term, ever Baške Oštarije (924 m a.s.l.) is 0.93°C, and be- since prehistory, bad agricultural practices on tween Baške Oštarije and Zavižan (1,594 m a.s.l.) this part of the Velebit Mt caused its deforestation. it is 0.57°C. According to Rogić (1958), heating Thus, today, on the SW slope and the mountain of carbonate rocks on the lower part of the SW crest, bare and semicovered karst prevails, which mountain flank (Karlobag has 109.6 warm and commonly passes into exposed karst. At the same 39.7 hot days annually) significantly promotes time, the NW mountain slope is largely forested, thermomechanical weathering of the rocks, and and the karst covered (Rogić, 1958). exceptionally strong evapotranspiration which causes their marked dryness. The intensity of corrosion and biocorrosion is lowered in these Types and formation of karren conditions, and it completely stops in thin soils during the prolonged droughts. This assumption The Velebit Mt karst contains all types of karren, was confirmed by the study of the intensity of but here we will deal only with solution runnels corrosion at the surface by using limestone tab- developed on Jelar-breccia. lets (Perica, 1998). The corrosion intensity on the Karren formed by water corrosion represent SW Velebit Mt slope is highest in its middle part; the most widespread karst forms on the Velebit it decreases insignificantly downslope, but signifi- Mt, and their various types occur from the sea- cantly upslope. The lower and middle part of the level up to the highest parts of the mountain crest. SW slope is characterized by significantly stronger However, they are most frequently and best devel- corrosion in the soil than at the surface. Thus, the oped on the lower and middle, bare parts of the corrosion intensity at the surface and in the soil in SW (coastal) part of the mountain slope, whereas Velika Paklenica (560 m a.s.l.) reaches 1:2.55, and their development in higher parts is restricted by on Babrovača (920 m a.s.l.) 1:3.49. This increase in thermal conditions. corrosion intensity can be explained by increase The formation of various types of karren results in biocorrosion which stems from a prolonged from interaction of: a) slope inclination, b) litholo- vegetative period. gy, and c) rock fracturing. Their shape is precondi- The crestal part of the Velebit Mt is character- tioned by the relationship of slope vs. bedding in- ized by large number of cold (160.9), icy (74.9) and clination, and degree of carbonate exposure under freezing (26.3) days which favour the freezing of vegetative and soil cover. These factors controlled water in fissures, so that cryogenic processes play the corrosion, which could have acted: a) directly 362 KRF•2 • OK.indd 362 15.12.2009 10:58:45 Dražen Perica and Tihomir Marjanac, Types of karren and their genesis on the Velebit mountain by corrosion of meteoric waters, b) by corrosion of waters which percolated through the soil, and c) by corrosion of ground water. Lithology also plays a major role in the formation of karren. They are absent or small in thin-bedded carbonates, but varied and very abundant in thick-bedded varie- ties. Thus, the Jelar-breccia is characterised by ex- tensive development of karren, essentially due to its lithological characteristics. On the Velebit mountain there occur several types of karren; furrow-like karren (locally called žlibe), fissure karren (locally called škrape), ka- menitzas (solution pans) and biocorrosion pits. Ford and Williams (1989) differentiate micro- karren (smaller than 1 cm), karren (1 cm‒10 m large), and bogaz or corridor (larger than 10 m). Microkarren are developed in homogeneous fine- grained rocks by corrosion of water under capil- lary pressure. Biocorrosion due to endolithic bac- teria, lichens and mosses, also plays a major role in formation of microkarren. Cyanobacteria form small 1 mm deep hollows, which are later invaded by other organisms (Verges, 1985) that promote corrosion by production of organic acids and emission of CO . Figure 2: Typical large rinnenkarren (solution runnels) 2 in Jelar-breccia. Paklenica National Park, south Velebit Rinnenkarren or runnels ( solution grooves) (Fig- mountain. Width of view is 5.5 m. ure 2) occur at a wide range of altitudes, from the Adriatic coast up to the Velebit Mt summit. Two genetic types of rinnenkarren can be differenti- ated; a) the variety formed by corrosion of atmos- eventually enriched in carbonate, it is still corro- pheric water, and b) those formed by corrosion of sive due to inflow of rain water. Thus, the grooves water enriched with biogenic CO . are wider and deeper downslope, whereas the 2 The first karren variety (a) is characteristic of ridges get narrower and sharper. Slope inclination steeper (> 20°) slopes on denuded Velebit Mt car- also controls their shape, so they are narrower and bonates. This karren type occurs on small lime- deeper on steeper rock faces, a consequence of re- stone blocks, sometimes not more than a few tens stricted lateral and enhanced regressive water ac- of sq. decimetres in area, but the maximal devel- tion, due to accelerated run-off and water inflow opment is on the Jelar-breccia at the Velebit Mt in the groove. Gentler slopes are characterised by SW slope 500‒1,200 m a.s.l. as well as on Upper slower run-off and thus enhanced lateral corro- Jurassic limestones in Paklenica National Park sion, so that the furrows are wider and shallower, (Perica et al., 1995) where bedding and slope in- and often gently curved. clination nearly coincide. The water strongly cor- The furrows commonly attain 50 cm in width, rodes bare carbonates, and flows down the rock- and 1 m in depth, but those developed on the face. This type of karren is most frequent on rock Jelar-breccia and Upper Jurassic limestones in the faces inclined at 30°‒70°. Although the water is Paklenica National Park can be much larger (Fig- 363 KRF•2 • OK.indd 363 15.12.2009 10:58:45 Karst Rock Features • Karren Sculpturing 1992). The end result is a decreased ratio between groove width and length, compared to the pri- mary furrow. The incision of the solution flutes of the second order provides a dog-tooth shape to the ridges. Furrows become irregular on the Jelar- breccia, as a result of differential corrosion due to clast inhomogeneity and clayey-limonitic matrix. Depressions are formed in places of corrosion- prone clasts, whereas harder clast lithologies re- main as elevated remnants. The second runnel variety (b) is formed by cor- rosion of water which is enriched in biogenic CO . 2 This karren type is much rarer than the first type, described above. On the Velebit Mt we can differ- entiate two sub-types. The first sub-type is covered karren, which was formed by subcutaneous cor- rosion under the soil cover. It is well exposed in Hajdučki Kukovi area, where vegetative cover was destroyed. The covered karren occurs individually, rarely in groups, and also occurs on gentler slopes. This type of karren ( rundkarren) is characterised by a gentle half-rounded shape, which is shallow and wide. It does not widen downslope as does the first karren type. When they occur in groups, Figure 3: Steep runnels with well developed serrated they are separated by wide ridges. This shape is a ridges showing solution flutes of the second order. result of slow and long-lasting water percolation Width of view is 3.5 m. in soil, which causes even wetting, and corrosion of the rocks. However, even here the water is con- centrated by channelized flow, although to a lesser ures 2, 3). These grooves attain a few tens of me- extent than on the bare rock faces. This karren tres in length, but in the Paklenica National Park type occurs on gentle slopes inclined just a few they exceed 100 m in length and 1 m in depth. The degrees, but its width/depth ratio at the change of formation of these large grooves is also control- slope inclination has identical characteristics as led by additional water supply by condensation of do the grooves formed by direct corrosion by the atmospheric water on the carbonate substrate (Pe- atmospheric water. rica and Orešić, 1999). Lateral corrosion narrows The second karren sub-type is the humus-water- the karren ridges, which become very sharp, fre- groove ( hohlkarren), which commonly occurs sin- quently ornamented by development of solution gly. This type of groove is formed by corrosion of flutes of the second order (after Bögli, 1980). As water which percolates from pedogenic cover, and they grow, these secondary grooves are progres- is enriched with biogenic CO . Their size is smal- 2 sively shallower and wider, compared to the pri- ler than the size of atmospheric water-generated mary groove which becomes steeper-sided. This is grooves, primarily because of restricted amount of a result of decreased water supply per unit of sur- CO in the water. Locally, in canyons of Velika and 2 face, and its quicker neutralization (and increase Mala Paklenica it is possible to see that for several in carbonate concentration; Perica and Kukić, metres there is precipitation of CaCO , indicating 3 364 KRF•2 • OK.indd 364 15.12.2009 10:58:48 Dražen Perica and Tihomir Marjanac, Types of karren and their genesis on the Velebit mountain Figure 4: Lateral biocorro- sion caused by mosses forms overhung kar- ren edges. Note residual clays fil ing the bottom. Bojinac, south Velebit mountain. Width of view is 3 m. carbonate-saturated water. Slow running of sub- A common characteristic of all these karren surface water is favourable for growth of lichens types is that at steep slopes (inclined 80° or more) and mosses which contribute to biocorrosion they become shallower and narrower, as well as that widens groove sides which become overhung, semicircular in cross-section. This is due to ac- whereas the bottom fills with residual clay which celerated run-off, increased rainfall on the surface insulates it from further corrosion (Figure 4). unit, and water neutralization with consequence 365 KRF•2 • OK.indd 365 15.12.2009 10:58:49 Karst Rock Features • Karren Sculpturing of decelerated corrosion. However, this karren Karren channels formed along diastromes are type is quite rare, and ridges are commonly lack- significantly longer, and commonly deeper than ing. If the rock face reaches 90°, or overhangs, other, primarily vertical channels. groove incision ceases. This wall-karren can be Exhumed karren formed under pedologic seen on steep rock faces, pillars and cliffs, which cover are characterised by numerous elliptical or are particularly impressive in Varnjača doline on rounded hollows. Subsoil channel widths exceed Hajdučki Kukovi, Bojinac, and Velika and Mala 30 cm, rounded margins are common, bottoms Paklenica canyons. are trough-shaped and channel sides smooth. On very gentle surfaces rinnenkarren are wider Formation of this type of subsoil channel is a or completely disappear. However, locally they consequence of corrosion by water enriched with develop as meandering grooves (meandering kar- humus-generated biogenic CO . This water acts 2 ren) which can be best seen on Bojinac, Hajdučki corrosively in all directions, which affects simul- Kukovi, Kiza and Alaginac on the Velebit moun- taneous lateral expansion and deepening of the tain. Karren meanders are typically asymmetric hollows. However, when the bottom of the chan- in cross-section, seldom wider and deeper than 10 nel is covered by residual clay, the corrosion is cm, and the length can exceed 10 and more metres. suppressed, as well as deepening of the channel. Karren features are particularly common on Destruction of the woods by burning, some- gentler slopes (< 12°). By their shape, fissure and times for expansion of cultivated areas, as well as network, karren types can be differentiated. By the dip of the slopes, caused strong soil removal their genesis we can differentiate karren formed (during abundant rainfall) and deflation (bora on bare rocks by corrosion of atmospheric water, wind action) which eventually resulted in bare and karren formed under pedologic and vegeta- mountain slopes (Simeonović, 1921; Cvijić, 1927; tive cover by subcutaneous corrosion. The shape Poljak, 1929a, b; Salopek, 1952; Rogić, 1958). Bögli of karren is also controlled by the lithology of the (1980) states that exhumed karren, exposed for host rocks. Fissure karren ( kluftkarren) are more 100‒200 years, are hardly recognizable due to cor- frequently formed in medium- and thick-bedded rosion by meteoric waters. The corrosion has been Cretaceous and Upper- and Lower Jurassic car- augmented also by strong heating of the rocks at bonates, whereas network karren almost always the surface by forest fires (Perica, 1998). predominate in the Jelar-breccia. Formation of Unlike the karren formed by the subcutaneous fissure karren in bedded carbonates is primarily corrosion, the ones formed by direct corrosion associated with diastromes and diaclases. The for- by atmospheric water have significantly narrower mation of network karren is primarily controlled sharp channels (kluftkarren and furrows) and by brachyclases and leptoclases which directed crests (pinnacles). The atmospheric water corrod- corrosion (Bognar and Blazek, 1986). The forma- ing the carbonate substrate, quickly runs through tion of network karren in areas built of Cretaceous the fissures, so their width/depth ratio is much breccias and Jelar-breccias is controlled by exten- larger than of other karren formed by the subcu- sive fracturing of unbedded carbonates, various taneous corrosion. Although the grikes depth usu- grain sizes and hardness of cemented debris, and ally exceeds 1 m, that is difficult to estimate due to particularly better solubility of carbonate cement. their small width. The result are large surfaces covered by irregular- Formation of transitional karren types is re- shaped network karren (Figure 5). Network kar- lated to carbonate substrate which is partly soil- ren formed on bedded rocks have more regular covered, as on the SW slope of Veliki Golić (1,265 shapes (Perica, 1998). Their development is pre- m a.s.l.) in the Paklenica National Park. Soil-filled ferred on steeply inclined rocks, where corrosion grikes are developed along subvertical diastromes, acts along the diastromes and secondary fissures. and covered with vegetation, whereas the beds 366 KRF•2 • OK.indd 366 15.12.2009 10:58:49 Dražen Perica and Tihomir Marjanac, Types of karren and their genesis on the Velebit mountain Figure 5: Network karren (largely developed into debris-karren) developed in Cretaceous limestones, south Velebit mountain. are bare and fractured. Pedologic cover is locally glacio-fluvial fans which comprise predominantly thick enough to support growth of big trees (Peri- small (up to a few cm) debris of Jurassic limestones ca et al., 1995). and dolomites, and only locally on vegetated and Particularly large grikes, which exceed 10 m stabilized colluvial fans. The bizarre shape of the in length, and attain depths of more than several vuggy rocks are a result of selective biocorrosion metres are locally called “škarovi” and “škripovi”. of glacio-fluvial breccias, as well as washing of the Karren have developed their ultimate stage on fine-grained debris and cement. The cavernous the central and lower SW Velebit Mt slope, where karren of this type are characterized by numerous they form debris karren (locally called grohot). On shallow depressions and fenestrae (commonly akin slopes dominated by karren developed on gently to honeycomb) which have been formed by corro- inclined limestone beds, the formation of debris sion of more soluble debris (Perica et al., 1995). karren is promoted by lateral corrosion which Solution pans ( kamenitzas) are developed on prograded along diastromes. On steeper slopes areas with bare Velebit Mt karst. They are several gravitational processes have formed colluvial centimetres up to several decimetres deep, and aprons made of debris karren. the width and length span from several centime- “Sige” -karren (Figure 6) represent a particular tres up to several metres. In exceptional cases their type of network karren, and are in Paklenica Na- depth reaches more than 1 m, and the width can tional Park known also as “vuggy rock”. They are exceed 10 m. Gams (1974) differentiated kameni- formed predominantly on gentler slopes (< 12°) of tzas which were formed under the pedologic cover 367 KRF•2 • OK.indd 367 15.12.2009 10:58:51 Karst Rock Features • Karren Sculpturing Figure 6: Tufa-like “sige”- karren (“vuggy rock”) is a particular type of karren formed in glacio-fluvial breccia in the Paklenica National Park. by subcutaneous corrosion, and those formed on The first phase in the growth of kamenitzas bare carbonate surfaces. The former are formed is characterised by prolonged wetting by atmos- by biocorrosional deepening of primary hollows, pheric water and its corrosive action in hollows and are characterised by a hemisphaerical cross- without pedological cover. The width of these section, and lack overhung margins which is oth- hollows is just several centimetres, and the depth erwise a characteristic feature. Formation of the only a few millimetres. Gradual lowering of the second type of kamenitzas, or true kamenitzas water level by desiccation, commonly enhanced sensu Gams (1974) is related to hollows on flat or biocorrosion which comes from decay of drifted gently inclined surfaces which were formed under organic matter (leaves, grass, algae, lichens and the pedological cover by subcutaneous corrosion. mosses) and promote corrosion which progrades These hollows Sweeting (1966) and Ford and Wil- towards the central and lower parts of the hollows liams (1989) attribute to karren as “solution pans”. by gradual steepening of their margins. As the One of their characteristics is their more frequent water gradually becomes more saturated by dis- occurrence on inhomogeneous rocks, which is the solved carbonate in its lower part, and the upper- cause of their irregular shape (Ford and Williams, most part absorbs atmospheric CO , the kamenit- 2 1989). Sweeting (1966) holds that in just 10 years za is widest in its central part. Gradual lowering of they can reach depth of 3‒5 cm. the water level, and increase in saturation by dis- Development of the second kamenitza type solved carbonates towards the kamenitza bottom, (true kamenitzas; Gams, 1974) can be attributed decrease the intensity of corrosion, so near the to four phases, the first two being constructional bottom kamenitzas are again narrower. Simulta- and the second two degradational. The size, pri- neously with kamenitza formation, there forms its marily the diameter, of this kamenitza type is con- outlet groove, which is in this first phase almost trolled by the dip of rock face on which the karren negligible. The groove is formed only during short is formed. They are significantly larger on flat or periods of water run-off from a kamenitza, during gently inclined surfaces, unlike on steeper slopes. and shortly after the rainfall, and occurs in the Gavrilović (1964) reported their occurrence even lowest marginal part of a kamenitza. on slopes steeper than 35°. The second phase in the formation of kameni- 368 KRF•2 • OK.indd 368 15.12.2009 10:58:53 Dražen Perica and Tihomir Marjanac, Types of karren and their genesis on the Velebit mountain Figure 7: Kamenitza “Jezerce”, Bojinac, south Velebit mountain. Note remnants of old kamenitza margins above the water level. Width of view is 7 m, in the middle. tzas is characterised by increasingly stronger lat- groove becomes more pronounced, the water level eral widening, caused by corrosion. Its vertical de- in the kamenitza progressively falls due to com- velopment is negligible or stagnant because of ac- bined run-off and evaporation, and the water be- cumulation of residual clays and decayed organic comes increasingly more saturated with dissolved matter at the bottom. Simultaneously, groove carbonate. The size (diameter) of kamenitzas is incision becomes increasingly stronger. Although controlled by the rate of groove incision, because the incision of grooves starts more slowly than when the bottom of a kamenitza levels with the the incision of kamenitzas, it gradually becomes groove, it stops growing and starts degrading. quicker than the incision of kamenitza bottoms. The third phase in the formation of kameni- At the end of this phase, the run-off through the tzas is characterised by the onset of their destruc- 369 KRF•2 • OK.indd 369 15.12.2009 10:58:54 Karst Rock Features • Karren Sculpturing Figure 8: Water-filled ka- menitza “Samogred”, south Velebit moun- tain, is 1 m deep and 8 m long. Width of view is 6 m. tion. This is caused by progressive widening of the The last, fourth phase in kamenitza develop- groove by lateral corrosion, and by destruction of ment, is characterised by complete destruction of the overhung part of kamenitza margin by regres- its overhung margin and widening of the groove, sive corrosion of the meteoric water. The kamenit- which commonly reaches (sometimes even ex- za margin gradually loses its overhung shape and ceeds) the kamenitza width. In this way kameni- becomes rounded, whereas the groove becomes tzas become shelf-like or amphitheatre shaped. wider and trough-shaped. The once overhung margin becomes a place of 370 KRF•2 • OK.indd 370 15.12.2009 10:58:56 Dražen Perica and Tihomir Marjanac, Types of karren and their genesis on the Velebit mountain Figure 9: Pot-like karren. See lens cap for scale. groove incision, which itself becomes meandering interest (Figures 7, 8). The remnants of overhung due to decreased slope inclination. This type of re- rims show that it was formed by merging of 5 ka- shaping of kamenitzas is found on Bojinac in the menitzas in various stages of development, which Velika Paklenica National Park. formed one within the other. The oldest seems to If a kamenitza reaches a fracture during its be 14 m long, 7 m wide, and about 1 m deep. The growth, it stops developing prematurely, because fourth kamenitza, which is at the transition be- the water widens the fracture and creates a tube- tween the second and third development phase, is shaped channel which eventually drains the water 7 m long, 4.5 m wide and 30 cm deep. The fifth away. In this way, a kamenitza stops developing, kamenitza is the youngest, at the first phase of de- but Gavrilović (1964) holds they may still form velopment, and is smaller than the previous ones. karst wel s. The frequency and size of kamenitzas is con- Kamenitzas developed on Jelar-breccias on the trolled, in addition to lithology, also by the cli- Velebit Mt are commonly irregular-shaped and mate. Kamenitzas are relatively rare on the lower characterised by uneven floors, as a result of clast parts of the SW slope of the Velebit mountain and as well as matrix/cement inhomogeneities. its summit, because the lower slope is character- The largest kamenitzas are developed on south- ized by strong evaporation (1,000 mm) and rela- ern Velebit Mt on Jagin Kuk and Prosenjak locali- tively low rainfall (1,200 mm) (Perica and Orešić, ties. Here they reach diameters of several metres 1999) which has a negative influence on corrosion and depths of more than one metre. The kameni- development. On the Velebit Mt summit, the low tza “Jezerce” (a pond in Croatian) is of particular temperature (average annual temperature reaches 371 KRF•2 • OK.indd 371 15.12.2009 10:58:59 Karst Rock Features • Karren Sculpturing Figure 10: Karren well “Čelinka”, Paklenica Na- tional Park. Width of view is 2.5 m. 3.5°C) presents a factor which limits development trap atmospheric water and will be subjected to of kamenitzas because at these temperatures cry- extended corrosion and gradually transform into ogenic processes are dominant. Kamenitzas are other types of karren (e.g. kamenitzas). The sec- most frequent, and largest, on the middle part of ond type of biokarst features is morphologically the Velebit Mt SW slope (400‒1,100 m a.s.l.) where identical to the first type, but is formed by biocor- mild temperatures prevail (average annual tem- rosion of bare rock faces by lower plants. Algae perature of 10°C), with relatively high rainfall receive moisture from the air and, in addition to (1,500‒2,000 mm) and lower evaporation. biogenic CO , they also receive atmospheric CO 2 2 Locally, there occur small depressions with from meteoric water (Ford and Williams, 1989). diameters of a few millimetres up to several cen- However, deepening of this type of karren is com- timetres which were formed by biogenic corro- monly obstructed by residual clay at their bottom. sion ‒ root karren. Two types of biokarst features The lower parts of the SW Velebit Mt slope can be differentiated: a) formed under the pedo- which are predominantly built of the Jelar-breccia, logic cover by corrosive action of plant roots, and are characterized by occurrence of pot-like karren b) formed on bare rocks by corrosion created by (Poljak, 1929a). The primary feature of this karren bacteria, lichens and mosses. The first type of type is the occurrence of chains of small pots (Fig- root karren formed by humic acids is quite rare. ure 9). The pots are rarely wider than 5 cm, and They are rapidly destroyed during denudation 2‒3 cm deep. They occur most often on rock faces of mountain slopes even by corrosion of atmos- inclined 20‒50°, which were previously exposed to pheric water. Bögli (1980) states that this type of corrosion. The pot-like karren are being formed karren will not be recognizable after a century of on rotated blocks where water remained trapped exposure. This is particularly true in the case of in small depressions, and induced development small depressions, whereas those larger forms will of small kamenitzas. Their development was also 372 KRF•2 • OK.indd 372 15.12.2009 10:59:01 Dražen Perica and Tihomir Marjanac, Types of karren and their genesis on the Velebit mountain Figure 11: Karren well “Čelinka”, diving revealed water depth of 4.2 m in early June. promoted by differential resistance of various ing. The karren well bottoms are filled with resid- clasts to erosion. Individual pots are later con- ual clays and silts, which completely seal fissures nected by small grooves. Pots may be sometimes in some of the wells, making them impermeable. connected by lateral biocorrosion. When two pots In this way some of the karren-wells permanently connect they are known as doublets or twins, and or semi-permanently hold water. The Čelinka kar- if three pots occur connected they are triplets. In a ren wel which occurs at 755 m a.s.l. near Vidakov case when more pots connect, they are called plat- Kuk, is 1.5 m long, 0.7 m wide and 4.5 m deep, ters or saucers (Rubić, 1936). Prolonged develop- and holds water throughout the year which sel- ment of grooves due to regressive and lateral cor- dom lowers below a water mark at 4 m. The walls rosion destroys the platters which remain only in of the Čelinka well are fluted down to the depth form of smaller flatter widenings. of 2.2 m, and smooth below. Its bottom is covered Specific karst features which were formed by by residual fine mud and clay with scattered small the sinking of larger quantities of water are kar- rock debris, which makes a layer locally more ren wel s (Figures 10, 11). They most often form in than 50 cm thick. The analysis of water from this places where rinnenkarren furrows develop along karren well (Perica, 1998) showed (at the time of a fissure in a semi-circular manner. Because the sampling in the dry season) pH 8.48, which does water flows down these gutters, it concentrates not explain uniform corrosion of the well walls, in a small area at the bottom, which promotes and smoothing of the edges of submerged flutes. stronger corrosion and, consequently, its deepen- 373 KRF•2 • OK.indd 373 15.12.2009 10:59:02 Karst Rock Features • Karren Sculpturing Conclusion Würm period, and that it locally reached even higher values due to south-western slope expo- Although the formation of karren is characteristic sure. Fissured carbonates prevented accumulation of parts of the Velebit Mt which are of carbonate of water, which would freeze at low temperatures, rocks, they are best exposed on its SW slope. This and surely cause mechanical destruction of the is primarily a result of anthropogenic deforesta- fissure- and network-type of karren. tion. Denudation of the mountain slope exposed During the Pleistocene glaciation, the high numerous types of subcutaneous karren. Their de- parts of the Velebit mountain were affected by per- velopment is still intensive below the pedologic iglacial, and the summit parts by glacial processes cover on other parts of the Velebit Mt, as shown which destroyed older fissure- and network-type by karren which become exposed on recently de- karren. Destruction of fissure- and network-type nuded surfaces. karren by periglacial processes was interpreted as The size of rinnenkarren features is larger on a major contributory to formation of a pediment the middle and lower parts of the SW Velebit Mt during the Quaternary on the SW Velebit Mt slope, than on its higher parts, which confirms slope (Bognar, 1992). Bögli’s (1980) hypothesis that their development The intensity of postglacial corrosion is dem- continued also during the last Ice-Age. This is onstrated by the development of large furrows in also confirmed by the development of furrow-like the highest parts of the Velebit mountain, which karren in the Julian Alps at altitudes up to 2,500 confirms observations from the Orjen Mt high m a.s.l. (Kunaver, 1985), at an average annual air parts (Riđanović, 1964, 1966). Development of temperature of –1.8°C (Bernot, 1985). It has been the large furrows on the Velebit Mt high parts can inferred that the Velebit Mt slopes at 500 m a.s.l. be explained by significantly stronger corrosion of had same average annual temperature during the water in cold mountain climates (Corbel, 1959). 374 KRF•2 • OK.indd 374 15.12.2009 10:59:02 Mid-MounTain Karrenfields 30 aT serra de TraMunTana in Mallorca island Joaquín GINÉS and Angel GINÉS Mallorca island is located roughly at the middle of level to above 1,400 m). The impact of human ac- the Western Mediterranean basin (39°N latitude tivity over the last 5 millennia, together with other and 3°E of Greenwich) being a fully representative mechanisms of natural deforestation, have pro- territory of this particular geographical macro- duced a complex evolutive history of the existing unit. The geological setting of the island, largely karrenfields within a mid-mountain bioclimatic formed of limestone rocks, and its typical medi- and geomorphological framework. terranean bioclimatic conditions produce several distinctive karst landscapes, the most outstand- ing being the Serra de Tramuntana range. This re- Geological and bioclimatic setting gion is the main mountain area in Mallorca, hav- ing a surface of approximately 1,000 km2, about The Serra de Tramuntana lies at the north of Mal- 65% of that being limestone outcrops. lorca island forming an abrupt mountain chain, Karren landforms of Serra de Tramuntana 90 km long and 15 km wide, elongated NE–SW gained the early interest of naturalists (Lozano, (Figure 1a). Their highest altitude is at Puig Major 1884). In the second half of the 20th century, some (1,445 m) with over fourteen other peaks higher researchers from central Europe (Mensching, than 1,000 m. It is composed of a complex system 1955; Bögli, 1976; Bär et al., 1986) pointed out the of folds and thrust sheets, formed by compressive spectacular nature and geomorphological rich- stresses pushing from the SE to the NW. These ness of the exokarst in the Mallorcan mountains. structures resulted from an alpine tectonic event More recently, from the nineties, a lot of literature that took place between the Late Oligocene and has been devoted to surface solutional features in Middle Miocene, involving rocks that range from Serra de Tramuntana, ranging from morphologi- Upper Palaeozoic to Lower Miocene (Gelabert, cal and morphometrical aspects to genetic or evo- 1998). In broad terms the structure of the Serra lutive ones; an exhaustive bibliography is given in is an assemblage of imbricated sheets piled up to- Ginés (1999a). wards the NW and aligned NE–SW, integrating The exokarst in the studied area is character- the emerged area of the so-called Balearic Prom- ized by a remarkable variety of solutional forms. ontory that is in fact a prolongation of the Betic These are the result of a wide diversity of environ- chains. The compressional alpine features are mental situations, basically controlled by climatic capped by Upper Miocene to Quaternary post- gradients linked to the altitude (ranging from sea orogenic sediments (Figure 1a). 375 KRF•2 • OK.indd 375 15.12.2009 10:59:02 Karst Rock Features • Karren Sculpturing ‚     ƒ †  ­€ ƒ„Ž ƒ†Ž ƒˆŽ ƒ„Ž ƒ„ ƒ† ƒ ƒ‹ ƒ‡ ƒ ƒˆ ƒ‹ ƒ„ ƒ† ƒ‰ Š€  ƒ ŒŒŒ Figure 1: Geographical information on Mal orca island: a. main Mal orcan litho-structural units and situation of sites re- ferred to in the text; b. distribution of the average annual rainfal and temperature values; c. simplified altimetry map. 376 KRF•2 • OK.indd 376 15.12.2009 10:59:03 Joaquín Ginés and Angel Ginés, Mid-mountain karrenfields at Serra de Tramuntana in Mal orca island The geologic structure briefly described above Snowfalls are scarce and now limited to the high- has formed a succession of high energy relief est elevations during a few winter days. Intense mountain alignments, that have steeper cliffs in stormy rain events (over 250 mm in 24 hours) are the northern side and relatively gentler slopes to not exceptional, particularly in autumn months, the south determined by the general dip of the due to sudden cold air irruptions – in the middle limestone beds towards the SE. The repetitive al- and upper parts of the troposphere – over a very ternation of limestones and marly or shale materi- hot Mediterranean Sea water mass. als due to the sheet imbrications, contributes to Mean annual temperatures range from 12°C the characteristic sawtooth profile that is seen in a in the central highest part of the mountain range, transverse NW–SE section of this range. to 17°C in the outermost Formentor and Andratx The lithologies are diverse (marls, shales, sand- ends. Seasonal variability is noticeable, with win- stones, gypsum, volcanic rocks, etc.) but the lime- ter mean temperatures below 10°C and summer stones are by far the dominant ones (Fornós and ones close to 25°C. In general terms, climate in Gelabert, 1995). Specifically, karst landscapes are Serra de Tramutana is that of mid latitudes but developed on Jurassic (Lower Lias) micritic lime- modified by the azonal anomaly provided by the stones and Lower Miocene (Burdigalian) calcar- Mediterranean Sea. Within this general context, eous conglomerates and calcarenites. The rocks great differences in rainfall and temperature val- from both stages are lithologically quite similar ues, linked to altitude, determine local microcli- and notably pure: insoluble residue ranges be- matic conditions ranging from humid to semiarid. tween 1 and 10% and Mg content is always < 4%. Vegetational stages found in the region are re- The only differences between them are textural; lated to this environmental variability imposed the Lower Miocene deposits are very coarse – but by macro- and micro-climatic controls. Dense supported by a micritic matrix – in comparison woods of holm oak ( Quercus ilex) are the com- with the fine-grained Jurassic limestones. The munity best adapted to the relatively humid, but karstifiable rocks in Serra de Tramuntana are me- seasonally dry conditions which dominate most chanically hard, giving Schmidt hammer values of this mountain range. Forests of Aleppo pines between 40 and 52. ( Pinus halepensis) are well developed in the drier An important role in karst development is environments, corresponding to the lower alti- played by the Triassic (Keuper) marly deposits. tudes and the outermost ends of the area. Finally, Firstly, these rocks have a relevant hydrogeologi- the tree line lies at about 800 metres a.s.l., forests cal function acting as impervious substratum for being replaced at the karstified summits by shrub the underground circulations, in addition to their formations very rich in endemic species: the so- above-mentioned contribution to the structural called “balearic stage”. and topographical configuration of the Serra. Sec- Cultivable lands are restricted to non-karstifia- ondly, these marls supply abundant loose material ble rock outcrops (mainly Triassic and Cretaceous that contribute significantly to the subsoil shaping marls), although a lot of stone wall terracing was of many karrenfields. made in some limestone slopes for olive tree cul- Climate in the area is typically mediterranean tivation. Due to the presence of extensive kar- (Guijarro, 1995), characterized by an important renfields in Serra de Tramuntana range, the ag- summer drought from June to September. Rainfall ricultural exploitation of these karstic landscapes reaches up to 1,400 mm/yr in the central, highest, is minimized by the lack of arable lands. For this portion of the range, decreasing strongly towards reason, the major primary human activities were the lower periphery of the mountain chain (< 500 historically extensive grazing (sheep and goats) mm/yr, in the SW and NE ends); thus, the rainfall together with periodic burning of brushwood in pattern follows that of the elevation (Figure 1b, c). order to renew the grazing cover (Ginés, 1999b). 377 KRF•2 • OK.indd 377 15.12.2009 10:59:03 Karst Rock Features • Karren Sculpturing Elementary karren features most impassable terrains jagged with sharp ridges represented in the area and pinnacles separated by deep grikes. The range of micro- and meso-forms is remarkable (Bär et Karren are the most striking and widespread al., 1986; A. Ginés, 1990, 1998b), many of them karstic landforms in Serra de Tramuntana region having attracted the attention of researchers in (Ginés and Ginés, 1995). Particularly, the lime- the last decades, as shall be discussed below. stones outcropping in the north-eastern half of the area (between Sóller and Pollença villages) constitute great expanses of bare rocks (Figure 2) Solution flutes and related features on which spectacular karrenfields extend contin- uously for several km2. These karren areas show Starting with the most simple and elementary so- their best examples at moderate altitudes (from lutional feature, ril enkarren are undoubtedly the 200 to 600 metres a.s.l.), frequently forming al- most investigated microforms in the Mallorcan Figure 2: Typical appear- ance of extensive kar- renfields in the heart of Serra de Tramuntana. The doline at the bottom of the image is named Clot de lÍnfern, being located near Torrent de Pareis, at Escorca munici- pality. 378 KRF•2 • OK.indd 378 15.12.2009 10:59:05 Joaquín Ginés and Angel Ginés, Mid-mountain karrenfields at Serra de Tramuntana in Mal orca island mountains, being practically ubiquitous over the always at elevations greater than 800 metres. These whole area (Figure 3). Nevertheless, it is clear that forms are interpreted as decantation flutes related this karren feature – as has been defined by Ginés to snow (Ginés, 1996b), which obviously was more (1996b) – has in Serra de Tramuntana an altitudi- frequent during Pleistocene cold events. That au- nal limit at around 800 metres a.s.l. A lot of work thor also refers to wide flutes (mean widths close has been done on the morphometry of rillenkarren to 2.5 cm) in particular environmental conditions, in this region (Bordoy and Ginés, 1990; Crowther, corresponding to a cleared forest which keeps up 1998; A. Ginés, 1990, 1996b, 1999a; Mottershead, a partial cover of holm oak ( Quercus ilex). In this 1996a) which provides some data that set the size last case, a greater water drop size caused by the limits for this form. Morphometric parameters forest canopy could be responsible for the anoma- established by those authors are as follows: width lous width of flutes. from 1.4 to 2.1 cm, length between 12 and 50 cm Some other morphological and evolutive pro- and depth ranging from 2.8 to 6.3 mm. perties of rillenkarren have been investigated, Several of the above-mentioned parameters are mainly in the renowned site of Lluc, at the heart strongly dependent on a very simple variable: the of Serra de Tramuntana (520 metres in altitude; altitude and the corresponding climatic gradation. mean annual precipitation > 1,000 mm). Motter- Thus, both rillenkarren length and depth show a shead (1996a) studied variations in rillenkarren clear negative correlation with altitude (Figure 4), cross-sections along their longitudinal profile, the rills becoming shorter and shallower as they noting a negative correlation between rill depth approach or surpass the tree line. However, width and slope angle of the rock surface. At the same is the most stable parameter (Ginés, 1996b), show- time it was observed that depth increases rapidly ing no dependence on elevation. It is worth men- downwards from the rock crests – mainly along tioning, however, the existence of solutional flutes the upper third part of the flute – decreasing after- wider than typical rillenkarren (from 2.5 to 4.0 wards more gently, whereas width remains con- cm) restricted to the highest summits of the range, stant over the whole longitudinal profile; this fact Figure 3: Well developed rillenkarren features in Lower Miocene rocks at the mid-mountain site of Mortitx (Escorca). 379 KRF•2 • OK.indd 379 15.12.2009 10:59:07 Karst Rock Features • Karren Scul ptu ring suggests a higher rate of flute channel lowering in its upper part. Moreover, Crowther (1998) report- ed a noteworthy asymmetry of rillenkarren cross- sections, which are composed of two independent parabolic half-sections. The observed asymmetry could cause a lateral migration of rillenkarren sets, maintaining still their form and size due to a kind of dynamic equilibrium. The current knowledge of Mallorcan rillenkar- ren shows that their morphometry is a very effec- tive environmental indicator, reflecting altitude- controlled climatic gradients: rainfall increasing and, particularly, temperature decreasing with elevation (A. Ginés, 1990, 1996b, 1999a). An additional interesting aspect of rillenkar- ren is the participation of biokarstic processes in shaping these elementary forms. Fiol et al. (1992, 1996) demonstrate that the mechanical removal of small limestone particles, detached by the im- pact of raindrops, is an efficient process affecting rillenkarren growth. This detachment is greatly favoured by the presence of algae that have previ- ously corroded the rock surface, with the subse- quent weakening of its crystalline structure.   Runnels and associated forms Solution channels of various types – generated by water runoff on limestones and decimetric to met- ric in size – are common where bare karren areas form the landscape. Their morphology is strong- ly controlled by some medium scale topographic effects, such as the slope of rock surfaces. Most common are the funnel cross-section runnels (a sort of rinnenkarren feature) more or less incised into the flanks of major karren macroforms such as pinnacles (Figure 5).  Runnels with an overall long profile gradient Figure 4: Morphometric parameters of rillenkarren in greater than 35° are the most common in the Serra, Serra de Tramuntana plotted against elevation of the usually reaching several metres in length and up to sampled localities. Length: average values of the 10 1 metre in depth. Rinnenkarren of those charac- longest flutes in 45 measured sites; depth: average teristics have their longitudinal profile interrupt- values from 14 different localities (n = 100 flutes/site); ed by gentle-sloped segments (from 5° to 20°) that width: average values from 26 different localities (n = 100 flutes/site). can extend up to as much as 1 m2, whereas their 380 KRF•2 • OK.indd 380 15.12.2009 10:59:08 Joaquín Ginés and Angel Ginés, Mid-mountain karrenfields at Serra de Tramuntana in Mal orca island Figure 5: Typical karst landscape in Serra de Tramuntana (Muntanya, Escorca). This charac- teristic karren assem- blage is a combination of rillenkarren, trittkar- ren, rinnenkarren and regenrinnenkarren features, sculpturing the ridges of pinna- cles (spitzkarren) which emerge over a cleared holm-oak forest. Eleva- tion of this site is 550 m a.s.l., and rainfall values surpass 1,000 mm/yr. plan development is rather straight. These flatter In the vast majority of cases, these present-day areas along the runnel΄s profile faithfully resem- rinnenkarren features derive from prior forms ini- ble trittkarren features as defined in the German tiated and partially evolved under soil cover. This literature (Bär et al., 1986); however, some authors fact is widespread in the general karrenfield evolu- prefer to simply refer to them as steps or stepped tion in the Serra de Tramuntana (A. Ginés, 1990, flats spaced along the rinnenkarren long profile 1995a; Ginés and Ginés, 1995), as will be argued (Crowther, 1997). In this last paper it is suggested later. In some particular sites (for example, near that small bevels or steps randomly evolve from Ses Basses de Mortitx, Escorca) there are fully- minor irregularities, due to the slower dissolution rounded runnels ( rundkarren) separated by met- occurring in horizontal segments compared with ric-sized smooth ridges, which have little or no su- the greater lowering of steeper slopes. This behavi- perimposition of bare karren features such as ril- our is attributed to the thicker boundary layer of lenkarren. Finally, it must be pointed out that the flow on the horizontal steps, that makes them walls of grikes and pinnacles are intensely sculp- slower developing forms in comparison with the tured by diverse vertical runnels and wide flutes, enclosing subvertical backwalls of the runnels. such as regenrinnenkarren (solution flutes of the Where the slopes of the rock surface are lower second order, after Bögli, 1980) and wandkarren; than 30° rinnenkarren are also present, but asso- these forms contribute strongly to the isolation of ciated with increased sinuosity of the runnels. In big pyramidal units of spitzkarren or pinnacles. this sense, Hutchinson (1996) observed increases Generally speaking, runnels are well developed of 0.3 in the sinuosity index with every 5° decrease at elevations from 200 to 800 metres a.s.l., where in the runnel slope; so, sinuosity values near 1.0 the best karren sites of the Serra occur. At greater correspond to high gradient runnels (> 35°) but elevations some localities (e.g. Puig Major massif) reach up to 1.5 in more gentle (25°) rinnenkarren. show modest rinnenkarren and mänderkarren, In still flatter areas, Hutchinson (1996) reports real developed on gentle slopes and associated with mäanderkarren on rock slopes between 7° and 14°. melting snow-patches. 381 KRF•2 • OK.indd 381 15.12.2009 10:59:10 Karst Rock Features • Karren Sculpturing Figure 6: Famous camel- like pinnacle – popularly known as Es Camell – existing in the karren- fields at the surround- ings of Lluc monastery (Escorca). Carbonate rocks are Lower Mi- ocene in age. Pinnacles and grikes Undoubtedly the most spectacular karren ex- amples in Mallorca are on the limestone rang- es of Escorca municipality, at the central part of Serra de Tramuntana. In that area, karrenfields like those situated between Lluc monastery and Menut farmhouse (inside a cleared holm oak for- est setting) or other localities such as Mortitx and Sa Calobra (in totally deforested conditions) form almost impassable areas, because an impressive pinnacle or spitzkarren landscape is the dominant feature (Figure 6). The sites referred to are always located below 700 metres a.s.l., being chiefly de- veloped on the Lower Miocene coarse limestones and conglomerates. The solutional forms which comprise these karrenfields consist of large pyramidal pinnacles – frequently more than 200 m2 in plan and exceed- ing 10 metres of height – separated by deep clefts or grikes whose walls are sculptured by abundant vertical runnels (Figure 7). The pinnacle flanks are Figure 7: Spectacular vertical runnels in Jurassic lime- cut by large stepped runnels, such as those previ- stones from the Es Castellots massif (Escorca). Elevation of this locality is around 500 m a.s.l. ously described. In detail, the ridges of pinnacles 382 KRF•2 • OK.indd 382 15.12.2009 10:59:14 Joaquín Ginés and Angel Ginés, Mid-mountain karrenfields at Serra de Tramuntana in Mal orca island (and rock crests in general) exhibit an extensive sculpturing by very sharp rillenkarren, making an extremely jagged and uninhabitable environ- ment. The hydrological efficiency of karrenfields of this kind is noteworthy, since rain is almost completely infiltrated through solutional widened fissures ( grikes or kluftkarren). The spot known as Sa Mitjania – also in Escorca municipality – is a typical example of landscape with minimized surface run-off due to karren development; it has large solutionally sculptured pinnacles separated by small dolines, grikes and even deep shafts (J. Ginés, 1990). Figure 8: Clearly smoothed and rounded karren features The individual features that constitute the (rundkarren) observable in Lower Miocene rocks at Ses spitzkarren or pinnacle extensions in Serra de Basses de Mortitx (Escorca). Tramuntana show clear evidences of subsoil dis- solution (Figure 8), which took place during the first stages of the evolution of these karrenfields areas located at low altitudes (< 200 metres a.s.l.; (A. Ginés, 1990a, 1995a, 1998b). Such cryptolapiaz annual precipitations < 700 mm) and in particu- (subcutaneous) inheritance is evident from the lar at its SW and NE extremes, such as the famous roughly rounded appearance of the ridges form- tourist place of Formentor. ing the emergent pinnacles and even in the rather Another karren microform linked in some rounded appearance of most of the rinnenkarren. manner to arid locations are ril ensteine ( micro- Modern subsoil sculpturing can be seen at the ril s). These very tiny millimetric rills and spikes bottom of the clefts between the pinnacles. These can be found in abundance at the drier ends of subsoil inheritances become progressively masked, the range, frequently related to marine spray dur- in many cases almost totally, by bare surface so- ing heavy storms. However, that rillensteine also lutional types (rainpits, kamenitzas, rillenkarren, occur in more humid conditions (karrenfields regenrinnenkarren, wandkarren, etc.) as has been situated from 500 to 800 metres a.s.l.) but asso- described at Son Marc site – between Lluc and Pol- ciated with water supplies in the form of dew. At lença village – by Smart and Whitaker (1996). greater elevations, lichen colonization can com- petitively hinder the generation of microkarren features, such as rillensteine or even larger flutes Other minor solutional sculpturing such as rillenkarren (Ginés, 1999a). In the high- est summits, subvertical or very steeply dipping Non-linear forms are represented by a wide range rock surfaces become sculptured by forms trans- of types whose distribution patterns can be quite verse to the water flow, consisting of horizontal different to the solutional features described cockling patterns which produce concave shapes above. For example, whereas solution pans (or ka- resembling ripples. These last features are located menitzas) may be considered ubiquitous in rela- at elevations above 1,000 metres a.s.l., with mean tively flat limestone outcrops, rainpits show a geo- annual precipitation that exceeds 1,000 mm and graphical distribution restricted to the more arid includes occasional snowfalls. spots of Serra de Tramuntana. Specifically, centi- An interesting aspect is related to the rough- metric cup-like rainpits are abundant in the pe- ness of the rock surfaces, a topic investigated in riphery of the mountain range, that is to say in the depth by Crowther (1996, 1997). That author 383 KRF•2 • OK.indd 383 15.12.2009 10:59:16 Karst Rock Features • Karren Sculpturing measured the so-called Mean Gradient Change of rocky surfaces with an approximate 1 mm resolu- tion, finding the lowest roughness values (MGC = 6°) in smooth features such as the stepped flats (trittkarren) embedded inside the rinnenkarren channels. On the other hand, the highest rough- ness is related to rillenkarren and rinnenkarren forms, with MGC values of 8.8° and 11° respec- tively; these high values were attributed to the turbulent flow occurring in runnels and flutes, compared with the laminar flow on flatter slopes. Finally, strongly etched surfaces characterized by a sharp microtopography can be observed in the more arid parts of the Serra, together with abun- dant rainpits and microrills. In some sites these etched rock surfaces appear associated with ma- Figure 9: Results of the geo-ecological factorial rine spray in the supralittoral zone. No data are analysis of karren assemblages, performed using available on roughness in these specific coastal 20 morphological karren-descriptors as well as the characteristic plant species from 100 sampled sites. and arid environments, but Crowther (1996) re- The distinguished assemblages are: OM. optimum fers to very high MGC values, between 15° and 25°, mediterranean mid-mountain karren; MS. mountain from supralittoral karren on the eastern coast of summits karren; SA. semi-arid karren. the island. The karren assemblages and their using a semi-quantitative method which takes ecological significance into account both the abundance of 20 karren types as well as the characteristic plant species Observation of karst landscapes throughout the found at each site. Vegetational descriptors were study area allows recognition of several distinc- the presence/absence of some species considered tive karren assemblages, whose distribution clear- good indicators of the environmental variability. ly shows strong regularities. For example, it has al- The data were treated by factorial analysis (Figure ready been reported in the literature that the solu- 9) and the results support the distinction of the tional forms present in the highest summits of the three main karren assemblages listed below: range are very different to those at the lowest el- 1 semi-arid karren, characterized by common evations, where a semi-arid climate occurs. Like- rainpits, rillensteine and irregularly etched sur- wise, any journey through the Serra shows that faces, located on southern exposures at the pe- the best developed karrenfields in Mallorca are riphery of the range (usually < 200 metres a.s.l.), found in quite specific environmental conditions: where rainfall do not reach 800 mm/yr and precipitation > 800 mm/yr; elevation between 200 xeric plant associations are dominant; and 700 metres a.s.l.; and other factors such as the 2 optimum mediterranean mid-mountain kar- presence of suitable lithologies. ren, exhibiting long rillenkarren and rinnen- A first approximation aimed at defining the karren, together with other types such as karren assemblages observable in Serra de Tra- trittkarren, regenrinnenkarren, etc., all inte- muntana was developed by Ginés (1996a). That grated into a spectacular spitzkarren landscape author sampled 100 sites, along the whole range, (Figure 5). This is found in mid-mountain lo- 384 KRF•2 • OK.indd 384 15.12.2009 10:59:17 Joaquín Ginés and Angel Ginés, Mid-mountain karrenfields at Serra de Tramuntana in Mal orca island calities (200 to 800 metres a.s.l.) with precipita- tion ranging from 800 to 1,000 mm/yr, mainly on southern exposures and associated in many cases with cleared holm-oak forests; 3 mountain summit karren, present at high el- evations (> 800 m a.s.l. on northern exposures, and > 1,100 m on southern ones), where more than 1,000 mm/yr of precipitation occur in- cluding some winter snowfalls. This assem- blage is defined by the dominance of small decantation flutes (wider than rillenkarren), kluftkarren and transversal cockling patterns, together with a few runnels and mäanderkar- Figure 10: Special karren assemblage of the highest ren (Figure 10). From the vegetational point of summits of Mallorcan mountains (> 800 m a.s.l.). view, this appears linked to the peculiar shrub At localities such as Puig de Massanel a (Escorca) formations of the so-called “balearic stage”. the predominant forms are a type of solution flute The above assemblages are complemented by wider than rillenkarren (decantation flutes) as well as kluftkarren (scale bar = 20 cm). two more karren associations related to small scale climatic variability linked – for instance – to differences in temperature and humidity between sunny and shady exposures, or to the availability annual rainfall less than 800 mm, showing of rock surfaces for colonization by lichens. There- poorly developed solutional features and rock fore, two more karren assemblages must be taken surfaces extensively colonized by xeric lichens, into account both characterized by biokarstic as happens in the Andratx area at the SW end weathering: of the range; 1b semi-arid biokarstic karren, on localities with 3b wet mountain biokarstic karren, restricted Figure 11: Distribution of the existing karren assemblages, shown on two idealized profiles of Serra de Tramuntana. 1. semi-arid karren; 2. optimum mediterranean mid-mountain karren; 3. mountain summits karren; 1b. semi-arid bio- karstic karren; 3b. wet mountain biokarstic karren. 385 KRF•2 • OK.indd 385 15.12.2009 10:59:19 Karst Rock Features • Karren Sculpturing Figure 12: Pinnacle landscape at Míner locality (Pollença). These karrenfields evolved from subcutaneous forms through progressive and diverse deforestation and soil loss mechanisms. to moist and shady northern exposures in the strong development of subcutaneous karren fea- high peaks of the Serra (precipitation > 800 tures corresponding to a former vegetal cover mm/yr), sites where intense colonization by li- adapted to the regional and local climatic condi- chens takes place. tions (Ginés, 1998b, 1999a). A subsequent phase The spatial distribution of these five karren as- consists of the progressive exhumation of the pre- semblages is represented in Figure 11 using two viously generated cryptolapiaz (Figure 12), as a idealised cross-sections of Serra de Tramuntana. result of the negative balance between generation This diagram shows how the location of the most and loss of soil (A. Ginés, 1995); this second stage spectacular karrenfields (2) corresponds to the includes the superimposition of a great variety of mid-altitude parts of the range; it is also evident bare solutional forms (rillenkarren, rainpits, ka- that climatic gradients related to topographical menitzas, etc.). relief are controlling the distribution of other well Smart and Whitaker (1996) followed that ap- differentiated assemblages both around the pe- proach in a case study on the karren assemblages riphery of the area (1, 1b) and on the highest sum- encountered at distinct elevations above the val- mits of the mountain chain (3, 3b). ley floor in Son Marc site, near Pollença. In that locality, exokarst development is characterized by the differential lowering of the soil-covered rock Karrenfield evolution surface, which preferentially occurs along subver- tical joints initially to form stripped rock ridges. When describing the different solutional forms Soil losses linked to incision of stream courses in- represented in Serra de Tramuntana, the subsoil dicate the advance of bedrock exposure, with the origin of most karrenfields was emphasized. This formation of incipient kamenitzas and rillenkar- is true in almost all of the environmental situa- ren. Later, rinnenkarren features are increasingly tions distinguished in this range, being evident developed on the steep-sided ridge flanks, at the a first morphogenetic stage characterized by a same time that differential relief is created be- 386 KRF•2 • OK.indd 386 15.12.2009 10:59:21 Joaquín Ginés and Angel Ginés, Mid-mountain karrenfields at Serra de Tramuntana in Mal orca island     Figure 13: Patterns observed in the evolution of Serra de Tramuntana karrenfields: a. two-stage model of karren de- velopment and possible factors responsible for deforestation and soil loss processes; b. karren features distribution related to the retreat of Triassic non-karstifiable materials from above the limestone, occurring at the front of main thrust sheets. 387 KRF•2 • OK.indd 387 15.12.2009 10:59:22 Karst Rock Features • Karren Sculpturing Figure 14: Karrenfield evolution associated with the fronts of thrust sheet structures. Both images are from Coma de ses Truges (Escorca). Above: pinnacle features in Lower Miocene rocks (left side of the photo) progressively emerg- ing with increasing distance from the contact with overlying Triassic marly materials (grassy slopes at the right side of the image); below: detail of the resulting spitzkarren landscape (square in the upper photo). 388 KRF•2 • OK.indd 388 15.12.2009 10:59:25 Joaquín Ginés and Angel Ginés, Mid-mountain karrenfields at Serra de Tramuntana in Mal orca island tween the rock crests and the rounded soil-filled play an important role in providing material for grikes. the subsoil evolution of the karrenfields during Related to this two-stage model of karren de- the first stage of their formation (Figure 14). The velopment (subcutaneous solutional shaping fol- retreat of these beds – which frequently overlie lowed by bare karren features), the main problem the Jurassic and/or the Lower Miocene carbon- lies in deducing the factors responsible for the ate rocks – allows the progressive exhumation of negative soil balance which causes the continuous the subsoil solutional features and the subsequent stripping of these subsoil sculptured solutional superimposition of bare karren forms; the end of forms. A. Ginés (1995) proposed three distinct – this evolutionary sequence is shown by the pin- but not mutually exclusive – causes that can pro- nacle karrenfields that occur in the areas furthest duce a complex karrenfield, starting from cryp- from the Triassic cover, which have been longest tolapiaz features and leading to the bare expanses exposed to subaerial conditions (Figure 13b). nowadays present in Serra de Tramuntana (Figure 13a). The proposed mechanisms are: • pleistocene cold-climate periods might have Conclusions stressed the existing forest communities, as well as severely damaging the soil mantle. The Mallorca island is a representative location for the replacement of forests by scrub and grass com- Mediterranean environment, including a mid-al- munities stimulated the natural degradation of titude mountain range – the Serra de Tramuntana soil profiles, mainly in the highest parts of the – that is notable for its spectacular karst landforms. range; Jurassic and Lower Miocene limestones, folded • human activity during the last five millennia during the alpine tectonic event, have experienced has undoubtedly contributed to the regression an intense exokarstic sculpturing which produces of holm oak and pine forests in the island. The extensive karrenfields with a remarkable richness traditional practice of burning brushwoods, in solutional micro- and meso-forms. The distri- in order to renew the grazing lands for cattle bution and morphometry of the different karren pasturing, is a very effective historical cause of types is controlled by the climatic gradients ex- plant cover reduction in the Mallorcan karst; isting over the range which are in turn linked to • the striking pinnacle landscapes found in the elevation, since the mountain heights span from central part of Serra de Tramuntana, seem to sea level to 1,445 metres above. Biokarst incidence result from a mechanism of “subsidence” of is remarkable, particularly in some specific eco- the natural soil and forest mantle as a whole, logical situations: the semi-arid periphery of the rather than corresponding to an authentic de- range and, particularly, the northern exposures forestation process. It could be a general ten- of the highest summits. Current karrenfields have dency of forest subsidence promoted by the evolved from previous subsoil types, being later progress of karstification, which produces im- exhumed by various deforestation and soil-loss portant vertical soil losses occurring down processes until reaching the conspicuous karren through the karst massif. The lowering of both pinnacles landscape, characteristic of the central soil and plant cover has resulted in a relative part of the area. steady rise of karren pinnacles above the level The Serra de Tramuntana is an excellent loca- of the forest. tion for karren investigation in mid-latitudes, In addition to this last point, it is worth point- owing to the great variety of environmental con- ing out that loose marl and clay particles, derived ditions as well as to the important – but relatively from the weathering of the Upper Triassic rocks, recent – human impact on the area. 389 KRF•2 • OK.indd 389 15.12.2009 10:59:25 Karst Rock Features • Karren Sculpturing Acknowledgement as well as for the improvement of the English text. This work was partially supported by the re- We gratefully acknowledge Ken Grimes (Hamil- search fund of Ministerio de Educación y Ciencia ton, Australia) for the helpful review of this paper – FEDER, CGL2006–11242–C03–01/BTE. 390 KRF•2 • OK.indd 390 15.12.2009 10:59:26 Tropical Monsoon Karren in ausTralia 31 Ken G. GRIMES Karren in tropical Australia are strongly deve- two structurally distinct provinces in the east and loped at all scales from microkarren to giant the north-west of the continent (Figure 1). A third grikes and pinnacled towers, but with decreasing province, the coastal dune limestones of south- intensity and variety as one moves into the drier ern Australia extends a short distance into the dry climates of the interior. However, the local effects tropics of western Australia but is not discussed of lithology, structure, cover and denudation his- here as the karren are poorly developed, and poor- tory can create considerable variation. ly documented (see chapter 42). There are also many areas of well-developed silicate karst, both as surface landforms (grikefields, “stone cities” The tropical karsts of Australia and pinnacles) and as caves, which will not be dis- cussed here (see bibliography in Wray, 1997). Re- Distribution cent reviews of Australia᾽s tropical karst are pro- vided in Spate and Little (1995) and Gillieson and The tropical karst of Australia can be divided into Spate (1998). ”‚ ‘ƒ’ “ ‚ Ž‡ € ‚ ‰‚ ‚„ •– ‚ €ƒ ‚ „  Œ Ž  ‚ ‹‡Š ‚ „ Š ‚  Š ‚ Figure 1: Location map of tropical karsts in Australia, showing the two main structural provinces. Karrenfields occur in the exposed karst, but not in the covered karsts.  39 1   KRF•2 • OK.indd 391 15.12.2009 10:59:27 €ƒ ‚„ € ‚   †‚„‚‡‡ ­   ­ ˆ  € ‚   €ƒ ‚ „ ‡   ”‚ ‘ƒ’ “ ‚ Ž‡ € ‚ ‰‚ ‚„ •– ‚ €ƒ ‚ „  Œ Ž  ‚ ‹‡Š ‚ „ Š ‚  Š ‚ Karst Rock Features • Karren Sculpturing    €ƒ ‚„ € ‚   †‚„‚‡‡ ­   ­ ˆ  € ‚   €ƒ ‚ „ ‡   Figure 2: Monthly and annual mean temperatures and rainfal , and elevations for selected climate stations on tropi- cal karsts. Station names are shown in parenthesis where this differs from the karst name. The Köppen climates for north Australia are also shown. Data from BOM (2005). *Temperature data for Chil agoe is from 3 years only (Robin- son, 1982). Geological setting the depth of the carbonate rock is limited to a few tens of metres. In some areas the carbonate rocks The East Australian Karst Province (Figure 1) is are well-exposed, with strongly karstified out- formed on strongly-folded, generally steep-dip- crops; others have extensive covers of Mesozoic ping, Palaeozoic limestones and occasional mar- and Tertiary sediments and younger soils (Figure bles. These usually form narrow linear outcrops. 1). There are also laterite and silcrete capped deep They are impounded karst (Jennings, 1985), in weathering profiles. Chert nodules and beds are which the drainage is largely controlled by allo- common in some of the carbonates and this can genic streams which cross over, or cut through, influence the degree of karst. the limestone belts with little loss of water under- ground. In the tropical part of this province the limestone beds tend to stand above the surround- Climate ing rocks as ridges and towers. In the North Australian Karst Province the Northern Australia has a tropical monsoon cli- host rocks are flat-lying to gently folded Protero- mate. The Köppen climate classes range from zoic dolomites and Palaeozoic limestones and do- humid Aw southwards through drier BShw to arid lomites. These form extensive regions, but in some BWhw. The rainfall has a pronounced seasonal- 392 KRF•2 • OK.indd 392 15.12.2009 10:59:27 Ken G. Grimes, Tropical monsoon karren in Australia Figure 3: Aerial view of the crest of a tower at Chil agoe showing large sculptured pinnacles and vertical wandkarren. ity with a five-month summer “wet” and a longer Vegetation winter dry season (see Figure 2, and BOM, 2005, for further details). On the east coast the season- Most of the region has a savanna woodland: dens- ality decreases southwards and the climate grades er and with more understory in the wetter parts, to the Cfa type. Most rain in the wet season falls and more open in the arid regions. Open grass- either in short intense thunderstorms, or in occa- land is found in the drier areas or where there is a sional cyclonic events lasting several days. Signifi- heavy clay soil cover. Deep-rooted deciduous vine cant variation in rainfall between years is a con- thicket may grow on the rocky limestone towers sequence of the “El Niño southern oscillation” ef- and karrenfields. fect. Potential evapotranspiration is substantially greater than actual rainfall throughout the region, giving a deficit in excess of 1,000 mm per annum. 393 KRF•2 • OK.indd 393 15.12.2009 10:59:28 Karst Rock Features • Karren Sculpturing Steep-dipping limestones, east had been given considerable emphasis in earlier Australia work. He noted that, in fact, the pediments con- stitute less than half the tower perimeters. How- Chillagoe and Mitchell-Palmer ever, they are still active in many places and have cut back the tower flanks, in some cases reduc- The Chillagoe area is one of the better document- ing the tower to a scatter of fragments and ruins. ed tropical karren in Australia (e.g. Lundberg, The lower “scree” slopes of the towers are partly 1977a; Ford, 1978; Pearson, 1982; Jennings, 1982; bedrock with a thin cover and Jennings (1982) re- Dunkerley, 1983). The area is best known for its ferred to these ramps as “Richer denudation slopes” serrated limestone towers ‒ or “bluffs” as they are (Figures 4, 5). Some towers have marginal depres- locally called (Figure 3) ‒ which can reach up to 90 sions, with active subsidence of the soil into the m high, though most are less than 50 m, and are epikarst, which are the result of aggressive water from 100 m to over a kilometre long. The overall runoff from tower surface (Pearson, 1982). size and distribution of the towers are structurally The towers may be quite old. Robinson (1978), controlled by the narrow lenses or fault-blocks of Jennings (1982), Webb (1996) and Gillieson et al. steep-dipping limestone which alternate with in- (2003), all discuss the age of the karst, noting the soluble chert and other sedimentary rocks that are presence of isolated outcrops of quartz sandstone less resistant to erosion in this setting (Figure 4) of possible Mesozoic age both on the tower tops, ‒ thus they are a special type of structural “tower and around their bases. The conclusion is that karst” (Ford, 1978; Jennings, 1981, 1982). Some the towers were already well-formed at the time towers are surrounded by a limestone pediment of their burial during the early Cretaceous trans- (Figure 5) or by alluvium, but others rise immedi- gression, and were exhumed and further dissected ately beside the (commonly faulted) contact with during the Cenozoic. the surrounding rocks. The Mitchell-Palmer karst is similar to Chillagoe, but more remote and has Karren forms larger towers but fewer pediments. While the climate would seem to be important Jennings (1981, 1982) discussed the pediments, for the overall abundance of karren forms in the which, along with climatic control of tower form, area, lithology has been an important control on Figure 4: Schematic profile of a limestone tower at Chil agoe, Queensland, based partly on a diagram in Robinson (1978). 394 KRF•2 • OK.indd 394 15.12.2009 10:59:29 Ken G. Grimes, Tropical monsoon karren in Australia Figure 5: Pediment (P) with clints and soil-filled grikes at Chil agoe. In background is a small tower with a debris- covered Richer slope (R). the detailed sculpturing of individual towers. This marble (41.3 ppm) were dissolving more rapidly was recorded quite early (e.g. Daneš, 1911) and has than the coarse marble (34.1 ppm). His detailed been discussed by many authors (Wilson, 1975; results have not yet been published. Marker, 1976b; Lundberg, 1977a, b; Ford, 1978; The following description draws mainly on Pearson, 1982; Jennings, 1982; and Dunkerley, Pearson (1982) and Jennings (1982). The white, 1983, 1988). Unfortunately, there has been a lack of coarsely-crystalline, marbles (“sugarstone”) form consistency in the lithological subdivisions reco- smoothly rounded domes with exfoliation sheets gnized, and in the terminology used. that occasionally are raised to form A-tents. Some Jennings (1982) summarized the lithological surfaces show a crazed pattern of fine cracks. Ril- control as producing poorly developed karren lenkarren do occur on the marble, but are less well on the coarse-grained “sugarstones” (a crumbly developed. Lundberg (1977a) tabulated the differ- coarse-grained marble) and heterogeneous lime- ences in character between the rillenkarren on the stones, and a much wider range of well-developed “sugarstone” and those on the “sparite” limestones karren on the fine-grained uniform limestones (see also summary in Jennings, 1982). The “sug- ‒ known variously as “sparite”, “fossil” or “reef” arstone” differed from the “sparite” limestone in limestone. He also noted that areas of excessive having pits and flutes that were narrower, shallow- fracturing inhibit those karren that result from er, more constant in form and less close-set and water flow over large surface areas. had more rounded ribs between them. Dunkerley (1988) measured the chemistry of The finer-grained marbles have karren forms runoff water and kamenitza waters from rock sur- that are more similar to the “sparite” limestone. faces on three lithological groups (coarse and fine- However, the grain size of the marble can be quite grained marble, and fine-textured “fossil” lime- variable over short distances, so the above distinc- stone) and found that the “fossil” limestone (water tions need to be applied with some care. of 41.9 ppm total hardness) and the fine-grained The more widespread “sparite” or “reef” lime- 395 KRF•2 • OK.indd 395 15.12.2009 10:59:30 Karst Rock Features • Karren Sculpturing Figure 6: Spitzkarren (S) grading down to deep vertical wandkarren (W), on a tower in the Mungana area, Chil a- goe. View is about 8 m high. stone towers are strongly dissected by solution numerous daylight holes. In places the grikes and contain large grikes, vertical sculptured walls, open out into karst corridors or deep steep-walled fields of spitzkarren and large sharp-sculptured dolines of both solutional and collapse origin. The pinnacles which make access difficult (Figure 3). grikes combine with rillenkarren and wandkar- The following description refers mainly to these ren to form intricately sculptured patterns of limestones (Jennings, 1982). Note that the term sharp pointed pinnacles 5 m or more high (Fig- “spitzkarren” is used here for small to medium- ure 3). Solution dolines on the tower tops tend to sized pointed pinnacles that are sculptured by ril- be irregular forms with fields of spitzkarren and lenkarren. In my usage these range from incipi- internal drainage via grikes. Some towers have a ent rosettes of rillenkarren a few decimetres high stepped relief with risers and treads. to large pinnacles several metres high. However, On the sides of the towers and the giant grikes I do not use the term for the “large sculptured extensive rillenkarren feed via steep runnels, pinnacles” which have more complex walls (with 10‒20 cm deep, into vertical wandkarren that can wandkarren) and are big enough to have clusters be up to a metre deep (into the wall) and 40 m of smaller spitzkarren pinnacles on their crests. long (Figures 3, 6). On steep slopes the rillenkar- Within the towers, giant grikes up to 10 m wide ren are modified by cockling patterns (Sweeting, and 30 m deep connect to fissure-maze caves with 1973). Rinnenkarren (runnels) are listed by several 396 KRF•2 • OK.indd 396 15.12.2009 10:59:31 Ken G. Grimes, Tropical monsoon karren in Australia Figure 7: Stereopair of phototropic spikes with coralloid overgrowths, a result of light-oriented algal corrosion, in the twilight zone of a cave entrance. Width of view is 65 cm. authors but some appear to use this term for re- 17.3 and 29.8 cm and widths of 16.9 to 18.5 mm genrinnenkarren or wandkarren. Occasional de- at three sites on the marble, whereas two “sparite” cantation runnels occur below horizontal joints sites had lengths of 31.3 and 35.6 cm and widths cutting into vertical walls. of 18.5 and 23 mm. On the more gentle slopes, which include steps Two types of horizontal solution ripple were and bevels, there are rainpits, localized rosettes described by Jennings (1982): on underhangs and of rillenkarren which grade to incipient spitzkar- in the twilight walls of cave entrances there are ren, short irregular runnels, and small (up to 1 m sharp-ribbed and deeply recessed symmetrical wide) solution pans ( kamenitzas). Wilson (1975) forms; whereas on steep surfaces exposed by soil reported that flat-floored solution pans are com- erosion of the pediment grikes there are more mon on tops of the towers, and noted that these rounded and asymmetrical ripples that might always have an outlet drain. have resulted from subsoil solution. The rillenkarren have been studied morpho- Jennings (1982) described phototropic karren metrically by Lundberg (1977a, b), Jennings (1982) which are grooves, sticks and spines oriented to- and Dunkerley (1983). Jennings (1982) measured wards the light and found in the twilight zone of rillenkarren lengths on the “sparite” that averaged the caves and deep grikes (Figure 7). These are 95‒100 cm at three sites, with a SD of 48. Dunker- a type of phytokarst eroded by algae. Individual ley (1983) summarized the results in Lundberg᾽s spikes and grooves are between 2‒50 mm across, (1977a) thesis, and also reported additional meas- but can be up to 400 mm long! Some spikes have urements giving flute lengths averaging between coral oid growths on their tips, or along their full 397 KRF•2 • OK.indd 397 15.12.2009 10:59:32 Karst Rock Features • Karren Sculpturing Figure 8: Solution pan on a pediment near Race- course Tower, Chil a- goe. It is formed from coalescence of smaller circular pans with cen- tral pits. Width of view is 50 cm. length. Jennings (1982) described small needles, 10 but show a greater size range and all gradations up mm high and 1‒2 mm thick, on the side of a rather to normal rainpits (10 mm or greater) can occur deep solution pan on top of one tower. on one outcrop. There is also a very fine etching Microkarren are more extensive than suggested of structures such as irregular cracks, the crystal by earlier reports (Jennings, 1981, 1982; Dunkerley, boundaries of the marbles, or the skeletal struc- 1983). The terminology used here is that suggested ture of fossil corals. by Grimes (2007). The microkarren are most com- These small features have been under-reported mon on the flatter surfaces, especially on the gen- because of their cryptic nature. They are most vis- tly rounded “clints” of the pediments and on the ible in areas lacking the ubiquitous thin grey algal steps of the towers. However, I found microrills coating, e.g. in the bare areas used by wallabies. and other forms on slopes up to 60 degrees. Some However, they seem too extensive to be solely a microkarren are superimposed on rillenkarren or consequence of corrosion by wallaby urine or rainpits (e.g. see photos 1‒3 in Dunkerley, 1983); dung, as suggested by Jennings (1981, 1982). So- these appear to be secondary features modifying lution by thin films of water, dew or light rain, the initial coarser form. Linear microril s grade to seems the most likely origin (see chapter 7). micro-networks of irregular, discontinuous ridges On the pediments there are smoothly rounded which in turn break up into arrays of tiny rasp- clints between soil-filled grikes (Figure 5). The clint like micro-teeth (Figure 9). I measured the follow- surfaces may carry small areas of rillenkarren, but ing size ranges from a set of enlarged photographs: rainpits or smooth surfaces are more common, the microrills range from 0.2 to 2.8 mm wide, av- along with a range of microkarren. Solution pans eraging 1.1 mm; the micro-teeth were spaced 0.5 (kamenitzas) are less common. In one area, which to 3 mm apart, averaging 1.5 mm. Vertical relief appears to be flooded regularly, there were com- is generally less than 1 mm; some microrills and posite pans formed from coalescing smaller cir- micro-networks are extremely shallow and visible cular pans with small deep conical holes in their mainly by a slight bleaching of the crests. Circu- centres (Figure 8). It would appear that these small lar micro-pits also occur as small as 1 mm across, pans have been draining downwards through fine 398 KRF•2 • OK.indd 398 15.12.2009 10:59:35 Ken G. Grimes, Tropical monsoon karren in Australia Figure 9: Microkarren on a pediment at Chil- lagoe. Left: micro-network of small furrows and bleached ridges; right: rasp-like micro- teeth. Scales in mm. cracks. Subsoil solution pipes also occur, typically may be slightly undercut. Commonly they are elongated along a joint. In places soil erosion has from 0.3 to 2 m wide and from 10 to 80 cm deep. exposed the grikes and other, generally rounded, The floor is generally bare limestone, with a thin rundkarren and subsoil karren. (1‒5 mm) coating of clay and organic material. There are several terraces visible on the channel floors with the presently active channel in places Broken River being a narrow slot within a broader channel. The higher terraces, which are commonly paired, now This region is similar to Chillagoe, but differs in have small rillenkarren, rainpits and kamenitzas that the limestones here are not as steeply dip- developing on them. ping, typically 50‒70 degrees, and rather than These channels appear to be dominantly so- high abrupt towers, they form long linear ridges lutional in origin. The wet season storms could dissected by grikes, spitzkarren and larger sculp- produce sufficient runoff to allow some hydraulic tured pinnacles. Microkarren here are restricted erosion, though there is no sediment to provide to bleached patches which are mainly the result of abrasive tools. The clay and algal material on the wallaby defecation. They comprise well-developed floors may have favoured undercutting of the microrills, micro-networks and micro-pits super- walls over down-cutting of the floor – as happens imposed on rillenkarren and rainpits. in kamenitzas. The Turtle Creek Tower has some features of special interest. This broad, but steep-sided tower is topped by a bare plateau, including several Fanning River broad solutional basins up to 100 m across, that are dissected into low spitzkarren and smoother This is a small karst area inland from Towns- areas of kamenitzas, rainpits and rillenkarren. ville that is developed on a 1 km wide low ridge An unusual set of “interconnected solution of gently dipping Devonian limestone. The rock rivulets” was first recorded within the basins by occurs in alternating zones of thick-bedded lime- Godwin (1988). The rivulets are large runnels that stone with good karst development and poorly ex- form a branching contributory system of small posed belts of interbedded limestone, sandstone flat-floored, and locally meandering, stream chan- and shale with no karst features (Grimes, 1990). nels incised into the limestone floors of the ba- Dips vary from 10 to 70 degrees. The thick-bedded sins (Figure 10). The drainage of the largest basin limestone has some grikes, rillenkarren, and ka- leaves the tower via an increasingly deep channel menitzas. However, surface solution sculpturing with some 2 m waterfalls; in the smaller basins the is not as well developed as in the Chillagoe and runnels sink underground into small shafts. The Broken River areas. channels have flat floors and steep sides which An unusual, dipping limestone pavement occurs 399 KRF•2 • OK.indd 399 15.12.2009 10:59:35 Karst Rock Features • Karren Sculpturing Figure 10: Meandering and branching flat-floored solutional runnels in a shallow bedrock basin on the crest of the Turtle Creek Tower, Broken River. Width of view is 5 m, at the lower edge. in one place. This is a 12 degree dip surface formed borderline example of tower karst as the hills tend by the stripping of a thinner-bedded muddy lime- to be conical with a scree-covered base, and ver- stone from above a thicker-bedded calcirudite. The tical cliffs are rare. The steep sides of the moun- pavement has scattered grikes (0.5‒2 m deep and tain are bare or covered with vine thicket and are 2‒20 m long), some of which connect with caves, strongly sculptured by a combination of rillenkar- some relatively deep kamenitzas with overflow ren and larger runnels to form spitzkarren. Large channels on the downslope side, and small patches rubble-choked grikes cut across the karrenfields. of rillenkarren and rainpits; but otherwise it is es- Cave entrances are associated with the grikes or sentially undissected. This may be a similar situa- with large, vertical, solution pipes. tion to that described below at the Gregory Karst ‒ where a surface has not been exposed for long enough to develop deep sculpturing. Other eastern areas Mount Etna is at the southern limit of the tropi- Mount Etna cal region, and lies just within the northern limit of the Cfa climate type. However, well-developed Mount Etna, rising 190 m above the surrounding spitzkarren are found as far south as Kempsey, lat- plain, is the largest of several limestone ridges and itude 31°S, in northern New South Wales, which hills that lie near the coast, just north of the Trop- has a Cfa climate with an annual rainfall of about ic of Capricorn (Shannon, 1970). These provide a 1,700 mm. 400 KRF•2 • OK.indd 400 15.12.2009 10:59:36 Ken G. Grimes, Tropical monsoon karren in Australia Flat-lying carbonates, north-west “merokarst” and excluded them from their main Australia discussion. Jennings and Sweeting (1963) described an evo- West Kimberley region, western Australia lutionary sequence of dissection for the pure and well-jointed limestones (but not the merokarst). The West Kimberley Karst Region of northwest- Progressive dissection and pediplanation has pro- ern Australia is also referred to in reports as the duced the following landforms on the pure lime- “Limestone Ranges”, “Napier Ranges” and “Fitzroy stones. Stripping of the original clay soil cover of Basin” regions. It is an extensive belt of exposed the plateau leaves a relatively smooth rock surface Devonian reef that has had little folding (Playford, with minor small karren features and scattered 1980). It lies at the junction between the rugged large, deep grikes (“giant grikeland”, Figure 11). Proterozoic ranges of the Kimberley region and Widening of the giant grikes forms box valleys, the flat plains of the Mesozoic Canning Basin with flat floor and vertical walls, which in turn to the south. Limestone ridges and plateau rise coalesce to leave isolated towers and large sculp- abruptly 30‒90 m above the plains and extend for tured pinnacles within a broad pediment. Dolines 290 km with a maximum width of 30 km. The pla- are relatively uncommon. teau top is a dissected planation surface that may date back to the Permian – Playford (2002) sug- Mesokarren forms gests that an Early Permian glacial palaeokarst, The surface sculpturing can be quite intense to exhumed during the Cenozoic, has been incorpo- form inhospitable jagged ridges and spires. In the rated into the modern karst topography in many undissected parts of the plateau the smooth sur- areas. Dissection of the plateau in the late Terti- face has kamenitzas, rainpits and small patches ary and Quaternary has created the present karst of rillenkarren. This pavement is cut by a widely landforms, along with gorges of superimposed spaced network of deep grikes with fluted vertical drainage that cut across the limestone ranges. walls (Figure 11). The giant grikes are up to 7 m wide, 33 m deep and hundreds of metres long and Large-scale karst landforms extend underground into fissure caves. As dissec- (macrokarren) tion becomes greater a rugged terrain of spitzkar- The surface karst landforms have been described in ren and larger sculptured pinnacles develops (cf. detail by Jennings and Sweeting (1963) and sum- Figure 18). Rillenkarren are ubiquitous, but their marized in later papers by Jennings (1967, 1969), intensity and character are controlled by the local Williams (1978), Goudie et al. (1989, 1990) and lithology, structure and slope (see below). On the Gillieson and Spate (1998). The main scarp is an vertical walls there are large vertical solution run- abrupt wall or cliff, deeply sculptured by various nels (wandkarren), 1‒2 m deep and wide and run- karren forms, as are the steep walls of the gorges, ning vertically for 30‒60 m (Figure 11). The pedi- box val eys ( bogaz) and giant grikes which extend ments have kamenitzas and occasional shallow into the plateau. In detail, the steepness and char- dolines and subsoil solution pipes. acter of these walls are controlled by the lithology Goudie et al. (1989) discuss some lithological and structure of the different reef facies (Allison and other factors controlling the development of and Goudie, 1990). In particular, Jennings (1967) the ril enkarren, which are only well-developed noted that the backreef facies tends to be impure on certain beds. The limestones are all hard and (due to terrigenous components) and that reduces have little primary porosity, but, of the factors the degree of karstification so that one finds more which Goudie and others studied, the important rounded hills and valleys typical of fluvial erosion. control on the occurrence of rillenkarren ap- Jennings and Sweeting (1963) called these areas peared to be the purity (insoluble residue) and the 401 KRF•2 • OK.indd 401 15.12.2009 10:59:36 Karst Rock Features • Karren Sculpturing Figure 11: A giant grike, at least 5 metres wide, with col apsed blocks and deep vertical wandkarren, above Mimbi Cave, West Kim- berley region. Note the relatively undissected plateau in the back- ground (photo by J. Jennings). homogeneity (as revealed by thin section study). slopes (0‒30°) tend to be pitted and have kameni- The difference is mainly one of the fabric and the tzas. Grikes are scattered across these “pavements” cement type: in particular, rillenkarren develop and some may be filled with tufa deposits. Mod- best in the absence of fabrics characterised by erate slopes (30‒55°) have bifurcating rillenkar- bioclasts, ooids, and an excess of ooids over in- ren which become parallel as the slope steepens traclasts. They are also related to an absence of (55‒80°). The steepest slopes (> 80°) are described micritic cement. Other factors statistically asso- by Goudie and others as having “boxy forms”. ciated with rillenkarren were low levels of dolo- Within the major river gorges which cross mite and a sparite cement that is less equant than the karst, wet season floods rise to heights of 10 elsewhere. In addition to the factors measured by m (Gillieson et al., 1991). The flooded sections of Goudie et al. (1989), Jennings and Sweeting (1963) the gorge walls show well-developed scal ops and noted the influence of bedding in disrupting ril- strong etching of bedding and vertical joints to lenkarren development and breaking it into pa- form cavernous slots ( splitkarren, sensu Ford and goda-like stacks of conical spitzkarren (cf. Figure Williams, 1989) and vertical grikes, along with 18). spongework. Goudie et al. (1989) also recorded that the kar- There is no information on the occurrence of ren types vary with gradient of slope: shallow microkarren in this area. 402 KRF•2 • OK.indd 402 15.12.2009 10:59:37 Ken G. Grimes, Tropical monsoon karren in Australia   Figure 12: Schematic profile of a karrenfield in the Gregory Karst, Northern Territory. cu. upper Skull Creek formation; S. Supplejack dolomite member; cl. lower Skull Creek formation; zone 1. incipient karren; zone 2. moderately-devel- oped grikes and spitzkarren; zone 3. deep grikes and large spitzkarren; zone 4. giant grikes, unroofed caves, sculp- tured pinnacles. Ningbing and Jeremiah Hills ing late Proterozoic Skull Creek formation. The Skull Creek formation is also dominantly carbon- There have been no karst-specific reports pub- ate, but less pure and thinner bedded (Sweet et al., lished on this area of gently-dipping Devonian 1974; Bannink et al., 1995). Apart from the Supple- and Carboniferous reef limestones, which is also jack member, the Skull Creek formation has only known as the East Kimberley. However, the geo- poorly developed mesokarren, but it has well-de- logical report by Veevers and Roberts (1968) has veloped microkarren, especially in the upper part. photographs of outcrops of the different carbon- Figure 12 illustrates the geological structure and ates which show the distinctive fluted, pinnacled the resulting karren. Extensive maze caves under- and cavernous tower structure seen in other areas. lie the dissected surface (Storm and Smith, 1991; There is also a suggestion of both lithological and Bannink et al., 1995). There is obvious joint and structural control on the character of the solu- bedding control of both the karren and the un- tional sculpturing. A photo of the fore-reef breccia derlying caves. of the gently dipping (10‒20°) Westwood member Only brief descriptions of the karren have shows unusual smooth-surfaced cones and pinna- been published previously (Dunkley, 1993; Ban- cles from 1 to 4 m high. These could be uncovered nink et al., 1995). The karrenfields on the Supple- subsoil features, but as the outcrop is in an area of jack member show a zonation which results from low relief in a prograding coastal plain the poten- progressively longer periods of exposure at the tial for soil erosion seems limited. surface. This starts with incipient karren develop- ment on recently exposed surfaces adjacent to the contact with the overlying Skull Creek formation Gregory Karst and continues through progressively deeper dis- sected karren to a final stage of “stone cities” of In this area karst and karren are largely restrict- isolated pinnacles at the outer edge (zones 1 to 4 ed to a thin (10‒18 m) but extensive dolomite on Figures 12, 13). The zones are gradational and unit, the Supplejack member, within the flat-ly- the boundaries shown on Figure 13 are only ap- 403 KRF•2 • OK.indd 403 15.12.2009 10:59:38 Karst Rock Features • Karren Sculpturing Figure 13: Aerial stereo- pair of a karrenfield in the Gregory Karst. Numbers 1–4 refer to the zones shown on Figure 12. The main karrenfield is on the Supplejack member; cl and cu are outcrops of the lower and upper Skull Creek formation, re- spectively. Note subsided areas in centre (S) – a re- sult of cave undermining. Width of view is 100 m. Original air-photos copy- right Northern Territory Government, 1989. proximate. This developmental sequence has sim- and deeper, averaging 2 m deep, but with con- ilarities to that described by Jennings and Sweet- siderable variation, including occasional narrow ing (1963) in the West Kimberley region, but at a connections to the cave passages below. smaller scale. The transition to zone 3 is quite gradual. Zone Zone 1 has well-preserved stromatolite domes 3 has wider and deeper grikes, and connections (up to 12 m wide and 2 m high) exposed by to the cave become more common, though still stripping of the overlying rock. The surfaces are narrow. Traversing the surface becomes difficult. smooth or sculptured by incipient rainpits and Spitzkarren are dominant and up to 2 m high. rillenkarren with superimposed microkarren Wandkarren appear on the grike walls and the (Figure 14). Etching of joints and bedding forms sides of the larger spitzkarren. splitkarren. There are scattered kamenitzas and In zone 4 the surface has become completely small grikes. dissected. Giant grikes 1‒5 m wide penetrate to Away from the contact, increasing dissection the cave floors 10‒15 m below and separate blocks produces small spitzkarren up to 0.3 m high, and of rock with strong spitzkarren on the tops and grades to zone 2. There the stromatolite domes are wandkarren, rillenkarren and cockles on the still recognizable locally, but are strongly dissec- walls. As the grikes widen, one gets a “stone city” ted by a variety of mesokarren, including numer- topography of isolated blocks, many of which are ous kamenitzas (up to 2 m wide and 0.4 m deep) tilted, and finally an abrupt change to a broad flat and spitzkarren up to 1 m high. Grikes are wider floored valley on the lower Skull Creek formation 404 KRF•2 • OK.indd 404 15.12.2009 10:59:38 Ken G. Grimes, Tropical monsoon karren in Australia with only scattered blocks and sculptured pinna- cles (Figure 15). Kamenitzas are common in zones 2 and 3 but also found in the other zones, reaching up to sev- eral metres wide and 0.4 m deep. These can form chains linked by short runnels. The flat floors have two types of surface associated with different algal types: smooth, bare rock floors are associ- ated with curled fragments of black algae, and pit- ted floors are coated by the usual thin, grey, hard film of algae that covers most of the rock surface. The pitted floors comprise both pits and cones, 2‒5 mm wide and 2 mm deep/high. Larger rainpits form hackly floors in places. Etched stromatolite structures make small ridges on some floors. Figure 14: Microril s superimposed on shallow rillenkarren In the twilight zone of cave entrances and at the on a horizontal slab in the upper Skull Creek formation, Gregory Karst region. Width of view is 15 cm. base of the giant grikes there are phototropic spikes and solution ripples similar to those described at Chillagoe (cf. Figure 7). Microkarren occur within the main karrenfield. They are common in zone 1, but also occur in the other zones, usually at the tops of spitzkarren and associated with rillenkarren and rainpits. How- ever, the best development of microkarren in the Gregory Karst is on the flaggy to slabby outcrops of dolomite in the upper Skull Creek formation, where there is little competition from mesokarren. This formation has the best array of microkar- ren found anywhere in tropical Australia; nearly every outcrop has examples. Microkarren are best developed on gentle slopes. They include microril s up to 60 cm long, typically 1‒2 mm wide, and straight to sinuous (Figure 14) or locally tightly meandering (Figure 16). Micro-pits have a full size range from less than 1 mm wide and deep up to 20 mm (i.e. they grade Figure 15: Isolated pinnacle, about 5 m high, at outer to rainpits). A broad range of sizes can occur edge of zone 4 in the Gregory Karst (photo by N. White, within a single outcrop. On gently-domed sur- 1992). faces micro-pits occur on the crest and grade to microrills on the slopes. Micro-teeth and micro- networks (as seen at Chillagoe, Figure 9) are less finely pitted or toothed floors. Micro-tessel ations common. Small shallow micro-pans are 2.5‒8 mm (spaced networks of shallow etched cracks; see wide but only 1‒2 mm deep and are superimposed photo 3 in Grimes, 2007) are also superimposed on pre-existing microrills (Figure 17). These have on other microkarren. 405 KRF•2 • OK.indd 405 15.12.2009 10:59:41 Karst Rock Features • Karren Sculpturing Figure 16: Tightly mean- dering microril s and a v-notch (V) following a joint, on a cobble in the upper Skull Creek for- mation, Gregory Karst region. Width of view is 50 mm. Figure 17: Shallow micro- pans, with finely pitted and toothed floors, su- perimposed on micro- ril s, Gregory Karst re- gion. Width of view is 50 mm. Katherine (Daly Basin) Cretaceous sandstone and claystone and younger alluvium (Figure 1). Karst features are mainly re- The Katherine area is at the northern end of the stricted to the exposed limestones at the northern Daly Basin, a broad area of flat-lying early Pal- and western margins of the basin. aeozoic limestone that has an extensive cover of Most outcropping limestone forms pavements 406 KRF•2 • OK.indd 406 15.12.2009 10:59:43 Ken G. Grimes, Tropical monsoon karren in Australia of grikes and clints or, in more strongly dissected pinnacles and deep shafts, which is exposed with- areas, widening and deepening of the grikes has in the occasional soil-subsidence doline. converted the clints to pinnacles and small tow- Further south, the rainfall is lower, and out- ers that are typically 1‒3 m high, but up to 30 m crops around the edge of the basin have only in places (Hamilton-Smith et al., 1989; Lauritzen hackly surfaces with deep rainpits ranging in and Karp, 1993; Karp, 2002). The surfaces of the width from 3 mm to 20 mm. pinnacles and towers are sculptured by deep rain- pits and rillenkarren grading to spitzkarren in the more dissected areas. Locally the pitting becomes Barkly Karst region (Georgina Basin) very intense to form a sharp fretted surface analo- gous to coastal phytokarst (Hamilton-Smith et al., This is the easternmost of the large covered karst 1989). Kamenitzas up to two metres across are basins (Figure 1). The rocks are mainly Palaeozoic also common and some have outlet channels. flat-lying dolomite with some gently folded lime- Microkarren occur but are not common. They stones around the basin margins which have the appear local patches of bleached outcrop or on best exposures of surface karst (Grimes, 1988). occasional distinctive thin beds of finer-grained, The climate ranges from semi-arid in the north to cream-coloured limestone – in contrast to the arid in the south (Figure 2) and karren are best usual more massive and coarser-grained grey developed in the wetter northern part. limestone. Beneath the sandy cover there is a well-devel- The dissected northeastern edge oped epikarst surface of narrow smooth-surfaced Much of the northeastern edge of the karst region Figure 18: The Colless Creek karrenfield, Barkly Karst, Queensland. In the foreground a thick-bedded limestone is dis- sected into deep grikes and spitzkarren. Beyond the gorge of Colless Creek is a plateau developed on a less pure and thinner-bedded limestone. This photo is typical of many outcrops of flat-lying thick-bedded carbonates in tropical Australia. 407 KRF•2 • OK.indd 407 15.12.2009 10:59:44 Karst Rock Features • Karren Sculpturing between the deep grikes is strongly dissected by rillenkarren, steep runnels and spitzkarren (Fig- ure 18). Kamenitzas and rainpits also occur on flatter surfaces. Gale et al. (1997) described another karren- field 12 km further west which had relatively thin (about 2 m) beds of pure and thick-bedded lime- stone interbedded with less pure, closer-jointed and medium-bedded cherty limestone beds. There, the grikes have widened to form small flat- floored “box valleys” (their usage) a few metres across and a “Lost City” of narrow walls and small “towers” up to 4.5 m high. The “towers” and walls are capped by the thick-bedded limestone, but the grikes have cut below this several metres into the underlying thinner bedded and less pure lime- stone. Gale et al. (1997) interpret the flat floors Figure 19: Fine etching of cracks (splitkarren) in dolomite as corresponding to an impermeable bed which near Camooweal, Barkly Karst, Queensland. Width of view is 20 cm. converted the downward erosion of the joints to lateral corrosion which widened the grikes. The southern, arid region is a dissected Tertiary plateau (Grimes, 1988; Wil- The more arid and less dissected parts of the Bark- liams, 1978) with the major streams incised as a ly Karst have relatively poor karren development. superimposed drainage pattern. Between these Possibly this is partly because much of this coun- is a dense pattern of modern dendritic surface try is developed on dolomite. drainage with v-section valleys and rounded in- Camooweal lies at the boundary between the terfluves which is developed on impure limestone semi-arid BShw and arid BWhw climates (Figure and dolomite with abundant chert as nodules and 2). Here, the dolomite strata are horizontal and thin beds. This is equivalent to the “merokarst” of thick to medium and occasionally thin bedded. the West Kimberley region (Jennings and Sweet- They have well developed vertical joints that result ing, 1963) and lacks significant karst or kar- in a blocky to slabby outcrop with narrow grikes ren, though there are scattered caves and dolines. and clints. Other karren are common rainpits and Within this terrain occasional distinctive dark fine etching of structures and occasional poorly- bands show up on the air photos ‒ these are kar- developed rillenkarren and runnels, usually only renfields developed on belts of pure, thick-bedded a few decimetres long. The rainpits occur mainly and well-jointed limestone (Figure 18). on slopes and vertical faces, and tend to follow One of the more accessible of these karrenfields the bedding structure. Sizes are variable, typi- is on the north side of Colless Creek a kilometre cally ranging from 4 to 25 mm across and they above its junction with Lawn Hill Creek, just west can form hackly surfaces. The flat tops of beds are of Lawn Hill Gorge (Grimes, 1978, 1988). There, generally smooth or finely etched or have various the flat-lying, thick-bedded, pure limestone bed is sizes of rainpits. The etching can be quite detailed, about 45 m thick and large grikes connect down following nets of very fine cracks, and forming to joint-controlled fissure caves. The structure is deeper v-notches ( splitkarren) in larger joints or similar to that at Gregory (Figure 12). The surface the bedding planes (Figure 19). Colour variations 408 KRF•2 • OK.indd 408 15.12.2009 10:59:44 Ken G. Grimes, Tropical monsoon karren in Australia in the cream-coloured dolomite indicate that recently from a number of areas, and are particu- weathering has penetrated a few mm in from the larly well-developed in the Gregory Karst, but be- major cracks. cause of their cryptic nature it is too early yet to Apart from etching and fine pitting, microkar- make deductions about their true distribution. ren are rare at Camooweal, but Reto Zollinger However, the local effects of lithology, structure, (personal comments, 2003) found a variety of cover and denudation history can create consider- well-developed microkarren in a more dissected able variation within that broad tropical theme. area 80 km to the north-east. These included microrills, micro-networks and rasp-like micro- teeth, as well as fine pitting. Micro-pans and A semi-arid tropical monsoon model? micro-tessellations were superimposed on the mi- crorills and micro-teeth. Jennings and Sweeting (1963), and Jennings in later Further south, in the drier area near Boulia papers, described a sequence of development for (Figure 2), I found some well-developed mi- the West Kimberley karst region, with gradation crokarren on cobbles of thin-bedded limestone. from undissected plateau through giant-grikes, The upper, horizontal surface had radiating mi- box valleys, and towers to pediment. Jennings crorills 0.5‒2.0 mm wide and linear microrills (1967) said “it may be that here there is a semi-arid also ran down the vertical sides, but became less tropical monsoonal karst type”. Some later writers pronounced downward suggesting a decantation have taken this sequence to be a purely climatic process. The underside of a loose specimen had model, arguing for or against it on that basis (e.g. fine pits (0.5‒2 mm) where it had been in contact Williams, 1978; Gale et al., 1997). But Jennings with the soil. In this arid region larger mesokar- and Sweeting (1963) also noted local lithological ren are restricted to grikes and rainpits (Andy and structural controls and specifically excluded Spate, personal comments). the impure limestones from their discussion as those formed a quite different “merokarst” terrain. The Jennings and Sweeting sequence is only appli- Conclusion cable to areas of hard, pure, thick-bedded, jointed and flat-lying limestones in a tropical monsoon Australia᾽s tropical monsoon karst have a number climate. Other constraints may also apply: e.g. of surface features in common (Spate and Little, time for the full sequence to evolve, and a dissect- 1995). There are extensive areas of bare limestone. ed plateau with limited vertical relief. Even within The macrokarren have a positive relief with up- such areas, variation in lithological and structural standing limestone towers, pinnacles, scarps and factors and the history of denudation can result in ridges, sometimes with adjoining pediments. The quite distinctive landforms and karren styles. mesokarren are very well developed and include When considering climate, Jennings (1967, extensive and deep grikes, spitzkarren, rillenkar- 1983) noted that although many of these areas are ren, kamenitzas, a variety of sharply-fretted pit- semi-arid the rainfall is concentrated into a short tings, and other forms. However, all these be- wet season and frequently falls as brief intense come less well-developed in the drier areas. Sub- storms. So intense solutional sculpturing of the soil forms such as subsoil grikes, rundkarren, and surfaces, comparable to that of more humid cli- smooth-surfaced pinnacles are locally exposed by mates, is not inexplicable ‒ one would not need soil erosion to form surface fields, as at Chillag- to invoke past wetter climates. However, this idea oe. Directional phytokarst structures and solution is opposed by theoretical modelling by Szunyogh ripples occur in the twilight zones of the caves (2005) which suggests that long periods of gentle and giant grikes. Microkarren have been recorded rain should be more effective for denudation than 409 KRF•2 • OK.indd 409 15.12.2009 10:59:44 Karst Rock Features • Karren Sculpturing short intense falls. Higher temperatures in tropi- Acknowledgements cal regions could also speed the reaction rates, and there may be a greater input from biological Andy Spate reviewed an early draft and provided activity ‒ including micro-organisms and algal information and photos of many of the areas. He coatings. However, in Australia what is probably also made available a collection of photos taken more important is the history of long periods of by Joe Jennings. Susan and Nicholas White, Mick exposure of many of the tectonically stable lime- Godwin and Reto Zollinger provided information, stone areas, which could compensate for the slow- photos and specimens from several areas. Les er rates of sculpturing. Pearson commented on the Chillagoe text. Angel However, Jennings (1981) argued for mode- Ginés discussed the nomenclature and Bernie ration in applying both climatic and other (e.g. Smith discussed etchings and rainpits. Jacques lithological) influences to karst morphology. The Martini discussed the karren zones in the Grego- truth will usually lie between the extreme views; ry Karst. Sources of photographic figures are ac- climate, lithology and structure have all contrib- knowledged in the captions, all unacknowledged uted to greater or lesser degree to the character of photos are my own. Australia᾽s tropical karst. Each area needs to be interpreted according to its local setting. 410 KRF•2 • OK.indd 410 15.12.2009 10:59:44 The Tsingy Karrenfields of Madagascar 32 Jean-Noël SALOMON Giant pointed karren forms, or tsingy, figure these karsts develop on monoclinal, slightly slop- among the most extraordinary karstic landscapes ing limestone plateaus with rocks dating from the of the planet. Because of their distinctiveness, spe- Jurassic to the Miocene. cial tsingy areas have been developed for tourism Rainfall (from 2,200 mm in the north, to less or scientific purposes, mainly as nature reserves. than 300 mm in the south) is concentrated over Madagascar has several tsingy karsts and those a period of a few months. The Madagascar karst of Bemaraha in particular have been included in areas are essentially famous for their extraordi- the Unesco World Natural Heritage list. This kind nary fields of tsingy, that is to say, giant, extremely of giant karren or megalapies, to be found only pointed and spectacular karren or megalapies in tropical zones, is restricted to rigid and highly (Figure 1). Those of the Bemaraha are registered pure limestone outcrops that are more often than on the Unesco World Natural Heritage list and not fractured. They require, moreover, a good deal those of the Ankarana are worthy of inclusion in of time to acquire sizeable dimensions. The study this list. Altogether, tsingy karsts cover about 800 of different tsingy karrenfields in Madagascar and km2. their comparison makes it possible to address the mechanisms of their genesis and their preserva- tion. In other respects, the extreme environments Tsingy karsts that make up tsingy karsts have allowed the de- velopment of both vegetal and animal endemic According to the people of Madagascar, the term forms which make them extremely rich reserves of tsingy (invariable noun) refers to a type of kar- of biodiversity which should at all costs be pro- ren landform that is particularly pointed and tected. which is said to “sing” when hit with a hammer, producing a sound transformed into an onomato- peia: “tsingy”! The tsingy karsts of Madagascar Not all the karsts of Madagascar include tsingy. Only a few have this particularity. They number Madagascar is an extraordinary country for the from north to south: the Ankarana, the Namoroka study of tropical karst phenomena since karst karst and the Bemaraha (Figure 2). Elsewhere, covers more than 30,000 km2, stretching 200 tsingy are only present in rare and confined areas km from the north of the island to the south. All (Narinda, the Kasinge Forest in the Kelifely, the 411 KRF•2 • OK.indd 411 15.12.2009 10:59:44 Karst Rock Features • Karren Sculpturing Figure 1: Extremely jagged karrenfield, called tsingy, in Madagascar. north of the Mikoboka). We shall, see further, on Tanzania), Bom Jesus da Lapa (Brazil), Ta Khli the reasons why. (Thailand), Mount Api (Mulu, Sarawak), Mount Other karstic regions have karren landforms as Kaijende (Papouasia-New Guinea), the Chillagoe tsingy, and apart from China (Ford et al., 1996; massif and the Fitzroy Limestone Ranges (Aus- Salomon, 1997) let us mention the best known: tralia), etc. Tsingy are great karren forms which certain parts of Cameroon, the Kouilou (Congo), are extremely pointed (Figure 3) and which de- the coastal area near Mombassa (Kenya and velop in groups, engendering extraordinary land- 412 KRF•2 • OK.indd 412 15.12.2009 10:59:46 Jean-Noël Salomon, The tsingy karrenfields of Madagascar  „  ‚ € € € ƒ        ­€ „ Figure 3: Tsingy are characterized by great karren forms which are extremely pointed. „ „„ are grooved with vertical scores and are studded with a honeycomb of cupola or micro cavities. As the crests separating two grooves are often in- Figure 2: Location of tsingy karsts in Madagascar. tersected, there result sharp reliefs which stand out clearly. Chiselling makes for edges that are extremely sharp and jagged (Figure 5), and dan- scapes. In my opinion, the word “pinnacle” often gerous in the case of a fall. The sizes of the tsingy used to describe them is not suitable: it is far too differ and range from a few metres to over 30 m inaccurate and qualifies countless morphologies for the biggest. that are often outside the domain of karst. Tsingy shapes do not greatly vary: they are al- ways nose cone-, turret-, or needle-shaped. The The Ankarana acute angles of the needle-shaped ones are always between 15 and 20 degrees, which, on a large The Ankarana massif is situated about 30 km to scale, produces a certain uniformity in an other- the north of Ambilobe and 75 km south/south wise chaotic landscape (Figure 4). In detail, tsingy west of Antsiranana (formerly Diego Suarez). It is 413 KRF•2 • OK.indd 413 15.12.2009 10:59:49 Karst Rock Features • Karren Sculpturing Figure 4: A typical appearance of tsingy karrenfields. composed of a set of limestone dating to the mid- traces of smoke on the walls and the abandoned dle Jurassic (Bajocian and Batonian) measuring fireplaces in the galleries. about 30 km in length and 8 km in width, some- The surface of the plateau is made up of lime- times intruded upon by quite recent basalt (Plio- stone capping rock on which lithology plays a fun- Pleistocene). The whole area covers around 150 damental part. The upper limestone layers, which km2. It is bordered on its north-western side by a are the most chalky, tend to produce reliefs with subvertical 200 m high fault scarp, the “Ankarana residual mounds and slightly depressed dolines. Wall”, which is over twenty kilometres long. However, there is very soon evidence of crystal- The Ankarana massif was not recognized as a line limestone with a change of morphology. The special reserve until 1956, undoubtedly due to the surface then becomes pockmarked with deep lack of knowledge about the area, for it is in fact collapse dolines, cluttered with block fields and unique, with its speleological network of more forming islands of xerophytic vegetation. than 100 km (Radofilao, 1977; Adamson et al., From a tectonic point of view, the Ankarana 1984), the largest in Africa, its wealth of tsingy and behaves like a rigid block. As early as the Jurassic a unique cave-dwelling fauna ( Crocodilus niloti- period, tectonic movements led the block to shift cus, bats, etc. (Wilson, 1985, 1987). However, the towards the west and this tendency continued in most visible caves (Andriabe, Ambatomanjamana the Cretaceous as is shown in the general incline and the underground waterway of the Mahajam- of the slopes in a westward direction. At the same ba) have long been known, as can be seen in the time, the massif was affected by a slight synclinal 414 KRF•2 • OK.indd 414 15.12.2009 10:59:51 Jean-Noël Salomon, The tsingy karrenfields of Madagascar Figure 5: Chisel ing by ril- lenkarren produces sharp reliefs on the tsingy pin- nacles. tendency (four-degree slopes both in the east and m above the surrounding savannah and canyon the west) with an axis situated approximately one beds corresponds to summit areas and zones of kilometre to the east of the cliffs. The consequence intense joint pattern. Indeed, the massif is crossed of these movements is the presence of a series of with deep bogaz and narrow, vertically-walled fault scarps, more or less parallel to the Great Wall gorges, oriented on the mainly NW-SE fractur- (Figure 6). ing. Some of them (the Forestier canyon) can be The actual tsingy karst which rises about 200 flooded up to a height of 30 m during periods 415 KRF•2 • OK.indd 415 15.12.2009 10:59:53 Karst Rock Features • Karren Sculpturing   † ­€ „ ƒ ‚ Figure 6: Morphologic sketch of the tsingy massif of Ankarana. of cyclonic rainfall. One part of these corridors latitude 16°20’ and 16°30’S. This 100 m thick table corresponds to the collapsing of the roofs of un- of crystalline limestone (but with variations tend- derground galleries oriented by faults; others are ing towards dolomitic) in many ways evokes the simply open fractures of tectonic origin that dis- Ankarana and contains anastomosed speleologi- solving and collapsing processes of the walls have cal networks (the Ambovonomby one has a de- consequently widened. Some peaks, separated by velopment of several kilometres). It is extremely profound crevices, measure over 20 m in height marked by often circular fracturing as can be seen and the landscape may look as if it is a turret karst. in the numerous corridors that cut through the All in all, it should be remembered that, on the massif, being colonized by xerophilous forest. In one hand, these tectonic movements can account detail, a network of open joints and grikes divides for a strong NW–SE joint pattern and that, on the surface of these massifs into blocks. Because of the other hand, the tsingy as well as the general this, an overhanging position of the upper blocks configuration of the cavities and the karst have is often to be found, giving to the landscape a very been considerably influenced by the rigid nature spectacular aspect. of brittle limestone. The tsingy chisel most of the Namoroka towers (some of which reach a height of 80 m) and some are more than 10 metres wide. They correspond The karst of Namoroka again to highly pure crystalline limestone (99.2% CaCO with only slight porosity: < 3%). As soon as 3 The karst of Namoroka (about 160 km2) is situated the limestone becomes more dolomitic (that is to at 150 km to the south-west of Majunga, between say, near the base), they disappear. 416 KRF•2 • OK.indd 416 15.12.2009 10:59:55 Jean-Noël Salomon, The tsingy karrenfields of Madagascar   † ‚ ­    ­­€‚ „­ ˆ„ ­ƒ„  ­ „­    €‡„ „ € Figure 7: Morphologic sketch of the tsingy plateau of Bemaraha. 417 KRF•2 • OK.indd 417 15.12.2009 10:59:56 Karst Rock Features • Karren Sculpturing The Bemaraha • the central plateau is dotted with kuppen and dolines. The kuppen correspond to hard lime- The Bemaraha karst lies on the west coast of Mada- stone with a thickness of some tens of metres gascar and covers a surface of about 4,000 km2, situated above levels of marly-limestone. The between latitude 17° and 20°S. It stretches from tops of the mounds are trenched with tsingy the Ranobe in the north as far as the great Tsiribi- from which a number of trees emerge, whereas hina river in the south. Its width varies from 4 to the marly-limestone floors are more water-re- 5 km in the north to 25 km in the south, between sistant and are covered in savannah. The actual the canyons of the Manambolo and the Tsiribihi- centre of Bemaraha is made up of a huge field of na where the plateau is the most spectacular. The kuppen separated by closed depressions which region is situated in a tropical climate area with are usually shallow since they are trapped on a a long dry season, but rainfall in season can be more impermeable level. They are large mounds heavy and concentrated. (up to 80 m high) with steep slopes (20 to 30°) From a geological point of view, the plateau laid out in groups of several individuals; corresponds to limestone outcrops of the Mid- • the Antsingy. The whole western border of the dle Jurassic. To the west, it is covered by marl and plateau which overhangs the western scarp marly-limestone of the Upper Jurassic and by the is the realm of an extraordinary landscape: Cretaceous continental sandstone. Its eastern and tsingy entablatures. The heart of the area, the western frontiers are well-defined since to the east Tsingy Nature Reserve of Bemaraha, covers there is a scarp face and to the west a scarp partly 152,000 hectares of a typical tsingy natural delimited by faults. On the south side, it is bor- biotope with caves, springs, a primary forest dered by the deeply embanked Tsiribihina valley, and calci-xerophilous growing formations whilst in the north it fades into small mounds of inhabited by a rich endemic fauna. It is actually ever decreasing height (Figure 7). Rossi (1980) dis- a limestone table of about 400 km2 that has tinguishes several areas according to lithology: been subjected to fracturing and is jointed in • the north of the plateau with its scattered large all directions (main directions: N20°W to N30° conical cupola-shaped mounds and its steep to N40°). They are open fractures that have cut slopes (mogotes). The smaller ones have a di- out tsingy blocks, engendering a landscape that ameter of around 100 m and are 20 to 30 m is not unlike that of the Ankarana but with a tall; the larger can have a major axis reaching greater scope. The sight when flying over the 2 km and a height of around 100 m. Between sharp blades of the tsingy is unforgettable: the mogotes, collapse-dolines are regularly knife blades, turrets, needles and tsingy follow present. Some summits are affected by tsingy each other in an endless succession, separated but the latter do not make up continuous ta- by deep corridors in which decalcification clay bles. Much of the karst has been covered with and vegetal products pile up. Similarly to the red clay coming from the breakdown of ba- Ankarana, the relief of these tsingy corresponds salt. Today, erosion causes the disappearance of to a thick table (150 to 200 m) of very pure, non- this cover and reveals crypto-corrosion shapes: dolomitic and non-porous (2 to 4%) micro- limestone blocks with smooth ovoid or ogival crystalline massive limestone. The density of shaped sides (rundkarren, small “stone teeth” fracturing would seem to be the fundamental or “dragon teeth”) that are sometimes pierced element as it is from the thick network of with cavities corresponding to what was once joints that the dissolving process carves out an opening made by roots. The limestone here tsingy. Even if some main directions may is very pure (> 95% CaCO ; 1 to 2% MgCO ) easily be identified (they are underlined by the 3 3 with low porosity (1 to 3%); appearance of volcanic seams) the density of the 418 KRF•2 • OK.indd 418 15.12.2009 10:59:56 Jean-Noël Salomon, The tsingy karrenfields of Madagascar Figure 8: Evolution of the various stages leading to the formation of tsingy, then to their deterioration: 1. crypto-karst stage below the soil and lateritic soil. Possible presence of forest. Karst develops due to diaclases and primary frac- tures; 2. dragon teeth stage. Upper layer removed by erosion, limestone is shaped (rundkarren); 3. pinnacle stage. Single towers and pil ars appear, rillenkarren and dissolution coupolas develop. Gorges and bogaz along the frac- tures and diaclases filled up by clay; 4. tsingy stage. This stage occurs when the limestone summits are very pure and massive and exposed to meteoric waters for a long time. In enlarged and developed bogaz some vegetation occurs; 5. developed stage. Landscape of isolated towers, enlarged gorges filled by breakdown and sediments. At certain low points phreatic zone appears. joints and grikes is such that all directions are and the Anjohy Kibojenjy network to 4,972 m. represented as orienting the networks. Among In all, more than a hundred cavities (Zohy in the latter, the Ming network extends to 1,050 m the Sakalaka language, or local language) have of galleries, that of Zohy Siramany to 1,950 m been counted (of which 12 are more than 1 km 419 KRF•2 • OK.indd 419 15.12.2009 10:59:57 Karst Rock Features • Karren Sculpturing long) and more than 53 km have been mapped in calcite, limestone becomes less and less porous (Delaty, 2000). as time goes by. Since 1996, several circuits have been devel- In Madagascar, it is the older (Jurassic) lime- oped starting out from Antsalova (“Petits tsingy”, stone that has tabular surfaces featuring tsingy. “Grands tsingy”). These circuits make it possible to All this should be set in the context of the perma- visit this magnificent karst and the gorges of the nence of a great structural stability of limestone Manambolo river. bastions, limiting mechanical erosion with time. Whatever the case, we must acknowledge a rela- tively ancient evolution in which water has always The genesis of the tsingy played an essential part and in which the tem- perature factor has always been secondary, being A comparison of the tsingy karsts described above subequal throughout the year. In particular, the makes it possible to identify a number of common absence of frost accounts for the fact that tsingy factors accounting for the genesis of the tsingy. have been preserved and have reached a consid- erable height (up to 20‒30 m in the Bemaraha). It can be observed that if tsingy are not present The role of the climate and the time factor in temperate or cold countries, it is because they are destroyed by frost as soon they reach a cer- In the case of Madagascar, most karst landforms, tain height. There are some outstanding pointed including therefore the tsingy, developed in a karren (lapies) in the Pyrenees (Arres d’Anie) and tropical climate with alternating seasons (rainfall in the Alps (Désert de Platée) but they only very exceeding 1,000 mm and a pronounced dry sea- seldom exceed 2‒3 m in height: they do not have son). The observation of other areas of the plan- the time to develop further. In arid and semi-arid et featuring tsingy karsts shows that the situation countries, drought and the predominance of me- is the same and this would seem to indicate that chanical erosion over chemical erosion account tsingy are subjected to tropical climates. Moreo- for the absence of tsingy. ver, the tropical climate atmosphere has long af- fected these regions, often since the Jurassic at least. Traces of this are numerous: tropical red The role of the structure: lithology and soils, strong weathered coverings and rubble fracturation (from broken up ferroaluminous crusts and so on). Only a few variations, such as the lengthen- The part played by lithology is clearly demonstrat- ing or shortening of the dry season or else of the ed in the development of different types of point- total annual rainfall have played a part. It can be ed karren, from the stage of dragon teeth (Ford et said that the main features of karst massifs that al., 1996) to the true tsingy (Figure 8). are visible today (tectonic movements, division As a general rule, morphologies (rundkarren, into large compartments, the development of net- dragon teeth) established under laterite (clay and works and so on) were established at the end of the colluvium silt) and a soil cover are similar what- Tertiary. The Plio-Pleistocene period brought only ever the lithology. However, once they are exposed slight alterations (volcanic intrusion, incision to the open air, the teeth formed in pure limestone of valleys and surface stripping) but which are gradually acquire pointed or ogival shapes caused not without importance as far as detailed minor by superficial dissolving. This can go as far as to forms are concerned, an example being the tsingy. produce the formation of the pointed pinnacles, The time factor also plays a part in so far as that, that are chiselled and carved of rillenkarren and through gradation neomorphism and enrichment honeycomb cupola. In detail, the reliefs are very 420 KRF•2 • OK.indd 420 15.12.2009 10:59:57 Jean-Noël Salomon, The tsingy karrenfields of Madagascar sharp. On the other hand, in impure and more • a relatively slight dip (< 5°). Indeed, a more pro- porous limestone (dolomitic) the forms that de- nounced dip would influence the flow of water velop are more rounded and of ruiniform type or and would limit the vertical penetration by the sometimes mushroom shaped. The sharpest and joint pattern; most spectacular shapes form tsingy which often • heavy and concentrated rainfall (everywhere supply the finishing touch on the tops of columns. more than 1,200 mm). When it rains, water In the end, the formation of tsingy requires condi- flows over the rock faces as limestone has low tions which are specific and always the same: porosity and is too water-resistant for rain to • very pure crystalline or microcrystalline lime- penetrate. Often the water will not get as far as stone (more than 95 or even 98% of CaCO ) with the base of the tsingy since it evaporates before- 3 low porosity (always less than 1 to 2%), which is hand in contact with the overheated rock face. the case of the tsingy formations of the Anka- Dissolving is just superficial and is decreasingly rana and of the Bemaraha but also of the “Stone effective from top to bottom. The rate of dis- Forest” of Lunan (China) and the karren spires solution depends first of all on the quantity of of Mulu (Sarawak). The nearly complete absence rain falling on a given limestone surface (sur- of porosity results in water being unable to pen- face and sides of the tsingy) and secondly on etrate the rocky mass and causes it all to run how aggressive the rain is (with a decrease from over the walls of the tsingy. Dissolving, facili- top to bottom due to the saturation and evap- tated by the extreme pureness of the limestone, oration). Dissolution only occurs on the sur- is only superficial and laminar: this results in face. But in fact, as the water is concentrated in the high development of the tsingy and the ver- cracks, the latter deepen more rapidly than the tical cracks which form ridges on their sides. It peaks of the tsingy. This leads to the formation also accounts for the absence of decalcification of tsingy which are higher and higher the older residues. However, differences in porosity of the they get as well as to the genesis of shapes that subhorizontal shelves can explain the develop- are more and more pointed, in some cases with ment of stratification joints and the many over- an angle of 10 to 20°. Once this angle has been hangs that are visible along the walls and that established, the faces seem to recede parallel to are often unstable, as well as the presence of col- themselves in such a way that the angle conse- lapsed blocks at the base of the latter when evo- quently remains the same. lution goes back a long time. The thickness of The total absence of vegetation, or at least its the limestone is of course an important element rareness, linked to the absence of soil or the weak- in explaining the height of the tsingy; ness of the corrosion in contact with the latter. • a very strong fracturation producing bogaz Tsingy develop in an environment of bare karst. and corridors lined up on the fracture as well But is this absence not a consequence of the pres- as a great many intersections of vertical frac- ence of the tsingy (the steepness of the slopes pre- tures that are often more than a metre wide and vents any soil accumulation)? On the other hand have an intense joint pattern. The fracturation bogaz soils may hold well grown vegetal forma- network is a prerequisite for the organization tions (e.g. xerophytic dry forests). of the patterns of the tsingy. This fracturation However, the broadening of joints makes it pos- was furthered in Madagascar by the interven- sible for organic debris to be trapped and some- tion of recent volcanic activity (intrusions) and times flowing water will reach the base of the the rigid nature of the limestone tables. The tsingy, thus accounting for the corrosion features widening of the cracks can also be explained by that may sometimes be observed there. Once the consecutive dissolving by water running after deepening stops (sometimes because of blocking heavy rainfall; at a more impermeable level), the bogaz widens 421 KRF•2 • OK.indd 421 15.12.2009 10:59:58 Karst Rock Features • Karren Sculpturing and it is possible to make one’s way without any the main factors accounting for their genesis. On a problem along the clay bottom stuffed with vege- relatively reduced time scale, tsingy require some tal debris, as is the case in the Bemaraha. millions or even tens of millions of years to de- velop. However, they are fragile forms, sensitive to tectonics and to gravity (overhangs and collapses). Conclusions These exceptional landscapes, which in Mada- gascar harbour treasures of biodiversity as far as The comparison of the various karsts of Madagas- both flora and fauna are concerned, are likely to car (as well as those of other regions of the world) be of increasing interest to the tourist industry makes it possible to underline the rareness of such as their existence becomes more widely known landscapes (Salomon, 1997). The pureness of the (through television programmes, sightseeing cir- limestone, its lack of porosity and the intense for- cuits, books of photographs, and so on). For this mation fracturing of the latter in addition to the reason, it is necessary to take immediate protec- permanence of a tropical climate atmosphere, are tive measures. 422 KRF•2 • OK.indd 422 15.12.2009 10:59:58 The pinnacle Karrenfields of Mulu 33 Mick DAY and Tony WALTHAM The pinnacles on the northwestern flanks of Gu- sequence that has been folded into an anticlino- nung Api, in the Gunung Mulu National Park, in rium and partially metamorphosed. At the core northern Sarawak, are perhaps the world’s largest of the anticlinorium and occupying the highest individual spitzkarren and certainly among the elevations are the slates, slatey shales and quartz- most impressive pinnacle or shilin karrenfields in itic sandstones of the Mulu formation (Liechti et any karst landscape. They were first described by al., 1960). These are overlain sequentially by the Wilford and Wall (1965), but remained virtually Melinau limestone, which crops out as a belt on unknown until 1977‒1978. They are now recog- the northwestern flanks of Gunung Mulu and by nized as the “type example” of pinnacle karst, but the Setap Shale, which forms the bulk of the west- their precise morphology, their exact distribution, ern lowlands and ridges. and their mode of formation are still known only The Mulu pinnacles are formed in the Melinau broadly, and they certainly warrant further study. limestone (Upper Eocene-Lower Miocene), which These “classic” pinnacles represent a dramatic sub- forms a dissected karst escarpment about 30 km set within the overall pinnacle assemblage within long and < 5 km wide, rising to 1,600 m from the the Gunung Mulu Park, and their relationship to alluvial lowlands on the northwest flank of the the broader karrenfield also warrants further re- metasedimentary spine of Gunung Mulu itself. search. The limestone is lenticular, 1,500 to 2,000 m thick, massively bedded, and dips steeply at 40‒50°, more in some locations, to the west-north-west, though Setting bedding planes are rarely visible in surface expo- sures (Osmaston and Sweeting, 1982; Waltham The Gunung Mulu National Park has a relatively and Webb, 1982; Waltham, 1997a). The limestone simple geological structure expressed by striking is dominantly a white to blue-grey calcilutite that topography that rises from less than 100 m a.s.l. to includes some fine calcarenite, is locally recrystal- over 2,400 m within a distance of less than 10 km lised to resemble a marble, and has a dolomite con- (Sweeting, 1980). The western 38% of the Park is tent ranging from 2 to 20%. The carbonate is pre- lowland with some ridges, but the eastern portion dominantly lagoonal, and the macrofauna is sparse, is mountainous, and about 25% of this is karst ter- with the most prolific fossils being foraminiferal rain. Essentially, the geology consists of a broad- tests (Adams, 1965; Waltham and Webb, 1982). ly conformable Paleocene-Miocene sedimentary The limestone belt is cut into several distinct 423 KRF•2 • OK.indd 423 15.12.2009 10:59:58 Karst Rock Features • Karren Sculpturing .Benarat. Sarawak Mulu .The Pinnacles. Borneo .Api. 45 0 3km . . Gunung Mulu river cave river Figure 1: Location of The Pinnacles on the limestone ridge of Gunung Api, within the Gunung Mulu National Park, Sarawak. blocks by the through-flowing allogenic rivers of Sweeting, 1982; Brook et al., 1982; Waltham, 1995, the Melinau, Melinau Paku and Medalam (Figure 1997a, b; Waltham and Brook, 1980a, b). 1). Within the karst blocks between the three river It is uncertain whether the limestone also ex- gorges, the peaks of Gunung Api and Gunung tends below the alluvial Melinau plain (Wilford, Benarat rise to elevations of about 1,700 m above 1961; Osmaston, 1980). Bordering the karst mas- very steep, vertical or undercut flanks. All the sifs, a series of gravel terraces are the outer por- karst blocks are drained internally by massive and tions of massive alluvial fans that head in the complex cave systems that are still not completely gorges where they are fed by denudation of the mapped (Eavis, 1980; Waltham, 1997b). Their sur- non-carbonates of Gunung Mulu itself. These faces are ubiquitous jagged pinnacle karrenfields, gravels have been transported largely above shrouded in dense rain forest and punctured by ground through the breaches in the limestone open shafts, large collapsed dolines, and a few escarpment, aggrading during wetter interglacial dismembered large dry valleys (Osmaston and periods and being incised during drier glacial epi- 424 KRF•2 • OK.indd 424 15.12.2009 10:59:59 Mick Day and Tony Waltham, The pinnacle karrenfields of Mulu Figure 2: Aerial view of pinnacles along the ridge of Gunung Api (photo Malaysian Tourist Board). sodes (Rose, 1982, 1984a, b). They also extend into tersecting rectilinear patterns in the vegetation the caves, and Farrant et al. (1995) have used them that appear to correspond to fracture patterns to date cave wall notches, which correlate with re- in the limestone itself. The limestone soils are surgence levels defined by interglacial aggradation lithosols thin, skeletal, highly organic silt or clay intervals, to demonstrate a mean base level lower- loams with limestone rubble (Baillie et al., 1982; ing rate of 0.19 m per 1,000 years over the past 2 Anderson and Chai, 1982). million years. Interpreting the rate of base level The climate accelerates contemporary carbon- lowering as the rate of isostatic uplift in response ate dissolution within the karst, producing an to regional denudation indicates that internal ero- inhospitable, sharply fretted surface that is very sion of the karst blocks has been minimal, and the difficult to traverse or investigate in detail. Many present karst topography has evolved over the past of the lower slopes are mantled by talus, much of 10 million years. which is cemented by calcite. Soil carbon dioxide The Mulu karst experiences an equatorial mon- levels may be locally high, but measurements are soonal climate, at a latitude of 4°N. Temperatures generally less than 0.1% (Friederich, 1980; Laverty, range within 20‒30°C, and mean annual rainfall 1980). Provisional data from limestone weight loss is about 5,000 mm, distributed evenly through- tablets (Day, 1981a) indicate that limestone weath- out the year (Walsh, 1982a). Vegetation is gener- ering rates within soils and in riverine locations ally dense rain forest, but that on the limestone are ca. 100 mm/ka, within the range measured is distinctive, with numerous endemic calcicolous elsewhere in tropical karst, but that rates on steep species exhibiting altitudinal zonation (Anderson rock faces are much lower, at around 10 mm/ka. and Chai, 1982). Osmaston (1980) pointed out in- Solute loads in the rapid run-off are low (Fried- 425 KRF•2 • OK.indd 425 15.12.2009 11:00:00 Karst Rock Features • Karren Sculpturing Figure 3: The Pinnacles, the classic site seen from the overlook high on Gu- nung Api. erich, 1980) and regional solute loads similarly are more gently sloping terrain of the summit ridges. unremarkable by tropical karst standards, gener- Further investigation is needed, with larger sized ally falling within the 80 to 150 mg/l range (Walsh, random samples. Ley (1980) measured some of the 1982b). By contrast, annual runoff is considerable, “classic” pinnacles at 35 m tall, and estimated that probably near 4,000 mm, and the chemical denu- some individuals might be up to 100 m in height, dation rate is about 125 mm/ka (Walsh, 1982b). though this would seem to be an overestimation, and such individuals would certainly represent statistical outliers. Many of the seriously inaccessi- Pinnacle morphology and ble high regions of the karst blocks have only been development observed from helicopters. Giant pinnacles pro- trude from the forest canopy over areas of the Api Pinnacle karrenfields are widespread within the and Benarat ridges (Figure 2) far beyond the well- Mulu karst, occupying at least 30% of the limestone known classic pinnacles on Api, and some of the surface on the Gunung Api and Gunung Benarat very steep (but not vertical) mountain flanks are blocks. Adequate sampling is almost impossible almost ladders of tall pinnacles half-hidden by the on the inaccessible and forest-clad mountains, but very large trees that grow between them. observations and limited, non-random sampling The pinnacles’ type example is a cluster known suggest that about half the pinnacles throughout as The Pinnacles high on the northwestern flank the area are less than 5 m tall, with 30% less than of the Gunung Api ridge (Figure 3). This site lies 2 m tall, while about 20% reach more than 20 m in in a shallow, steeply inclined valley aligned north- height, and the type examples on the flanks of Gu- east-southwest at about 1,200 m, and contains nung Api may reach to 50 m. This suggests that the perhaps no more than 100 individual pinnacles, size distribution is not normally distributed, and within an area of just a few hectares (Osmaston, supports previous observations about the size dis- 1980; Waltham, 1995, 1997a). Between 30 and 50 tribution (Osmaston, 1980; Waltham, 1995, 1997a). m in height, these enormous individuals protrude The available evidence also suggests that the larg- 10‒20 m above the forest canopy. Many pinnacles er pinnacles generally are at the higher elevations, are connected to form contiguous but dissected but this may be due to their prominence on the arêtes or lines of summits (Ley, 1980) and others 426 KRF•2 • OK.indd 426 15.12.2009 11:00:02 Mick Day and Tony Waltham, The pinnacle karrenfields of Mulu Figure 4: Pinnacles that appear to be clustered around deep shafts, on the edge of The Pinna- cles group. are grouped in circular pattern, resembling the Although surface run-off down the pinnacles remnants of near-vertical shaft margins (Figure has not been documented, by inference and anal- 4). The dominant pinnacle summit alignment is ogy it occurs frequently, rapidly, and with dissolu- close to north-south, so that the pinnacles cut tion limited by the restricted residence time. Thus across the minor valley, with their summit eleva- water reaching the pinnacle bases remains aggres- tions decreasing towards the valley axis. Osmas- sive, and may be rendered more so by contact with ton (1980) reported that the dominant fracture the suspended organic root mat, and dissolution sets over most of Gunung Api are orientated at 22° is focused within the epikarst below the pinnacle and 85°, but that these were displaced by about 10° bases, though this is still far above the known ac- clockwise near The Pinnacles. tive cave systems (Brook et al., 1982). 427 KRF•2 • OK.indd 427 15.12.2009 11:00:03 Karst Rock Features • Karren Sculpturing Figure 5: Immature rillenkarren grooves cut into the Figure 6: Deeply fretted honeycomb karren near the rounded flanks of a rundkarren dome since the latter crest of one of the Gunung Api pinnacles. was exhumed from its soil cover, beside the trail up to The Pinnacles. Ley (1980) postulated the potential role of high steep to maintain any cover of macro-vegetation, humidity, cloud condensation and mist in main- and even algae may be scoured by run-off and taining the sharpness of the pinnacles. Noting desiccated under the hot sun in intervening dry that mists often wetted pinnacle summits, and periods. The lack of blue-green algae is evident that rainfall run-off registered solute loads of in the stark whiteness of the largest pinnacles, in only 10 mg/l, compared to 60 mg/l for condensa- contrast to the blackened surfaces that character- tion run-off, Ley suggested that the ril enkarren ise most smaller pinnacles in China and elsewhere. themselves were produced by torrential stormwa- Given the significant role of organic agencies in ter run-off, but that their razor-sharp edges were other tropical karst environments (Viles, 1984, maintained by condensation corrosion. Although 1988), it would be surprising if the pinnacles did this contention provoked initial skepticism (Ley, not have at least some biogenic component, but 1980; Osmaston, 1980), it is noteworthy that the this requires further investigation. At the transi- role of condensation in carbonate dissolution has tion from the fluted upper pinnacles to their more since received increased attention (e.g. Dublyan- blocky and rounded pedestals, there is a sus- sky and Dublyansky, 2000), and its role in shaping pended organic mat where vegetation thrives on a the pinnacles may warrant further attention. tangled root mass of skeletal organic soils, besides The remnant limestone blocks have sides too rooting in soils that clog the diminishing widths 428 KRF•2 • OK.indd 428 15.12.2009 11:00:04 Mick Day and Tony Waltham, The pinnacle karrenfields of Mulu Figure 7: Razor-sharp crests on some of the smaller pinnacles on Gunung Api. of the intervening fissures. From the root mass, what controls the position of the mat, and whether trees rise to heights in excess of 20 m, with the it marks a formerly less dissected surface between white-grey fluted pinnacle spires emergent above the individual pinnacles themselves. the canopy with their rock faces inclined at 70‒85°. At their bases, although generally obscured by The precise role that the suspended organic root soil and vegetation, the pinnacles have broadly mat plays in pinnacle evolution has to date been blocky, sometimes rounded and often densely pit- conjectural, and there remain questions, such as ted margins (Figure 5), and typical sub-soil rund- 429 KRF•2 • OK.indd 429 15.12.2009 11:00:05 Karst Rock Features • Karren Sculpturing karren are widespread where the vegetation mat had collapsed” (Osmaston, 1980). Fallen lime- has been stripped away from the rock by micro- stone blocks occur elsewhere too, especially close earthslides (Waltham, 1995, 1997a). Above their to steep, undercut cliffs along the flanks of the bases, however, the pinnacles are upward tapering limestone belt, and some of these appear as iso- and with a grossly blade-like morphology, which lated pinnacle rising out of the alluvial soil cover. Osmaston (1980) attributed to a tendency for the All across the forested limestone slopes, small two upper corners of elongate pinnacle blocks to pinnacles project from the soil and understory be eroded. The upper portions of the pinnacles, vegetation, and many are distinguished by razor- above the vegetation mat, are deeply fluted by sharp edges (Figure 7) ‒ sharp enough to slice to dissolutional runnels, producing exaggerated ril- the bone the thigh of a falling walker. Many of the lenkarren with razor-sharp edges. Osmaston and pinnacles in the shadows of the forest lack the dra- Sweeting (1982) recorded a range of flute sizes, be- matic proportions of their larger equivalents, but tween 16 and 590 mm in width, 2 to 500 mm in details of their morphology need further investi- depth, and up to 15 m in length. Ley (1980) noted gation. that rillenkarren were particularly well developed on the more exposed southern and western faces of the “classic” pinnacles. Pinnacle geology At their uppermost extremities, above the ril- lenkarren zone, most pinnacles have sharp, point- Other factors may have played some role in pin- ed spires, but some have grotesque, labyrinthine nacle development, particularly in enlargement shapes reflecting assemblages of what Osmaston of the classic site of The Pinnacles on the north- (1980) terms honeycomb karren (Figure 6). These western flank of Gunung Api. Frost action dur- have received little attention, and it would seem ing the Pleistocene is unlikely to have occurred, that the pitting could be due to either biokarstic much less had a pronounced influence, and sur- processes or to variations within the rock lithology. face profiles mean that unloading stress relief can It would be interesting to compare their morphol- only have had some local influence on joint acces- ogy to that of the karren developed on the ped- sibility. Local focusing of surface drainage, either estals beneath the vegetative mat. Another unre- on the early limestone surface or from an overly- solved issue is the role of bedding, if any, in in- ing non-carbonate surface (developed on the Ter- fluencing pinnacle micromorphology. Osmaston tiary setap shales) may also have been a factor. In (1980) noted that there were “faint signs of bed- many areas of mature fluviokarst, deep fissures ding on the pinnacles and only occasional fracture on joints that parallel the contours of steep valley planes with moderate easterly dips”, and that the sides have a jagged morphology that emulates pin- latter “typically show as an undercut ledge with a nacles. Consideration of comparisons with the de- rough surface that is suggestive of a stylolitic pres- velopmental history of the Lunan shilin promot- sure contact.” ed the conjecture that “there is a clear possibili- There has been only limited research on the ty that these shales overlay a Tertiary fossil karst, broader array of pinnacles within Gunung Mulu, which was exhumed to provide an initial stage of but Osmaston (1980) made some general observa- development of the modern pinnacles” (Waltham, tions and reported on a grouping that had been 1995). Alternatively, the apparent location of the investigated by Hans Friederich (Friederich, 1980) largest pinnacles at relatively high elevations may in a small doline low on the southwestern slopes suggest that they are simply the oldest such fea- of Gunung Api. There, too, were “sharp crested tures; it is difficult to assess the age of The Pinna- monoliths, but much smaller, due to the closer cles, but a realistic timescale is within 10‒100 ka fractures, and at the bottom of the doline many (Waltham, 1995). 430 KRF•2 • OK.indd 430 15.12.2009 11:00:05 Mick Day and Tony Waltham, The pinnacle karrenfields of Mulu One factor in the development and mainte- ity and time and an adequate rate of dissolution, nance of the dramatic pinnacle morphology is the pinnacles will be derived from the walls of deep- mechanical strength of the limestone itself (Day, ening wedge-shaped fissures, with the slopes of 1980, 1981b, 1982b). Schmidt hammer hardness, the walls depending on the ratio of the rates of a surrogate for compressive strength, is 56 (Day, vertical and lateral dissolution; his model has the 1980) ‒ but this is within the upper end of the range intersections of fissures producing wedge-shaped for limestones that support many other karst ter- pinnacles, the heights of which are proportional rains. The limestone is very massive, facilitating to the fracture spacing, and whose relationship to the formation and persistence of pinnacles within the initial form of the upper surfaces of the lime- very thick individual beds; the singularly mas- stone blocks is minimal. sive structure of the Melinau limestone appears Osmaston considered two aspects of Mulu’s to account for it containing some of the world᾽s pinnacles: 1) the general formation of such sharp- largest cave chambers, in addition to some of the crested protrusions, and 2) the specific develop- world᾽s tallest pinnacles. The limestone also has a ment of the exceptionally large “classic” pinna- very low primary permeability, and is extremely cles. With respect to the latter, he concluded that pure, with non-carbonate contents less than one “…the local factor which is responsible for the percent (Waltham and Webb, 1982). Osmaston formation of The Pinnacles is the exceptionally (1980) suggested that the “classic” pinnacles might widely spaced, rectangular, vertical fracture pat- be formed in an area of particularly pure calcite, tern coupled with an absence of other fractures but this suggestion has not yet been investigated or bedding. This has a threefold importance: it further. leaves strong intact blocks that are large in plan; it The pinnacles are assumed to have developed leaves wide fissures between them; and, because by dissolution focused on nearly vertical joints ‒ the blocks are stable, they have time to assume though definitive evidence of this awaits access a profile of dynamic equilibrium which can de- to, and mapping within, cave passages immedi- scend with the general landscape surface” (Os- ately beneath some of the taller pinnacles. Mean maston, 1980). spacing of the defining joints at right angles to each other in the “classic” Pinnacles area has been estimated at between 10 and 20 m (Osmaston, Analogous features 1980). It may be somewhat less elsewhere, thereby restricting formation of the larger pinnacles to The Mulu pinnacles are broadly analogous to locations where wider spacing results in the fo- giant spitzkarren karst at various other locations. cusing of larger volumes of rainwater into each The shilin (stone forests) of China, including the fracture (Osmaston and Sweeting, 1982). Ley type locality at Shilin, Yunnan (Song et al., 1997), (1980) invoked bedding planes dipping at about tend to cover much larger areas with unbroken 60° to the northwest and joints dipping at 45° to seas of pinnacles that do not reach the heights of the east in explaining the shape of the pinnacles, those in Mulu. The tsingy of Madagascar (Rossi, but also portrayed vertical development (presum- 1974; Salomon, 1997) are even more extensive, but ably down another set of joints) in his develop- only small areas have pinnacles heights compara- mental model. The role of vertical joint sets was ble to those in shilin karst. The assegai landforms raised in subsequent debate (Waltham, in Ley, of Palawan, Philippines (Longman and Brownlee, 1980) and in other reports (Osmaston, 1980), but 1980) are larger than shilin, and similar terrains the issue has not been fully resolved. Osmaston forming the arête karst around Mount Kaijende, (1980) presented a simplified model of pinnacle New Guinea (Jennings and Bik, 1962; Beck, 2003) evolution suggesting that, given sufficient stabil- appear to have pinnacles that equal or exceed the 431 KRF•2 • OK.indd 431 15.12.2009 11:00:05 Karst Rock Features • Karren Sculpturing dimensions of those in Mulu. Waltham (1995, Context 1997a) draws a distinction between normal pinna- cle karst, such as that in Mulu, which has formed The Mulu pinnacles are landforms of significance on steep slopes in high limestone mountain rain to geomorphology on a world scale. They are pro- forests, and the shilin sub-type, which has evolved tected both by their inaccessibility and by the es- through multiple phases on gently dipping lime- tablishment (since 1974) of the Gunung Mulu Na- stones. There are also more distant analogies to tional Park, which was inscribed as a U.N. Natural the pinnacles formed in indurated calcareous aeo- World Heritage Site in 2000. Along with the caves lianites in the Nambung National Park in western and other karst features, they represent a signifi- Australia (Ford and Williams, 1989), and to non- cant attraction of the park, and deserve appropri- karstic pinnacles formed in soils and other rocks ate attention in the development of future man- (James, 1997). agement strategies (Anderson et al., 1982; Day, 1979, 1982a, 1983). 432 KRF•2 • OK.indd 432 15.12.2009 11:00:05 arÊTe and pinnacle KarsT 34 of MounT KaiJende Paul W. WILLIAMS Mount Kaijende is located in the Western High- humid or seasonally humid at sea level to glacial lands district of Papua New Guinea (Figure 1) above the snow line at about 4,600 m a.s.l. Mean between the settlements of Laiagam and Porgera annual temperatures decrease with elevation at at latitude 5°30’S. Arête and pinnacle karst oc- a rate of about 5.8°C/1,000 m, consequently the curs near its summit, which rises to about 3,500 upper slopes of Mount Kaijende are cool with a m above sea level, the regional tree-line being at mean annual temperature near the summit of around 3,800 m a.s.l. The karst of the area was about 10°C and the tropical location implies lit- first brought to the attention of scientists through tle seasonal variation. Nevertheless, there is a sig- the work of Jennings and Bik (1962) and investi- nificant diurnal variation, with a range of about gated further by Williams (1971, 1972). Other oc- 12°C being measured at 2,800 m a.s.l. Night time currences of arête and pinnacle karst have been frosts are common above 2,430 m, so freezing observed from aircraft further west in Papua New conditions must often affect the arête and pinna- Guinea and Papua, but these sites have not been cle terrain. The mountain is almost always cloud scientifically investigated. This account draws on covered, thus fog-drip must make a large contri- the work cited above and on a recent review by bution to annual precipitation. Rainfall at Porgera, Williams (2004). which is 8 km to the west and at 2,200 m a.s.l. in Although Jennings and Bik named the mor- a nearby valley, has been measured at about 3,700 phology “arête-and-doline karst”, Williams (1972) mm, so it is likely to be considerably more at the considered it misleading to imply that the en- summit. The arête surfaces are essentially bare on closed depressions resemble dolines as they are their crests, being in an exposed hostile environ- normally understood, because of their very steep ment near the upper limit of montane forest. Frost rock sides and intimate connection to the arête shattering was probably common during the Last ridges; so he preferred to omit the term doline Glacial Maximum, because at that time the snow- from the description of the terrain, which he de- line was at about 3,550 m (Löffler, 1977). scribed as “arête and pinnacle karst” (Figure 2). At an elevation of about 3,200 m a.s.l., Williams (1971) commented that “thick, virtually impen- etrable moss-forest, swirling in mist, clings like a Bioclimatic environment sodden cloak to the rugged slopes”. Daytime tem- perature at that height was 11°C and water drip- In Papua New Guinea climates vary from tropical ping from moss-draped branches had a pH of 3.9. 433 KRF•2 • OK.indd 433 15.12.2009 11:00:05 Karst Rock Features • Karren Sculpturing ††‡ †‡ † ‡ † ‰‡ † Š‡ ‰‡  ‚ƒ„­ € ­ Š‡     ƒ ‹‡  ††‡ †‡ †ˆ‡ † ‡ Figure 1: Location of Mount Kaijende in Papua New Guinea (from Wil iams, 1972). Solutional denudation rates have not been meas- Geological setting ured, but must be at the upper end of international estimates. Glacial pedestals on mountain Jaya in Karst in the southern fold belt of Papua New Guin- neighbouring Irian Jaya indicate surface lowering ea is developed mainly in Oligocene to Miocene rates on limestones of 32 mm/ka over the last 9.5 limestones and covers an area of about 20,000 km2 ka (Peterson, 1982). (Löffler, 1977). Around Mount Kaijende it is devel- On the lower slopes of the mountain at around oped in Lower Miocene limestone of more than 2,900 m a.s.l. arête and pinnacle karst progressive- 1,000 m stratigraphic thickness. Exposures in cliff ly gives way to polygonal karst. This is completely faces show it to be extremely massively bedded clothed in rainforest except for the bottoms of (order of tens of metres). Mount Kaijende is a tri- some of the larger depressions which are covered angular block faulted on its NW, SW and E sides, with coarse kunai grassland and tree-ferns, tem- and tilted gently to the SSE. Fault scarps vary in perature inversion and associated frost drainage height, some attaining almost 1,500 m. Thus local having excluded the forest. relief is considerable and the vadose zone is deep. The region was probably first uplifted and exposed 434 KRF•2 • OK.indd 434 15.12.2009 11:00:06 Paul W. Wil iams, Arête and pinnacle karst of Mount Kaijende Figure 2: Sketch of arête and pinnacle karst on Mount Kaijende (from Wil iams, 1972). to denudation in the early Pliocene, but the region isolate large blocks, the edges of which have been is still tectonically active. incised by huge sub-vertical solution gutters. The converging heads of these solution channels from different flanks of the blocks impart a sinuosity to Regional morphology the arête ridges. The bare pinnacle tops are prob- ably sharp and fluted with rillenkarren, but inves- The altitudinal zonation of karst forms in Papua tigation has not been close enough to determine New Guinea was discussed by Williams (1973). details. There are no measurements of the height Arête and pinnacle karst occurs between 2,600 of the pinnacles, but they are of the order of 100 m and 3,500 m a.s.l., but its occurrence appears to or more and the near vertical solution gutters are have more to do with structural-topographic several metres wide. Drainage down the gutters considerations than climate. Arête and pinnacle converges in deep shaft-like enclosed basins (Wil- terrain is known in three small localities at the liams, 2004). northern extremity of the limestone country in Although the arêtes are bare along their crests, the neighbourhood of Mount Kaijende, and oc- the limestone faces gradually obtain an increas- curs mostly within an area of 8‒10 km2 around ingly dense plant cover as the surfaces descend the summit. The arêtes are naked, reticulated, into the more sheltered environment of joint can- saw-topped ridges with spires, and with practical- yons, along which the bottoms of closed depres- ly vertical side slopes that have a local relief of up sions are aligned. Thus most depressions within to about 120 m. A good photograph of the area is the arête and pinnacle terrain are vegetated at available in Löffler (1977). The ridges are crude- their base. However, authoritative observations ly aligned, dominantly NNE–SSW, and are de- are unavailable, because the terrain is extremely termined by master joints striking at 25° crossed inaccessible (Figure 3) and so has never been pen- by other sets at 110° and 155°. These lineaments etrated very far and subjected to a field survey. 435 KRF•2 • OK.indd 435 15.12.2009 11:00:09 Karst Rock Features • Karren Sculpturing Figure 3: General aerial view from the south of arête and pinnacle karst on Mount Kaijende. Figure 4: Oblique aerial view of arête and pinnacle karst on Mount Kaijende. 436 KRF•2 • OK.indd 436 15.12.2009 11:00:12 Paul W. Wil iams, Arête and pinnacle karst of Mount Kaijende Figure 5: View of Mount Kaijende from the southeast with arête and pinnacle karst on the skyline. Discussion in Mulu National Park. On Gunung Api the pin- nacles are sharper than observed on Mount Kai- The landforms described are an extreme case of jende, but long arête ridges are less common. The karrenfield (Figure 4); extreme because of their gradual clothing with vegetation of the pinnacle height and steepness. This is made possible by sides as they descend into the forest is a feature of the combination of considerable local relief, the both areas (Figure 5). The tsingy of Madagascar strength of very massive and widely jointed lime- also have some similarity to the karren landforms stones, and the rapid dissolution engendered by of Mount Kaijende, but the lowland context is dif- the very humid environment. The tropical context ferent and the extent of the tsingy area is greater. has no special significance, except that occasional The shilin (stone forest) of Lunan in China also frosts are not severe enough to destroy the kar- has tall sharp spire features that resemble those ren and Pleistocene glaciers have not eroded the found in Madagascar, although the pinnacled area; however, glaciation has affected other karsts ridge relief of Mount Kaijende is much higher and in the region (Hope, 1976). Similar landforms are more dramatic. Ford et al. (1996) compare the fea- found in Sarawak in the pinnacles of Gunung Api tures of these three areas. 437 KRF•2 • OK.indd 437 15.12.2009 11:00:14 KRF•2 • OK.indd 438 15.12.2009 11:00:14 liThological characTerisTics, shape, 35 and rocK relief of The lunan sTone foresTs Martin KNEZ and Tadej SLABE The Lunan stone forests – shilin are a unique form et al., 1991; Sweeting, 1995). The carbonate rock of karst karren (Chen et al., 1998; Knez and Slabe, on which karren developed was covered by thick 2001a, b, 2002; Kogovšek et al., 1999). The kar- layers of sediment that decisively influenced the ren, which is criss-crossed by fissures along the occurrence and shape of the stone forest. Accord- fractures, is composed of rock pil ars (Song, 1986; ing to the thickness of the sediment, a stone forest Habič, 1980) or stone teeth (Song, 1986; Song and can be barren, covered, or buried. Hantoon (1997) Liu, 1992). Stone teeth are smaller protuberances describes the stone forests as an epikarst form. less than five metres tall; “high” teeth are taller Mangin (1997) believes that epikarst of the stone than three metres, and “low” teeth are less than forests extends to a depth of 100 metres. one metre tall (Song and Liu, 1992). According to The Lunan stone forests were formed predomi- their shape, they are divided (Song and Liu, 1992) nantly through the dissolving of rock below the into conical, angular, and oblong. The pillars are soil and sediments. The water increases the width between five and fifty metres high and are of vari- of the fissures and separates rocks. Under sedi- ous shapes. Large stone forests are a characteris- ment with acidic water, wide and deep cracks de- tic feature of subtropical climate conditions (Song, veloped between the pillars with deep channels 1986). on their walls (Yuan, 1997). Uncovered carbonate According to their location, Song (1986) dis- rock is transformed by rainwater. Teeth develop tinguishes three types of stone forests: valley, hill first and from them, the forests form (Song, 1986). top and hill slope. In lowlands and valleys, large Originally, the limestone, which was already in stone forests occur with intermediate dolines and the process of karstification (Yu and Yang, 1997; depressions. Underground waters flow beneath Song and Wang, 1997), was covered by Permian them, so they are periodically flooded or water basalt and tuff that influenced its shaping and flows through them. Stone forests on the tops of in places metamorphosed the rock (Song and Li, the hills are lower (10–30 metres), their pillars 1997; Ford et al., 1996). Water permeated through grow from a common foundation, and the cover the basalt and tuff, and underground karst began of sediment above them was thin. Stone forests on to develop. In the Mesozoic, part of the lime- hillslopes are an intermediate form between the stone was exposed (Song and Wang, 1997). In other two types. the Oligocene and Miocene, rock blocks rose and The Lunan stone forests are often described as lowered, and in lower parts the karst relief was a form of covered karst (Chen et al., 1986; Maire transformed by erosion (Yu and Yang, 1997). In 439 KRF•2 • OK.indd 439 15.12.2009 11:00:14 Karst Rock Features • Karren Sculpturing Figure 1: Major stone forest. the Eocene, the Lunan graben subsided, and thick removal of sediments from the surface, and the layers of lake sediments were deposited (Chen more rapid growth of the stone forests. et al., 1986; Zhang, Geng et al., 1997; Song and The Major stone forest spreads over 80 hec- Wang, 1997). In the tropical climate, thick layers tares, while the other larger and smaller stone of laterite soil developed on the sediment (Sweet- forests cover some 350 square kilometres (Chen ing, 1995; Ford et al., 1997). In the Pliocene, the et al., 1986; Zhang, Geng et al., 1997). The Major current stone forests began to develop (Yu and stone forest lies at 1,625–1,875 metres above sea Yang, 1997). In the Quaternary, a great part of the level and is located in a valley system. The un- sediment was removed but some remained in the derground water, which is just below the surface, fissures. rises ten metres following abundant precipitation. The level of the underground water played an Most (70–80%) of the annual precipitation of 936.5 important role in the development of the stone mm falls between June and October (Chen et al., forests (Ford et al., 1997). With the development 1986). The average temperature is 16.3°C, with a of underground water courses, pillars began to range between –2°C and 39°C. The pillars are tall- develop from teeth (Zhang, Geng et al., 1997). est in the central part of the valley system where The fluctuating underground water widened the the surface waters flow into the underground and fissures (Yuan, 1991). Below the forests today is a there is more sediment at the edges of the forest comprehensive and diverse system of water caves (Sweeting, 1995). Habič (1980) calls this “shallow (Zhang et al., 1997). Tectonic action resulted in karst.” In the lower part of the stone forest, waters lowering the level of the underground water, the also run on the surface. 440 KRF•2 • OK.indd 440 15.12.2009 11:00:16 Martin Knez and Tadej Slabe, Lithological characteristics, shape, and rock relief of the Lunan stone forests Figure 2: Naigu stone forest. The tourist areas of the Lunan stone forests are Selected examples of stone forests visited annually by more than a million people. This is a unique and integral natural and cultural Naigu stone forest landscape where the Sani minority lives, many of whom work in the tourist industry (Figure 1). The Naigu stone forest (Figure 2) lies twenty kil- The shape of the pillars in the stone forests and ometres east of the Major stone forest and is an their height are characteristic of certain types of important tourist site. This stone forest is com- rock and their topographical position (Zhang, posed of larger rock masses (Figure 3) and small- Geng et al., 1997). er pillars that stand together or individually. The Numerous examples of stone forests that have unique form of the forest is defined primarily by developed in almost identical conditions confirm the fracturedness and texture (Figure 4) of the that the diverse shapes of the pillars are primarily various beds of rock from which the stone forest the consequence of the distribution and density formed at different levels. The dimensions of the of the joints and fissures that cut the carbonate pillars are dictated by the joints and fissures that rock and the rock’s varying stratification and tex- vertically criss-cross the layers of rock. The shape ture. We must also add the importance of the effi- of the pillars, which are frequently undercut, and ciency of their formation by underground factors their rock relief clearly reflect the importance of and their reshaping by rainwater that determined subsoil formation and the reshaping by rainwater the course of their development in different pe- progressing slowly down the pillars. riods. The stone forest lies alongside two ridges slight- 441 KRF•2 • OK.indd 441 15.12.2009 11:00:17 Karst Rock Features • Karren Sculpturing Figure 3: Larger stone masses of Naigu stone forest. ly elevated by tectonics. The joints that border the and hollowing of the most porous part of the rock, joint zone are extremely strong, and the interme- whose surface disintegrates relatively quickly. The diate joints that largely run in a northwest-south- pillars whose tops are in porous and heavily dolo- east direction are several kilometres long and mitized beds are narrower and mostly without deep. The pillars formed on a package of Lower characteristic or regular shapes dictated by the Permian carbonate rock of Qixia formation more factors of their development. Stratified and non- than one hundred metres thick. The properties of porous limestone often forms wider bases of the rock throughout the geological cross-section are pillars composed of porous and heavily dolomi- very different, and from the morphogenic view- tized and massive and striped dolomitized lime- point we therefore divide the groups of layers stone. The shape of subsoil rock teeth as a rule into three groups from the bottom up: a) layered does not reflect the different texture of the rock. micrite and non-porous limestone, b) porous and The most distinctive rock forms are subsoil and heavily dolomitized limestone (Figure 5), and c) composed. Subsoil rock forms include large chan- massive and striped dolomitized limestone. nels on the walls of pillars and channels on the The pillars developed on different levels of the broader tops. Composed forms include the chan- described rock beds and their shapes correspond nels leading from the subsoil channels or subsoil to this rock variation. The most characteristic cups found on the tops. The deepening of subsoil shape of the pillars is mushroom-like, and there cups and water flowing along the channels caused are distinct notches along the porous and heav- the dissection of the tops of the pillars, particu- ily dolomitized beds (Figure 4). This notching is larly the larger ones, into points with funnel-like the consequence of faster underground corrosion notches between them (Figures 4, 5). 442 KRF•2 • OK.indd 442 15.12.2009 11:00:19 Martin Knez and Tadej Slabe, Lithological characteristics, shape, and rock relief of the Lunan stone forests Figure 4: Individual pil ars in Naigu. Subsoil rock forms, as a rule the larger ones, influenced by the texture of the rock. Subsoil developed on all types of rock in the Naigu stone rock forms developed on the majority of beds of forest. The rock influenced their shape, especially different rock, but only a few are found on po- that of the smallest, which on dolomitized rock rous and heavily dolomitized beds. Here we find often have jagged edges. Flutes hollowed by rain subsoil tubes. When these beds of rock are found drops are a less distinctive rock form in Naigu. at the tops of pillars, smaller rock forms hol- Their occurrence and development is primarily lowed by rainwater hardly occur. In places these 443 KRF•2 • OK.indd 443 15.12.2009 11:00:21 Karst Rock Features • Karren Sculpturing Figure 5: Dolomitized beds of Naigu stone forest. Width of view is 1.5 m, in the middle. are only rainpits or the rainwater shapes larger the development of the caves below them, is also subsoil rock forms. The rock relief therefore de- confirmed by the traces of the development in veloped relative to the position of the beds in the the Bayun cave in its central part. From the cave pillars. sediments and the rock relief we can distinguish The gradual and diverse development of the several periods of development in the epiphreatic stone forest, which of course is connected with part of the aquifer, then a rapid drop of the un- 444 KRF•2 • OK.indd 444 15.12.2009 11:00:23 Martin Knez and Tadej Slabe, Lithological characteristics, shape, and rock relief of the Lunan stone forests Figure 6: The top of Pu Chao Chun stone forest. derground water that probably caused the faster The rock changes little across the geological growth of the stone forest (Šebela et al., 2001). profile. Throughout, we trace mainly biomicrosp- aritic limestone with an almost hundred per cent proportion of CaCO , limestone that in this pro- 3 Pu Chao Chun stone forest file shows similar sedimentation conditions and that regardless of the thickness of the beds shows The Pu Chao Chun (Figure 6) is a smaller stone the same response to the influence of erosion and forest located fifteen kilometres south of the corrosion processes. The thickness of the beds has Major stone forest. Its rock pillars are located on a a decisive influence and clearly reflects the mor- ridge, where their network is the densest, and on phological appearance of the individual rock pil- the hillslope below. Its shape is defined primarily lars. by the unique distribution of variously thick beds In the upper part of the stone forest, the pillars of rock, mostly thin in the upper parts, on which are mostly individual and of smaller diameters, the stone forest developed at different levels. The and the rock beds are the thinnest here. The lower dimensions and oblong form of the pillars were parts of the pillars, which are formed on the thick- dictated by the joints and fissures that vertically er beds of the rock, are stouter and stand closer criss-cross the rock beds. The shape of the pillars together. Along the thinner beds (Figure 7) there and its subsoil rock forms clearly reflect their sub- are distinct notches. Here, the beds disintegrated soil formation and their transformation by rain- faster and the tops left beneath them are often flat, water slowly progressing down the pillars. while if the beds above them were thicker, the 445 KRF•2 • OK.indd 445 15.12.2009 11:00:24 Karst Rock Features • Karren Sculpturing Figure 7: Characteristical pil ars of Pu Chao Chun stone forest. tops are sharp. In the lower part of the stone for- by rainwater. The composed rock forms are chan- est, where there are fewer rock pillars, the pillars nels that lead from subsoil channels and cups at formed on thick beds of rock and as a rule there- the tops and hollows between the bedding planes fore are wider at the bottom and narrower toward of the rock. Funnel-shaped mouths formed on the the top. edges. Exposed subsoil rock forms were trans- All the types of subsoil rock forms that reveal formed by rainwater that hollowed flutes, chan- the evolution of the stone forest are well developed. nels, and rainpits, and on vertical and overhang- These include large subsoil channels and subsoil ing walls rain scal ops are present that occurred scal ops, as well as channels and subsoil cups on due to water trickling down the rough surface of the wider tops. The latter channels often devel- the rock. Solution pans most often occur on the oped from subsoil tubes along bedding planes and bottoms of exposed subsoil rock forms. were uncovered when the upper beds disintegrat- ed. Subsoil notches developed where long-lasting layers of soil surrounded the walls. A distinct pro- Lao Hei Gin stone forest portion of the rock relief consists of composed rock forms. These are divided into those that occurred The Lao Hei Gin stone forest (Figure 8) lies eighte- due to direct interaction of subsoil factors and en kilometres north of the Major stone forest. In- rainwater and those whose unique shape is the dividual and clustered rock pillars and larger consequence of transformation of subsoil forms blocks of rock transformed by corrosion and ero- 446 KRF•2 • OK.indd 446 15.12.2009 11:00:26 Martin Knez and Tadej Slabe, Lithological characteristics, shape, and rock relief of the Lunan stone forests Figure 8: Lao Hei Gin stone forest. sion occupy about two square kilometres. Where lars are roughly built of dolosparites and dolomi- the pillars are clustered, there are only cracks or crosparites of the grainstone type. fissures between them. The pillars developed on An important difference in individual packages various levels of almost horizontal rock beds, and of layers was in the determination of various de- their shapes correspond to this. The larger clusters grees of secondary porosity and recrystallization, of stone pillars are composed of several dozen pil- which are also reflected in the morphological ap- lars. On the relatively large area of the stone forest, pearance of the stone pillars. The lower parts of there are only individual pillars and rock teeth. the pillars are composed of dolosparites to dolom- Some pillars have the shape of square towers and icrosparites of the grainstone type in which sec- others of mushrooms. They are often composed of ondary porosity is barely noticeable. The central several blocks (Figure 9), the remains of rock beds part of the pillars is composed of very secondary- between bedding planes and fissures. The indi- porous dolomites. On average, the crystals of vidual pillars are either relatively large, wide, and dolomite are smaller than the crystals in the lower high or low (1–2 m) and wide. package of the layers and at the same time are In the area of the Lao Hei Gin stone forest, less pure. The upper parts of the pillars are again the original limestone is heavily diagenetically composed of secondary almost non-porous lime- altered: under the microscope, we can observe stone and dolomites. Only the tops are composed subhedral to euhedral grains of dolomite in the of recrystallized secondary non-porous limestone. rock that form hipidiotopic to idiotopic structures. Dolomite rock disintegrates mostly into grains. With the exception of the upper part, the rock pil- The strongly secondary porous central parts of 447 KRF•2 • OK.indd 447 15.12.2009 11:00:27 Karst Rock Features • Karren Sculpturing Figure 9: Mushroom-like pil ars in Lao Hei Gin stone forest. pillars below the ground as well as on the surface in the lower non-porous rock and protrude from weather and disintegrate faster. As a rule, tall pil- the ground remain. The heavily porous rock in lars therefore have a distinct mushroom shape the central part of the pillars is often hollowed by because the non-porous beds are more durable subsoil tubes that have been formed by rainwater and extensive. In places, the upper parts of pillars trickling down the pillars. The rare tops of pillars no longer exist, and only low pillars that formed formed on such rock have non-uniform shapes in most cases. The rock relief is composed of all the charac- teristic groups of subsoil rock forms, forms hol- lowed by rainwater, and composed rock forms. To a considerable extent, the texture of the various rock beds determines their features. The first complex of subsoil forms includes various subsoil channels that occurred due to the continuous flowing of water along the contact of the wall and the sediment that covered the rock and filled cracks along the vertical fissures. The diameter of the largest channels reaches several Figure 10: Rough surface of the dolomitized rock in Lao metres. They dissect all four different complexes Hei Gin stone forest. Width of view is 50 cm. of beds. On the tops of the tallest pillars, they were 448 KRF•2 • OK.indd 448 15.12.2009 11:00:30 Martin Knez and Tadej Slabe, Lithological characteristics, shape, and rock relief of the Lunan stone forests Figure 11: Chert nodules from Naigu stone forest. Height of view is 2.5 m. transformed by rainwater, while the porous mid- is, due to the permeation of water through the soil dle beds weathered too quickly for the channels and it’s flowing along the contact with the rock. to remain on them for a longer time. Thus the The larger channels on the upper parts of pillars channels are mostly a characteristic of lower pil- are composed rock forms. They occur due to the lars and rock teeth. Subsoil scallops, which occur water flowing from subsoil channels on the wide at relatively permeable contacts between the rock tops of pillars or lead from funnel-shaped notches. and sediments, are also preserved as a rule on the At the bottoms of the latter there are or used to be nonporous beds or on only recently uncovered subsoil cups. At the edges of the tops are therefore heavily porous beds. larger or smaller funnel-shaped mouths most fre- The wider tops of pillars and teeth are dissected quently reshaped by rainwater. They are especially by medium-sized and smaller subsoil channels distinct on the non-porous beds, or when the top and subsoil cups (Slabe, 1999) that developed is in a limestone rock, reaching to lower lying under the soil that partially covered the rock, that heavily porous beds. Their distribution and shape 449 KRF•2 • OK.indd 449 15.12.2009 11:00:32 Karst Rock Features • Karren Sculpturing Figure 12: Typical shapes of pil ars in Lunan stone forests (drawing by Tamara Korošec). 450 KRF•2 • OK.indd 450 15.12.2009 11:00:33 Martin Knez and Tadej Slabe, Lithological characteristics, shape, and rock relief of the Lunan stone forests ‒ relatively narrow and deep ‒ are determined by stone, which in individual horizons contains up the fracturedness of the rock and the indentation to several decimetres of thick quartz chert nod- of the rock circumference by the texture of the ules (Figure 11). The main lithological features of rock. the Maokou formation are roughly similar to the Half-bel s occur at the longer lasting levels of Qixia formation, the only difference being that in soil and sediments surrounding the pillars (Slabe, the Maokou carbonates we do not trace any major 1998, 1999). impact of late diagenetic dolomitization and in Rock forms hollowed by rainwater, especially some places there is considerable secondary po- smaller ones like flutes and rainpits, do not occur rosity. In both formations, we notice heavy diage- on the rock described above. The rock is coarsely netic variability of the foundation rock, which is rough, and only those rock forms whose size ex- undoubtedly the consequence of the intensive vol- ceeds individual elements of texture and structure canic (basalt lava) activity during the transition occur on it (Figure 10). The exception is the small- from the Paleozoic to the Mesozoic. The rock has er, highest-lying part of the stone forest where an exceptionally high percentage of carbonate. flutes occurred on the tops of limestone teeth. The In the area studied, we found considerable vari- dissection of most of the tops is therefore deter- ations in the thickness, porosity, degree and man- mined by the texture and fracturedness of the ner of dolomitization, containment of inclusions, rock. From subsoil cups distinctly dissected and and colour of individual layers reflected in the for- rough solution pans occur, and only the bottoms mation of the stone forests (Knez, 1998). of those covered by a thin layer of sediment and Morphological characteristics (Figure 12) are are overgrown remain flat and relatively smooth. the reflection of various factors, of which the most On the upper part of steep walls, there are rough important are geological factors. One of the basic features similar to channels, which in most cases factors is undoubtedly the fissuring of the rock, are very narrow and relatively deep, of angular which influences the shape of the forest and the shape; their diameters measure one to ten centi- dimensions of the stone pillars. The distribution metres and they are two to three metres long. of the pillars (ground-plan of the stone forests) matches the fracturedness of the rock. The pillars can be joined in rows between distinct joint sec- Lithological and morphological tions and stand closely side by side, or the stone characteristics and rock relief forests or their parts are composed of individual wide or narrow pillars. Pillars with smaller diam- The area of the stone forests is composed of Lower eters as a rule occur along dense networks of fis- Permian carbonates of the Qixia and Maokou for- sures, while larger rock masses with broader tops mations. These formations are two of the more occur along thinner networks. important basal formations on which numerous An especially important factor is the stratifica- stone forests developed in the southern Yunnan tion of rock, which affects the shape of the stone region of Lunan. Characteristics of the Qixia for- pillars. The beds have virtually no impact on pil- mation are micrite limestone with intercalations lars that developed on thicker beds and beds with of dolomite and dololimestone with intermediate an even rock texture. However, the vertical cross- sheets of schist. In the lower part of the Maokou sections of pillars on thin beds of rock are often formation, the limestone alternates with dolomite jagged because they are dissected by the notches and dolomitic limestone. In the upper part, we that occur along bedding planes, and the unequal trace the sequence of limestone, which is thinly resistance of different beds of rock is reflected in bedded in places and elsewhere composed of sev- their external shape. eral metres thick beds, as well as massive lime- Enumerating geological factors, we should not 451 KRF•2 • OK.indd 451 15.12.2009 11:00:33 Karst Rock Features • Karren Sculpturing Figure 13: Stone teeth. forget the texture of the rock. The rock texture, es- bells and subsoil channels and cups on broader pecially if it is diverse, can decisively influence the tops, while composed rock forms include the shaping of stone pillars, as much the shape of their channels leading from subsoil channels or solu- vertical cross-sections as the size of the cross-sec- tion cups and they dissect the walls of the pillars. tions. Porous beds are often hollowed and disin- Many pillars are subsoil undercut, and their tops tegrate faster, while beds of rock with less soluble are transformed by secondary subsoil rock forms components most often protrude from the walls. and shapes hollowed by rainwater. A unique rock For better understanding both the regional and relief is found on the larger rock pillars, especially local development of stone pillars, we must also those that have wide tops, either on thick beds of point out the influence of subsoil processes that rock where secondary subsoil rock forms occur or fostered the development of pointed tops of sub- on the tops that developed due to the disintegra- soil teeth (Figure 13) and the undercut shape of tion of thin rock beds, where the subsoil tubes that the pillars. Rainwater sharpens the tops of pillars occurred along bedding planes are transformed and transforms the traces of their original subsoil into subsoil forms or large channels reshaped by formation. The unique development of the stone rainwater. Both features also indirectly influence forests is reflected in their rock relief. Rainwater the shaping of the pillar sides due to the flowing gradually transforms subsoil rock relief. Subsoil of the water and the hollowing of channels. As a and composed rock forms, especially the largest, rule, smaller rock forms do not occur on dolomite are the most distinctive. Subsoil rock forms in- rock, on very porous rock, or on rock with large clude scallops, large channels, notches, and half- inclusions. 452 KRF•2 • OK.indd 452 15.12.2009 11:00:35 TWO IMPORTANT EVOLUTION MODELS 36 OF LUNAN SHILIN KARST Linhua SONG and Fuyuan LIANG The Shilin National Park, with a total area of 350 The karst features were formed by subsoil dissolu- km2, is located in Lunan, 80 km east of Kunming, tion. On hilltops, the limestone blocks have been the capital of Yunnan Province. The shilin land- corroded by rainwater dissolution along fractures scape is a special type of pinnacle karst developed and other openings, with a result that is similar in thick, gently dipping, and mostly pure Lower to the tsingy karst landforms of Madagascar. On Permian carbonate rocks in a humid, tropical cli- slopes, most rainwater runs off and washes the mate. It includes subjacent karst, subsoil karst and soil and insoluble materials onto lower land such subaerial karst. The main shilin karst landscape as depressions and basins. On the slopes, isolated consists of stone columns or pil ars, and stone teeth, stone columns or pillars and stone teeth with sub- with extensive subsoil solution features such as soil solution features are displayed. In the depres- through caves, wall niches, and solution grooves, sions and basins, the soil thickness reaches several as well as subaerial solution forms like different metres to tens of metres. The subsoil solution that types of karren, vertical solution columns, depres- formed the stone columns was highly irregular, sions, and pits. The most typical joint-oriented fea- and the distance between stone columns varies tures, both subsoil and subaerial, are vertical wells from 10 to 50 m. The lower parts of the columns about 10 m deep. are buried by fluvial deposits. Field studies show two evolutionary types for the shilin landscape. In the Shilin National Park the area was covered by basalts in the Late Per- Development of the Shilin Park mian, and by conglomerates and mudstones in the Tertiary period. When the basalt covered the The shilin landscape is the special type of karst Lower Permian limestone, the basalt fractured landscape in the Lunan Area, Yunnan. The name as it cooled and solidified, and intense weather- shilin is from the Chinese, shi = stone or rock, ing took place. Rainwater was able to penetrate and lin = forest. The term applies only to stone through the basalt into the limestone and cause pillars that stand above the ground at least 5 m intense dissolution of the limestone. This process like a stone forest (Yuan, 1982; Song, 1986). Three produced the high and magnificent shilin land- thousand years ago, Qiu Yuan searched for the scape beneath the basalt cover. Another type of Shilin, but nobody could tell him the location shilin landscape developed in the carbonate area (Wang et al., 1994). Four hundred years ago, Xiu that has not been covered by non-carbonate rocks. Xiake, the great geographer of the Ming Dynasty, 453 KRF•2 • OK.indd 453 15.12.2009 11:00:35 Karst Rock Features • Karren Sculpturing traveled through 11 provinces from the east to the Williams, Derek Ford, Timothy Atkinson and southwest of China for about 50 years to visit karst Anthony Waltham, also confirmed that the shilin landforms, caves, springs and minority societies. karst is mainly caused by subsoil solution (Song, He summed up his great learning in his “Travels 1986). of Xiu Xiake”. But there is no specific description How the shilin karst formed by subsoil solution, of the shilin landscape in his diary, although he has been systematically studied since 1999. The wrote about similar features. In 1931, Lu Yun, the Institute of Geography, Chinese Academy of Sci- president of Yunnan Province, visited Shilin, and ences, and administration of Shilin National Park used the term “shilin” to describe the stone for- coordinated in establishing the Shilin Research est landscape. The Shilin Park was approved and Center and Shilin Research Foundation to sup- directly managed by the Yunnan government. To port the scientific study of the karst evolution, ec- develop the wild karst park into a tourist site, the osystem, and environment. This paper describes government developed a brief plan and allocat- some of the studies of how subsoil dissolution has ed money to build trails, a small visitors’ centre, produced the shilin scenery. and a sightseeing pavilion overlooking the Shilin landscape. Since then, the Shilin Karst Park has become well known in China. The first scientif- Field studies ic study of the shilin landforms was conducted by Ma (1936). Since then, with the development During field trips since 1984, two contrasting of tourism in China, many scientists have stud- types of landscape have been distinguished in the ied their origin and evolution (Qing, 1977; Song, Lunan shilin karst area: 1986; Zhang, 1984; Waltham, 1984; Ford, 1997; Lin, 1997; Song and Li, 1997). In the 20th century, people recognized that Type 1: Naigu shilin and Puduocun the shilin was the result of rainwater dissolution (Qing, 1977). That point of view predominated Naigu shilin is located in the northern Shilin Na- throughout China. In 1979, in cooperation be- tional Park. The Lower Permian Qixia limestone tween the Chinese Academy of Sciences and the and the thick Maokou limestone, both with sili- Yugoslavian Academy of Arts and Sciences, Dr. ceous soil and residuum, extend throughout the Peter Habič and Dr. Rado Gospodarič visited area. The Maokou limestone block is elevated China. We carefully observed the characteristics about 50 m above the ground surface and is cross- of the shilin karst and discovered that subsoil so- cut by NE and NW joint sets. Karren and rocky lutional features such as through holes solutional dolines controlled by joint intersections are well conduits, spongework, and benches are typical. developed on the tops of limestone blocks. We learned that subsoil dissolution is the main On the slopes of limestone hills, isolated pil- process by which the shilin landscape forms. In lars about 10‒15 m high, as well as stone teeth, 1982, when Dr. Marjorie Sweeting visited Shilin, are well developed. Subsoil solution features in- we studied the karren on the tops of stone teeth clude through holes conduits, solutional furrows and pillars, the gullies and flutes on their sides, and spongework. Red soil up to 0.5 m thick covers the contact between the limestone teeth and the the limestone, and it fills the limestone fissures to surrounding red mud, and the subjacent karst depths of more than 1 metre. (Chen et al., 1986). In 1984, Prof. Shouyue Zhang Depressions are floored by red and brown soil published a paper that described the character- with a thickness of 2 m or more. A few isolated istics of subsoil dissolution that forms the shilin stone pillars about 7‒10 m high stand on the de- landscape. During visits in 1983 and 1984, Paul pression floors (Figure 1). Some depressions are 454 KRF•2 • OK.indd 454 15.12.2009 11:00:35 Linhua Song and Fuyuan Liang, Two important evolution models of Lunan shilin karst floored only by red soil with no protrud- a ing columns or teeth. Type 2: Songmaoshan karst and Major shilin karst About 1.5 km south of the Major shilin, basalt tuff constitutes the upper part of Songmaoshan hill. On the southeastern hillslope, intense soil erosion has removed the brown red weathered soil, loose tuff and debris. It is clear that the basalt and tuff once covered the limestone and filled the limestone fissures. Stone teeth 3‒5 m high and small pillars 5‒7 m high stand b above the fractured basalt tuff. Solution holes, pits, and karren ril s have devel- oped on the upper parts of the stone teeth, whereas the columns are distinguished by smooth surfaces. The lower parts re- main as truncated features surrounded by coarse black debris and rhombic cal- cite crystals 50 mm in diameter in small solutional holes. The limestone has been metamorphosed to marble. About 30 m eastward from the metamorphosed stone teeth and columns, limestone pillars 10‒15 m high and stone teeth are very well devel- oped in depressions. In the opposite direc- c tion, limestone fissures also retain basalt. It should be mentioned that the Major shilin was also developed in relation to the basalt. The basalt is distributed on the top and southern slope of Shilin Hotel hill, which is the Major shilin on the south side. The magnificent Major shilin, 10‒40 m high, is characterized by blocks of limestone separated by narrow vertical fissures. Pinnacles are well developed on the tops of the columns, whereas vertical ril s are common on the sides. Solution pockets, benches, conduits, and niches are developed in the lower parts of the col- Figure 1: Evolution of the shilin karst landscape from the hil s to umns. Landform development in the Shi- the depressions: a. on hill tops; b. on slopes; c. in depressions. 455 KRF•2 • OK.indd 455 15.12.2009 11:00:36 Karst Rock Features • Karren Sculpturing Figure 2: Sketch of karst geomorphological profile in Naigu Shilin Park. lin Hotel zone is very similar to that in the Song- limestone contact. In thin soil, the CO value is 2 maoshan. low (e.g. A1, at 1,000 ppm). If the nearby soil is thick, the CO content will be affected by the CO 2 2 content of the thick soil (e.g. A2, with 2,200 ppm). Two models of shilin karst Table 1 also shows that the maximum CO content development 2 occurs at depths of 40‒60 cm below the surface, with the highest value reaching 17,820 ppm. The Developmental model of the shilin effect of vegetation on CO content decreases in 2 landscapes the following order: irrigated lawns, cypress, rare grassland, pine, bush and cultivation, and land The development of Type 1 shilin karst is summa- without vegetation. rized in Figure 2. The limestone block on the high Experimental results with standard limestone plateau is assumed to be cut by joints and other tablets with diameters of 50 mm and thickness of fractures and to have suffered mechanical and 5 mm, made from the Maokou limestone in Shi- chemical weathering. Soluble materials such as lin Park, show that they lost 0.04 mg·cm-2·100d-1 Ca+2, Mg+2 and HCO ‒ are carried away with the in soil 20 cm thick by dissolution. The dissolution 3 water on and/or in the limestone, while residual rates in the soil on slopes was 0.58, 0.16, and 0.04 materials are transported by water to lower sites mg·cm-2·100d-1 at 20, 60, and 100 cm depths in the such as dolines and depressions. soil, respectively. However, the average dissolu- The CO content in soil air varies with the tion rates in wet soil reached 4.46 mg·cm-2·100d-1 2 depth of soil and the type of vegetation (Figure 3). at depths of 20‒80 cm. The CO in irrigated loam is much higher than in The shilin landscape on karst hilltops is guided 2 other soil and vegetative covers. The lowest CO is by the spacing of fractures, where soil and re- 2 in soil with no vegetal cover ‒ only 2,400 ppm at siduum are very thin. But the soil and residuum the depth of 120 cm. may also enhance the limestone dissolution along Soil thickness also strongly affects the soil CO fractures, even though it may be slow. The soil on 2 (Table 1). CO was measured down to the soil/ the slope varies from 0 to 0.5 m on the limestone, 2 456 KRF•2 • OK.indd 456 15.12.2009 11:00:37 Linhua Song and Fuyuan Liang, Two important evolution models of Lunan shilin karst Table 1: Soil air CO2 between stone teeth near the Stone Screen spot (Liang et al., 2000). No. Soil CO (ppm) at different depths (cm) 2 10 20 30 40 50 60 70 80 100 A1 1,000 A2 2,200 B1 3,820 10,180 11,200 18,840 B2 13,340 9,420 14,260 12,730 B3 3,820 C1 6,370 16,240 17,820 8,150 7,640 C2 6,110 12,220 15,280 15,780 5,600 9,670 Table 2: Properties and contents of microbes and CO2 in weathered basalt (D) and limestone soils. No. Soil depth (cm) Soil temperature pH Humidity Microbes Soil CO (ppm) 2 (0C) (%) (x104/g dry soil) D1–6–1 –20 10.4 6.71 33.092 20.8 18,000 D1–5–2 –40 10.8 6.87 33.636 12.6 12,000 D1–4–1 –60 11.4 7.33 32.593 28.5 20,000 D1–3–1 –80 11.9 7.51 34.084 24.4 14,000 D1–2–1 –100 12.3 6.85 40.805 7.55 12,000 D1–1–1 –120 15.5 6.95 46.697 20.2 20,000 S1–5–1 –20 9.6 7.06 40.813 20.6 3,400 S1–4–1 –40 10.3 7.24 40.746 15.3 4,500 S1–3–1 –60 10.6 7.37 45.614 18.3 3,000 S1–2–1 –80 11.1 7.48 46.355 6.56 4,500 S1–1–1 –100 11.4 7.68 43.888 8.70 4,000 Table 3: The subsoil dissolution rate (mg·cm-2·y-1) on limestone tablets in shilin. Soil depth (cm) D1 D2 S1 X1 B1 E1 T1 –20 5.9495 2.117 2.117 15.111 0.146 1.1315 6.0225 –40 5.6575 1.898 0.5475 20.367 –60 6.424 1.679 0.584 13.870 –80 5.2925 0.219 15.768 –100 9.6725 0.146 –120 10.1105 Mean rate 7.1832 1.898 0.7227 16.279 0.146 1.1315 6.0225 and beneath it the limestone can corrode at a rate case, the pillars become smaller, evolve into stone of 0.16‒0.58 mg·cm-2·100d-1. The isolated stone pil- teeth, and finally disappear completely. lars with the solution pockets, benches, conduits and niches may appear on the slope. After heavy rain, the slope flow carries detrital materials into Subjacent development model of the shilin depressions. Many depressions form temporary landscape lakes after storms. Their sediments cover the un- derlying karst features. The wet soil will strongly After the development of palaeokarst in the early dissolve the stone teeth and pillars at the rate of Permian, basaltic lava of Late Permian age cov- 4.46 mg·cm-2·100d-1, after the experiments. In this ered the karst features and filled fissures. Karst 457 KRF•2 • OK.indd 457 15.12.2009 11:00:37 Karst Rock Features • Karr C en S O culpturing 2 concentration (ppm) 0 4000 8000 12000 16000 features that developed beneath the non-carbon- 0 ate rocks have been defined as subjacent karst (Chen et al., 1986). Measurements show that the water penetrat- ing into basalts has a low pH value of about 7. Mi- –40 crobes reproduce vigorously in the weathered ma- terials (Liang et al., 2003). Table 2 shows that the th (cm) microbes and CO concentrations in weathered 2 basaltic soil are much richer than in the limestone il depso soil. –80 The high CO content in the weathered basalt 2 soil air causes strong corrosion of the underlying limestone (Song and Liang, 2001). Table 3 demon- strates that the subsoil dissolution rate in weath- –120 ered basalt is 4‒6 times of the rates on soil-covered limestone or in soil-filled limestone fissures. It is grassland lawn no vegetation cypress pine clear the karst development in limestone under the basalt is stronger than in soil-covered lime- Figure 3: CO2 concentration in the atmosphere of soil stone. covered by various types of vegetation. Figure 4: Subjacent karst evolution model for the limestone covered by basalt in Songmaoshan region. The upper diagram shows the status of the shilin landscape in relation to the basalt. 1. shows the basalt cover on the lime- stone; 2. subjacent karst developed beneath the basalt; 3. basalt weathered away and the shilin landscape devel- oped. 458 KRF•2 • OK.indd 458 15.12.2009 11:00:40 Linhua Song and Fuyuan Liang, Two important evolution models of Lunan shilin karst The shilin development in the basalt area is The water in the deep weathered materials in de- sumarized by Figure 4. The upper part shows the pressions dissolves the limestone tablets at a mean relations between the basalt and the palaeokarst rate of 1.968 mg·cm-2·100d-1, with a range of 1.45 features (stone teeth and pillars). The lower three to 2.77 mg·cm-2 ·100d-1. In this case, the fissures in diagrams illustrate the evolution of karst features limestone were enlarged and began to partition covered by the basalt. the limestone beneath the basalt. The initial phase After the basalt lava flows covered the lime- of development of the stone teeth and pillars were stone, fractures, especially vertical ones, de- as shown in Figure 4. veloped during the cooling process. The basalt With this manner of subjacent karst develop- became a fractured aquifer. During and after ment, the shilin landscape rapidly develops along rainfall, water penetrates through the fractured fractures. As the basalt on the proto-shilin is erod- basalt aquifer and into the underlying limestone ed away, the proto-shilin landscape is exposed. along fractures. The water absorbs CO from the Vegetation grows on the basalt soil and the strong 2 weathered material, so that it can rapidly dissolve subsoil solution greatly increases the rate of de- the underlying limestone. The water retains its velopment of the shilin landscape. Subsoil solution solutional capacity for limestone, as the basalt is features are developed on the buried parts of the relatively insoluble. The limestone tablet experi- shilin landforms, and subaerial solution features ments show that water in loose weathered mate- such as rillenkarren, solution pans, and pinnacles rial on the rims of basalt depressions can dissolve are developed on the exposed shilin. This evolu- limestone at the rate of 0.46‒0.58 mg·cm-2·100d-1. tionary stage is shown in Figure 4, diagram 3. 459 KRF•2 • OK.indd 459 15.12.2009 11:00:40 KRF•2 • OK.indd 460 15.12.2009 11:00:40 soluTion raTes of liMesTone TableTs 37 and cliMaTic condiTions in Japan Kazuko URUSHIBARA-YOSHINO, Naruhiko KASHIMA, Hiroyuki ENOMOTO, Takehiko HAIKAWA, Masahiko HIGA, Zenshin TAMASHIRO, Tokumatsu SUNAGAWA and Eisyo OOSHIRO Many karstologists have been interested in karst by Urushibara-Yoshino (2003). In this paper, the terrain, which is the surface which has been results of 10 years from 1992 to 2002 will be dis- formed over a long period through the dissolu- cussed. tion of the rock by physical-chemical process- es. However, this feature per se cannot inform us of the time scale of the dissolution process. So Methods and study areas many karstologists have tried to clarify the rates at which this process of limestone dissolution Limestone areas occupy only 1,764 km2 (less than had occurred, and so identify it as the main fac- 0.05% of the land) in Japan (Urushibara-Yoshino, tor causing karstification. Measurements of solu- 1996). The limestone areas are distributed from tion rates using limestone tablets were made by north to south and developed under different cli- Trudgill (1975) and Jennings (1977). matic conditions. Seven limestone areas were se- Since 1980 the denudation commission of the lected for study (from 1992‒1997), from Hokkaido International Speleological Union has measured in northern Japan to Okinawa southwest Japan. solution rates globally using the Slovenian Creta- They were in Toma, Abukuma, Chichibu, Aki- ceous limestone (Gams, 1985). Similar limestone yosidai, Shikoku, Ryugado and Minamidaito. tablets made of Guilin Permian limestone were From 1998 to 2002, the observations continued at measured in China (Yuan, 1991). The solution all limestone sites except Chichibu. The locations rates of Slovenian Cretaceous limestone in Japan of these sites are shown in Figure 1 and their envi- were measured from 1989 to 1990 by Urushibara- ronmental conditions in Table 1. Yoshino (1991). In addition the solution rates of The tablets used at these sites measured 40 mm Slovenian Cretaceous limestone, Guilin Permi- in diameter with a thickness of 4 mm. Before plac- an limestone and Japanese limestone have been ing the tablets, their surface was ground to 300 measured since 1992. The results of these observa- meshes and weighted after dry up in 7 days in silli- tions during the three years 1993, 1994 and 1995 cagel. The weights of tablets ranged from 12.5‒13.5 were published (Urushibara-Yoshino et al., 1998) g. Every 10 years, the tablets were cleaned, dried as they were during the five years from 1993 to and weighted. The weight loss was measured to be 1997 by Urushibara-Yoshino et al. (1999b). The so- in the range of 10-5 g. At all observation sites, the lution rates of limestone tablets from 1992 to 2001 upper surfaces of the tablets were corroded more in Minamidaito island is discussed in an article than the bottom surfaces. From the air, rain water 461 KRF•2 • OK.indd 461 15.12.2009 11:00:40 Karst Rock Features • Karren Sculpturing     ‚ €­ €  €  €  €‚ ­    ƒ  ƒ  ƒ ­ ƒ€‚ ƒ€  ƒ€ ƒ€ ƒ€­ ƒ‚ ƒ  Figure 1: Limestone areas and location of observation sites in Japan. drops onto the upper surface, and water in the stone dissolution between the upper and lower sur- soil also percolates through to the upper surface. face, the decrease in thickness was not measured. Because of this, there is a differential rate of lime- Only the loss of weight was measured. At each site 462 KRF•2 • OK.indd 462 15.12.2009 11:00:41 K. Urushibara-Yoshino, N. Kashima, H. Enomoto, T. Haikawa, M. Higa, Z. Tamashiro, T. Sunagawa, E. Ooshiro: Solution rates of limestone tablets … Figure 2: The ob- ured because their solution rates are extremely servation site at high in soils. In this study, the CO in A and B 2 3 2 Abukuma in the horizons, where the tablets were set, was meas- air (1.5 m). ured for at least 4 seasons at each locality over a period of 10 years. The following three methods were used: namely, the Non Dispersive Inbrared Gas Analyzer, Gastec (Urushibara-Yoshino et al., 1998) and the Dräger method (Miotke, 1974). For an understanding of the limestone solution processes, the water balance in each year from 1993 to 2002 was calculated using the method of Thorn- thwaite (1948). This method was chosen because it considers the soil moisture in the water balance. Results Solution rates of limestone tablets The original weight of the tablets was in the range of 12.5‒13.5 g. The mean annual solution rate of the tablets were made from Slovenian Cretaceous the four tablets was used for the discussion of limestone, Guilin Permian limestone, Chichibu differences between location, and year-to-year Triassic limestone and the local limestone. These fluctuation. The correlation matrix for 5 years four tablets, made from different lithologies, were (1993‒1997) between the solution rates of lime- set in the air 1.5 m above the ground (Figure 2) stone and the climatic factors of annual precipi- and in the A and B horizons of local soils. The tation, water surplus (WS), water surplus – water 3 2 depth of the buried tablets at each site was differ- deficit (WS–WD) and precipitation – evapotran- ent. This was because of the different thickness of spiration (P–E) was examined (Urushibara- soil horizons found at each site. Yoshino et al., 1999). In the air, the correlation The CO concentrations in the soils were meas- coefficient is high between the solution rate and 2 Table 1: Location and environment of observation sites of limestone areas. ASAHIKAWA SHIKOKU (ONOGA- ABUKUMA AKIYOSHI RYUGADO MINAMI-DAITO (TOMA) HARA) 43°49'30" N 37°20' 30" N 31°14'50" N 33°28' 24" N 33° 54'54" N 25°30'7" N Location 142°37'30" E 140°40'40" E 131°17'40" E 132°52'51" E 133°44'53" E 131°14' 1" E Carboniferous Quaternary Geology Permian limestone Cretaceous marble and Permian Permian limestone Triassic limestone limestone limestone needle and evergreen de- subtropical ever- deciduous forest deciduous forest evergreen Vegetation deciduous ciduous forest and green tree and and grassland and grassland forest grassland forest grassland grassland Annual mean temperature (°C) 7.0 10.5 13.7 9.1 14.9 23.3 Annual precipitation (mm) 1118.2 1216.2 1913.1 2262.9 2291.7 1727.2 lO years mean water surplus (mm) 558 552 1192 1645 1504 615 463 KRF•2 • OK.indd 463 15.12.2009 11:00:42 Karst Rock Features • Karren Sculpturing            ­€  ­€ € €­ € €­      Figure 3: Trends of solution rates at 1.5 m above the Figure 4: Trends of solution rates at 1.5 m above the ground from 1993 to 1997. ground from 1998 to 2002. WS–WD. In A horizon and in B horizon in the 3 2 soil, the correlation coefficients are highest be- tween solution rate and annual precipitation. This means that the degree of dry condition (WS – WD) affects the difference of the solution rate in the air. However, the annual precipitation (total wet condition) of the area effects to the solution rates in the soils. The same correlation matrix for the 5 years (1998‒2002) has also been examined. The   correlation coefficient was highest with the same      ­ factor as the case with the results for 1993‒1997.   €‚ƒ ƒ ƒ­‚ The solution rates of limestone tablets were plot- € ted in Figure 3 from 1993 to 1997 and Figure 4 from 1998 to 2002. The solution rates of the tab- lets in the air showed the highest correlation co- efficient between WS – WD during the two peri- ods. And the solution rates at the observation sites Figure 5: Trends of solution rates at 1.5 m above the show strong year-to-year fluctuations. In Figure ground at each observation site. Extreme cases of dry 3, the tendency curves of the solution rates dur- year and wet year are shown. ing the five years are shown by several lines. Fur- 464 KRF•2 • OK.indd 464 15.12.2009 11:00:45 K. Urushibara-Yoshino, N. Kashima, H. Enomoto, T. Haikawa, M. Higa, Z. Tamashiro, T. Sunagawa, E. Ooshiro: Solution rates of limestone tablets …      ­€­€ €‚ƒ­€„€ ƒ ƒ­€   †­‡ ‡ ‡‚†„  € €„€ Figure 6: The CO measurement in Minamidaito island   2 (upper Dräger method, lower Gastec method).   ­  €‚€‚  ­‚ƒ„€‚ ‚ „ „€‚­­ †‡€ˆ thermore, the driest year 1994 and the wettest year ˆ ˆƒ ‡   ‚†‚ ‚ 1993 are shown by heavy lines from north to south during the five years (1993‒1997). In Figure 4, the same tendency curves at the observation sites and the wettest years line (1999), as well as the driest year (2002), are also shown by heavy lines during 5 years (1998‒2002). Only the Asahikawa (Toma) values are too small to fit into the tendency curve for those years. During these 10 years, the driest examples are 2002 and 1994 and the wettest ex- amples are in the years 1999 and 1993. These tendency curves are combined in Figure 5. In Figure 5 the tendency curves of solution rates of the driest year 2002 and wettest year 1993 in the 10 years period are shown. In 2002 WS – WD are extremely low, this was a dry year, especially in southern part of Japan, therefore the solution rates show the smallest values. In 1993, WS – WD are high in the air, this was an extremely wet year, and Figure 7: The CO in B horizon in August and in February 2 2 therefore the solution rates were high at every site. from 1998–2002. 465 KRF•2 • OK.indd 465 15.12.2009 11:00:48 Karst Rock Features • Karren Sculpturing Figure 8: The tablets inserted in A and B horizons of soil 3 2 in Minamidaito island. CO in soils tween the soils particles after heavy rain (over 20 2 mm). However, this gas cannot escape into the air In this part of the study, carbon dioxide in A and or into the underground water. This is because 3 B horizons was measured at 7 sites in 1993‒1997 the pressure is higher in B horizon than in the 2 2 and at 6 sites in 1998‒2002. At all sites, the con- A horizon. Carbon dioxide concentrations of 3 centrations of carbon dioxide values were higher the B horizon in summer (August) and in win- 2 in the B horizons than in the A horizons. The ter (February) are shown in the article by Urush- 2 3 reason why in the B horizon, the concentrations ibara-Yoshino et al. (1999b). The carbon dioxide, 2 of carbon dioxide were higher than in the A ho- being measured by the Dräger and Gastec method 3 rizon, is that the density of soils is higher in the in Minamidaito, is shown in Figure 6. Carbon di- B horizon than in the A horizon. At all the sites oxide concentrations of the B horizon in summer 2 3 2 where measurements were taken, a high concen- (August) and in winter (February) are shown for tration of carbon dioxide occurs in the spaces be- the 5 years (1998‒2002) in Figure 7. The high CO 2 466 KRF•2 • OK.indd 466 15.12.2009 11:00:48 K. Urushibara-Yoshino, N. Kashima, H. Enomoto, T. Haikawa, M. Higa, Z. Tamashiro, T. Sunagawa, E. Ooshiro: Solution rates of limestone tablets … concentration in the B horizon during the warm tion rates in the air 1.5 m above the ground shows 2 periods of the year can support higher solution the smallest range with WS–WD in 1994 and rates of limestone tablets in soils. However, solu- 2002, but the largest range with WS–WD during tion activity in winter is extremely low, because of the 10 years in 1993 and 1999. the low CO in the B horizon. From these observations, the sites in Shikoku 2 2 Figure 8 shows the four tablets in the A and in and Akiyoshidai, in the monsoonal temperate 3 B horizons. The solution rates of limestone tablets zone, have the best water balance conditions for 2 were 3 to 5 times higher in B horizon than in the solution of limestone in the air. Because of these 2 air in Minamidaito Island (Urushibara-Yoshino et good water balance conditions, the karstification al, 2003). One of the reasons why in B horizon high of the limestone plateau with a dense pattern of 2 solution rates could be obtained, might be high dolines developed to a greater extent in these re- concentration of carbon dioxide in B horizon. gions of Japan. Of course, for the development of 2 dolines, beddings planes, the thickness of layers and the fissure pattern of limestone also assist Conclusion karstification. In the soil, the solution rates are 3 to 5 times Over the 10 years period, the correlation coeffi- higher than in the air. One reason might be high cients between the solution rates and WS–WD are concentration of carbon dioxide in the soil. This highest in the air. On the other hand, correlation means that the surface covered soils will be de- coefficients between the solution rate and the an- nuded much faster than limestone outcrop areas nual precipitation are highest in the A and B ho- in Japan. 3 2 rizons in the 10 years observed. Over the 10 years period, solution rate of lime- stone at the 6 sites showed annual fluctuations. The Acknowledgement solution rates in the air 1.5 m above the ground show a clear trend with WS–WD. At each obser- This study was supported by the Ministry of Edu- vation site, the solution rate increased in accord- cation, Science and Culture in Japan (No 0768090 ance with WS–WD, over the range 1,000‒1,600 1995‒1997). We would like to express our sincere mm. Solution rates then decreased above 1,600 thanks to Prof. I. Gams and Prof. D. Yuan who mm in WS–WD. supplied the tablets of Slovenian limestone, and At the observation sites, the trend of the solu- Guillin limestones, respectively. 467 KRF•2 • OK.indd 467 15.12.2009 11:00:49 KRF•2 • OK.indd 468 15.12.2009 11:00:49 The rocK ciTies of rosso aMMoniTico 38 in The VeneTian prealps Ugo SAURO In the Venetian Prealps there are characteristic on contiguous thicker rock units. The Rosso Am- rocky landscapes developed in a thin formation of monitico formation plays a very important role in micritic limestone, dissected by a system of wide- the control of the subterranean hydrology and in ly spaced fractures. These structurally controlled the development of both dolines and lithological forms produce rock cities or giant limestone pave- contact caves. The rocky landscapes of the Rosso ments (Figures 1, 2, 3) which contrast strongly Ammonitico with their associated features (Fig- with the soil-covered, smooth surfaces developed Figure 1: A typical rock city developed in the upper Rosso Ammonitico unit (Val Marisa, Monti Lessini). Note the sys- tem of corridors isolating large blocks. 469 KRF•2 • OK.indd 469 15.12.2009 11:00:50 Karst Rock Features • Karren Sculpturing Figure 2: The upper and lower units of the Rosso Ammonitico exert a strong structural con- trol on slope evolu- tion (Costeggioli, Monti Lessini). The “corridor karst” is better devel- oped in the upper unit. The large blocks to the right have been involved during the last cold phases of Pleistocene by solifluction. Figure 3: Detail of the Valle delle Sfingi (La Busa di Camposilvano, Monti Lessini). The relict blocks developed in the lower Rosso Ammonitico are widely spaced and simi- lar to mushrooms. ure 4) are the consequence of both karst and per- Sette Comuni (also called the Asiago plateau), is iglacial processes. that which developed on the limestone formation called “Rosso Ammonitico” (ammonitic red). On this limestone are structurally controlled rocky A structurally controlled rocky landscapes of the “rock city” or “giant limestone landscape pavement” types, contrasting with the soil-cov- ered, smooth surfaces developed on contiguous A scenic landscape in the Venetian Prealps, espe- rock units (Figures 1, 2) (Corrà and Benetti, 1966; cially of the Monti Lessini and the Altopiano dei Sauro, 1973c, 1977; Perna and Sauro, 1978, 1981). 470 KRF•2 • OK.indd 470 15.12.2009 11:00:54 Ugo Sauro, The rock cities of Rosso Ammonitico in the Venetian Prealps Figure 4: The Ponte di Veja (Val del a Marchio- ra, Monti Lessini), devel- oped in the lower Rosso Ammonitico, is the rem- nant of the roof of a large cave, now partial y col- lapsed. The Rosso Ammonitico formation, which is tabular summits, structural terraces and rocky about 30 m in thickness and mostly made up of benches on the slopes, bedding slopes and wide pink, nodular, micritic limestones rich in iron ox- valley bottoms. Due to the lower resistance to ero- ides, was deposited during the Middle and Upper sion of the intermediate beds, in the slopes where Jurassic in an oxidizing marine environment, at horizontal or infacing strata are exposed the a slow sedimentation rate (about 1 m per million Rosso Ammonitico extends in two superimposed years). The Rosso Ammonitico formation is very lines of benches corresponding to the lower and massive in its lower part, well stratified, cherty upper parts of the sequence (Sauro, 1973c). and more clayey in its intermediate part and The vertical succession of rocks with different again relatively massive in its upper part. The rock characteristics influences the evolution of karst formation is crossed by systems of widely spaced landforms and caves (Sauro, 1973c, 1974). Under- fractures (with meshes ranging between a few me- ground water flowing diffusely in the dense net- tres and some tens of metres). work of discontinuities of Maiolica encounters Above the Rosso Ammonitico is the Maiolica a barrier with only a few main fractures when it (also called Biancone), a marly limestone similar comes into contact with the Rosso Ammonitico. to chalk, about 150 m in thickness, closely strati- The focused flow through the Rosso Ammonitico fied and densely fractured. Below is the Calcari results in the development of discrete caves. In del Gruppo di San Vigilio (San Vigilio Group specific morpho-structural settings following limestone), a pure, diffusely fractured limestone, the lowering of the topographical surface, funnel up to 200 m in thickness. shaped dolines develop in the lower Maiolica just The Rosso Ammonitico is much more resist- above the points of focused drainage within the ant to denudation than the adjacent formations underlying Rosso Ammonitico. Collapse dolines and outcrops over relatively large areas mostly in the lower Rosso Ammonitico are caused by the as structural and sub-structural surfaces: small breakdown of cave roofs. 471 KRF•2 • OK.indd 471 15.12.2009 11:00:56 Karst Rock Features • Karren Sculpturing Figure 5: Karren developed on bedding surfaces in the upper Rosso Ammonitico unit (Monte Grol a, Monti Lessini). Both rounded karren and runnels are present. Evolution of the rocky landscape of Ammonitico takes place within the epikarst, by the Rosso Ammonitico underground dissolution, before the rock is ex- posed by denudation. Recently outcropping sur- The rocky landscape of the Rosso Ammonitico is faces are characterized by bedding planes inter- the consequence both of the much greater resist- sected by networks of grikes and corridors, which ance of this rock in comparison with the confin- have greatly varying widths, between a few centi- ing rock units (Maiolica and Calcari del Gruppo metres and several metres and are partially filled di San Vigilio) and of the opening and widening with soil and sediment. of the main fractures which evolved as dissolution The upper surfaces of the isolated blocks present grikes and corridors. These dissolution features different situations due to the lithological control isolate large blocks with tabular bedding plane and the character of the local slope. Some surfac- upper surfaces, delimited laterally by the rims of es are made up of bare rock (Figure 5), others are nearly vertical rock faces. These rocky landscapes characterized by a weathered layer of rock where may be considered special types of rock cities the nodular structure of the limestone is evident where the large clint blocks are delimited by a geo- and a thin soil cover has developed. Many of the metrical pattern of corridors corresponding to the nodules isolated by dissolution and biokarstic network of fractures. processes are ammonite moulds. On the bare The early widening of the fractures in the Rosso surfaces it is possible to find rundkarren, rinnen- 472 KRF•2 • OK.indd 472 15.12.2009 11:00:58 Ugo Sauro, The rock cities of Rosso Ammonitico in the Venetian Prealps Figure 6: Detail of the moved parallelepipedal blocks (Costeggioli, Monti Lessini). karren, kamenitzas, and other karren features. fluction (Figures 2, 6). These isles evolved as a pe- Ril enkarren are relatively rare, probably because culiar type of rock glacier following the filling of of the nodular structure of the rock. the grikes and other discontinuities caused by the The evolution of the rocky landscapes of the ice (Sauro, 2002). These forms no longer appear to Rosso Ammonitico is the consequence not only of be active. dissolution but also of periglacial processes such To conclude, the rocky landscapes of the Rosso as cryoclastic weathering and solifluction. These Ammonitico with their associated features may processes operated very effectively during the be classified as rock cities or giant limestone pave- cold phases of the Pleistocene. On bedding slopes ments, which are strongly controlled by the geo- of 10°‒15° (Sauro, 1976a), both isolated blocks and logical structure and evolved mainly by karst and “isles” of the Rosso Ammonitico moved by soli- periglacial processes. 473 KRF•2 • OK.indd 473 15.12.2009 11:01:00 KRF•2 • OK.indd 474 15.12.2009 11:01:00 coasTal eogeneTic Karren of 39 san salVador island John E. MYLROIE and Joan R. MYLROIE Analysis of eogenetic coastal karren on Holocene Geographic and geologic setting and late Pleistocene eolianites from San Salvador island, Bahamas, indicates that the karren devel- The Bahama islands are a 1,400 km long por- op quickly, and similarly, on all outcrops studied. tion of a NW–SE trending archipelago that ex- The Holocene eolianites provide a tight time win- tends from Little Bahama bank east of the coast dow to overall karren development, and major of Florida to Great Inagua island, just north of storm-erosion features indicate that karren are the coast of Cuba (Figure 1). The archipelago ex- routinely removed and re-initiated. Comparison tends farther southeast as the Turks and Caicos of lagoonal versus ocean-exposed outcrops indi- islands, and Mouchoir, Silver, and Navidad banks cate that higher wave energies create a faster recy- that are a separate political entity. The northeast- cle time for the ocean-exposed karren. Both the ern Bahama islands are isolated landmasses that Holocene and late Pleistocene eolianites are eo- project above sea level from two large carbonate genetic in nature, and their overall rock youth is platforms, Little Bahama bank and Great Bahama the single most important aspect of coastal karren bank. To the southeast, beginning in the area of development, over-riding both the age and ocean San Salvador island (Figure 2), the Bahamas com- exposure differences among the study sites. The prise small isolated platforms, many of which are coastal karren are expressed as pitting and sur- capped by islands that make up a significant por- face irregularity produced in the sea-spray zone; tion of the available platform area. The Bahami- rocks in coastal settings but protected from sea an platforms have been sites of carbonate depo- spray are smoothed by storm events and show lit- sition since at least the Cretaceous, with a mini- tle meteoric karren development. The presence of mum thickness of 5.4 km (Meyerhoff and Hatten, coastal sea-spray produced karren, and the paucity 1974) and perhaps as much as 10 km (Uchupi et of meteoric karren indicate both the rapidity with al., 1971). The large platforms to the northwest are which coastal sea-spray karren can develop, and dissected by deep channels and troughs, and the also the cycle time of major storm events, which isolated platforms of the southeastern Bahamas must be faster than meteoric karren production. are surrounded by deep water (Melim and Masa- 475 KRF•2 • OK.indd 475 15.12.2009 11:01:00 Karst Rock Features • Karren Sculpturing   ‰Š    ‰ŠŠŠ   Š ‹ŠŠ ‰ŠŠ  ‚    Œ ˆ†     ˆ ‡ ˆ   ˆˆ  † „  ­ €      ­ ‚ „   ƒ       Figure 1: A map of the Bahamas, showing the location of San Salvador island. ferro, 1997). Water depth on the platforms is gen- ca. 125,000 years ago) are represented solely by eo- erally less than 10 metres. There is no evidence for lianites (Carew and Mylroie, 1995a, 1997). These active tectonics in the Bahamas (Carew and Myl- oldest eolianites form the Owl᾽s Hole formation roie, 1995a). (Figure 3) and are made of biopelsparites on San Salvador. The Owl᾽s Hole formation contains eoli- anites of at least two sea level highstands prior to Stratigraphy the last interglacial (Panuska et al., 1999), as has been noted elsewhere in the Bahamas (Kindler The exposed carbonates of San Salvador island and Hearty, 1995). consist of late Quaternary limestones that were Overlying the Owl’s Hole formation, and sepa- deposited during glacio-eustatic highstands of sea rated from it by a palaeosol or other erosion sur- level. Each highstand event produced transgres- face are deposits of the last interglacial (Marine sive-phase, stillstand-phase, and regressive-phase Isotope substage 5e), the Grotto Beach formation units. Because of slow platform subsidence, Pleis- (Figure 3). The Grotto Beach formation contains a tocene carbonates deposited on highstands prior complete sequence of subtidal, intertidal, and eo- to the last interglacial (Marine Isotope substage 5e, lian carbonates, consisting of two members. The 476 KRF•2 • OK.indd 476 15.12.2009 11:01:02 John E. Mylroie and Joan R. Mylroie, Coastal eogenetic karren of San Salvador island      ­€ ­„ ‚ƒ   ­   † †    ‚  ‚ ƒ ‡ ‰ Š ˆ  Figure 2: A map of San Salvador island, showing major features and location of the sample sites (at the north and south ends of the island, in large type). 477 KRF•2 • OK.indd 477 15.12.2009 11:01:03 Karst Rock Features • Karren Sculpturing tirely of eolianites, whose foreset beds can com- monly be followed at least 2 m below modern sea level. Whole rock carbon-14 measurements from the North Point member indicate particle ages centred around 5,000 years BP. Laterally adjacent, but rarely in an overlying position, is the younger Hanna Bay member. This unit consists of intertid- al facies and eolianites deposited in equilibrium with modern sea level. The eolianite grains have radiocarbon ages that range from approximately 3,300 to 400 years BP (Carew and Mylroie, 1987; Boardman et al., 1989). While weakly-developed ooids have been reported from the early stages of North Point member deposition (Carney and Boardman, 1991), the Rice Bay formation is pre- dominantly peloidal and bioclastic on San Salva- dor. The North Point member is currently being attacked by wave erosion. Sea caves, inland cliff-  line talus, and coral-encrusted wave-cut benches  of the North Point member exist. The relationship   of the deposits to sea level, and the carbon-14 ages, indicate that the North Point member was depos- Figure 3: Stratigraphic column for surface rocks on San ited as a transgressive unit after sea level rose at Salvador island (from Panuska et al., 1999). the end of the last glaciation, but that deposition was complete prior to sea level reaching its cur- rent elevation. Sea level must have been about 10 French Bay member is a transgressive eolianite. In m or less below current sea level, as the platform some places, transgressive eolianites are marked had to be partially flooded to allow allochem pro- by an erosional platform on which later still-stand duction that later became the source material for fossil corals are found (Carew and Mylroie, 1995b, the eolianites. But sea level must have been below 1997; Halley et al., 1991). The Cockburn Town modern position, as evidenced by the foreset beds member is a complex arrangement of still-stand of the North Point member eolianites dipping ~2 subtidal and intertidal facies overlain by regres- m below modern sea level. Given the tectonic sta- sive eolianites. During Grotto Beach time ooids bility of the Bahamas, the time and sea-level posi- were produced in great numbers, and the vast ma- tion window for the North Point member rocks jority of eolianites in the Grotto Beach formation is tightly constrained. By the time of the Hanna are either oolitic (up to 80‒90% ooids) or contain Bay eolian deposition, beginning ~3,000 years ago, appreciable ooids. sea level had reached and stabilized at its current Overlying the Grotto Beach formation, and position. separated from it by a palaeosol or other erosion The three formations are separated from one surface is the Rice Bay formation (Figure 3) that another by well-developed palaeosols and other has been deposited during the Holocene. The Rice erosion surfaces that formed during sea-level low- Bay formation is divided into two members, based stands; however, as previously noted, the Owl᾽s on their depositional history relative to Holocene Hole formation contains at least one palaeosol sea level. The North Point member consists en- within its section representing a glacioeustatic 478 KRF•2 • OK.indd 478 15.12.2009 11:01:04 John E. Mylroie and Joan R. Mylroie, Coastal eogenetic karren of San Salvador island a c b d Figure 4: a. overview of the Grahams Harbour side of North Point. Note the “dark zone” at the shore and the “light zone” farther inland on the low, wide bench; b. diagrammatic representation of the view in a, showing transect lo- cation; c. overview of the Atlantic side of North Point. The steep topography makes the “dark zone” and “light zone” bands less obvious than in a; d. diagrammatic representation of the view in c, showing transect location. 479 KRF•2 • OK.indd 479 15.12.2009 11:01:08 Karst Rock Features • Karren Sculpturing sea-level lowstand (Panuska et al., 1999). The tropical limestone coasts, include champignon stratigraphic column shown in Figure 3 contains surface (Stoddard et al., 1971), phytokarst (Folk magnetozones that allow palaeosols of different et al., 1973), lacework morphology (Bull and Lav- ages to be identified based on the signal of secu- erty, 1982), biokarst (Viles, 1988) and coastal kar- lar variation contained in their palaeomagnetic ren (Mylroie and Carew, 1995). Considered one record (the rocks are too young to display magnet- of the most variable and least understood karren ic reversals). Despite the subdivisions made above, types (White, 1988), its appearance is thought to the carbonates of the Bahamas are remarkably be largely influenced by endolithic and epilithic uniform in texture and petrography, especially organisms (Jones, 1989). This was first reported by when compared to ancient rocks found in conti- Folk et al. (1973), who named the features “phy- nental settings. For a further review of Bahamian tokarst” based on the assumption that boring by geology and stratigraphy see Carew and Mylroie filamentous algae is responsible for observed mor- (1995b, 1997), and Melim and Masaferro (1997), phology. However, the impact of organisms on and the references therein. topography has never been unequivocally demon- strated (Viles, 2001) and this karren is most likely polygenetic, affected by a variety of concurrently Eogenetic coastal karren of San operating processes in addition to biological cor- Salvador island, Bahamas rosion. In coastal areas, these mechanisms can in- clude abrasion by marine invertebrates (Schneider Eogenetic karren and Torunski, 1983; Trudgill, 1987), wave action, wetting and drying, salt weathering and hydra- The karren of carbonate islands and coasts are dis- tion (Ford and Williams, 1989), and salt spray tinct from karren of inland settings of continents and rain water mixing. This great variety of proc- or large islands. Reef, lagoonal and eolian lime- esses is superimposed on the highly heterogene- stones that form most young carbonate islands ous texture and high primary porosity of young are eogenetic, meaning they have not undergone limestones, which may be the crucial factors con- significant diagenesis and still exhibit high pri- trolling the development of this type of karren by mary porosity and extreme heterogeneity (Myl- making the rocks predisposed to the development roie and Vacher, 1999; Vacher and Mylroie, 2002). of multitudes of irregular pits and hol ows. These lithologic qualities appear to favour the de- Taboroši et al. (2004) report that while the velopment of highly irregular and composite kar- impact of endolithic microorganisms on rock ren, and preclude the development of many other morphology, for example, may be considerable types, particularly the hydrodynamically-shaped at small scales, the large scale pinnacles and pits forms (Taboroši et al., 2004). These coastal karren are more likely a result of differential erosion due contain deep pits, extremely sharp points, knife- to metre-scale heterogeneities, such as variations edge ridges, and completely penetrating holes. in mineralogy and cementation (Trudgill, 1976b) Despite being the characteristic karren type and structural weaknesses (Viles and Spencer, on young eogenetic limestones, there is not yet 1986). Thus, the term “phytokarst” is inappropri- a unique and accurate geomorphic term for this ate not simply because the organisms originally type of sculpturing, both the centimetre-scale thought to be responsible are not plants (Jones, karren, and metre-scale pinnacles, found on 1989), but also because a variety of other proc- youthful, or eogenetic, carbonate coasts. Taboroši esses are involved, and their relative contributions et al. (2004), have reviewed the situation, as fol- vary by location and scale. The terms “eogenetic lows. The terms applied to the most intensely cor- karren” for this type of small-scale dissolutional roded variant of this karren, often reported from morphology, and “eogenetic karrenfeld” for result- 480 KRF•2 • OK.indd 480 15.12.2009 11:01:08 John E. Mylroie and Joan R. Mylroie, Coastal eogenetic karren of San Salvador island ant metre-scale topography has been proposed by a Taboroši et al. (2004). We believe that these terms proposed by Taboroši et al. (2004) are most ap- propriate because they emphasize the eogenetic nature of host limestone as the common factor controlling the development of all karren forms of this type, while avoiding references to genetic mechanisms (which are too numerous, variable, and poorly known) as well as settings (which are not necessarily island nor coastal). Placing “coast- al” in front of the general term “eogenetic kar- ren” provides a location indicator that does not predispose a genetic mechanism for the karren development. Thus, this paper will discuss coastal b eogenetic karren of Holocene coastal rocks on San Salvador island. Coastal eogenetic karren: the Holocene of San Salvador island San Salvador island has the most complete pub- lished geologic record of any of the Bahamian is- lands, and perhaps of any carbonate island in the world, as a result of the Gerace Research Centre (formally the Bahamian Field Station) on the is- land, which has hosted scientific enquiry for over 30 years. The geologic map of San Salvador (Carew and Mylroie, 1995b) shows that Holocene rocks, primarily eolianites, are found in numer- ous coastal locations. Two areas of the Holocene Figure 5: a. overview of the West of Gulf area; b. diagram- North Point member of the Rice Bay formation matic representation of the view in a, showing transect (Figure 3) were examined at the type section of location. North Point (Figures 2, 4). An additional area of late Pleistocene-aged Cockburn Town member of the Grotto Beach formation (Figure 3) at the south end of the island, called West of Gulf for its loca- The southern coast location, West of Gulf, being tion just west of a landmark called The Gulf, was in an ocean-facing, but late Pleistocene aged out- studied as a control for the work (Figures 2, 5). At crop (Figure 5a, b), was selected to determine if all three locations the bedrock is eolian in origin. any age-related differences might be discernable. The North Point locations are divided into an Examination of the three sample outcrops ocean-facing outcrop, the Atlantic side (Figure demonstrates that they contain a suite of etched 4c, d) and a lagoon-facing outcrop, Grahams Har- and pitted surfaces (Figure 6) that are blackened bor (Figure 4a, b), to help differentiate wave and by endolithic algal growth (“dark zone” of Figures spray magnitude issues in a uniform lithology. 4b, d, 5b), with a sharp boundary with a relatively 481 KRF•2 • OK.indd 481 15.12.2009 11:01:10 Karst Rock Features • Karren Sculpturing Table 1: Summary numerical data from the transect measurements. Small pits Location width SD width confidence width mean depth SD depth confidence depth mean Atlantic side North Point 0.865 0.159 1.35 0.590 0.108 0.68 Graham's Harbour 1.057 0.198 1.71 0.902 0.169 0.81 West of Gulf 0.690 0.116 1.17 0.451 0.076 0.65 Large pits width SD width confidence width mean depth SD depth confidence depth mean Atlantic side North Point 11.039 3.122 17.39 6.412 0.814 9.68 Graham's Harbour 9.575 4.845 26.46 4.104 2.077 8.36 West of Gulf 13.564 4.203 24.55 4.702 1.457 6.20 All measurements are in cm. smooth, grey limestone inland of the pitted sur- 7). The mean values for the small pit widths and face (“light zone” of Figures 4b, d and 5b). The depths from the protected Grahams Harbour lo- boundary between the light and dark coloured cation on North Point were wider and deeper than zones is very abrupt. This abrupt boundary, while for the less protected Atlantic side at North Point, sometimes coincident with local relief, commonly and at West of Gulf. is not. Field observations indicate that the bound- As shown in Figure 4, the Atlantic side of North ary coincides with the margin of sea-spray wetting Point is steep and rugged in its gross morphology. during normal wave conditions. The Bahamas are The Grahams Harbour side of North Point has a in a low-amplitude tidal environment, with tides broad erosional bench separating the coast from a ranging ~1 m. As a result, sea spray is scattered cliff of erosionally-truncated eolianites. The West over a very similar range during the tidal cycle. of Gulf site (Figure 5) is in a basin-like feature Small pits are the dominant karren feature in the flanked by higher rocky outcrops. The Grahams dark zone. Harbour bench indicates that while normal waves Measurements of the pitted surface indicate do not cross it, major storm and hurricane waves that the pits fall into two arbitrary categories, must do so, as the bench has been carved since small (0.5‒6 cm width) and large (6‒65 cm width). the stabilization of sea level at its current position At each outcrop location, a transect made up of ~3,000 years ago. Both the Atlantic side at North grid boxes was placed perpendicular to the shore- Point, and West of Gulf at the south end of the line, from the shoreline inland into the grey or island, have more rugged coasts reflecting their light limestone area (Figures 4, 5, 6d). Circles of exposure to wave energies routinely higher than 15 cm diameter were dropped into each grid box that found in the sheltered Grahams Harbour site. at random and the small pits in each circle meas- ured in terms of width and depth, and catalogued. All large pits in each grid box were measured in Conclusions terms of width and depth, and catalogued. The re- sults from these measurements are shown in Table The first major conclusion that can be reached is 1 and in Figure 7. None of the width-to-depth dis- that the pitted surface found at North Point is tributions have any statistical correlation (r2 < 0.4), Holocene in origin, as it must be younger than although the large pits at West of Gulf had a width the Holocene rocks on which that surface is de- to depth r2 = 0.52. The width to depth ratio for al- veloped. The similarity of pitting seen at the late most all pits, however, is less than one, indicating Pleistocene section at West of Gulf indicates that that they are wider than they are deep (Figure those pits are also Holocene in age, and have not 482 KRF•2 • OK.indd 482 15.12.2009 11:01:11 John E. Mylroie and Joan R. Mylroie, Coastal eogenetic karren of San Salvador island a b c d Figure 6: a. closer view of scene in Figure 4a, note smooth “light zone” to the right, and pitted, “dark zone” to the left; b. close up view of the pitted surface found in the “dark zone” areas on San Salvador island, top to bottom linea- tions are bedding of the eolianite; c. rock hammer sits on surface that had covering rock peeled off by a storm in the winter of 2003. The peeled surface is white and unpitted, but as this surface is in the spray zone, it will soon become dark and pitted; d. transect being measured on the Graham Harbour site. Note that it extends from the dark zone into the light zone. been inherited from an earlier time (such as the ocean-exposed sites, may indicate that the recycle last interglacial sea-level highstand). The age dif- time is longer in the protected lagoon, allowing ference in the rocks at North Point and West of the pitting to progress more completely before a Gulf is expressed most obviously in the more ir- storm strong enough to agitate the lagoon occurs. regular large-scale topography at West of Gulf, The occasional major storm event that has which most likely reflects differential cementation carved the Grahams Harbour bench must at the and diagenesis in those rocks since their deposi- same time strip the pitted surface. After the storm tion 120,000 years ago. At the small scale, however, passes, sea spray begins to re-initiate pitting on the rocks at all sites appear to behave similarly, re- those surfaces within spray reach, while the wave- flecting their overall eogenetic nature. The larger polished surface farther inland, away from sea depth and width of the small pits at the protected spray, remains unaltered. Given that the eolianite Grahams Harbour site, relative to the other two high ground has been eroded inland at least 10 m 483 KRF•2 • OK.indd 483 15.12.2009 11:01:16 Karst Rock Features • Karren Sculpturing  ­     Figure 7: Plots of width versus depth data for pits measured in the transects at the three sites: abowe. small pits; below. large pits. 484 KRF•2 • OK.indd 484 15.12.2009 11:01:17 John E. Mylroie and Joan R. Mylroie, Coastal eogenetic karren of San Salvador island in the last 3,000 years, the outcrop must be rou- rapidity with which coastal sea-spray karren can tinely, in geologic time terms, swept and eroded develop, and also the cycle time of major storm by storm waves. Therefore, the pitting seen on events, which must be faster than meteoric karren the coastal “dark zone” must develop relatively production. quickly. Most studies reported in the literature of young coastal carbonate outcrops deal with rocks that Acknowledgments are Pleistocene in age. The work presented here from San Salvador, using Holocene rocks, pro- The authors wish to thank the Gerace Research vides time constraints not available in those other Centre for assistance with the field research, con- localities. While age differences of a hundred ducted under GRC Research Number G-102. Field thousand years manifest themselves on the large assistance by the Teachers in Geoscience Bahami- scale, they do not do so at the small scale. Rather, an Field Class of 2003 (Christene Anderson, Tracy exposure to wave energies seems to be more im- Brown, Neal Becker, Deborah Buffingham, Melis- portant. The data further reinforce the idea that it sa Kellerman, Patrick Leonard, Lisa Keith-Lucas, is the eogenetic aspect of the rocks that is critical. Rebecca Murray, Debra Neuhaus-Palmer, Chris- The data also indicate that these coastal eogenetic tine Oxenford, Thomas Rozycki, Tricia Schafe- karren forms develop quickly, and rapidly renew book, Patti Jean Simpson and Rebecca Vowell) is their micro-morphology as major erosion events greatly appreciated. Data reduction by Lica Ersek, reshape the overall coastal landscape. The pre- Monty Keel, Nono Lascu and Monica Roth was sence of coastal sea-spray produced karren, and extremely valuable. Discussions with Larry Davis the paucity of meteoric karren indicate both the and Jim Carew were helpful. 485 KRF•2 • OK.indd 485 15.12.2009 11:01:18 KRF•2 • OK.indd 486 15.12.2009 11:01:18 coasTal Karren in The balearic islands 40 Lluís GÓMEZ-PUJOL and Joan J. FORNÓS Coastal karren, understood as an assemblage of karren features, the approach to the study of this small (down to millimetres) to large-scale (up to topic has changed significantly. Earlier workers several metres) mainly dissolutional features de- focused their effort on morphological descrip- veloped on carbonate coasts, is a topic with a large tions and spatial zonations across the coast pro- but not with an abundant tradition in geomor- files (Emery, 1946; Corbel, 1952; Guilcher, 1953; phological literature. Since Wentworth (1939), Dalongeville, 1977; Mazzanti and Parea, 1979), one of the earliest papers concerned with coastal whereas modern ones put their effort into try- ‚ƒ „ † ‡     ­     € ˆ‰ Figure 1: Location inset and geological map of Balearic islands. Dotted line indicates coastline with conspicuous coastal karren features. CG. Cala d’en Guixar; SA. S’Alavern; PS. Punta des Savinar; CM. Cala Murada; PF. Punta des Faralló; FC. Far de Ciutadel a; CA. Cap d’Artrutx; CT. Cala Turqueta; CF. Cap d’en Font; BN. Binibèquer; CS. Cala Sant Esteve. 487 KRF•2 • OK.indd 487 15.12.2009 11:01:19 Karst Rock Features • Karren Sculpturing ing to identify and understand which processes Study area and agents operate on carbonate coasts (Folk et al., 1973; Schneider, 1976; Trudgill, 1976a, 1987; Mallorca and Menorca are the two biggest islands Viles et al., 2000; Lundberg and Lauritzen, 2002; of the Balearic archipelago, which is located at Moses, 2003). the centre of the Western Mediterranean (Fig- The earliest references to the coastal karren of ure 1). They have a typical mediterranean climate the Balearic islands appear in a nineteen-century with hot dry summers and mild wet winters. The naturalist’s works (Habsburg-Lorena, 1884‒1891), mean annual temperature is approximately 17°C, although the first description from a scientific with mean winter and summer values of 10 and point of view is developed by Walter-Levy et al. 25°C respectively; the mean annual precipitation (1958). Butzer (1962), Butzer and Cuerda (1962) is about 500 mm and is mostly concentrated in and Ginés (2000) consider briefly Mallorcan autumn (Guijarro, 1986). The Western Mediterra- coastal karren forms in their Quaternary stratig- nean presents a temperate, oligotrophic, clear sea raphy research, but the first paper exclusively con- environment. Waves rarely exceed 8 m in height cerned with these forms is published by Rosselló and 50 m in wavelength; these values are consid- (1979) who attempts to assess the type of forms erably reduced near the shore where a maximum and their spatial organization in the south-east- height of 4 m is achieved only during 6‒8 Beau- ern Mallorca coast. More specific studies linking fort scale gales (Butzer, 1962). Forcing by tides coastal karren forms and bioerosion are devel- is almost negligible in the Mediterranean with a oped by Kelletat (1980, 1985) in the north-eastern spring tidal range of less than 0.25 m, although coast of Mallorca. Moses and Smith (1994) char- changes in atmospheric pressure and wind stress acterize the spatial domain of inorganic proc- can account for a considerable portion of sea level esses – salt weathering and solution – operating fluctuations (Basterretxea et al., 2004). on Mallorca’s southern coasts, by means of SEM These islands are the eastern emergent parts of and XRD analysis. Since 1998 there has been an the Balearic promontory; a thickened continental increase in coastal karren knowledge related to crustal unit forming the NE continuation of the investigations on limestone rock coast evolution Alpine Betic thrust and fold belt (Alonso-Zarza in the Balearic islands. Recent works character- et al., 2002; Gelabert et al., 1992) built during the ize these features from an integrated approach Middle Miocene. The major topographic heights linking forms and morphological zonation to of both islands are horsts formed during post- different processes, agents and their erosion Middle Miocene extension and expose deformed rates (Fornós and Gómez-Pujol, 2002). The most Palaeozoic to Middle Miocene carbonate rocks. detailed inventory of coastal karren forms and The intervening grabens are flat areas, filled by their organization is by Gómez-Pujol and Fornós an Upper Miocene carbonate shelf and Quater- (2001) who link forms to hydrodynamic and bio- nary alluvial fan and aeolianite deposits (Gelab- logical gradients as well as to major features of ert, 1998, 2003). Cliff coasts are characteristic of a cliff profiles. Recently, Gómez-Pujol and Fornós large part of the Mallorcan and Menorcan littoral. (2004b) have assessed Menorcan coastal karren They are almost exclusively associated with deeper through morphometrical, SEM and mineralogi- water offshore, and the –20 m isobath is generally cal studies. found at distances considerably less than 500 m The aim of this paper is to review and summa- from the shoreline. Cliff form is closely related to rize the characteristic assemblages, organization the main characteristics of the large-scale mor- and processes involved in coastal karren develop- phostructural units of each island. Abrupt cliff ment in the Balearic islands. faces are characteristic of horsts, and grabens host the major beach-barrier systems. Tabular 488 KRF•2 • OK.indd 488 15.12.2009 11:01:19 Lluís Gómez-Pujol and Joan J. Fornós, Coastal karren in the Balearic islands     ­€‚ ƒ „ „  ­ ­     „ „     † ‡ ‡ Figure 2: Balearic islands coastal karren selected cross-section profiles showing main morphological features, hydro- dynamic gradient and biological zonation. See locations in Figure 1. 489 KRF•2 • OK.indd 489 15.12.2009 11:01:21 Karst Rock Features • Karren Sculpturing relief added to the major horsts and grabens are trol the geometry of each karren feature. SEM bounded by Upper Miocene-Quaternary listric observations according to Viles (1987) and Tay- faults (Gelabert, 1998), which result in medium lor and Viles (2000) have been developed at each to low vertical sea cliffs. Thus the general picture coast profile. is one of plunging and composite cliffs that af- fect from Palaeozoic to Middle Miocene folded outcrops. Cliff faces vary locally from 3 to 30 m Classification of coastal karren forms in height and extend from 5 to 10 m below sea level. In these folded outcrops shore platforms The classification adopted here is based on mor- and rock coasts sculptured by coastal karren ap- phologies with some subdivision based on genetic pear patchily and are closely related to lithology factors. Although generic classification of karren and structure control. On the other hand, in post- forms is to be preferred, the genesis of many forms orogenic Upper Miocene outcrops, cliffs present a is not completely understood. Much of the vari- composite profile with a step-like form closely re- ety in karren occurs because two or more differ- lated to higher Pleistocene sea levels (Butzer, 1962). ing processes combine to produce a polygenic form These steps are enhanced by the geometry of the (Ford and Williams, 1989). This is especially true Upper Miocene tabular strata and differences in for coastal karren where different processes, such geomechanical properties between depositional as salt weathering, inorganic solution or bioero- levels (Pomar and Ward, 1999). Where cliffs fall sion, contribute to the final shape of forms like ba- vertically, their height range is 3 to 30 m. Shore sins or pinnacles. platforms, although, patchily distributed, are here more continuous than in the folded rock coasts. Coastal karren forms are more common and con- Negative forms tinous in the Upper Miocene outcrops. Circular plan forms Micropits: A wide variety of small depressions and Method differential etching forms commonly less than 1.0 cm in characteristic dimension. Micropit walls The coastal karren was surveyed from low water are smooth and are partly or entirely covered by mark at the shore platforms up to the terrestri- cyanophytes, fungi or lichens. They may be iso- al transition zone using a tachometer TOPCOM® lated, in coalescence, or aligned following a mi- CTS210. On each profile the extent of hydrody- crofracture. Micropits cover other major forms namic and biological zonation (Figure 2) was such as pinnacles, and also are present on bare marked according to colouration and key-species and vertical rock surfaces. These depressions do presence (Torunski, 1979; Schneider, 1976). The not present any kind of preferred orientation, and surveys documented the size and shape of kar- gravitational control is not dominant. For this rea- ren features as well as comments focusing on any son, although a recognized solution microtopog- biological action. The morphometrical approach raphy, it is believed that biological agents also play designed by Johansson et al. (2001) has been de- an important role on micropit development (Folk veloped just for basin pools. It consists of taking et al., 1973; Danin et al., 1982). Most species of cy- the major morphometrical parameters – length, anophytes, fungi and lichens are surface dwell- width, and depth ‒ and classifying the shape ac- ers (epiliths), but in stressful environments – such cording to a set of shape types; of assessing the as rock coasts – some of them bore into rocks to connectivity degree between different basin pools; depths of ~1.0 mm (Jones, 1989; Viles et al., 2000). and of evaluating the number of joints that con- They contribute to micropit creation or enlarge- 490 KRF•2 • OK.indd 490 15.12.2009 11:01:21 Lluís Gómez-Pujol and Joan J. Fornós, Coastal karren in the Balearic islands Figure 3: a. basin pools isolated and coalesced at Binibèquer (BN). Note the control of joints on biggest basin pool shape and limits, width of view is 7 m; b. big basin pools developed on supratidal zone of the profile of Cala Sant Esteve (CS). Inside the basin remnants of former basin pools wal s can be identified; c. fretted pinnacle complete- ly covered by microscopic algae at Cap d’en Font (CF), width of view is 15 cm; d. zone of pinnacles just above the notch and shore platform at Cap d’Artrutx (CA); e. splitkarren perpendicular to the coastline at Punta des Savinar (PS), width of view is 1.5 m; f. rillenkarren developed on Jurassic breccia in the splash zone at Cala d’en Guixar (CG), width of view is 15 cm; g. microril s on a fine grained limestone located in the spray domain, Punta Prima (Menor- ca), width of view is 5.5 cm. Order of pictures from top: a, b, c, d, e, f, g. 491 KRF•2 • OK.indd 491 15.12.2009 11:01:23 Karst Rock Features • Karren Sculpturing ment by organic acid excretion or CO production 10 cm and depth 1.0 to 5.0 cm occur on both verti- 2 (Pomar et al., 1975; Gehrmann et al., 1992; Peyrot- cal and horizontal surfaces between pinnacles and Clausade et al., 1995) or by the physical action of pans. lichen hyphae (Moses and Smith, 1993; Chen et al., Basin pools: These depressions display an el- 2000). Larger borers, such as gastropods, can cre- liptical or irregular plan view and a flat or nearly ate pits directly by physical and chemical action flat bottom that is usually horizontal (Figure 3a, and indirectly by forcing cyanophytes and lichen b). The walls are steep and may display a basal to bore more deeply in order to avoid being con- corrosion notch. Individual basin pools attain di- sumed (Torunski, 1979; Spencer, 1988). It is quite ameters of several metres and depths greater than common in the Balearic islands to find snails such one metre. Coalescence of adjoining pools is com- as Melaraphe neritoides or Melaraphe punctata in- mon, creating larger features with crenulated or side micropits. Their densities in areas intensively irregular plan form. The origin of these features colonized by cyanobacteria can reach 200 to 2,000 is largely attributed to dissolution (Ford and Wil- individuals per m2 (Gómez-Pujol et al., 2002; Kel- liams, 1989), but biochemical processes are really letat, 1980). Some depressions seem to have a pro- important in their formation by causing undersat- portional relation between snail body size and mi- uration of water in basins with respect to CaCO 3 cropit dimensions. at night (Emery, 1946; Schneider, 1976; Trudgill, Pits: Small, closely spaced, roughly circular to 1976a) or directly as a result of biological corro- elliptical depressions of a few centimetres in dia- sion and erosion of the substrate by cyanophytes meter and depth. Such features are attributed to a (Dalongeville et al., 1994; Torunski, 1979) and number of origins including those described for snails, limpets or sea urchins which also attack micropits. Mechanical weathering has been ar- the rock mechanically (Hodgkin, 1970; Trudgill gued in combination with inorganic and organi- et al., 1987; Andrews and Williams, 2000). cally induced dissolution. Salt crystallisation has In the Balearic islands basin pools range in been observed by scanning electron microscopy width from 1.0 dm to 6.0 m and in depth from and X-R diffraction analysis on carbonate rock 4 cm to 1.6 m (Table 1). But considerable differ- coasts in Japan (Matsukura and Matsuoka, 1991) ences between morphometrical parameters can and in Mallorca (Balearic islands) by Moses and be identified across the coast profile. Thus basin Smith (1994). Pits appear in the zone subjected to pools nearest to the sea are narrower than those the salt-laden spray; depressions of diameter 1.0 to that are far away; for instance, in the southern Table 1: Summary of reported basin pool form properties measurement in Balearic islands. Location Lithology Width range Depth range Source Cala Pudent (Mallorca) Quaternary carbonate aeolianite 0.90 to 2.30 m 0.07 to 0.38 m Rosselló (1979) Cala d’en Guixar (Mallorca) Jurassic limestone brecchia 0.45 to 1.20 m 0.12 to 0.33 m Gómez-Pujol and Fornós (2001) Cala d’en Guixar (Mallorca) Quaternary carbonate aeolianite 0.48 to 1.58 m 0.22 to 0.43 m Gómez-Pujol and Fornós (2001) S’Alavern (Mallorca) Miocene reefal calcarenites 0.30 to 1.70 m 0.04 to 0.35 m Gómez-Pujol and Fornós (2001) Cala Figuera (Mallorca) Miocene reefal calcarenites 0.30 to 4.00 m 0.25 to 0.70 m Gómez-Pujol and Fornós (2001) Punta des Sivinar (Mallorca) Miocene reefal calcarenites 0.10 to 6.00 m 0.20 to 1.20 m Gómez-Pujol and Fornós (2001) S’Estret des Temps (Mallorca) Quaternary carbonate aeolianite 0.40 to 2.00 m 0.17 to 0.50 m Gómez-Pujol and Fornós (2001) Es Caló (Mallorca) Quaternary carbonate aeolianite 0.50 to 1.30 m 0.14 to 0.47 m Gómez-Pujol and Fornós (2001) Far de Ciutadella (Menorca) Miocene reefal calcarenites 0.40 to 4.00 m 0.11 to 0.67 m Gómez-Pujol and Fornós (2004) Cap d’Artrutx (Menorca) Miocene reefal calcarenites 0.56 to 3.20 m 0.10 to 0.53 m Gómez-Pujol and Fornós (2004) Cala Turqueta (Menorca) Miocene reefal calcarenites 0.54 to 1.30 m 0.16 to 0.53 m Gómez-Pujol and Fornós (2004) Cap d’en Font (Menorca) Miocene reefal calcarenites 0.49 to 3.12 m 0.12 to 0.45 m Gómez-Pujol and Fornós (2004) Cala Sant Esteve (Menorca) Miocene reefal calcarenites 0.38 to 3.20 m 0.10 to 1.60 m Gómez-Pujol and Fornós (2004) 492 KRF•2 • OK.indd 492 15.12.2009 11:01:23 Lluís Gómez-Pujol and Joan J. Fornós, Coastal karren in the Balearic islands zone zone zone zone shore platform A B C D snails lichens spray own algae ns nacles blue-green algae bar splash red & br sea urchi limpets swash biological presence swash splash spray Figure 4: Morphological, biological and hydrodynamic zonation at coastal karren systems in Balearic islands. Menorca Miocene calcarenites, the mean diame- at the bottom of tubes thin layers of soils or beach ter for basin pools nearest to the sea is 0.62 m and sands may accumulate. Subsoil tubes measured in landward this parameter rises to the 1.66 m. If we the Balearic islands range from a minimum of 4 x compare maximum values between seaward and 4 x 9 cm to a maximum of 26 x 26 x 50 cm. Ford landward basin pools, the same pattern applies: at and Williams (1989) and A. Ginés (1995) identify Cap d’Artrutx study site near the sea, basin pools this kind of form as a feature developed beneath have a maximum width of 2.20 m and landward soil and vegetation cover and exposed after soil of 3.20 m; the same couple for Cala Sant Esteve erosion. In coastal profiles exposed to sea waves, is 1.23 m and 3.20 m. This pattern is also clear in subsoil tubes only appear landward, sometimes the Mallorca coast: in Jurassic limestone, seaward between isolated basin pools; on bare surfaces basin pools reach 0.9 m in mean width and 2.30 m close to the soil layer similar forms partially ex- at landward. The same is true for Upper Miocene posed can be observed. calcarenites and Quaternary carbonate aeolian- ites where width increases from 0.7 to 4.0 m and from 0.4 to 1.2 m respectively. Basin pool depth Linear plan forms fracture controlled is quite variable, and although all features are characteristically tapered, depth values change Microfissures: Small channels guided by micro- from one basin to another according to changes joints, tapering with depth. Microfissures in the in facies and lithology. Basin pools can be isolated Balearic islands limestone rock coast may be some (more likely close to the land) or coalesced (more centimetres wide but rarely exceed more than 5 likely close to the sea): close to the sea basin pools mm in depth. They are more common in the land- are connected in 60% to 90% of cases; far away ward areas. Ford and Lundberg (1987) understand from the sea, basin pools are isolated in 70% to microfissures as a typical inorganic solution form. 90% of cases (Figure 4). Splitkarren: Depressions that are elongated Subsoil tubes or shafts: These features are deeper along joints, veins or fractures. These features than wide, and in plan form rounded or ellipti- range in length from one centimetre to several cal. Tubes evolve under gravitational control en- metres and down to 20 cm in width; with the same hanced by rock discontinuities and joints systems. magnitude scale order for the depth (Figure 3e). If Their vertical walls are smooth and sinuous, and a rock outcrop displays several families of joints, 493 KRF•2 • OK.indd 493 15.12.2009 11:01:24 Karst Rock Features • Karren Sculpturing splitkarren can result in a pseudo-meandering fractures or basin pools. Their walls appear dis- channel. Solution is the main process operating in sected with nested concavities bounded by knife- these features (Ford and Williams, 1989) although edges that often penetrate the rock (Figure 3c, d). the splitkarren walls are carpeted by cyanobacte- Micropits are not gravitationally oriented, giv- ria and may also be related to microkarren forms ing to the pinnacles a spongy appearance and such as pits or micropits. the rock surface is coated by a layer of blue-green algae. Pinnacles are the only coastal karren form positive in topographic terms; Trudgill (1979) ar- Linear plan forms hydrodynamically gues that the rugged topography could be related controlled to the differential solubility of clasts and cements mainly in bioclastic limestone. So rock structure Microril s: Small channels ~1.0 mm in width nor- has an overriding importance and microorgan- mally rounded in cross-section with depth. They ism processes just enhance surface irregularities are sinuous or anastomizing on gentle slopes and produced by inorganic dissolution. Alternatively, straighter on steep slopes. Lengths are up to a few other authors (e.g. Jones, 1989) point out that bio- centimetres. In the Balearic islands microrills in logical processes outweigh solutional disintegra- coastal karren environments occur landward in tion while at the same time lithological variations coastal profiles; they are widespread in the spray control microfloral growth. Moses (2003) links domain (Figure 3g). They are associated with fine biological and salt weathering to pinnacle devel- grained limestones. Capillary flow is believed to opment. Pinnacles in the Balearic islands range explain these features, although wind or gravity in height from 20 cm to 1.3 m and are better de- forces also may play a significant role (Laudermilk veloped on carbonate aeolianite and also on Mi- and Woodford, 1932; Ford and Lundberg, 1987). ocene calcarenite outcrops. They do not appear on Microrills have been identified in different locali- Jurassic breccia outcrops, nor in marls. ties, both in fine-grained Miocene calcarenite and on Jurassic mudstones from Mallorca and Menor- ca (see chapter 7). Organization of coastal karren forms Ril enkarren: Shallow channels with round- bottomed troughs, packed side by side, separated The sequence of coastal karren development in by sharp crested ridges, and commencing at crest the Balearic islands is a general picture where in of slope (Ford and Williams, 1989) (Figure 3f). a meso-scale order of analysis four morphological They are not really frequent in Balearic coastal domains can be separated (Figure 4). After a sub- karren assemblages, although they appear on dif- horizontal shore platform, 1 to 6 m wide, sculp- ferent rock types such as Miocene calcarenite or tured by sea-urchin hollows and completely car- Jurassic breccias. They show width ranges from 0.7 peted by green and brown algae and sometimes to 1.4 cm, and length from 4 to 12 cm. They may preceded by a notch profile, we find the coastal occur on dissected relief, between discrete basin karren forms from the swash domain to inland pools or on splitkarren walls and in splash and until the transition to fully terrestrial environ- swash rush domain. ments. Zone A: This domain is characterized by the transition from the swash to the splash hydrody- Positive remnant forms namic zone in quiet conditions. Waves reach this surface during storms. From a morphological point Pinnacles: Upward-pointing pyramid or projec- of view the presence of isolated pinnacles is the tile-shaped bodies of rock separated by widened clearest feature, those near to the sea are sharper 494 KRF•2 • OK.indd 494 15.12.2009 11:01:24 Lluís Gómez-Pujol and Joan J. Fornós, Coastal karren in the Balearic islands than those that are far away. Their surface is com- Zone D: Here basin pools are the characteristic pletely fretted by micropits and the rock surface has feature. They are generally isolated (around 85% of a dark brown to blue-black coloration due to the cases). The rock surface between them is smooth intense colonization by cyanophytes and lichens or rounded inland where lichens (mainly Verru- ( Rivularia sp., Pynerocol ema sp.). It is easy to find caria sp.) are present. A smooth rock surface is individuals of Melaraphe neritoides or Melaraphe common on basin walls and floors. Winkles such punctata; limpets (mainly Patel a rustica and Pa- as Melaraphe neritoides are abundant; densities of tel a caerulea) are abundant at horizontal surfaces 271 to 897 ind/m2 on basin walls are common. M. between pinnacles or vertical wal s just above the neritoides individuals in Zone D show bigger sizes scarp after the shore platform as wel as snails such than in the other zones (Gómez-Pujol et al., 2002), as Monodonta turbinata and/or Monodonta articu- this fact is related to physical resistance to seawa- lata. Joints widened by solution (splitkarren) that ter impact and also to their reproduction strategy remain in zone A are colonized by filtering barna- (Bosch and Moreno, 1982). Beyond the upper limit cles ( Chthamalus depressus and Chthamalus stel a- of Zone D, where micro-relief is not conspicuous, tus) special y where waves and run-off water flow. forms such as microrills or subsoil tubes appear. The rock surface is very rough and abundant salt These last features lie in the zone of lichen and efflorescence occurs during dry periods. halophytes, just in the transition to the spray do- Zone B: This area corresponds to the extension main to the fully terrestrial environments. affected by wave splash. Pinnacles are the domi- nant feature although they are not isolated. Pinna- cles are joined at their bases by a small wall, result- Major controls on coastal karren ing in a configuration of shallow basins flanked by forms triangular bodies of pinnacles. The general aspect, although a different order of magnitude, is remi- Coastal karren landforms are complex systems; niscent of “cockpit” karst. Densities of Melaraphe many agents and factors act on morphologies neritoides and M. punctata winkles increase and configuration and erosion processes. The proc- limpets and barnacles decrease. ess magnitude, order and frequency is quite var- Zone C: This is a zone when sea spray is the iable according to the focus of interest (Goudie dominant hydrodynamic condition, only sub- and Viles, 1999); different controls can be identi- jected to splash in storms. Basin pools have a fied if the investigation is concerned with bioero- greater degree of connectivity, between 60 to 90% sion or with distribution of forms along the coast. of cases; and those nearest to Zone B share walls So, scale issues are fundamental to the study the between them. Many of them show overhanging coastal karren assemblages, being a fundamental sidewalls, usually fretted by micropits, where den- key to identify the controls of the system under sities of Melaraphe neritoides reach the most dense analysis (Viles, 2001). of the profile. Some localities can reach between 200 to 600 ind/m2, although on lithologies such as carbonate aeolianites, densities of 1,700 ind/m2 Macro-scale controls have been counted (Palmer et al., 2003; Kelletat, 1980). All basin pools have basal coverings of cy- The chemical nature of the carbonate rock is the anophytes while some also contain a layer of salt main key for coastal karren development, but tex- crystals. The rock surface between basins presents ture and structure are also important. Structure a rough texture and much of it is colonized also plays a double role because it guides the outcrop by cyanophytes giving to the rock a characteristic of lithologies susceptible to coastal karren devel- blue to grey color. opment, and also governs the shape of the profile. 495 KRF•2 • OK.indd 495 15.12.2009 11:01:24 Karst Rock Features • Karren Sculpturing      Figure 5: Typology of coast profiles in Balearic islands and coastal karren associated forms. This last feature is also enhanced by lithology and stone and grainstones that are characteristic of discontinuities. Reefal unit (Pomar et al., 2002). So, vertical cliffs The map of coastal karren distribution along are dominant on Lower Bar unit while there are coast of Mallorca and Menorca (Figure 1) shows abundant examples of coastal karren assemblag- that there is not a homogeneous presence of es on the Reefal unit. coastal karren assemblages between and within The control of tectonics and lithology on coast- geological units. The interplay of structure and al karren development is also quite evident in lithology can explain much of this patchy distri- Mallorca. For instance the south-eastern coast bution. For instance, in southern Menorca most is built of Upper Miocene reefal and oolitic cal- rock outcrops affected by coastal karren appear carenites affected by a pre-Holocene tectonic tilt- eastward and westward, the central sector of the ing towards the SE (Fornós et al., 2002). This fact island being poor in this kind of micro-relief. explains the presence of vertical low cliffs in the Southern Menorca is built of Miocene limestone north-east cut into Reefal unit rocks alone, and on which is affected by a gentle anticline (Gelabert, stepped profiles in the south-east built by the suc- 2003; Gelabert et al., 2005); for this reason lower cession of Reefal unit calcarenites to oolitic and Miocene units (Lower Bar unit) appear in the mangrove calcarenites. central southern coast of the island and Upper Based on macro-scale controls, an array of pos- Miocene bodies (Reefal unit) appear to the east sible profiles can be drawn (Figure 5). Completely and west. Muddy calcarenites are the most im- vertical or subhorizontal profiles just allow the portant bulk constituent of the Lower Bar unit development of micropits or pits and some basin and erode more easily than the red-algae rud- pools; however, stepped profiles display the clas- 496 KRF•2 • OK.indd 496 15.12.2009 11:01:25 Lluís Gómez-Pujol and Joan J. Fornós, Coastal karren in the Balearic islands sical succession of forms described above and are the sedimentary environment are also important. the necessary condition for pinnacle development. This is quite evident in coastal karren assemblages developed on Upper Miocene sediments in both Menorca and Mallorca. Upper Miocene is com- Meso-scale controls posed of bioclastic calcarenites and coral reefs growths interlayered with levels of oolitic lime- The hydrodynamic gradient also plays a role as a stones and stromatolites (Fornós and Gelabert, meso-scale control. It is largely responsible for the 1995; Pomar et al., 2002). These materials corre- morphological zonation: differences in sea water spond to the development of a coral reef barrier input and contact with rock control the amount type with wide lagoon, which evolves to a more of physical impact, dissolution or salt deposition littoral environment. Although the extent of the (Moses and Smith, 1994). For instance, the pinna- outcrops are built up by calcarenites or calci siltites, cle domain is closely related to the extent of pro- there are considerable lateral variations according file extension affected by splash; micropits and to the depositional environment (Fornós, 1999). fretted hollows, that overlie rock surfaces in this This fact can imply that in the same profile some domain, contain a layer of salt crystals. Exposed features can evolve easily because the rock facies profiles tend to present bigger areas of rock out- enhances the differential erosion. crop affected by karren sculpturing than sheltered ones. On Mallorcan Cala d’en Guixar (CG on Fig- Micro-scale controls ure 1) Jurassic folded limestones, exposed profiles show karren development for 24 m inland, where- SEM examination showed abundant evidence for as in sheltered profiles coastal karren landforms biological weathering processes along coastal kar- extand only 4 to 10 m landward. This also applies ren profiles of Mallorca and Menorca limestone to Menorcan coastal karren. In exposed profiles coasts. Viles and Moses (1998) note that circular such as in Cap Artrutx (CA) or Cap d’en Font (CF), etch pit and tunnel nanomorphologies, as well as coastal karren development extends more than 20 lichen hyphae and biological patinas, are relat- m inland. In sheltered profiles such as Cala Tur- ed to biological weathering action; whereas crys- queta (CT) or Binibèquer (BN), sculptured surfac- tal boundary widening, V-in-V etching, blocky es extend only 8 to 10 m inland. Obviously sea- etching, grain rounding and deposition of salts water is essential for basin pool development. The or crystal growth are related to inorganic solu- hydrodynamic gradient is also important for bi- tion (Figure 6). Thus, studying the relative abun- ological colonization and zonation. Palmer et al. dance of nanomorphologies can assess the relative (2003) demonstrate that rock surface wetness is a magnitude of these processes across the profile. significant control on both biological film (endo For instance, in Menorcan coastal karren (Table and epilithic algae, lichens and fungi) and graz- 2) there are abundant examples of circular etch- ing organisms (limpets, snails, etc.), the greatest ing in the pinnacles domain and connected basin density of organisms or biomass being near the pools (zones A and B). These forms are the mor- sea edge of the profile. This means that bioerosion phological result of cyanophyte biological activity can be greater seaward than landward; thus biot- (Jones, 1989; Schneider, 1976) and affect both the ic and abiotic factors, such as the hydrodynamic rock grains and the cement between them. Cyan- gradient, reinforce the same pattern maximizing ophyte activity increases rock porosity; so there is the erosion rate at sites exposed and located just a major effective surface attack where addition- above the sea level. al agents and processes (e.g., salt weathering or From a meso-scale control the role of texture wave impact) can act (Fiol et al., 1996). This fact and composition differences on lithology due to can help to explain the fretted surface of coast- 497 KRF•2 • OK.indd 497 15.12.2009 11:01:25 Karst Rock Features • Karren Sculpturing al karren seaward forms. Differential dissolution tures are related to the chemical dissolution that between grains aid detachment by seawater im- occurs below the soil (Ginés, 1999a). When the pact, and rain or wind action leaves depressions soil is eroded, these forms become exposed to on the rock surface. Landward (zones C and D), the weathering agents that slowly give to the rock biological action appears by means of lichen ac- surface a roughened appearance (e.g. microrills, tion. Some act as protective agents because they cockling). If the rock surface remains free of soil carpet the rock surface, and others may play an cover, then it is colonized by lichen and blue-green erosive role because their hyphae grow between algae, which carpet the rock and protect it from grains voids (Moses and Smith, 1993; Chen et al., the physical action of rain drops and sea splash 2000). Dissolution due to inorganic processes or and spray. At the same time the lichen texture in- biochemically-driven processes increases land- creases surface roughness, allowing the retention ward (shown by the rounding of grains). It is less of thin water films, which contribute to chemical important in the pinnacle zone than on surfaces weathering. In addition, weathering is enhanced between isolated basin pools or in most terrestrial by physicochemical action of lichens closely re- environments. However, not all surfaces are fret- lated to their physiological activity (Chen et al., ted: smooth rock surface is also a feature in mm 2000; Viles, 1987). Basin pools have been de- to cm scale. Dissolution processes act, in spatial scribed simply as typical solution forms (Trudgill, terms, in a more homogeneous fashion than the 1987), but this is not necessarily true. Although processes and agents described above. In fact, the these depressions are often filled up by fresh water, spatial continuity of the thin boundary layer of aggressive enough for calcium carbonate dissolu- static water on rock surface and their low turbu- tion, this is not the habitual situation. Most of the lence and laminar behaviour favours the develop- time basin pools are filled by seawater that is sat- ment of smooth surfaces, at least at the nanoscale urated with carbonate, so dissolution is not pos- (Ford and Williams, 1989; Trudgill, 1985). This sible by means of inorganic processes. Schneider fact can explain much of the smooth rock surface (1976) and Trudgill (1987) point out that in basin behaviour in the spray domain. pools, undersaturation may occur during dark- ness hours caused by the physiological activity of blue-green algae colonizing rock surfaces. Dur- Processes, zonation and evolution of ing the day, the biological cover consumes CO by coastal karren 2 means of photosynthesis; when sunlight decreas- es and there is not enough light for photosynthe- Of the many sites documented, a general zona- sis, this process stops and CO content increases 2 tion may be delineated using morphologies dis- in basin pool standing water. So the basin pool tribution and attributes. Evidence for “classical” water becomes undersatured with CaCO . Thus 3 dissolution processes was limited to rounded and the solution potential increases and inorganic so- smooth rock surface and subsoil tubes. These fea- lution by itself can clearly be excluded as the ge- Table 2: Summary of nanomorphology abundances at different zones of coastal karren profiles. 0 form not evident; + form evident; + + form well developed; + + + abundant form. Zone Morphological Circular etch pits Crystal boundary V – in – V Blocky etching Rounding Deposition, Biological feature and tunnels widening etching crystal growth patina A Pinnacle edge + + + + 0 + + 0 0 A Pinnacle base + + + + + 0 + + + + + + B Horizontal surface + + + + + + + + + B Basin pool floor + + + 0 0 + + + + + + + C Horizontal surface + + + 0 + + + + + + + + 498 KRF•2 • OK.indd 498 15.12.2009 11:01:25 Lluís Gómez-Pujol and Joan J. Fornós, Coastal karren in the Balearic islands a b c d e f g h i j k l Figure 6: Scanning electron micrographs: a. fracture surface showing circular etch pits, attributed to boring algae activity. Note the “V in V” disolution etching of calcite crystal; b. detail of pitted surface of pinnacle with two gene- rations of solution pittings. Surface is overlaid by case-hardened limestone. This is sculptured by larger hollows which host smaller pits that also appear on the crusted surface; c. detail of endolithic algal cel ; d. circular etch pits developed on rock cement and calcite crystals affected by typical solution and crystallographical y controlled na- nomorphologies; e. biological patina covering rock surface in spray domain; f. blocky etching of the rock cement grains. Note the rounding of the grains; g. salt and gypsum deposition inside rock pores at base of pinnacles; h. combination of salt action, biological patina and filament shaped trenches on a surface between basin pools; i. biological circular etch pits affecting rock cement on a basin pool wall surface; j. high magnification view of base of individual etch pit showing a biological patina carpeting the depression surface; k. blocky etching and “V in V” solution of calcite cristals. Note the widening of the contact between cement grains that enhances rock granular disintegration; l. lichen hyphae and circular etch pits on a basin pool floor sample. 499 KRF•2 • OK.indd 499 15.12.2009 11:01:29 Karst Rock Features • Karren Sculpturing A B C SO BW SW BR process operating on profile line thickness indicates the magnitude and the role of the process 1m process operating but of minor importance Figure 7: Suggested process zonation in the Balearic islands according to SEM observations and morphological or- ganization. SO. dissolution; BW. biological weathering; SW. salt weathering; BR. bioerosion. netic mechanism of basin pool formation (Schnei- weathering affects most of the profile, although it der, 1976; Trudgill, 1987). This biologically-driven is really important and intensive in zones A and B process is probably responsible for the gross form and with a minor role in C. The third vector, bio- of micropits and pinnacles, although salt weather- logical erosion, caused mainly by grazers (limpets, ing (Moses, 2003; Moses and Smith, 1994) and the sea urchins and snails), is constrained mainly to biological action should not be forgotten (Jones, zone A and to the vertical walls of sea edge pro- 1989). On the seaward edge, the region of major files. Finally, salt weathering affects mainly zone biological colonization (Palmer et al., 2003), graz- A and decreases landward according to the extent ing erosion and biochemical weathering play a of the splash and spray domains. major role, while the physical action of waves con- A general model for coastal karren develop- tributes to rock physical erosion. ment in the Balearic islands is proposed below At least four main weathering and/or erosion (Figure 8). We consider a first stage (T1 on Fig- vectors can be drawn on coastal karren systems ure 8) when the rock surface is not yet affected by according to the organization of coastal kar- marine agents and still suffers mainly terrestrial ren forms and the controls exerted by structure, weathering processes. The soil layer retreats grad- geological history and hydrodynamic gradient ually, extending the surface exposed to marine (Figure 7). The first of these corresponds to dis- agents (Figure 8.T1). The second stage represents solution weathering understood as inorganically- the initial development of basin pools by means of driven dissolution. This vector decreases seaward some inorganic dissolution (fresh rainwater and and has its main morphological expression in run-off) but mainly by biological control of blue- subsoil exhumed forms. It is especially important green algae that colonizes rock surfaces. Basin in zones D and in C. Biological weathering or bi- pool development may be enhanced by rock joints ologicaly-driven dissolution is the second vector and discontinuities (Figure 8.T2). In stage 3, the and the most important weathering agent along soil retreats landward and leaves smooth depres- coastal karren profiles. Because the biochemical sions to be colonized where seawater is added by action of organisms controls the chemical prop- wave splash. At the same time older basin pools erties of water standing in basin pools, biological become wider and deeper and some of them share 500 KRF•2 • OK.indd 500 15.12.2009 11:01:30 Lluís Gómez-Pujol and Joan J. Fornós, Coastal karren in the Balearic islands subaerial weathering 1m subaerial weathering zone A B C D marine agents soil retreats basin pool development zone A B C D marine agents basin pool enlargement isolated basin pool soil retreat pinacle and connection enlargement development subsoil weathering zone A B C D pinnacle and connected basin pool development and enlargement subsoil forms reworked subsoil tubes zone A B C D coastline retreat subaerial weathering pinnacle and connected basin pool isolated basin pool development and enlargement enlargement and connection pinnacle destruction subsoil tubes zone A B C D Figure 8: Evolution model for coastal karren landforms in Balearic islands. 501 KRF•2 • OK.indd 501 15.12.2009 11:01:31 Karst Rock Features • Karren Sculpturing their walls. Those basin pools, that are nearest to from landward to seaward zones, different stages the sea edge suffer the physical action of wave im- of basin pool evolution in the same profile. pact during storms and their thinner walls break Despite the major imprints of geological history down. According to the density of joints and to and structural control on profiles, coastal karren rock properties some points or basin pools walls systems are organized according to the biological may be more resistant and will remain as pyrami- zonation and the hydrodynamic gradient; thus it dal bodies; this is the initial pinnacle develop- is an ecological zonation. The ecosystems present ment (Figure 8.T3). Generally the rock surfaces along the coast profile control and in some cases nearest to the sea increase their roughness due to govern by their ethology the rock morphology biological weathering and erosion combined with and it may occur at different scales and orders of salt weathering. Previous basin pools developed magnitude; from facts as the fluctuation of water in zone B have evolved to isolated pinnacles, and chemical properties in basin pools to rock fatigue those previously isolated basin pools become wider caused by the physical erosion enhancement by and share the walls between them. Some of them boring and biological increase in porosity. Thus, coalesce and initial elliptical plan forms evolve the coastal karren cannot be understood as clas- to complex forms. In stage 4 landward, isolated sical exokarstic landforms. Dissolution is the basin pools enlarge. Soil retreat allows the inter- dominant process on limestone rocks, but this is action between marine forms and subsoil forms. induced directly or indirectly by biological activ- At the seaward edge a narrow shore platform and ity. So, coastal karren should be understood as a a notch develop and those former isolated pinna- complex example of biokarst in the sense of Viles cles nearest to the sea edge are destroyed, leaving (1984). some irregular topography on the notch roof (Fig- ure 8.T4). The following stage corresponds to the visor break and this fact implies the displacement Acknowledgements of the coastal karren system landward at the same time that the shore platform enlargement occurs We would like to thank Angel Ginés, Joyce Lun- (Figure 8.T5). The model follows in a cycle as the dberg, Guillem X. Pons and Lluís Fiol for helpful evolution of the coastal karren system follow, as comments, as well as the work of the technical we are once again in the third described stage. staff of the SEM unit of the Balearic islands Uni- Two features should be pointed out from this versity. Also we are in debt to Pau Balaguer, Joan model that combines morphological zonation and Miquel Carmona, Marta Asensi and Mariana controls on coastal karren systems at different Baldo for the logistical support during fieldwork. scales. The first is that pinnacles are not a mor- Financial support was received from DGI-FED- phological feature by themselves because they are ER project of the Spanish Government BTE2002- the remnant of the basin pool evolution. In fact 04552-C03-02 and CGL2006-1242-C03-01/BTE. they are just the only one topographically positive L. Gómez-Pujol was in receipt of a FPI scholar- form described in the coastal karren forms classi- ship from the Direcció General de R+D+I, Govern fication. Also the hydrodynamic gradient creates, de les Illes Balears. 502 KRF•2 • OK.indd 502 15.12.2009 11:01:31 coasTal and lacusTrine Karren 41 in WesTern ireland David DREW More than 40% of the island of Ireland is under- yet some of them, particularly those in the west of lain by karstifiable limestone rocks almost all of Ireland, exhibit karren forms in the zone of sea- which are of Carboniferous age. The limestones sonal fluctuation of lake water levels. Two lakes, form the bedrock along the coast in comparative- Muckross and Leane, in County Kerry in the ex- ly restricted zones (Figure 1). However, limestone treme southwest of Ireland, are located on steeply coasts do exist, especially in the east of Ireland, dipping limestone bedrock but are fed by streams north of Dublin; in the west of Ireland on the east- with acidic waters derived from adjacent non- ern and southern shores of Galway Bay and on the carbonate rocks. Inorganic dissolution has gene- Atlantic coast of County Clare, and in isolated rated notches, small caves, dissolution platforms parts of the south coast. Littoral (inter-tidal zone) and limited areas of scal ops, but not the zone of karren are developed to some extent in many of intense karren development found on lakes with these areas of coastal limestone outcrop. saturated water (Priesnitz, 1985). Littoral karren have been investigated by Burke The best developed and best documented exam- (1994) on the low-lying limestone coast of north ples of littoral and lacustrine karren, in Counties County Dublin. The zone occupied by karren Clare and Mayo respectively, are described in this forms is wide as tidal range is some 3 m. In this chapter. area karren development is correlated with the purity of the limestone. However, the features are considerably smoothed and modified by sand Coastal karren, Burren, County Clare abrasion and this, together with the lesser biodi- versity, mean that the littoral karren are less well The Burren karst plateau of County Clare is developed than those found on the Atlantic coast bounded to the north by Galway Bay and to the of Ireland described below. west by the open Atlantic Ocean. Over a 20 km Some 2% of the land area of Ireland is occu- long stretch of shoreline, mainly on the western- pied by lakes, many of which are located on the facing Atlantic coast, pure, highly karstifiable Vi- limestone floored central plain of Ireland (Figure sean limestones comprise the bedrock. The tidal 1). The origin of the lakes is presumed to be due range along this high-energy coastline is in ex- in part to glacial erosion and deposition and in cess of 4 m. The extent of the inter-tidal zone var- part to solution. The waters of most of these lakes ies. The limestones dip to the south at 2‒5° and on are saturated with respect to calcium carbonate dip-slope coastal exposures the littoral zone may 503 KRF•2 • OK.indd 503 15.12.2009 11:01:31 Karst Rock Features • Karren Sculpturing   Figure 1: The distribution of Carboniferous limestone in Ireland and the locations of the main coastal and lacustrine karren zones. 504 KRF•2 • OK.indd 504 15.12.2009 11:01:32 David Drew, Coastal and lacustrine karren in western Ireland Characteristics 1 mean diameter 270 mm 2 depth 80 mm 3 occupied by pools 36 % 4 mean ph 8.5 Characteristics 1 mean diameter 240 mm 2 depth 165 mm 3 occupied by pools 44 % lichens Characteristics 4 mean ph 8.2 1 division = 10 cm 1 mean diameter 250 mm 2 depth 150 mm 3 area occupied by pools 50 % algae and patellidae Characteristics 4 mean ph 8.0 water clint with subaerial karren level 1 mean diameter 430 mm chthomalus 1 division = 10 cm 2 depth 400 mm 3 occupied by pools joint-controlled grike 80 % 4 mean ph 7.1 patellidae and algae pockets of sand water and shell debris soil and vegetation level 1 division = 10 cm mytilus 1 division = 10 cm unkarrenised limestone water emerging from beneath level till cover paracentrotus 1 division = 10 cm mean high water level mean bedding planes jointing low water level Figure 2: Idealized transect of karren forms in the inter-tidal zone on the County Clare coast (adapted from Lund- berg, 1977c). be more than 100 m in extent, whereas on up-dip sion of the bedrock beneath and thus the newly exposures the intertidal zone may be compressed exposed limestone surface is smooth and wholly into less than 10 m horizontally. lacking in dissolution features. At a distance of The karren zones were first described by Lund- less than a metre from the till margin subaerial berg (1977c), with subsequent studies by Trudgill solution has already enlarged the joints and so- (1987), Trudgill and Crabtree (1987) and Trudgill lution hol ows are beginning to form. The wave- et al. (1987) focussing upon the role of marine splash zone is characterized by isolated, shallow organisms in forming the karren features. The pans with lichens being the dominant life form. typical sequence of karren forms in the littoral Within the intertidal zone Littorina, Chthamalus zone is shown in Figure 2 (Drew 2001, adapted and Mytilus-Paracentrotus dominated zones at from Lundberg, 1977c). At the type site shown successively lower levels have been distinguished in Figure 2, at Poulsallagh Bay, morainic depos- (Figures 4, 5, 6). The mean depth of the hollows its blanket the limestone bedrock (Figure 3). The increases from 80 mm to 400 mm and the per- moraine has been eroded by wave action to a centage of the area of limestone occupied by hol- level above high tide and the limestone bedrock lows increases from 36% to 80% from the upper- is being progressively exposed. The calcareous most to the lowest zone. nature of the till has prevented solutional ero- Lateral erosion is eroding away the remnants of 505 KRF•2 • OK.indd 505 15.12.2009 11:01:34 Karst Rock Features • Karren Sculpturing Figure 3: A general view of the limestone foreshore at Poulsal agh, County Clare. In the background is a morainic depo- sit which is gradual y being eroded, exposing intact limestone which is then occupied by karren under sub-aerial conditions. In the centre of the picture is the sloping limestone surface in the inter-tidal zone with the suite of bio- erosion karren, and to the right is an isolated limestone block with Chthamalus zone karren on its top surface. Figure 4: The upper Mytilus zone in which remnants of the original rock surface are still preserved. 506 KRF•2 • OK.indd 506 15.12.2009 11:01:37 David Drew, Coastal and lacustrine karren in western Ireland Figure 5: The lower Chthamalus-Mytilus zone in which almost all of the original limestone surface has been destroyed, leav- ing only residual stumps of rock. the original limestone surface in each zone and Lacustrine karren, Lough Mask, that process, combined with the vertical erosion, County Mayo means that the zones are progressively migrating landwards. Enlarged joints which are a prominent Loughs Corrib (190 km2 in area), Mask (90 km2 feature of the uppermost zones are not apparent in area), and Carra (20 km2 in area) are located in any of the lower zones where bio-erosion and in the east of Counties Galway and Mayo (Figure the development of kamenitzas are all-important. 1). Loughs Corrib and Mask occur on the bound- Examples of the littoral karren zones are given in ary between limestone to the east and non-calcar- Figures 3, 4, 5 and 6. The karren suite shown in eous rocks to the west whilst Lough Carra is lo- Figure 2 is that which occurs on sections of coast cated wholly on limestone. Lough Mask, up to 60 where the limestone dip slope is stepped with suc- m deep in its western part, has a catchment area cessively lower bedding planes hosting the differ- of approximately 1,000 km2, 60% of which is lime- ent karren assemblages. On uniformly sloping stone floored. Over the greater part of the lime- coastal sections the zones grade into one another stone shoreline of these lakes the bedrock is blan- whilst on cliffed sections of coast some or most of keted beneath till, peat or marl. However, such the zones may be absent. Lundberg (1977c) regard- deposits are absent along part of the shoreline of ed solution as being of great importance in devel- Lough Carra, the northeastern shore of Lough oping the pits in the littoral zone. Trudgill (1987) Corrib and the southeastern shore of Lough Mask regarded boring by marine organisms as being of and gently dipping Carboniferous limestone greatest significance, especially in the lower zones forms the lake margins. Loughs Mask and Cor- where he records boring rates of 1.2‒10 mm per rib are linked by subterranean channels; numer- annum. Simms (1990), discussing the existence of ous sink in the southeastern part of Lough Mask photokarren in caves on the foreshore and its rela- draining to a series of very large springs in the vil- tion to the karren zonations, also favours boring lage of Cong on the shores of Lough Corrib (Drew as the primary mechanism. and Daly, 1993). It is in this 8‒10 km long reach of the shoreline of Lough Mask, including the swal- 507 KRF•2 • OK.indd 507 15.12.2009 11:01:39 Karst Rock Features • Karren Sculpturing Figure 6: A large pool near low tide level, occupied by and eroded by Para- centrotus lividus which is enlarging lateral y to consume the higher Chthamalus-Mytilus dom- inated zone of pools and spires. low hole zone, that the lacustrine karren are best the zone of seasonal fluctuation of lake water level. developed. The strata dip at 2‒5° to the southeast This fluctuation, originally averaging some 2.9 m, and comprise mainly pure bioclastic limestones is associated with the capacity of the underground with some dolomites. The development of karren drainage system from Lough Mask in compari- does not appear to be noticeably influenced by li- son to the inflow to the lake from surface rivers. thology. In places on the dip slopes the width of A complicating factor is that in 1855 a canal (the the karren-bearing limestones exceeds 50 m but Cong canal) was completed linking Loughs Mask there are also numerous scarps 1‒5 m in height. and Corrib via a surface channel, which had the The lake-shore karren have developed within effect of lowering the height of maximum water 508 KRF•2 • OK.indd 508 15.12.2009 11:01:41 David Drew, Coastal and lacustrine karren in western Ireland Figure 7: The lakeshore of southeastern Lough Mask at low summer water level. In the fore- ground the limestone is covered by sediment derived from till and kar- ren have not developed. The centre-background shows the gently in- clined bedrock surfaces covered with small so- lution pits. The isolated boulders have pits on their upper surfaces and rohrenkarren on the un- derhangs. Figure 8: Small solution pits on 350 x 500 mm area of limestone bedrock in the zone of seasonal in- undation on the shore of Lough Mask. levels in Lough Mask in winter and allowing sum- growth beyond the reach of lake water throughout mer lake levels to fall lower than hitherto. At a cer- the year. The karren phenomena were first noted tain lake water level the canal becomes dry. Thus in passing by Kinahan and Nolan (1870) and sub- since 1855 the zone of active karren development sequently by Ford and Williams (1989). However, associated with seasonal lake level fluctuations the only scientific studies of the phenomena are has been shifted down by ca. 1 m, allowing previ- those made by Quigley (1984) and Simms (2002). ously permanently inundated areas of limestone Two distinct karren forms occur in the zone of to experience water level oscillations but placing seasonal inundation (Figure 7). On horizontal or the highest levels of the zone of former karren gently sloping surfaces, whether of bedrock or on 509 KRF•2 • OK.indd 509 15.12.2009 11:01:45 Karst Rock Features • Karren Sculpturing boulders, every bare limestone surface is pitted chemical, taking place when water vapour in the with hollows, 20‒120 mm in diameter and 10‒120 air trapped by rising lake levels, condenses at times mm in depth ( lacustrine pittings). On overhang- when the rock is colder than the water. He argues ing surfaces, whether on the underside of isolated that the condensate trickles downwards and cor- boulders or on the upper surfaces of enlarged bed- rodes the limestone. Initiation of these features re- ding planes on the small scarps, are found tubu- quires calm and saturated lake water. lar features, tapering upwards to a rounded apex. In the lake-shore zones where the karren forms These features, termed rohrenkarren ( tube karren) are best developed, individual limestone beds may by Simms (2002), occur at the same density as the have coalesced pits on their upper surface and solution hollows. The occurrence of the karren rohrenkarren coalesced to develop downward forms is summarized in Table 1. pointing spires, on their lower surfaces with some According to Quigley (1984), the smaller pit- rohrenkarren having penetrated the full thick- tings (Figure 8) range from conical to cylindrical ness of the bed to form tubes linking the upper hemispherical in shape. When they are closely and lower surfaces of the bed. spaced they coalesce leaving a surface of sharp The effects of the lowering of water levels in residual pinnacles. As the pits enlarge so they Lough Mask due to the construction of the canal become more variable in size within a particular (Figure 11) are evident in the pittings at the high- area. There is a good correlation between altitude est levels – those levels that are not now inundated (a surrogate for inundation duration) and mor- in winter. The pits have developed into pans (Fig- phology (depth and diameter) except for the pits ure 12) with lateral erosion outstripping vertical at the lowest levels (longest periods of inundation) erosion. They are mutating into wholly subaerial which may not have had time to attain their final forms of kamenitza. It is probable that in post-gla- form. The pits are essentially sub-aerial features cial times the drainage channels that link Lough ‒ the pittings at higher levels contain algae ( Cy- Mask with Lough Corrib have become progres- anophyta and Chlorophyta predominantly) whose sively enlarged and that this increased transmis- presence must enhance erosion in the hollows. sivity has further lowered the inundation zone of Rohrenkarren are described by Simms (2002) as Lough Mask. However, no research has been un- commonly having diameters of less than 30 mm, dertaken to determine if ancient solution pits can but with lengths of hundreds of mm (Figures 9, be detected at elevations above the pre-canal inun- 10). They occur at densities comparable to those of dation level. The location of rohrenkarren means the solution pits. The two may coalesce where ro- that their development will effectively have ceased hrenkarren penetrate a limestone bed to a surface within zones that are no longer seasonally inun- above occupied by pits. Simms regards the devel- dated. All of these karren features are almost cer- opment of rohrenkarren as being purely physico- tainly Holocene in age ‒ some 15 ka ± 1 in this area. Table 1: Zonation of lacustrine karren forms on the shores of Lough Mask in relation to lake water level before and since the lowering of lake water levels in 1855. Innundation regime Karren forms Lake water levels Never submerged Whol y subaerial runoff, pit and pan karren types. 22.65 m a.s.l. (original max.) Former lacustrine solution pits how being modified subaerial y into pans, vegetated pits and Seasonal under original conditions runoff karren. Decaying rohrenkarren. 21.21 m a.s.l. (present day max.) Well developed small solution pits on the upper surfaces of all exposed bedding planes and boulders. Rohrenkarren well developed on lower surfaces of undercut bedding planes. 19.72 m a.s.l. (original min.) Seasonal under present day conditions Incipient solution pits and possibly incipient rohrenkarren on bare rock surfaces. 18.41 m a.s.l. (present day min.) Always submerged Few or no karren forms. Slight solutional enlargement of some joints on bare rock surfaces. 510 KRF•2 • OK.indd 510 15.12.2009 11:01:45 David Drew, Coastal and lacustrine karren in western Ireland Figure 9: Rohrenkarren on the lower surface of an enlarged bedding-plane. Many of the tubes have coalesced to form down- ward projecting spires. Southeastern shore of Lough Mask. Width of view is 70 cm. Figure 10: Rohrenkarren developed on the un- derside of a boulder which has subsequent- ly been tilted from its original position. South- eastern shore of Lough Mask. Discussion described above, are consistent in their morphol- ogy and in the existence of zonation with those As is apparent from the paper by Lundberg (see described from elsewhere. It remains uncertain as chapter 20), coastal karren are well documented to the relative importance that should be attached from a wide variety of limestone coastal envi- to corrosion as against dissolution in explain- ronments and different climatic zones. The litto- ing their origin. Although the present evidence ral karren on the Atlantic coast of County Clare, emphasizes the primacy of bioerosion, the con- 511 KRF•2 • OK.indd 511 15.12.2009 11:01:49 Karst Rock Features • Karren Sculpturing   Figure 11: Seasonal inundation durations for the lake- shore of southeast Lough Mask. trast between the limestone foreshore in County Figure 12: Solution pits on the shore of Lough Mask in Clare with its high degree of karren development the zone that has not been subject to seasonal inunda- and nearby shorelines located on non-calcareous tion since lake levels were lowered by drainage in 1855. rocks with few kamenitzas is striking. The pits are now enlarging lateral y and forming pans. Far less common, certainly as far as descrip- tions in the literature are concerned, are the lake- shore karren (the features on the shores of Lake Of the karren forms considered: one is wholly Huron in Canada being one of the few other ex- physico-chemical in origin (rohrenkarren), one is amples; Vajoczki and Ford, 2000). The initiation largely bioerosional ‒ direct or indirect ‒ (coastal of both the solution pits and the rohrenkarren is karren), and one is the result of both biologically uncertain but their subsequent development is and physically driven solution (lacustrine pit- complementary ‒ both require alternating sub- tings). Both the coastal and the lacustrine karren aerial and submerged conditions, though for ro- seem to require oscillations in water level to de- hrenkarren the period of emergence can be brief velop, though in the case of the coastal site the in- as they do not develop subaerially whilst for the undation cycle is on a diurnal scale whilst for the pits the period of submergence can be brief as they lacustrine site the cycle is annual. do not develop subaqueously. 512 KRF•2 • OK.indd 512 15.12.2009 11:01:52 soluTion pipes and pinnacles 42 in syngeneTic KarsT Ken G. GRIMES Solution pipes (or dissolution pipes), as described beach or shallow marine sands), in finer-grained here, are the small, vertical, smoothly cylindrical material such as chalk and in coarser coquina or pipes found in soft (poorly cemented) porous cal- reef rubble. It is distinguished from the classic (te- carenites, and usually associated with a modern or logenetic) karsts in that the host limestone has ancient soil or a calcrete band. They are typically never been deeply buried and indurated by me- about 0.5 m wide and 2‒5 m deep, though there is sogenetic diagenesis (Choquette and Pray, 1970). significant variation. Pinnacles are associated fea- Apart from being only weakly cemented, a critical tures, but less common. A recent detailed review feature of these limestones is that most of them of solution pipes was given by Lundberg and Tag- still have a significant primary matrix porosity gart (1995) – who advocated “dissolution pipe” as – up to 30%. Within these soft sediments many being a more correct term. of the karst features, including the pipes, have Solution pipes are also known as solution chim- formed at the same time as the sand was being ce- neys, shafts, pits, geological organs, and Lundberg mented into a rock and the term syngenetic karst and Taggart (1995) list other names. The confu- has been applied to that process (Jennings, 1968; sion of terminology is increased by many of those Grimes, 2002, 2006). White et al. (2007) discuss terms also being used for similar features in the the usage of the terms syngenetic and eogenetic epikarst of hard telogenetic limestones, where the karst, which overlap in most situations, and sug- lack of matrix porosity and greater structural con- gest that “syngenetic” be used as the general term, trol require a different genesis. and that “eogenetic karst” be restricted to the sub- set of syngenetic karst which postdates the depo- sitional cycle in which the sediments were formed. Syngenetic and eogenetic karst This chapter deals with solution pipes formed in The development of syngenetic karst soft, porous limestones. These limestones form a special type of karst that has been referred to In calcareous dunes, percolating rain water grad- as syngenetic or eogenetic karst (Jennings, 1968; ually converts the unconsolidated sand to lime- Mylroie et al., 2001; Grimes, 2002, 2006; White stone by dissolution and redeposition of calcium et al., 2007). Syngenetic karst occurs in dune carbonate. Initial solution at the surface forms a limestone ( aeolianite) and other calcarenites (e.g. terra rossa or similar soil depleted in carbonate 513 KRF•2 • OK.indd 513 15.12.2009 11:01:52 Karst Rock Features • Karren Sculpturing but enriched in the insoluble grains (e.g. quartz). in the chalk of Europe, which is finer grained, but At the base of the soil, precipitation of carbon- still relatively soft and porous (Burnaby, 1950; ate forms a cemented and locally brecciated cal- Ford, 1984; Rodet, 1992). crete layer or hard-pan, also known as a cap-rock. Climate appears to be less important than the Within and below this the downward percolat- nature of the host rock, although the global dis- ing aggressive water becomes focussed to dissolve tribution of dune calcarenites seems to be partly characteristic vertical solution pipes, and simulta- controlled by climate and oceanography (Gardner, neously the carbonate dissolved at the surface and 1983; McKee and Ward, 1983). Many aeolianites within the pipes cements the surrounding sand. occur between 20‒40 degrees of latitude, either in Calcrete hard-pans and solution pipes both ap- coastal “mediterranean” climates that have cool pear quite early in the syngenetic sequence, long wet winters and hot dry summers, or in hotter or before the sand is sufficiently cemented to support more arid settings. However, there are exceptions a cave roof (Bastian, 1964). However, the pipes in cooler and wetter climates. continue to develop and deepen as cementation of the host sand continues. Early cementation tends to be localized about roots to form distinctive rhi- The nature of solution pipes zomorphs or rhizocretions. Surface karren forms are rare in syngenetic Form karst, mainly because there is little hard rock available for their formation. Where soil stripping Typically, solution pipes form smooth vertical cyl- exposes the calcrete hard-pan, rainpits and small inders which may narrow towards a rounded base grikes may form, and sharply pitted phytokarst oc- (“cigar shaped” is a common description) or ter- curs in coastal exposures. Subsoil karren are also minate abruptly in a hemisphere (Figures 1, 2). uncommon, apart from the pipes and pinnacles Conical pipes are less common. The pipes have a discussed in this chapter. The top of the hard-pan range of widths, averaging about 0.5 m, but can may show irregular hollows, but it is difficult to be smaller than 0.2 m or over 1 m, although the be sure whether these are solutional, or merely ir- wider ones tend to be less regular, and some may regularities in the top of the cemented zone. Rhi- be due to coalescence of several smaller pipes. zomorphs are common. Depths are typically 2‒5 m, but they can be up to 20 m deep and some may connect with underlying caves (Figure 3). They can occur as isolated indi- Occurrence of solution pipes viduals, widely spaced sets (e.g. 5‒10 m spacing) or in dense fields with spacings that can be closer Solution pipes have been reported from porous than one metre (Figure 1). At Cape Bridgewater, limestones in many parts of the world, in particu- Victoria, Webster (1996) measured densities of 0.7 lar from the dune limestones, also known as dune to 2.8 pipes per m2 (average 1.8) in ten 3 x 3 m calcarenite or aeolianite (Gardner, 1983; McKee quadrats; and pipe diameters ranging from 0.27 m and Ward, 1983). Examples include: the west- to 0.54 m (average 0.40 m). In one 5 x 5 m quadrat ern and southern coasts of Australia (Fairbridge, at the same site Grimes (2004) measured a den- 1950; Boutakoff, 1963; Jennings, 1968; Grimes, sity of 2.1 pipes per m2, a mean inside diameter 1994, 2004, 2006; White, 2000), southern Africa of 0.27 +/– 0.09 m, and mean distance to nearest (Coetzee, 1975), the Mediterranean (Day, 1928; neighbour of 0.46 +/– 0.013 m. Herwitz (1993) re- Marsico et al., 2003), the Caribbean (Lundberg ported mean diameters between 0.2 and 0.37 m and Taggart, 1995; Mylroie and Carew, 1995) and from sites at Bermuda, however his densities were Bermuda (Herwitz, 1993). Similar pipes also occur much less at between 0.33 and 0.60 per m2, though 514 KRF•2 • OK.indd 514 15.12.2009 11:01:52 Ken G. Grimes, Solution pipes and pinnacles in syngenetic karst Figure 1: Stereopair of a cluster of pipes at “The Petrified Forest”, Cape Bridgewater, western Victoria. Note the cemented rims. Width of view is 6 m. Figure 2: Stereopair of the cigar-shaped base, with thin cemented rim, of a pipe near “The Petrified Forest”, Victoria. Width of view is 60 cm. he mentioned densities in localised areas exceed- spaced sequences occur, solution pipes may ter- ing 1.2 per m2. minate on reaching the underlying palaeosoil, or Solution pipes are commonly associated with may drill through it and continue through the un- a present or past soil horizon; either descending derlying dune unit. from it (Figure 4), or cutting through a hard band of pedogenic calcrete that could be a subsoil hard- pan. In stacked dune sequences one commonly Related features sees several levels of palaeosoils, each with a set of associated soil-filled solution pipes. Where closely In the Bahamas the term pit cave has been applied 515 KRF•2 • OK.indd 515 15.12.2009 11:01:55 Karst Rock Features • Karren Sculpturing both to solution pipes, and to larger pits, up to 7 m in diameter and 10 m depth, that have less-regular forms (Pace et al., 1993; Mylroie and Carew, 1995; González et al., 1997). Some of these have hori- zontal or inclined extensions at depth. In some cases these larger pits appear to be due to coales- cence of smaller solution pipes, but many are too irregular to have that origin. Pinnacles, such as those at Nambung in west- ern Australia, may be an extreme case resulting from the coalescence of closely spaced solution pipes in a calcrete band, or they may be due to focussed cementation. They are discussed later in this chapter. Rims and fill material Solution pipes commonly, but not always, have a calcareous cemented rim around them that is a few centimetres thick. Thin concentric micrit- ic calcrete laminae can also line the pipes. Lun- dberg and Taggart (1995) describe in detail the rims, fil s and host rocks at two sites in Puerto Figure 3: A deep, open solution pipe that forms a cave Rico: the rims there were of micrite and micro- entrance. Ladder rungs are spaced 30 cm; photo by R. K. Frank. spar, and there was also replacement of bioclasts Figure 4: A red palaeosoil and soil-filled pipes beneath a younger sand dune exposed in a cliff at Canunda National Park, south Australia. These pipes lack a cemented rim. Width of view is 10 m. 516 KRF•2 • OK.indd 516 15.12.2009 11:01:58 Ken G. Grimes, Solution pipes and pinnacles in syngenetic karst by those cements. Porosities were much less than in the host rock, typically 0‒5%. Cemented rims and fills can be etched out by erosion of the sur- rounding softer sands (Figure 1). Some pipes appear to be filled with a modi- fied version of the original host sediment, and relict structures of the original bedding may be preserved (the “ghost tubes” of Pace et al., 1993). Most, however, are filled with a downward exten- sion of the overlying red or pale brown soil (typi- cally a terra rossa that has been enriched in in- soluble components of the host sediment). Some of the associated soils are partly allogenic rather than entirely residual (e.g. Herwitz, 1993). Pipes can be emptied by loss of their fill downward into an underlying (younger) cave, where they may Figure 5: Concentric laminae in the partly cemented fill form soil cones, or by erosional undermining, of a solution pipe near “The Petrified Forest”, Victoria. or by excavation by sea water or a stream. These Width of view is 40 cm. empty pipes may later be refilled by younger al- logenic material, for example by a younger dune, or during a subsequent marine transgression. Palaeokarst Secondary fills are common in palaeokarst ex- posures, where complex multi-generation fills Solution pipes can be preserved in palaeokarsts can occur (e.g. see Figure 3 of Mylroie and Carew, and are an important clue to the existence of sub- 1995). Fills can be massive, or crudely bedded, or aerial disconformities and hardgrounds in the have concentric cemented layers or calcrete lami- geological record (e.g. Ford, 1984; Wright, 1988; nae (Figure 5). Brecciated material and calcare- Sandler, 1996). The fill material in palaeokarst ous veins occur in some pipes. Many pipe fills pipes may be an important record of deposition have traces of thin calcareous root structures events that have been destroyed elsewhere during (rhizomorphs) embedded in them; as does the the subsequent transgression (e.g. Walkden and surrounding host sand. Davies, 1983). Rhizomorphs Mode of formation Rhizomorphs (or rhizocretions) are hard calci- An early suggestion, by Boutakoff (1963) among fied root structures that are commonly associat- others, was that the pipes were petrified forests; ed with the pipes. Rhizomorphs are common in that is, moulds of buried tree trunks. This had calcareous dunes and have an obvious branching some initial support from workers in Bermuda, root structure. They form from carbonate that has where the pipes were regarded as moulds of pal- been precipitated around the root, and are thus metto stumps; however, recent work has discred- thicker than the original root – which may be ited this (Herwitz, 1993; Grimes, 2004). identifiable as a thin hollow core if that has not Lundberg and Taggart (1995) note that dissolu- been infilled by younger cement. tion by focussed vertical vadose flow of under-sat- urated rain or soil water through the porous sedi- 517 KRF•2 • OK.indd 517 15.12.2009 11:02:00 Karst Rock Features • Karren Sculpturing ment can explain all the features of the pipes: the tree intersect rain, and direct it down the branch- uniform, vertical cylindrical form, the dense clus- es so that it is concentrated at the base of the trunk. tering in places, and the cemented rims (where The concentrated inflow would cause localized dissolved material is re-precipitated at the edges of solution and pipe development (Figure 6a). Her- the pipe). The associated rhizomorphs are formed witz (1993) measured stem-flow under a variety of around rootlets that have penetrated the sands trees in Bermuda and showed that it could gener- from above, possibly following the soil-filled pipes ate significant concentrations of water and noted by preference and radiating out from them. As the that multiple generations of trees could produce pipes are developing downward from the surface the dense spacing of pipes which is observed in or from a soil cover the overlying material can places. progressively fill them as they deepen. But why is the downward water flow focussed into narrow routes rather than travelling evenly Roots throughout the uniformly porous sand? In hard, non-porous, limestone pipes usually form where The influence of tree roots was suggested by Jen- flow is concentrated along the intersections of nings (1968) and Brink and Partridge (1980). joints or steeply-dipping bedding planes. But in Roots generate organic acids and raised CO levels 2 soft sandy limestone there are no vertical joints, that enhance solution in their vicinity (Figure 6b). and the inter-granular porosity is uniform apart A vertical tap-root could therefore form an initial from occasional horizontal hard-bands – the dune thin pipe which would enhance water flow and en- cross-bedding seems to have little effect on flow large with time. This is a self-perpetuating process directions. Three methods of concentrating the as a pipe, with soil fill, would be a preferred place flow have been suggested by Lundberg and Tag- for continuing root growth and organic activity. gart (1995), drawing on earlier authors: surface hollows, roots and stem-flow; to those Grimes (2004) added a fourth: areas of higher porosity Surface hollows within the developing soil hard-pan (Figure 6). In passing, it is worth noting that similar verti- Surface hollows were suggested by Coetzee (1975) cal pipes occur in the giant podsols that develop as a way of concentrating inflow (Figure 6c). If on the porous quartz sand dunes of the Queens- hollows exist (on a partly indurated surface, or on land coast (e.g. Thompson and Bowman, 1984). the top of the soil hard-pan) then water will accu- These have a deep, leached, white A2 horizon mulate in these and the base of the hollows will be over a dark, humic-rich, less permeable B hori- lowered by solution at a faster rate than the sur- zon. Pipes of the leached A2 material from a few rounding higher areas – the process becomes self- centimetres to nearly half a metre wide penetrate perpetuating. several metres down into the enriched B horizon. Spontaneous focussing of downward water flow through the porous sand seems to be involved in Variations in hard-pan porosity that setting also. Solution pipes also occur in later- ite karsts, as discussed in the section on pinnacles. Uneven cementation of the developing hard-pan is a possible fourth process (Grimes, 2004). Rain dissolves carbonate grains as it penetrates the soil, Stem-flow and some of this is re-precipitated lower down to form a hard-pan or calcrete band near the base of Stem-flow is the process whereby the leaves of a the soil. In the initial stages this cemented band 518 KRF•2 • OK.indd 518 15.12.2009 11:02:00 Ken G. Grimes, Solution pipes and pinnacles in syngenetic karst Figure 6: Alternative ways in a) Stemflow b) Tap-roots c) Surface hol ows d) Porous patches which the downward flow of water can become fo- cussed to generate solution pipes (see text). Black arrows are aggressive water flow and red arrows are saturated water. Note, the alternatives are not mutual y exclusive, they could all contribute in different settings. KGG 1-2005. Figure 7: Stages in which a solution pipe deepens and develops a cemented rim. A possible further stage in  which the fill is cemented is  not shown. would not develop evenly (Figure 6d). The better- of saturated water out of the pipe would form the cemented areas would tend to deflect flow laterally cemented rim and also contribute to the general to places which retained more of their original po- cementation of the sand body (Figure 7). Lund- rosity and concentrated inflow would occur there, berg and Taggart (1995) noted that the linings inhibiting further cementation, and allowing so- have many features of pedogenic calcretes. Where lution pipes to form below. pipes become emptied, case-hardening of the ex- posed pipe walls would also contribute to rim cementation. Some fills show “ghost structures” Ongoing evolution of the pipe which indicate that the host sand has had its po- rosity enhanced, without being actually removed. In all four cases, once the inflow is concentrated at Most fills are associated soils that have subsided a point, solution will progressively deepen a verti- into the pipe as it formed, or later allogenic ma- cal pipe beneath the focal point. Lateral movement terial that has entered an empty pipe. These fills 519 KRF•2 • OK.indd 519 15.12.2009 11:02:04 Karst Rock Features • Karren Sculpturing Figure 8: A composite conical pinnacle at Nambung, western Australia, that shows the dune cross-bedding and sec- tions of several small filled solution pipes that have been intersected by the pinnacle. Height is about 2 m. can also be cemented and may show structures of shape of the pit. For example, the inclined pits may pedogenic calcretes. be following indurated dune cross-bedding, and the irregular vertical profiles may reflect various degrees of cementation in the host rock. Some pit Special cases caves seem to show joint control. Some special cases include the larger of the pit caves of the Bahamas and the pinnacles of the Pinnacles Nambung area in western Australia. The larger pit caves are distinguished by their less regular form. The pinnacles at Nambung and other parts of the Instead of smooth cylinders they have irregular coastal dune limestone in western Australia may outlines and may be inclined or bell out at depth. be an extreme case resulting from the coalescence Pace et al. (1993) attributed the Bahamas pit caves of closely spaced solution pipes in a calcrete band to the “concentration of meteoric water by surface (Lowry, 1973; McNamara, 1995), or they may be and subcutaneous channelization”; the same proc- due to focussed cementation. ess described above. However, the more complex These are generally discrete pinnacles with a forms of these larger pits do not agree with the con- conical form (Figure 8), or are cylindrical with a cept of simple focussed flow through a uniformly round top (Figure 9). A few are hollow. They are porous sand. Possibly the larger pit caves are late up to 3 m high and 0.5 to 3 m wide. The broader syngenetic features where the more strongly ce- pinnacles are composite structures with multiple mented limestone exerts structural controls on the peaks (Figure 8). They are the dissected remnants 520 KRF•2 • OK.indd 520 15.12.2009 11:02:06 Ken G. Grimes, Solution pipes and pinnacles in syngenetic karst Figure 9: Smooth cylindrical pinnacles at Nambung developed in the hard calcrete band. Height of the pinnacles is about 2 m. of a cemented band. The upper part of this band together the calcrete may form a phallic bulb at is a hard pedogenic calcrete in which the primary the top of the pinnacle. Sections of an earlier gen- depositional structures have been obliterated, but eration of small solution pipes (0.1 to 0.4 m wide) it grades down into a cemented dune sand where with a hard concentric fill are exposed in both the the dune bedding is still visible. At the base ce- calcrete and the bedded material (Figure 8). The mented rhizomorphs extend downward into the tops of the pinnacles show a summit conform- soft parent sand. Those pinnacles developed in ity which would be the sharp upper surface of the calcrete have smooth surfaces (Figure 9), but the original calcrete band. Where exposed, their those developed below have rough surfaces re- bases may end abruptly or, more usually, grade sulting from the fretting of the dune bedding and downward into less-cemented material character- rhizomorphs (Figure 8). Where both types occur ised by abundant rhizomorphs (Figure 10). Figure 10: A fallen pin- nacle shows a smooth, strongly cemented, upper part and a rough- er area below that is less cemented, and main- ly composed of rhizo- morphs. Width of view is 4 m. 521 KRF•2 • OK.indd 521 15.12.2009 11:02:08 Karst Rock Features • Karren Sculpturing Genesis pipes, to saturated water flow might reflect a cli- mate change. Lowry (1973) and McNamara (1995) suggested that Supporting evidence of this process is given by the pinnacles at Nambung may be residual features some calcrete hard-pans which have bulbous ce- resulting from coalescence of densely spaced solu- mented pendants descending from them into the tion pipes that dissected a cemented calcrete band. softer sand below (Figure 11). These inverted pin- The genesis is complicated by the presence of an nacles could result from focussed cementation. earlier generation of solution pipes, with cemented The focussed cementation process differs from concentric-banded fill, that is exposed in the sides that of the solution pipes in that the pipes are self- of the later pinnacles (Figure 8). Lowry (1973) sug- perpetuating and can drill down to great depths, gested the following stages in development of the whereas the vertical cemented zones would reduce Nambung pinnacles: the permeability and deflect the flow so that the • formation of the dunes as loose calcareous sand; cemented area spreads horizontally and eventually • development of a hard cap-rock (hard-pan) cements the whole dune. Perhaps pinnacles are comprising cemented calcarenite, recrystallised less common than pipes because we only see them micritic limestone and banded secondary lime- where the cementation is incomplete. stone (calcrete). Solution pipes develop and be- Both of the suggested processes, coalescing come filled with concentric layers of calcrete; solution pipes and focussed cementation, could • continued leaching sculptures the cement- be valid. The cylindrical pinnacles (Figures 9, 10) ed limestone into pinnacles up to 4‒5 m high, might have formed by focussed cementation, as which cut across the earlier structures of dune would the hollow pinnacles which would be due to bedding, rhizomorphs, cemented solution pipes, cementation around a solution pipe. However, the and calcrete. The pinnacles are covered by 4‒5 composite pinnacles (Figure 8) might be the result m of loose yellow quartz sand; of coalescing pipes. • erosion of the loose sand has exposed the pinna- cles. McNamara (1995) extended Lowry᾽s model to Other pinnacles suggest that some of the more cylindrical pinna- cles might have formed by cementation around tap In France, Rodet (1992) described subsoil pinna- roots in zones up to 1 m wide. He also noted that cles in the chalk, exposed at the coast and known some of the small pinnacles could be the cemented as bonshommes de craie. He attributed these to co- fill of prior solution pipes. alescence of conical solution pipes, his racines du An alternative origin for the pinnacles could be manteau d’altération. Waltham (2001) described as a result of focussed cementation – the focuss- 2‒4 m high pinnacles in chalk in the Egyptian ing would be in a similar way to that described for desert and attributed them to the same solutional solution pipes, but in this case instead of the down- processes that produce stone teeth in hard lime- flowing water being aggressive, it was saturated stones. The Egyptian chalk pinnacles have been and so cemented the sand in vertical cylindrical modified by sand-blasting and thermal shattering patterns. The source of the saturated water would and are larger than those at Nambung, so it is dif- be the topsoil of the dune, or possibly younger ficult to compare the two areas. dune sands which buried the initial dune. The lat- Pinnacles are also reported as epikarst features ter situation could explain the earlier generation buried beneath phosphate deposits on several oce- of solution pipes exposed within the pinnacles at anic islands (e.g. Jacobson et al., 1997), but unfor- Nambung. Alternatively, the change from unsatu- tunately there is generally insufficient information rated water that produced the earlier generation of on the character of the host limestone (in particu- 522 KRF•2 • OK.indd 522 15.12.2009 11:02:08 Ken G. Grimes, Solution pipes and pinnacles in syngenetic karst Figure 11: Cemented lobes descending from a hard-pan layer at Naracoorte, south Australia, suggest focussed ce- mentation by downward moving water. Width of view is 5.5 m. lar, its matrix porosity and cement) to allow com- but continuing to deepen and evolve after the sand parison with the Nambung pinnacles. On Christ- has been converted to a soft limestone. They can mas island, in the Indian Ocean (Grimes, 2001), contain a variety of fill materials, which may give the pinnacles beneath the phosphate are formed clues to the history of the karst surface and are on a hard, micritic limestone that has minimal ma- particularly useful in the interpretation of palae- trix porosity. Those pinnacles are best classed with okarst exposures. epikarst features on hard, telogenetic limestones; Solution by focussed vertical vadose seepage they are not the same as the syngenetic pinnacles through the porous sand can account for both iso- on the calcarenites at Nambung. lated pipes, and the dense fields of pipes. Note that There are analogies with laterite karsts. In the four alternative modes of focussing water flow northern Australia deep weathering profiles and discussed above are not presented as mutually ex- associated ferruginous and siliceous cemented clusive hypotheses – all could act, either together duricrusts show both pinnacles and solution pipes or separately, according to the local situation in (Grimes and Spate, 2008). These are also “syn- any area. The associated pinnacles may be an ex- genetic” in that they formed at the same time as treme case in which solution pipes cutting through the weathering profile, and they also appear to a cemented band have coalesced to leave residual have formed by focussed cementation (the pinna- areas of hard limestone; or they may be the result cles) and solution (the pipes). A significant number of focussed cementation by down-flowing saturat- of laterite pinnacles are hollow, which suggests ce- ed vadose water. mentation adjacent to a pipe. Acknowledgements Conclusion My colleague, Susan White, has contributed to Solution pipes are distinctive features of soft po- many discussions on the nature of these and other rous limestones, in particular dune calcarenites. features of the calcareous dunes. Andy Spate com- They are syngenetic karst features, developing in mented on an early draft of this paper. I also thank the early stages of cementation of the loose sand, my wife, Janeen Samuel, for assistance in the field. 523 KRF•2 • OK.indd 523 15.12.2009 11:02:10 KRF•2 • OK.indd 524 15.12.2009 11:02:10 The Karren landscapes in 43 The eVaporiTic rocKs of sicily Giuliana MADONIA and Ugo SAURO The karren in the evaporitic rocks of Sicily show Setting wide distribution and variety of shapes relating to the large extension of the rocky outcrops, to the The most complete and extensive sequence of different lithofacies and to the climate. Karren evaporitic rocks in the Mediterranean basin out- features are largely present in all kinds of evap- crops in Sicily, covering a total area of more than orites: macrocrystalline selenitic gypsum, detritic 1,000 km2 (Macaluso et al., 2003). This series con- gypsum with various grain size, microcrystalline sists mostly of gypsum rocks of Messinian age gypsum and in salts such as halite and kainite. showing nearly all the possible lithological vari- Both the origin and the evolution of the karren eties such as selenitic (or macro-crystalline), ala- are controlled by the dynamics of several process- bastrine, laminated balatino and various detritic es such as solution and recrystallization, granu- lithofacies. The range of crystal sizes is really wide, lar disintegration, carbonation, and phenomena from tenths of millimetres to several decimetres, linked to biological activity. and the porosity of the rocky mass also varies The karren features have different dimensions, widely. ranging from the nano- and the micro-forms to The gypsum rocks present a multiplicity of mor- very large forms, and develop both on the exposed phostructural settings and natural environments. surfaces and under permeable covers. Karren are From the morphostructural point of view it is present on extensive outcrops, such as denuded possible to distinguish tabular plateaux, homocli- slopes and hilly summits, and even on the exposed nal ridges, fault scarps and faulted blocks, folded faces of little stones and isolated blocks. Peculiar ridges, isolated large gypsum blocks floating on environments where some specific types of karren clays, and different types of landslides. From the have been recognized are the fluvial and coastal point of view of the natural environments, the geo-ecosystems, some artificial and semi-artificial slope geo-ecosystems, the fluvial geo-ecosystems, geo-ecosystems such as quarries, the dumps of the lacustrine geo-ecosystems, the coastal geo- mines and dry walls. ecosystems and the hypogean geo-ecosystems can Generally several analogies can be drawn be- be distinguished. High mountain environments tween the gypsum karren in Sicily and limestone do not exist, because the highest gypsum outcrops karren, despite important differences. do not reach 1,000 m a.s.l. 525 KRF•2 • OK.indd 525 15.12.2009 11:02:10 Karst Rock Features • Karren Sculpturing The climate regime is of mediterranean type slopes and to the consequences of human impact with a rainy season (autumn–winter, where 80% on the fragile environments characterized by the of the precipitation amount occurs) and a summer typical mediterranean climate. The land use of the dry season, even lasting in some southern areas steeper slopes, as pasture for sheep and goats, has up to 7 months. The annual temperature range caused desertification starting in prehistoric times is moderate, especially in the coastal areas. The (Figure 1). Soil erosion has been effective and total highest summer temperatures (more than 42°C) also on relatively gentle slopes and hilly summits, have been recorded in the southern-central areas. due to the scarcity of grikes and holes normally acting as traps for soil sediments. On the other hand, inside most of the closed depressions such Gypsum karren in Sicily as dolines there are consistent amounts of fill- ings, made up of clays and other residual materi- In nearly all environments exposed rock surfac- als present on the gypsum surfaces (Sauro, 1996). es are common, due both to the gradient of many On the exposed gypsum surfaces there is a Figure 1: Desertified gypsum slope exposing pseudo- Figure 2: Pavements with runnels and ril s in macro- bedding planes resulting by tensional relaxation (Palma crystal ine gypsum (Palma di Montechiaro, Agrigento). di Montechiaro, Agrigento). Width of view is 5 m. Figure 3: Association of different types of karren devel- Figure 4: Mini-craters (rainpits) with slopes interested by oped on alabastrine gypsum: rillenkarren, heelprint kar- the development of microril s. ren and solution levels (Montallegro, Agrigento). 526 KRF•2 • OK.indd 526 15.12.2009 11:02:15 Giuliana Madonia and Ugo Sauro, The karren landscapes in the evaporitic rocks of Sicily great density of karren-like features ranging and some present significant differences (Macalu- from the nano- and the micro-forms to very large so and Sauro, 1996a; Macaluso et al., 2001). forms. Even if large forms are present only on the In particular, most of the spotted, linear and extensive outcrops, such as on the denuded slopes small planar forms are hydrodynamically control- and hilly summits, it is possible to identify micro- led, like the minute rain craters or rainpits (Figure and small karren, often covering nearly all the 4), the micro-karren ( microril s, etc; Figures 4, 7), surface, on the exposed faces of little stones and the solution rills ( ril enkarren), the meandering isolated blocks. The karren are often organized ril s. Some types of heelprint karren and of scal- in complex associations in which different forms, lop-like karren are very well expressed in the fine shapes and dimensions overlap over one and other grained gypsum lithofacies. Beside these forms, (Figures 2, 3, 4). In particular, on solution level- on macro and microcrystalline gypsum and salt, like meso-forms, ril s, heelprints and small runnels solution levels can be easily observed. These forms are present; on the exposed faces of little stones are produced by diffuse solution from a homoge-and isolated blocks rain craters and rills are com- neous water sheet flowing slowly on the surface or mon; on the minute rain craters and on the rills, by water stagnation (Figures 3, 6) (Macaluso and micro-ril s, micro-meanders and micro-ridges can Sauro, 1996a, b); their extent ranges from some be observed. decimetres to several metres. The largest of these Besides the gypsum karren, assemblages of salt forms may be considered meso-forms. karren are present, developed on some salt out- Among the relatively rare or uncommon forms crops related to the evaporitic series, of which only both the covered type forms, like the covered kar- one is natural. The others are linked to salt mines. ren ( rundkarren) and the linear forms (fracture Here solution forms may even develop during controlled) like the grikes, are worthy of note. a single rainfall, but these karren are ephemeral Typical rundkarren, which have evolved in a forms. The more frequent features are rills, minute covered environment, are relatively rare. In fact, spitz, heelprints and small solution levels (Figures they do not develop under common soil which 5, 6). acts as a protective cover, but only below perme- Almost every type of karren already recognized able covers made up of insoluble clastic material. in limestone is present here (Table 1) but some are The best developed hidden forms ( crypto-karren) much better expressed, others are relatively rarer, have been found on the Ciminna plateau (Paler- Figure 5: Short ril s organized in a comb pattern, Figure 6: Small solution levels delimited by small scarps associated with summit minute craters, developed on and crests developed on salt blocks (dump of Case salt blocks (dump of Case Ranieri Mine, Caltanissetta). Ranieri Mine, Caltanissetta). Width of view is 80 cm. Width of view is 45 cm. 527 KRF•2 • OK.indd 527 15.12.2009 11:02:18 Karst Rock Features • Karren Sculpturing Table 1: The main types of karren present in the evaporitic rocks of Sicily (after Macaluso and Sauro, 1996a, modi- fied). mf = meso-forms; mc = micro-forms; sf = small forms; b, d, l = width x depth x length; diam. = diameter; d. = depth; div. = divers; al. = alabastrine; bal. = laminated balatino; disint. = disintegration; tensl. = tensional slackening; biocon = biological control; diffsol = differential solution. Size Nomenclature Relief Dimensions Lithology Geometry Processes Control Environment (b,d, l) mm mc micro-rills negative 1, 1, 50-200 al. bal. gypsum linear solution hydrodynamical bare rock mc micro-ridges positive 0.5-2, 1, 5 al. bal. gypsum linear solution hydrodynamical bare rock mc micro-meanders negative 1-4, 2, 50-400 al. bal. gypsum, salt linear solution hydrodynamical and semicovered rock decantation mc micro-loops negative 0.2, 0.2, 2-5 al. bal. gypsum, salt linear solution hydrodynamical and semicovered rock decantation mc micro-pits negative 3-10, 3-6 (diam., d.) div. gypsum planar circular solution hydrodynamical bare rock mc micro-conduits negative 1-5, variable bal. gypsum planar circular solution structural (fractures) various sf minute rain craters negative 10-20, 5-30 al. bal. gypsum, salt planar circular solution hydrodynamical rocky spikes sf rills negative 3-30, 2-20, 200-1000 div. gypsum, salt linear solution hydrodynamical bare rock sf minute-spitz positive 20, 10-30 al. bal. gypsum, salt linear solution hydrodynamical bare rock sf solution levels negative large variability al. bal. gypsum, salt planar solution hydrodynamical bare rock sf heelprint karren negative 50-200, 5-30, 50-200 div. gypsum planar solution hydrodynamical bare rock sf scallops negative 10-80, 5-20, 20-100 gypsum al., bal. circular solution hydrodynamical bare rock sf meandering rills negative large variability gypsum al., bal. linear solution hydrodynamical and semicovered rock decantation sf runnels negative 30-300, 30-150, div. gypsum linear disint. and solution hydrodynamical and semicovered rock 200-400 m decantation sf meandering runnels negative 4-20, 5-15, 50-700 gypsum al., bal. linear solution hydrodynamical and semicovered rock decantation sf small knobs positive 10-500, 20-200 div. gypsum planar circular biocon; diffsol complex various sf pans in knobs negative 5-30, 10-200 div. gypsum planar circular biocon; diffsol complex various sf pans in boxes negative 10-200, 5-30, 10-300 div. gypsum, salt planar circular diffsol lithological (veins of various calcite) sf, mf grikes negative large variability div. gypsum linear tensl., solution structural (fractures) various sf, mf pits negative 30-500 (diam.) div. gypsum planar circular solution, disint. structural (fractures) various mf rundkarren positive large variability div. gypsum areal solution and structural (fractures) covered and weathering semicovered rock mf pavements positive large variability div. gypsum areal solution and complex various weathering mf pinnacle karst positive large variability div. gypsum areal interface solution complex covered surface mo) where permeable sediments cover a macro- and Sauro, 1996; Macaluso et al., 2001) separated crystalline gypsum. Along slopes where the cover by deep trenches (Figure 9). has been eroded there are many deep and wide The rarest solution features in the gypsum are cutters with rounded bottoms isolating small the grikes and all the minute-shafts penetrating pinnacles (Figure 8). In the Siculiana area (south- inside the mass of the rock. Because of the dy- ern Sicily) a more typical “stone forest” is present namic of the outer gypsum layer (see below), these which evolved under a permeable cover, which last forms only evolve under a permeable cover of favoured the development of an interface karst. insoluble materials and are exposed after its re- Where later erosion, also due to quarrying activ- moval. ity, exhumed this karst, unusual landforms have In terms of karren presenting significant differ- been revealed like upstanding rocky peaks (Forti ences in comparison with limestone karren, the 528 KRF•2 • OK.indd 528 15.12.2009 11:02:18 Giuliana Madonia and Ugo Sauro, The karren landscapes in the evaporitic rocks of Sicily Figure 7: Small block of arenitic gypsum engraved by nu- Figure 8: Crypto-karren with deep and wide cutters often merous radial y arranged microril s (S. Caterina Vil armo- with rounded bottoms isolating small pinnacles, devel- sa, Caltanissetta). Width of view is 2 m. oped under permeable covers subsequently eroded (Ciminna basin, Palermo). Figure 9: Crypto-karren of stone forest type developed under permeable covers exposed by quarrying activity near Siculiana (Agrigento). Width of view is 8 m. solution runnels (rinnenkarren), the small closed basins, selective solution forms and in general all the solution forms influenced by biological proc- esses are worthy of note. In effect, the runnel-like features are not so common and not always so clearly defined as those of the limestones (Figure 2). These forms mainly develop on selenitic gyp- Figure 10: Runnels oriented according to the slope on a surface in selenitic gypsum in the Santa Ninfa area (Ar- sum or on detritic macro-crystalline gypsum, chivio R.N.I. Grotta di Santa Ninfa). sometimes occupying the whole slope (Figure 10). Their development is frequently influenced by the crystalline structure or by the bedding planes. In crystals tend to lengthen along the direction of the particular, runnels on gypsum with iso-oriented long axes of the crystals and thus to stray from the 529 KRF•2 • OK.indd 529 15.12.2009 11:02:24 Karst Rock Features • Karren Sculpturing Figure 11: Runnels and scal- lops by wave splashing and surf erosion devel- oped on a coastal scarp in pelitic gypsum (Marina di Palma di Montechiaro, Agrigento). slope. They also tend to lengthen along the direc- parallel runnels, which are often associated with tion of the bedding on gypsum with subvertical rills, meandering rills and scallop-like forms (Fig- bedding planes (Gatani et al., 1989; Macaluso et ure 11). al., 2001). On gypsum surfaces, classical solution pans On coastal cliffs runnels develop downslope of originated by biological processes, like the ka- the belts affected by wave splashing and the sub- menitzas, do not exist. Howewer closed basins sequent decantation of the water. They are small with diameters ranging from some centimetres to Figure 12: Boxwork and polygon pans evolved by selective solution on a surface of arenitic gyp- sum. The fissures are filled by calcite (S. Cate- rina Vil armosa, Caltanis- setta). Width of view is 90 cm. 530 KRF•2 • OK.indd 530 15.12.2009 11:02:29 Giuliana Madonia and Ugo Sauro, The karren landscapes in the evaporitic rocks of Sicily a few decimetres and, usually, depths of less than one decimetre are frequent. These basins can be isolated or arranged in groups to form a honey- comb karst-like mini-relief. Their evolution is due to the combination of solution by water stagna- tion, bio- and thermoclastic phenomena and slow draining of the water through the intra-crystal- line porosity (Macaluso and Sauro, 1996a). Among the selective solution forms some result from the rock structure, others from differential covers of some species of lichens. Boxwork-like forms are present on gypsum outcrops containing veins of less soluble materials, like calcite result- ing from carbonation (precipitation of calcite). Between these forms, polygon basins and solu- tion levels delimited by enclosures of calcite veins may be easily found. These forms are common on microcrystalline gypsum and on salt outcrops (Figure 12). Between the forms originated by the protective influence of lichen colonies small knobs and enclosures are common on surfaces of micro- crystalline gypsum. The enclosures present a cir- cular pattern and surround closed depressions. Figure 13: Minute ril s and sinuous meandering ril s deve- The ample variability of gypsum karren is loped on steep surfaces fed by slow and protracted also linked with the wide range of environmen- water decantation from soil patches (Montallegro, Agri- tal conditions. According to the local micro-en- gento). vironments it is possible to find gypsum surfaces with very different conditions of soil or detritus cover, also changing in time. These are: clay as a residuum of the solution processes; clay and soil while during dry summers the oversaturated solu- sediments transported by the overland flow on tion migrates upwards and causes the precipitation the gypsum slopes; gypsum crystals originating of gypsum crystals or the growth of pre-existing from the granular disintegration of the rock; and, crystals. This results in an increase of volume of in the coastal environment, also sand brought by the outer layer with a rearrangement of its crystal the breakers. The presence of these thin and often structure. This in turn causes a plastic deformation scattered covers influences the evolution of the of the rock causing the closure of nearly all the plan karren (Macaluso and Sauro, 1996a; 1998b), espe- discontinuities, like the fissures and consequently cially of the sinuous linear forms, like the mean- impeding the development of the grikes. In fact, dering rills and runnels (Figures 13, 14). the grikes are present only on the rocky surfaces On the bare gypsum surfaces a phenomenon previously buried under covers of permeable mate- that interferes in some way with the morphological rials. The forms resulting from the dynamic proc- evolution is the development of a weathering crust esses of the gypsum crust are the gypsum bubbles linked with the cycles of the pore water hosted in- ( tumulos), the pressure ridges, the pressure humps, side the rock (Ferrarese et al., 2002). During wet etc., and also larger forms, like the dome-like hil s winters the rock becomes saturated with water, (Macaluso and Sauro, 1998a; Ferrarese et al., 2002; 531 KRF•2 • OK.indd 531 15.12.2009 11:02:31 Karst Rock Features • Karren Sculpturing Figure 14: Meandering ril s developed on balatine gypsum in the coastal scarp where the wave splashing feeds water decantation (Marina di Palma di Montechiaro, Agrigento). Width of view is 40 cm. Sauro, 2003). On the surfaces of these “pressure” all the elementary types of karren, already recog- forms solution forms are uncommon probably be- nized on the limestones (Perna and Sauro, 1978), cause of both the rearrangement of the gypsum proves that the development of these karren is crystals and the detachment of some of these. mostly controlled by the dynamics of the water flowing above the rocky surfaces (Macaluso and Sauro, 1996a). In fact, the differences in the solu- Conclusions tion processes between the gypsum and the lime- stones are insignificant from the point of view of The presence on the gypsum surfaces of nearly the genesis of the free surface karren. 532 KRF•2 • OK.indd 532 15.12.2009 11:02:32 Giuliana Madonia and Ugo Sauro, The karren landscapes in the evaporitic rocks of Sicily Only the kamenitzas and the rounded karren these forms. The Sicilian Region Authority has al- which have developed in the limestones at the soil/ ready set up some small natural reserves in these rock interface, are strongly influenced by the hy- areas, mostly linked to the presence of show caves. drochemical environment. Anyway, here we do But the exceptional nature of these landscapes is not find typical forms except for those which are worthy of the creation of a special park devoted structurally controlled. both to the natural aspects and to the characteris- The areas of evaporitic rocks in Sicily represent tic nature of the human landscapes (Bianco et al., really extraordinary environments both for their 2003). In the park it would be possible to maintain original landscapes and for the observation and or to revive some of the traditional human activi- study of the solution forms. They are also ideal for ties linked to the use of these distinctive natural carrying out experiments on the development of resources. 533 KRF•2 • OK.indd 533 15.12.2009 11:02:32 KRF•2 • OK.indd 534 15.12.2009 11:02:32 references Abensperg-Traun M., Wheaton G. A., Eliot I. G., 1990: ture mechanics and stress analysis. Geological Soci- Bioerosion, notch formation and micromorphol- ety of London, Special Publication 92: 149–174. ogy in the intertidal and supratidal zones of calcare- Anderson J. A. R., Chai P. P. K., 1982: Vegetation. In: ous sandstone stack. Journal of the Royal Society of Jermy A. C., Kavanagh K. P. (Eds.), Gunung Mulu Western Australia 73: 47–56. National Park, Sarawak. Sarawak Museum Journal Adams C. G., 1965: The Foraminifera and stratigraphy (Special Issue 2) 30, 51: 195–206. of the Melinau Limestone, Sarawak, and its impor- Anderson J. A. R., Jermy A. C., The Earl of Cranbrook, tance in Tertiary correlation. Quarterly Journal of 1982: Gunung Mulu National Park. A management the Geological Society of London 121: 283–338. and development plan. Royal Geographical Society, Adamson A. P., Boase M. J., Howarth C. J., Wilson J. M., London, 345 p. Wilson M. E., 1984: Southampton University Mada- Andrews C., Williams R. B. G., 2000: Limpet erosion of gascar Expedition. University of Southampton, 136 chalk shore platforms in southeast England. Earth p. Surface Processes and Landforms 25: 1371–1382. Alexandersson T., 1976: Actual and anticipated petro- Aubert D., 1969: Phénomènes et formes du Karst jurass- graphic effects of carbonate undersaturation in shal- ien. Eclogae Geologicae Helvetiae 62, 2: 325–399. low water. Nature 262: 653–658. Audra P., 2000: Le karst haut alpine du Kanin (Alpes Allen J. R. L., 1972: The origin of cave flutes and scallops Juliennes, Slovénie-Italie). Karstologia 35: 27–38. by enlargement of inhomogeneities. Rassegna Spe- Auerswald K., 1998: Bodenerosion durch Wasser. In: leologica Italiana 24, 1: 3–20. Richter G. (Ed.), Bodenerosion. Analyse und Bilanz Allison R. J., Goudie A. S., 1990: The form of rock slopes eines Umweltproblems, Darmstadt, 31–42. in tropical limestone and their associations with Bahun S., 1974: Tektogeneza Velebita i postanak Jelar- rock mass strength. Zeitschrift für Geomorphologie naslaga. Geološki vjesnik 27: 35–51. 34, 2: 129–148. Baillie I. C., Liong T. Y., Pang L. C., Phang C. M. S., Alonso-Zarza A. M., Armenteros I., Braga J. C., Muñoz 1982: Soils. In: Jermy A. C., Kavanagh K. P. (Eds.), A., Pujalte C., Ramos E., Aguirre J., Alonso-Gavilán Gunung Mulu National Park, Sarawak, Sarawak G., Arenas C., Baceta J. I., Carballeira J., Calvo J. P., Museum Journal (Special Issue 2) 30, 51: 183–193. Corrochano A., Fornós J. J., González A., Luzón A., Balogh Z., 1998: Saroknyomkarrok vizsgálata az ausz- Martín J. M., Pardo G., Payros A., Pérez A., Pomar L., triai Totes-Gebirgében. Karsztfejlődés 2: 149–167. Rodríguez J. M., Villena J., 2002: Tertiary. In: Gib- Bannink P., Bannink G., Magraith K., Swain B., 1995: bons W., Moreno T. (Eds.), The Geology of Spain. Multi-level maze cave development in the Northern The Geological Society, London, 293–334. Territory. In: Baddeley G. (Ed.), Vulcon Preceedings Ameen M. S., 1995: Fracture characterization in the of the 20th Conference of the Australian Speleologi- Chalk and the evolution of the Thanet monocline, cal Federation. Victorian Speleological Association, Kent, southern England. In: Ameen M. S. (Ed.), Melbourne, 49–54. Fractography: Fracture topography as a tool in frac- Bär W. F., Fuchs F., Nagel G., 1986: Lluc/Sierra Norte 535 KRF•2 • OK.indd 535 15.12.2009 11:02:32 Karst Rock Features • Karren Sculpturing (Mallorca) – Karst einer mediterranen Insel mit nelle rocce evaporitiche: problemi e prospettive. In: alpidischer Struktur (UIS International Atlas of Madonia G., Forti P. (Eds.), Le aree carsiche gessose Karst Phenomena, sheet 5). Zeitschrift für Geomor- d’Italia. Memorie Istituto Italiano di Speleologia 2, phologie, Supplementband 59: 27–48. 14: 115–120. Bartrum J. A., Mason P. A., 1948: Lapiez and solution Biese W. B., 1956: Über Karstvorkommen in Chile. Die pits in basalts at Hokianga, New Zealand. New Zea- Höhle 7, 4: 91–96. land Journal of Science Technology 30, B: 165–172. Biese W. B., 1957: Auf der Marmor-Insel Diego de Basterretxea G., Orfila A., Jordi A., Casas B., Lynett P., Almagro (Chile). Natur und Volk 87, 4: 123–132. Duarte C. M., Tintoré J., 2004: Seasonal dynamics Bird E. C. F., Guilcher A., 1982: Observations prélimi- of a microtidal pocket beach with Posidonia oce- naires sur les récifs frangeants actuels du Kenya et anica seabeds (Mallorca, Spain). Journal of Coastal les formes littorales associées. Revue de Géomor- Research 20: 1155–1164. phologie Dynamique 31: 113–125. Bastian L., 1964: Morphology and development of Bird J. B., Richards A., Wong P. P., 1979: Coastal subsys- caves in the southwest of Western Australia. Helic- tems of Western Barbados, West Indies. Geografiska tite 2, 4: 105–119. Annaler 61A, 3–4: 221–236. Battistini R., 1981: La morphogenèse des plateformes Boardman M. R., Neumann A. C., Rasmussen K.A., de corrosion littorale dans les grés calcaires (plate- 1989: Holocene sea level in the Bahamas. In: Mylroie forme supérieure et plateforme à vasques) et le prob- J. E. (Ed.), Proceedings of the 4th Symposium on the lème des vasques, d’après des observations faites à Geology of the Bahamas. Port Charlotte, Bahamian Madagascar. Revue de Géomorphologie Dynamique Field Station, Florida, 45–52. 3: 81–94. Bögli A., 1951: Probleme der Karrenbildung. Geo- Battistini R., Guilcher A., 1982: Les plates-formes litto- graphica Helvetica 3: 191–204. rales à vasques en roches calcaires: répartition dans Bögli A., 1960a: Kalklösung und Karrenbildung. Zeit- le monde, mer Méditerranée non comprise. Karst schrift für Geomorphologie, Supplementband 2: Littoraux, Comité National Français de Géogra- 4–21. phie, May 1982, Actes du Colloquium de Perpignan Bögli A., 1960b: Les phases de dissolution du calcaire 1: 1–11. et leur importance pour les problèmes karstiques. Bauer F., 1962: Nacheiszeitlliche Karstformen in den Rassegna Speleologica Italiana 4: 16 p. Österreichischen Kalkhochalpen. Actes du Deux- Bögli A., 1961: Karrentische, ein Beitrag zur Karst- ième Congrès International de Spéléologie 1: morphologie. Zeitschrift für Geomorphologie 5, 3: 299–328. 185–193. Beck H. M., 2003: Beneath the Cloud Forests – a His- Bögli A., 1964: Le Schichttreppenkarst. Un exemple de tory of Caving in Papua New Guinea. Caving Publi- complexe glaciokarstique. Revue Belge de Géogra- cations International, Allschwil, 352 p. phie 1, 2: 64–82. Bedi N., Rao K. L. V. R., 1984: Nature and evolution of Bögli A., 1970: Le Hölloch et son karst [Das Hölloch a part of Saurashtra coast, Gujarat, India. Zeitschrift und sein Karst]. Baconnière, Neuchâtel, 109 p. für Geomorphologie 28: 53–69. Bögli A., 1972: Gips in Höhlen. Urner Mineralienfre- Beek W. J., Muttzall K. M. K., 1975: Transport Phenom- und 10, 6: 77–84. ena. Wiley, London, New York, 298 p. Bögli A., 1973: Der alpine Karst in der Zentralschweiz. Belloni S., 1969: Alcune osservazioni sulle acque e Proceedings of the 6th International Congress of Spe- sui depositi al fondo delle “vaschette di corrosione” leology, Olomouc. Academia, Praha, 39–42. (kamenitza) della località Borgo Grotta Gigante Bögli A., 1976: Die wichtigsten Karrenformen der (Carso Triestino). Atti e Memorie della Commis- Kalkalpen. In: Karst Processes and Relevant Land- sione Grotte “Eugenio Boegan” 9: 33–62. forms. Department of Geography, Philosophical Belloni S., Orombelli G., 1970: Osservazioni e misure Faculty, Ljubljana, 141–149. su alcuni tipi morfologici nei campi solcati del Carso Bögli A., 1978: Karsthydrographie und physische Triestino. Atti della Società Italiana di Scienze Nat- Speläologie [Karst hydrology and physical speleol- urali e Museo Civico di Storia Naturale Milano 110, ogy]. Springer-Verlag, Berlin, Heidelberg, New York, 4: 317–372. 292 p. Bernot F., 1985: Podnebje. In: Fabjan I. (ed.), Triglavski Bögli A., 1980: Karst Hydrology and Physical Speleol- narodni park. Vodnik, Bled, 57–61. ogy. Springer-Verlag, Berlin, 284 p. Bianco D., Panzica La Manna M., Sauro U., 2003: Bögli A., 1981: Solution of limestone and karren forma- Tutela e valorizzazione delle aree carsiche italiane tion. In: Sweeting M. M. (Ed.), Karst Geomorphol- 536 KRF•2 • OK.indd 536 15.12.2009 11:02:32 References ogy. Benchmark Papers in Geology 59, Hutchinson Buhmann D., Dreybrodt W., 1985: The kinetics of cal- Ross Publishing Company, 64–89. cite dissolution and precipitation in geologically Bögli A., 1987: Der Karst der Innerschweiz. Mit- relevant situations of karst areas: 1. Open system. teilungen Naturforschende Gesellschaft Luzern 29: Chemical Geology 48: 189–211. 225–235. Bull P. A., Laverty M., 1982: Observations on phy- Bognar A., 1992: Pedimenti južnog Velebita. Geograf- tokarst. Zeitschrift für Geomorphologie 26: 437– 457. ski glasnik 54: 19–32. Burke M. A., 1994: A quantitative analysis of marine Bognar A., Blazek I., 1986: Geomorfološka karta kamenitza on the Carboniferous limestone between područja V. Paklenice 1:25.000. Simpozij o kraškem Skerries and Loughshinny, Co. Dublin. Thesis sub- površju. Acta Carsologica 14, 15: 19–206. mitted as part of B. A. degree, Geography Depart- BOM (Bureau of Meteorology, Australia), 2005: Cli- ment, Trinity College, Dublin, 198 p. mate. http:/ www.bom.gov.au/climate/averages. Burnaby T. P., 1950: The tubular chalk stacks of Sher- Bordoy M., Ginés A., 1990: Observaciones morfométri- ingham. Proceedings of the Geological Association cas sobre la profundidad de estrías de lapiaz (Ril- 61: 226–241. lenkarren) en Mallorca. Endins 16: 21–25. Buser S., 1976: Geological, geomorphological and hydr- Bosch M., Moreno I., 1982: Estructura de las pobla- ogeological investigations of Kanin Mts. Archives of ciones y crecimiento de Littorina neritoides (L. 1758) Geological Institute of Slovenia, Ljubljana, 56 p. (Mollusca, Gastropoda) en las costas de las Islas Bal- Buser S., 1986a: The basic geological map 1:100.000, L eares. Bolletí de la Societat d’Història Natural de 33–64, L 33–63, Tolmin in Videm (Udine). Beograd. Balears 31: 57–66. Buser S., 1986b: Commentary to the map Tolmin and Boutakoff N., 1963: The geology and geomorphology Videm (Udine), L 33–64, L 33–63. Beograd, 103 p. of the Portland area. Geological Survey of Victoria, Butzer K. W., 1962: Coastal geomorphology of Majorca. Memoir 22: 52–58. Annals of the Association of American Geographers Bradley W. C., Hutton J. T., Twidale C. R., 1978: Role of 52: 191–212. salts in development of granitic tafoni, south Aus- Butzer K. W., Cuerda J., 1962: Coastal stratigraphy of tralia. Journal of Geology 86: 647–654. southern Mallorca and its implications for the Pleis- Brandt C. J., 1989: The size distribution of through- tocene chronology of the Mediterranean Sea. Jour- fall drops under vegetation canopies. Catena 16: nal of Geology 70: 398–416. 507–524. Calaforra J. M., 1996: Some examples of gypsum karren. Brandt C. J., 1990: Simulation of the size distribution In: Fornós J. J., Ginés A. (Eds.), Karren Landforms. and erosivity of raindrops and throughfall drops. Universitat de les Illes Balears, Palma de Mallorca, Earth Surface Processes and Landforms 15: 687–698. 253–260. Brink A. B. A., Partridge T. C., 1980: The nature and Calder I. R., 1996: Dependence of rainfall interception genesis of solution cavities (Makondos) in Transvaal on drop size: 1. Development of the two-layer sto- cave breccias. Palaeontology Africa 23: 47–49. chastic model. Journal of Hydrology 185: 363–378. Bromley R. G., 1978: Bioerosion of Bermuda Reefs. Pal- Cardell C., Delalieux F., Roumpopoulos K., Moropou- aeogeography, Palaeoclimatology, Palaeoecology 23: lou A., Auger F., Van Grieken R., 2003: Salt-induced 169–197. decay in calcareous stone monuments and buildings Brook D. B., Eavis A. J., Lyon M. K., Waltham A. C., in a marine environment in SW France. Construc- 1982: Caves in the limestone. In: Jermy A. C., Kavan- tion and Building Materials 17: 65–179. agh K. P. (Eds.), Gunung Mulu National Park. Carew J. L., Mylroie J. E., 1987: A refined geochronol- Sarawak Museum Journal (Special Issue 2) 30, 51: ogy for San Salvador Island, Bahamas. In: Curran H. 95–119. A. (Ed.), Proceedings of the 3rd Symposium on the Brown M. A., 1990: Regional lithofacies and depo- Geology of the Bahamas. Fort Lauderdale, CCFL sitional environments of the Salem Limestone Bahamian Field Station, Florida, 35–44. (Valmeyeran, Mississippian), south-central Indiana. Carew J. L., Mylroie J. E., 1995a: Quaternary tectonic In: Thompson T. A. (Ed.), Architectural elements stability of the Bahamian Archipelago: Evidence and paleoecology of carbonate shoal and intershoal from fossil coral reefs and flank margin caves. Qua- deposits in the Salem Limestone (Mississippian) in ternary Science Reviews 14, 2: 145–153. south-central Indiana. Indiana Geological Survey, Carew J. L., Mylroie J. E., 1995b: Depositional model Guidebook 14: 1–12. and stratigraphy for the Quaternary geology of the Bryan K., 1920: Origin of Rock Tanks and Charcos. Bahama Islands. In: Curran H. A., White B. (Eds.), American Journal of Science 50, 4: 163–174. Terrestrial and Shallow Marine Geology of the Baha- 537 KRF•2 • OK.indd 537 15.12.2009 11:02:32 Karst Rock Features • Karren Sculpturing mas and Bermuda. Geological Society of America, change, and stability of phases in the calcite-seawa- Special Paper 300: 5–32. ter system. Marine Chemistry 5: 75–92. Carew J. L., Mylroie J. E., 1997: Geology of the Baha- Corbel J., 1952: Les lapiaz marins. Revue Géographique mas. In: Vacher H. L., Quinn T. M. (Eds.), Geology de Lyon 37: 379–380. and Hydrogeology of Carbonate Islands. Elsevier, Corbel J., 1959: Érosion en terrain calcaire (Vitesse Amsterdam, 91–139. d’érosion et morphologie). Annales de géographie Carney C., Boardman M. R., 1991: (abstract), Oolitic 366: 97–120. sediments in a modern carbonate lagoon, Graham’s Corrà G., 1972: Morfologie carsiche nella zona di Harbor, San Salvador, Bahamas. Geological Society Canale in Val Lagarina (Val d’Adige meridionale). of America Abstracts with Programs 23, 5: A 225. Studi Trentini di Scienze Naturali A 49, 2: 127–160. Carter N. E. A., Viles H. A., 2003: Experimental investi- Corrà G., Benetti A., 1966: Morfologie carsiche nella gations into the interactions between moisture, rock zona lessinea compresa tra il M. Purga di Velo e il M. surface temperatures and an epilithic lichen cover in Bellocca. Natura Alpina 17: 115–129. the bioprotection of limestone. Building and Envi- Crowther J., 1996: Roughness (mm-scale) of lime- ronment 38: 1225–1234. stone surfaces: examples from coastal and mountain Cecioni G., 1982: El fenómeno cárstico en Chile. Infor- karren features in Mallorca. In: Fornós J. J., Ginés maciones geográficas 29: 57–79. A. (Eds.), Karren Landforms. Universitat de les Illes Cerdá A., 1997: Rainfall drop size distribution in the Balears, Palma de Mallorca, 149–159. Western Mediterranean basin, Valencia, Spain. Crowther J., 1997: Surface roughness and the evolution Catena 30, 2–3: 169–182. of karren forms at Lluc, Serra de Tramuntana, Mal- Chaix E., 1895: Topographie du Désert de Platé. Le lorca. Zeitschrift für Geomorphologie 41, 3: 393–407. Globe, Société de Géographie de Genève 34: 44 p. Crowther J., 1998: New methodologies for investigat- Chen J., Blume H. P., Beyer L., 2000: Weathering of ing rillenkarren cross-sections: a case study at Lluc, rocks induced by lichen colonization, a review. Mallorca. Earth Surface Processes and Landforms Catena 38: 121–146. 23, 4: 333–344. Chen X., Gabrovšek F., Huang C., Jin Y., Knez M., Cucchi F., Forti F., Finocchiaro F., 1987: Carbonate sur- Kogovšek J., Liu H., Petrič M., Mihevc A., Otoničar face solution in the Classical Karst. International B., Shi M., Slabe T., Šebela S., Wu W., Zhang S., Journal of Speleology 16, 3–4: 125–138. Zupan Hajna N., 1998: South China Karst I. Založba Cucchi F., Forti F., Marinetti E., 1996: Surface degrada- ZRC, Ljubljana, 247 p. tion of carbonate rocks in the karst of Trieste (Clas- Chen Z., Song L., Sweeting M. M., 1986: The pinnacle sical Karst, Italy). In: Fornós J. J., Ginés A. (Eds.), karst of the stone forest, Lunan, Yunnan, China: an Karren Landforms. Universitat de les Illes Balears, example of a subjacent karst. In: Paterson K, Sweet- Palma de Mallorca, 41–51. ing M. M. (Eds.), New Directions in Karst. Geo- Cucchi F., Radovich N., Sauro U., 1990: I campi solcati books, Norwich, 597–607. di Borgo Grotta Gigante nel Carso Triestino. Inter- Choppy J., 1996: Les cannelures et rigoles sont des indi- national Journal of Speleology 18, 3–4: 117–144. cateurs climatiques. In: Fornós J. J., Ginés A. (Eds.), Curl R. L., 1966: Scallops and flutes. The Transaction of Karren Landforms. Universitat de les Illes Balears, Cave Research Group of Great Britain 7, 2: 121–160. Palma de Mallorca, 137–148. Cvijić J., 1924: The evolution of lapiés. A study in karst Choquette P. W., Pray L. C., 1970: Geologic Nomencla- physiography. Geographical Review 14: 26–49. ture and Classification of Porosity in Sedimentary Cvijić J., 1926: Geomorfologija 2. Beograd, 505 p. Carbonates. American Association of Petroleum Cvijić J., 1927: Škrape. Glasnik Srpskog geografskog Geologists, Bulletin 54: 207–250. društva 13: 17–29. Cílek V., 1989: Rosná koroze rápencu z pustrich oblastí Dalongeville M., 1977: Formes littorales de corrosion Iránu a Pákistánu. Československy Kras 40: 135–137. dans les roches carbonatées au Liban: Étude Mor- Coetzee F., 1975: Solution pipes in coastal aeolianites phologique. Méditerranée 3: 21–33. of Zululand and Moçambique. Transactions of the Dalongeville M., Guilcher A., 1982: Les plates-formes Geological Society of South Africa 78: 323–333. à vasques en Méditerranée, notamment leur exten- Conca J. L., Rossman G. R., 1985: Core softening in cav- sion vers le nord. Karst Littoraux, Comité National ernous weathered tonalite. Journal of Geology 93: Français de Géographie, May 1982, Actes du Collo- 59–73. quium de Perpignan 1: 13–22. Cooke R. C., 1977: Factors regulating the composition, Dalongeville R., Le Campion T., Fontaine M. F., 1994: Bilan bioconstruction-biodestruction dans les 538 KRF•2 • OK.indd 538 15.12.2009 11:02:33 References roches carbonatées en mer Méditerranée: étude Day M. J., 1982b: The influence of some material prop- expérimentale et implications géomorphologiques. erties on the development of tropical karst terrain. Zeitschrift für Geomorphologie 38: 457–474. Transactions of the British Cave Research Associa- Daneš J. V., 1911: Physiography of some limestone areas tion 9, 1: 27–37. in Queensland. Proceedings of the Royal Society of Day M. J., 1983: Karst-related management considera- Queensland 23: 75–86. tions in the Gunong Mulu National Park, Sarawak, Danin A., 1983: Weathering of limestone in Jerusalem East Malaysia. In: Dougherty P. H. (Ed.), Environ- by cyanobacteria. Zeitschrift für Geomorphologie mental Karst. GeoSpeleo Publications, Cincinnati, 27: 413–421. 55–76. Danin A., Caneva G., 1990: Deterioration of limestone Day M. J., 1997: Conservation issues in Stone Forest walls in Jerusalem and marble monuments in Rome karst. In: Song L., Waltham T., Cao N., Wang F. caused by cyanobacteria and cyanophilous lichens. (Eds.), Stone Forest: A Treasure of Natural Herit- International Biodeterioration 26: 397–417. age. Proceedings of the International Symposium Danin A., Gerson R., Garty J., 1983: Weathering pat- for Lunan Shilin to Apply for World Natural Herit- terns on hard limestone and dolomite by endolithic age Status. China Environmental Science Press, Bei- lichens and cyanobacteria: Supporting evidence for jing, 17–21. eolian contribution to Terra Rossa soil. Soil Science Delaty J. N., 2000: Expédition Bemaraha 1998. Expedi- 136: 213–217. tion Malagasy 1999. Spelunca 78: 11. Danin A., Gerson R., Marton K, Garty J., 1982: Patterns Doornkamp J. C., Brunsden D., Jones D. K. C., 1980: of limestone and dolomite weathering by lichen and Geology, Geomorphology and Pedology of Bahrain. blue-green algae and their paleoclimatic signifi- Geobooks, Norwich, 443 p. cance. Palaeogeography, Palaeoclimatology, Palae- Drew D. P., 2001: Classic Landforms of the Burren oecology 37: 221–233. Karst. The Geographical Association, Sheffield, 51 p. Darabos G., 2003: Observation of microbial weathering Drew D. P., Daly D., 1993: Groundwater and Karstifi- resulting in peculiar “exfoliation” features in lime- cation in Mid Galway, South Mayo and North Clare. stone from Hirao-dai karst, Japan. Zeitschrift für Report Series, Geological Survey of Ireland RS 93, 3: Geomorphologie, Supplementband 131: 33–42. 86 p. Davies T. T., Sutherland A.J., 1980: Resistance to flow Dreybrodt W., 1988: Processes in Karst Systems. Phys- past deformable boundaries. Earth Surface Proc- ics, Chemistry and Geology. Springer Series in Phys- esses 5, 2: 175–179. ical Environments 4, Berlin, New York, 288 p. Davis W. E., 1957: Rillenstein in Northwest Greenland. Dreybrodt W., 1990: The role of dissolution kinetics National Speleological Society Bulletin 19: 40–46. in the development of karst aquifers in limestone: Day A. E., 1928: Pipes in the Coast Sandstone of Syria. A model simulation of karst evolution. Journal of Geological Magazine 65: 412–415. Geology 98: 639–655. Day M. J., 1979: The contribution of geomorphology Dreybrodt W., 1996: Principles of early development of to the formulation of a management plan for the karst conduits under natural and man-made condi- Gunong Mulu National Park, Sarawak, East Malay- tions revealed by mathematical analysis of numeri- sia. Applied Geography Conferences 2: 193–202. cal models. Water Resources Research 32: 2923–2935. Day M. J., 1980: Rock hardness: Field assessment and Dreybrodt W., Gabrovšek F., Romanov D., 2005: Proc- geomorphic importance. Professional Geographer esses of speleogenesis: a modeling approach. Založba 32, 1: 72–81. ZRC, Ljubljana, 375 p. Day M. J., 1981a: Limestone hardness and tropical karst Dreybrodt W., Kaufmann G., 2007: Physics and chem- terrain. Proceedings of the 8th International Con- istry of dissolution on subaerialy exposed soluble gress of Speleology 1, Bowling Green, 327–329. rocks by flowing water films. Acta Carsologica 36, 3: Day M. J., 1981b: Rock hardness and landform develop- 357–367. ment in the Gunong Mulu National Park, Sarawak, Duane M. J., Al-Mishwat A. T., Rafique M., 2003: East Malaysia. Earth Surface Processes and Land- Weathering and biokarst development on marine forms 6, 2: 165–172. terraces, Northwest Morocco. Earth Surface Proc- Day M. J., 1982a: Geomorphological considerations in esses and Landforms 28: 143–144. a management plan for the Gunong Mulu National Dubljanszkij J. V., 1987: Teoreticseszkoje modelirov- Park. In: Frazier J. W. (Ed.), Applied Geography: anije dinamiki formirovanija gidrotermokarsz- Selected Perspectives. Prentice-Hall International, tovüh polosztyej – Metodi i izucssenyija geologics- 197–218. eszkih javlenyij [Теоретическое моделирование 539 KRF•2 • OK.indd 539 15.12.2009 11:02:33 Karst Rock Features • Karren Sculpturing динамики формирования гидротермокарстовых thesis submitted in partial fulfilment of the require- полостей - Методи и изучения геологических ment for the degree of Master of Science, M. Sc. явлений]. Novoszibirszk, 97–111. Thesis, University of Cincinnatti, Cincinnatti, 84 p. Dublyansky V. N., Dublyansky Y. V., 2000: The role Ewers R. O., 1982: Cavern Development in the Dimen- of condensation in karst hydrogeology and speleo- sion of Length and Breadth. Ph. D. Thesis, McMaster genesis. In: Klimchouk A. B., Ford D. C., Palmer A. University, Hamilton, Ontario, 398 p. N., Dreybrodt W. (Eds.), Speleogenesis: Evolution Fabre G., Nicod J., 1982: Lapiés, modalités et rôle de la of Karst Aquifers. National Speleological Society, corrosion crypto-karstique. Mémoires et documents Huntsville, Alabama, 100–112. de géographie, Phénomènes karstiques 3: 115–131. Dunkerley D. L., 1979: The morphology and develop- Fairbridge R. W., 1950: The geology and geomorphol- ment of rillenkarren. Zeitschrift für Geomorpholo- ogy of Point Peron, Western Australia. Journal of gie 23, 3: 332–348. the Royal Society of Western Australia 34: 35-72. Dunkerley D. L., 1983: Lithology and micro-topogra- Farrant A. R., Smart P. L., Whittaker F. F., Tarling D. phy in the Chillagoe karst, Queensland, Australia. H., 1995: Long-term Quaternary uplift rates inferred Zeitschrift für Geomorphologie 27, 2: 191–204. from limestone caves in Sarawak, Malaysia. Geology Dunkerley D. L., 1988: Solution and precipitation of 23: 357–360. limestone in the Chillagoe karst and opportuni- Favre A., 1867: Recherches géologiques dans les parties ties for dating landscape development. 17th Biennial de la Savoie, du Piemont et de la Suisse voisine du Conference of the Australian Speleological Federa- Mont Blanc 3, 71 p. tion, Sydney, 112–117. Ferrarese F., Macaluso T., Madonia G., Palmeri A., Dunkley J. N., 1993: The Gregory Karst and Caves, Sauro U., 2002: Solution and re-crystallization proc- Northern Territory, Australia. Proceedings of the esses and associated landforms in gypsum outcrops 11th International Congress of Speleology, Beijing, of Sicily. Geomorphology 49: 25–43. 17–18. Fiol L., Fornós J., Ginés A., 1992: El Rillenkarren: un Dzulynski S., Gil E., Rudnicki J., 1988: Experiments on tipus particular de Biocarst? Primeres dades. Endins kluftkarren and related lapis forms. Zeitschrift für 17, 18: 43–49. Geomorphologie 32, 1: 1–16. Fiol L., Fornós J., Ginés A., 1996: Effects of biokarstic Eavis A. J., 1980: Caves of Mulu ‘80. Royal Geographi- processes on the development of solutional ril- cal Society, London, 52 p. lenkarren in limestone rocks. Earth Surface Proc- Eckert M., 1898: Die Karren oder Schratten. Peter- esses and Landforms 21, 5: 447–452. manns Mitteilungen, Gotha, 69–71. Focke J. W., 1977: The effect of a potentially reef-build- Eckert M., 1902: Das Gottesackerplateau. Ein Kar- ing Vermetid-coralline algal community on an erod- renfeld im Allgäu. Studien zur Lösung des Karren- ing limestone coast, Curaçao, Netherlands Anti- problems. Wissenschaftliche Ergänzungshefte zur lles. Proceedings of the 3rd International Coral Reef Zeitschrift des Deutschen und Österreichischen Symposium 1: 239–245. Alpenvereins 1, 3: 1–108. Focke J. W., 1978a: Limestone cliff morphology on Elert G., 2001: Diameter of a Raindrop. http:/ hyper- Curaçao (Netherlands Antilles) with special atten- textbook.com/facts/2001/IgorVolynets.shtml. tion to the origin of notches and vermetid/cor- Emery K. O., 1946: Marine Solution Basins. The Jour- alline algal surf benches (“cornices”, “trottoirs”). nal of Geology 54: 209–228. Zeitschrift für Geomorphologie 22: 329–349. Emery K. O., 1962: Marine geology of Guam. U.S. Geo- Focke J. W., 1978b: Limestone cliff morphology and logical Survey Professional Paper 403-B, 76 p. organism distribution on Curaçao (Netherlands Emmett W. W., 1970: The hydraulics of overland flow Antilles). Leidse Geologische Mededelingen 51, 1: on hillslopes. U.S. Geological Survey Professional 131–150. Paper 662-A, 45 p. Folk R. L., Roberts H. H., Moore C. H., 1973: Black Escobar F., 1980: Mapa Geológico de Chile. 1:1.000.000 phytokarst from Hell, Cayman Islands, British West (feuille sud). Servicio Nacional de Geología y Min- Indies. The Geological Society of America Bulletin ería, Departamento de Geología General, Santiago. 84: 2351–2360. Ewers R. O., 1966: Bedding plane anastomoses and their Ford D. C., 1980: Threshold and limit effects in karst relation to cavern passages. Bulletin of the National geomorphology. In: Coates D. L., Vitek J. D. (Eds.), Speleological Society 28, 3: 133–141. Thresholds in Geomorphology. Allen Unwin, Ewers R. O., 1972: A model for the development of sub- London, 345–362. surface drainage routes along bedding planes. A Ford D. C., 1996: Features of karren development on 540 KRF•2 • OK.indd 540 15.12.2009 11:02:33 References dolostones in Canada. In: Fornós J. J., Ginés A. lorca: Unidad Calizas de Santanyí (Complejo Termi- (Eds.), Karren Landforms. Universitat de les Illes nal). In: El Terciario de las Islas Baleares. Guía de las Balears, Palma de Mallorca, 161–162. excursiones del 10th Congreso Nacional de Sedimen- Ford D. C., 1997: Preface. In: Song L., Waltham T., Cao tología, Menorca. Universidad de Palma de Mal- N., Wang F. (Eds.), Stone Forest: A Treasure of Natu- lorca, 177-206. ral Heritage. Proceedings of the International Sym- Forsythe R., 1981: Geological investigations of pre-Late posium for Lunan Shilin to Apply for World Natural Jurassic terranes in the Southernmost Andes. Ph. D. Heritage Status. China Environmental Science Press, Thesis, Columbia University, New York, 298 p. Beijing, 1–4. Forsythe R., Mpodozis C., 1983: Geología del basa- Ford D. C., Lundberg J., 1987: A review of dissolutional mento pre-jurásico superior en el archipiélago rills in limestone and other soluble rocks. Catena Madre de Dios, Magallanes, Chile. Servicio Nacional Supplement 8: 119–140. de Geología y Minería, Boletin No. 39, 63 p. Ford D. C., Salomon J. N., Williams P. W., 1996: Les Forti F., 1972: Le “vaschette di corrosione”. Atti e Mem- Forêts de Pierre ou Stone forests de Lunan (Yunnan, orie della Commissione Grotte “Eugenio Boegan” Chine). Karstologia 28: 25–40. 11: 37–65. Ford D. C, Salomon J. N., Williams P. W., 1997: The Forti P., 1996: Erosion rate, crystal size and exokarst Lunan Stone forest as a potential world herit- microforms. In: Fornós J. J., Ginés A. (Eds.), Karren age site. In: Song L., Waltham T., Cao N., Wang F. Landforms. Universitat de les Illes Balears, Palma de (Eds.), Stone Forest: A Treasure of Natural Herit- Mallorca, 261–276. age. Proceedings of the International Symposium Forti P., Cucchi F., Ayub S., 2001: Le “marmitte di cor- for Lunan Shilin to Apply for World Natural Herit- rosione” della Grotta Perolas (San Paolo, Brasile). Le age Status. China Environmental Science Press, Bei- Grotte d’Italia 5, 2: 15–24. jing, 107–123. Forti P., Sauro U., 1996: Gypsum karst of Italy. In: Ford D. C., Williams P. W., 1989: Karst Geomorphol- Klimchouk A., Lowe D., Cooper A., Sauro U. (Eds.), ogy and Hydrology. Unwin Hyman, London, 601 p. Gypsum Karst of the World. International Journal of Ford D. C., Williams P. W., 2007: Karst Hydrogeology Speleology 25, 3–4: 239–250. and Geomorphology. Willey, Chichester, 576 p. Fras F. J., 1835: Vollständige Topographie der Karl- Ford T. D., 1978: Chillagoe – a tower karst in decay. städter Militärgrenze. Državna štamparija, Zagreb. Trans British Cave Research Association 5: 61–84. Friederich H., 1980: The water chemistry of the unsatu- Ford T. D., 1984: Palaeokarsts in Britain. Cave Science rated zone in the Melinau Limestone. Geographical 11, 4: 246-264. Journal 146, 2: 246–258. Fornós J. J., 1999: Karst collapse phenomena in the Fry E. J., 1927: The mechanical action of crustaceous Upper Miocene of Mallorca (Balearic Islands, West- lichens on substrata of shale, schist, gneiss, lime- ern Mediterranean). Acta Geologica Hungarica 42: stone and obsidian. Annals of Botany 41: 437–460. 237–250. Fry J. C., Swineford A., 1947: Solution features on creta- Fornós J. J., Gelabert B., 1995: Lithology and tectonics ceous sandstone in central Kansas. American Jour- of the Majorcan karst. In: Ginés A., Ginés J. (Eds.), nal of Science 245: 366–379. Karst and caves in Mallorca. Endins 20, Mono- Gabrovšek F., 2000: Evolution of early karst aquifers: graphs of the Natural History Society of the Balearic from simple principles to complex models. Inštitut Islands 3: 27–43. za raziskovanje krasa ZRC SAZU, Postojna, 150 p. Fornós J. J., Gelabert B., Ginés A., Ginés J., Tuccimei Gale S. J., Drysdale R. N., Scherrer N. C., Fischer M. J., P., Vesica P., 2002: Phreatic overgrowths on spele- 1997: The Lost City of North-west Queensland: a test othems: a useful tool in structural geology in littoral of the model of giant grikeland development in semi- karstic landscapes. The example of eastern Mallorca arid karst. Australian Geographer 28, 1: 107–115. (Balearic Islands). Geodinamica Acta 15: 113–125. Gams I., 1971: Podtalne kraške oblike [Subsoil karst Fornós J. J., Ginés A., 1996: Karren Landforms. Univer- forms]. Geografski vestnik 43: 27–45. sitat de les Illes Balears, Palma de Mallorca, 450 p. Gams I., 1974: Kras. Slovenska matica, Ljubljana, 358 p. Fornós J. J., Gómez-Pujol L., 2002: Estudio integrado Gams I., 1976: Forms of subsoil karst. In: Proceedings del lapiaz costero de Mallorca dentro del Proyecto of the 6th International Congress of Speleology, Olo- ESPED: metodología y resultados preliminares. mouc. Academia, Praha, 169–179. Boletín de la Sociedad Española de Espeleología y Gams I., 1985: Mednarodne primerjalne meritve povr- Ciencias del Karst 3: 106–107. šinske korozije s pomočjo standardnih apneniških Fornós J. J., Pomar L., 1983: Mioceno Superior de Mal- tablet [International comparative measurements of 541 KRF•2 • OK.indd 541 15.12.2009 11:02:33 Karst Rock Features • Karren Sculpturing surface solution by means of standard limestone George T. N., Johnson G. A. L., Mitchell M., Prentice tablets]. In: Grafenauer S., Pleničar M., Drobne K. J. E. I., Ramsbottom W. H. C., Sevastopulo G. D., (Eds.), Zbornik Ivana Rakovca. Slovenska akademija Wilson R. B., 1976: A correlation of Dinantian Rocks znanosti in umetnosti, 361–386. in the British isles. Geological Society of London Gams I., 1990: Depth of rillenkarren as a measure of Special Report 7, 87 p. deforestation age. Proceedings of the International Gèze B., 1973: Lexique des termes français de spéléolo- Conference on Anthropogenic Impact and Envi- gie physique et de karstologie. Annales de Spéléolo- ronmental Changes in Karst. Studia Carsologica 1: gie 28, 1: 1–20. 29–36. Gil M. V., 1992: Quantitative analysis of solution flutes Gams I., 1997: Climatic and lithological influence on in La Safor karst, Valencia, Spain. Zeitschrift für the cave depth development. Acta Carsologica 26, 2: Geomorphologie, Supplementband 85: 89–100. 321–336. Gillespie P. A., Walsh J. J., Watterson J., Bonson C.G., Gams I., Kunaver J., Radinja D., 1973: Slovenska kraška Manzocchi T., 2001: Scaling relationships of joint terminologija [Slovene Karst Terminology] (in Slov- and vein arrays from The Burren, Co. Clare, Ireland. ene, English, French and German). Kraška termi- Journal of Structural Geology 23: 183–201. nologija jugoslovanskih narodov, knjiga I. Oddelek Gillieson D. S., Nethery J., Webb J., Godwin M., O’Keefe za geografijo Filozofske fakultete, Ljubljana, 76 p. C., Atkinson A., 2003: Field Guidebook: Limestone Gardner R. A. M., 1983: Aeolianite. In: Goudie A. S., and Lava. 15th Australasian Conference on Cave and Pye K. (Eds.), Chemical sediments and geomorphol- Karst Management, Chillagoe and Undara, Queens- ogy. Academic Press, London, 265–300. land. Australasian Cave and Karst Management Gatani M. G., Laureti L., Madonia P., Pisano A., 1989: Association, 43 p. Caratteri e distribuzione delle microforme carsiche Gillieson D. S., Smith D. I., Greenaway M., Ellaway nel territorio di S. Ninfa (Trapani). In: Agnesi V., M., 1991: Flood history of the limestone ranges in Macaluso T. (Eds.), I gessi di Santa Ninfa: studio the Kimberley region, Western Australia. Applied multidisciplinare di un’area carsica. Memorie Isti- Geography 11: 105–123. tuto Italiano di Speleologia 2, 3: 49–58. Gillieson D. S., Spate A. P., 1998: Karst and caves in Gavrilović D., 1964: Kamenice – mali korozivni oblici Australia and New Guinea. In: Yuan D., Liu Z. (Eds.), na krečnjaku. Glasnik Srpskog geografskog društva Global Karst Correlation. Science Press, Beijing PRC 44, 1: 53–60. and VSP Press, Utrecht, 229–256 p. Gavrilović D., 1968: Kamenice, kleine Korrosionfor- Ginés A., 1990: Utilización de las morfologías de lapiaz men im Kalkstein. 4th International Congress of Spe- como geoindicadores ecológicos en la Serra de Tra- leology 3: 127–133. muntana (Mallorca). Endins 16: 27–39. Gehrmann C. K., Krumbein W. E., Petersen K., 1992: Ginés A., 1993: Morfologies exocàrstiques. In: Alcover J. Endolithic lichen and the corrosion of carbonate A., Ballesteros E., Fornós J. J. (Eds.), Història Natural rocks, a study of biopitting. International Journal of de l’Arxipèlag de Cabrera. Monografia de la Societat Mycology and Lichenology 5: 37–48. d’Història Natural de les Balears 2: 153–160. Gelabert B., 1998: La estructura geológica de la mitad Ginés A., 1995: Deforestation and karren development occidental de la isla de Mallorca. Instituto Tecnológ- in Majorca, Spain. Acta Universitas Szegediensis, ico GeoMinero de España, Colección Memorias, Acta Geographica 34: 25–32. Madrid, 129 p. Ginés A., 1996a: An environmental approach to the Gelabert B., 2003: La estructura geológica de Menorca: typology of karren landform assemblages in a Med- las zonas de Tramuntana y Migjorn. In: Rosselló V., iterranean mid-mountain karst: the Serra de Tra- Fornós J. J., Gómez-Pujol L. (Eds.), Introducción a la muntana, Mallorca, Spain. In: Fornós J. J., Ginés A. Geografía Física de Menorca. Monografia de la Soci- (Eds.), Karren Landforms. Universitat de les Illes etat d’Història Natural de Balears 10: 39–48. Balears, Palma de Mallorca, 163–176. Gelabert B., Fornós J. J., Pardo J. E., Rosselló V. M., Ginés A., 1996b: Quantitative data as a base for the Segura F., 2005: Structurally controlled drainage morphometrical definition of rillenkarren features basin development in the south of Menorca (West- found on limestones. In: Fornós J. J., Ginés A. (Eds.), ern Mediterranean, Spain). Geomorphology 65: Karren Landforms. Universitat de les Illes Balears, 139–155. Palma de Mallorca, 177–191. Gelabert B., Sàbat F., Rodríguez-Perea A., 1992: A struc- Ginés A., 1998a: Dades morfomètriques sobre les tural outline of the Serra de Tramuntana of Mal- estries de lapiaz dels Alps calcaris suïssos i la seva lorca (Balearic Islands). Tectonophysics 203: 167–183. 542 KRF•2 • OK.indd 542 15.12.2009 11:02:33 References comparació amb les estries de la Serra de Tramun- reference to North West England. Unpublished Ph. tana. Endins 22: 109–118. D. Thesis, Oxford University, Oxford. Ginés A., 1998b: L’exocarst de la serra de Tramuntana Goldie H. S., 1986: Human influences on landforms: de Mallorca. In: Fornós J. J. (Ed.), Aspectes geològics the case of limestone pavements. In: Paterson K., de les Balears, Universitat de les Illes Balears, Palma Sweeting M. M. (Eds.), New Directions in Karst. de Mallorca, 361–389. Geobooks, Norwich, 515–540. Ginés A., 1999a: Morfología kárstica y vegetación en la Goldie H. S., 1993: The legal protection of limestone Serra de Tramuntana. Una aproximación ecológica. pavements in Great Britain. Environmental Geology Unpublished Ph. D. Thesis, Departament de Biologia 21: 160–166. Ambiental, Universitat de les Illes Balears, Palma de Goldie H. S., 1995: Major protected sites of limestone Mallorca, 581 p. pavement in Great Britain. Acta Geographica Szege- Ginés A., 1999b: Agriculture, grazing and land use diensis, Special Issue T 34: 61–92. changes at the Serra de Tramuntana karstic moun- Goldie H. S., 1996: The limestone pavements of Great tains. International Journal of Speleology 28 B, 1–4: Asby Scar, Cumbria, U.K. Environmental Geology 5–14. 28, 3: 128–136. Ginés A., 2004: Karren. In: Gunn J. (Ed.), Encylopedia Goldie H. S., 2005: Erratic judgments: re-evaluating of Caves and Karst Science. Fitzroy Dearborn, New solutional erosion rates of limestones using erratic- York, London, 470–473. pedestal sites, including Norber, Yorkshire. Area 37, Ginés A., Ginés J., 1995: The exokarstic landforms of 4: 433–442. Mallorca Island. In: Ginés A., Ginés J. (Eds.), Karst Goldie H. S., 2006: Mature intermediate-scale surface and caves in Mallorca. Endins 20, Monographs of karst landforms in NW England and their relations the Natural History Society of the Balearic Islands to glacial erosion. In: Kiss A., Mezősi G., Sümeghy 3: 59–70. Z. (Eds.), Táj, Környezet és Társadalom [Landscape, Ginés J., 1990: El modelat càrstic de sa Mitjania Environment and Society]. Studies in Honour of (Escorca, Mallorca). Endins 16: 17–20. Prof. Ilona Bárány-Kevei, Szeged, 225–238. Ginés J., 1995: Mallorca’s endokarst: the speleogenetic Goldie H. S., Cox N. J., 2000: Comparative morphom- mechanisms. Endins 20: 71–86. etry of limestone pavements in Switzerland, Britain Ginés J., 2000: El karst litoral en el Levante de Mallorca: and Ireland. Zeitschrift für Geomorphologie, Sup- una aproximación al conocimiento de su morfogé- plementband 122: 85–112. nesis y cronología. Unpublished Ph. D. Thesis, Uni- Golubić S., Friedmann E., Schneider J., 1981: The litho- versitat de les Illes Balears, Palma de Mallorca, 595 p. biontic niche, with special reference to microorgan- Ginés J., Fornós J.J., 2004: Caracterització del carst del isms. Journal of Sedimentary Petrology 51: 475-478. Migjorn. La seva contribució al modelat del territori. Golubić S., Perkins R. D., Lukas K. J., 1975: Boring Monografia de la Societat d’Història Natural de les micro-organisms and microborings in carbonate Balears, 11: 259–274. substrates. In: Frey R. W. (Ed.), The Study of Trace Ginsburg R. N., 1953: Intertidal erosion on the Florida Fossils. Springer-Verlag, Berlin, 229–259. Keys. Bulletin of Marine Science Gulf and Carib- Gómez-Pujol L., Balaguer P., Baldo M., Fornós J. J., bean 3, 1: 55–69. Pons G. X., Villanueva G., 2002: Patrones y tasas de Gladysz K., 1987: Karren on the Quatsino Limestone, erosión de Melaraphe neritoides (Linneo, 1875) en el Vancouver Island. B. Sc. Thesis, McMaster Univer- litoral rocoso de Mallorca: resultados preliminares. sity, Hamilton, Ontario. In: Pérez-González A., Vegas J., Machado J. (Eds.), Glew J. R., 1976: The simulation of rillenkarren. Unpub- Aportaciones a la Geomorfología de España en el lished M. Sc. Thesis, McMaster University, Hamil- Inicio del Tercer Milenio. ITGE, Madrid, 351–354. ton, Ontario, 116 p. Gómez-Pujol L., Fornós J. J., 2001: Les microformes de Glew J. R., Ford D. C., 1980: A simulation study of the meteorització del litoral calcari de Mallorca: aproxi- development of rillenkarren. Earth Surface Proc- mació a la seva sistematització. Endins 24: 169–185. esses 5, 1: 25–36. Gómez-Pujol L., Fornós J. J., 2004a: Forma, procesos Godwin M., 1988: Broken River Karst: a speleological y zonación en el lapiaz-karren-litoral del Sur de field guide, North Queensland. Unpublished report Menorca, 1: aproximación morfométrica. In: Benito by Chillagoe Caving Club and Queensland National G., Díaz A. (Eds.), Contribuciones Recientes sobre Parks and Wildlife Service, Cairns, 134 p. Geomorfología. SEG, CSIC, Madrid, 347–355. Goldie H. S., 1976: Limestone pavements: with special Gómez-Pujol L., Fornós J. J., 2004b: Les microformes litorals del Migjorn. In: Fornós J. J., Obrador A., 543 KRF•2 • OK.indd 543 15.12.2009 11:02:33 Karst Rock Features • Karren Sculpturing Rosselló V. M. (Eds.), Història Natural del Migjorn Grimes K. G., 1990: Fanning River Karst Area: Notes de Menorca. Bolletí de la Societat d’Història Natural on the geology and geomorphology. Queensland de Balears 11: 245–258. Department of Mines, Record 1990/7 (unpublished), González L. A., Ruiz H. M., Taggart B. E., Budd A. F., 31 p. Monell V., 1997: Geology of Isla de Mona, Puerto Grimes K. G., 1994: The South-East Karst Province Rico. In: Vacher H. L., Quinn T. (Eds.), Geology and of South Australia. Environmental Geology 23: Hydrogeology of Carbonate Islands. Developments 134–148. in Sedimentology 54, Elsevier Science BV, 327–358. Grimes K. G., 2001: Karst features of Christmas Island Goodchild M. F., Ford D. C., 1971: Analysis of scallop (Indian Ocean). Helictite 37, 2: 41–58. pattern by simulation under controlled conditions. Grimes K. G., 2002: Syngenetic and Eogenetic Karst: an Journal of Geology 79, 1: 52–62. Australian Viewpoint. In: Gabrovšek F. (Ed.), Evo- Gosden M. S., 1968: Peat deposits of Scar Close, Ingle- lution of karst: from prekarst to cessation. Karst borough, Yorkshire. Journal of Ecology 56, 1: Research Institute ZRC SAZU, Postojna, 407–414. 345–353. Grimes K. G., 2004: Solution Pipes or Petrified Forests? Goudie A. S., 1974: Further experimental investigation Drifting sands and drifting opinions! The Victorian of rock weathering by salt and other mechanical Naturalist 121, 1: 14–22. processes. Zeitschrift für Geomorphologie, Supple- Grimes K. G., 2006: Syngenetic Karst in Australia: a mentband 21: 1–12. review. Helictite 39, 2: 27–38. Goudie A. S., 1990: The Landforms of England and Grimes K. G., 2007: Microkarren in Australia – a Wales. Blackwell, Oxford, 394 p. request for information. Helictite 40, 1: 21–23. Goudie A. S., 1999: Experimental salt weathering of Grimes K. G., Spate A. P., 2008: Laterite Karst (Andy- limestones in relation to rock properties. Earth Sur- sez No 53). Australasian Cave and Karst Manage- face Processes and Landforms 24: 715–724. ment Association Journal 73: 49–52. Goudie A. S., Bull P. A., Magee A. W., 1989: Lithological Gross M. R., Fischer M. P., Engelder T., Greenfield R. control of rillenkarren development in the Napier J., 1995: Factors controlling joint spacing in inter- Ranges, Western Australia. Zeitschrift für Geomor- bedded sedimentary rocks: integrating numerical phologie, Supplementband 75: 95–114. models with field observations from the Monterey Goudie A.S., Migon P., Allison R.J., Rosser N., 2002: Formation, USA. In: Ameen M. S. (Ed.), Fractog- Sandstone geomorphology of the Al-Quwayra area raphy: fracture topography as a tool in fracture of south Jordan. Zeitschrift für Geomorphologie 46: mechanics and stress analysis. Geological Society of 365–390. London, Special Publication 92: 215–233. Goudie A. S., Viles H. A., 1997: Salt Weathering Haz- Guijarro J. A., 1986: Contribución a la bioclimatología ards. Wiley, Chichester, 241 p. de las Baleares. Unpublished Ph. D. Thesis, Universi- Goudie A. S., Viles H. A., 1999: The frequency and tat de les Illes Balears, Palma de Mallorca. magnitude concept in relation to rock weathering. Guijarro J. A., 1995: Bioclimatic aspects of karst in Mal- Zeitschrift für Geomorphologie, Supplementband lorca. In: Ginés A., Ginés J. (Eds.), Karst and caves 115: 175–189. in Mallorca. Endins 20, Monographs of the Natural Goudie A. S., Viles H. A., Allison R. J., Day M. J., Liv- History Society of the Balearic Islands 3: 17–26. ingstone I. P., Bull P. A., 1990: The geomorphology Guilcher A., 1952: Formes et processus d’érosion sur of the Napier Range, Western Australia. Transac- les récifs coralliens du nord du banc Farsan (Mer tions of the Institute of British Geographers 15, 3: Rouge). Revue de Géomorphologie Dynamique 6: 308–322. 261–274. Granger D. E., Fabel D., Palmer A. N., 2001: Pliocene- Guilcher A., 1953: Essai sur la zonation et la distribu- Pleistocene incision of the Green River, Kentucky, tion des formes littorales de dissolution du calcaire. determined from radioactive decay of cosmogenic Annales de Géographie 62: 161–179. 26Al and 10Be in Mammoth Cave sediments. Geo- Guilcher A., 1958: Coastal Corrosion Forms in Lime- logical Society of America Bulletin 113, 7: 825–836. stones Around the Bay of Biscay. The Scottish Geo- Grimes K. G., 1978: Colless Creek and Lawn Hill graphical Magazine 74, 3: 137–149. Gorge, Down Under. Newsletter of the University of Gunn J., 2004: Encyclopedia of Caves and Karst Sci- Queensland Speleological Society 17, 2: 45–49. ence. Fitzroy Dearborn, New York, 902 p. Grimes K. G., 1988: The Barkly Karst Region, North- Gustavson T. C., Holliday V. T., Hovorka S. D., 1995: west Queensland. 17th Biennial Conference of the Origin and Development of Playa Basins, Source of Australian Speleological Federation, Sydney, 16–24. Recharge to the Ogallala Aquifer, Southern High 544 KRF•2 • OK.indd 544 15.12.2009 11:02:33 References Plains, Texas and New Mexico. Report of Investiga- propagation across layer interfaces in sedimentary tions 229, Bureau of Economic Geology. University rocks. Journal of Structural Geology 13, 8: 897–911. of Texas at Austin, 45 p. Herwitz S. R., 1993: Stemflow influences on the forma- Habič P., 1980: S poti po kitajskem krasu. Summary: tion of solution pipes in Bermuda eolianite. Geo- From the way to the Chinese karst. Geografski vest- morphology 6: 253–271. nik 52: 107–122. Higgins C. G., 1980: Nips, notches and the solution Habsburg-Lorena L. A., 1884–1891: Die Balearen in of coastal limestone: an overview of the problem Wort und Bild Geschildert. Brockhaus, Leipzig. with examples from Greece. Estuarine and coastal Hacquet B., 1785: Physikalisch-politische Reise aus marine science 10: 15–30. den Dinarischen, durch die Julischen, Carnischen, Hinsinger P., Fernandes Barros O. N., Benedetti M. F., Rhätischen in die Nordischen Alpen, im Jahre 1781 Noack Y., Callot G., 2001: Plant-induced weathering und 1783 unternommen. Böhme, Leipzig, 377–380. of a basaltic rock: Experimental evidence. Geochim- Hall R. L., Calder I. R., 1993: Drop size modification by ica et Cosmochimica Acta 65: 137–152. forest canopies: measurements using a disdrometer. Hobléa F., Jaillet S., Maire R., 2001: Érosion et ruissel- Journal of Geophysical Research 90: 465–470. lement sur karst nu en context subpolaire sur les iles Halley R. B., Muhs D. R., Shinn E. A., Dill R. F., Kind- calcaires de Patagonie. Karstologia 38: 13–18. inger J. L., 1991: [abstract], A +1.5-m reef terrace Hodgkin E. P., 1964: Rate of erosion of intertidal lime- in the southern Exuma Islands, Bahamas. GSA stone. Zeitschrift für Geomorphologie 8, 4: 385–392. Abstracts with Programs 23, 1: 40. Hodgkin E. P., 1970: Geomorphology and biological Halpern M., 1973: Regional geochronology of Chile, erosion of limestone coasts in Malaysia. Bulletin of south of 50° latitude. Geological Society of America, the Geological Society of Malaysia 3: 27–51. Bulletin 84, 7: 2407–2422. Holbye U., 1989: Bowl-karren in the littoral karst of Hamilton S. E., Holland E., Mott K., Spate A., 1989: Nord-Arnøy, Norway. Cave Science 16: 19–25. Cutta Cutta Caves Nature Park – Draft Plan of Man- Hope G. S., 1976: Vegetation. In: Hope G. S., Peterson J. agement. Unpublished report by the Australasian A, Radok U., Allison I. (Eds.), Equatorial Glaciers of Cave Management Association for the Conservation New Guinea. Balkema, Rotterdam, 113–172. Commission of the Northern Territory. Hopley D., Mackay M. G., 1978: An investigation of Hantke R., 1961: Tektonik der helvetischen Kalkalpen Morphological Zonation of Beach Rock erosional zwischen Obwalden und dem St. Galler Rheintal. features. Earth Surface Processes 3: 363–377. Vierteljahresschrift Naturforschende Gesellschaft Horbury A. D., 1987: The relative roles of tectonism and Zürich 106, 1: 210 p. eustacy in the deposition of the Urswick Limestone Hantoon P. W., 1997: Definition and characteristics of in South Cumbria and North Lancashire. In: Arthur- Stone forest epikarst aquifers in south China. Pro- ton R. S., Gutteridge P., Nolan S. C. (Eds.), The Role ceedings of the 12th International Congress of Spe- of Tectonism in Devonian and Carboniferous Sedi- leology 1, 8: 311–314. mentation in the British Isles. Yorkshire Geological Hartlieb M., 1999: Der oberirdische Karst im Gebiet des Society, Occasional Publication 6: 153–169. Dachsteins. http:/ www.sbg.ac.at/geog/studenten/roth/ Horoi V., 2001: L’influence de la Géologie sur la Karsti- se/manfred/manfred.html. fication: Étude Comparative Entre le Massif Obar- Haserodt K., 1965: Untersuchungen zur Höhen- und sia Closani – Piatra Mare (Roumanie) et le Massif Altersgliederung der Karstformen in den Nördli- d’Arbas (France). Thèse, Laboratoire souterrain de chen Kalkalpen. Münchner Geographische Hefte 27, Moulis (CNRS) and Institut Emile Racovitza de 114 p. Bucarest, 174 p. Hauer F., 1867–1871: Geologische Übersichtskarte Humbert S. E., Driese S. G., 2001: Phased develop- Österreichisch-Ungarischen Monarchie. Geologi- ment of a subaerial paleokarst plane in upper Pen- schen Reichsanstalt, Wien. nington Formation limestones (upper Mississipian) Heim A., 1878: Über die Karrenfelder. Jahrbuch des and associated paleokarst features. Poster presented Schweizer Alpenclub 13: 421–433. during G. S. A. Big South Fork National Recreation Heinemann U., Kaaden K., Pfeffer K. H., 1977: Neue Area, Tennessee, Annual Meeting, M. S. November Aspekte zum Phänomen der Rillenkarren. Abhan- 5–8, 2001. M. Sc. Thesis, 143 p. dlungen zur Karst- und Höhlenkunde, Reihe A, 15: Hume W. F., 1925: Geology of Egypt. Government 56–80. Press, Cairo, 408 p. Helgeson D. E., Aydin A., 1991: Characteristics of joint Hunt C. O., 1996: Tafoni (pseudokarst) features in the Maltese Islands. Cave and Karst Science 23, 2: 57–62. 545 KRF•2 • OK.indd 545 15.12.2009 11:02:33 Karst Rock Features • Karren Sculpturing Hutchinson D. W., 1996: Runnels, rinnenkarren and Cooleman Plains, New South Wales, Australia, and mäanderkarren: form, classification and relation- their implications. Abhandlungen zur Karst- und ships. In: Fornós J. J., Ginés A. (Eds.), Karren Land- Höhlenkunde, Reihe A, 15: 26–38. forms. Universitat de les Illes Balears, Palma de Mal- Jennings J. N., 1981: Morphoclimatic control: a tale lorca, 209–223. of piss in the wind or a case of the baby out with Ivanović A., Sakač K., Marković S., Sokač B., Šušnjar the bath water? Proceedings of the 8th International M., Nikler L., Šušnjara A., 1973: Osnovna geološka Congress of Speleology 1, Bowling Green, 61–73. karta SFRJ 1:100.000, List Obrovac L 33–140. Institut Jennings J. N., 1982: Karst of northeastern Queensland za geološka istraživanja, Zagreb, Savezni geološki reconsidered. Tower Karst, Chillagoe Caving Club, zavod, Beograd. Occasional Paper 4: 13–52. Jacobson G. J., Hill P. J., Ghassemi F., 1997: Geology Jennings J. N., 1983: The disregarded karst of the arid and Hydrogeology of Nauru Island. In: Vacher H. L., and semiarid domain. Karstologia 1: 61–73. Quinn T., (Eds.), Geology and Hydrogeology of Car- Jennings J. N., 1985: Karst Geomorphology. Basil Black- bonate Islands. Developments in Sedimentology 54, well, New York, 293 p. Elsevier Science HV, 707–742. Jennings J. N., Bik M.A., 1962: Karst morphology in Jaillet S., Hobléa F., and l’équipe Ultima Patagonia, Australian New Guinea. Nature 194: 1036–1038. 2000: Une morphologie originale liée au vent: les Jennings J. N., Sweeting M. M., 1963: The limestone “fusées” ou “crêtes éoliennes de lapiaz” de l’île Madre ranges of the Fitzroy Basin, Western Australia, a de Dios (Archipel Ultima Esperanza – Patagonie – tropical semi-arid karst. Bonner Geographische Chili). Actes 10ème rencontre d’octobre du Spéléo- Abhandlungen 32: 1–60. Club de Paris. Jeschke A. A., Vosbeck K., Dreybrodt W., 2001: Sur- Jakucs L., 1977: Morphogenetics of karst regions. face controlled dissolution rates of gypsum in aque- Akadémiai Kiadó, Budapest, 284 p. ous solutions exhibit nonlinear dissolution kinetics. Jakucs L., Keveiné-Bárány I., Mezősi G., 1983: A Geochimica et Cosmochimica Acta 65: 27–34. karsztkorrózió korszerű értelmezése. Földrajzi Johansson M., Migon P., Olvmo M., 2001: Development Közlemények 31, 107, 3–4: 213–217. of joint controlled rock basins in Bohus granite, SW Jakucs P., 1956: Karrosodás és növényzet. Földrajzi Sweden. Geomorphology 40: 145–161. Közlemények 3: 241–249. Jones B., 1989: The role of microorganisms in phy- James J., 1997: A comparison of the Stone Forest of tokarst development on dolostones and limestones, Lunan with pinnacle karsts of the World. In: Song Grand Cayman, British West Indies. Canadian Jour- L., Waltham T., Cao N., Wang F. (Eds.), Stone Forest: nal of Earth Sciences 26: 2204–2213. A Treasure of Natural Heritage. Proceedings of the Jones B., Hunter I. G., 1992: Very large boulders on the International Symposium for Lunan Shilin to Apply coast of Grand Cayman: the effects of giant waves for World Natural Heritage Status. China Environ- on rocky coastlines. Journal of Coastal Research 8, mental Science Press, Beijing, 22–29. 4: 763–774. Jennings J. N., 1967: Some karst areas of Australia. In: Jones B., Pemberton S. G., 1987: The role of fungi in the Jennings J. N., Mabbutt J. A. (Eds.), Landform Stud- diagenetic alteration of spar calcite. Canadian Jour- ies from Australia and New Guinea. Australian nal of Earth Sciences 24: 903–914. National University Press, Canberra, 256–292. Jones R. J., 1965: Aspects of the biological weathering of Jennings J. N., 1968: Syngenetic Karst in Australia. In: limestone pavements. Proceedings of the Geologists’ Williams P. W., Jennings J. N. (Eds.), Contributions Association 76: 421–433. to the study of karst. Australian National University, Jones W. K., Culver D. C., Herman J. S., 2004: Epikarst. Department of Geography Publication G 5: 41–110. Karst Waters Institute, Charles Town, Special Publi- Jennings J. N., 1969: Karst of the seasonally humid cation 9, 160 p. tropics in Australia. In: Štelcl O. (Ed.), Problems of Karp D., 2002: Land degradation associated with sink- the Karst Denudation. Supplement for the 5th Inter- hole development in the Katherine region. Resource national Speleological Congress, Institute of Geog- Assessment Branch, Northern Territory Department raphy, Brno, 149–158. of Infrastructure, Planning and Environment, Tech- Jennings J. N., 1971: Karst. Australian National Univer- nical Report No. 11/2002, 53 p. sity Press, Canberra, 252 p. Kashima N., Urushibara-Yoshino K., 1996: Karren Jennings J. N., 1973: Karst. The M. I. T. Press, Cam- development. Solutional erosion measurements bridge, Massachusetts, London, 253 p. by the limestone-tablet method in Shikoku Island, Jennings J. N., 1977: Limestone tablets experiments at South-West Japan. In: Fornós J.J., Ginés A. (Eds.), 546 KRF•2 • OK.indd 546 15.12.2009 11:02:33 References Karren Landforms. Universitat de les Illes Balears, D., Palmer A., Dreybrodt W. (Eds.), Speleogenesis: Palma de Mallorca, 65–73. Evolution of Karst Aquifers. National Speleological Kaufmann G., Dreybrodt W., 2007: Calcite dissolu- Society, Huntsville, Alabama, 45–53. tion kinetics in the system CaCO -H O-CO at high Klimchouk A. B., 2000c: The formation of epikarst and 3 2 2 undersaturation. Geochimica et Cosmochimica its role in vadose speleogenesis. In: Klimchouk A. Acta 71, 6: 1398–1410. B., Ford D. C., Palmer A. N., Dreybrodt W. (Eds.), Kaye C. A., 1959: Shoreline features and Quaternary Speleogenesis: Evolution of Karst Aquifers. National shoreline changes, Puerto Rico. U. S. Geological Speleological Society, Huntsville, Alabama, 91–99. Survey Professional Paper 317-B: 49–140. Klimchouk A. B., Ford D. C., Palmer A. N., Dreybrodt Kelletat D., 1980: Formenschatz und Prozessgefüge W., 2000: Speleogenesis: Evolution of Karst Aquifers. des “Biokarstes” an der Küste von Nordost-Mal- National Speleological Society, Huntsville, Alabama, lorca (Cala Guya). Berliner Geographische Studien 527 p. 7: 99–113. Knez M., 1998: Lithologic Properties of the Three Kelletat D., 1985: Bio-destruktive und bio-konstruktive Lunan Stone Forests (Shilin, Naigu and Lao Hei Formelemente an den spanischen Mittelmeerküsten. Gin). In: Chen X., Gabrovšek F., Huang C., Jin Y., Geodynamik 6: 1–20. Knez M., Kogovšek J., Liu H., Petrič M., Mihevc Kelletat D., 1991: The 1550 B.P. tectonic event in the A., Otoničar B., Shi M., Slabe T., Šebela S., Wu W., Eastern Mediterranean as a basis for assessing the Zhang S., Zupan Hajna N., 1998: South China Karst intensity of shore processes. Zeitschrift für Geomor- I. Založba ZRC, Ljubljana, 30–43. phologie, Supplementband 81: 181–194. Knez M., Slabe T., 2001a: Oblika in skalni relief stebrov Kerr A., 1983: Rock temperature measurements from v Naigu kamnitem gozdu (JZ Kitajska). Acta Carso- southeast Morocco, and their implications for logica 30, 1: 13–24. weathering and weathering forms. Unpublished B. Knez M., Slabe T., 2001b: The lithology, shape and rock Sc. Thesis, Queen’s University, Belfast, 136 p. relief of the pillars in the Pu Chao Chun stone forest Kershaw S., Guo L., 2001: Marine notches in coastal (Lunan stone forests, SW China). Acta Carsologica cliffs: indicators of relative sea-level change, Pera- 30, 2: 129–139. chora Peninsula, Central Greece. Marine Geology Knez M., Slabe T., 2002: Lithologic and morphologi- 179: 213–228. cal properties and rock relief of the Lunan stone for- Kinahan G. H., Nolan J., 1870: Explanation to accom- ests. In: Gabrovšek F. (Ed.), Evolution of karst: from pany map sheet 95 of the Geological Survey of Ire- prekarst to cessation. Karst Research Institute ZRC land. Memoirs of the Geological Survey of Ireland, SAZU, Postojna, 259–266. Dublin. Kogovšek J., Kranjc A., Slabe T., Šebela S., 1999: South Kindler P., Hearty P. J., 1995: Pre-Sangamonian eolian- China Karst 1999, Preliminary research in Yunnan. ites in the Bahamas? New evidence from Eleuthera Acta Carsologica 28, 2: 225–240. Island. Marine Geology 127: 73–86. Komac B., 2001: The karst springs of the Kanin massif Kinnell P. I., 1987: Rainfall Energy in Eastern Australia: [Kraški izviri pod Kaninskim pogorjem]. Geograf- Intensity-Kinetic Energy Relationships. Australian ski zbornik 41: 7–43. Journal of Soil Research 25: 547–553. Kondic L., 2003: Instabilities in Gravity Driven Flow of Kirchhofer W., 1982: Mittlere Jahrestemperaturen. Kli- Thin Fluid Films. Society for Industrial and Applied maatlas der Schweiz, sheet 6.1. Mathematics Review 45, 1: 95–115. Kirchhofer W., Sevruk B., 1991: Mittlere jährliche kor- Körner C., 1999: Alpine Plant Life – Flunctional Plant rigierte Niederschlagshöhen 1951–1980. Hydrolo- Ecology of High Mountain Ecosystems. Springer- gischer Atlas der Schweiz, sheet 2.2. Verlag, Berlin, Heidelberg, 343 p. Kirigin B., 1967: Klimatske karakteristike Sjevernog Kunaver J., 1973a: O razvoju slovenske terminologije Velebita. Zbornik radova X kongresa klimatologa za mikroreliefne kraške oblike [On the development Jugoslavije, Kopaonik, Beograd, 189–206. of the Slovene terminology of the Karst microrelief]. Klimchouk A. B., 2000a: Dissolution and conversion of In: Gams I., Kunaver J., Radinja D. (Eds.), Slovenska gypsum and anhydrite. In: Klimchouk A. B., Ford D. kraška terminologija [Slovene Karst Terminology]. C., Palmer A. N., Dreybrodt W. (Eds.), Speleogene- Oddelek za geografijo Filozofske fakultete, Ljubljana, sis: Evolution of Karst Aquifers. National Speleologi- 68–76. cal Society, Huntsvile, Alabama, 160-168. Kunaver J., 1973b: The High Mountainous Karst of Klimchouk A. B., 2000b: Types of karst and evolution the Julian Alps in the System of Alpine Karst. IGU of hydrogeologic sections. In: Klimchouk A. B., Ford 547 KRF•2 • OK.indd 547 15.12.2009 11:02:33 Karst Rock Features • Karren Sculpturing European regional conference, Symposium on karst- Ley R. G., 1977: The Influence of Lithology on Marine morphogenesis, Papers, Budapest, 209–225. Karren. Abhandlungen zur Karst- und Höh- Kunaver J., 1976: Kotlič – a Specific Depression Form of lenkunde, Reihe A, 15: 81–100. Subnival Alpine Karst. Proceedings of the 6th Inter- Ley R.G., 1979: The Development of Marine Karren national Congress of Speleology, Academia, Praha, Along the Bristol Channel Coastline. Zeitschrift für 223–230. Geomorphologie, Supplementband 32: 75–89. Kunaver J., 1983: Geomorphology of the Kanin Moun- Ley R. G., 1980: The pinnacles of Gunung Api. Geo- tains with special regard to the glaciokarst. Geograf- graphical Journal 146, 1: 14–21. ski zbornik 22: 197–346. Liang F., Song L., Tang T., 2003: Microbial produc- Kunaver J., 1984: The high mountains karst in the Slov- tion of CO in red soil in Stone Forest National Park. 2 ene Alps. Geographica Yugoslavica 5, Ljubljana, Journal of Geographical Sciences 13, 2: 250–256. 15–22. Liang F., Song L., Wang F., Zheng B., Zhang L., 2000: Kunaver J., 1985: Relief. In: ˝Fabjan I. (ed.), Triglavski The case study of subsoil solution features and soil narodni park. Vodnik, Bled, 29–56. CO concentration in Stone Forest Region, Lunan, 2 Kunaver J., 1991: Corrosion terraces as geoecological Yunnan, China. Carsologica Sinica 19: 187–193. response to postglacial development of glaciokarstic Liechti P., Roe R. W., Haile N. S., 1960: The geology rock surface. In: Proceedings of the International of Sarawak, Brunei and the western part of North conference on environmental changes in karst areas. Borneo British Territory of Borneo. Bulletin of the 13 Quaderni del dipartimento di geografia, Univer- Geological Survey Department 3 (2 volumes), 360 p. sità di Padova, 325–331. Lin J., 1997: Genesis of Lunan Stone Forest and its Geo- Kunaver J., 1998: Karst of the Kanin-mountain. morphological Significance. In: Song L., Waltham T., Alpine Karst, Guide-booklet for the excursions and Cao N., Wang F. (Eds.), Stone Forest: A Treasure of abstracts of the papers. 6th International Karstologi- Natural Heritage. Proceedings of the International cal School, Classical karst. Speleological Association Symposium for Lunan Shilin to Apply for World of Slovenia and Karst Research Institute ZRC SAZU, Natural Heritage Status. China Environmental Sci- Postojna, 12–16. ence Press, Beijing, 30–33. Kurtz H. D., Netoff D. I., 2001: Stabilization of friable Linton D. L., 1963: The Forms of Glacial Erosion. sandstone surfaces in a desiccating, wind-abraded Transactions of the Institute of British Geographers environment of south-central Utah by rock surface 33: 1–28. microorganisms. Journal of Arid Environments 48: Linton D. L., 1964: The origin of the Pennine tors – an 89–100. essay in analysis. Zeitschrift für Geomorphologie 8: Kuščer I., Kodre A., 1994: Matematika v fiziki in teh- 5–24. niki. Društvo matematikov, fizikov in astronomov Löffler E., 1977: Geomorphology of Papua New Guinea. Slovenije, Ljubljana, 194–199. Scientific and Industrial Research Organization in Laczay L., 1982: A folyószabályozás tervezésének mor- association with Australian National University fológiai alapjai. Vízügyi Közl 25: 235–254. Press, Canberra, 195 p. Langridge D., 1971: Limestone pavement patterns on Longman M. W., Brownlee D. N., 1980: Character- the island of Inishmore, County Galway. Irish Geog- istics of karst topography, Palawan, Philippines. raphy 3: 282–293. Zeitschrift für Geomorphologie 24: 299–317. Laudermilk J. D., Woodford A. O., 1932: Concern- Louis H., 1968: Allgemeine Geomorphologie. Walter ing rillensteine. American Journal of Science 223: de Gruyter, Berlin, 522 p. 135–154. Lowry D. C., 1973: Origin of the Pinnacles. Australian Lauritzen S. E., 1981: Simulation of Rock Pendants Speleological Federation Newsletter 62: 7–8. – Small Scale Experiments on Plaster Models. Pro- Lozano R., 1884: Anotaciones físicas y geológicas de ceedings of the 8th International Congress of Speleol- la Isla de Mallorca. Excma. Diputación Provincial ogy, Georgia, 407–411. de Baleares. Imprenta de la Casa de Misericordia, Lauritzen S. E., Karp D., 1993: Speleological assessment Palma de Mallorca, 68 p. of karst aquifers developed within the Tindall Lime- Lundberg J., 1974: The karren of the littoral zone, stone, Katherine N.T., Power and Water Authority, Burren District, Co. Clare, Ireland. Unpublished B. NT. Report 63/1993 (unpublished). Sc. Thesis, Trinity College, Dublin. Laverty M., 1980: Water chemistry in the Gunung Mulu Lundberg J., 1977a: The geomorphology of Chillagoe National Park, including problems of interpretation limestones: variations with lithology. Unpublished and use. Geographical Journal 146, 2: 232–245. 548 KRF•2 • OK.indd 548 15.12.2009 11:02:34 References M. Sc. Thesis, Australian National University, Can- limestone on the coasts of some islands in the Red berra, 175 p. Sea. Geographical Journal 75: 27–34. Lundberg J., 1977b: An analysis of the form of rillenkar- Maire R., l990: La haute montagne calcaire. Karsts, cav- ren from the tower karst of Chillagoe, Northern ités, remplissages, quaternaire, paleoclimats. Karst- Queensland, Australia. Proceedings of the 7th Inter- ologia-Mémoires 3, Association Française de Kars- national Congress of Speleology, Sheffield, 294–296. tologie et Fédération Française de Spéléologie, La Lundberg J., 1977c: Karren of the littoral zone, Burren Ravoire, 731 p. District, Co. Clare, Ireland. Proceedings of the 7th Maire R., 1999: Les glaciers de marbre de Patagonie. Un International Speleological Congress, Sheffield, karst subpolaire océanique de la zone australe. Kars- 291–293. tologia 33: 25–40. Lundberg J., 2004: Coastal Karst. In: Gunn J. (Ed.), Maire R., 2004: Patagonia marble karst (Chile). In: Encyclopedia of Cave and Karst Science. Fitzroy Gunn J. (Ed.), Encyclopedia of Cave and Karst Sci- Dearborn, New York, London, 231–233. ence. Fitzroy Dearborn, New York, 57–577. Lundberg J., Lauritzen S. E., 2002: The search for an Maire R., Pernette J. F., Fage L. H., 1999: Les "glaciers arctic coastal karren model in Norway and Spitzber- de marbre" de Patagonie, Chili; un karst subpolaire gen. In: Hewitt K., Byrne M. L., English M., Young G. océanique de la zone australe [The "marble glaciers" (Eds.), Landscapes of Transition: Landform assem- of the Chilean Patagonia; example of a subpolar oce- blages and transformations in cold regions. Kluwer anic karst in the austral zone]. Karstologia 33: 25–40. Academic Publishers, The GeoJournal Library 68: Maire R., Zhang S., Song S., 1991: Génese des karsts 183–203. subtropicaux de Chine du sud (Guizhou, Sichuan, Lundberg J., Taggart B. E., 1995: Dissolution pipes in Hubei). Grottes et karsts tropicaux de Chine Méridi- northern Puerto Rico: an exhumed paleokarst. Car- onale, Karstologia mémoires 4: 162–186. bonates and Evaporites 10, 2: 171–183. Malis C. P., 1997: Littoral karren along the western Ma X., 1936: Preliminary investigation on the Lunan shore of Newfoundland. Unpublished M. Sc. Thesis, stone forest, Yunnan, from the geomorphological McMaster University, Hamilton, Ontario, 310 p. view. Theoretical Review 1, 1. Malis C. P., Ford D. C., 1995: Littoral karren along the Macaluso T., Madonia G., Palmeri A., Sauro U., 2001: western shore of Newfoundland. Geological Society Atlante dei karren nelle evaporiti della Sicilia [Atlas of America, Abstracts with Programs 27, 6: A 9–56. of the karren in the evaporitic rocks of Sicily]. Quad- Malott C. A., 1945: Significant features of the Indiana erni del Museo Geologico “G.G. Gemmellaro” 5, karst. Indiana Academy of Science Proceedings 54: Dipartimento di Geologia e Geodesia, Università 8–24. degli Studi di Palermo, 143 p. Mamužić P., Milan A., Korolija B., Borović I., Majcen Macaluso T., Madonia G., Sauro U., 2003: Le forme di Ž., 1969: Osnovna geološka karta SFRJ 1:100.000, soluzione nei gessi. In: Madonia G., Forti P. (Eds.), List Rab L 33–114. Institut za geološka istraživanja, Le aree carsiche gessose d’Italia. Memorie Istituto Zagreb, Savezni geološki zavod, Beograd. Italiano di Speleologia 2, 14: 55–64. Mangin A., 1997: Some features of the Stone forest of Macaluso T., Sauro U., 1996a: The karren in evapor- Lunan, Yunnan, China. In: Song L., Waltham T., itic rocks: a proposal of classification. In: Fornós J. J., Cao N., Wang F. (Eds.), Stone Forest: A Treasure of Ginés A. (Eds.), Karren Landforms. Universitat de Natural Heritage. Proceedings of the International les Illes Balears, Palma de Mallorca, 277–293. Symposium for Lunan Shilin to Apply for World Macaluso T., Sauro U., 1996b: Weathering crust and Natural Heritage Status. China Environmental Sci- karren on exposed gypsum surfaces. International ence Press, Beijing, 68–70. Journal of Speleology 25, 3–4: 115–126. Mariko S., Bekku Y., Koizumi H., 1994: Efflux of carbon Macaluso T., Sauro U., 1998a: Aspects of weathering dioxide from snow covered forest floors. Ecological and landforms evolution on gypsum slopes and Research 9: 343–350. ridges of Sicily. Geografia Fisica Dinamica Quater- Marker M. E., 1976a: Note on some South African naria, Supplement 3, 4: 91–99. pseudokarst. Bollettino Sociedad Venezolana Espe- Macaluso T., Sauro U., 1998b: I Karren nei gessi di Ver- leologia 7: 5–12. zino. In: Ferrini G. (Ed.), L’area carsica delle Vigne Marker M. E., 1976b: A Geomorphological Assessment di Verzino. Memorie Istituto Italiano di Speleologia of the Chillagoe Karst Belt, Queensland. Helictite 14, 2, 10: 35–45. 1: 31–49. Macfayden W. A., 1930: The undercutting of coral reef Marker M. E., 1985: Factors controlling micro-solu- 549 KRF•2 • OK.indd 549 15.12.2009 11:02:34 Karst Rock Features • Karren Sculpturing tional karren on carbonate rocks of the Griqualand Mihevc A., 2001: Speleogeneza Divaškega krasa. Zbirka West sequence. Cave Science 12, 2: 61–65. ZRC 27, Založba ZRC, Ljubljana, 180 p. Marsico A., Selleri G., Mastronuzzi G., Sanso P., Walsh Mikulas R., 1999: The protective effect of lichens and N., 2003: Cryptokarst: a case-study of the Quater- the origin of modern bulge-like traces on weakly nary landforms of southern Apulia (southern Italy). lithified sandstones (Hilbre Islands, Wirral Penin- Acta Carsologica 32, 2: 147–159. sula, Great Britain). Ichnos 6: 261–268. Martel E. A., 1921: Nouveau traité des eaux souter- Miotke F. D., 1974: Carbon dioxide and the soil atmos- raines. Doin, Paris, 838 p. phere. Abhandlungen zur Karst- und Höhlenkunde, Martini I. P., 1978: Tafoni weathering with examples Reihe A, 9: 1-49. from Tuscany, Italy. Zeitschrift für Geomorphologie Moe D., Johannessen P. J., 1980: Formation of cavities 22: 44–67. in calcareous rocks in the littoral zone of northern Matičec D., Fuček L., Oštrić N., Sokač B., Velić I., Norway. Sarsia 65: 227–232. Vlahović I., Ženko T., 1999: Geološka građa Velebita Monroe W. H., 1970: A Glossary of Karst Terminol- duž trase cestovnog tunela “Sv. Rok”. Field-trip ogy. Geological Survey Water-Supply Paper, 1899-K: guidebook. Geotrip, Hrvatsko geološko društvo, 1–26. Zagreb, 1–20. Moseley F., 1972: A tectonic history of northwest Eng- Matsukura Y., Matsuoka N., 1991: Rates of tafoni land. Journal of the Geological Society of London weathering on uplifted shore platforms in Nojima- 128: 561–598. zaki, Boso Peninsula, Japan. Earth Surface Processes Moses C. A., 2003: Observations on coastal biokarst, and Landforms 16: 51-56. Hells Gate, Lord Howe Island, Australia. Zeitschrift Matsukura Y., Matsuoka N., 1996: The effect of rock für Geomorphologie 47: 83–100. properties on rates of tafoni growth in coastal envi- Moses C. A., Smith B. J., 1993: A note on the role of Col- ronments. Zeitschrift für Geomorphologie, Supple- lema auriforma in solution basin development on a mentband 106: 57–72. Carboniferous limestone substrate. Earth Surface Mazari R. K., 1988: Himalayan karst – karren in Kash- Processes and Landforms 18: 363–368. mir. Zeitschrift für Geomorphologie 32, 2: 163–178. Moses C. A., Smith B. J., 1994: Limestone weathering Mazzanti R., Parea G. C., 1979: Erosione della “panch- in the supra-tidal zones: an example from Mallorca. ina” sui littorali di Livorno e di Rosignano. Bolletino In: Robinson D. A., Williams R. B. G. (Eds.), Rock Società Geologica Italiana 96: 457–489. Weathering and Landform Evolution. John Wiley McCabe A. M., 1987: Quaternary deposits and glacial and Sons, Chichester, 433–451. stratigraphy in Ireland. Quaternary Science Reviews Moses C., Spate A. P., Smith D. I., Greenaway M. A., 6: 259–299. 1995: Limestone weathering in eastern Australia. McKee E. D., Ward W. C., 1983: Eolian Environment. Part 2: Surface micromorphology study. Earth Sur- In: Scholle P. A., Bebout D. E., Moore C. H. (Eds.), face Processes and Landforms 20: 501–514. Carbonate depositional environments. American Moses C. A., Viles H. A., 1996: Nanoscale morpholo- Association of Petroleum Geologists Memoir 33: gies and their role in the development of karren. 131–170. In: Fornós J. J., Ginés A. (Eds.), Karren Landforms. McNamara K. J., 1995: Pinnacles (revised edition). Universitat de les Illes Balears, Palma de Mallorca, Western Australian Museum, Perth, 24 p. 85–96. Melim L. A., Masaferro J. L., 1997: Geology of the Baha- Mottershead D. N., 1996a: A study of solution flutes mas: Subsurface geology of the Bahamas banks. In: (Rillenkarren) at Lluc, Mallorca. Zeitschrift für Geo- Vacher H. L., Quinn T. M. (Eds.), Geology and Hydr- morphologie, Supplementband 103: 215–241. ogeology of Carbonate Islands. Elsevier, Amsterdam, Mottershead D. N., 1996b: Some morphological prop- 161–182. erties of solutional flutes (Rillenkarren) at Lluc, Mal- Mensching H., 1955: Karst und Terra-Rossa auf Mal- lorca. In: Fornós J. J., Ginés A. (Eds.), Karren Land- lorca. Erdkunde 9: 188–196. forms. Universitat de les Illes Balears, Palma de Meyerhoff A. A., Hatten C. W., 1974: Bahamas salient Mallorca, 225–238. of North America: Tectonic framework, stratigra- Mottershead D. N., Lucas G. R., 2000: The role of lichens phy, and petroleum potential. American Associa- in inhibiting erosion of a soluble rock. Lichenologist tion of Petroleum Geologists, Bulletin 58: 1201–1239. 2000: 601–609. Migon P., Dach W., 1995: Rillenkarren on granite out- Mottershead D. N., Lucas G. R., 2001: Field testing of crops, S.W. Poland, age and significance. Geografiska Glew and Ford's model of solution flute evolution. 77 A, 1–2: 1–9. Earth Surface Processes and Landforms 26: 839–846. 550 KRF•2 • OK.indd 550 15.12.2009 11:02:34 References Mottershead D. N., Lucas G. R., 2004: The role of in Bermuda and measurements of the boring rate of mechanical and biotic processes in solution flute the sponge Cliona lampa. Limnology and Oceanog- development. In: Smith B. J., Turkington A. V. (Eds.), raphy 11: 92–108. Stone decay: its causes and control. Donhead Pub- Newell N. D., 1961: Recent terraces of tropical lime- lishing, London, 273–291. stone shores. Zeitschrift für Geomorphologie, Sup- Mottershead D. N., Moses C. A., Lucas G. R., 2000: plementband 3: 87–106. Lithological control of solution flute form: a com- Newell N. D., Imbrie J., 1955: Biogeological recon- parative study. Zeitschrift für Geomorphologie 44, naissance in the Bimini area, Great Bahama Bank. 4: 491–512. Transactions of the New York Academy of Sciences Mottershead D. N., Pye K., 1994: Tafoni on coastal 18: 3–14. slopes, South Devon, U. K. Earth Surface Processes Newson M. D., 1970: Studies in chemical and mechani- and Landforms 19: 543–563. cal erosion by streams in limestone terrains. Ph. D. Mottershead D. N., Viles H. A., 2004: Experimental Thesis, University of Bristol, Bristol. studies of rock weathering by plant roots: Updat- Nicod J., 1976: Corrosion de type crypto-karstique ing the work of Julius Sachs (1832-1897). In: Mitch- dans les karst mediterrannéens. In: Karst Processes ell D. J., Searle D. E. (Eds.), Stone Deterioration in and Relevant Landforms. Department of Geography, Polluted Urban Environments. Scientific Publish- Philosophical Faculty, Ljubljana, 171–179. ing, Plymouth, 61–72. Ogorelec B., 1996: Dachstein Limestone from Krn in Murphy P. J., Lord T., 2003: Victoria Cave, Yorkshire, Julian Alps (Slovenia). Geologija 39: 133–157. UK: new thoughts on an old site. Cave and Karst Sci- Ollier C., 1984: Weathering. Longman, London, New ence 30, 2: 83–88. York, 270 p. Murray A. N., Love W. W., 1930: Action of organic Oluić M., Grandić S., Haček M., Hanich M., 1972: Tek- acids upon limestone. The American Association of tonska građa vanjskih Dinarida Jugoslavije. Nafta 23, Petroleum Geologists Bulletin 13: 1467–1475. 1–2: 3–16. Mustoe G. E., 1983: Cavernous weathering in the Capi- Osmaston H. A., 1980: Patterns in trees, rivers and tol Reef Desert, Utah. Earth Surface Processes and rocks in the Mulu Park, Sarawak. Geographical Landforms 8: 517–526. Journal 146, 1: 33–50. Myers T. G., 2002: Modelling laminar sheet flow over Osmaston H. A., Sweeting M. M., 1982: Geomorphol- rough surfaces. Water Resources Research 38, 11: ogy. In: Jermy A. C., Kavanagh K. P. (Eds.), Gunung 1230. Mulu National Park, Sarawak. Sarawak Museum Mylroie J. E., Carew J. L., 1995: Karst development on Journal (Special Issue 2) 30, 51: 75–93. carbonate islands. In: Budd D. A., Harris P. M., Saller Pace M. C., Mylroie J. E., Carew J. L., 1993: Investiga- A. (Eds.), Unconformities and Porosity in Carbonate tion and review of dissolution features on San Sal- Strata. American Association Petroleum Geologists, vador Island, Bahamas. In: White B. (Ed.), Proceed- Memoir 63: 55–76. ings of the 6th Symposium on the Geology of the Mylroie J. E., Jenson J. W., Taboroši D., Jocson J. M. U., Bahamas. Bahamian Field Station, Port Charlotte, Vann D. T., Wexel C., 2001: Karst Features of Guam Florida, 109–123. in Terms of a General Model of Carbonate Island Palmer A. N., 1975: The origin of maze caves. National Karst. Journal of Cave and Karst Studies 63, 1: 9–22. Speleological Society Bulletin 57: 56–76. Mylroie J. E., Vacher H. L., 1999: A conceptual view of Palmer A. N., 1991: Origin and morphology of lime- carbonate island karst. In: Palmer A. N., Palmer M. stone caves. Geological Society of America Bulletin V., Sasowsky I. D. (Eds.), Karst Waters. Karst Waters 103: 1–21. Institute Special Publication 5: 48–57. Palmer M., Fornós J. J., Balaguer P., Gómez-Pujol L., Narr W., Suppe J., 1991: Joint spacing in sedimentary Pons G. X., Villanueva G., 2003: Spatial and seasonal rocks. Journal of Structural Geology 3, 9: 1037–1048. variability of the macro-invertebrate community of Naylor L. A., Viles H. A., 2002: A new technique for a rocky coast in Mallorca (Balearic Islands): implica- evaluating short-term rates of coastal bioerosion tions for bioerosion. Hydrobiologia 501: 13–21. and bioprotection. Geomorphology 47: 31–44. Palmer M., Palmer A. N., 1975: Landscape development Nelhans G., Svensson P., 1997: Some small-scale weath- in the Mitchell Plain of southern Indiana. Zeitschrift ering structures in the Sidi Bouzid area, central für Geomorphologie 19, 1: 1–39. Tunisia. The virtual Geomorphology, http:/ user. Panuska B. C., Mylroie J. E., Carew J. L., 1999: Paleo- tninet.se/~bgb354w/home/Tunisien.htm. magnetic evidence for three Pleistocene paleosols Neumann A. C., 1966: Observations on coastal erosion on San Salvador Island, Bahamas. In: Curran H. A., 551 KRF•2 • OK.indd 551 15.12.2009 11:02:34 Karst Rock Features • Karren Sculpturing Mylroie J. E. (Eds.), Proceedings of the 9th Sympo- to measure water film thickness and the study of sium on the Geology of the Bahamas, Bahamian flow hydraulics and dissolutional characteristics on Field Station, San Salvador Island, Bahamas, 93–100. plaster of Paris rillenkarren channels. B. Sc. Thesis, Papida S., Murphy W., May E., 2000: Enhancement of School of Geographical Sciences, University of Bris- physical weathering of building stones by micro- tol, Bristol. bial populations. International Biodeterioration and Peyrot-Clausade M., Le Campion T., Harmelin M., Biodegradation 46: 305–317. Romano J. C., Chazottes V., Pari N., Le Campion J. Paradise T., 1997: Disparate sandstone weather- L., 1995: La bioérosion dans le cycle des carbonates: ing beneath lichens, Red Mountain, Arizona. essais de quantification des processus en Polyné- Geografiska Annaler 79: 177–184. sie française. Bulletin de la Société Géologique de Parry J. T., 1960: Limestone pavements of North West France 1: 85–94. England. Canadian Geographer 15: 41–21. Phelps H. O., 1975: Shallow laminar flows over rough Paterson K., Sweeting M. M. (Eds.), 1986: New Direc- granular surfaces. Journal of the Hydraulics Divi- tions in Karst. Geobooks, Norwich, 613 p. sion 101, 3: 367–384. Pearson L. M., 1982: Chillagoe karst solution and Pigott C. D., 1962: Soil formation and development on weathering. Tower Karst, Chillagoe Caving Club, the carboniferous limestone of Derbyshire. I. Parent Occasional Paper 4: 58–70. materials. Journal of Ecology 50: 145–156. Perica D., 1998: Geomorfologija krša Velebita. Doctoral Pigott C. D., 1970: Soil formation and development on Thesis, University of Zagreb, Zagreb, 251 p. the carboniferous limestone of Derbyshire. II. The Perica D., Kukić B., 1992: Karren on the South Velebit relation of soil development to vegetation on the Range. International Symposium “Geomorphology plateau near Coombs Dale. Journal of Ecology 58: and Sea”, Mali Lošinj. Zagreb, 153–157. 528–541. Perica D., Kukić B., Trajbar S., 1995: Egzokrške osobine Pirnat J., 2002: Speleology on Kanin. Soški razgovori Nacionalnog parka Paklenica. Paklenički zbornik 1, 1, Proceedings for Regional Studies of the History Starigrad-Paklenica, 65–69. Section of the Cultural Association Golobar, Bovec, Perica D., Marjanac T., Aničić B., Mrak I., Juračić 77–98. M., 2004: Small karst features (karren) of Dugi Playford P. E., 1980: Devonian “Great Barrier Reef” of Otok island and Kornati Archipelago coastal karst Canning Basin, Western Australia. American Asso- (Croatia). Acta Carsologica 33: 117–130. ciation of Petroleum Geologists, Bulletin 64, 6: Perica D., Marjanac T., Mrak I., 1999: Vrste grižina i 814–840. njihov nastanak na području Velebita. Acta Geo- Playford P. E., 2002: Palaeokarst, pseudokarst, and graphica Croatica 34: 31–58. sequence stratigraphy in Devonian reef complexes Perica D., Orešić D., 1997: Prilog poznavanju klimat- of the Canning Basin, Western Australia. In: Keep skih obilježja Velebita. Acta Geographica Croatica M., Moss S. J. (Eds.), The Sedimentary Basins of 32: 45–68. Western Australia 3. Petroleum Exploration Society Perica D., Orešić D., 1999: Klimatska obilježja Velebita i of Australia, Symposium, Perth, W. A., 763–793. njihov utjecaj na oblikovanje reljefa. Senjski zbornik Pluhar A., Ford D. C., 1970: Dolomite karren of the 26: 1–50. Niagara Escarpment, Ontario, Canada. Zeitschrift Perna G., Sauro U., 1978: Atlante delle microforme di für Geomorphologie 14, 4: 392–410. dissoluzione carsica superficiale del Trentino e del Plummer L. N., Wigley T. M. L., 1976: The dissolution Veneto. Memorie del Museo Tridentino di Scienze of calcite in CO -saturated solutions at 25°C and 1 2 Naturali 22, Trento, 176 p. atmosphere total pressure. Geochimica et Cosmo- Perna G., Sauro U., 1981: Types morphologiques du chimica Acta 40: 191–202. Rosso Ammonitico dans le Trentino et le Veneto. Plummer L. N., Wigley T. M. L., Parkhurst D.L., 1978: In: Farinacci A., Elmi S. (Eds.), Rosso Ammonit- The kinetics of calcite dissolution in CO -water sys- 2 ico Symposium Proceedings. Tecnoscienza, Roma, tems at 5° to 60°C and 0.0 to 1.0 atm CO . American 2 541–546. Journal of Science 278: 179–216. Peter C., 2001: Deep into the land of extremes. Probing Poljak J., 1929a: Geomorfološki oblici krednih kršnika Chile’s wild coast. National Geographic, 2–19. Velebita. Vijesti Geološkog zavoda 3: 53–85. Peterson J. A., 1982: Limestone pedestals and denuda- Poljak J., 1929b: Planinarski vodič po Velebitu. Hrvat- tion estimates from mountain Jaya, Irian Jaya. Aus- sko planinarsko društvo, 277 p. tralian Geographer 15: 170–173. Pomar L., Esteban M., Llimona X. M., Fontarnau R., Petterson O., 2001: The development of a technique 1975: Acción de líquenes, algas y hongos en la telodi- 552 KRF•2 • OK.indd 552 15.12.2009 11:02:34 References agénesis de las rocas carbonatadas de la zona prelito- Riđanović J., 1966: Orjen. Radovi Geografskog insti- ral catalana. Instituto de Investigaciones Geológicas tuta u Zagrebu 5: 5–103. 30: 83–117. Robinson T., 1978: A question of age. Tower Karst, Pomar L., Obrador A., Westphal H., 2002: Sub-wase- Chillagoe Caving Club, Occasional Paper 2: 18–36. base cross-bedded grainstone on distally steepened Robinson T., 1982: Limestone solution experiment carbonate ramp, Upper Miocene, Menorca, Spain. at Chillagoe. Tower Karst, Chillagoe Caving Club, Sedimentology 49: 139–169. Occasional Paper 4: 71–76. Pomar L., Ward W. C., 1999: Reservoir-scale heteroge- Rodet J., 1992: La Craie et ses Karsts. Centre de Géo- neity in depositional packages and diagenetic pat- morphologie du Centre National de la Recherche terns on a reef-rimmed platform, Upper Miocene, Scientifique, Caen, 560 p. Mallorca, Spain. American Association of Petro- Rodríguez-Navarro C., Doehne E., Sebastián E., 1999: leum Geologists, Bulletin 83: 1579–1773. Origins of honeycomb weathering: The roles of Powell R. L., 1961: Caves of Indiana. Indiana Geologi- salt and wind. Bulletin of the Geological Society of cal Survey 8: 1–127. America 111: 1250–1255. Powell R. L., 1964: The origin of the Mitchell Plain. Indi- Rodríguez-Navarro C., Rodríguez-Gallego M., Ben ana Academy of Science, Proceedings 73: 177–182. Chekroum K., González-Muñoz M. T., 2003: Con- Priesnitz K., 1985: Nichtbiogene Kalklösung am Lough servation of ornamental stone by Myxococcus xan- Leane und am Mackross Lake (SE-Irland). Berliner thus-induced carbonate biomineralization. Applied Geographische Studien 16: 55–69. and Environmental Microbiology 69: 2182–2193. Pye K., Mottershead D. N., 1995: Honeycomb weather- Rodríguez-Perea A., 1984: El Mioceno de la Serra Nord ing of carboniferous sandstone in a sea wall of Wes- de Mallorca. Estratigrafía, Sedimentología e Impli- ton-Super-Mare, UK. Quarterly Journal of Engi- caciones Estructurales. Ph. D. Thesis, University of neering Geology 28: 333–347. Barcelona, Barcelona, 532 p. Qing S., 1977: Lunan stone forest – a spectacular karst Rogić V., 1958: Velebitska primorska padina. Radovi geomorphology. Geographical Knowledge 4. Geografskog Instituta u Zagrebu 2: 1–114. Quigley M., 1984: A study of micro-solution pits on the Rögner K., 1986: Temperature measurements of rock limestone pavements at Lough Mask, Co. Mayo. M. surfaces in hot deserts (Negev, Israel). International Sc. Thesis, Geography Department, University Col- Geomorphology 2: 1271–1286. lege, Dublin, 225 p. Rose J., 1982: The Melinau River and its terraces. Cave Quinif Y., 1973: Contribution à l’étude morphologique Science 9, 2: 113–127. des coupoles. Annales de Spéléologie 28, 4: 565–573. Rose J., 1984a: Contemporary river landforms and sed- Radofilao J., 1977: Bilan des explorations spéléologiques iments in an area of equatorial rain forest, Gunung dans l’Ankarana. Annales de lÚniversité de Mada- Mulu National Park, Sarawak. Transactions of the gascar, 14: 159–204. Institute of British Geographers 9: 345–363. Raistrick A., 1947: Malham and Malham Moor. Dales- Rose J., 1984b: Alluvial terraces of an equatorial river, man, Clapham, 122 p. Melinau drainage basin, Sarawak. Zeitschrift für Rasmusson G., 1959: Karstformen im Granit des Fich- Geomorphologie 28, 2: 155–177. telgebirges. Die Höhle 1: 1–4. Rose L., Vincent P. J., 1986a: Some aspects of the mor- Ray J.A., 1996: Fluvial features of the karst-plain ero- phometry of grikes; a mixture modeling approach. sion surface in the Mammoth Cave region. Proceed- In: Paterson K., Sweeting M. M. (Eds.), New Direc- ings of the 5th Annual Science Conference, Mam- tions in Karst. Geobooks, Norwich, 497–514. moth Cave National Park, Kentucky, 137–156. Rose L., Vincent P. J., 1986b: The kamenitzas of Gait Revelle R., Emery K. O., 1957: Chemical erosion of Barrows National Nature Reserve, north Lancashire, beach rock and exposed reef rock. U.S. Geological England. In: Paterson K., Sweeting M. M. (Eds.), New Survey Professional Paper 260-T: 699–706. Directions in Karst. Geobooks, Norwich, 473–496. Richmond I. A., 1933: Castle Folds, by Great Asby. Rosselló V. M., 1979: Algunas formas kársticas lito- Transactions of the Cumberland and Westmorland rales de Mallorca. In: Barceló B. (Ed.), Actas del VI Antiquarian and Archaeological Society, 233–237. Coloquio de Geografía. AGE, Palma de Mallorca, Riđanović J., 1964: Glacijalni relikti kao kriterij za kro- 115–121. nološko određivanje morfogeneze prevladavajućih Rossi G., 1974: Morphologie et évolution d’un karst oblika krša. Zbornik VII. kongresa geografa SFRJ, en milieu tropical: l’Ankarana. In: Phenomènes Zagreb, 279–290. Karstiques II. Mémoires et Documents 14, Centre National de la Recherche Scientifique, 279–298. 553 KRF•2 • OK.indd 553 15.12.2009 11:02:34 Karst Rock Features • Karren Sculpturing Rossi G., 1980: L’Extrême-Nord de Madagascar. Edisud, des Hauts Lessini Veronais. Actes du Symposium Aix-en-Provence, 440 p. sur les versants en Pays Méditerranéens – Aix en Rubić I., 1936: Mali otoci na obalnom reljefu istočnog Provence, Centre d’Études de Géografie et Écologie Jadrana. Geografski vestnik 12, 13: 3–53. de la Region Mediterraneenne 5: 43–47. Rudnicki J., 1960: Experimental work on flutes devel- Sauro U., 1976b: The geomorphological mapping of opment. Speleologia 2, 1: 17–30. “Karrenfelder” using very large scales: an exam- Rust D., Kershaw S., 2000: Holocene tectonic uplift pat- ple. In: Gams I. (Ed.), Karst Processes and Relevant terns in northeastern Sicily: evidence from marine Landforms. Department of Geography, Philosophi- nocthes in coastal outcrops. Marine Geology 167: cal Faculty, Ljubljana, 189–199. 105–126. Sauro U., 1977: Le città di roccia. Economia Trentina Sachs J., 1865: Handbuch der experimental Physiologie 26, 1: 73–80. der Pflanzen. Verlag von Wilhelm Engelmann, Leip- Sauro U., 1996: Geomorphological aspects of gypsum zig, 514 p. karst, with special emphasis on exposed karst. In: Safriel U. N., 1966: Recent vermetid formation on the Klimchouk A., Lowe D., Cooper A., Sauro U. (Eds.), Mediterranean shore of Israel. Proceedings of the Gypsum Karst of the World. International Journal of Malacological Society London 37: 27–34. Speleology 25, 3–4: 105–114. Sakač K., Benić J., Fuček L., Miknić M., Vlahović I., Sauro U., 2002: Quando in Lessinia c’era il grande gelo. 1993: Stratigraphic and tectonic position of pale- Quaderno Culturale – La Lessinia ieri oggi domani, ogene Jelar beds in the Outer Dinarides. Natura Verona, 85–94. Croatica 2, 1: 55–72. Sauro U., 2003: Aspetti evolutivi del paesaggio carsico Salles C., Poesen J., Sempere-Torres D., 2002: Kinetic nei gessi in Italia. In: Madonia G., Forti P. (Eds.), Le energy of rain and its functional relationship with aree carsiche gessose d’Italia. Memorie dell’Istituto intensity. Journal of Hydrology 257, 1–4: 256–270. Italiano di Speleologia 2, 14: 41–45. Salomon J. N., 1997: Comparaison entre les “Stone Schmalz R. F., Swanson F. J., 1969: Diurnal variations Forests” du Lunan (Yunnan–Chine) et les karsts a in the carbonate saturation of seawater. Journal of “tsingy” de Madagascar. In: Song L., Waltham T., Sedimentary Petrology 39: 255–267. Cao N., Wang F. (Eds.), Stone Forest: A Treasure of Schneider J., 1976: Biological and inorganic factors in Natural Heritage. Proceedings of the International the destruction of limestone coasts. Contributions Symposium for Lunan Shilin to Apply for World to Sedimentology 6: 1–112. Natural Heritage Status. China Environmental Sci- Schneider J., Le Campion Alsumard T., 1999: Con- ence Press, Beijing, 124–136. struction and destruction of carbonates by marine Salopek I., 1952: O gornjem permu Velike Paklenice and freshwater cyanobacteria. European Journal of u Velebitu. Rad Jugoslavenske akademije znanosti i Phycology 34: 417–426. umjetnosti 289: 5–26. Schneider J., Torunski H., 1983: Biokarst on limestone Sandler A., 1996: A Turonian subaerial event in Israel: coasts, morphogenesis and sediment production. karst, sandstone and pedogenesis. Geological Survey Marine Ecology 4: 45–63. of Israel, Bulletin 85, 56 p. Schoeneich Ph., Reynard E., Pierrehumbert G., 1998: Sauro U., 1973a: Forme di corrosione su rocce monto- Geomorphological mapping in the Swiss Alps and nate nella Val Lagarina meridionale. L’Universo 53, Prealps. Wiener Schriften zur Geographie und Kar- 2: 309–344. tographie, Band 11: 145–153. Sauro U., 1973b: Observations on some great solution Sellier D., 1997: Utilisation des mégalithes comme runnels with nested pans of the Venetian Prealps. marqueurs de la vitesse de l’érosion des granites Proceedings of the 6th International Congress of Spe- en milieu tempéré: enseignements apportés par les leology, Olomouc. Academia, Praha, 353–361. alignemets de Carnac (Morbihan). Zeitschrift für Sauro U., 1973c: Il paesaggio degli Alti Lessini. Studio Geomorphologie 41, 3: 319–356. geomorfologico. Memorie del Museo Civico di Storia Shannon C. H. C., 1970: Geology of the Mt. Etna area. Naturale di Verona, Memorie fuori serie 6, 161 p. In: Sprent J. K. (Ed.), Mount Etna Caves. Univer- Sauro U., 1974: Aspetti dell’evoluzione carsica legata a sity of Queensland Speleological Society, Brisbane, particolari condizioni litologiche e tettoniche negli 11–21. Alti Lessini. Bollettino Società Geologica Italiana 93: Simeonović R., 1921: O škrapama. Glasnik Srpskog 945–969. geografskog društva 5: 142–155. Sauro U., 1976a: Aspects et influences de la Formation Simms M. J., 1990: Phytokarst and photokarren in Ire- du Rosso Ammonitico sur l'évolution des versants land. Cave Science 17: 131–133. 554 KRF•2 • OK.indd 554 15.12.2009 11:02:34 References Simms M. J., 2002: The origin of enigmatic, tubular, debris in a hot desert environment. Geomorphology lake-shore karren: a mechanism for rapid disso- 1: 355–367. lution of limestone in carbonate saturated waters. Smith B., McAllister J. J., 1986: Observations on the Physical Geography 23, 1: 1–20. occurrence and origins of salt weathering phenom- Slabe T., 1992: Naravni in poskusni obnaplavinski ena near Lake Magadi, southern Kenya. Zeitschrift jamski skalni relief [Natural and experimental cave für Geomorphologie 30: 445–460. rocky relief on the contact of water and sediments]. Smith B., Moses C. A., Warke P. A., 1996: Modification Acta Carsologica 21: 7–34. of karren in arid environments: a case study from Slabe T., 1994: Dejavniki oblikovanja jamske skalne southern Tunisia. In: Fornós J. J., Ginés A. (Eds.), površine [The factors influencing on the forma- Karren Landforms. Universitat de les Illes Balears, tion of the cave rocky surface]. Acta Carsologica 23: Palma de Mallorca, 123–134. 369–398. Smith D. I., 1969: The solutional erosion of limestone Slabe T., 1995a: Cave Rocky Relief and its Speleoge- in an arctic morphogenetic region. In: Stell O. (Ed.), netical Significance. Zbirka ZRC 10, Založba ZRC, Problems of Karst Denudation. Proceedings of Ljubljana, 128 p. the 5th International Speleological Congress, Brno, Slabe T., 1995b: Experimental modelling of cave rocky Czechoslovakia. Studia Geographica 5. forms in Paris plaster. Atti e Memorie della Com- Sokač B., 1973: Geologija Velebita. Ph. D. Thesis, Uni- missione Grotte “Eugenio Boegan” 32: 65–83. versity of Zagreb, Zagreb, 151 p. Slabe T., 1996: Karst features in the motorway section Song L. H., 1986: Origination of stone forest in China. beetwen Čebulovica and Dane. Acta Carsologica 25: International Journal of Speleology 15, 1–4: 3–33. 221–240. Song L. H., Liu H., 1992: Control of geological struc- Slabe T., 1998: Rock relief of pillars in the Lunan Stone tures over development of cockpit karst in south Forest. In: Chen X., Gabrovšek F., Huang C., Jin Y., Yunnan, China. Tübingen Geographische Studien Knez M., Kogovšek J., Liu H., Petrič M., Mihevc 109: 57–70. A., Otoničar B., Shi M., Slabe T., Šebela S., Wu W., Song L., Li Y., 1997: Definition of Stone forest and its Zhang S., Zupan Hajna N., 1998: South China Karst evolution in Lunan County, Yunnan, China. In: I. Založba ZRC, Ljubljana, 51–67. Song L., Waltham A. C., Cao N., Wang F. (Eds.), Slabe T., 1999: Subcutaneous rock forms. Acta Carsolo- Stone Forest: A Treasure of Natural Heritage. Pro- gica 28, 2: 255–271. ceedings of the International Symposium for Lunan Slabe T., 2005: Two experimental modelings of karst Shilin to Apply for World Natural Heritage Status, rock relief in plaster: subcutaneous “rock teeth” and China Environmental Science Press, Beijing, 37–45. “rock peaks” exposed to rain. Zeitschrift für Geo- Song L., Liang F., 2001: Distribution of CO in Soil Air 2 morphologie 49, 1: 107–119. Affected by Vegetation in the Shilin National Park. Smart P. L., Dawans J. M., Whitaker F., 1988: Carbon- Acta Geologica Sinica 75, 3: 288–293. ate dissolution in a modern mixing zone. Nature Song L., Waltham A. C., Cao N., Wang F. (Eds.), 1997: 335: 811–813. Stone Forest: A Treasure of Natural Heritage. Pro- Smart P. L., Whitaker F. F., 1996: Development of ceedings of the International Symposium for Lunan karren landform assemblages – a case study from Shilin to Apply for World Natural Heritage Status, Son Marc, Mallorca. In: Fornós J. J., Ginés A. (Eds.), China Environmental Science Press, Beijing, 136 p. Karren Landforms. Universitat de les Illes Balears, Song L., Wang F., 1997: Lunan Shilin Landscape in Palma de Mallorca, 111–122. China. Proceedings of 12th International Congress of Smith B. J., 1978: The origin and geomorphic implica- Speleology, Symposium 8, 1: 433–435. tions of cliff foot recesses and tafoni on limestone Spate A. P., Little L., 1995: Is the conventional approach hamadas in the North West Sahara. Zeitschrift für to karst area management appropriate to tropical Geomorphologie 22: 21–43. Australia. In: Henderson K., Houshold I., Middleton Smith B. J., 1986: An integrated approach to the weath- G. (Eds.), Proceedings of the 11th Australasian Con- ering of limestone in an arid area and its role in ference on Cave and Karst Management. Australa- landscape evolution: a case study from south- sian Cave and Karst Management Association, Carl- east Morocco. In: Gardiner V. (Ed.), International ton South, 68–84. Geomorphology 2. John Wiley and Sons, London, Spencer T., 1985a: Marine erosion rates and coastal 637–657. morphology of reef limestones on Grand Cayman Smith B. J., 1988: Weathering of surficial limestone Island, West Indies. Coral Reefs 4: 59–70. Spencer T., 1985b: Weathering rates on a Caribbean 555 KRF•2 • OK.indd 555 15.12.2009 11:02:34 Karst Rock Features • Karren Sculpturing Reef limestone: results and implications. Marine and Co. Clare. Étude de Travaux de Méditerranée Geology 69: 195–201. 7: 210–209. Spencer T., 1988: Limestone coastal morphology: the Szunyogh G., 2000a: Differential Equations Describing biological contribution. Progress in Physical Geog- the Changes of Shape Caused by Karst Corrosion of raphy 12: 66–101. any Arbitrary Limestone Surface. Karsztfejlődés 4: Spencer T., Viles H., 2002: Bioconstruction, bioerosion 151–174. and disturbance on tropical coasts: coral reefs and Szunyogh G., 2000b: The Theoretical-Physical Study of rocky limestones. Geomorphology 48: 23–50. the Process of Karren Development. Karsztfejlődés Stanton W. I., 1984: Snail holes in Mendip limestones. 4: 125–150. Proceedings of the Bristol Naturalists’ Society 44: Szunyogh G., 2005: Theoretical investigation of the 15–18. duration of karstic denudation on bare, sloping Stenson R. E., Ford D. C., 1993: Rillenkarren on gypsum limestone surfaces. Acta Carsologica 34, 1: 9–23. in Nova Scotia. Géographie Physique et Quaternaire Szunyogh G., Lakotár K., Szigeti I., 1998: Nagy területet 47, 2: 239–243. lefedő karrvályúrendszer struktúrájának elemzése. Stoddart D. R., Taylor J. D., Farrow G. R., Fosberg F. R., Karsztfejlődés 2: 125–147. 1971: The geomorphology of Aldabra. In: Westoll T. Šebela S., Slabe T., Kogovšek J., Liu H., Pruner P., 2001: S., Stoddart D. R. (Eds.), A discussion on the results Baiyun Cave in Naigu Shilin, Yunnan Karst, China. of the Royal Society expedition to Aldabra, 1967– Acta Geologica Sinica 75, 3: 279–287. 1968. Philosophical Transactions of the Royal Soci- Šušnjar M., Bukovac J., Nikler L., Crnolatac I., Milan ety of London, London, 31-65. A., Šikić D., Grimani I., Vulić Z., Blašković I., Storm R., Smith D. I., 1991: The caves of Gregory 1970: Osnovna geološka karta SFRJ 1:100.000, List National Park, Northern Territory, Australia. Cave Crikvenica L 33–102. Institut za geološka istraživanja, Science 18, 2: 91–97. Zagreb, Savezni geološki zavod, Beograd. Svensson U., Dreybrodt W., 1992: Dissolution kinet- Taboroši D., Jenson J. W., Mylroie J. E., 2004: Karren ics of natural calcite minerals in CO -water-systems features in island karst: Guam, Mariana Islands. 2 approaching calcite equilibrium. Chemical Geology Zeitschrift für Geomorphologie 48, 3: 369–389. 100: 129–34. Tari Kovačić V., Mrinjek E., 1994: The role of Paleo- Sweet I. P., Mendum J. R., Bultitude R. J., Morgan C. gene clastics in the tectonic interpretation of north- M., 1974: The geology of the southern Victoria River ern Dalmatia (southern Croatia). Geologia Croatica Region, Northern Territory. Bureau of Mineral 47, 1: 127–138. Resources, Australia, Report 167, 143 p. Taylor M. P., Viles H. A., 2000: Improving the use of Sweeting M. M., 1955: Landforms in North-West microscopy in the study of weathering: sampling County Clare, Ireland. Transactions of the Institute issues. Zeitschrift für Geomorphologie, Supple- of British Geographers 21: 218–249. mentband 120: 145–158. Sweeting M. M., 1966: The weathering of limestones. Thomas T. M., 1970: The limestone pavements of the In: Dur G. H. (Ed.), Essays in Geomorphology. Hein- North Crop of the South Wales coalfield with spe- eman, London, 177–210. cial reference to solution rates and processes. Trans- Sweeting M. M., 1972: Karst landforms. MacMillan actions of the Institute of British Geographers 50: Press, London, 362 p. 87–105. Sweeting M. M., 1973: Karst Landforms. Columbia Thompson C. H., Bowman G. M., 1984: Subaerial denu- University Press, New York, 362 p. dation and weathering of vegetated coastal dunes in Sweeting M. M., 1980: The geomorphology of Mulu: an eastern Queensland. In: Thom B. G. (Ed.), Coastal introduction. Geographical Journal 146, 1: 1–7. Geomorphology in Australia. Academic Press, Sweeting M. M., 1995: Karst in China. Its Geomorphol- Sydney, 263–290. ogy and Environment. Springer-Verlag, Berlin, New Thornthwaite C. W., 1948: An approach toward a York, 265 p. rational classification of climate. Geographical Sweeting M. M., Lancaster N., 1982: Solutional and Review 38: 55–94. wind erosion forms on limestone in the Central Tjia H. D., 1985: Notching by abrasion on a limestone Namib Desert. Zeitschrift für Geomorphologie 26, coast. Zeitschrift für Geomorphologie 29, 3: 367–372. 2: 197–207. Toledo X., Zapater E., 1991: Geografía General y Sweeting M. M., Sweeting G. S., 1969: Some aspects of Regional de Chile. Editorial Universitaria, Colec- the Carboniferous limestone in relation to its land- ción Imagen de Chile, Santiago, 443 p. forms with particular reference to NW Yorkshire Tomaselli L., Lamenti G., Bosco M., Tiano P., 2000: Bio- 556 KRF•2 • OK.indd 556 15.12.2009 11:02:34 References diversity of photosynthetic micro-organisms dwell- cover and the evolution of limestone pavements, ing on stone monuments. International Biodeterio- Malham District, north Yorkshire. In: Paterson K., ration and Biodegradation 46: 251–258. Sweeting M. M. (Eds.), New Directions in Karst. Torunski H., 1979: Biological erosion and its signifi- Geobooks, Norwich, 461–471. cance for the morphogenesis of limestone coasts Trudgill S. T., 1987: Bioerosion of intertidal lime- and for nearshore sedimentation (Northen Adriatic). stone, Co. Clare, Eire 3: zonation, process and form. Senckenbergiana Maritina 11: 193–265. Marine Geology 74: 111–121. Tóth G., 2003: Karrenmorphologische Forschungen im Trudgill S. T., Crabtree R. W., 1987: Bioerosion of inter- Dachstein und im Toten-Gebirge – Gmundner Geo- tidal limestone, Co. Clare, Eire 2: Hiatella arctica. Studien 2. Beiträge zur Geologie des Salzkammer- Marine Geology 74: 99–109. guts, 191–198. Trudgill S. T., Inkpen R., 1993: Impact of Acid Rain Tóth G., 2007: A mérsékeltövi mészkö magashegységek on Karst Environments. Catena Supplement 25: fedetlen karros celláinak osztályozása és fejlödése 199–218. [Classification and development of bare karren cells Trudgill S. T., Smart P. L., Friederich H., Crabtree R. W., in calcareous high mountains]. BDF Természet- 1987: Bioerosion of intertidal limestone, Co. Clare, földrajzi Tanszék Kiadó, Szombathely, 116 p. Eire 1: Paracentrotus lividus. Marine Geology 74: Tóth G., Balogh Z., 2000: Karrmeanderek lefűződésének 85–98. vizsgálata a Júliai-Alpok Héttó-völgyének egy karsz- Tućan F., 1911: Die Oberflächenformen bei Carbon- tos térszíne alapján. Karsztfejlődés 5: 139–142. atgesteinen in Karstgegenden. Centr. F. Min. Geol. Trenhaile A. S., 1987: The Geomorphology of Rocky und Paleont., 343–350. Coasts. Clarendon Press, Oxford, 384 p. Uchupi E., Milliman J. D., Luyendyk B. P., Brown C. Trenhaile A. S., 2002: Rock coasts, with particular O., Emery K. O., 1971: Structure and origin of the emphasis on shore platforms. Geomorphology 48: southeastern Bahamas. American Association of 7–22. Petroleum Geologists, Bulletin 55: 687–704. Trenhaile A. S., Pepper D. A., Trenhaile R. W., Dali- Udden J. A., 1925: Etched Potholes. University of Texas monte M., 1998: Stacks and notches at Hopewell Bulletin 2509: 5–9. rocks, New Brunswick, Canada. Earth Surface Proc- Uijlenhoet R., Stricker J. N. M., 1999: A consistent rain- esses and Landforms 23: 975–988. fall parameterization based on the exponential rain- Tricart J., 1972: Landforms of the Humid Tropics: For- drop size distribution. Journal of Hydrology 218: ests and Savannas. Longman, London, 306 p. 101–127. Trimmel H., 1965: Speläologisches Fachwörterbuch. Underhill J. R., Gayer R. A., Woodcock N. H., Don- Landesverein für Höhlenkunde in Wien und Nied- nelly R., Jolley E. J., Stimpson I. G., 1988: The Dent erösterreich, Wien, 109 p. Fault System, Northern England – reinterpreted as Trudgill S. T., 1975: Measurement of erosional weight- a major oblique slip zone. Journal of the Geological loss of rock tablets. British Geomorphological Society of London 145, 2: 303–316. Research Group, Technical Bulletin 17: 13–19. Urushibara-Yoshino K., 1991: The regionality of solu- Trudgill S. T., 1976a: The marine erosion of limestones tion rate of limestone tablets in karst areas of Japan on Aldabra Atoll, Indian Ocean. Zeitschrift für Geo- (in Japanese). Institute for Applied Geography, morphologie, Supplementband 26: 164–200. Komazawa University, Regional Views 4: 107–117. Trudgill S. T., 1976b: The subaerial and subsoil ero- Urushibara-Yoshino K., 1996: Karst – environment sion of limestones on Aldabra Atoll, Indian Ocean. and human activities (in Japanese). Taimeido Ltd., Zeitschrift für Geomorphologie, Supplementband Tokyo, 325 p. 26: 201–210. Urushibara-Yoshino K., 2003: Karst terrain of raised Trudgill S. T., 1979: Spitzkarren on calcarenites, coral islands, Minamidaito and Kikai in the Nansei Aldabra Atoll, Indian Ocean. Zeitschrift für Geo- Islands of Japan. Zeitschrift für Geomorphologie, morphologie, Supplementband 32: 67–74. Supplementband 131: 17–31. Trudgill S. T., 1983: Preliminary estimates of intertidal Urushibara-Yoshino K., Miotke F. D. and Research limestone erosion on Tree island, Southern Great Group of Solution Rates in Japan, 1998: Solution Barrier Reef, Australia. Earth Surface Processes and rates of limestone tablets and CO measurement in 2 Landforms 8: 189–193. soils in limestone areas of Japan. Geografia Fisica e Trudgill S. T., 1985: Limestone geomorphology. Long- Dinamica Quaternaria, Supplement 4: 35–39. man, London, New York, 196 p. Urushibara-Yoshino K., Research Group of Solution Trudgill S. T., 1986: Limestone weathering under a soil Rates in Japan, 1999a: Interannual variation of lime- 557 KRF•2 • OK.indd 557 15.12.2009 11:02:34 Karst Rock Features • Karren Sculpturing stone solution rates in Japan. In: Bárány-Kevei I., fogenetikájához – Zirci Természettud. Múzeum Gunn J. (Eds.), Essays in the ecology and conserva- Közleményei 20: 7–46. tion of karst. Special Issue of Acta Geographica Sze- Veress M., Barna J., 1998: Karrmeanderek morfoló- gediensis, Acta Geographica 36: 201–211. giai térképezésének tapasztalatai. Karsztfejlődés 2: Urushibara-Yoshino K., Miotke F. D., Research Group 59–73. of Solution Rates in Japan, 1999b: Solution Rate of Veress M., Kocsis Z. S., 1996: A Szentbékkállai kőtenger Limestone in Japan. Physics and Chemistry of the madáritatóinak morfogenetikai csoportosítása. Pro- Earth A 24, 10: 899–903. ceedings of the 6th International Symposium on Usdowski E., 1982: Reactions and equilibria in the sys- Pseudokarst, Galyatető, 90–97. tems CO -H O and CaCO -CO -H O (0-50° C), a Veress M., Lakotár K., 1995: Saroknyom karrok mor- 2 2 3 2 2 review. Jahrbuch für Mineralogie, Abhandlungen fogenetikai csoportosítása Totes Gebirge-i példák 144: 148–171. alapján. Karsztfejlődés 1: 89–102. Vacher H. L., Mylroie J. E., 2002: Eogenetic karst from Veress M., Nacsa T., 1998: The study of the history of the perspective of an equivalent porous medium. dissolution of karren ground surfaces developing Carbonates and Evaporites 17: 182–196. to karren monadocks and inselbergs – with some Vajoczki S., Ford D., 2000: Underwater dissolutional examples. Karsztfejlődés 4: 77–108. pitting on dolostones, Lake Huron-Georgian Bay, Veress M., Nacsa T., Széles Gy., Dombi L., 1995: Néhány Ontario. Physical Geography 21, 5: 418–432. totesi karros forma domborzatrajzi ábrázolása. Vanstone S. D., 1998: Late Dinantian palaeokarst of Karsztfejlődés I. (Totes Gebirge karrjai) Pauz Kiadó, England and Wales: implications for the exposure Szombathely, 31–40. surface development. Sedimentology 45: 19–37. Veress M., Szabó L., 1996: Adatok a velemi Kalapos-kő Veevers J. J., Roberts J., 1968: Upper Palaeozoic rocks, morfogenetikájához. Vasi Szemle, 50: 211–234. Bonaparte Gulf Basin of Northwestern Australia. Veress M., Szabó L., Zentai Z., 1996: Evolution of hats Bureau of Mineral Resources, Australia, Bulletin 97, on the greenschist terrain of Kőszegi Mountain. 155 p. In: Eszterhás I., Sárközi Sz. (Eds.), Proceedings of Veni G., 1994: Hydrogeology and evolution of caves and the 6th International Symposium on Pseudokarst, karst in the southwestern Edwards Plateau. In: Elliot Galyatető, 98–104. W. R., Veni G. (Eds.), The Caves and Karst of Texas. Veress M., Szunyogh G., Tóth G., Zentai Z., Czöpek I., National Speleological Society, Huntsville, Alabama, 2006: The effect of the wind on karren formation on 13–30. the Island of Diego de Almagro (Chile). Zeitschrift Veress M., 1993: Egy Totes Gebirge-i nagy karrvályú für Geomorphologie 50: 425–445. kioldódástörténeti vázlata. Karszt és Barlang 1–2: Veress M., Tóth G., 2001: Formes et micro-reliefs de 21–28. lapiés: Approche morphométrique, application au Veress M., 1995: Karros folyamatok és formák rendsze- massif de Totes-Gebirge (Autriche). Karstologia 37, rezése Totes Gebirge-i példák alapján. Karsztfejlődés 1: 47–53. 1: 7–30. Veress M., Tóth G., 2002: Egy dachsteini réteglapos tér- Veress M., 1998: Adatok karrvályúk meanderfejlődésé- színrészlet karros fejlődéstörténete. Karsztfejlődés 7: hez. Karsztfejlődés 2: 75–90. 187–204. Veress M., 2000a: Karrformák összeoldódása. Karszt- Veress M., Tóth G., 2004: Types of meandering karren. fejlődés 5: 143–158. Zeitschrift für Geomorphologie 48, 1: 53–77. Veress M., 2000b: The main types of karren develop- Veress M., Tóth G., Czöpek I., 2003: Falikarrok morfo- ment of limestone surfaces without soil covering. genetikája dachsteini példák alapján. Karsztfejlődés Karsztfejlődés 4: 7–30. 8: 197–212. Veress M., 2000c: Adalékok karrványúk működéséhez. Veress M., Tóth G., Péntek K., 2001b: Adalékok karrfor- Hidrológiai Közlöny 80, 4: 207–209. mák kialakulási korához és fejlődési sebességéhez a Veress M., 2000d: The morphogenetics of karren mean- Hallstatt-gleccser jégmentes völgytalpán. Karsztfej- der and its main types. Karsztfejlődés 4: 41–76. lődés 6: 161–169. Veress M., 2002: Talaj nélküli sziklafelszínek néhány Veress M., Tóth G., Zentai Z., Kovács G., 2001a: Study karros jelensége és ezek hatására képződő karr- of a new method for characterising karren surfaces formák. Földr Értesítő, 51: 41–71. based on alpine researches. Revue de Géographie Veress M., 2003: Adalékok a homokkő anyagú Alpine 89: 49–62. kőtengerek (Káli-medence) pszeudokarrjainak mor- Verges V., 1985: Solution and associated features of 558 KRF•2 • OK.indd 558 15.12.2009 11:02:35 References limestone fragments in calcareous soil (lithic calcix- Reading’s IAS Lecture tour ‘99 (Croatia). Field-trip eroll) from southern France. Geoderma 36: 109–122. guidebook, Hrvatsko geološko društvo, Zagreb, Viles H. A., 1984: Biokarst: Review and Prospect. 23–25. Progress in Physical Geography 8: 523–542. Wagner G., 1950: Rund um Hochifen Gottesackerge- Viles H. A., 1987: Blue-green algae and terrestrial lime- biet. Öhringen, 72–80. stone weathering on Aldabra Atoll: a SEM and light Walkden G. M., 1972: The mineralogy and origin of microscope study. Earth Surface Processes and interbedded clay wayboards in the Lower Carbon- Landforms 12: 319–330. iferous of the Derbyshire Dome. Geological Journal Viles H. A., 1988: Organisms and karst geomorphology. 8: 143–160. In: Viles H. A. (Ed.), Biogeomorphology. Blackwell, Walkden G. M., 1974: Palaeokarstic surfaces of upper Oxford, 319–350. Viséan (Carboniferous) limestones of the Derby- Viles H. A., 1995: Ecological perspectives on rock sur- shire Block, England. Journal of Sedimentary Petrol- face weathering: towards a conceptual model. Geo- ogy 44: 1232–1247. morphology 13: 21–35. Walkden G.M., Davies J., 1983: Polyphase erosion of Viles H. A., 2001: Scale issues in weathering studies. subaerial omission surfaces in the late Dinantian of Geomorphology 41: 63–72. Anglesey, North Wales. Sedimentology 30: 861–878. Viles H.A., Moses C.A., 1998: Experimental produc- Walsh R. P. D., 1982a: Climate. In: Jermy A. C., Kavan- tion of weathering nanomorphologies on carbonate agh K. P. (Eds.), Gunung Mulu National Park, stone. Quarterly Journal of Engineering Geology 31: Sarawak. Sarawak Museum Journal (Special Issue 2) 347–357. 30, 51: 29–67. Viles H. A., Spencer T., 1986: Phytokarst, blue-green Walsh R. P. D., 1982b: Hydrology and water chemis- algae and limestone weathering. In: Paterson K., try. In: Jermy A. C., Kavanagh K. P., (Eds.), Gunung Sweeting M. M. (Eds.), New directions in Karst. Mulu National Park, Sarawak. Sarawak Museum Geobooks, Norwich, 115–140. Journal (Special Issue 2) 30, 51: 121–181. Viles H. A., Spencer T., Teleki K., Cox C., 2000: Obser- Walter-Levy L., Frécaut R., Strauss R., 1958: Contribu- vations on 16 years of microfloral recolonization tion à l’étude de la zone littorale des îles Baléares. data from limestone surfaces, Aldabra Atoll, Indian Biologie et chimie des algues calcaires. Formes Ocean: Implications for biological weathering. Earth du relief qui leur sont liées. Revue Algologique 3: Surface Processes and Landforms 25: 1355–1370. 202-228. Viles H. A., Trudgill S. T., 1984: Long term remeasure- Waltham A. C., 1984: Some features of karst geomor- ments of micro-erosion meter rates, Aldabra Atoll, phology in south China. Cave Science 11: 185–199. Indian Ocean. Earth Surface Processes and Land- Waltham A. C., 1995: The pinnacle karst of Gunung forms 9: 89–94. Api, Mulu, Sarawak. Cave and Karst Science 22, 3: Vincent P. J., 1983a: The morphology and morphome- 123–126. try of some arctic Trittkarren. Zeitschrift für Geo- Waltham A. C., 1997a: Pinnacle karst of Gunung Api, morphologie 27, 2: 205–222. Mulu, Sarawak. In: Song L., Waltham A. C., Cao N., Vincent P., 1983b: The dissolving landscape: step karren. Wang F. (Eds.), Stone Forest: A Treasure of Natural Geographical Magazine 55: 508–510. Heritage. Proceedings of the International Sympo- Vincent P. J., 1995: Limestone pavements in the British sium for Lunan Shilin to Apply for World Natural Isles: a review. Geography Journal 161: 265–274. Heritage Status, China Environmental Science Press, Vincent P., 1996: Rillenkarren in the British Isles. Beijing, 52–55. Zeitschrift für Geomorphologie 40, 4: 487–497. Waltham A. C., 1997b: Mulu – the ultimate in cavern- Vincent P., 2004: Loess on the Burren. North Munster ous karst. Geology Today 13, 6: 216–222. Antiquarian Journal 44: 59–67. Waltham A. C., Brook D. B., 1980a: Cave development Vincent P., Lee M., 1981: Some observations on the in the Melinau Limestone of the Gunung Mulu loess around Morecambe Bay, North-West England. National Park. Geographical Journal 146, 2: 258–266. Proceedings of the Yorkshire Geological Society 43: Waltham A. C., Brook D. B., 1980b: Geomorphologi- 281–294. cal observations in the limestone caves of Gunung Vlahović I., Velić I., Tišljar J., Matičec D., 1999: Lithol- Mulu National Park, Sarawak. Transactions of the ogy and Origin of Tertiary Jelar Breccia within the British Cave Research Association 7, 3: 123–139. Framework of Tectogenesis of Dinarides. In: Some Waltham A.C., Simms M. J., Farrant A. R., Goldie H. Carbonate and Clastic Successions of the External S., 1997: Karst and Caves of Great Britain. Chapman Dinarides: Velebit Mountain, Island of Rab. Harold and Hall, London, 358 p. 559 KRF•2 • OK.indd 559 15.12.2009 11:02:35 Karst Rock Features • Karren Sculpturing Waltham A. C., Webb B., 1982: Geology. In: Jermy A. C., cial reference to western Ireland. Transactions of the Kavanagh K. P. (Eds.), Gunung Mulu National Park, British Geographers 40: 155–172. Sarawak. Sarawak Museum Journal (Special Issue 2) Williams P. W., 1971: Illustrating morphometric anal- 30, 51: 68–74. ysis of karst with examples from New Guinea. Waltham T., 2001: Pinnacles and barchans in the Egyp- Zeitschrift für Geomorphologie 15: 40–61. tian desert. Geology Today 17, 3: 101–104. Williams P. W., 1972: Morphometric analysis of polyg- Waltham T., Hamilton-Smith E., 2004: Ha Long Bay, onal karst in New Guinea. Bulletin of the Geological Vietnam. In: Gunn J. (Ed.), Encyclopedia of Cave Society of America 83: 761–796. and Karst Science. Fitzroy Dearborn, New York, Williams P. W., 1973: Variations in karst landforms London, 413–414. with altitude in New Guinea. Geographische Ward S. D., Evans D. F., 1976: Conservation assessment Zeitschrift 32: 25–33. of British limestone pavements based on floristic cri- Williams P. W., 1978: Interpretations of Australasian teria. Biological Conservation 9: 217–233. Karsts. In: Davies J. L., Williams M. A. J. (Eds.), Webb J., 1996: Chillagoe Field Trip Excursion Guide. 7th Landform Evolution in Australia. Australian Australia and New Zealand Geomorphology Group National University Press, Canberra, 259–286. Conference, Cairns, 21 p. Williams P. W., 1983: The role of the subcutaneous Webb S., 1995: Conservation of limestone pavement. zone in the karst hydrology. Journal of Hydrology Cave and Karst Science 21: 97–100. 61: 45–67. Weber H., 1967: Die Oberflächenformen des festen Williams P. W., 2004: Mountain Kaijende arête and Landes. B. G. Teubner Verlaggesellschaft, Leipzig, pinnacle karst, Papua New Guinea. In: Gunn J. (Ed.), 367 p. Encyclopedia of Caves and Karst, Fitzroy Dearborn, Webster C. L., 1996: The Intrigue of the Question about New York, 467–468. the Bridgewater “Fossil Forest”, Victoria, Australia. Williams P. W., 2008: The role of the epikarst in karst Origins 23, 1: 50–60. and cave hydrogeology: a review. International Jour- Wentworth C. K., 1939: Marine bench-forming proc- nal of Speleology 37, 1: 1–10. esses II, solution benching. Journal of Geomorphol- Wilson J. M., 1985: Ecology of the Crocodile Caves of ogy 2: 3-25. Ankarana, Madagascar. Cave Science 12: 135–138. Wentworth C. K., 1944: Potholes, pits and pans subaer- Wilson J. M., 1987: The Crocodile Caves of Ankarana: ial and marine. Journal of Geology 52: 117–130. Expedition to Northern Madagascar. Cave Science Werner E., 1975: Solution of calcium carbonate and the 14: 107–119. formation of karren. Cave Geology 1: 3–28. Wilson P. A., 1975: Observations on the geomorphol- White S., 2000: Syngenetic Karst in Coastal Dune ogy of the Chillagoe limestones. Proceedings of the Limestone: A Review. In: Klimchouk A. B., Ford D. 10th biennial conference of the Australian Speleo- C., Palmer A. N., Dreybrodt W. (Eds.), Speleogene- logical Federation, Sydney, 69–73. sis: Evolution of Karst Aquifers. National Speleologi- Winkler E. M., 1979: Role of salts in development of cal Society, Huntsville, Alabama, 234–237. granitic tafoni, South Australia: a discussion. Jour- White S., Grimes K. G., Mylroie J. E., Mylroie J. R., nal of Geology 87: 119–120. 2007: The earliest time of karst cave formation. Pro- Woodruff C. D., Stoddard D. R., Harmon R. S., Spencer ceedings of the Time in Karst Conference, Karst T., 1983: Coastal morphology and Late Quaternary Research Institute, Postojna (on a CD-ROM), 5 p. history, Cayman Islands, West Indies. Quaternary White W. B., 1988: Geomorphology and hydrology of Research 19: 64–84. karst terrains. Oxford University Press, New York, Wray R. A. L., 1997: A global review of solutional 464 p. weathering forms on quartz sandstones. Earth-Sci- White W. B., Jefferson G. L., Haman J. F., 1966: Quar- ence Reviews 42: 137–160. zite karst in Southeastern Venezuela. International Wright V. P., 1988: Paleokarsts and paleosoils as indi- Journal of Speleology 2: 309–314. cators of paleoclimate and porosity evolution: a Wilford G. E., 1961: The geology and mineral resources case study from the Carboniferous of South Wales. of Brunei and adjacent parts of Sarawak. Geological In: James N. P., Choquette P. W. (Eds.), Paleokarst. Survey Department (British Territories in Borneo), Springer-Verlag, New York, 329–341. Memoir 10: 319 p. Yonge C. M., 1963: Rock-boring organisms. In: Sogn- Wilford G. E., Wall J. R. D., 1965: Karst topography in naes R. F. (Ed.), Mechanics of soft tissue destruction. Sarawak. Journal of Tropical Geography 21: 44–70. American Association for the Advancement of Sci- Williams P. W., 1966: Limestone pavements with spe- ence 75: 19–24. 560 KRF•2 • OK.indd 560 15.12.2009 11:02:35 References Young A. R. M., 1987: Salt as an agent in the develop- their morphogenetic analysis. Carsologica Sinica 13, ment of cavernous weathering. Geology 15: 962–966. 3: 270–280. Yu Y., Yang B., 1997: Paleoenvironment during forma- Zhang F., Geng H., Li Y., Liang Y., Yang Y., Ren J., Wang tion of Lunan Stone Forest. In: Song L., Waltham A. F., Tao H., Li Z., 1997: Study on the Lunan stone C., Cao N., Wang F. (Eds.), Stone Forest: A Treas- forest karst, China. Yunnan Science and Technology ure of Natural Heritage. Proceedings of the Inter- Press, 155 p. national Symposium for Lunan Shilin to Apply for Zhang F., Wang F., Wang H., 1997: Lunan Stone forest World Natural Heritage Status, China Environmen- landscape and its protection and conservation. In: tal Science Press, Beijing, 63–67. Song L., Waltham A. C., Cao N., Wang F. (Eds.), Yuan D., 1982: Glossary of Karstology. Geological Pub- Stone Forest: A Treasure of Natural Heritage. Pro- lishing House, Beijing, 55p. ceedings of the International Symposium for Lunan Yuan D., 1991: Karst of China. Geological Publishing Shilin to Apply for World Natural Heritage Status. House, Beijing, 224 p. China Environmental Science Press, Beijing, 71–77. Yuan D., 1997: A global perspective of Lunan Stone Zhang S., 1984: The development and evolution of forest. In: Song L., Waltham A. C., Cao N., Wang Lunan stone forest. Carsologica Sinica 3: 78–87. F. (Eds.), Stone Forest: A Treasure of Natural Her- Zotov V. D., 1941: Pot-holing of limestone by develop- itage. Proceedings of the International Symposium ment of solution cups. Journal of Geomorphology 4: for Lunan Shilin to Apply for World Natural Herit- 71–73. age Status, China Environmental Science Press, Bei- Zseni A., 1999: A talaj szerepe a mészkőjárdák jing, 68–70. kialakulásában [The role of soil cover in the evo- Zamora E., Santana A., 1979: Características climáticas lution of karrenfields]. CD proceeding of the 4th de la costa occidental de la Patagonia entre las lati- National Conference of Geographical Ph. D. Stu- tudes 46°40' y 56°30' S. Anales del Instituto de la Pat- dents, Szeged. agonia 10: 109–154. Zseni A., 2002a: The role of soil cover in the evolution Zeller J., 1967: Meandering channels in Switzerland. of karrenfelds. In: Gabrovšek F. (Ed.), Evolution of International Union of Geodesy and Geophysics, karst: from prekarst to cessation. Karst Research International Association of Scientific Hydrology Institute ZRC SAZU, Postojna, 299–306. (IUGG/IASH), Symposium on River Morphology, Zseni A., 2002b: Karrmezők talajainak vizsgálata Bern, 174–186. magyarországi és angol területeken [Investigation Zentai Z., 2000: Karrvályúk fejlődésének sajátossá- of soils of karren fields in Hungarian and English gai néhány Héttó-völgyi (Júliai-Alpok, Szlové- karst]. Karsztfejlődés 7: 281–295. nia) mintaterület adatainak felhasználásával. Zseni A., Goldie H., Bárány-Kevei I., 2003: Limestone Karsztfejlődés 5: 127–137. pavements in Great Britain and the role of soil cover Zezza F., 1994: Minoan palace of Knossos: weather- in their evolution. Acta Carsologica 32, 1: 57–67. ing of gypsum stones. In: Fassina V., Ott H., Zezza Zseni A., Bárány-Kevei I., 2000: Nagy-Britannia F. (Eds.), Proceedings of the 3rd International Sym- mészkőjárdái és a talaj hatása azok fejlődésében posium on the Conservation of Monuments in the [Limestone pavements of Great Britain and the influ- Mediterranean Basin, Venice, 635-646. ence of soil in the evolution of them]. Karsztfejlődés Zhang D., 1994: Distribution of Tibetan karren and 5: 181–194. 561 KRF•2 • OK.indd 561 15.12.2009 11:02:35 KRF•2 • OK.indd 562 15.12.2009 11:02:35 9 789612 541613