Acta carsologica

Citation
Acta carsologica

Material Information

Title:
Acta carsologica
Series Title:
Acta Carsologica
Alternate Title:
Krasoslovni zbornik
Creator:
Inštitut za raziskovanje krasa (Slovenska akademija znanosti in umetnosti)
Publisher:
Inštitut za raziskovanje krasa
Publication Date:
Language:
Multiple Languages

Subjects

Subjects / Keywords:
Geology ( local )
Genre:
serial ( sobekcm )

Notes

Restriction:
Open Access - Permission by Publisher
Original Version:
Vol. 36, no. 3 (2007)

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
K26-00158 ( USFLDC DOI )
k26.158 ( USFLDC Handle )
5219 ( karstportal - original NodeID )
0583-6050 ( ISSN )
8894944 ( OCLC )

USFLDC Membership

Aggregations:
Added automatically
Karst Information Portal

Postcard Information

Format:
serial

Downloads

This item is only available as the following downloads:

Physics And Chemistry Of Dissolution On Subaerialy Exposed Soluble Rocks By Flowing Water Films / Wolfgang Dreybrodt - Georg Kaufmann - ( .pdf )

Fractal Analysis Of The Distribution Of Cave Lengths In Slovenia Timotej Verbovšek ( .pdf )

The Mineralogical Study On The Cueva De Las Velas (Naica, Mexico) / Paolo Forti - Ermanno Galli - Antonio Rossi ( .pdf )

Submerged Speleothems - Expect The Unexpected. Examples From The Eastern Adriatic Coast (Croatia) / Maša Suri - Branko Jalži - Donat Petricioli ( .pdf )

Proposed Methodology Of Vulnerability And Contamination Risk Mapping For The Protection Of Karst Aquifers In Slovenia / Nataša Ravbar - Nico Goldscheider ( .pdf )

Questions Of Dry Valleys In Karst: Case Study Of Mali Dol, Kras (Slovenia) / Jure Košutnik ( .pdf )

Land Cover Mapping Using Landsat Satellite Image Classification In The Classical Karst - Kras Region / Žiga Kokalj - Krištof Oštir ( .pdf )

A Review Of The Forest Management History And Present State Of The Haragistya Karst Plateau (Aggtelek Karst, Hungary) / Eszter Tanács - Ferenc Szmorad - Ilona Bárány-Kevei -- ( .pdf )

Fossil Vertebrates And Paleomagnetism Update Of One Of The Earlier Stages Of Cave Evolution In The Classical Karst, Slovenia: Pliocene Of RnotiE Ii Site And RaIška PeIna Cave / Ivan HoráEk - Andrej Mihevc - Nadja Zupan Hajna - Petr Pruner - Pavel Bosák ( .pdf )

Early Late Pliocene Paleokarstic Fillings Predating The Major Plio-Pleistocene Erosion Of The Quercy Table, Sw-France / Jean-Pierre Aguilar - Jacques Michaux - Thierry Pélissié - Bernard Sigé ( .pdf )

Fossil Population Structure And Mortality Of The Cave Bear From The Mokrica Cave (North Slovenia) / Irena Debeljak ( .pdf )

Results Of Paleostomatological Analysis Of Material From The Cave Near Gornji Vakuf (South-West Bosnia) / Amila Zukanovi - Jasminko Mulaomerovi - Damir Marjanovi ( .pdf )

Biodiversity And Ecology Of Fauna In Percolating Water In Selected Slovenian And Romanian Caves / Oana T. Moldovan - Tanja Pipan - Sanda Iepure - Andrej Mihevc - Janez Mulec ( .pdf )

Subterranean Waters Described In Carniolan Letters Mailed To A. Kircher (* 1601; †1680) / Stanislav Južni ( .pdf )

Sound Properties Of Planinsko Polje (Slovenia) / Josip Korošec - Boštjan Perovšek - Dejan VonIna ( .pdf )


Full Text

PAGE 1

P H Y SICS AND CHEMISTR Y OF DISSOLUTION ON SUBAERIAL Y E X POSED SOLUBLE ROCKS B Y FLOWING WATER FILMS F IZIKA IN KEMIJA RAZTAPLJANJA ATMOSFERI IZPOSTAVLJENIH VODOTOPNIH KAMNIN POD TANKO VODNO PLASTJO Wolfgang D RE Y BRODT 1 & Georg K AUFMANN 2 Izvleek UDK 551.44:54.056 Wolfgang Dreybrodt & Georg Kaufmann: Fizika in kemija raztapljanja atmosferi izpostavljenih vodotopnih kamnin pod tanko vodno plastjo Skalne oblike na kamnina h izpostavljenim atmosferi so po sledica raztapljanja tanki h vodni h plasti, ki tejo po povrini kamnine. Hitrost raztapljanja apnenca oz. sadre je podana z zakonom F = (c eq -c), kjer je c eq -c razlika med koncentracijo raztopljeni h mineralov v vodnem lmu in ravnoteno kon centracijo glede na ustrezen mineral. Pri sadri je koecient doloen z molekularno difuzijo. Za apnenec pa ekperimentalni podatki kaejo, da je pri mono podnasieni raztopini (c<0.3c eq ) kinetini zakon podan s F = (0.3c eq -c) pri emer je za red velikosti veji kot pri koncentracija h c>0.3c eq Kinetine za kone uporabimo pri raunu denudacijske stopnje na kamniti h povrina h izpostavljenim razlinim intenzitetam deja. Nae ugotovitve se ujemajo tudi eksperimentalnimi podatki. Z ozi rom na tudijo razvoja deni h lebiev, ki sta jo predstavila Glew in Ford (1980), predlagamo novo razmerje med dolino lebiev in naklonom povrine. To razmerje uporabimo tudi na terenski h podatki h, ki sta ji h pridobila J. Lundberg in A.Gines. V lui tevilni h parametrov, ki vplivajo na razvoj deni h lebiev so dobljene korelacije zadovoljive. Kljune besede: kras, kinetika raztapljanja, deni lebi. 1 Institute of Experimental P h ysics, Karst Processes Researc h Group, University of Bremen, D-28334 Bremen, Germany 2 Fac h bereic h Geowissensc h aen,Fac hric h tung Geop h ysik, Haus D, Freie Universitaet Berlin, Malteserstr. 74-100, D-12249 Berlin, Germany Received/Prejeto: 01.09.2007 COBISS: 1.01 ACTA CARSOLOGICA 36/3, 357-367, POSTOJNA 2007 Abstract UDC 551.44:54.056 Wolfgang Dreybrodt & Georg Kaufmann: Physics and chemistry of dissolution on subaerialy exposed soluble rocks by owing water lms e basic process active in t h e formation of subaerial features on karst rocks is c h emical dissolution of limestone or gypsum by water lms owing on t h e rock surface. e dissolution rates of limestone and gypsum into t hin lms of water in laminar ow are given by F = (c eq -c), w h ere (c eq -c) is t h e dierence of t h e actual concentration c in t h e water lm and t h e equilibrium concentration c eq wit h respect to t h e corresponding mineral. W h ereas for gypsum is determined by molecular diusion t h e situation is more complex for limestone. Experiments are pre sented, w hic h s h ow t h at for hig h undersaturation, c<0.3c eq t h e rate law is F = ( 0.3c eq -c) ,and becomes hig h er by about a fac tor of ten t h an for t h e rates at c>0.3c eq ese rate laws are used to calculate denudation rates on bare rock surfaces exposed to rainfall wit h diering intensity. e estimations are in reason able agreement to eld data. Starting from t h e experiments on t h e formation of Rillenkarren on gypsum performed by Glew and Ford (1980), we suggest a new relation between t h eir lengt h from t h e crest to t h e Ausgleic h sc h e and t h e inclination of t h e rock surface. is is also applied to eld data of Rillenkarren on limestone provided by J. Lundberg and A. Gines. In view of t h e many parameters inuencing t h e formation of Rillenkarren t h ese correlations can be considered as satisfactory. Key words: karst, dissolution kinetics, Rillenkarren.

PAGE 2

ACTA CARSOLOGICA 36/3 2007 358 Karren is t h e generic term for dissolution features on ex posed soluble rock surfaces. Because of t h eir variety of s h apes and also t h eir regularity karren h ave been a fasci nating object of interest for geomorp h ologists. Alt h oug h a large body of observations and descriptions of karren h as been accumulated, knowledge on t h e p h ysical and c h emical processes on t h eir formation by dissolution is scarce. In t his paper we will focus on processes occur ring on bare rock surfaces suc h as limestone or gypsum exposed to t h e atmosp h ere, and covered by owing water lms. Two basic ingredients control t h e dissolution proc ess, t h e h ydrodynamics of t hin water lms owing down inclined surfaces, and t h e dissolution kinetics of t h e CO 2 -containing rainwater on limestone or gypsum rock surfaces. Aer discussion of t h ese two topics, we will use t his for an interpretation of t h e data of Glew and Ford (1980), w h o performed experimental simulations on t h e formation of Rillenkarren on inclined surfaces of plaster of Paris exposed to articial rainfall. Using t h ese results an interpretation of existing eld data on lengt h s of rillenkarren is presented. Also recent data by Petterson (2001) on dissolution on Rillenkarren from Plaster of Paris will be discussed. Finally t h e dissolution kinetics of limestone will be used to explain surface denudation on bare limestone sur faces. T HE FLUID D Y NAMICS OF WATER FILMS ON SMOOTH AND ROUGH SURFACES I NTRODUCTION W h en rain wit h intensity q (cms -1 1 mm/h our = 2.8 10 -5 cms -1 ) falls onto an inclined smoot h surface wit h slope angle a t hin layer of water is establis h ed (see Fig. 1). Its ow rate Q in cm 3 /s per unit widt h is given in cm 2 /s. Aer distance x'= down t h e surface of t h e rock Q is Qx qq l co s (1) e t hickness h (in cm) of t h e water lm is related to ow Q (Myers, 2002) by Q gh 3 3 si n (2) w h ere g is eart h s gravitational acceleration, is t h e density of water, and its viscosity. By using eqns. 1 and 2, we obtain t h e lm t hickness h q g = 3 3 l ta n (3) For rainfall intensities of 1 mm/h our onto a surface sloping wit h 45 and at a distance =50 cm, a fairly t hin lm of h=3.6 -3 cm develops. For 40 mm/h our rainfall intensity as used by Glew and Ford (1980), t h e lm t hick ness h is 1.2 -2 cm. e ow velocity u (in cms -1 ) is obtained from uh=Q by inserting eqns. 2 and 3 and one nds u gQ gq = = 2 3 22 2 3 3 3 si n c os si n l (4) Note t h at t h e velocity increases wit h ow distance Assuming a rainfall of 10 mm/h our a ow distance of 1 m, and a slope angle of 45, t h e velocity is 2.1 cms -1 If rainfall is reduced to 1 mm/h our one nds 0.5 cms -1 ese velocities are of importance because t h ey give t h e time of residence during w hic h a water parcel can dis solve bedrock. W h en t h e surface is roug h a correction factor must be introduced (Myers, 2002), w hic h is given by (5) Fig. 1: Water lm on inclined rock surface. W OLFGANG D RE Y BRODT & G EORG K AUFMANN

PAGE 3

ACTA CARSOLOGICA 36/3 2007 359 GY PSUM By use of rotating disc experiments Jesc h ke et al., (2001) h ave found t h at t h e surface reaction rates of gypsum (in mmol cm -2 s -1 ) are given by Rk cc cc with kc ss se qs eq s s se q (/ )( ) / 1 wit h Rk cc cc with kc ss se qs eq s s se q (/ )( ) / 1 (8) Here, c s is t h e calcium-concentration at t h e surface and t h e rate constant is k s =1.1 -4 mmol cm -2 s -1 e equilibrium concentration c eq wit h respect to gypsum is 15,4 -3 mmol cm -3 Ca 2+ and SO 4 2-ions released from t h e mineral surface are transported away from t h e sur face into t h e solution by molecular diusion. erefore concentration gradients exist and t h e surface concentra tion cs diers from t h e concentration c in t h e bulk. e transport rate R D by molecular diusion is given by Rk cc cc with kc DD eq De q D De q (/ )( ) / 1 wit h Rk cc cc with kc DD eq De q D De q (/ )( ) / 1 (9) w h ere k D is t h e transport constant and c is t h e average concentration of t h e bulk solution. Since due to mass conservation R S must be equal to R D one nds an eec tive rate law (Dreybrodt, 1988). Rk cc with k kk kk ef f e q e ff sD sD (/ ) 1 (10) or Rc cw ith ef fe q e ff sD sD () W h en k s >> k D k e becomes close to k D and rates are controlled by diusion. On t h e ot h er h and if k s << k D k e becomes close to ks and t h e rates are surface controlled. In t h e region w h ere ks and k D are of similar magnitudes bot h processes control dissolution. For a laminar water lm of t hickness h, t h e trans port coercient k D is given by (Beek & Muttzall, 1975) kD ch or Dh De q D 2 2 /, / (11) w h ere D is t h e coercient of diusion (110 -5 cm -2 s -1 ). For h=0.01cm one obtains D =1 -3 cm -1 and t h e rates are controlled by diusion. However, raindrops impinging on t h e water lm may cause mixing, w hic h could in crease t h e eective diusion constant. Only a factor of 10 surces to obtain surface control and a value of a e 7 -3 cms -1 To convert t h e rates from mmol cm -2 s -1 into retreat of rock in cm /year for gypsum one h as to multiply by a factor of 2.3 6 L IMESTONE Water lms running down rock surfaces under natural rainfall conditions h ave a comparatively small dept h of a few tent h s of a millimetre. In contrast to gypsum, w h ere dissolution rates are determined by bot h, surface reaction and molecular diusion, t h e situation on limestone is more complex. Fig. 2 sc h ematically depicts t hree regimes of dissolution rates. For hig h ly undersaturated solutions, 0 c c app rates are hig h and decline steeply wit h slope 1 to an apparent equilibrium concentration c app = 0.3c eq w h ere c eq is t h e true equilibrium concentration wit h re spect to calcite. e values of 1 are almost independent on t h e lm t hickness h for 0.005 cm < h < 0.03 cm, and 1 =5 -4 cms -1 (Kaufmann and Dreybrodt, 2007). To a good approximation t h e rates found by t h eo retical modelling can be expressed (Kaufmann, 2004) by Rc c Ia pp 1 () for c c eq (12) For hig h er calcium concentrations a second linear region wit h signicantly lower slope 2 arises, until close to equilibrium in region 3 for c c sw above t h e switc h concentration c sw =0.9c eq in hibition occurs and t h e rates are controlled by slow surface reactions. D ISSOLUTION KINETICS w h ere k is t h e roug hness of t h e surface and h t h e lm t hickness of t h e layer on a smoot h surface, as given by eqn. 3 (P h elps, 1975). is dimensionless factor relates t h e ow velocities u and u r of t h e smoot h to t h e roug h surfaces respectively. uf u rc (6) Because uh=Q t h e lm t hickness values are related by h f h r c 1 (7) For k/h = 2, a reasonable number, we obtain fc 0.4, and ow velocities are lower. Film t hickness values are hig h er by a factor of 2.5. P H Y SICS AND CHEMISTR Y OF DISSOLUTION ON SUBAERIAL Y E X POSED SOLUBLE ROCKS B Y FLOWING WATER FILMS

PAGE 4

ACTA CARSOLOGICA 36/3 2007 360 e dissolution rates in regions 2 and 3 are well understood (Plummer et al ., 1978; Bu hmann and Drey brodt, 1985; Svensson and Dreybrodt, 1992). ree basic c h emical reactions control t h e dissolu tion of CaCO 3 1. HC aC OC aH CO 3 2 3 2. HC OC aC OC aH CO 23 3 2 3 2 3. Ca CO HO Ca CO HO Ca HC OO H 32 2 3 2 2 2 3 Ca CO HO Ca CO HO Ca HC OO H 32 2 3 2 2 2 3 For all t hree reactions CO 2 dissolved in t h e solution must be h ydrated into carbonic acid, w hic h rapidly reacts to H + +HCO 3 4. HO CO HC O 2 2 23 5. CO OH HC O 2 3 e pH-values of t h e solution in region 2 are be tween 7.5 and 8.3. For suc h pH-values conversion of CO 2 is slow (Usdowski 1982, Dreybrodt 1988) and for t hin lms below 0.02 cm control by CO 2 -conversion limits t h e rates. For lm t hickness between 0.01 cm up to 0.04 cm slope values are about 2 3 -5 cms -1 lower by about one order of magnitude t h an 1 =5 -4 cms1 e reason for t h e hig h rates in region 1 are reac tions (1) and (3). W h en no calcite h as yet been dissolved t h e initial pH of t h e solution in equilibrium wit h CO 2 in t h e atmosp h ere is 5.7. Since reaction (1) is very fast pro tons are rapidly consumed by dissolving calcite. Furt h ermore dissolution of calcite produces OH ions. erefore pH increases to values of about 11. Be cause of t h e hig h concentration of OH conversion of CO 2 is fast by reaction 5. Wit h increasing Ca-concentra tion pH drops, and consequently slow conversion of CO 2 by reaction (4) takes over in controlling t h e rates. As a conclusion we state t h at for low concentrations c t h e rates are given by t h e relation Rc cc c eq eq 1 03 00 3 (. ); (13) Rc cc c eq eq 2 03 60 9 () ;. (14) EX PERIMENTAL DETERMINATION OF DISSOLUTION RATES IN REGION 1 W h en a t hin water layer of widt h W ows down a smoot h, plane limestone surface wit h inclination angle it dissolves calcite and t h e concentration c(x) of calcium along its ow pat h increases. Note t h at in t his section for simplicity we use x instead of x for t h e ow pat h on t h e rock surface. e amount of calcite dissolved during one second between positions x and x + dx is given by 1 (c app c(x))dxW. Due to mass conservation t his must be equal wit h Q total dc, w h ere dc is t h e increase in concentration from x to x + dx, and Q total is t h e total ow rate in cm 3 s -1 From t his a dierential equation is found dc dx W cc ap p 1 Q tota l () (15) Its solution is cx c Q x ap p () ex p 1 1 (16) w h ere Q=Q total /W is t h e amount of ow in one cm widt h of t h e lm. We use eqn. 16 to determine 1 experimentally. To t his end, we h ave constructed a c h annel of 5 cm widt h and 1.2 m lengt h by employing acryl rims xed to a plate of limestone. e inclination is = 3.2. At t h e end of t h e c h annel a funnel of acryl-glass c h annels t h e water into a h ole from w h ere it runs into a bottle. e experiment is illustrated in Fig. 3a, w hic h provides a view from above. To guide t h e water into a stable lm t h e c h annel at its up per end is blocked by a piece of acryl-glass, w hic h leaves a narrow space of a few tent h s of a millimetre between t h e limestone surface and its lower plane face (see Fig. 3b). Distilled water in equilibrium wit h t h e p CO 2 in t h e at mosp h ere by use of a peristaltic pump is introduced into Fig. 2: Dissolution rates of limestone by CO 2 -containing water. ree regimes of very fast (Region 1), moderate (Region 2), and inhibited dissolution rates (region 3) are clearly distinguishable. Only the fast dissolution rate in region 1 is relevant in this paper. W OLFGANG D RE Y BRODT & G EORG K AUFMANN

PAGE 5

ACTA CARSOLOGICA 36/3 2007 361 t h e upper compartment and a lm of constant t hickness moves down in laminar ow at ambient temperature of 20C. is lm is establis h ed by drawing down t h e water along t h e limestone surface by use of a wet paper strip as wide as t h e lm is desired to be. e water lm does not touc h t h e acryl walls but is kept by surface tension. It does not c h ange its s h ape, even w h en its dept h varies by a factor of t hree. e surface of t h e lm is absolutely plain as can be seen by a mirror like reection of lig h t. e ow rate Q is measured by collecting 10 ml of water at t h e out let h ole at t h e end of t h e c h annel, and measuring t h e time needed. e calcium concentration c end of t his sample is t h en measured for various values of Q. Furt h ermore wa Fig. 3: Experimental set up to measure limestone dissolution rates in Region 1 (top and side view). Length of channel 120 cm, width of channel 5 cm, average width W of water lm 4 cm. ter in equilibrium wit h atmosp h eric p CO 2 and calcite is used to measure c eq e calcium concentrations are de termined by measuring electrical conductivity, w hic h for suc h low concentrations is linear wit h calcium concen tration. e experiment was performed at 25 C. Eqn. 16 can be rewritten to (17) Fig. 4 s h ows t h e plot of t h e experimental data in terms of versus 1/Q total is can be tted tted wit h a straig h t line by using c app =0.3c eq =0.17 mmol/cm 3 From t h e slope 0.129 of t h e line one nds 1 =2.6 -4 cms -1 w hic h is in reasonable agreement to t h e t h eoreti cal predictions of 1 t h = 5 -4 cms -1 and c t h app =0.36 c eq Fig. 4: Calcium concentration versus inverse of ow rate for experimental data (squares). Q total is the total ow rate of the lm. e straight line is a least square t to the data. SOLUTION ON BARE ROCK SURFACES W h en rain falls onto an inclined surface t h e ow rate downstream increases (see Fig. 1). If at x=0 t h e ow rate is Q 0 ; t h en at a later position x it is given by Q = Q 0 + q xcos = Q 0 + qx (18) Mass conservation demands t h at (19) w h ere c is t h e average concentration at position x, and W is t h e widt h of t h e lm. f f w h ere f is a cor rection factor considering t h e roug hness of t h e rock sur P H Y SICS AND CHEMISTR Y OF DISSOLUTION ON SUBAERIAL Y E X POSED SOLUBLE ROCKS B Y FLOWING WATER FILMS

PAGE 6

ACTA CARSOLOGICA 36/3 2007 362 R ILLENKARREN EX PERIMENTS ON FORMATION OF RILLENKARREN ON G Y PSUM Glew and Ford (1980) experimentally simulated t h e for mation of Rillenkarren on gypsum by exposing inclined surfaces of plaster of Paris to a rainfall intensity of 38 mm/h our, w hic h lasted for 500 h. ey obtained well de veloped Rillenkarren. eir average lengt h from t h e crest to t h e Ausgleic h sc h e was dependent on t h e angle of inclination, as s h own in Fig. 5. Ford and Glew argued, t h at t h e Ausgleic h sc h e could form only w h en t h e wa ter lm exceeds a critical t hickness h c w hic h s h ould be hig h er t h an t h e roug hness k of t h e rock. Wit h t his as sumption by use of eqns. 1 and 2 one nds (24) erefore, by plotting versus tan one s h ould nd a straig h t line. is indeed is t h e case for t h e Glew & Ford (1980) data, as s h own by Fig. 5. e slope of t h is line is 14 cm, from w h ic h one nds a critical t h ickness h c =7.710 -3 cm if one assumes a smoot h surface. For a roug h surface wit h k=h c one nds a value of 10 -2 cm. Glew and Ford mea sured a value below (1.5.5) -2 cm, w hic h is in good agreement. ey also measured dissolution rates of 4 -3 cm/h. For t h eir experimental data one nds c 0.66 c eq from eqn. 17. e amount of ow leaving a rock of widt h W at x is equal to t h e amount of rainfall w hic h falls to t h e area Wx. It carries away t h e mass of rock qc x w hic h is dis solved from t h e rocks surface area Wx = Wx/cos. Con verting t h e mass of dissolved material to its volume one nds t h e retreat of rock face. If one assumes t h at t h e rock surface consists of small h alf sp h eres densely packed, instead of a smoot h plane, t h e surface area available for dissolution will increase by f =2. is gives an estimation on t h e order of magnitude of f Equation 19 states t h at t h e outow of calcium at position x + dx is given by t h e inow at position x plus t h e amount of calcium ions dissolved per time between x and x + dx. Neglecting terms wit h dxdc one nds a dierential equation. (20) wit h solution (21) For large values of x t h e concentration approac h es t h e value (22) 90% of t his value is reac h ed at a distance (23) For Q 0 =0 t h e concentration c is establis h ed imme diately. erefore dissolution rates are uniform down stream if one assumes t h at f is independent of t h e t hick ness of t h e water s h eet. is is not true for gypsum. A reasonable approximation is to use average values. For gypsum is maximal 7.1 -3 cms -1 if t h e rates are con trolled by surface reactions and at a s h eet t hickness of 0.1 mm it is 1.56 -3 cms -1 (see eqn. 10). At a s h eet t hickness of 0.5 mm one nds =3.8 -4 cms -1 Fig. 5: Length of experimental rillenkarren versus slope, tan e squares are experimental data from Glew and Ford (1980). e line represents eqn. 24 with hc=7.7 -3 cm W OLFGANG D RE Y BRODT & G EORG K AUFMANN

PAGE 7

ACTA CARSOLOGICA 36/3 2007 363 Rc q D g co s/ (25) w h ere g g/cm 3 is t h e density of gypsum. Wit h t h e experi mental conditions of Glew and Ford one nds R D =2.7 3 cos (cm/h). is ts reasonably well into t h eir data set. However, it represents a lower limit because one assumes laminar ow. Splas hing raindrops may disturb t his ow and cause mixing of t h e solution by w hic h t h e eective diusion constants increase. A factor of 10 is surcient to rise c to 0.9 c eq In a recent work Petterson (2001) h as exposed Rillen karren c h annels modelled from real limestone Rillenkar ren by plaster of Paris, to articial rain of 115 mm/h our intensity. By using an optical tec hnique h e measured t h e t hickness of t h e laminar owing water lms along t h e karren rills. e t hickness of t h ese lms, measured at a distance of 5 cm to 40 cm from t h e upper edge, range from 0.2 mm up to 0.8 mm, w h en t h e karren model was tilted by 30. Water samples collected from t h e karren at various distances from t h e crest were used to measure t h e calcium concentration prole along t h e karren. Petterson found an almost linear increase from 75 mg/ of calcium at 5 cm to a value of 105 mg/ at 40 cm. e average value was 90 mg/ mg/ Wit h an average lm t hickness of 0.5 mm one nds f = 7.610 4 cms -1 Wit h a rainfall intensity of 115 mm/h = 3.2 -3 cms -1 by use of eqn. 22 one obtains a value c =118 mg/ In view of t h e approximations t his can be regarded as good agreement to experiment and proves our t h eo retical considerations. I NTERPRETATION OF FIELD DATA OF RILLENKARREN A large body of data h as been collected, w hic h relates t h e lengt h s of Rillenkarren to t h e slope of t h e rock surface Fig. 6: Length of natural rillenkarren on limestone versus slope tan From J Lundberg and A. Gines, private communication. e straight lines are ts to =Atan P H Y SICS AND CHEMISTR Y OF DISSOLUTION ON SUBAERIAL Y E X POSED SOLUBLE ROCKS B Y FLOWING WATER FILMS

PAGE 8

ACTA CARSOLOGICA 36/3 2007 364 e value of h c 3 /q is close to t h at found from t h e de pendence of lengt h on slope in t h e previous example. As a nal example we discuss t h e data presented in Fig. 8 w hic h relates t h e average lengt h s of rillenkarren in t h e Serra de Tramuntana as a function of altitude above sea level, taken from: Lundberg and Gines ( 2006). ere is a clear decrease of lengt h wit h altitude h, w hic h can be caused by two reasons. First t h ere is a linear relation between altitude and temperature. e up most abszissa s h ows t h e corresponding temperature given by T = 17 0.0065 H (C), (29) w h ere t h e altitude H is in m. Furt h ermore mean annual precipitation q av is related to altitude H by q av = 461 + 0.4 H [mm/year] (30) See upper abscissa in Fig 8. We now assume t h at t h e actual rainfall to t h e rock is related to q av by q=f q q av w h ere f q is a constant. Bot h q and viscosity depend on altitude. Using eqns. 26, 27, 29, and 30 one can calculate t h e lengt h as a function of altitude. Wit h h c 3 /q as a tting parameter one obtains t h e curve in Fig. 8. e curve underestimates t h e large lengt h s, but s h ows t h e general trend. W h et h er it is a reasonable es timation must be judged from t h e value of h c 3 /q(H). If one assumes t h at 1000 mm/year correspond to an av erage actual precipitation of 10 mm/ h our one obtains h c =0.005. cm and correspondingly h c 3 /q(600)=6.710 -4 cm 2 s. is value is also close to t h ose found in t h e pre w h ere t h ey grow. From eqn. 24 one expects a linear rela tion of lengt h and slope. wit h A g q h c 3 3 (26) Fig. 6 s h ows average lengt h s of Rillenkarren ver sus slope (tan) for several areas (Gines and Lundberg, 2006). e straig h t line represents a least square t by t h e relation =consttan to t h e data points wit h 46(tan n2). Alt h oug h t h e scatter of points, w hic h could be caused by diering values of precipitation q at dierent sites and times is signicant one nds A = 12.6 cm for all plots. From t his by use of eqn (1) one obtains h c 3 / q=(3.9.2) -4 cm 2 s. Fig. 7 s h ows t h e relations hip of lengt h wit h mean annual temperature as reported by Lundberg and Gines (2006). e data can be tted by a relation = 0.5 T + 12.6 (cm), w h ere T is in C. e variation of in tempera ture could result from t h e temperature dependence of w hic h can be presented wit h an accuracy wit hin 2% by t h e empirical relation (27) Introducing t h is into eqn (26) one nds using h c 3 /q= 6.11 -4 cm 2 s and tan = 1 one nds t h at is valid between 0C and 25C. (28) Fig. 7: Length of natural rillenkarren on limestone versus mean annual temperature. From J Lundberg and A. Gines, private communication. Fig. 8: Length of natural rillenkarren on limestone (M allorca) versus altitude above sea level. From J Lundberg and A. Gines, private communication. e curve represents the t discussed in the text. W OLFGANG D RE Y BRODT & G EORG K AUFMANN

PAGE 9

ACTA CARSOLOGICA 36/3 2007 365 vious examples. Assuming an average actual precipita tion of 10 mm/ h dominant in t h e formation of karren one nds h c =0.0059 cm from t h e lengt h -slope relation and h c =0.0065 cm from t h e lengt h -temperature rela tion. In all t hree examples we h ave assumed an average precipitation of about 10 mm/h our during t h e formation or rillenkarren. is is a value, w hic h seems possible. For hig h er precipitation t h e lengt h would be smaller and would be overprinted by lower precipitation yielding longer karren. At low precipitation rates (1mm/h our) t h e karren become very long (2 m) and will form very slowly, suc h t h at t h ey may not be detected. In summary Glews and Fords idea t h at karren lengt h is determined by a critical t hickness h c of t h e down owing water lm can be used to explain eld data. One s h ould keep in mind t h at at a precipitation rate of 10 mm/h a lm t hickness of 0.006 cm is attained aer 27 cm on a smoot h rock surface inclined by 45. We do not know at present t h e p h ysical reason, w h y t his critical t hickness avoids furt h er growt h of rillen karren. is requires experimental observations of ow rates and c h emical composition of t h e water owing on natural karren on limestone during rain storms of vari ous intensities. D ENUDATION RATES IN THE FIELD GY PSUM Denudation rates on subaerial exposed gypsum samples h ave been reported by Gucc hi et al (1996). In an observa tion station close to Triest (Italy) wit h a yearly rainfall of 1350 mm t h ey found 0.9 mm/year as an average during an observation time of eig h t years. At rainfall intensities of 40 mm/h our t h e solution running o t h e rock h as a concentration of 0.5c eq At low er rainfall intensities of 4 mm/h our one nds c = 0.9c eq erefore it is reasonable to take an average value c = 0.75c eq for all t h e water during one years rainfall. From t his one nds a denudation rate of 1 mm/year. L IMESTONE For dissolution under linear kinetics wit h a rate law Rc c eq () (26) t h e time T, w hic h is needed until a volume element wit h initial concentration zero attains concentration of 0.63 c eq is given by Th / (27) For limestone wit h a lm t hickness of 0.2 mm one nds T1=10 -2 / f t =20s to attain c=0.64c app In t h e slower region 2, f b =2 -5 cms -1 and t h e time to reac h c=0.63c eq is T 2 =500 s. Under natural rainfall ow velocities are on t h e order of 1 cms -1 erefore dissolution will be ef fective only in region 1. Even w h en t h e water dissolved limestone in region 2 t h e dissolution rates were about two orders of magnitude lower. In ot h er words, all t h e water, w hic h falls to t h e rock surface, will leave it wit h concentration c derived from dissolution in region 1. Wit h f t =10 -3 cms -1 one nds c p c ap p 10 10 28 10 3 3 5 .c os (28) w h ere p is t h e rainfall intensity in mm/h. Fig. 9: Karren formation, from which water was collected. e grey line marks the ow path. e water was collected at the end of this line. P H Y SICS AND CHEMISTR Y OF DISSOLUTION ON SUBAERIAL Y E X POSED SOLUBLE ROCKS B Y FLOWING WATER FILMS

PAGE 10

ACTA CARSOLOGICA 36/3 2007 366 D ISCUSSION AND CONCLUSION We h ave presented some basic principles of ow dy namics of t hin water lms t h at can approximate ow on natural rock surfaces under rainfall conditions. Alt h oug h t h ese approximations are crude t h ey can be used for re alistic estimations. To understand t h e formation of geomorp h ologic features on rock surfaces basic knowledge of t h e disso lution rates by owing water s h eets is needed. Water in equilibrium wit h t h e p co 2 of t h e atmosp h ere dissolves limestone quickly up to a concentration of c app 0.3c eq For hig h er concentrations t h e dissolution rates drop rapidly. e time to reac h t h e concentration c app under natural rainfall conditions is on t h e order of 10 seconds, surciently s h ort, t h at all dissolution will be aected in t his regime of concentrations. Even if t h e solution would reac h concentrations hig h er t h an c app t h en dissolution rates drop to suc h low values t h at t h ey become insigni cant. We h ave presented experimental data, w hic h con rm t his be h aviour. It is also possible to understand from t h ese kinetics denudation rates of limestone measured in t h e eld. For gypsum dissolution rates are controlled by mixed kinetics of surface reactions and molecular dif fusion. erefore, t h e rates become dependent on t h e t hickness of t h e owing water s h eet. It is possible, h ow ever, to predict denudation rates on gypsum, as obtained from eld data. Furt h ermore experimental ndings on Rillenkarren can be explained. It s h ould be noted t h at we h ave neglected tempera ture dependence and h ave used 20C as standard. Since many of t h e constants used depend on temperature, h owever, some temperature dependence on t h e denuda tion rates is expected. In view of t h e many approxima tions t his is not of hig h signicance. We h ave not addressed t h e issue of Rillenkarren for mation. At present one may only speculate. e surface of t h e rocks acts to ow like a two-dimensional porous medium. In suc h an in h omogeneous environment c h an nelling can occur and parallel ow pat h s can arise, w h ere t h e ow rates are hig h er. For limestone t h en t h e concen tration c decreases and dissolution rates corresponding ly increase. In gypsum t h e solution is close to saturation and t h erefore t h e amount of dissolved rock is propor tional to t h e volume of t h e owing water. One t h erefore could imagine t h at Rillenkarren could only originate at roug h rocks. is issue can be h andled experimentally by simulating karren formation experimentally on polis h ed and roug h samples of plaster of Paris. An object of furt h er researc h s h ould be to mea sure ow velocities on limestone surfaces under natural conditions in dependence of rainfall intensity, and also to take samples of t h e water at various locations on t h at surface to obtain calcium concentrations. Suc h experi mental data could be of utmost use for a better under standing. One of t h e purposes of t his work is to stimulate suc h researc h. At low slope angles (cos) and for rainfall inten sities of 1 mm/h, c =0.97c app =0.29 mmol/ At 10mm/h, c =0.24 mmol/ and for extreme intensities of 40 mm/h c =0.14 mmol/ Cucc h i et al., (1996), by using micrometers, mea sured surface denudation rates on a h uge number of limestone samples wit h slope angles of about 15 degrees in t h e karst of Triest. ey found average dissolution rates sampled over eig h t years of 0.015 0.01 mm/year. At an average rainfall of 1350 mm/year in t h is region one needs an average run-o concentration c =0.5 c eq to explain t h is number. A closer inspection of t h e distribution of rain fall-dept h distribution is t h erefore necessary to verify t h is number. Anyway, our ndings support t h at denudation on bare rock by t h e dissolutional action of rainwater is caused by fast dissolution in region 1 of Fig. 2. We h ave performed a rst attempt to measure con centrations of rainwater owing from t h e surface of a karren formation of limestone from Lipica, Slovenia, ex hibited in front of t h e Postojna cave. Aer two days of h eavy rainfalls, cleaning t h e rock from dust, water was collected during a medium strong rainfall of a few mil limeters/h by use of an aluminum foil attac h ed to t h e rock. Fig. 9. s h ows t h e experimental situation. e water h ad own on top of t h e formation, w hic h exhibits only a slig h t inclination of about 10 degree for about one me ter, t h en down one h alf meter, almost vertically, w h ere it was c h annelled by t h e foil and collected into a beaker. is owpat h is depicted by t h e grey line. Measures were taken to prevent dilution of t h e sample by rainwater drip ping into it. In parallel a sample of rainwater was collect ed. e specic conductivities were measured in t h e eld. e conductivity of rain water was 6 S/ cm, w h ereas t h e water from t h e karren exhibited 57 S/cm. Analysis for calcium in t h e lab yielded a value of 0.25 mmol/liter, 38% of t h e saturation value of 65 mmol/liter at 10 C, t h e tem perature during collection of t h e sample. is result is in good agreement to w h at one expects from eqn. 28. W OLFGANG D RE Y BRODT & G EORG K AUFMANN

PAGE 11

ACTA CARSOLOGICA 36/3 2007 367 A CKNOWLEDGEMENT We t h ank Joyce Lundberg and Angel Gines for providing t h eir eld data in Figures 6, 7 and 8. R EFERENCES Beek, W. J., & Muttzall, K.M.K., 1975: Transport P h e nomena. Wiley, London and New Y ork. Bu hmann, D., & Dreybrodt, W., 1985: e kinetics of calcite dissolution and precipitation in geologically relevant situations of karst areas: 1. Open system.C h emical Geology 48, 189-211. Dreybrodt, W., 1988: Processes in Karst Systems.P h ys ics, C h emistry, and Geology, Springer, Berlin and New Y ork. Cucc hi, F., Forti, P., & Marinetti, E., 1996: Surface deg radation of carbonate rocks in t h e karst of Trieste. In J.J., Fornos, & Gines, A., (editors): Karren land forms, Universitat de les Illes Balears, Palma. Gines, A., & Lundberg, J., 2006: Rillenkarren, private communication. Glew, J.R., & Ford, D.C., 1980: A Simulation Study of t h e development of Rillenkarren.Eart h Surface Processes, 5, 25-36. Jesc h ke, A.A., Vosbeck, K., & Dreybrodt, W., 2001: Sur face controlled dissolution rates in aqueous solu tions exhibit nonlinear dissolution kinetics.Geo c himica et Cosmoc himica Acta, 65, 13-20. Kaufmann, G., & Dreybrodt, W., 2007: Calcite dissolu tion kinetics in t h e system CaCO 3 H 2 O-CO 2 at hig h undersaturation.Geoc himica et Cosmoc himica Acta, 71 (6), 1398-1410. Myers, T.G., 2002: Modeling laminar s h eet ow over roug h surfaces.Water Resources Researc h, Vol. 38, No. 11: 1230 (12-1 12-12). Petterson, 0., 2001: e development of a tec hnique to measure water lm t hickness and t h e study of ow h ydraulics and dissolutional c h aracteristics on plas ter of Paris rillenkarren c h annels, B.Sc.-esis, Uni versity of Bristol, Sc h ool of geograp hical Sciences, U.K. P h elps, H.O., 1975: S h allow laminar ows over roug h granular surfaces.J. Hydraul. Div. Am. Soc. Civ. Eng. 10 (HY3): 367-384. Plummer, L.N., Wigley, T.M.L., & Park h urst, D.L., 1978: e kinetics of calcite dissolution in CO 2 -water sys tems at 5 to 60 C and 0.0 to 1.0 atm CO 2 .Am. J. Sci. 278, 179-216. Svensson, U., & Dreybrodt, W., 1992: Dissolution kinet ics of natural calcite minerals in CO 2 -water-systems approac hing calcite equilibrium.C h emical Geol ogy 100, 129-34. Usdowski, E., 1982: Reactions and equilibria in t h e sys tems CO 2 H2O and CaCO 3 CO 2 -H 2 O (0-50 C). A review.Jb. Miner. Ab h. 144, 148-171. P H Y SICS AND CHEMISTR Y OF DISSOLUTION ON SUBAERIAL Y E X POSED SOLUBLE ROCKS B Y FLOWING WATER FILMS



PAGE 1

F RACTAL ANAL Y SIS OF THE DISTRIBUTION OF CAVE LENGTHS IN S LOVENIA F RAKTALNA ANALIZA PORAZDELITVE DOLIN JAM V S LOVENIJI Timotej V ERBOVEK 1 Izvleek UDK 551.435.84:51-7 Timotej Verbovek: Fraktalna analiza porazdelitve dolin jam v Sloveniji Dolina jam v Sloveniji je porazdeljena po potennem zakonu, ki je znailen za fraktalne objekte. Fraktalna dimenzija jam se giblje okoli vrednosti 1.07 in se spreminja glede na tektonsko in hidrogeoloko okolje. Odstopanja od idealne premice nastanejo zaradi podcenjenega tevila jam, saj je kraji h jam ve, kot ji h je dejansko zabeleeni h. Analiza tektonskega in hidrogeolokega okolja kae, da so najvije vrednosti fraktalne dimenzije znailne za kamnine s krako-razpoklinsko in razpoklinsko poroznostjo ter najnije za slabo prepustne kamnine. Bliina tektonski h struktur zelo vpliva na porazdelitev dolin jam, vpliv pa je veji pri jama h, ki leijo blije prelomom in narivom. Vrednosti di menzij jam so manje kot dimenzije mre razpok ali prelomov, najverjetneje zaradi koncentriranja tokov (kanalski h efektov) po mrea h razpok, kar posledino zmanja fraktalno dimen zijo. Fizikalni vzroki, ki povzroajo potenno odvisnost in vari acije fraktalni h dimenzij (eksponentov potennega zakona), so e vedno delno nepojasnjeni. Vseeno pa la h ko nastanek mre razpok pripiemo fraktalni fragmentaciji kamnin, ki deluje neodvisno od merila, jame pa nato ob nastajanju podedujejo doloene fraktalne lastnosti razpok. Kljune besede: dolina jam, fraktalna dimenzija, Slovenija, kraka hidrogeologija. 1 University of Ljubljana, Faculty of Natural Sciences and Engineering, Department of Geology, Akereva 12, Ljubljana, Slovenia p h one: +386 1 4704644, fax: +386 1 4704560, e-mail: timotej.verbovsek@guest.arnes.si Received/Prejeto: 01.10.2007 COBISS: 1.01 ACTA CARSOLOGICA 36/3, 369-377, POSTOJNA 2007 Abstract UDC 551.435.84:51-7 Timotej Verbovek: Fractal analysis of the distribution of cave lengths in Slovenia e lengt h s of t h e Slovenian caves follow t h e power-law distri bution t hroug h several orders of magnitude, w hic h implies t h at t h e caves can be considered as natural fractal objects. Fractal dimensions obtained from distribution of all caves are about 1.07, and vary wit hin dierent tectonic and h ydrogeological units. Some deviations from t h e ideal best t line in log-log plots (i.e. lower and upper cut-o limits) can be explained by underestimation, as many very s h ort caves are not registered. e study of tectonic and h ydrogeological setting indicates t h at t h e greatest dimensions occur in t h e rocks wit h karstic-fracture and fracture porosity and t h e lowest in low-permeability rocks. Proximity to major tectonic structures s h ows a detectable eect on t h e cave lengt h distribution, and t h e inuence is greatest for t h e caves closer to t h e faults and t hrust fronts. Dimensions are lower t h an t h ose of fracture networks and faults, w hic h can be most probably explained by ow c h anneling along t h e fracture networks, w hic h causes t h e decrease of fractal dimension. e p h ysical causes of power law scaling and variations in fractal dimensions (power law exponents) are still poorly understood, but t h e be h aviour of fracture networks is believed to be caused by a scale-independent fractal fragmentation of t h e blocks, and during t h e process of forming t h e caves in h erit some fractal geometrical properties of t h e networks. Key words: cave lengt h, fractal dimension, Slovenia, karst h y drogeology.

PAGE 2

ACTA CARSOLOGICA 36/3 2007 370 Fractals are dened as geometric objects wit h a self-simi lar property, w hic h implies t h at t h ey do not c h ange t h eir s h ape wit h scale (Feder, 1988). is statement is valid only for strictly self-similar mat h ematical fractals, like Koc h curve or Sierpinski carpet. One s h ould note t h at natural fractals dier from t h e ideal ones, as alt h oug h t h ey ap pear self-similar or self-arne at some scales, t h ere always exist a natural lower and upper cut-o scale, and frac tal analyses of t h ese objects are valid only wit hin t h ese two values. Fractal approac h es are appropriate w h ere classical geometry is not suitable for describing t h e ir regular objects found in nature. Generally t h ese cannot be modelled by easily-dened mat h ematical objects for example t h e clouds are not sp h eres, mountains are not cones, coastlines are not circles, and bark is not smoot h, nor does lig h tning travel in a straig h t line (Mandelbrot, 1983). e fundamental property of fractals is t h eir frac tal dimension (D), w hic h represents t h e ability of an ob ject to ll t h e space (in one, two or t hree dimensions). It can occupy non-integer values, compared to t h e integer values c h aracteristic of Euclidean objects, suc h as 3-D cubes or 2-D planar surfaces. As an example, an object wit h a fractal dimension of 1.4 exhibits properties of bot h 1-D and 2-D objects, as it lls t h e more space t h an a line (D = 1) and less space t h an a surface (D = 2). e caves form during t h e selective enlargement of fractures, bedding planes, faults and ot h er discontinuities in t h e soluble rock and only a few presolutional openings develop in larger passages (Palmer, 1991, Ford & Wil liams, 2007). e degree of a cave to ll t h e neig h bor ing rocks can be described quantitatively wit h t h e fractal dimension D. Bot h caves (Curl, 1999) and consequently cave lengt h s (Laverty, 1987) h ave been found to exhibit fractal properties. A study of Curl (1966) was performed for distribution of cave lengt h s and t h e number of en tranceless for t h e proper caves t h ose of accessible size including t h ose wit h no entrances. However, t h e inu ences of dierent lit h ologic properties, h ydrogeologic and tectonic settings on t h e distribution of cave lengt h s h ave not been yet discussed in detail. e goal of t his paper is to analyze and discuss t h e distribution of lengt h s of t h e caves in Slovenia in dier ent tectonic and h ydrogeological environments plus t h e inuence of t h e distance of t h e caves to t h e most obvious tectonic structures. As already noted by Curl (1986), t h e fractal interpretations probably do not directly reveal any details about geomorp hic processes responsible for t h e distribution of lengt h s of caves, but t his distribution does contain information about t h e geometry of caves and possibly constrains ideas about geomorp hic processes. I NTRODUCTION M ATERIALS AND METHODS ree dierent inuencing factors on t h e cave lengt h dis tribution were studied, as mentioned above (tectonic and h ydrogeological position plus t h e distance to t h e major tectonic structures). e data for 7552 caves were ana lyzed (spatial coordinates in t h e national Gauss-Krueger system and cave lengt h s), as recorded in t h e national cave register. e lengt h s are based on survey lengt h s, as recorded in t h e register. ere exist many ot h er ways of measuring cave lengt h s besides classical survey, includ ing 3-D measurements wit h sp h erical linked modular elements (Curl, 1986; 1999) and measuring in 2-D plane (plan lengt h) instead of performing classical total survey lengt h s in all t hree dimensions (Laverty, 1987). Never t h eless, regardless on met h od used, cave lengt h s distribu tion exhibits fractal properties. Also, as caves are usually long compared to passage breadt h, t h e classical approac h is acceptable. Unfortunately t h ere exists no data on sur veying met h od in t h e register, so t h e lengt h values are taken directly from register. is approac h is similar to t h e one of Curl (1966), w h ere if t h e lengt h of a cave was only stated in t h e report, t his value was used. An impor tant factor w hic h can aect t h e results of analyzed cave lengt h s is t h e number of entranceless caves, studied in detail by Curl (1966). e number of entranceless caves in Slovenia is not known, but probably it is hig h, as pre dicted by Curl. However, h e noticed t h at t h e average lengt h s of entranceless caves are more like t h ose of caves wit h one or more entrances t h an like t h e predicted aver age lengt h of entranceless caves. erefore t h e eect on t h e greater number of entranceless caves s h ould be uni formly distributed along a complete cumulative curve of cave lengt h s and s h ould not aect t h e s h ape of t h e curve, but s h ould only s hi it upwards. e register was imported into relational database program (MS Access) and t h e data was furt h er analyzed wit h GIS and statistical soware. Some basic statistics were also calculated, suc h as minimum and maximum lengt h and median. e median was used instead of mean or geometric mean, as t h e data does not follow nei t h er normal nor lognormal distribution. T IMOTEJ V ERBOVEK

PAGE 3

ACTA CARSOLOGICA 36/3 2007 371 R ESULTS T ECTONIC SETTING Caves were grouped into seven tectonic units according to t h eir location in t h e structural-tectonic map (Placer, 1999; Poljak, 2000; Fig. 1). Wit h minor deviation in t h e le-h and side of t h e plot, cave lengt h s follow power law distribution (linear line in log-log plot), c h aracteristic for fractal be h aviour. e median values of lengt h s (Tab. 1) are quite similar, except for t h e group of Adriatic fore land, and h ave t h e value around 23 m. e fractal dimensions enable more appealing in sig h t into t h e cave lengt h properties t h an t h e classical sta tistical approac h using t h e median or ot h er statistics, and t h ey vary among t h e tectonic units (Tab. 1). All results exhibit a very hig h value of R 2 Note t h at t h e values of D and R 2 in t h e table are valid only for t h e linear part, not for t h e complete curve. e lowest values can be found in t h e tectonic units of Periadriatic igneous rocks and Internal Dinarides, and t h e hig h est in t h e unit of Exter nal Dinarides and also in Sout h ern Alps. e discussion of t h e results is given in t h e next section. e number of analyzed caves (N=9) in t h e Adriatic foreland is too small to comment reliably, and deviations of t h e curve can be also seen in t h e plot (Fig. 2), so t h e D could not be calculated. HY DROGEOLOGIC SETTING Similar be h aviour of cave lengt h distribution can be ob served in t h e plot (Fig. 4) for t h e dierent h ydrogeologi cal units (Fig. 3). e hig h est values (Tab. 2) are found in aquifers wit h karstic and fracture porosity and t h ose wit h fracture porosity (D=1.06) and lowest in t h e aquifers and beds wit h intergranular porosity (D=0.87, D=0.86). De viations occur only for t h e group Beds wit h low poros ity, as D is greater t h an expected, about 1.08. is curve does not s h ow suc h a linear trend as t h e ot h ers, and t h e number of t h e data is muc h smaller. D ISTANCE TO THE MAJOR TECTONIC STRUCTURES Caves were grouped into t hree classes (m, m and m), w h et h er t h ey fell into t h e 300m, 500m or 1000m wide belt around t h e fault or t hrust front (Fig. 5), as s h own on t h e structural-tectonic map (Poljak, 2000). Similar be h aviour of general cave lengt h distribution as for t h e tectonic and h ydrogeological units can be ob served in t h e plot for t h e t hree groups, as t h e lengt h s fol low a linear t line in t h e log-log plots. e median values are similar, approximately 23 m. As for t h e tectonic units, t h e units wit h hig h er D contain longer caves, w hic h is reasonable for t h ose caves wit h fractal dimension larger t h an one compared to t h ose wit h D lower t h an one. Nevert h eless, a gap of number of caves occurs in t h e rig h t-h and side of all t hree plots (Fig. 6), for example at L = 3000m (logL = 3.5) for t h e m distance group. is indicates t h at t h e number of caves long about 3000m is muc h lower t h an in case w h ere all t h e caves are consid ered regardless of distance to t h e faults. e inuence of t h e tectonic structures is greater w h en t h e caves are clos er to t h e structures, as t h e gap is more noticeable for t h e m group and slowly disappears towards t h e m group. For t h e determination of tectonic setting, t h e struc tural-tectonic map of Slovenia (Poljak, 2000) was digi tized into a GIS s h ape le and t h e tectonic unit names were assigned to polygons. Caves belonging to a selected polygon (i.e. tectonic unit) were consequently selected from t h e complete dataset. For t h e determination of h y drogeologic setting, t h e s h ape le wit h t h e polygons of dierent h ydrogeological units was obtained from t h e Eu roWaterNet project website (h ttp://nfp-si.eionet.eu.int/ ewnsi), and t h e process of grouping t h e caves was similar to t h e grouping into tectonic units. e major faults and t hrust fronts were digitized from t h e same structural-tec tonic map (Poljak, 2000) and using t h e GIS soware t h e caves were grouped into t hree classes (m, m and m), w h et h er t h ey fell into t h e 300m, 500m or 1000m wide belt around t h e fault or t hrust front. Subsequently t h e relations hip between t h e numbers of caves N in t h e specic setting wit h lengt h greater t h an L was establis h ed, and t h e correlations were inspected in t h e log-log plots. For example, caves belonging only to t h e tectonic unit of External Dinarides were selected as explained in t h e former paragrap h, and t h eir distribution was analyzed in t h e following way. According to equa tion D = log N(s) / log L (Bonnet et al., 2001), t h e fractal dimension D was calculated as t h e negative slope of t h e linear regression best-t line of log Nlog L plot. e pro cess of calculation of D was repeated for all ot h er caves belonging to dierent units or groups of distance to t h e major tectonic structures. e number of steps for t h e lengt h s interval was c h osen as t h e power of 2 (1, 2, 4, 8 ...), wit h some major additional steps in between (10, 50, 100 etc). F RACTAL ANAL Y SIS OF THE DISTRIBUTION OF CAVE LENGTHS IN S LOVENIA

PAGE 4

ACTA CARSOLOGICA 36/3 2007 372 Fig. 2: Log-log distribution plot for the number of caves (N) longer than a specic length (L) in dierent tectonic settings Tab. 1: Results for fractal dimension of cave lengths in dierent tectonic units (D=fractal dimension, R 2 =coecient of determination, N=number of caves.e same notation is valid for the Tab. 2. Tectonic setting D R 2 N median min max Adriatic foreland 9 10.0 5 876 Southern Alps 1.00 0.9974 1744 21.5 1 10870 Internal Dinarides 0.74 0.9934 60 20.0 4 1726 External Dinarides 1.10 0.9970 5166 24.0 1 19555 Eastern Alps 0.92 0.9940 44 18.0 5 2057 Tc and Q sediments 0.89 0.9950 158 18.5 3 1300 Periadriatic igneous rocks 0.60 0.9741 13 20.0 7 205 Total 1.08 0.9993 7194 23.0 1 19555 T IMOTEJ V ERBOVEK Fig. 1: Structural-tectonic setting of the caves

PAGE 5

ACTA CARSOLOGICA 36/3 2007 373 Fig. 3: H ydrogeological setting of the caves Fig. 4: Log-log distribution plot for the number of caves (N) longer than a specic length (L) in dierent hydrogeological settings Tab. 2: Results for fractal dimension of cave lengths in dierent hydrogeological environments Hydrogeologic setting D R 2 N median min max Aquifers with intergranular porosity 0.87 0.9957 263 20.0 2 8057 Aquifers with karstic-fracture porosity 1.06 0.9975 5872 23.0 1 19555 Aquifers with fracture porosity 1.06 0.9954 510 24.5 4 5800 Beds with intergranular & fracture por. 0.86 0.9943 404 23.0 3 2780 Beds with low porosity 1.08 0.9852 77 25.0 7 1159 Total 1.07 0.9991 7126 23.0 1 19555 F RACTAL ANAL Y SIS OF THE DISTRIBUTION OF CAVE LENGTHS IN S LOVENIA

PAGE 6

ACTA CARSOLOGICA 36/3 2007 374 Fig. 5: Settings of the caves according to distance to the major faults and thrust fronts Fig. 6: Log-log distribution plot for the number of caves (N) longer than a specic length (L) in three groups of distance to the major tectonic structures T IMOTEJ V ERBOVEK

PAGE 7

ACTA CARSOLOGICA 36/3 2007 375 Cave lengt h distribution can be described as fractal. Re markably similar be h aviour of curves in t h e plots is ob served, as a linear plot of number of caves, longer t h an specic lengt h in t h e log-log plots. e fractal approac h provides a better insig h t into t h e cave geometry by ana lyzing t h e fractal dimension D instead of median or ot h er common statistics values. e fractal dimension calculated from t h e distribu tions can not be directly interpreted as a fractal dimension of t h e caves t h emselves, i.e. used as a direct measurement of t h e geometry of t h e caves, as t h ese two dimension are obtained in a dierent way. e rst one is calculated as a negative slope of t h e distribution of cave lengt h s, and t h e second one is usually obtained by a Ric h ardsons (yard stick) or box-counting met h od (Feder, 1988). However, t h ese distributions probably h ave a natural source, and t h e dierences between t h e fractal dimensions are clearly observable, as discussed below. e lowest values can be found in t h e tectonic units of Periadriatic igneous rocks and Internal Dinarides, w hic h are comprised mostly of low-porosity and especially of low-permeability rocks. e hig h est fractal dimensions (D=1.10) appear in t h e unit of External Dinarides. is unit is represented mostly by carbonates of Dinaric car bonate platform, w hic h are intensely fractured and karst ied. Similar explanation is valid for t h e unit of Sout h ern Alps (D=1.00), also consisting of karstied and fractured carbonates. e number of analyzed caves (N=9) in t h e Adriatic foreland is too small to comment reliably, and deviations of t h e curve can be also seen in t h e plot (Fig. 2), so t h e D could not be calculated. e rocks represented in t his unit are clastic (ysc h) sediments, and caves occur in t h e relatively t hin-bedded layers of calcarenite. Value of D in Tertiary and Quaternary sediments is lower t h an one, w hic h can indicate t h at t h e caves formed in t his unit could resemble objects wit h geometries between a point and a line, and not t h e branc hing c h annels wit h D hig h er t h an one. e fractal dimension closer to zero resembles point-like objects, t h e one closer to one linear objects and t h e one closer to two planar-lling objects. Values of D lower t h an one are t h erefore possible, as dimension is ob tained from t h e distribution and not from t h e geometric properties of t h e caves. Anot h er explanation for t h e low er values of D, alt h oug h less possible, could be found in t h e surveying met h od, as t h e caves are usually surveyed by classical linear met h od. One s h ould be t h erefore very careful w h en applying t h e results for fractal dimension obtained from t h e lengt h distribution to geometric prop erties of t h e caves. Nevert h eless, t h e value of dimension less t h an one clearly indicates t h at t h ese cave lengt h s are dierent from t h e ones wit h t h e hig h er dimension, and interpretation of t h ese values is still possible by fractal met h ods. e fractal dimension is lower in less soluble and less erodable rocks, like igneous rocks (D=0.60) or rocks of Internal Dinarides (D=0.74), w hic h were af fected by lower degree of fracturing and h ave generally lower permeability t h an t h e igneous rocks. e lowest values are found in Periadriatic group. e h ardness of t h ese rocks is greater compared to t h e ot h ers, and con sequently t h ey are h ard to erode (Kusumayud h a et al ., 2000), so t h e cave passages cannot develop in suc h extent as in more soluble carbonates or clastic rocks. Similar to t h e explanation of tectonic setting, t h e hig h er D for hydrogeologic setting could correspond to t h e rocks h aving been aected by fractal fracturation and subsequent dissolution along t h e fracture networks. e hig h est values (Tab. 2) are found in aquifers wit h karstic and fracture porosity and t h ose wit h fracture porosity (D=1.06) and lowest in t h e aquifers and beds wit h inter granular porosity (D=0.87, D=0.86). Deviations occur only for t h e group Beds wit h low porosity, as D is great er t h an expected, about 1.08. Possible explanation is t h at rocks wit h quite dierent h ydrogeological and lit h ologi cal properties occur wit hin t his group, w hic h inuences t h e fractal dimension. e vicinity of tectonic structures t h erefore h as a no ticeable eect on cave lengt h distribution, and t his can be most likely interpreted as tectonic dissection of lon ger caves into s h orter ones, and t h e tectonic eects can be manifested by displacement or collapse of t h e caves. is eect is also seen on t h e middle part of t h e plot (to t h e le side of t h e gap), w h ere a lower slope indicates t h e greater number of s h orter caves, w hic h are uniformly distributed along t h e line. Some points in t his part lie hig h er above t h e linear t line t h an expected and t h ese represent t h e increased number of s h orter caves, w hic h form by fragmentation of t h e longer ones. e deposited cave sediments can also inuence t h e results, as t h ese obstruct t h e traversable passages and can t h erefore di vide t h e cave into smaller segments. However, t his pro cess could h ardly be seen on t h e cumulative distribution plot for all caves, as t h e eect is more or less random and s h ould t h us be distributed along t h e complete plot and in addition it s h ould not be inuenced by distance to t h e tectonic structures. e fractal dimension obtained from t h e distribu tion of all caves is about 1.07 and varies among dierent tectonic and h ydrogeological units. e usual explanation of fractal dimension D hig h er t h an 1 indicates t h at caves wit h suc h dimension ll more space t h an t h ose wit h ideal dimension of 1.00 (for example a straig h t line), and t h e geological constraints limit t h e dimension to be lower D ISCUSSION AND CONCLUSIONS F RACTAL ANAL Y SIS OF THE DISTRIBUTION OF CAVE LENGTHS IN S LOVENIA

PAGE 8

ACTA CARSOLOGICA 36/3 2007 376 t h an 2. is is strictly true for dimensions calculated by Ric h ardsons or box-counting met h ods, and possibly not directly applicable to t h e ones obtained by distribution analysis, t h oug h t h e results are in very good agreement wit h t h e ot h er studies, as follows. Kusumayud h a et al (2000) obtained t h e dimension D = 1.04-1.08 .01 for caves in dierent lit h ologic environments in Indonesia and h ave used t h e box-counting met h od. uteri (1983) calculated t h e value of D = 1.08 for t h e cave Dimnice in Slovenia by Ric h ardsons (yardstick) met h od and similar approac h was used by Laverty (1987), w h o noted t h at cave lengt h exhibits fractal be h aviour wit h dimensions between 1.0 and 1.5 for caves in Sarawak and Spain. Frac tal dimension based on calculation from t h e distribution was determined by Curl (1986), w h o calculated a slig h tly hig h er value D = 1.4 t h an in t his study for caves in dif ferent environments. e dierences from t h e analyses of Curl (1986) can be attributed to t h e facts t h at in his study only t h e caves in limestone, marble and magnesitic limestone were analyzed and t h ose in dolomite, insoluble rock and gypsum were excluded. e dimensions are valid for t h e caves situated in specic regions in t h e USA, and t h e two exceptions from t h ese values are found in t h e Austrian and Iris h limestones. e geological, h ydrologi cal and tectonic settings certainly inuence t h e distribu tions, but t h ere is no available data to precisely compare t h e eects of t h e dierent environments. e fractal be h aviour of cave lengt h s distribution can be possibly explained as t h e dissolution occurs along t h e fractures, bedding planes, faults and ot h er disconti nuities in t h e soluble rock. It is well known t h at fracture networks are fractal, and t h eir dimension in 2D varies from around 1.3 to 1.7 (Bonnet et al., 2001). Faults are also fractal objects wit h rat h er lower dimensions, around 1.0 1.5. Results of t his study s h ow t h at t h e cave lengt h s distributions exhibit lower dimensions (D = 1.08) t h an t h e faults or t h e fracture networks. Alt h oug h t h e dimen sions can not be directly compared, lower values can be explained by c h anneling of ow t hroug h t h e fracture networks and especially bedding planes, w hic h serve as pat h ways for t h e water. It h as been observed t h at w h en a preferential way is dissolved t hroug h t h e network, t h e ow increases due to larger c h annels, t h e obliteration of irregular s h ape of t h e c h annel by erosion is faster and consequently t h e fractal dimension t h erefore decreases wit h larger ow rates (Kusumayud h a et al., 2000). e lower slope of t h e distribution curves on t h e le-h and side of t h e plots can be explained by unders ampling (Villemin et al., 1995), as below some t hres h old values t h e number of caves is underestimated. Similar trends were observed by t hree dierent studies. Curl (1966) analyzed t h e cave lengt h s, w h ere for t h e observed curves for natural data, t h e le part of t h e plots exhibited a lower slope and t h e modeled curves s h owed muc h uni form slope. He also noted for his data, t h at t h e cumula tive distributions s h ould be smoot h er if enoug h accurate data were available and all caves were considered. Loucks (1999) observed t his eect for t h e cave widt h s, w h ere deviations appeared for widt h below a t hres h old of few meters. Finally, Villemin et al (1995) noticed t his eect for fault lengt h s. e caves wit h lengt h s lower t h an few meters are merely not considered as caves (t h ey are not recorded in t h e register), and t h us t h eir number is muc h hig h er in t h e nature t h an actually recorded. e problem of cave denition can be raised h ere and was already dis cussed by Curl (1986). Generally t h e cave is regarded as suc h if it is traversable by h umans. Cave spaces evidently exist at all scales, but are not registered, and t h ese voids in t h e rocks are present from microns to h undreds of meters (Curl, 1999). e number of caves N wit h lengt h about 1 m s h ould t h us be muc h hig h er, around 107,000 and not around 7,200 as seen from example of t h e all units in t h e Fig. 1. is number can be simply estimated by inserting t h e value of L = 1 m into t h e best linear-t equation log N = 1.082 log L + 5.029 for all units. is is only a quick estimation, as t h e entranceless caves are not considered in t his study due to t h e lack of data in t h e register. e grap h could also be extended to a muc h low er scale (fart h er to t h e le), and t h e rock porosity (disso lution, fenestral, vug) can be also interpreted as a cave, but obviously not traversable by h umans. Extrapolation to t h e longer side is contrarily not possible, as in t his case t h e number of caves becomes less t h an one, and t h e curve also rapidly deviates from t h e linear t line. Similar observations were made by Curl (1966), w h ere t h e ob served (natural data) lengt h distributions exhibited more curvature on t h e plots t h an t h e modeled t h eoretical ones, so t h e proper basis for comparison of dierent cave set tings is t h e use of all caves. Alt h oug h t h e exact values of D can not be interpret ed directly by morp h ology of t h e caves, t h e larger fractal dimensions can be most probably interpreted by t h e abil ity of t h e caves to form complex longer passages, most probably along t h e initial fracture networks and also bedding planes. e more soluble and fractured rocks exhibit greater fractal dimensions, larger t h an one, and rocks wit h intergranular porosity (generally t h ose wit h low porosity, low solubility and small degree of fractur ing), s h ow D below one. ese variations probably h ave a natural source, and t h e dierences between t h e dimen sions are clearly observable, Larger values of D could be expected in anastomotic or networks caves, and lesser values in branc h work or single-passage caves (Palmer, 1991). e p h ysical causes of power law scaling and varia tions in fractal dimensions (power law exponents) are T IMOTEJ V ERBOVEK

PAGE 9

ACTA CARSOLOGICA 36/3 2007 377 A CKNOWLEDGMENTS e aut h or t h anks all t h e cave explorers for t h e eorts encountered during t h e cave measurements, France uteri for debate, David J. Lowe for smoot hing t h e Englis h version of t h e text and Lee Florea for useful com ments w hic h improved t h e quality of t h e text. still poorly understood (Bonnet et al ., 2001). e be h aviour of fracture networks is believed to be caused by fractal fragmentation of blocks (Turcotte and Huang, 1995), w hic h is scale-independent. Caves develop along t h e fractures and bedding planes, so t h ey in h erit t h e geometrical properties to some degree by dissolution of fractal networks. However, t h e processes w hic h lead to t h e values of fractal dimensions of fracture networks and fractal be h aviour of distribution of cave lengt h s and t h eir dependence are still a c h allenge to be analyzed. R EFERENCES Bonnet, E., Bour, O., Odling, N.E., Davy, P., Main, I., Cowie, P., Berkowitz, B., 2001: Scaling of fracture systems in geological media.Reviews of Geop h ysics, 39, 3, 347-383. Curl, R.L., 1966: Caves as a Measure of Karst. Journal of Geology, 74, 5, 798-830. Curl, R.L., 1986: Fractal Dimensions and Geometries of Caves.Mat h ematical Geology, 18, 2, 765-783. Curl, R.L., 1999: Entranceless and Fractal Caves Revis ited.In: Palmer, A.N., Palmer, M.V., Sasowsky, I.D., eds: Karst Modeling, Special Publication 5, Karst Water Institute, C h arlottesville, Virginia, 183-185. Feder, J., 1988: Fractals .Plenum Press, p. 283, New Y ork. Ford, D. & Williams, P., 2007: Karst H ydrogeology and Geomorphology .Jo hn Wiley & Sons, Ltd., p. 576. Kusumayud h a, S. B., Zen, M. T., Notosiswoyo, S., Gau tama, R. S., 2000: Fractal analysis of t h e Oyo River, cave systems, and topograp h y of t h e Gunungsewu karst area, central Java, Indonesia.Hydrogeology Journal, 8, 271-278. Laverty, M., 1987: Fractals in Karst.Eart h Surface Pro cesses and Landforms, 12, 475-480. Loucks, R.G., 1999: Paleocave Carbonate Reservoirs: Ori gins, Burial-Dept h Modications, Spatial Complex ity, and Reservoir Implications.AAPG Bulletin, 83, 11, 1795-1834. Mandelbrot, B., 1983: e Fractal Geometry of Nature. W. H. Freeman & Co., p. 468, New Y ork, NY Palmer, A. N., 1991: Origin and morp h ology of limestone caves.Geological Society of America Bulletin, 103, 1-21. Placer, L., 1999: Contribution to t h e macrotectonic sub division of t h e border region between Sout h ern Alps and External Dinarides.Geologija 41, 191-221. Poljak, M., 2000: Structural-Tectonic M ap of Slovenia. Mladinska knjiga & Geological Survey of Ljubljana, Ljubljana. uteri, F., 1983: Determination of t h e unknown cave passages lengt h by means of fractal analysis.In: Janaik, A., ed.: Nove smri ve speleologii (New trends in speleology), 24.-28.10.1983 (Proceedings), 61-62. uteri, F., 1992: Delovni seznam jam jugovzh odne Slo venije.Nae jame 34, 74-108. Turcotte, D. L. & Huang, J., 1995: Fractal Distributions in Geology, Scale Invariance, and Deterministic C h a os. In: Barton, C. C. and La Pointe, P. R. (eds): e Fractals in the Earth Sciences, 1-40. Villemin T., Angelier, J., Sunwoo, C., 1995: Fractal Distri bution of Fault Lengt h and Osets: Implications of Brittle Deformation Evaluation e Lorraine Coal Basin. In: Barton, C. C. & La Pointe, P. R. (eds): e Fractals in the Earth Sciences, 205-226. F RACTAL ANAL Y SIS OF THE DISTRIBUTION OF CAVE LENGTHS IN S LOVENIA



PAGE 1

THE MINERALOGICAL STUD Y ON THE CUEVA DE LAS VELAS NAICA, ME X ICO MINERALOKE RAZISKAVE JAME CUEVA DE LAS VELAS NAICA, MEHIKA Paolo F ORTI 1 Ermanno G ALLI 2 Antonio R OSSI 2 Izvleek UDK 552.54:551.44(72) Paolo Forti, Ermanno Galli & Antonio Rossi: Mineraloloke raziskave jame Cueva de Las Vegas (Naica, Mehika) Cueva de las Velas je zadnja odkrita jama na nivoju -290 v rud niku Naica. Votlina je je v zaetku leta 2005 presekala rudniko galerijo. Ena od zanimivosti jame je razirjenost diagenetski h mineralov, ki na debelo odloeni na jamski h stena h pred raz vojem kristalov sadre. Te obloge so iz kompleksni h, le redko kristaliozirani h elezo-mangano-svinevi h hidroksidov s prisotnostjo karbonatov, sulfatov in silikatov. Rast ostali h, v glavnem sulfatni h mineralov, se je zaela takoj po odvodnjavan ju tega dela rudnika pred 20 leti. Do sedaj smo nali 17 razlini h min eralov, od kateri h je pet taki h, ki so bili prvi najdeni v jamskem okolju. Raziskave te h mineralov, so poleg prisotnosti povsem novi h vrst kristalov sadre, omogoile nova spoznanja o speleogenezi te jame. Ta je verjetno precej bolj kompleksna od geneze ostali h jam na tem nivoju (-290), saj so ji botrovali razlini speleogenetski me h anizmi, kot npr. termalna korozija, neravnoteje sistema sadra/an hidrit, delovanje razlini h kislin, ter kapilarna migracija in izh lapevanje. Kljune besede: jame v rudniki h, jamski minerali, speleogenetski me h anizmi, Me hika. 1 La Venta Exploring Team & Istituto Italiano di Speleologia, paolo.forti@ unibo.it 2 Dipartimento di Scienze della Terra, Universit di Modena e Reggio Emilia Received/Prejeto: 22.08.2007 COBISS: 1.01 ACTA CARSOLOGICA 36/3, 379-388, POSTOJNA 2007 Abstract UDC 552.54:551.44(72) Paolo Forti, Ermanno Galli & Antonio Rossi: e mineralogi cal study on the Cueva de Las Vegas (Naica, Mexico) e Cueva de las Velas is t h e last cave unveiled at -290 level wit hin t h e Naica Mine; t h e cavity h as been intercepted by a mine gallery at t h e beginning of 2005. One of its peculiarities is t h e widespread t hick deposits of diagenetic minerals deposited over t h e cave walls before t h e beginning of t h e evolution of t h e giant gypsum crystals. ese deposits consist of complex, oen scarcely crystalline iron-manganese-lead oxides-h ydroxides, but carbonates, sulp h ates and silicates are also present. Ot h er minerals, mainly sulp h ates, started developing just aer t his area of t h e mine was dewatered some 20 years ago. Presently 17 dierent minerals h ave been observed, 5 of w hic h (orien tite, starkeyite, szmolnokite, szmikite and woodrurte) are completely new for t h e cavern environment. e study of t h ese minerals, toget h er wit h t h e presence of a completely new type of gypsum crystals, allowed to improve t h e knowledge on t h e speleogenetic evolution of t his cave, w hic h seems to be by far more complex t h an t h at of t h e ot h er cavity of t h e -290 level. Its complexity is reected by t h e activity of a larger number of dierent speleogenetic mec h anisms. Among t h em are wort h of mention t h e t h ermal corrosion/dissolution, t h e an h ydritegypsum disequilibrium, t h e acid aggression, and t h e capillary migration and evaporation. Keywords: Mine caves, cave minerals, speleogenetic mec h a nisms, Mexico.

PAGE 2

ACTA CARSOLOGICA 36/3 2007 380 e systematic study of mine caves h as emp h asized t h e hig h scientic interest of t h e minerogenetic processes active t h erein (De Waele & Naseddu, 2005; Forti et al., 2005). Fig. 2: General view of the giant gypsum crystals (caver for scale) (photo Archive Speleoresearch & lm and La Venta Exploring team). I NTRODUCTION From t his point of view, t h e natural cavities crossed by mine galleries in Naica (C hi h ua h ua, Mexico) (Fig. 1) h ave been world renowned for over a century, due to t h e dimension and purity of t h eir gypsum crystals (Hill & Forti, 1997) (Fig.2). Beside Cueva de las Espadas (Swords cave), unveiled at t h e be ginning of t h e 20t h century at t h e -120 level, w h ere crystals up to 2 meters in lengt h exist (Degou tin, 1912; Fos h ag, 1927). In t h e last 5-6 years mine galleries at t h e -290 level h ave intercepted several natural cavities, t h e most important of w hic h are Cueva de los Cristales (Crystal Cave), Ojo de la Reina (Queens Cave) (S h an gun, 2001) and Cueva de las Ve las (Sails Cave) (Fig. 3). All t h ese caves h ost gypsum crystals muc h bigger t h at t h ose in t h e Cave of t h e Swords, but t h ere are many ot h er features wort h y of study (Forti, 2006). e Cueva de las Velas is t h e latest large mine cave found at Naica. is cavity was intercepted in 2005 and it was explored and mapped by La Venta Exploring Team in April 2006 (Badino & Forti, 2007). e cave, w hic h h as a total lengt h of about 80 m, is oriented nort h-sout h and consists of two main large rooms par tially superimposed (Fig. 3). e average hig h of t h e c h ambers is 3-4 m, w hile t h eir widt h is oen over 10 m. Cueva de las Velas immediately proved to be ex tremely interesting due to peculiarity of t h e h osted c h em ical deposits. W h at makes t his cavity absolutely special is t h e presence of many small, t hin and very delicate spe leot h ems (t h e sails), developed on top of t h e gypsum crystals in t h e upper room of t h e cave (Fig. 4) (Bernabei et al., 2006), but t his is not t h e single peculiarity of Cueva de la Velas. In fact it is t h e single cavity in w hic h most of t h e carbonate rock, w h et h er covered or not by gypsum crystals, is overlain by a t hick (up to 20-50 cm) deposit of metallic oxides-h ydroxides (Fig. 5). ese deposits are muc h more frequent and t hick in t h e lower room, w hic h t h erefore exhibits a muc h darker aspect. e present study is focused on t h e minerals of t h ese deposits and it is a part of a general researc h project Fig. 1: Sketch of the Naica mine with the locations of the main natural cavities: on the right the stratigraphic sequence as derived by a 1150 m long drilling (aer Badino & Forti, 2007, modied). P AOLO F ORTI E RMANNO G ALLI & A NTONIO R OSSI

PAGE 3

ACTA CARSOLOGICA 36/3 2007 381 Fig. 4: A characteristic gypsum crystal (called vela, sail), which gave the name to the Cueva de las Velas (P hoto Archivio La Venta & S/F). Fig. 3: P lan and vertical sections of the Cueva de las Velas with the sample locations. w hic h t h e owner of t h e mine, t h e Peoles Company, decided to commit to La Venta Exploring Team from Italy and Speleo researc h & Films of Mexico City in 2006 (Forti, 2006). Fig. 5: A portion of the cave wall in which a thick deposit of metallic oxides-hydroxides are exposed (P hoto Archivio La Venta & S/F). THE MINERALOGICAL STUD Y ON THE CUEVA DE LAS VELAS NAICA, ME X ICO EX PERIMENTAL OBSERVATIONS e general geology, t h e structure and t h e ore evolution of Naica region is well known and documented (Stone, 1959; Megaw et al., 1988; Lang, 1995; Garca-Ruiz et al., 2007), t h erefore it would be useless to discuss t h em in detail. e Naica mine opens on t h e nort h-western side of a 12 km long and 7 km wide dome-s h aped structure NW-SE oriented and aected by secondary foldings, faults and erosion. is structure h as an average h eig h t of 1700 m above sea level and is formed, almost entirely, by calcare ous rocks (limestone, dolomitic limestone and carbonate dolostone) t h at h ave settled during a dozen million years, starting from t h e Albian (125 Myr BP). e sulp hides mineralization (Pb, Zn and Ag) h as formed due to h ydrot h ermal circulation, produced by tertiary dykes(26,2-25,9 My) some 3 km below t h e Naica surface. e structural control over t h e arrangement of t h e mineralized masses, t h e water circulation and t h e karst development was exerted by a system of faults and frac

PAGE 4

ACTA CARSOLOGICA 36/3 2007 382 Fig. 6: Limestone ceiling of the cave close to the entrance where small corrosion domes induced by acid aggression developed (P hoto Archivio La Venta & S/F). Fig. 7: a) Lower surface of gypsum mega-crystals incrusted by a dark grey to reddish brown thin scarcely cemented powder; ESEM images: b) well formed transparent celestine crystals; c) toroidal grain of earthy milky white rozenite; d) whitish earthy hard aggregate of starkeyite; e) layered structure of thin lms rich in rozenite and starkeyti. P AOLO F ORTI E RMANNO G ALLI & A NTONIO R OSSI tures, parallel to t h e longer axis of t h e dome, NW-SE oriented and dipping towards SW all t h e way to verticality (Fig. 1). Finally, t h e mining area is still under t h er mal anomaly: t h e water t h at spurts into t h e mine bottom reac h es a temperature close to 59C. e morp h ology of t h e few parts of t h e Cueva de las Velas free of gypsum crystals s h ows t h at karst evolution was mainly controlled by bedding planes, w hile, at least in one developing stage before t h e deposition of gypsum, acid aggression induced by sulp hide oxidation was t h e main corrosion factor (Fig. 6). Nine samples were taken from seven spots in t h e wall of t h e cave w h ere t h e primary minerals h ave been transformed to give rise to alteration compounds: all of t h em were taken in t h e lower part of t h e caves w h ere t h e widespread lack of gypsum cover allows an easy c h oice (Fig. 3). Four samples were obtained by scratc hing t h e inner part of t hree wall fractures in t h e mine gallery (spots 1, 2, 3 of Fig. 3) w hic h are clearly connected wit h t h e cave and in one fracture (spot 6 of Fig. 3) in t h e wall between low er and upper part of t h e cavity. Inside all t h ese fractures t h ere were clear evidences of active diagenesis, probably induced by h ydration and/or oxidation due to t h e pres ence of an atmosp h ere as a consequence of mine dewa tering since 17 years. All t h ese samples consist of rock fragments (calcite and/or dolomite 1 and 2; and gyp sum 3 and 6) covered by alteration material. One side of sample 1 is covered by a t hin lm of sp h eroid aggregates of small scaleno h edral semitransparent calcite crystals, w hic h are in turn partially covered by greyis h powder. e ot h er side is covered by a deposit of grey to reddis h eart h y material. Bot h sides of sample 2 are covered by t hin silver s hining cauliower s h aped very fragile blades, w hic h loose t h eir s hining and turn into eart h y coal black powder w h en scratc h ed. Bot h samples from spot 3 and 6 consist of gypsum fragments covered by an eart h y deep grey, cinder like, powder very similar to t h at of sample 2 even if wit h bigger grains. Because t h ese four samples h ave similar textural and structural c h aracteristics (t hin powder from w hite to greyis h to pale brown) and evidenced similar mineral ogical composition, t h ey h ave been considered as a single sample, referred as Wf (wall fractures) in Table 1. e ot h er ve samples came from t h e t hick black deposit underlying t h e gypsum crystals and were taken in t h e spots 4, 5, 7 of Fig. 3. All of t h em consist of black, s hining black, red to reddis h-brown, eart h y yellow, some times partially cemented powder (Fig. 7a). Always inside t h e powder t h ere are several small transparent sub-mil limetric sized crystals. Due to t h eir rat h er constant mor

PAGE 5

ACTA CARSOLOGICA 36/3 2007 383 Tab. 1. Cave minerals of Cueva de Las Velas (Naica) Sample type Mineral Chemical formula System Group Fig. No References 1 Wf Anglesite PbSO 4 Orth. Barite HB 5, 24 Wf Bd Celestine SrSO 4 Orth. 7b; 9f HB 5, 122 Wf Szmolnokite* FeSO 4 H 2 O Mon. Kieserite HB 5, 688 Wf Kieserite MgSO 4 H 2 O Mon. HB 5, 358 Wf Szmikite* MnSO 4 H 2 O Mon. HB 5, 687 Wf Bd Gypsum CaSO 4 2H 2 O Mon. 9e; 9f HB 5, 271 Wf Rozenite FeSO 4 4H 2 O Mon. Rozenite 7c, e HB 5, 602 Wf Starkeyite* MgSO 4 4H 2 O Mon. 7d, e HB 5, 663 Wf Jarosite K 2 Fe 3+ 6 (SO 4 ) 4 (OH) 12 Trig. Alunite 8a HB 5, 330 Wf Bd Calcite CaCO 3 Trig. Calcite HB 5, 101 Wf Dolomite CaMg(CO 3 ) 2 Trig. Dolomite HB 5, 191 Wf Bd Goethite -Fe 3+ O(OH) Orth. 8d, e, f; 8e HB 3, 223 Wf Fluorite CaF 2 Cub. HB 3, 205 Bd Coronadite Pb(Mn 4+ Mn 2+ ) 8 O 16 Mon. Cryptomelane 9a, b, c, d, e, f HB 3, 138 Wf Woodrute* ZnMn 3 O 7 2H 2 O Tetr. 8b HB 3, 606 Bd Opal SiO 2 nH 2 O Amor. Wf Orientite* Ca 2 Mn 2+ Mn 3+ 2 Si 3 O 10 (OH) 4 Orth. 8c HB 2/2, 602 1 HB: Ant h ony et al. Handbook of Mineralogy, Volume, Page. New cave mineral. THE MINERALOGICAL STUD Y ON THE CUEVA DE LAS VELAS NAICA, ME X ICO p h ology and mineralogical composition also t h ese sam ples h ave been considered as a single one and referred in t h e Table 1 as Bd (Black deposit). EX PERIMENTAL METHODS A detailed analysis of all t h e samples by t h e stereoscopic microscope was performed to distinguis h and to separate t h e dierent mineralogical p h ases present in eac h sam ple. en t h e single p h ases were analysed by a powder diractometer (P hilips PW 1050/25), w h en t h e material was quantitatively enoug h and h omogeneous, or by a Gandol camera (: 114.6 mm, exposition: 24/48 hrs), w h en t h e material was scarce or in h omogeneous. Always t h e experimental conditions were: 40Kv e 20 mA tube, CuK Ni ltered radiation ( = 1.5418 ). Rat h er all t h e samples analyzed by Gandol camera were later used to obtain images and c h emical qualita tive analyses t hroug h an electron scanning microscope (ESEM FEI Quanta 200) wit h an electronic microprobe (EDS Oxford INCA 350) at t h e C.I.G.S. (Centro Interdi partimentale Grandi Strumenti) of t h e Modena and Reg gio Emilia University. A NAL Y TICAL RESULTS e diagenetic minerals observed in t h e rst group of samples (Wf) are by far t h e majority (see Tab. 1) and t h ey consist mainly of h ydrated sulp h ates, carbonates, oxides and silicates. Sample 1, and in particular its frac tion consisting of cinder grey eart h y material proved to be very interesting. In fact in t his sample beside common cave minerals like celestine (Ant h ony et al., 2003c), one of t h e most abundant mineral w hic h is present mainly as equant cross-section (Fig. 7b) but also as millimetric, perfectly transparent, well-formed lat h like crystals, tabu lar on {001} and anglesite (Ant h ony et al., 2003a), t h ere are also very rare h ydrated sulp h ates of Fe like szmolnok ite (Ant h ony et al., 2003m), and rozenite (Ant h ony et al., 2003h), of Mg come kieserite (Ant h ony et al., 2003g), and starkeyite (Ant h ony et al., 2003i),) and of Mn like szmikite (Ant h ony et al., 2003l). ree of t h em (szmol

PAGE 6

ACTA CARSOLOGICA 36/3 2007 384 A few minerals (celestine, gypsum, calcite and goe t hite) are common to bot h t h e sample types (Wf and Bd). Goet hite (Ant h ony et al., 1997c), is by far t h e most com mon and widespread compound. It is normally scarcely crystalline and occurs normally as sub-millimetric par tially empty sp h eres. e outer part is reddis h-brown to black w hile t h e inner part consist of eart h y red aggregates. e ESEM-EDS analyses evidenced t h e presence of tus of laments consisting of a sequence of micro-sp h eres of dierent size w hic h give rise to complex dendritic struc tures (Figgs. 8d, e, f): it is evident t h at goet hite fossilized biogenic structures, probably t h ose responsible for t h e oxidation of t h e ore bodies. Coronadite (Ant h ony et al., 1997a), is t h e most pe culiar and rare mineral of t h e Bd samples. is Pb and Mn oxides is rat h er common wit hin t h e Sout h America ore bodies even if it was rarely well-c h aracterized. Its name comes from t h e rst Spanis h explorer of t h e American nokite, starkeyite and smikite) are new for t h e cavern environment. Rozenite (Fig. 7c) and starkeyite (Fig. 7d) are always deeply mixed toget h er giving rise to millimetre sized, eart h y, milky w hite sp h eres wit h a cotton ball wrinkled surface. Bot h t h ese mineral are monocline and exhibit very similar crystallograp hic constants (rozenite: a o = 5.799 b o = 13.650 ; c o = 7.977 = 90.43; starkey ite: a o = 7.902 b o = 13.594 ; c o = 5.920 = 90.89). eir X-ray powder diraction patterns are rat h er co incident because t h e t h ey are isostructural (Baur, 1960, 1961; Jambor & Traill, 1963). erefore t h eir identica tion was possible only analysing t h e same samples of t h e Gandol camera by EDS microprobe on ESEM. Figure 6e s h ows alternating bands respectively ric h in rozenite or in starkeyite. It h as been extremely dircult also to identify t h e t hree mono h ydrated sulp h ates szmolnokite (FeSO 4 H 2 O), kieserite (MgSO 4 H 2 O) and szmikite (MnSO 4 H 2 O). ey were undistinguis h able at t h e stereo-microscope, because all t h ese t hree minerals give rise to small eart h y aggregates of micrometric crystals wit h slig h tly dierent colours (pale brown masses: szmolnokite; milky w hite, greasy sp h eres: kieserite; botryoidal milky w hite to pale pink masses: szmikite). erir X-ray powder diraction are very similar, because all t h e t hree minerals belongs to t h e same group and h ave similar crystallograp hic con stants (szmolnokite: a o = 7.624 b o = 7.468 ; c o = 7.123 = 115.9; kieserite: a o = 7.511 b o = 7.611 ; c o = 6.921 = 116.17; szmikite: a o = 7.766 b o = 7.666 ; c o = 7.120 = 115.85). eir identication was made by c h emical analyses performed on t h e same samples used for t h e X-ray diraction: szmolnokite and szmikite proved to be very rare and h ave been identied surely only once, w hile kieserite is muc h more abundant. Jarosite, (Ant h ony et al., 2003f) t h e iron h ydroxylsulp h ate belonging to t h e alunite group, is rat h er com mon as eart h y lemon yellow so crumbly grains (Fig. 8a) or as minutely crystalline crust over metallic grains over w hic h oen diagenetic calcite and dolomite may be found. Woodrurte (Ant h ony et al., 1997d), a compound w h ic h is c h aracteristic of t h e oxidized zone of Ag-Pb-Zn deposits, h as been h ere reported for t h e rst time as cave mineral. It is anyway very rare and it occurs as tus of t h in small, silver s h ining to old gold yellow, blades (Fig 8b). e last of t h e ve new cave minerals, orientite (Ant h ony et al., 1995), is a Mn and Ca h ydrous silicate present as t hin silver blades, w hic h at hig h er enlargement proved to be tus of t hin acicular crystals (Fig. 8c): it is always mixed to woodrurte and t h ey were found exclu sively in a small lens lling a crack of t h e sample taken in t h e wall fracture close to t h e spot 2. Fig. 8: ESEM images: a) detail of a earthy lemon yellow globular aggregate of jarosite; b) thin emi-transparent tabular crystals of woodrute; c) tus of acicular silver grey to gold shining orientite crystals; d) thin layer of goethite reddish-brown to dark grey micro-spheres: most of them are empty inside, clearly fossilizing biological masses; e) strange aggregate of goethite micro-spheres; f) detail of the structure of the goethite consisting of thin small blades resembling a wool skein. P AOLO F ORTI E RMANNO G ALLI & A NTONIO R OSSI

PAGE 7

ACTA CARSOLOGICA 36/3 2007 385 sout h west, Francisco Vasquez de Coronado. Coronadite was described as cave mineral only recently from a single cave, Santa Barbara mine cave, Italy (Forti et al., 2005). In Cuevas de las Velas it occurs as: a) radial aggregates (Fig. 9a) of microscopic tabular blades wit h colour c h anging from s hining gold yellow at t h e bottom to s hining silver grey at t h e top; or b) as eart h y so deep grey to reddis h material. At strong enlargement (Fig. 9b) t h e tabular blades of t h e radial aggregates consist of a t hick network of t hin elongated iso-oriented micrometric crystals (Fig. 9c), w hile brous radial aggregates of prismatic crystals, far s h orter t h an t h e previous ones, constitute t h e eart h y material (Fig. 9d). In a single sample it was possible to detect opal as small sp h eres oen growing over t h e Feoxidesh ydroxides. From paragenetic point of view no deposition se quence can be dened wit hin t h e Wf samples: it is t h ere fore hig h ly probable t h at all t h e 15 dierent minerals are presently growing simultaneously. On t h e contrary t h e Bd samples evidenced a clear depositional sequence. Goet hite and coronadite are t h e more abundant and rst deposited minerals, t h en opal and calcite started forming just w h en t h e deposition of t h e oxides was close to t h e end. Aer t h at eu h edral celes tine crystals started to develop t h us covering t h e pre-ex isting minerals. Finally w h en t h e deposition of celestine was over gypsum begun to form, later giving rise to t h e development of t h e giant gypsum crystals. D ISCUSSION From t h e mineralogical point of view t h e black deposits s h ow a far lower variability (6 minerals) if compared wit h t h at of t h e samples coming from t h e cave walls fractures (15 minerals). Moreover only two minerals are peculiar of t h e black deposits (coronadite and opal), w hile t h e ot h er 4 are present also in t h e cave wall fractures. Among t h e six minerals of t h e black deposits 5 are very common: only coronadite may be considered a rare cave mineral, h aving been detected until now in a single cavern envi ronment (Forti et al., 2005). e strong oxidation processes of t h e ore bodies are testied by t h e t hickness of t h e black deposits and by t h e h uge amounts of biogenic structure fossilized inside t h em. Anyway t h e scarcity of mineral variability wit hin t h e black deposits are clearly a direct consequence of t h e fact t h at t h ese processes occurred wit hin a h uge t h ermal reservoir, w h ere no s h arp variation of pH, nor supersat uration wit h respect to soluble salts may be induced by oxidizing processes. W hile goet hite and coronadite are direct products of t h ese processes, t h e opal formation was induced by a even scarce pH lowering caused by t h e oxidation of t h e H 2 S to H 2 SO 4 Later, t h e increase of SO 4 2caused t h e su persaturation wit h respect to Celestine and aer to gyp sum, t h us conrming t h e mineralogical sequence of Fig ure 9f. Finally calcite is always very rare and it is present as small aggregates of crystals wit h complex s h apes: its ori gin was likely controlled by variation in t h e activity of t h e biogenic masses ruling t h e oxidation of t h e ore bodies, w hic h caused simultaneous variation in CO 2 concentra tion wit hin t h e t h ermal water. e number of minerals (15) present in t h e pow der scratc h ed from t h e cave wall and from its fractures is very hig h and testies t h e minerogenetic erciency of t h e oxidation processes of t h e primary minerals, dispersed in t h e h ost rock, induced by t h e presence of an atmosp h ere. Unlikely to t h ose of occurred in t h e black deposits, t h e oxidation processes wit hin t h e wall fractures are abso lutely uncontrolled by microorganisms, and t h erefore no biomasses h ave been observed inside t h em. All t h e wall fracture reactions occur wit hin very small amount of condensation water, w hic h is in turn subject to fast evaporation due to t h e forced ventilation of t h e mine. erefore t h e supersaturation is reac h ed not only for low solubility sulp h ates like anglesite, celestine, jarosite and gypsum, but also for some very soluble com pound: it is t h e case of szmolnokite, kieserite, smikite, rozenite, and starkeyite. Moreover, in suc h an environ ment, t h e variation in t h e p h ysico-c h emical parameters is very hig h and t h e pH oen reac h es values close to 1 or 0 (Forti & Salvatori, 1988; Forti & Mocc hiutti, 2004). is fact induces t h e deposition of minerals normally unstable in a cavern environment like jarosite. In order to understand t h e deposition of rare and/or unusual minerals, anot h er point h as to be considered: t h e very s h ort time span since t h e dierent minerogenetic reactions were active (less t h an 20 years). is fact justify t h e presence of only two crystalline oxide (goet hite and woodrurte), w hile most of t h e iron and manganese is still in amorp h ous oxide-h ydroxides compounds. More over t h e scarcity of time hindered many elements (in par ticular iron and manganese) to reac h t h e hig h est possible oxidation, t h us inducing t h e deposition of minerals in THE MINERALOGICAL STUD Y ON THE CUEVA DE LAS VELAS NAICA, ME X ICO

PAGE 8

ACTA CARSOLOGICA 36/3 2007 386 Fig. 9 ESEM images: a) ower consisting of thin bended blades of coronadite, the color of which is gold yellow at the basis and shining silver on top; b) radial tus of small prismatic acicular crystals of coronadite; c) detail of a coronadite tu showing the quadrate section of the prismatic crystals; d) detail of the earthy coronadite structure; e) transparent prismatic tabular celestine crystals with overgrowth of iso-oriented acicular gypsum crystals and goethite micro-spheres; f) Small opale cavity covered by laments and micro-spheres of goethite on the le, and dendritic aggregates of coronadite on the right. w hic h t h ese elements h ave a lower valence (i.e. szmol nokite, smirkite, rozenite, coronadite, orientite). e presence of calcite and, wit h a biogenic me diation, of dolomite is t h e normal consequence of CO 2 STAGES IN THE DEVELOPMENT OF THE CAVE On t h e basis of t h e actual knowledge on t h e mineralogy of t h e Cueva de las Velas it is now possible to reconstruct t h e main evolutionary steps, t h e cavity underwent from its genesis up to t h e present days. Its rst stage of development, w hic h must be clearly contemporary to t h at of t h e ot h er 3 caves of Naica, h as to be referred to t h ermal water upli along t h e main faults, w hic h, still now, are responsible for t h e water circulation inside t h e Naica structure. ese faults partially displaced t h e mineral deposits and t h erefore t h ey are far younger t h an t h e ore bodies, w hic h started to develop 26,2-25,9 My BP (Megaw et al., 1988). e existence inside t h e ore bodies of dierent kinds of uid inclusion wit h temperature ranging from 680 to 130C (Erwood et al., 1979) suggest t h at, until t h e tem perature of t h e t h ermal uids was hig h, no karst void was developed: during all t his period, in fact, t h e t h ermal uids were in a condition of net deposition or at least of balance between deposition-corrosion processes, as conrmed by t h e scarcity and t h e small size of t h e open voids inside t h e mineral bodies. Surely t h e cooling down of t h e uids lasted a very long time due to t h e fact t h at no external spring was related to t h em and t h e contribute of meteoric seepage scarce if any. Aer t h e end of ore bodies development tectonic movements took place causing t h e partial displacement of t h e mineral deposits: t h ese displacements were con trolled by t h e same faults w hic h later allowed t h e water upli w hic h gave rise to t h e speleogenesis of t h e Naica caves and t h en t h e development of t h eir gigantic gypsum crystals. P AOLO F ORTI E RMANNO G ALLI & A NTONIO R OSSI diusion in a solution saturated wit h respect to gypsum (Forti et al., 2007; Vasconcelos et al., 1995). Finally t h e fast evaporation of small volumes of water may be t h e cause of t h e deposition of uorite, t h e uorine ions coming from t h e widespread uorite wit hin t h e mineralised masses. It is not possible to give a depositional c hronology, like t h at obtained for t h e black deposits, for t h e samples scratc h ed from t h e cave walls: t his because eac h powder grain is composed by a single mineral or, eventually, by t h e minerals of t h e same group. It is hig h ly probable t h at t h e number of t h e mineral actually forming wit hin t h e fractures of t his cave s h ould be hig h er, in fact a preliminary mineralogical analysis performed over a few scratc h ed samples from fractures of t h e Ojo de la Rejna cave (a few tens of meter far from Cueva de las Velas) evidenced t h e presence not only of several already observed minerals (gypsum, starkeyite, calcite, coronadite) but also of four new ones (bloedite, quartz, bassanite, and epsomite).

PAGE 9

ACTA CARSOLOGICA 36/3 2007 387 erefore t h e rst stage of t h e karst development s h ould h ave started only a few millions years BP, w h en t h e c h aracteristics of t h e h ot waters, wit h temperatures lower t h an 130C, allowed corrosion to prevail deposition. is rst stage of deep karst development was surely s h ort and t h e corrosion process not very eective: in fact, t h e presently known caves are small and always t h ey cor respond to scarcely widened fractures (Ojo de la Reina and Cueva de las Espadas) or bedding planes (Cueva de las Velas). Aer t his rst speleogenetic stage, w hic h was com mon to all t h e 4 caves of Naica, t h eir evolution was slig h t ly dierent from eac h ot h er, depending on local factors. Cueva de las Velas was interested by widespread oxi dation of sulp hide minerals w hic h were transformed into oxides-h ydroxides: t his process was clearly controlled by micro-organisms, as testied by widespread biogenic structures preserved wit hin t h e deposits (Fig. 8d.e). During t his stage, w hile a t hick black deposit cov ered t h e cave oor and most of its walls, t h e roof clearly underwent acid aggression as testied by t h e corrosion cupolas still visible in t h e rst part of t h e cavity (Fig. 6). During t h e latest stage of oxides-h ydroxides de position, t h e upliing water became oversaturated wit h respect to celestine and, later, wit h respect to gypsum: t his sequence is clearly recorded in t h e black deposit as s h own by ESEM-EDS (Fig. 9f). e appearance of gypsum crystals is an indirect proof t h at t h e temperature of t h e t h ermal water reac h ed a value lower t h an 59C, because above t his value an h y drite s h ould be t h e single forming mineral (Garca-Ruiz et al., 2007). W h en t h e sulp hide oxidation was over, t h e devel opment of t h e giant gypsum crystals started: t his pro cess was completely controlled by t h e an h ydrite-gyp sum equilibrium and t h e needed calcium sulp h ate was provided by t h e slow dissolution of t h e an h ydrite lenses widespread wit hin t h e h ost rock. e existence of a lot of small gypsum crystals wit hin t h e black deposits, w hic h acted as crystallization nuclei, justies t h e fact t h at, despite t h e new nucleation probability was extremely low (Garca-Ruiz et al., 2007) Cueva de las Velas is t h e single cave of Naica in w hic h only a few of giant crystals developed, w hile most of its oor and walls are covered by a very hig h number of rela tively small crystals (10-20 cm in size). e gypsum deposition went on until, some 20 years ago, t h e mine exploitation caused t h e complete dewater ing of t h e cave: anyway t his fact did not represent t h e end of t h e development of t h e cave, w hic h was c h aracterized by two furt h er steps. e rst one started just during t h e dewatering of t h e cave giving rise to t h e sails (Bernabei et al., 2006), but it stopped immediately aer t h e cave was completely dried. e latest stage, w hic h is still active now, was in duced by t h e presence of moist air, t h e condensation of w hic h induced t h e diagenesis (oxidation) of some of t h e ore bodies exposed on t h e cave walls. is process, even extremely young and still in progress, allowed t h e development a large number of minerals, ve of w hic h are completely new for t h e cavern environment. Because t his stage is induced by t h e direct contact between rock and air, it is hig h ly probable t h at t h e same process is ac tive also in ot h er caves of t h e -290 level or even deeper wit hin t h e Naica Mine: next investigation will test suc h an h ypot h esis. THE MINERALOGICAL STUD Y ON THE CUEVA DE LAS VELAS NAICA, ME X ICO e mineralogical study of t h e c h emical deposits of Cue va de las Velas wit hin t h e Naica mine h as evidenced t h e existence in t h e cave history of two distinct periods in w hic h several cave minerals were deposited mainly due to t h e oxidation of t h e ore bodies. In t h e rst one, w hic h occurred deep inside t h e t h ermal aquifer before t h e dep osition of t h e giant gypsum crystals, a large quantity of material was deposited but, due to t h e scarce variability of suc h an environment, only a few minerals developed. e second one, w hic h started under aerate conditions less t h an 20 years ago and it is still active now, induced t h e deposition of scarce material but wit h an extremely hig h mineralogical variability. From t his point of view t h e Cueva de Las Velas is an extremely important cave because it represent a natu ral laboratory in w hic h can be experimentally tested t h e boundary conditions for t h e development of several even extremely rare cave minerals. Unfortunately t h is cave, as all of t h e ot h er karst p h enomena at t h e 290 level of Naica mine will remain visible only for a few years, and as soon as t h e mining activities will stop (an event t h at is expected wit h in 7-10 years), t h e upliing of groundwater will submerge t h em under some 170 m of water. It is t h erefore important t h at all t h e studies still in progress on Naica and its in credible mine caves will be completed in t h e s h ortest possible time in order to preserve for future generations at least a full knowledge of t h ese astonis h ing natural p h enomena. F INAL REMARKS

PAGE 10

ACTA CARSOLOGICA 36/3 2007 388 A CKNOWLEDGEMENTS e Aut h ors t h anks Peoles Company for allowing t h e access inside t h e Naica Mine and for any kind of h elp giv en during t h e eld work, Roberto Villasuso for t h e useful geological discussions and doctor Pier Luigi Fabbri of t h e Centro Interdipartimentale Grandi Strumenti (C.I.G.S.) of t h e University of Modena and Reggio Emilia for t h e precious h elp in t h e ESEM-EDS analyses. R EFERENCES Ant h ony J.W., R.A. Bideaux, K.W. Blad h, & M.C. Nic h ols, 1995: H andbook of M ineralogy Vol. II (Silica, Sili cates). Mineral Data Publis hing, Tucson, Arizona. Ant h ony J.W., R.A. Bideaux, K.W. Blad h, & M.C. Nic h ols, 1997: H andbook of M ineralogy Vol. III (H a lides, H ydroxides, Oxides). Mineral Data Publis h ing, Tucson, Arizona: a)138; b) 205; c) 223. Ant h ony J.W., R.A. Bideaux, K.W. Blad h, & M.C. Nic h ols, 2003: H andbook of M ineralogy Vol. V (Borates, Carbonates, Sulfates). Mineral Data Publis hing, Tucson, Arizona: a) 24; b) 101; c) 122; d) 191; e) 271; f) 330; g) 358; h) 602; i) 663; l) 687; m) 688. Baur, W.H., 1960: Die Kristallstruktur von FeSO 4 H 2 O. Naturwissenschaen 47, 467. Baur, W.H., 1961: Die Kristallstruktur des Leon h ardtits, MgSO 4 H 2 O, und des Rozenits, FeSO 4 H 2 O. Forts. M in., 39, 333-334. Badino G. & P. Forti, 2007: e Exploration of t h e Caves of t h e Giant Crystals (Naica, Mexico) NSS News 65 (2), 12-18. Bernabei T., P. Forti, & R. Villasuso, 2006: A new type of gypsum crystal from Naica (Mexico). Int. J of Spele ology 36(1), 23-30. De Waele J. & A. Naseddu (Ed.), 2005: Le grotte di miniera tra economia mineraria ed economia turistica. IIS Mem. X VII. Degoutin N., 1912 : Les grottes a cristaux de gypse de Naica. Soc. Cient. Antonio Alzate Rev. 32, 35-38. Erwood, R.J., S.E. Kesler, & P.L. Cloke, 1979: Compo sitionally distinct, saline h ydrot h ermal solutions, Naica Mine, C hi h ua h ua, Mexico. Economic Geology 74, 95-108. Forti P., 2006: Una foresta di cristalli di gesso nel pro fondo della miniera di Naica. Geoitalia 18, 29-34. Forti P. & A. Mocc hiutti, 2004: Le condizioni ambientali c h e permettono levoluzione di speleotemi di zolfo in cavit ipogenic h e:nuovi dati dalle grotte di Capo Palinuro (Salerno, Italia). Grotte dItalia 5, 4 (2003), 39-48. Forti P. & S. Salvatori, 1988: Nuovi minerali di grotta: la broferrite di Grotta Ferrata. Riv. M ineral. Italiana 1988, 4, 219-226. Forti P., E. Galli, & A. Rossi, 2006: Peculiar mineroge netic environments: t h e Cuatro Cinegas karst area (Co h auila, Mexico) Acta Carsologica 35,1, 79-98. Forti P., E. Galli, & A. Rossi, 2007: Il sistema Gesso-Cal cite-Aragonite: nuovi dati dalle concrezioni del Liv ello -570 della Miniera di Naica (Messico). Congr. Naz. Speleol., Iglesias, in press. Forti P., A. Pagliara, E. Galli, A. Rossi, J. De Waele, A. Naseddu, & S. Papinuto, 2005: Studio morfologico e mineralogico di dettaglio del concrezionamento del sistema carsico di Santa Barbara (Miniera di San Giovanni). Atti Simposio Le grotte di miniera tra economia mineraria ed economia turistica, Iglesias 2004, IIS Mem. X VII, 57-68. Fos h ag, W., 1927: e selenite caves of Naica, Mexico. Amer. M ineral., 12, 252232. Garca-Ruiz J.M., R. Villasuso, C. Ayora, A. Canals, & F. Otlora, 2007: e Formation of Gypsum Megac rystals. Geology 35,4, 327-330 Hill C.A. & P. Forti, 1997: Cave minerals of the World. National Speleological Society, Huntsville. Jambor, J.L. & R.J. Traill, 1963: On rozenite and siderotil. Canad. M ineral. 7, 751-763. Lang, J.R., 1995: A geological evaluation of the Naica de posit, Chihuahua, M exico. Internal Report of Com paia Fresnillo. 109 pp. Megaw, P.K.M., J. Ruiz, & S.R. Titley, 1988: Hig h-temper ature, carbonate-h osted Pb-Zn-Ag (Cu) deposits of nort h ern Mexico. Economic Geology 83, 1856-1885. S h agun C.L., 2001: Naicas subterranean Marvels. NSS News 59, 6, 166-169. Stone, J.G., 1959: Ore genesis in t h e Naica District, C hi h ua h ua, Mexico. Economic Geology 54, 1002-1034. Vasconcelos C., J.A. Mckenzie, S. Bernasconi, D. Crujic, & A. Tiens, 1995: Microbial mediation as a possible mec h anism for natural dolomite formation at low temperature. Nature 377, 220-222. P AOLO F ORTI E RMANNO G ALLI & A NTONIO R OSSI



PAGE 1

SUBMERGED SPELEOTHEMS E X PECT THE UNE X PECTED. E X AMPLES FROM THE EASTERN ADRIATIC COAST CROATIA P OTOPLJENA SIGA PRESENETLJIVI REZULTATI ANALIZ VZORCEV IZ VZHODNE OBALE J ADRANA H RVAKA Maa S URI 1 Branko J ALI 2 Donat P ETRICIOLI 3 Izvleek UDK 551.44(497.5) Maa Suri, Branko Jali & Donat Petricioli: Potopljena siga presenetljivi rezultati analiz vzorcev iz vzhodne obale Jadra na (Hrvaka) Z namenom rekonstrukcije sprememb morske gladine v poz nem Pleistocenu in Holocenu v vzh odnem Jadranu, smo vzorili sigo v potopljeni h jama h in brezni h na razlini h lokacija h. Pri tem smo predvideli, da so obdobja rasti in stagnacije sige povezana s poloajem morske gladine. Ker je prirastek stalag mitov vedno pravokoten na obstojeo podlago, nudijo stalag miti primerneji zapis dogodkov kot stalaktiti. zato smo vzorili predvsem stalagmite. Pri dve h izmed sedemnajsti h vzorcev, L-1, nabran na globini 1,5 m v Medvei jami na otoku Loinj in vzorec M-25, vzet iz globine 25 m v breznu pri otoku Iki Mrtovnjak, smo opazili precej nenavadno notranjo strukturo. Oba sta bila na jamskem dnu v poloaju stalagmita, a sta tako preni in podolni rez pokazala, da gre v resnici za stalaktit. To so potrdile tudi Uin 14 C datacije. Vzrok in nain padca obe h kapnikov je bil razlien. M-25 je skupaj z ve kapniki obtial v morskem sedimentu v izvirni legi, medtem ko je bila lega L-1 obrnjena, kapnik pa je e naprej rastel v obdobju nizke gladine morja. Podobne dogodke la h ko priakujemo tudi v jama h na kopnem. V vsakem primeru pa take kapnike v podmorskem okolju teje opazimo saj sta nji h ova prvotna oblika in poloaj velikokrat zamaskirana z razlinimi morskimi sedimenti. Kljune besede: potopljeni kapniki, Jadransko morje, H r vaka. 1 Department of Geograp h y, University of Zadar, Tumanova 24 i, 23000 Zadar, Croatia, Fax.: +385 23 311 282, e-mail: msuric@unizd.hr 2 Croatian Natural History Museum, Demetrova 1, 10000 Zagreb, Croatia, e-mail: bjalzic@ya h oo.com 3 DIIV, Obala Petra Lorinija bb, 23281 Sali, Croatia, e-mail: dpetricioli@diiv.hr Received/Prejeto: 11.08.2007 COBISS: 1.01 ACTA CARSOLOGICA 36/3, 389-396, POSTOJNA 2007 Abstract UDC 551.44(497.5) Maa Suri, Branko Jali & Donat Petricioli: Submerged spe leothemes expect the unexpected. Examples from the eastern Adriatic coast (Croatia) Wit h t h e intention of reconstructing Late Pleistocene Holo cene sea-level c h anges along t h e Eastern Adriatic coast, a series of speleot h ems were collected from several submerged caves and pits, in order to constrain periods of t h eir deposition and ceased growt h related to sea-level uctuations. For t h at pur pose, stalagmites provide more reliable records t h an stalactites, due to t h eir successive layers deposited perpendicularly to t h e growt h direction. erefore, stalagmites h ave been collected preferably. But, two of 17 speleot h ems displayed unexpected interior morp h ology speleot h em L-1 collected at t h e dept h of 1.5 m in Medvjea spilja Cave on Loinj Island, and speleot h em M-25 from Pit near Iki Mrtovnjak Islet collected at t h e dept h of 25 m. Bot h of t h e samples were taken from t h e cave oor, in t h e growt h position of t h e stalagmite. But t h e insig h t into t h e perpendicular cut wit h evident central tube revealed t h eir true (stalactitic) origin and additional conrmations were obtained by longitudinal cut and Uand 14 C dating. Just as t h e causes of t h eir breakdowns were probably dierent, so were t h eir falls; speleot h em M 25 (toget h er wit h several ot h er speleot h ems around it) stuck in t h e marine sediment in its primary posi tion, w hile L-1 turned upside-down and even continued crys tallizing during t h e lower sea level. ese events are possible in t h e continental caves, as well. Evidently, it is muc h easier to recognize and avoid t h ese problems in air-lled caves t h an in t h e submarine ones w h ere t h e speleot h ems are almost always covered wit h marine overgrowt h, w hic h disguises t h eir outer morp h ology. Additionally, t h e bases of t h e stalagmites are also sometimes covered wit h marine sediment, w hic h makes correct estimation rat h er dircult. Keywords: submerged speleot h ems, Adriatic Sea, Croatia.

PAGE 2

ACTA CARSOLOGICA 36/3 2007 390 Speleot h ems are typical subaerial features, so t h eir oc currence in presently submerged (submarine) caves is irrefutable evidence of former low sea levels. Since t h e 1970s (Spalding & Mat h ews, 1972; Harmon et al., 1978; Gascoyne et al., 1979), a series of exploration h ave been based on submerged speleot h ems as indicators of low sea-level stands. e.g. on Ba h amas (Ric h ards et al., 1994; Lundberg & Ford, 1994), on Balearic Islands (Forns et al., 2002; Onac et al., 2006), on t h e Eastern Adriatic coast (Vr h ovec et al., 2001; Suri et al., 2005a), etc. Hig h sea-level stands, i.e. t h e latest possible time of cave ood ing caused by sea-level rise can be determined by dating biogenic overgrowt h t h at usually covers t h e speleot h ems in marine environment (Alessio et al., 1992; Antonioli & Oliverio, 1996). Furt h ermore, if submerged speleot h ems contain remnants of marine organisms wit hin t h e spe leot h em body, episodes of hig h sea level can be revealed (Fig. 1). Based on t h at fact, 215-ka history of sea-level os cillations was reconstructed on t h e Tyrr h enian Sea coast (Bard et al., 2002; Antonioli et al., 2004). In order to constrain t h e periods of speleot h em deposition and ceased growt h connected wit h sea-level uctuations, radiometric dating of submerged speleo t h ems is usually employed. For t h at purpose, stalagmites provide better stratigrap hic resolution and more reliable records t h an stalactites due to t h eir successive layers de posited perpendicularly to t h e growt h direction (Hill & Forti, 1997). Besides, to avoid problems wit h ion mobili sation c h aracteristic for stalactites, in various geoc hrono logical studies and palaeoenvironmental reconstructions, stalagmites are collected preferably (Ric h ards & Dorale, 2003). Hence, we used suc h sampling strategy in recent palaeoenvironmental study on t h e eastern Adriatic coast wit h t h e aim to reconstruct Late Pleistocene Holocene relative sea-level c h anges (Suri, 2006). Y et, in spite of careful sampling, we encoun tered a problem of mislead ing collapsed stalactites t h at we h ad not expected. An overview of t h e researc h es on collapsed speleot h ems is given by Forti (1997), but only t h ose related to eart h quake reconstructions, not as a sampling problem. INTRODUCTION Fig. 1: Deposition of carbonate owstone in subaerial conditions during low sea-level stand (a and c) and its cessation with biogenic encrustation (red layer) in marine environment in the periods of high sea level (b and d). STUD Y AREA AND SAMPLING Eastern Adriatic coast is locus typicus of Dalmatian coast a ooded concordant coast c h aracterized by c h ains of islands formed by Late Pleistocene Holocene sea-level inundation of coast-parallel ridges and valleys (Kearey, 1996). Its genesis started in Mesozoic in form of carbon ate platform w hic h was subsequently disintegrated and uplied during t h e Alpine orogeny due to subduction of African plate under t h e Eurasian continent (Vla h ovi et al., 2002, 2005). Emerged carbonate complex was inten sively karstied and repeatedly (partially) ooded during t h e Quaternary sea-level uctuations. Late Pleistocene Holocene sea-level rise of 121 5 m (Fairbanks, 1989) submerged vast part of karstied area and formed t h e re cent coast. Along wit h ot h er presently submerged karst forms like karrens, submerged karst springs, river can yons, etc. (Suri, 2005), more t h an 230 caves were dis covered along t h e eastern Adriatic coast, and over 140 of t h em contain speleot h ems (Suri, 2006), potential mate rial for subsequent researc h es. Wit hin a larger study, in order to encompass a wide range of sea-level stands, speleot h ems were collected from t h e dept h s of 1.5 m below mean sea level (b.m.s.l.) to 41.5 m b.m.s.l. from seven submerged karst features: two caves, four pits and one vrulja (submarine spring), M AA S URI B RANKO J ALI & D ONAT P ETRICIOLI

PAGE 3

ACTA CARSOLOGICA 36/3 2007 391 all located along t h e Croatian coast (Suri, 2006). Sampling was provided by SCUBA divers. Among 17 collected speleot h ems, two of t h em s h owed unexpected interior morp h ology. ose were spe leot h em L-1 from Medvjea spilja Cave on Loinj Island, and speleot h em M-25 from t h e Pit near Iki Mrtovnjak Islet (Fig. 2). Fig. 2: Study area with the investigated locations: M edvjea spilja Cave on Loinj Island and P it near Iki M rtovnjak Islet. MEDVJEA SPILJA CAVE Medvjea spilja Cave, situated on t h e eastern coast of Loinj Island, was formed along vertical ssure in Upper Cretaceous well bedded limestones (Jali, 2005). It is an anc hialine cave wit h t h e entrance 17.5 m above mean sea level, 55 m from t h e coast (Fig 3a). Presumed primary entrance is 8 m b.m.s.l., buried wit h collapsed material, and t h e circulation wit h t h e open sea, wit h evident tidal oscillations, is establis h ed t hroug h ssures and c h annels in t h e karstied bedrock. Speleothem L-1 (21.5 cm long) was collected from t h e dept h of 1.5 m b.m.s.l. in t h e growt h position of sta lagmite, c h osen between several similar speleot h ems from t h e same base ow stone. But already perpen dicular section wit h central tube (Fig. 3c) s h owed t h at t h e origin of t h at speleot h em was stalactitic, and longitudinal cut conrmed it, as well (Fig 3b). According to t h e dier ences in morp h ology and t h e colour of t h e carbonate, we can presume two growt h p h ases. e inner part was deposited as a stalactite, w hile t h e outermost brig h t part of t h e speleot h em grew in t h e acicular (needle-like) form, apparently in dierent con Fig. 3: a) Cross-section of M edvjea spilja Cave showing the l o cation of speleothem L-1. SUBMERGED SPELEOTHEMS E X PECT THE UNE X PECTED. E X AMPLES FROM THE EASTERN ADRIATIC COAST CROATIA

PAGE 4

ACTA CARSOLOGICA 36/3 2007 392 According to t h e radiometric data and inner and outer morp h ology of t h e speleot h ems L-1 and M-25, we can assume t h e scenarios of t h eir evolution as follows: Speleot h em L-1 grew in t h e form of stalactite at least during MIS 5. Aerwards, it appears t h at t h e part of t h e owstone, toget h er wit h several adjacent stalactites, ditions. Major part of t h e central tube of t his speleot h em was pre served unlled, w hic h can be traced all along t h e longitudinal sec tion as a small c h annel. Only at t h e top of t h e speleot h em ca 0.5 cm of t h e central tube was lled wit h t h e youngest carbonate t h at forms needle-like cover. Lack of marine overgrowt h 3c 3d Fig. 3: b) longitudinal section of L-1 with obtained ages in ka (scale bar 10 cm); c) perpendicular section with central tube (scale bar 5 cm); d) penetration of younger carbonate into the central tube. indicates alternating fres h water/brackis h conditions in t his s h allow part of t h e cave. Obtained Uage of 120.4 ka of inner part (Fig. 3b) suggests t h at t h e primary position of t h e speleot h em L-1 in t h e rst (stalactitic) growt h p h ase was above t h e sea level associated wit h t h e MIS 5 hig h stand t h at was hig h er t h an present location of speleot h em at 1.5 m b.m.s.l. (Suri, 2006). According to 14 C measurement, subsequent needle-like crystallization went on around 3.3 ka BP (Fig. 3b) (Suri, 2006). PIT NEAR IKI MRTOVNJAK ISLET Pit near Iki Mrtovnjak Islet was formed in Upper Creta ceous rudist limestones, and wit h its entrance 5 m b.m.s.l., 12 m os h ore, it is completely wit hin marine environ ment (Fig. 4a). It is ric h in speleot h ems t h at are entirely covered wit h marine organisms belonging to t h e bioce nosis of caves and ducts in complete darkness (Jurai et al., 2002). Ages of t h e analysed speleot h ems from t h e dept h s of 14 m, 19 m and 23 m b.m.s.l. range from 39 ka to 202 ka (Suri, 2006), and it seems t h at t hroug h out t h e Late Pleistocene t his pit experienced numerous submerg ings and recommencements of speleot h em deposition. Speleothem M-25 (46 cm long) was collected from t h e cave oor at t h e dept h of 25 m b.m.s.l. (Fig. 4a). It was in t h e growt h position of a stalagmite, nearby several similar, relatively long and t hin speleot h ems (Fig. 4b). e evidence of stalactitic origin of speleot h em M-25 is central tube visible in t h e perpendicular cut (Fig. 4c), as well as its s h ape somew h at wider upper part (Fig. 4d) t h at is c h aracteristic for carrot-s h aped stalactite, and not for stalagmites. e dierence in outer morp h ology be tween real stalagmites, w hic h are t hicker and s h orter, and fallen stalactites w hic h are t hinner and longer is visible on Fig. 4b. is speleot h em was completely covered wit h marine biogenic overgrowt h, including t h e uppermost part w h ere t h e speleot h em h ad broken o. 3b M AA S URI B RANKO J ALI & D ONAT P ETRICIOLI DISCUSSION AND CONCLUSIONS

PAGE 5

ACTA CARSOLOGICA 36/3 2007 393 4c SUBMERGED SPELEOTHEMS E X PECT THE UNE X PECTED. E X AMPLES FROM THE EASTERN ADRIATIC COAST CROATIA broke o from t h e roof and fell on t h e bottom in upsidedown position (Fig. 5a) w h ere t h e carbonate deposition recommenced in acicular form. Penetration of younger brig h t carbonate into t h e central tube (Fig. 3d) indicates t h at t h e breakdown and immediate recommencement of carbonate deposition probably h appened in subaerial en vironment. Finally, t h e speleot h em L-1 was submerged by t h e last, Late Pleistocene Holocene sea-level rise. Speleot h em M-25, toget h er wit h several ot h ers, aer t h e stalactitic growt h apparently fell down and di rectly stuck into t h e cave or marine mud in t h eir original position (Fig. 5b). e cause of t h eir breakdown was ei t h er of seismic origin or t h e removal of t h e buoyant sup port in t h e period of sea-level fall. Subaerial carbonate deposition aer t h e breakdown was not noticed, more over, it looks like marine overgrowt h covered t h e fres h cut soon aer t h e breakdown. Breakdowns of speleot h ems and recommencements of carbonate deposition in new position are not unusual events in caves. Rarely, speleot h ems keep vertical posi tion aer t h e fall. Y et, it is possible, so it could be mislead ing w hile sampling, especially in case of a group of fallen stalactites as in aforementioned cases. If t h e stalactites fall toget h er wit h t h e owstone layer and turn upside-down (Fig. 5a), it looks like it is t h eir primary position, and t h e carrot-like stalactites could even resemble stalagmites in s h ape. In ot h er cases, if t h e falling speleot h ems stuck di rectly in t h e mud (Fig. 5b), t h e junctures are invisible. In marine environment, speleot h em sampling is additionally complicated due to encrusting organisms. Namely, wit hin submarine caves t h e speleot h ems are very oen covered wit h marine overgrowt h, w hic h hides t h eir outer morp h ology and possible breakage. In addition, t h e bases of t h e speleot h ems on t h e cave bottom are also sometimes overlaid wit h marine sediment, w hic h makes t h e estimation w h et h er it is a real stalagmite or a fallen stalactite rat h er dircult. Marine overgrowt h also disables t h e assessment of quality of t h e speleot h em fabric w hic h is essential for most geoc h emical (and geoc hronological) analyses, since t h e boring organisms can h eavily damage speleot h em body (Fig. 6) even if t h e outer appearance of t h e sample suggests t h at it could be appropriate for t h e furt h er measurements and analyses. Finally, in palaeoenvironmental studies, especially in palaeotemperature reconstructions, one of t h e most important requirements is t h at t h e speleot h ems are col Fig. 4: a) Cross-section of P it near Iki M rtovnjak Islet showing the location of speleothem M 25; b) speleothems in P it near Iki M rtovnjak Islet 25 m b.m.s. l. Note the similar shape and dimensions of stalactite hanging from the roof and speleothems stuck in the mud (arrow points to the speleothem M-25). Unlike them, real stalagmites (front and right) are shorter and thicker; c) perpendicular section of speleothem M-25 with central tube (scale bar 5 cm); d) outer surface covered with marine overgrowth (scale bar 10 cm). 4a 4d 4b

PAGE 6

ACTA CARSOLOGICA 36/3 2007 394 ACKNOWLEDGEMENTS Ministry of Science, Education and Sport of Republic of Croatia supported t his researc h (Projects: 269-26930841177 Geographical features in the development of Croatian coastal regions, 119-0362975-1226 Research of threatened marine habitats in the coastal submerged karst of Croatia, 119-1191152-1169 Recent sediments and fossil environ ments of the Adriatic coastal zone). We express our grati tude to Sreko Trajbar and Vedran Jali for t h e h elp during eld work and for drawings. We would also like to t h ank t h e reviewers for useful comments and sugges tions. lected from t h e deep interior of t h e cave, in order to avoid t h e eect of kinetic isotopic fractionation and obtain h o mogenized 18 O signal (Ford, 1999). But, due to t h e ob jective risks of speleo-diving in distant parts of t h e caves, speleot h ems in t his study were mostly sampled from ap proac h able parts, so t h e recorded 18 O signal could not h ave been regarded as palaeotemperature record (Suri et al., 2005b). Fig. 5: Stalactites in the growth positions of stalagmites aer the breakdown: a) overturned during the fall; b) stuck directly in primary position. Fig. 6: Longitudinal section of stalagmite B-28 from the P it in Luice Bay on Bra Island (from the depth of 28 m b.m.s.l.) heavily devastated by boring marine organisms (scale bar 5 cm) (from Suri, 2006). In conclusion, all t h e limits and problems present in t h e process of speleot h em sampling in air-lled caves are even more pronounced wit hin t h e submerged ones, and wit h t h e addition of unexpected traps, submarine explorations can sometimes be quite uncertain and in decisive. Of course, by cautious approac h and attentive sampling, aforementioned situations could be recognized and avoided. M AA S URI B RANKO J ALI & D ONAT P ETRICIOLI

PAGE 7

ACTA CARSOLOGICA 36/3 2007 395 SUBMERGED SPELEOTHEMS E X PECT THE UNE X PECTED. E X AMPLES FROM THE EASTERN ADRIATIC COAST CROATIA REFERENCES Alessio, M., Allegri, L., Antonioli, F., Belluomini, G., Fer ranti, L., Improta, S., Manfra, L. & Proposito, A., 1992: Risultati preliminari relativi alla datazione di speleo temi sommersi nelle fasce costiere del Tirreno cen trale.Giornale di Geologia, ser. 3., 54/2, 165-193. Antonioli, F. & Oliverio, M., 1996: Holocene sea-level rise recorded by a radiocarbon-dated mussel in a submerged speleot h em beneat h t h e Mediterranean Sea.Quaternary Researc h, 45, 241-244. Antonioli, F., Bard, E., Potter, E.K., Silenzi, S. & Improta, S., 2004: 215-ka history of sea-level oscillations from marine and continental layers in Argentarola Cave speleot h ems (Italy).Global & Planetary C h ange, 43(1-2), 57-78. Bard, E., Antonioli, F. & Silenzi, S., 2002: Sea-level dur ing t h e penultimate interglacial period based on a submerged stalagmite from Argentarola Cave (Ita ly).Eart h and Planetary Science Letters, 196 (3-4), 135-146. Fairbanks, R. G., 1989: A 17000-year glacio-eustatic sea level record: inuence of glacial melting rates on t h e Y ounger Dryas event and deep-ocean circulation.Nature, 342, 637-642. Ford, D., 1997: Dating and Paleo-Environmental Studies of Speleot h ems. In: C. Hill & P. Forti (Eds.), Cave M inerals of the World, 2 nd ed., National Speleological Society, Huntswille, 271-284. Forti, P., 1997: Speleot h ems and Eart h quakes. In: C. Hill & P. Forti (Eds.), Cave M inerals of the World, 2 nd ed., National Speleological Society, Hunt swille, 284-285. Forns, J. J., Gelabert, B., Gins, A., Gins, J., Tuccimei, P. & Vesica, P., 2002: P hreatic overgrowt h s on speleo t h ems: a useful tool in structural geology in littoral karstic landscapes. e example of eastern Mallorca (Balearic Islands).Geodinamica Acta, 15, 113-125. Gascoyne, M., Benjamin, G.J., Sc h warcz, H.P. & Ford, D.C., 1979: Sea-Level Lowering During t h e Illinoian Glaciation: Evidence from a Ba h ama Blue Hole.Science, 205, 806-808. Harmon, R.S., Sc h warcz, H.P. & Ford, D.C., 1978: Late Pleistocene Sea Level History of Bermuda.Quater nary Researc h, 9, 205-218. Hill, C. & Forti, P., 1997: Cave Minerals. In: C. Hill & P. Forti (Eds.), Cave M inerals of the World, 2 nd ed., Na tional Speleological Society, Huntswille, 117-227. Jali, B., 2005: Najdua bua Jadrana (Medvjea pilja, Loinj).More, 124, 84-89. Jurai, M., Bakran-Petricioli, T. & Petricioli, D., 2002: Cessation of Karstication Due to t h e Sea-level Rise? Case Study of t h e Y -Cave, Dugi otok, Croa tia. In: Grabovek F. (Ed.): Evolution of Karst: From P rekarst to Cessation, Zaloba ZRC, Postojna, Lju bljana, 319-326. Kearey, P., 1996: e New P enguin Dictionary of Geology, Penguin Books, London, p. 366. Lundberg, J. & Ford, D. C., 1994: Late Pleistocene sea lev el c h ange in t h e Ba h amas from mass spectrometric U series dating of submerged speleot h em.Quater nary Science Reviews, 13 (1), 1-14. Onac, B., Forns, J.J., Gins, J., Gins, A., Tuccimei, P., Peate, D.W. & Bjrck, S., 2006: Sea-level position at ~80 ka based on p hreatic overgrowt h s on speleo t h ems from Mallorca. In: Onac, B., Tma, T., Con stantin S., Peroiu, A. (Eds.): Archives of Climate Change in Karst, Proceedings of t h e symposium Climate C h ange: e Karst Record (IV), 26-29 May 2006, Bile Herculane, Romania, Karst Water Insti tute, 189-191. Ric h ards, D. A., Dorale, J. A., 2003: Uranium-series C hronology and Environmental Applications of Speleot h ems. In: Bourdon, B., Henderson, G. M., Lundstrom, C. C., Turner, S. P., (Eds.): Uranium Series Geochemistry, Reviews in Mineralogy & Geo c h emistry, Vol. 52, Geoc h emical Society, Mineral ogical Society of America, 407 460. Ric h ards, D.A., Smart, P.L. & Edwards, R.L., 1994: Maxi mum sea levels for t h e last glacial period from U series ages of submerged speleot h ems.Nature, 367, 357-360. Spalding, R.F. & Mat h ews, T.D., 1972: Stalagmites from Caves in t h e Ba h amas: Indicators of Low Sea Level Stand.Quaternary Researc h, 2, 470-472. Suri, M., 2005: Submerged karst dead or alive? Ex amples from t h e Eastern Adriatic coast (Croatia).Geoadria, 10/1, 5-19. Suri, M., 2006: Late P leistocene H olocene palaeoenvi ronmental changes records from submerged speleo thems from the Eastern Adriatic Sea (Croatia). P hD t h esis, Faculty of Science, University of Zagreb, p. 213, Zagreb, (in Croatian). Suri, M., Jurai, M., Horvatini, N. & Krajcar Broni, I., 2005a: Late Pleistocene Holocene sea-level rise and t h e pattern of coastal karst inundation: records from submerged speleot h ems along t h e Eastern Adriatic Coast (Croatia).Marine Geology, 214 (13), 163-175.

PAGE 8

ACTA CARSOLOGICA 36/3 2007 396 Suri, M., Horvatini, N., Suckow, A., Jurai, M. & Barei, J., 2005b: Isotope records in submarine spe leot h ems from t h e Adriatic coast, Croatia.Bulletin de la Socit Gologique de France, 176/4, 363-373. Vla h ovi, I., Tiljar, J., Veli, I. & Matiec, D., 2002: e Karst Dinarides are Composed of Relics of a Single Mesozoic Platform: Facts and Consequences.Geo logia Croatica, 55/2, 171-183. M AA S URI B RANKO J ALI & D ONAT P ETRICIOLI Vla h ovi, I., Tiljar, J., Veli, I. & Matiec, D., 2005: Evo lution of t h e Adriatic Carbonate Platform: Palaeo geograp h y, main events and depositional dynam ics.Palaeogeograp h y, Palaeoclimatology, Palaeo ecology, 220, 333-360. Vr h ovec, T., Mi h evc, A., Lauritzen, S.E. & Lundberg, J., 2001: O starosti potopljeni h stalaktitov v jami pri otoku Galiola, Dalmacija, Hrvatska, (On t h e ages of t h e submerged stalactites from t h e pit near islet Galiola, Dalmatia, Croatia).Nae jame, 43, 31-36.



PAGE 1

PROPOSED METHODOLOG Y OF VULNERABILIT Y AND CONTAMINATION RISK MAPPING FOR THE PROTECTION OF KARST AQUIFERS IN SLOVENIA PREDLOG METODOLOGIJE KARTIRANJA RANLJIVOSTI IN TVEGANJA ZA ONESNAENJE VODA ZA VAROVANJE KRAKIH VODONOSNIKOV V SLOVENIJI Nataa R AVBAR 1 & Nico G OLDSCHEIDER 2 Izvleek UDK 551.44:556.388(497.4) Nataa Ravbar & Nico Goldscheider: Predlog metodologije kartiranja ranljivosti in tveganja za onesnaenje voda za va rovanje krakih vodonosnikov v Sloveniji Upotevajo posebnosti slovenskega krasa smo na podlagi smernic evropskega projekta COST Action 620 predlagali sploen pristop k ocenjevanju ranljivosti in tveganja kraki h voda za onesnaenje. Tako imenovani Slovenski pristop us treza slovenski okoljski zakonodaji in omogoa primerjavo z razmerami v Evropi. Metoda pri ocenjevanju ranljivosti pod talnice upoteva asovno hidroloko spremenljivost, ponuja monost povezovanja zaite povrinski h in podzemni h voda ter predlaga nove smernice za zaito podzemne vode in vod ni h virov, za kar je bil razvit nov faktor K (pretakanje kraki h voda v zasieni coni). Slovenski pristop predvideva obirno analizo tveganja, ki temelji na oceni naravne ranljivosti, dejan ski h in potencialni h obremenjevalcev ter pomembnosti vod nega vira oziroma podzemne vode. Predlagana metodologija je bila prvi uporabljena v zaledju vodnega vira Podstenjek v jugoza h odni Sloveniji, rezultati pa preverjeni s pomojo dve h sledilni h poizkusov, ki sta skupno zajemala est injicirni h tok. Rezultati kart naravne ranljivosti, obremenjevalcev in tveganja so zadovoljivi. Preverjanje potrjuje ocenjeno naravno ranljivost reprezentativni h tok, izbrani h za injiciranje sledila. Konne karte omogoajo izpopolnjeno razmejitev vodovarstveni h pa sov ter oznaujejo obmoja neustreznega ravnanja, nudijo pod lago za reorganizacijo dejavnosti in za bolje reitve v pri h odnjem nartovanju. Kljune besede: Slovenski pristop, ranljivost podzemne vode, tveganje za onesnaenje, kraki vodonosnik, zaita in upravljanje z vodnimi viri, preverjanje kart ranljivosti. 1 Karst Researc h Institute SRC SASA, Titov trg 2, SI-6230 Postojna, Slovenia, e-mail: natasa.ravbar@zrc-sazu.si 2 Centre of Hydrogeology, University of Neuc h tel, 2009 Neuc h tel, Switzerland, e-mail: nico.goldsc h eider@unine.c h Received/Prejeto: 14.09.2007 COBISS: 1.01 ACTA CARSOLOGICA 36/3, 397-411, POSTOJNA 2007 Abstract UDC 551.44:556.388(497.4) Nataa Ravbar & Nico Goldscheider: Proposed methodology of vulnerability and contamination risk mapping for the pro tection of karst aquifers in Slovenia On t h e basis of work accomplis h ed by t h e European COST Action 620, a compre h ensive approac h to groundwater vulner ability and contamination risk assessment is proposed, taking into account t h e special c h aracteristics of Slovene karst aqui fer systems. e Slovene Approac h is consistent wit h national environmental legislation and enables comparison across Eu ropean countries. e met h od integrates temporal h ydrologi cal variability in t h e concept of groundwater vulnerability and oers a new possibility to combine surface and groundwater source and resource protection, w hic h required t h e develop ment of a new K factor (karst groundwater ow wit hin t h e saturated zone). e risk analysis considers intrinsic vulnera bility, contamination h azards and t h e importance of t h e source or resource. It h as been rst applied to t h e Podstenjek springs catc hment in sout h western Slovenia and validated by means of two multi-tracer tests wit h a total of six injection points. e resulting vulnerability, h azard and risk maps are plausible, and t h e validation conrmed t h e vulnerability assessment at t h e representative sites t h at were selected for tracer injection. e maps provide improved source protection zones and make it possible to identify land mismanagement and to propose better practices for future planning. Keywords: Slovene Approac h, groundwater vulnerability, contamination risk, karst aquifer, water source protection and management, validation of vulnerability maps.

PAGE 2

ACTA CARSOLOGICA 36/3 2007 398 Karst springs are t h e most important drinking water sources in Slovenia, because of t h eir good quality and suf cient amount. Unfortunately, in t h e Slovene legislation on water sources protection, t h e special c h aracteristics of karst aquifers are insurciently taken into consideration (Ravbar & Kovai 2006). Alt h oug h t h e quality of karst waters is relatively hig h, individual examples of contami nation illustrate t h e s h ortcomings of water management even in unin h abited alpine karst areas, w hic h are ordi narily very favourable for protection. In some ot h er countries, groundwater vulnerability and risk maps are used for protection zoning and land use planning in karst. us, dierent met h ods h ave al ready been developed and implemented in numerous test sites worldwide. Moreover, in some European countries, vulnerability mapping h as been integrated in t h e state protection legislation. However, in Slovenia experiences of suc h applications are very modest only two karst spring vulnerability studies h ave been done so far (Jana & Prestor 2002; Petri & ebela 2004). In order to provide comparable guidelines for t h e protection of carbonate aquifers in individual regions of Europe, COST Action 620 developed a general concep tual framework for vulnerability and risk mapping (Daly et al 2002; Zwa h len 2004). ere are two types of vul nerability: t h e intrinsic vulnerability only depends on t h e h ydrogeological c h aracteristics of an area determining its protective function against contamination; t h e specif ic vulnerability additionally considers t h e properties of specic contaminants. Vulnerability and risk maps can be prepared for a groundwater resource or for a specic source, suc h as a tapped spring or a pumping well. On t his basis, a compre h ensive approac h for ground water vulnerability and contamination risk assessment is proposed as an alternative to t h e existing karst ground water management in Slovenia. e Slovene Approac h includes t hree elements: t h e intrinsic vulnerability map, t h e h azard map and t h e contamination risk map. e met h odology takes into account special c h aracteristics of Slovene karst aquifer systems; it is consistent wit h na tional environmental legislation and enables comparison across European countries. e Slovene Approac h oers a new possibility to in tegrate surface and groundwater protection. In addition, it includes two new aspects for t h e intrinsic vulnerability mapping, w hic h h ave not yet been surciently addressed in t h e previous met h ods. Temporal h ydrologic variabil ity is particularly important for contaminant transport (Ravbar & Goldsc h eider 2006; Gppert & Goldsc h eider 2007) and h as t h us been integrated in t h e concept of groundwater vulnerability assessment. Furt h ermore, t h e met h od considers groundwater ow and transport pro cesses wit hin t h e saturated zone t h at are crucial for source protection. e European Approac h was completed by including t h e importance of a water resource or source into t h e risk analysis. e proposed approac h includes relatively detailed assessment sc h emes and is t h us most appropriate for aquifers and spring catc hments w h ere ex tensive data are available or can be obtained, and w h ere small-scale land use planning is considered important. However, for t h e application in data-poor environments and/or on larger scales, t h e met h od can be generalised and adapted to t h e local conditions. INTRODUCTION SLOVENE APPROACH TO GROUNDWATER VULNERABILIT Y MAPPING According to t h e European Approac h groundwater vulnerability mapping is founded on t h e assessment of basic factors t h at control inltration of water and con taminants from t h e land surface towards t h e groundwa ter, suc h as Overlying layers (O), Concentration of ow (C) and Precipitation regime (P). ere are two general approac h es of water protection: resource protection aims to protect t h e w h ole groundwater body, w h ile source protection aims to protect a particular spring or well. In t h e rst case, t h e mostly vertical seepage of water t h roug h t h e unsaturated zone to t h e uppermost groundwater surface is considered; in t h e second case, t h e lateral ow route wit h in t h e saturated zone s h ould be included as well. us an additional factor for t h e Karst saturated zone (K) h as to be considered (Daly et al. 2002). e Slovene Approac h to intrinsic vulnerability as sessment is partly based on t h e Spanis h COP met h od (Vas et al. 2006; Andreo et al. 2006), w hic h represents an integral interpretation of t h e European Approac h. Al t h oug h t h e COP met h od h as been successfully applied in dierent karst areas, it h as some weaknesses. erefore, it h as been modied, complemented and extended for source vulnerability mapping. e resource vulnerability map is obtained by combining t h e O, C and P factors, w h ereas t h e source vulnerability map is obtained by su N ATAA R AVBAR & N ICO G OLDSCHEIDER

PAGE 3

ACTA CARSOLOGICA 36/3 2007 399 perimposing t h e resource vulnerability map and t h e K factor (Fig. 1). O VERL Y ING LA Y ERS O FACTOR e O factor indicates t h e eectiveness of layers overly ing t h e groundwater surface to protect it against poten tial contamination (Daly et al. 2002). e O factor assess ment t h erefore takes into account t h e residence time of t h e percolating water (and/or contaminant) t hroug h t h e soil and t h e rocks composing t h e unsaturated zone, con siderably aected by t h e t hickness, porosity and perme ability of eac h layer. e evaluation of t h e soil protection function, w h ere present, is based on its texture (i.e. grain size distribu tion), soil structure (i.e. t h e presence of aggregates and macropores) and its t hickness. W h en assessing soil dept h (especially of patc h y soils) t h e percolation time t hroug h t h e soil, providing its eective t hickness, s h ould be con sidered. Furt h ermore, t h e protection function of t h e unsaturated zone is quantied by t h e lit h ological c h aracteristics and t hickness of eac h stratum (determining h ydrogeological properties, eective porosity and h ydrau lic conductivity), as well as by t h e degree of fracturation and/or karstication of t h e carbonate rocks. In Slovene karst regions, deep karst plateaux prevail, for w hic h an immedi ate inltration of rainwater and fast verti cal draining are c h aracteristic. e dept h of t h e unsaturated zone can reac h 1500 m and more. In general, t h e protective cover of soil and sediments is t hin or completely absent. Alt h oug h great t hicknesses of t h e unsaturated zone may provide some de gree of protection, bare karrenelds con nected wit h deep s h as (e.g. t h e Kaninski Podi, t h e Kriki Podi, t h e Rombonski Podi in t h e Alps and t h e drocle on t h e Snenik mountain) can provoke rapid percolation bypassing t h e overlying layers. A very low protective value is consequently assigned to suc h areas. Conned circumstance of t h e aqui fer is considered as well. In cases w h ere an aquifer under consideration is overlain by anot h er relevant aquifer, t h e vulnerabil ity of t h e hig h est one h as to be considered and grap hically symbolized on t h e map, as it is done in t h e PI met h od (Goldsc h eider 2005). C ONCENTRATION OF FLOW C FACTOR e C factor distinguis h es areas of dierent inltration conditions. It identies t h e existence of allogenic point rec h arges and expresses t h e degree to w hic h t h e overly ing layers are bypassed. e evaluation of t h e C factor is based on t h e zoning of t h e swallow h ole rec h arge area, and t h e rest of t h e area. Wit hin t h e catc hment of sinking surface waters, t h e distance to t h e swallow h ole and t h e distance to t h e sink ing stream or lake are considered. Most existing met h ods classify swallow h oles, sinking streams and t h eir catc h ment areas as zones of hig h or extreme vulnerability. Examples from t h e Slovene karst s h ow t h at, due to fast and strong groundwater level oscillations, some swallow h oles are frequently or permanently active, w hile ot h ers operate only during exceptional h ydrological events, sometimes less t h an once per year. However, only in case of a permanently active point inltration, would a con taminant release always and rapidly reac h t h e groundwa Fig. 1: P rinciple of the Slovene Approach to groundwater vulnerability mapping. PROPOSED METHODOLOG Y OF VULNERABILIT Y AND CONTAMINATION RISK MAPPING FOR THE PROTECTION OF KARST ...

PAGE 4

ACTA CARSOLOGICA 36/3 2007 400 ter wit h out signicant attenuation, w h ereas it mig h t not directly enter t h e karst groundwater in case of an occa sionally active one (Ravbar & Goldsc h eider 2006). If t h e swallow h oles are not permanently active, t h e temporal variability sub-factor (tv) s h ould be considered, reducing vulnerability in dependence on t h e frequency and dura tion of t h e swallow h ole activity. Furt h ermore, sinking streams in Slovenia are some times tens of kilometres long and drain catc hments of h undreds of km 2 (e.g. t h e Reka river, t h e Temenica river); t h ere are also examples of large lakes drained by swal low h oles (e.g. t h e lake of Cerkniko Jezero). Regarding t h e concept of swallow h oles and sinking surface waters being extremely vulnerable, t his would lead to extremely large areas to be protected at t h e hig h est level. We propose to assign a lower degree of vulnerability more t h an 5 km upstream from t h e swallow h ole, w h ere surface waters and t h eir catc hments s h ould be protected independently from groundwater vulnerability issues, as proposed by existing European and national water protection poli cies. Areas t h at drain out of t h e karst system under con sideration, eit h er wit h out contact to t h e groundwater or via gaining streams, s h ould be assigned a low degree of vulnerability. In t h e autogenic rec h arge area, surface karst land forms (karren, dolines and ot h ers) as well as hig h ly frac tured areas s h ould be identied, as t h ose represent zones of preferential inltration and ow concentration (Ford & Williams 2007). W h en sediments and soils overly t h ese landforms, t h e protection is increased. e inltration of water is also controlled by slope inclination, vegetation cover and ow type, t h e last of t h ese h aving t h e strongest impact on t h e vulnerability evaluation. ese aspects are considered in t h e sv subfactor, applicable for bot h point or diuse inltration conditions. e dominant ow process is controlled by t h e permeability of t h e layer at or closely below t h e sur face. W h ere layers are less permeable, surface or s h al low subsurface ow oen occurs on very at and even h orizontal surfaces. It may eventually inltrate in more or less concentrated mode, e.g. into a swallow h ole. On t h e ot h er h and, even steep slopes of permeable ground may drain underground and direct inltration predomi nates. Furt h ermore, t h e sv sub-factor attributes hig h er vulnerability to steeper slopes and sparser vegetation. Denser vegetation always provides protection to ground water due to less runo, more intermediate storage and t h us slower inltration. P RECIPITATION REGIME P FACTOR e precipitation regime inuences t h e rates of inltra tion, percolation and groundwater ow and t h us con taminant transport in t h e aquifer (Daly et al. 2002). e P factor considers t h e quantity and intensity of precipita tion events based on t h e daily precipitation amount for at least a 30-year period. Hig h er rainfall quantities and intensities mean more surface ow, hig h er transport ve locities, s h orter transit times, more turbulent ow, more eective transport of sediments, microbial pat h ogens and particle-bound c h emical contaminants, mobilisation of DNAPLs (Dense Non-Aqueous P h ase Liquids), etc., and t h us hig h er vulnerability. Two sub-factors s h ould be considered to describe t h e number of major and extreme precipitation events t h at enable signicant contaminant mobilisation and rapid transport: e rd sub-factor indicates t h e num ber of days wit h major rain quantities (20 mm/day), w hile t h e se sub-factor indicates t h e number of extreme storm events (> 80 mm/day). K ARST SATURATED ZONE K FACTOR e K factor represents t h e predominantly lateral groundwater ow in t h e saturated zone of t h e karst aq uifer towards t h e spring or well. is factor needs to be considered for source vulnerability mapping, toget h er wit h t h e t hree factors included in t h e resource vulner ability assessment (O, C and P). e K factor does not only consider t h e degree of karstication, but mainly describes t h e dynamics of groundwater ow and h ow a particular zone is connected to t h e spring or well, em p h asized and implemented also by Andreo et al. (manu script in preparation). In karst aquifers, underground ow pat h s are oen not known. Also t h e classication of t h e karstication degree can be very subjective. erefore, t h e detailed dis tribution of t h e K factor is dircult to map. Size, connection and density of groundwater ow passages is a reliable source of information on t h e karst network, h owever, even a relatively small degree of karst ication (e.g. conduits 5 cm wide and inaccessible to cav ers) can result in very rapid ow and contaminant trans port wit h out signicant attenuation. Groundwater divides in karst aquifer systems are oen not identical to topograp hic divides and are of ten dircult to determine. Furt h ermore, t h e position of groundwater divides is oen not stable but may vary for several h undreds of metres or even kilometres as a func tion of t h e h ydraulic conditions (Ravbar & Goldsc h eider 2006). However, t h e size of a catc hment is particularly crucial for source vulnerability assessment. In addition, catc hments of several individual springs oen overlap or are h ydraulically connected over long distances. e K factor s h ould ideally reect t h e following aspects (aer Goldsc h eider et al. 2001; Daly et al. 2002; Brouyre 2004): N ATAA R AVBAR & N ICO G OLDSCHEIDER

PAGE 5

ACTA CARSOLOGICA 36/3 2007 401 t h e travel time of a contaminant arrival at t h e source, t h e relative quantity of contaminants t h at arrive at t h e source, t h e contaminant concentration at t h e source and t h e duration of a contamination of t h e source. e evaluation of t h e K factor considers t hree subfactors and includes information on groundwater travel time, variability of t h e drainage divides and information on underground water ow pat h s. Duration of a contam ination could be an optional aspect. e t sub-factor considers t h e groundwater travel time in t h e saturated zone during hig h-ow conditions. e proposed classes are < 1 day, 1 days and > 10 days, but t h ese limits could be adapted to national legislation. Travel times can best be obtained from articial tracer tests, if possible supplemented by geological, speleologi cal and natural tracer data. Due to aquifer h eterogeneity, it is very dircult to draw precise isoc hrones. Nevert h e less, travel times s h ould be included in t h e assessment sc h eme using t h e best possible estimates. e r sub-factor expresses t h e degree of connection and contribution of dierent parts of t h e aquifer system to t h e source. e proposed assessment sc h eme consid ers t h e h ydrogeological structure of t h e aquifer system by distinguis hing between inner, intermediate and outer zones. e Iris h Groundwater Protection Sc h emes use a similar system, but dene 100 days as t h e limit between an inner and an outer zone (DoELG/EPA/GSI 1999). e inner zone comprises parts of t h e system t h at always con tribute to t h e spring and t h at are directly connected to t h e spring. e inner zone is classied as most vulnerable. e outer zone comprises parts of t h e system t h at contribute only marginally to t h e spring disc h arge (e.g. because t h ey essentially drain towards anot h er spring), areas t h at contribute only temporarily (e.g. during hig h water conditions), areas t h at are indirectly connected to t h e spring (e.g. because t h ey are separated by an aqui clude), as well as areas for w hic h we are not sure if t h ey contribute to t h e source. Very remote parts of a spring catc hment could also be included into t h e outer zone. e outer zone is classied as less vulnerable; a moderate vulnerability is assigned to intermediate situations. e information required for t h e r sub-factor assessment can also be obtained from tracer tests and general h ydrogeo logical considerations. e n sub-factor indicates t h e presence of an active conduit network and assigns hig h er vulnerability to t h e wider area above t h ose conduits. In many cases, suc h in formation is not available, and it is widely known t h at t h e absence of explorable conduits does not mean t h at t h ere are no conduits. However, if t h ere is direct evidence about active groundwater ow pat h s, t h e vulnerability assessment can be improved by including t his informa tion. S OURCE INTRINSIC VULNERABILIT Y ASSESSMENT Source vulnerability is obtained by combining t h e re source vulnerability assessment and t h e K factor vulner ability evaluation (Fig. 2). e source vulnerability map can be used as a basis for t h e delineation of source pro tection zones. SLOVENE APPROACH TO CONTAMINATION RISK MAPPING Vulnerability mapping is not always a surcient criterion for proper land use planning, since it does not s h ow t h e degree to w hic h t h e aquifer is already under pressure, and h ow important t h e groundwater is for water sup ply or ot h er purposes. On t h e basis of t h e conceptual framework proposed by COST Action 620 (De Ketelaere & Daly 2004), t h e Slovene Approac h provides a com pre h ensive risk analysis, w hic h takes into account t hree elements: t h e intrinsic vulnerability of t h e groundwater resource or source, t h e contamination h azards, and t h e importance of t h e resource or source. H AZARD ASSESSMENT e goal of h azard mapping is to identify and illustrate t h e locations and types of h uman activities t h at pose a t h reat to groundwater quality. e h azard evaluation considers t h e type, noxiousness and quantity of t h e contaminants, as well as t h e likeli h ood of a contaminant release (De Ketelaere et al 2004). e h azard level is ac h ieved by assessing t h e socalled h azard weig h ting, ranking and reduction values. e weig h ting factor (H) distinguis h es h azard h armfulness on t h e basis of a qualitative comparison of t h e potential damage to t h e groundwater or source. e main criteria for weig h ting dierent h azards concern t h e toxicity of relevant substances associated wit h eac h type of h azard as well as t h eir properties regarding solubility and mobility. A detailed table of t h e weig h ting values for dierent h azard types, ranging between 10 and 100, can be found in t h e Final Report of COST Action 620, p. 9596 (De Ketelaere et al. 2004). PROPOSED METHODOLOG Y OF VULNERABILIT Y AND CONTAMINATION RISK MAPPING FOR THE PROTECTION OF KARST ...

PAGE 6

ACTA CARSOLOGICA 36/3 2007 402 Fig. 2: e Slovene Approach intrinsic vulnerability assessment scheme. N ATAA R AVBAR & N ICO G OLDSCHEIDER

PAGE 7

ACTA CARSOLOGICA 36/3 2007 403 Hazards of t h e same type but wit h dierent c h ar acteristics (size, etc.) can be compared using a h azard ranking classication. According to COST Action 620, t h e ranking factor (Qn) ranges between 0.8 and 1.2. e Slovene Approac h provides ranking values for selected h uman activities, developed for Slovene circumstances (Fig. 3). e reduction factor (Rf) expresses t h e probability of a contamination event to occur. erefore, t h e tec h ni cal status, level of maintenance, surrounding conditions, security measures and ot h er factors s h ould be consid ered. e values s h ould range between 0 and 1. e re duction factor is 1 w h en no suc h information is available. Lower values imply positive information concerning t h e reduction of t h e likeli h ood. However, it is recommended to use small deviations from 1 in order to avoid minimi zation of t h e eects of h azards wit h h ig h toxic potential (De Ketelaere et al 2004). For eac h activity its h azard level is assessed by mul tiplying t h e h azard weig h ting value H, t h e ranking factor Qn and t h e reduction factor Rf (Fig. 4). E VALUATING THE IMPORTANCE OF A GROUNDWATER SOURCE OR RESOURCE e importance of a groundwater source or resource de pends on its actual or potential use for drinking water purpose, agriculture or ot h er h uman activities, and on t h e ecological value of t h e ground or spring water and associated surface waters. e quantity of t h e used water and t h e size of t h e population and animal stock depend ing on t his water also determine t h e importance. e ir replaceability also needs to be considered, i.e., is it t h e Fig. 3: H azard ranking. PROPOSED METHODOLOG Y OF VULNERABILIT Y AND CONTAMINATION RISK MAPPING FOR THE PROTECTION OF KARST ...

PAGE 8

ACTA CARSOLOGICA 36/3 2007 404 Fig. 4: H azard level assessment. only possible water source, or are t h ere ot h er available options? On t h e basis of t h ese criteria, a simple assess ment sc h eme for t h e importance of karst water sources and resources is proposed (Fig. 5). Four sub-factors need to be evaluated and summed up. e importance evalu ation enables prioritisation procedure for protection and sanitation programme on a local or regional scale. R ISK ASSESSMENT e risk intensity is obtained by combining t h e vul nerability and t h e h azard assessment. e hig h est risk of groundwater contamination is present w h ere danger ous h azards occur in a hig h ly vulnerable zone. e total risk additionally considers t h e importance of t h e water resource and source (Fig. 6). A hig h risk consequently in dicates t h e necessity to act by c h anging land use practices or removing existing h azards. FIRST APPLICATION AND VALIDATION OF THE SLOVENE APPROACH HY DROGEOLOGICAL CHARACTERISTICS OF THE TEST SITE e catc hment of t h e Podstenjek springs stretc h es over about 9 km 2 in t h e Upper Pivka valley in sout h west ern Slovenia. e catc hment area occupies moderately karstied Cenomanian limestones and limestone brec cias, and Palaeogene limestones t h at are over-t hrusted to t h e impermeable Eocene ysc h (iki & Pleniar 1975), t h us forming a s h allow karst aquifer. Due to t h e fast and strong h ydrologic variations in response to precipitation events or snowmelt, groundwater table uctuates for sev eral tens of meters. ere is no permanent surface stream rec h arging t h e springs h owever, owing to groundwater uctuations and weak connections between dierent karst conduits, two intermittent lakes (embijsko Jezero, Narie) appear w h enever groundwater level is surcient ly hig h. In t h e area of t h e intermittent lakes, alluvial de posits can be found, and in t h e uplied dry valley at t h e outskirts of t h e catc hment, t h ere is periglacial material deposited in t h e dolines. S h allow c hromic Cambisol interwoven wit h Ren dzina layers appears in patc h es. e t hickest soil layers can be found in t h e bottom of t h e concave relief s h apes, w hile t h e rest of t h e surface is pretty rocky. Most of t h e area is overgrown wit h forest and meadows or is used for low-intensity agriculture. For t h e catc hment delineation and application of t h e Slovene Approac h, a compre h ensive study was done, including tracer tests, detail structural-lit h ological and geomorp h ological mapping, electrical resistivity imag ing, as well as h azard mapping. Continuous monitoring of t h e springs p h ysico-c h emical c h aracteristics h as been performed for t h e h ydrograp h analyses, water balance calculation and aquifer be h aviour compre h ension (Rav bar 2007). A PPLICATION OF THE S LOVENE A PPROACH AND RESULTS e proposed Slovene Approac h h as been rst applied to t h e Podstenjek springs catc hment, t h us allowing its development, completion and testing. Based on t h e geo logical and geomorp h ological settings, tracer test results and t h e springs h ydrodynamic be h aviour t h e catc hment can be subdivided into an inner and an outer zone. In t h e area t h at is always, directly and fully contributing to t h e disc h arge of t h e springs (i.e. t h e inner zone) t h e geomor p h ological features (karren, hig h ly fractured areas, caves, karst edge) and outcrops along t h e roads w h ere soil cover is absent or rarely exceeds 20 cm were identied as hig h ly vulnerable. Hig h degree of vulnerability is also assigned to t h e estavelle (s h own in t h e zoomed inset) w h ere oc casional point rec h arge occurs. Moderate vulnerability h as been assigned to t h e bare karst landscape or karst covered by s h allow soils, as well as to t h e karren and dry valleys in t h e area of t h e partial or occasional contribu tion to t h e springs (i.e. t h e outer zone). e bottoms of intermittent lakes and dolines covered by t hicker soils or sediments are of low vulnerability, as well as rest of t h e outer zone (Fig. 7). N ATAA R AVBAR & N ICO G OLDSCHEIDER

PAGE 9

ACTA CARSOLOGICA 36/3 2007 405 Fig. 5: Assessment scheme to evaluate the importance of a karst groundwater resource or source. Fig. 6: e Slovene Approach to contamination risk assessment. PROPOSED METHODOLOG Y OF VULNERABILIT Y AND CONTAMINATION RISK MAPPING FOR THE PROTECTION OF KARST ... In t h e embije village, t h e only settlement in t h e catc hment, about 200 in h abitants live. e h ouses h ave been linked to t h e public sewage system since 1998 and connected to t h e wastewater treatment plant located out side t h e catc hment. e intensity of agricultural activity is relatively low. ere are some smaller waste disposal and excavation sites. e h azards found in t h e test site are mainly classied as low or very low; h owever, in more t h an h alf of t h e area no h azards h ave been identied (Fig. 8). e Podstenjek water source supplies only a few h undred people and is scantily used for animal breed ing and gardening. However, it is t h e only water source. Furt h ermore, t h e presence of P roteus Anguinus h as been

PAGE 10

ACTA CARSOLOGICA 36/3 2007 406 reported in t h e nearby cave (Krivic et al. 1987). ere fore, importance of t h e source h as been evaluated as medium. By superimpos ing source vulnerability, h az ard and source importance maps, t h e total contamina tion risk evaluation h as been obtained. In general, t h e risk degree strongly depends on t h e h azard level and its dis tribution. Most of t h e catc h ment is exposed to low risk; only urban areas, roads, dumps and excavation sites represent medium degree of contamination risk (Fig. 9). R ELIABILIT Y OF THE RESULTS e source vulnerability map s h ows zones of low, medium and hig h degree of vulner ability, w hic h can be t h e ba sis for t h e protection zoning. However, vulnerability maps are conservative simplica tions of natural conditions and t h e results are inuenced by diverse aspects (e.g. quali ty and accuracy of data, t h eir interpretation, selection and evaluation of dierent pa rameters, etc.). erefore t h e results need to be validated. Goldsc h eider et al. (2001) proposed using trac er tests for t h e validation of vulnerability maps and con sidered t hree criteria t h at can be obtained from tracer breakt hroug h curves: t h e peak time (time of maxi mum tracer concentration), t h e recovery rate (R), and t h e maximum concentration normalised by t h e injected tracer mass (C/M). is approac h h as t hree minor drawbacks: t h e time of rst arrival is oen more relevant for problems of water con Fig. 7: Source vulnerability map of the P odstenjek springs catchment, maps of the individual parameters and location of the validation points. Fig. 8: H azard map of the P odstenjek springs catchment. N ATAA R AVBAR & N ICO G OLDSCHEIDER

PAGE 11

ACTA CARSOLOGICA 36/3 2007 407 tamination; R and C/M are interdependent; and R de pends directly on t h e spring disc h arge Q. erefore, we propose to use only two modied validation criteria: t h e time of rst tracer detection, and t h e normalized tracer recovery R N w hic h is dened as: It is a way of expressing t h e tracer recovery indepen dent of t h e spring disc h arge. W h en R N is used for validation, t h e same degree of vulnerability would be attributed to a small spring and to a big spring if t h e tracer breakt h roug h curves at bot h springs are similar, i.e. similar maximum concentration and duration of t h e tracer (or potential con tamination event) appearance. e origin (injection point) presents h ig h vulnerability for t h e observed target (most commonly a source), if rapid inltration and fast ow in conduits are t h e dominant conditions. Resulting travel times are t h us very s h ort, minimizing also t h e sorption, degradation, cation exc h ange, dispersion and dilution of a solute matter. In suc h conditions t h e possible contami nation would reac h t h e water source very rapidly and its concentration at t h e outlet, as well as t h e relative quantity of t h e recovered tracer, would be h ig h In contrast, t h e ori gin (injection point) presents low vulnerability for t h e ob served target (most commonly t h e source), if t h e tracer is mostly absorbed in t h e sediments and soil. Consequently, t h e possible contaminant arrival is retarded and its con centration signicantly reduced or t h e contaminant does not arrive at all. Intermediate situations correspond to me dium vulnerability (Fig. 10). By carrying out two multi-tracer tests in t h e stud ied area, we examined and veried t h e adequacy of t h e obtained vulnerability map and gained additional infor mation on t h e mec h anism of potential contaminant trans port under dierent h ydro logical conditions. e rst tracer test, car ried out in Marc h 2006, was made under h ig h water con ditions and was followed by several intense precipitation events so t h at immediate inltration and transport of tracers towards t h e springs took place. Two tracers were injected; 94 g of sulfor h oda mine B was injected in an es tavelle t h at was empty at t h e time of injection (injection point A) and 500 g of eosin was injected in karren (injec tion point B). e estavelle was c h aracterised as h ig h ly vulnerable and t h e area below t h e Milanka mountain as of low vulnerability (Figs. 7 and 11). Bot h tracers more or less simultaneously appe a red t h ree days aer t h e injection in t h e Podstenjek springs. Sulfor h odamine B was detected in t h e springs samples for few days wit h maximal concentration of 1.65 ppb and ap peared again in lower concentrations aer t h e subsequent Fig. 9: Total risk map of the P odstenjek water source. Fig. 10: Diagram setting up the tracer test results for source vulnerability validation purposes. Due to the lack of consensus regarding the vulnerability classication between dierent methods a universal class distribution is provided. To enable general applicability of the diagram denite boundaries between vulnerability classes are omitted. PROPOSED METHODOLOG Y OF VULNERABILIT Y AND CONTAMINATION RISK MAPPING FOR THE PROTECTION OF KARST ...

PAGE 12

ACTA CARSOLOGICA 36/3 2007 408 rainy events. Altoget h er 52.5% of t h e sulfor h odamine B h as been recovered. On t h e contrary, t h e eosin appear ance does not s h ow a typical breakt h roug h curve and only few samples were eosin positive. e peak concentration of 0.2 ppb and t h e total recovery of 0.95% were observed at t h e Podstenjek springs. e greater portion (81.2%) owed to t h e nearby Bistrica spring (Figs. 11 and 12). e second experiment, carried out in November 2006, was made under low water conditions. A more in Fig. 11: Results of the two multi-tracer tests performed in the P odstenjek spring catchment during high and low water conditions. tense rainy event did not occur until 15 days aer t h e in jection. Four tracers were injected in four locations. On t h e bottom of t h e embijsko Jezero 500 g of uranine was spread over several metres t h ick soil and sediment cover (injection site 1). Over t h e Narie w h ere soil and sediments occur in pockets and t h e limestone rock base outcrops in places 400 g of sulfor h odamine G was spread (injection site 2). Two tracers were spread over t h e limestone surface, partially covered by scarce soil and vegetation cover. At in jection site 3 a total of 5 kg of lit h ium c h loride was used and at t h e injection site 4 a total of 5 kg of potassium iodide was used. e rst t h ree injection sites are classied as of low and t h e last one as of moderate vul nerability (Fig. 7). Only iodide, used in in jection site 4 was detected at t h e Podstenjek springs (Fig. 12). It was rst detected two days aer t h e injection and its appearance lasted for addi tional two days wit h maximal concentration of 3.2 ppb. Al toget h er 0.63% of t h e injected iodide was recovered. Lit hium was only detected in t h e Piv ka spring and even aer six mont h s of sampling no uo rescent tracers h ave been de tected in any of t h e observed springs (Fig. 11). ese results conrm t h e vulnerability assessment; t h e tracer injected in t h e area classied as hig h ly vulnerable rapidly reac h ed t h e spring, its concentrations and recovered quantities were hig h. e trac er injected on t h e area classi ed as moderately vulnerable rapidly reac h ed t h e spring, t h e concentrations were hig h, but t h e recovered quantities very low. e tracers injected on areas classied as low vulner ability zones did not arrive at t h e spring at all or t h e tracer only arrived in small propor tions and was detected in low concentrations. N ATAA R AVBAR & N ICO G OLDSCHEIDER

PAGE 13

ACTA CARSOLOGICA 36/3 2007 409 Fig. 12: P articular tracer appearance at the P odstenjek spring (the location of the injection sites are shown on Fig. 7). CONCLUSION e proposed Slovene Approac h is so far t h e most com plete interpretation of t h e European Approac h to vul nerability and risk mapping, as it can be used for bot h resource and source vulnerability mapping and also in cludes an assessment of contamination h azards, an eval uation of t h e value or importance of t h e groundwater, and dierent types of risk maps. e resulting maps can be used as a basis for t h e delineation of protection zones and ot h er land use planning issues. Suc h maps could also be used to focus pollution investigations and pollution prevention inspections of hig h risk premises. e test site application and validation wit h tracer tests conrmed t h at t h e intrinsic vulnerability map is plausible and reliable. Furt h ermore, t h e validation also justies t h e integration of h ydrological variability into vulnerability mapping: t h e tracers injected in sites of occasional direct inltration during low ow condi tions were not detected in any of t h e springs, but were absorbed by t h e soil and sediments, w hile t h e tracer in jected during hig h ow conditions arrived at t h e springs. e h azard and risk maps also s h ow t h at t h e water source is not at hig h risk. e few water quality analyses s h ow ing its hig h quality conrm t his evaluation. e study h as s h own t h e new Slovene Approac h gives justied results and provides improved source pro tection zoning. Furt h ermore, we identied land misman agement and proposed better practices for future plan ning. us, t h e Slovene Approac h could be proposed as t h e basis for t h e delineation of karst source protection zones and included to t h e state protection sc h emes. Al t h oug h t h e met h od considers karst-specic inltration conditions, it is not restricted solely to karst but can also be used in non-karst areas. Moreover, since we believe t h at vulnerability met h ods s h ould not be limited to t h e individual countries borders, t h e Slovene Approac h could be a basis for t h e furt h er work concerning ground water protection elsew h ere. PROPOSED METHODOLOG Y OF VULNERABILIT Y AND CONTAMINATION RISK MAPPING FOR THE PROTECTION OF KARST ...

PAGE 14

ACTA CARSOLOGICA 36/3 2007 410 R EFERENCES Andreo, B., Goldsc h eider, N., Vadillo, I., Vas, J.M., Neu kum, C h., Sinreic h, M., Jimnez, P., Brec h enmac h er, J., Carrasco, F., Htzl, H., Perles, J.M. & F. Zwa h len, 2006: Karst groundwater protection: First applica tion of a Pan-European Approac h to vulnerabil ity, h azard and risk mapping in t h e Sierra de Lbar (Sout h ern Spain).Science of t h e Total Environ ment 357, 54-73. Brouyre, S., 2004: A quantitative point of view of t h e concept of vulnerability. In: Zwa h len, F. (ed.), COST Action 620. Vulnerability and Risk Mapping for t h e Protection of Carbonate (Karstic) Aquifers. Final report COST Action 620.European Commission, Directorate-General for Researc h, 10-15, Brssel, Luxemburg. Daly, D., Dassargues, A., Drew, D., Dunne, S., Golds c h eider N., Neale, S., Popescu, I.C. & F. Zwa h len, 2002: Main concepts of t h e European approac h to karst-groundwater-vulnerability assessment and mapping.Hydrogeology Journal 10, 340-345. De Ketelaere, D. & D. Daly, 2004: Vulnerability in a Risk Framework. In: Zwa h len, F. (ed.), COST Action 620. Vulnerability and Risk Mapping for t h e Protection of Carbonate (Karstic) Aquifers. Final report COST Action 620.European Commission, DirectorateGeneral for Researc h, 84-85, Brssel, Luxemburg. De Ketelaere, D., Htzl, H., Neukum, C h., Civit, M. & G. Sappa, 2004: Hazard Analysis and Mapping. In: Zwa h len, F. (ed.), COST Action 620. Vulnerability and Risk Mapping for t h e Protection of Carbonate (Karstic) Aquifers. Final report COST Action 620.European Commission, Directorate-General for Researc h, 106-107, Brssel, Luxemburg. DoELG/EPA/GSI, 1999: Groundwater protection sc h emes.Department of Environment and Local Government, Environmental Protection Agency and Geological Survey of Ireland, 24 pp., Dublin. Ford, D.C. & P.W. Williams, 2007: Karst h ydrogeology and geomorp h ology.Jo hn Willey & Sons, Ltd, 562 pp., London. Goldsc h eider, N., 2005: Karst groundwater vulnerability mapping application of a new met h od in t h e Swabian Alb, Germany.Hydrogeology Journal 13, 555-564. Goldsc h eider, N., Htzl, H., Fries, W. & P. Jordan, 2001: Validation of a vulnerability map (EPIK) wit h tracer tests.7 t h Conference on Limestone Hydrology and Fissured Media, Besanon 20 t h -22 nd September 2001, Sci. Tec h. Envir., Mm. H S. n 13, 167-170. Gppert, N. & N. Goldsc h eider, 2007: Solute and colloid transport in karst conduits under low and hig h ow conditions.Ground Water, publis h ed online (DOI: 10.1111/j.1745-6584.2007.00373.x). Jana, M. & J. Prestor, 2002: Ocena naravne ranljivosti vodonosnika v zaledju izvira Riane po metodi SIN TACS.Geologija, 45/2, 401-406, Ljubljana. Krivic, P., Juren, A., Bizjak, M. & B. Ravnikar, 1987: Hidrogeoloke raziskave zaledja Zgornje Pivke, IV. faza.Geoloki zavod Ljubljana, TOZD-I geologija, geote hnika, geozika, 21 pp., Ljubljana. Petri, M. & S. ebela, 2004: Vulnerability mapping in t h e rec h arge area of t h e Korentan spring, Slovenia.Acta Carsologica, 33/2, 151-168, Ljubljana. Ravbar, N. & G. Kovai, 2006: Karst water management in Slovenia in t h e frame of vulnerability mapping.Acta Carsologica, 35/2, 73-82, Ljubljana. Ravbar, N. & N. Goldsc h eider, 2006: Integrating tem poral h ydrologic variations into karst groundwater vulnerability mapping examples from Slovenia.8 t h Conference on Limestone Hydrogeology, 229233, Neuc h tel. Ravbar, N., 2007: e protection of karst waters : a com pre h ensive Slovene approac h to vulnerability and contamination risk mapping.ZRC Publis hing, 254 pp., Ljubljana. N ATAA R AVBAR & N ICO G OLDSCHEIDER e study was founded by t h e Slovene Ministry of Hig h er Education, Science and Tec hnology in t h e frame of t h e Y oung Researc h ers Programme, and is a contribution to t h e UNESCO/IUGS IGCP 513 Project Global Study of Karst Aquifers and Water Resources. Special t h anks to Dr. Pierre-Andr Sc hnegg (University of Neuc h tel), Hans Kra (Nationalpark Berc h tesgaden) and Borut Peric (Park kocjanske jame) for lending us eld uor ometers and an automatic sampler. e aut h ors also acknowledge t h e h elp of Silvia Guglielmetti (University of Neuc h tel), Dr. Jesus Vas (University of Malaga) and Gregor Kovai (University of Primorska). A CKNOWLEDGEMENTS

PAGE 15

ACTA CARSOLOGICA 36/3 2007 411 iki, D. & M. Pleniar, 1975: Osnovna geoloka karta Jugoslavije. Tolma za list Ilirska Bistrica.Zvezni geoloki zavod Beograd, 51 pp., Beograd. Vas, J.M., Andreo, B., Perles, J.M., Carrasco, F. & I. Va dillo, 2006: Proposed met h od for groundwater vul nerability mapping in carbonate (karstic) aquifers: t h e COP met h od. Application in two pilot sites in Sout h ern Spain.Hydrogeology Journal 14, 6, 912925. Zwa h len, F., 2004: Vulnerability and Risk Mapping for t h e Protection of Carbonate (Karstic) Aquifers. Fi nal report COST action 620.European Commis sion, Directorate-General for Researc h, 297 pp., Brssel, Luxemburg. PROPOSED METHODOLOG Y OF VULNERABILIT Y AND CONTAMINATION RISK MAPPING FOR THE PROTECTION OF KARST ...



PAGE 1

QUESTIONS OF DR Y VALLE Y S IN KARST: CASE STUD Y OF MALI DOL, KRAS SLOVENIA V PRAANJE SUHIH DOLIN NA KRASU PRIMER M ALEGA DOLA K RAS S LOVENIJA Jure K OUTNIK 1 Izvleek UDK 911.2:551.435.83(497.4 Mali dol) Jure Koutnik: Vpraanje suhih dolin na Krasu, primer Male ga dola, Kras (Slovenija) V lanku je predstavljena su h a dolina Mali dol, ki je ena izmed su hi h dolin, ki prekajo 45 km dolgo in do 15 km iroko uravnavo Krasa. Mali dol je 10 km dolga in med 50 in 75 metrov globoka su h a dolina, ki meandrira po Komen skem Krasu v prenodinarski smeri preno na vse reliefne in geoloke strukture. Zaradi sedimenta in debeleje odeje prsti so se v dno dolinaste oblike poglobile globoke in velike koro zijske vrtae. Te vrtae so veje in globlje od vrta na sosed njem, uravnanem krakem povrju. Lokalno pretrte cone, ko menski ploati apnenci z roenci in debela odeja sedimenta dovoljujejo povrinsko odtekanje vode s poboij v dno doline. Na nekateri h mesti h so se zato oblikovali globoki in strmi ero zijski jarki. Geomorfoloke znailnosti, starost dolinske oblike in debelina sedimenta govorijo o poligenetskem nastanku su h e doline. Oblika je nastala med uravnavanjem platoja v viini piezometrinega nivoja. Kljune besede: Geomorfologija, su h a dolina, krako uravna vanje, Mali dol. 1 Petrieva 14, SI-1230-Ljubljana-rnue, jurekosutnik@gmail.com Received/Prejeto: 17.08.2007 COBISS: 1.01 ACTA CARSOLOGICA 36/3, 425-431, POSTOJNA 2007 Abstract UDC 911.2:551.435.83(497.4 Mali dol) Jure Koutnik: Questions of dry valleys in Karst: Case study of Mali dol, Kras (Slovenia) In t h e article t h e geomorp h ology of t h e Mali dol dry valley is presented. It is one of at least two dry valleys t h at cross about 45 km long and up to 15 km wide Kras plateau. Mali dol is a 10 kilometers long, 50 to 75 meters deep dry valley. It meanders on t h e leveled surface across t h e Kras plateau in direction NESW, perpendicularly to all structures. Because of t h e sediment and t hicker soil cover deep and large solution dolines inside t h e basic valley-like s h ape were formed. ese dolines are larger and deeper t h an t h e dolines formed on t h e nearby surface of t h e plateau. Local tectonically crus h ed zones, limestones wit h c h erts and t hick sediment cover allow some overland ow into t h e valley. In suc h places some smaller but deep and steep ero sion gullies were formed. From t h e morp h ology, t h e position of t h e dry valley and t h e sediments we can infer t h e polygenetic origin of t h e valley. It was formed during t h e planation of Kras in t h e h eig h t of t h e piezometric level. Keywords: Geomorp h ology, dry valley, karst planation, Mali dol. I NTRODUCTION Kras, because of its historic importance for karstology, also called classical Kras and because it lies in t h e hin terland of Trieste also called Trieste Kras, is around 500 km2 big limestone plateau in SW Slovenia (Gams, 1974). Geologically Kras plateau is an anticlinorium leveled by a longer period of planation and denudation (Gams, 1974). It is part of t h e Dinaric Mountains. In Mesozoic Kras was part of t h e dinaric carbonate platform. e oldest and most widely spread rocks on t h e surface of Kras are cretaceous limestones. In t h e smaller extent also Paleocene and lower Eocene limestones can be found. Result of t h e deepening of sedimentation envi ronment (Jurkovek, 1996) and erosion in Alps are t hick Eocene ysc h beds w hic h covered SW Slovenia (Placer, 1981; Habi, 1984; Gams, 2003) and are t h e youngest rocks in t h e area.

PAGE 2

ACTA CARSOLOGICA 36/3 2007 426 Kras inclines evenly from SE towards NW; from 435 meters above see level at Divaa to 98 meters above sea level at Doberdo. ree hig h er elevation ridges, wit h h eig h ts above 450 meters are present (Slovenija: pokra jine in ljudje, 2001). Trstelj hills (Trstelj 643 m), t h e nort h ern most part of Kras borders wit h Vipava valley. e second ridge starts near Sesljan (Sistiana), it reac h es t h e biggest h eig h ts above Trieste and continues into Slo venia. On t h e top of t h e middle ridge Volnik hills (Vol nik 546 m) t h e state border between Italy and Slovenia is running. Volnik hills divide two main plateaus, t h e nar rower Nabreina-Bazovica (Aurisina-Basovizza) and t h e wider Seana-Komen plateau (Gams, 2003). Micro relief in Kras is represented in numerous dolines and scarce low, rounded hills. Depression forms are represented also wit h large rounded dolines of collapse origin and valleylike features, wit h additional, sometimes prolonged, so lution dolines in t h e bottom. e primary goal of t h e researc h was to determinate t h e origin of t h e valley-like features. Alt h oug h uvial erosion seemed to be t h e most probable process t h at created t h e forms, also ot h er possibilities were considered. Karst h as its beginning and its end, its history in w hic h it goes t hroug h a variety of evolu tion p h ases. e denition of karst we know from t h e literature but if we take a historical time line of a certain karst region and we slice it somew h ere in its past evolution and analyze t h e state of t h e karst and make a denition out of our ndings, we will many times nd t h at our denition is just opposite of t h e one from t h e text books. If we are aware of t his and if we know t h at karst is in a way conservative, t h at it to some extend preserves t h e sur face forms, t h e combination of dierent historical land forms in karst surface is somet hing we must seriously consider during our studies. I can say t h at dry valleys are c h aracteristic and important karst land form. Dry valleys ap pear also in t h e uvial geomor p hic system but t h eir identica tion in karst is easier because of t h eir longer life spend. eir im portance lies in t h eir key-role in interpretation of development of certain karst surface. Researc h ers named dry valleys in Kras doli (singu lar: dol). In h abitants of Kras use t h e expression dol for bigger dolines, w hic h are common also in t h e bottom of dry valleys (Radinja, 1965). e word dol is of preslavic origin and meant cave, cavern, valley and also into t h e valley and into t h e cave (Snoj, 2003). Mali dol got its name aer t h e village of t h e same name in its bottom, so we can distinguis h it from Brestovica and Doberdo dry valleys. For t h e purpose of understanding t h e processes in t h e dry valley-like form and its origin detailed geomor p h ologic mapping was done, w hic h resulted in a geomor p h ologic map in t h e scale 1:5000 wit h legend. Mali dol meanders on t h e leveled surface across Kras plateau in t h e direction NE-SW, perpendicularly to lit h o logical and structural units of prevailing dinaric direc tion. It starts above Branica river valley wit h a pass in t h e Trstelj hills on 293 meters above sea level. It h as a distinc Fig. 1: Kras plateau with M ali dol dry valley (Slovenija: pokrajine in ljudje, 1999). J URE K OUTNIK

PAGE 3

ACTA CARSOLOGICA 36/3 2007 427 tive prolonged depressional s h ape, c h aracteristic for river valleys. In its bottom large solution dolines were deepened. Transition between t h e surrounding planated karst surface and Mali dol is quick but not s h arp. e dierence in elevation between Komen plateau and t h e bottom of Mali dol is in its nort h ern part 40, in its central part between 50 and 60 and in t h e sout h ern part up to 75 meters. e bottom of t h e valley-like form is at its largest extent more t h an 250 and in its smallest extant only 40 meters wide. e total lengt h of Mali dol is 10 kilometers. In Mali dol, t hroug h out its lengt h t h e rocks are of upper cre taceous age. Between Tomaevica and Mali dol village c h aracteristic t hin-layered limestones wit h c h erts can be found. At t h e edge of Kras anticlinorium in Trstelj hills also layers of Paleocene and Eocene age appear. On t h e opposite end of Mali dol in t h e Divaa fault lower creta ceous limestones and dolomites ap pear (Osnovna geoloka karta SFRJ, 1973). Age of rocks generally de creases in t h e NE direction. R ELIEF FORMS IN M ALI DOL Evolution of relief forms in Mali dol presumably h appened gradually, all forms are not of t h e same age and all forms were not made only by cor rosion. e most important relief forms are dolines and erosion gul lies. Up to 20 meters deep solution dolines are deepened into valleys bottom. Dolines are in average wid er t h an dolines on t h e surrounding karst plateau and usually prolonged in t h e general valley direction. Ex tremely wide irregular dolines can also be found in t h e bottom of Mali 50 0 40 0 30 0 20 0 15 0 10 0 50 0 500 400 300 200 150 100 50 0 100 300 500 700 900 1100 (m ) Branica river valle y Kome n Veliki do l Trstelj hill s Trstelj hills Veliki do l Komen Branica river valley Longitudinal profile of Mali do l Longitudinal profile of Komen plateau Fig. 2: Longitudinal prole of M ali dol dry valley. Geomorphologic mapping of M ali dol Fig. 3: DEM of M ali dol and its surrounding. QUESTIONS OF DR Y VALLE Y S IN KARST: CASE STUD Y OF MALI DOL, KRAS SLOVENIA

PAGE 4

ACTA CARSOLOGICA 36/3 2007 428 dol. According to t h eir size we can call t h em uvalas but t h ey can be also of collapse origin. e number of dolines on a square kilometer of planated surface near Komen and on t h e square kilometer of Mali dol is at a ratio of 5 to 1. In t h e slopes of Mali dol, in t h e slopes of dolines in t h e bottom of Mali dol and between dolines erosion gullies were developed. Big ger gullies are less steep and wider t h an smaller ones, commonly related to tectonically fractured zones. Usually t h ey end in dolines in t h e bottom of Mali dol, w h ere alluvial fens were formed. Gullies in t h e slopes of Mali dol reac h t h e planated Komen plateau and can still bee traced as a series of s h allow dolines. Gullies were probably formed, because lo cally t hicker, less permeable soil and sediment cover allows surface runo. Evidence of t h at was commonly found. Surface water ow h as an ability to car ry sediment and organic parti cles from t h e plateau and slopes to t h e bottom of Mali dol. Fig. 5: Legend of the geomorphologic map. Fig. 4: Geological map of M ali dol (Osnovna geoloka karta, list Gorica). Fig. 6: P art of the geomorphologic map of M ali dol (TTN 1:5000). J URE K OUTNIK

PAGE 5

ACTA CARSOLOGICA 36/3 2007 429 ple removed t h e sediments and used t h em elsew h ere, even aer several meters bedrock was not reac h ed. e origin of t h e sedi ment is polygenetic. Part is t h e so called insoluble limestone residue, w hic h is more important only w h ere ot h erwise very pure limestone contains c h erts. Origin of t h e ot h er part of t h e sediment can be related wit h impermeable ysc h layers, from w hic h particles were broug h t on today already denudated karst surface by sinking streams. Two important processes are enabled by t h e sediment presence. Firstly it allows surface run-o wit h S EDIMENTS AND FLOWSTONE IN M ALI DOL Bottom of Mali dol is covered by a t hick layer of sedi ments and soil. At some locations in dolines, w h ere peo Fig. 7: P rolonged doline in the bottom of M ali dol. Fig. 8: Doline with anthropogenicaly changed bottom. Fig. 9 and Fig. 10: Erosion gullies in the slopes of M ali dol. Fig. 11: ick sediment in the bottom of M ali dol. Fig. 12: Rock particles and pebbles found in the sediment. QUESTIONS OF DR Y VALLE Y S IN KARST: CASE STUD Y OF MALI DOL, KRAS SLOVENIA

PAGE 6

ACTA CARSOLOGICA 36/3 2007 430 tra n sportation to lower parts of t h e valley and formation of gullies. Secondly its capability of water retention in creases t h e accelerated corrosion on t h e contact between sediment and limestone. Aloc h tonous and autoc h t h onous rock particles were found in t h e sediment of Mali dol. Autoc h t h onous are t h e semi-rounded parts of c h ert rubble and some sub cutaneously rounded limestone particles, bot h deriving from underling bedrock. Aloc h tonous rock particles are pebbles of quartz and c h erts w hic h were transported and rounded by sinking streams from ysc h. Quartz pebbles are w hite, non-transparent and partly limonited. C h ert pebbles, reddis h to pale grey, almost w hite in color were made from ysc h sandstone and h ave visible layers and grainy texture (Herlec, 2006). More or less rounded par ticles of limonite iron ore can be found. All speleot h em occurrences were noted because of t h e interpretation of t h e origin of t h e valley. Compact speleot h em blocks were found twice, t h e rst time in t h e Bobovec doline nort h of village Mali dol w h ere de nudation in t h e slope revealed a cave, possibly an abyss lled wit h cave sediments and owstone. e second time owstone was found NW of t h e village ko. A cave opens in t h e SE slope of doline called Jevine. Distinc tive yellowis h cave sediment and t hick speleot h em come to t h e surface. T HE QUESTION OF DEVELOPMENT ORIGIN AND AGE OF M ALI DOL According to older Slovene geomorp h ologists Mali dol was created by uvial erosion in t h e so called pre-karst p h ase of surface development. e valley is a remnant of t h e last p h ase of river down-cutting before it gradually sinked underground in correlation wit h sudden start of karstication process (Radinja, 1965). Gams wrote t h at Mali dol is t h e result of t h e last p h ase of tectonic lowering of t h e western part of Kras as a result of counter-clockwise rotation of t h e Adriatic micro plate and general tectoni cal lowering of Po river basin (Gams, 1998). Tectonical origin of t h e valley is also accepted by Habi. He wrote t h at Mali dol was created on t h e contact between dier ent tectonical structures in association wit h t h e c h ange in general ridge direction of Kras from dinaric (NW-SE) to alpine (W-E; Habi, 1984). In addition to possible uvial and tectonical origin of Mali dol also t h e development from a denudated cave system s h ould be considered. By inspection of t h e geomorp h ologic map and by t h e met h od of exclusion we can state, t h at Mali dol, as we see it today, was made by uvial erosion. Meandering s h ape excludes its tectonical origin, t h e lack of t hick-bed ded speleot h ems and unproportionaly smaller dimen sions of t h e largest Slovene unroofed cave (unroofed cave in Loza) exclude its unroofed development. Only expla nation by t h e uvial erosion development remains. e valley-like feature Mali dol in Kras was at some time in t h e past an active river valley w hic h is now dry. C ONCLUSION Already in t h e early stages of researc h preparation, it be came clear t h at t h e task is going to be h arder t h an ex pected. In a seemingly simple equation new, puzzling variables emerged. Buy geomorp hic mapping polygenetic origin of Mali dol was conrmed. Dierent processes, acting one by anot h er and simultaneously, h ave aected t h e creation and later development of Mali dol during longer period of time. All aective geomorp hic processes are closely connected wit h sediments. Less permeable sediment, w h et h er it is insoluble limestone residue, remnant of Eo cene ysc h cover or lling of denudated caves, h as a key role in t h e geomorp h ological development of Mali dol. e sediments impermeability, t hickness and retention capability aected t h e accelerated corrosion process and contributed to preservation of t h e land form. Numerous questions remain open. e time-line, age of planation, beginning of t h e last p h ase of karstica tion and t h e valley formation s h ould be conrmed wit h absolute datations. As far as we know t h e most promising datation met h od is paleomagnetism w hic h already gave some results on t h e age of cave sediments in Slovenia. New data about tectonic activity on t h e Miocene/Plio cene border is being gat h ered, along wit h possible gra dient c h anges on Kras during t h e messinian crisis and mineralogical c h aracteristic of karst sediments. Specic accelerated corrosion impact and detailed sediment c h aracteristic of Mali dol are jet to be studied. J URE K OUTNIK

PAGE 7

ACTA CARSOLOGICA 36/3 2007 431 is article is part of my diploma t h esis presented at Geograp h y department of Ljubljana University. I would like to t h ank Dr. Andrej Mi h evc for guidance and sup port. A CKNOWLEDGEMENT R EFERENCES Dravna topografska karta RS 1:25 000. 145 Opatje selo, 1996. 1. izd. 1:25 000. Ljubljana, Republika Slovenija Ministrstvo za obrambo in Ministrstvo za okolje in prostor, Geodetska uprava RS. Dravna topografska karta RS 1:25 000. 127 Miren. 1997. 1. izd. 1:25 000. Ljubljana, Republika Slovenija Ministrstvo za obrambo in Ministrstvo za okolje in prostor, Geodetska uprava RS. Gams. I., 1974: Kras.-Slovenska matica, p. 359, Ljublja na. Gams, I., 1986: Kontaktni uviokras.-Acta carsologica, 14/15, 71, Ljubljana. Gams, I., 1998: Geomorp h ogenetics of t h e Classical Karst Kras.-Acta carsologica, 17, 2, 181, Ljubljana. Gams, I., 2003: Kras v Sloveniji.-Zaloba ZRC, ZRC SAZU, p. 516, Ljubljana. Gregori, A., 2005: Analiza razvoja paleonapetosti v dinarskem narivnem sistemu za h odne Slovenije (okolica Nove Gorice).-Univerza v Ljubljani, NTF, Oddelek za geologijo, p. 84, Ljubljana. Habi, P., 1984: Reliefne enote in strukturnice matinega Krasa.-Acta carsologica, 12, 5, Ljubljana. Habi, P., 1986: Povrinska razlenjenost Dinarskega Krasa.-Acta carsologica, 14/15, 39, Ljubljana. Herlec, U., 2006: Litoloka sestava proda.-Personal source, Ljubljana. Jurkovek B., et al 1996: Formacijska geoloka karta junega dela Trako-komenske planote.-Intitut za geologijo, geote hniko in geoziko, p. str, Ljubljana Kras, pokrajina-ivljenje-ljudje. 1999. Kranjc, A., (ur.). Ljubljana, Zaloba ZRC, ZRC SAZU, 321 str. Mi h evc, A., 1999: Pomen brezstropi h jam za geoloko, geomorfoloko in geografsko prouevanje krasa.Geoloki zbornik, 14, 27, Ljubljana. Osnovna geoloka karta SFRJ. Tolma lista Gorica. 1973. Beograd, Zvezni geoloki zavod, 50 str. Osnovna dravna karta, SR Slovenija 1:5 000. B22 Ko men 19. 1968. 1:5 000. Ljubljana, Zvezna geodetska uprava SFRJ. Placer, L., 1981: Geoloka struktura jugoza h odne Slo venije.-Geologija, 24, 1, 27, Ljubljana. Radinja, D., 1965: Nova morfogenetska dognanja na Kra su.-Univerza v Ljubljani, Oddelek za geograjo, p. 133, Ljubljana. Radinja D., 1966: Morfogenetska problematika matinega Krasa.-Geografski obzornik, 13, 3, 108, Lju bljana. Radinja D., 1974: Matini Kras v lui irega reliefnega razvoja. Acta carsologica, 6, 22, Ljubljana. Slovenija: pokrajine in ljudje. 2001. Perko, D., Oroen Adami, M. (ur.). Ljubljana, Zaloba Mladinska knjiga, 735 str. Snoj, M., 2003: Slovenski etimoloki slovar. 2. pregledana in dopolnjena izd.-Modrijan, p. 1022, Ljubljana. ebela, S., 1999: Kras v geoloki preteklosti. V: Kras, pokrajina-ivljenje-ljudje, Kranjc, A., (ur.). Ljublja na, Zaloba ZRC, ZRC SAZU, str. 19. uteri, F., 1998: Interaction between a cave system and t h e lowering karst surface case study: Laki ravnik.Acta carsologica, 27, 2, 115, Ljubljana. Temeljni topografski nart, SR Slovenija 1:5 000. B22 Ko men 28. 1981. 1:5 000. Ljubljana, Geodetska uprava SRS. Temeljni topografski nart, SR Slovenija 1:5 000. B22 Ko men 29. 1981. 1:5 000. Ljubljana, Geodetska uprava SRS. Temeljni topografski nart, SR Slovenija 1:5 000. B22 Ko men 37. 1981. 1:5 000. Ljubljana, Geodetska uprava SRS. Temeljni topografski nart, SR Slovenija 1:5 000. B22 Ko men 38. 1981. 1:5 000. Ljubljana, Geodetska uprava SRS. Temeljni topografski nart, SR Slovenija 1:5 000. B22 Ko men 47. 1981. 1:5 000. Ljubljana, geodetska uprava SRS. Temeljni topografski nart, SR Slovenija 1:5 000. B22 Ko men 48. 1981. 1:5 000. Ljubljana, geodetska uprava SRS. QUESTIONS OF DR Y VALLE Y S IN KARST: CASE STUD Y OF MALI DOL, KRAS SLOVENIA



PAGE 1

L AND COVER MAPPING USING L ANDSAT SATELLITE IMAGE CLASSIFICATION IN THE C LASSICAL K ARST K RAS REGION U GOTAVLJANJE POKROVNOSTI K RASA S KLASIFIKACIJO SATELITSKIH POSNETKOV L ANDSAT iga K OKALJ 1 & Kritof O TIR 1 Izvleek UDK 528.85:711.14(497.4) iga Kokalj & Kritof Otir: Ugotavljanje pokrovnosti Krasa s klasikacijo satelitskih posnetkov Landsat Kras je ekoloko zelo raznolika in obutljiva regija zato potrebu je posebno premiljeno gospodarjenje z naravnimi viri in kul turno dediino. Pomemben kazalnik, ki omogoa analizo nji h ovega stanja in spremljanje razvoja, je pokrovnost. Daljin sko zaznavanje oziroma napredna klasikacija satelitski h pos netkov sta natanna in cenovna ugodna alternativa klasinim te hnikam kartiranja pokrovnosti. V prispevku so opisane metode za pridobitev zanesljive in uporabne karte pokrovnosti zemeljskega povrja. Kompleksnost obmoja narekuje kombi nacijo razlini h virov podatkov, kot so satelitski posnetki Land sat, digitalni model viin, ortofoto posnetki in obstojee topo grafske in tematske karte. Kot glavni klasikacijski algoritem je bila uporabljena metoda najveje verjetnosti, natannost pa je bila poveana z uporabo me h ke klasikacije, omejevanjem z viino in nagibom ter dodatnimi sloji podatkov. Kljune besede: daljinsko zaznavanje, pokrovnost, raba tal, klasikacija, satelitski posnetki, Kras. 1 Institute of Ant hropological and Spatial Studies ZRC SAZU, Novi trg 2, Si-1000 Ljubljana, Slovenia, e-mail: zkokalj@zrc-sazu.si, kristof@zrc-sazu.si Received/Prejeto: 29.03.2007 COBISS: 1.01 ACTA CARSOLOGICA 36/3, 433-440, POSTOJNA 2007 Abstract UDC 528.85:711.14(497.4) iga Kokalj & Kritof Otir: Land cover mapping using Land sat satellite image classication in the Classical Karst Kras region Suc h a diverse and sensitive eco-region as Karst needs to be managed wit h special attention and consideration of its natural and cultural resources. Land cover is an important indicator, w hic h enables t h e analysis of t h eir condition and development monitoring. Advanced satellite images classication represents an accurate and cost-eective alternative to t h e classical tec h niques of land cover mapping. e met h ods used to produce a reliable land cover map are presented in t his paper. e com plexity of t h e area requires a combination of various data suc h as Landsat satellite images, digital elevation model, digital or t h op h otos as well as existing topograp hic and t h ematic maps. e maximum likeli h ood algorit hm was used as t h e main clas sier and t h e accuracy of results was furt h er improved by fuzzy classication, altitude and inclination ltering and auxiliary data integration. Key words: remote sensing, land cover, land use, classication, satellite imagery, Classical karst. e Karst or classical karst is an extensive limestone pla teau, well distinguis h ed from t h e nearby regions due to its steep rise above t h e neig h bouring predominantly ysc h areas. It is a distinct border region, w hic h is evident in several c h aracteristics. It lies in t h e vicinity of t h e Adriat ic Sea; nevert h eless t h e steep elevation gradient prevents t h e seas soot hing eects to reac h it. Due to t h e vicinity of t h e hig h karst plateaux in t h e nort h t h ere are substantial continental inuences. e transition between t h e Medi terranean and continental impact is present in t h e hig h winds; Burja, a strong nort h wind is common during t h e winter (Perko et al 1999). Due to its typical water and soil c h aracteristics t h e Karst landscape is extremely sensitive to pollution and t h erefore special attention h as to be paid to its manage ment. All available means s h ould be employed in t h e ef I NTRODUCTION

PAGE 2

ACTA CARSOLOGICA 36/3 2007 434 Remote sensing applications for karst environment studies were rst focused on geological lineaments ex traction (Suzen et al 1998). Sabins (1997) describes tec h niques for geological and geomorp h ological surveillance of tropical karst using radar images, w hile Hung et al (2002) used image fusion of Landsat images and edge detection ltering for fault and lineament extraction, serving for cave development analysis in t h e tropical karst area of nort h-western Vietnam. Hung et al (2003) presented an environmental analysis consisting of met h ods for image transformation, image fusion, lineament extraction, time series, and c h ange detection for study ing land cover c h anges. e groundwater rec h arge and disc h arge zones were dened by t h e tec hnique of image transformation. Furt h er h ydrological applications were realised by estimating water rec h arge potential by deter mination of lineaments and drainage frequency density, lit h ologic c h aracter, karstic domains and land cover/land use wit h utilization of Landsat and SPOT imagery and aerial p h otos (S h aban et al 2006). Kresic (1995) describes Dinaric karst in t h e Balkans as a favourable area for ap plication of h ydrogeological remote sensing tec hniques, due to t h e geomorp h ologic c h aracteristics, in particular t h e specic surface drainage and karst forms, t h e varying vegetation t h at most oen reects t h e existence of dif ferent geologic formations on t h e surface, and distinct tectonic features. He proposes a met h od for determina Fig. 1: e Karst plateau and its vicinity as seen on the 1999 Landsat satellite image. IGA K OKALJ & K RITOF O TIR fective management and monitoring of natural resourc es. e interaction between mankind and t h e environ ment h as to be taken into consideration, since it is t h e greatest t hreat to sensitive areas suc h as t h e Karst region. Landscape observation met h odologies t h at oer accurate results and enable historical, e.g. annual, comparisons s h ould be employed. Remote sensing surveys provide a rapid means of data collection t h at can ac hieve complete coverage of large areas, wit h far lower costs t h an t h ose associated wit h eld survey. Remote sensing can detect features un seen on t h e surface, map t h em accurately, and oer inter pretations based on t h eir form, distribution, and context. Image interpretation and processing h ave now become standard tools, and t h e use of aerial p h otograp h s, satel lite imagery and ot h er remote sensing tec hniques h ave become increasingly sop histicated particularly because digital spatial imagery h as become ever more ubiquitous (Kvamme 2005). Aerial p h otograp h y is t h e oldest domain of remote sensing of karst landscapes and sill receives a great focus, especially for its detail, but ot h er sensing devices h ave been placed in t h e air in recent decades, including pas sive multispectral and t h ermal sensors, and active radar and laser altimeter systems, making aerial remote sens ing truly multidimensional (Kvamme 2005). A number of satellite systems h ave played a signicant role in mod elling end exploring karst landscapes. Landsat was t h e rst satellite program for collecting repetitive, synoptic, multi-spectral imagery for monitoring and analysing Eart h s resources and environment. Early studies focused on environmental zones or land-cover mapping, because spatial resolution was too coarse to detect smaller karst features. However, relatively recent introduction of hig h (spatial) resolution satellite imagery, wit h a sell size in t h e range of 1 m, enabled detection and mapping of individ ual karst features, especially w h en combined wit h lidar tec hnology. Ikonos and QuickBird are t h e two most oen used hig h resolution commercial satellites, oering mul tispectral data at 4 m and 2.4 m spatial resolutions, wit h panc hromatic data at 1 m and 0.61 m respectively. Lidar (LIg h t Detection And Ranging) is t h e optical equivalent of radar, an active instrument capable of rapidly generat ing hig h ly accurate digital models of topograp h y as well as t h e vertical structure of ot h er surfaces (buildings, trees) from t h e air. Lidar is a tec hnology providing remarkable surface detail, wit h absolute vertical accuracy up to a few centimetres, even in vegetated areas, and h orizontal sampling densities well below a meter. e potential for mapping karst features is immense as lidar can penetrate forest canopy and t h us provide information on features t h at are not identiable on eit h er topograp hic maps or aerial p h otograp h y.

PAGE 3

ACTA CARSOLOGICA 36/3 2007 435 tion of fractures and faults, as well as ground water ow direction from processing of satellite and aerial imagery. An interesting study was conducted by Peng et al (2000), w h o employed remote sensing to investigate karst land scape of sout h-eastern C hina for t h e potential of building t h e worlds largest radio telescope. Suc h investigations of topograp h y can be readily assisted by t h e S h uttle Radar Topograp h y Mission (SRTM) digital elevation model. e mission, own on t h e space s h uttle in 2000, created a digital elevation model covering 80 % of t h e Eart h s land surface by radar interferometry. e model wit h a resolu tion of 30 m provides an excellent tool for regional topo grap hic analyses. In t h e past classication of satellite and aerial im agery h as proved to be a good alternative to eld obser vations of large areas for it enables a detailed classica tion into ten or more land cover classes, as well as rapid execution and temporal comparisons (S h engtian et al 1999). e image classication tec hnology is well known and oen used, but h as to be applied specically to every observed environment and p h enomena. In classifying t h e Karst land cover, special attention was paid to all nec essary steps, from data selection to accuracy assessment. As basic classication does not provide surcient accura cy in t h e presented study we used advanced classication met h ods, suc h as post classication modelling. L AND COVER CLASSIFICATION e main purpose of satellite and ot h er imagery classi cation is t h e recognition of objects on t h e Eart h s surface and t h eir presentation in t h e form of t h ematic maps. Land cover is determined by t h e observation of grey values in t h e imagery. Classication is one of t h e most important steps in h andling remote sensing imagery and represents important input data for geograp hic information systems (GIS) (Otir 2006). e rst step in t h e classication is t h e selection of suitable data (images). We h ad several reasons to use Landsat satellite images in our project. ey h ave an ex cellent price-quality ratio, good spectral (seven bands from visible to t h e infrared spectre) and spatial c h ar acteristics (30 m resolution). Two multispectral images h ave been applied, one from Landsat 5 (18/8/1992) and t h e ot h er from Landsat 7 (15/9/1999, Fig. 1). Bot h cover t h e entire study area and are cloudless, w h ic h enables simple processing and accurate classication. Images were georeferenced to t h e Gauss-Krueger co-ordinate system, using multiple control points (86 for Landsat 5 and 102 for Landsat 7) wit h a h ig h er density in t h e Karst region. e ac h ieved average positional error is 32 m, w h ic h is approximately t h e pixel size. Bot h images were merged into a single multilayer data le, omitting t h e sixt h t h ermal band because of t h e lower resolution and its minimal contribution to t h e quality of t h e clas sication. e classication of satellite and ot h er images is di vided into supervised and unsupervised. e main dier ence between t h e two is in t h e way t h e spectral signatures are created. Wit h supervised classication t h e operator determines t h e areas, w h ere a distinct particular type of land cover is present and t h en t h e computer computes t h e spectral signatures. On t h e ot h er h and, in t h e unsu pervised classication t h e computer creates t h e spectral signatures using mat h ematical data clustering in t h e multidimensional feature space. e determination of t h e used land cover classes was inuenced by t h e previous classication of entire Slovenia (Otir et al 2000). is enabled t h e compari son and dierence analysis of t h e nal results. Unlike t h e previous classication t h e t hree urban classes (urban, densely built-up, and scarcely built-up) were united into built-up areas. e following categories were used: coniferous forest, deciduous forest, mixed forest forest wit h approximately t h e same proportion of coniferous and deciduous trees, bus h es and overgrowt h bus h es, transition from forest to meadow, overgrowing meadow, low (mainly Karst) forest, open meadows and pastures, agriculture elds, vineyards, orc h ards, built-up areas cities, villages, industrial areas, wider roads and parking places, construction sites, water sea, rivers, lakes, salt-pans. e rst and most important step in supervised classication is t h e selection of training samples. e operator digitalises t h e areas wit h known land cover on t h e screen. Image processing soware t h en computes t h e spectral signatures of t h e land cover types. e process runs interactively, as t h e quality of t h e training samples h as to be constantly evaluated and usually some h ave to be improved or even discarded. e mixed forest class was not used as a sample; it was obtained by unsupervised classication, as described bellow. On t h e basis of digital ort h op h otos, unsupervised classication into 20 classes and local area knowledge L AND COVER MAPPING USING L ANDSAT SATELLITE IMAGE CLASSIFICATION IN THE C LASSICAL K ARST K RAS REGION

PAGE 4

ACTA CARSOLOGICA 36/3 2007 436 t hree to ve training samples were determined for eac h category. eir suitability was c h ecked by t h e grap hic presentation of t h e spectral response, separability and contingency evaluation, blending in t h e spectral space and test-classication. e maximum likeli h ood met h od was c h osen as t h e main classication algorit hm. Comparing t h e classica tion result to t h e digital ort h op h otos situation s h owed t h at t h e built-up areas were in places classied as ploug h ed elds or ot h er open ground areas and vice versa. Elimi nating t h e problem wit h extra ploug h ed eld training samples did not solve t h e problem, as at t h at point a con siderable part of t h e built-up areas was classied as ag ricultural areas. Using t h e normalised dierence veg etation index (NDVI) also proved to be inappropriate, due to t h e same problem too many misinterpreted pixels. We decided to use t h e fuzzy classication ap proac h. e observed area was classied into two lay ers. e rst layer deter mined t h e most probable land cover class, w hile t h e second determined t h e next most probable class. For t h e built-up areas t h at were determined as agri cultural areas in t h e second layer of t h e classication t h e attribute was c h anged into agricultural areas. By t his action t h e above men tioned problem was solved in most cases (Fig. 2). For t h e classication of forests, training samples of coniferous and decidu ous forests were selected, w hile t h e mixed forest was initially supposed to be ob tained by considering areas wit h a similar proportion of bot h. Unfortunately t h e procedure failed, since t h e acquired percentage of mixed forests was too small. e inuence of t h e rugged terrain on t h e sur face illumination created an additional problem as t h e deciduous forest on t h e s h ady slopes s h owed similar spectral values as t h e coniferous one on t h e sunny slopes. Topograp hic normalisation was used wit h out any success. Better results were obtained by t h e unsupervised classi cation of forest surfaces only. Once more t h e masked (by forests) satellite image was classied by t h e unsupervised classication into 3, 6, and 10 classes. In ideal circum stances t h e classication into t hree classes would distin guis h coniferous, deciduous and mixed forests. Unfortu nately, t his was not a straig h tforward process and t h e best results were obtained by interpreting t h e classications in 6 and 10 forest types. Aerial p h otograp h y interpretation h elped to c h oose t h e 10-class classication, as it was t h e Fig. 2: Land cover map of the Karst region and its vicinity from the combined T M (18/8/1992) and ET M+ (15/9/1999) images. IGA K OKALJ & K RITOF O TIR

PAGE 5

ACTA CARSOLOGICA 36/3 2007 437 best for distinguis hing between t h e forest types. Certain classes were t h en merged into t hree basic categories and t h eir value was added to t h e initial classication. Distinguis hing elds, bus h es and meadows caused additional problems (Fig. 3). For example, it occurred t h at t h e class agricultural area covered bus h es or meadows. e problem, w hic h is a result of t h e fragmented land di vision, dierent land use types and rapid overgrowing of t h e Karst region, is especially evident, because meadows can be detected t hroug h t hin bus h es t h at create a similar spectral signature in all t hree categories. One also h as to consider t h e acquired data from t h e satellite images, as in t h e late summer a large amount of t h e elds is already barren, and due to t h e drier grounds it is h arder to distin guis h between dierent kinds of vegetation. Due to t h e insurcient spatial resolution of t h e Landsat satellite imagery only larger rivers are directly detectable. In order to avoid a discontinuous riv er sc h eme and to improve t h e accuracy we used t h e existing vector h ydrology data. e nal land cover map was masked by includ ing rivers wider t h an 5 m, as well as lakes, swamps, salt-pans and t h e sea. Furt h er quality im provement of t h e classica tion was made by consider ing (limiting) t h e altitude and inclination. e inter ferometric digital elevation model (DEM) wit h t h e resolution of 25 m (Otir et al 2002) was used as t h e source of t h e altitude data. e Gams altitude borders study (1960) and t h e incli nation denitions accord ing to t h e farming suitabil ity of t h e ground (Kladnik 1999) were also considered. e altitude of 1450 m was selected as t h e forest border, t h e altitude of 850 m and a 22 inclination were determined for agriculture, w hile t h e altitude of 900 m and a 25 inclination was consid ered for built-up areas. Everyt hing hig h er and steeper t h an t h e determined limits was reclassied into an open category. Due to t h e fact t h at an accurate DEM was not available at t h e time of t h e study t h e area on t h e Italian side more t h an 5 km from t h e state border was excluded in t h e post-classication. Furt h er spatial ltering was used to eliminate t h e noise from t h e results and ac hieve partial generalisation. An adapted majority lter (Kokalj and Otir 2006) of 3 by 3 pixels was used in order to detect t h e isolated pixels and assign t h em t h e prevalent class in t h eir vicinity. Fig. 3: e problematic category division resulting in the great variability of land use is one of the main characteristics when classifying satellite imagery of the Karst region. In the picture one can see the area south of Volji Grad village. e land cover layer is partially transparent, the background is a digital orthophoto (Source: DOF, 2002, GURS). L AND COVER MAPPING USING L ANDSAT SATELLITE IMAGE CLASSIFICATION IN THE C LASSICAL K ARST K RAS REGION R ESULTS e attribute accuracy of t h e classication was evaluated by using 120 control points wit h known land cover, de termined from t h e aerial p h otograp h y. e nal accuracy is very hig h and exceeds 90 %. roug h t h e analysis of t h e quality it was establis h ed t h at agricultural areas and open categories were mostly misclassied, for in almost all cases t h ey s h ould h ave been found in t h e bus h es and overgrowt h category. However, taking into account t h e

PAGE 6

ACTA CARSOLOGICA 36/3 2007 438 rapid overgrowing p h enom enon, t h e amount of young low forests and t h e very vari able land use c h anging over s h ort distances (extreme fragmentation) t his is under standable. Some misinter pretation can also be found in t h e forests, w hic h suggests t h at additional attention is necessary for accurate distin guis hing. e comparison wit h a previous classication (Fig. 4, Tab. 2), created for t h e plan ning of t h e mobile telep h one network, s h ows dierences mainly in t h e built-up and forest categories (table 2). e previous classication distinguis h ed between t hree classes of built-up areas, i.e. urban centres wit h out veg etation, densely and scarcely built-up areas wit h vegetation and gardens. In t h e presented classication t h ese classes were merged and are all pre sented by built-up areas wit h out dense vegetation. us, greater dierences appear in t h e countryside w h ere t h e newer classication does not consider t h e small built-up areas in t h e hills w h ereas t h e previous one did. However, t h e previous classication also considers a number of non built-up areas, mainly agricultural in t h e Koprska Brda region as well as in t h e neig h bour h oods of big cit ies. Fig. 4: M ultitemporal classication is especially useful for designating built-up areas. Comparison of the old (le) and new (right) classication reveals major improvements. e land cover layer is partially transparent, the background is a digital orthophoto (Source: DOF, 2003, GURS). Class Reference points Classied points Accurately classied Users accuracy Coniferous forest 18 18 17 94 % Deciduous forest 21 22 21 96 % Mixed forest 14 12 12 100 % Bushes and overgrowth 23 16 16 100 % Open 16 19 16 84 % Agriculture 17 21 17 81 % Built-up 5 6 5 83 % Water 6 6 6 100 % Total 120 120 110 Table 1: Accuracy assessment of the land cover classication. Classication for Mobitel (1999) Classication for Aquadapt (2004) ha % of total area ha % of total area Urban 1700 0.6 Densely built-up 6000 2.1 Scarcely built-up 16,800 5.9 Built-up total 24,500 8.6 17,200 6.0 Coniferous forest 23,000 8.0 31,100 10.9 Deciduous forest 71,400 25.0 57,300 20.1 Mixed forest 14,500 5.1 22,900 8.0 Forest total 108,800 38.1 111,300 39.0 Table 2: Comparison of built-up areas and forests between the single temporal and multitemporal classication. IGA K OKALJ & K RITOF O TIR

PAGE 7

ACTA CARSOLOGICA 36/3 2007 439 C ONCLUSIONS It h as been proven t h at t h e classication of satellite im ages is an ercient tool w h en determining t h e land cover in t h e Karst region. By using suc h a classication one can perform t h e mapping of a larger area and observe its tem poral development in a relatively s h ort period of time. e Landsat imagery used in t his study proved to be ap propriate for distinguis hing approximately ten to twenty land cover categories wit h a spatial resolution of 30 m. Due to t h e uniqueness of t h e Karst landscape, especially t h e hig h fragmentation of land owners hip, complex cul tivation patterns, variable soil conditions and rapid over growing, it is dircult to distinguis h some of t h e classes and more advanced met h ods need to be applied, e.g. post-classication. e use of various image data, ltering according to altitudes and inclination, and t h e combination of super vised and unsupervised classication enable a signicant improvement in t h e quality of t h e nal land cover map. e attribute accuracy of t h e produced classication assessed by comparing a larger number of test points wit h ground data collected from aerial ort h op h otos is hig h and exceeds 90 %. e categories agricultural land and open spaces are most commonly misclassied; h ow ever some misclassication also occurs wit hin t h e forest types. e classication of two (merged) images eases t h e distinction between t h e classes; especially t h e built-up areas can be easily distinguis h ed from t h e ot h er catego ries. W h en comparing t h e two classications dierences could mainly be noticed in forests and built-up areas. In order to determine t h e benets of individual classica tions additional eldwork would be required. is study conrmed our assumption t h at a simple land cover classication can not be used if we wis h to ac hieve hig h accuracy. It h as been proven t h at t h e land cover determination depends on its location as it is nec essary to consider all local c h aracteristics, t h e natural p h enomena and occurrences as well as man made ob jects. We can clearly claim t h at land cover classication s h ould be performed for smaller landscape units. e Karst region wit h its specics demands a detailed re searc h, supported by terrain limiting training samples selection and detailed result quality c h ecking. One of t h e most interesting questions w h en considering future studies is t h e usefulness of hig h-resolution (spatial and spectral) satellite and airborne sensors, especially lidar, for detailed researc h of natural and ant hropogenic Karst c h aracteristics. L AND COVER MAPPING USING L ANDSAT SATELLITE IMAGE CLASSIFICATION IN THE C LASSICAL K ARST K RAS REGION e discrepancy in t h e percentage of certain for est types occurs due to t h e dierences in obtaining t h e raw forest data. e previous classication determined t h e coniferous and deciduous forests t hroug h supervised classication, and t h e mixed forest data was obtained by considering t h eir relative proportions. e new classi cation determined forests t hroug h unsupervised classi cation and by later merging of t h e classes. In spite of t his t h e dierence in t h e total s h are of forests is rat h er small w h en t h e two classications are compared. Decid ing w hic h classication is better suited as regards t h e available means (area knowledge, digital ort h op h otos) is a dircult task, h owever, t h e forest map, w hic h can be found at t h e Slovenian Ministry of Forestry, can be used (wit h some hindrances). We believe extensive eldwork is necessary in order to verify certain sample areas.

PAGE 8

ACTA CARSOLOGICA 36/3 2007 440 Gams, I., 1960: O viinski meji naseljenosti, ozimine, gozda in snega v slovenski h gora h.Geografski vest nik, 32, 59-69, Ljubljana. Hung, L.Q. & Batelaan, O., 2003: Environmental geologi cal remote sensing and GIS analysis of tropical karst areas in Vietnam.Geoscience and Remote Sensing Symposium, IGARSS Proceedings, vol. 4, IEEE International, 2964-2966. Hung, L. Q., Din h, N.Q., Batelaan, O., Tam, V. T. & La grou D., 2002: Remote sensing and GIS-based analy sis of cave development in t h e Suoimuoi Catc hment (Son La NW Vietnam).Journal of Cave and Karst Studies, 64, 1, 23-33, Alabama. Kladnik, D., 1999: Leksikon geograje podeelja. Intitut za geograjo, p. 318, Ljubljana. Kokalj, & Otir, K., 2006: Ugotavljanje pokrovnosti Slovenije iz satelitski h posnetkov Landsat.Geo grafski vestnik, 78, 2, 85-95, Ljubljana. Kresic, N., 1995: Remote sensing of tectonic fabric con trolling groundwater ow in Dinaric karst.Remote Sensing of Environment, 53, 2, 85-90, New Y ork. Kvamme, K. L., 2005: Terrestrial Remote Sensing in Ar c h aeology.Handbook Of Arc h aeological Met h ods (H. Masc hner in C. C hippindale, ur.), AltaMira Press, 423-477, Lan h am. Otir, K., 2006: Daljinsko zaznavanje. Zaloba ZRC, p. 250, Ljubljana. Otir, K., Stani, Z., Podobnikar, T. & Ve h ovar, Z. 2000: Pridobivanje in uporaba prostorski h podatkov vi soke loljivosti pri nartovanju omreja mobilne telefonije.Geografski informacijski sistemi v Slo veniji 1999-2000, 143-152, Ljubljana. Peng, B. Nan, R. & Su, Y ., 2000. Proposed worlds largest single dis h, FAST.Radio Telescopes (Butc h er, H. R., ur.), Proceedings of SPIE, vol. 4015, 45-54. Perko, D. & Oroen Adami, M. (ur.) 1999: Slovenija pokrajine in ljudje .Mladinska knjiga, p. 735, Ljubljana. Sabins, F. F., 1997: Remote sensing. P rinciples and Inter pretation.W. H. Freeman & co., p. 494, New Y ork. S h aban, A., K h awlie, M. & Abdalla h, C., 2006: Use of re mote sensing and GIS to determine rec h arge poten tial zones: t h e case of Occidental Lebanon.Hydro geology Journal, 14, 4, 433-443. S h engtian, Y & Qijiang, Z., 1999: e Researc h on t h e Land cover and Vegetation Degradation Rate In t h e Karst Area using Remote Sensing.Asian Confer ence on Remote Sensing, Proceedings. h ttp://www. gisdevelopment.net/aars/acrs/1999/ps5/ps500a.asp Suzen, M. L. & Toprak, V., 1998: Filtering of satellite images in geological lineament analyses: an application to a fault zone in Central Turkey.International Journal of Remote Sensing, 19, 6, 1101-1114. R EFERENCES IGA K OKALJ & K RITOF O TIR



PAGE 1

A REVIEW OF THE FOREST MANAGEMENT HISTOR Y AND PRESENT STATE OF THE HARAGIST Y A KARST PLATEAU AGGTELEK KARST, HUNGAR Y PREGLED UPRAVLJANJA Z GOZDOVI NA PLANOTI HARAGIST Y A AGGTELEKI KRAS, MADARSKA Eszter T AN CS 1 Ferenc S ZMORAD 2 Ilona B R N YK EVEI 3 Izvleek UDK 551.44:630*6(439) Eszter Tancs, Ferenc Szmorad, Ilona Brny-Kevei: Pregled upravljanja z gozdovi na planoti Haragistya (Aggteleki kras, Madarska) Vegetacijo Madarskega krasa v glavnem sestavljajo listnati meani gozdovi, zato je upravljanje z gozdovi ena glavni h tem trajnostnega upravljanja s krasom. V tej tudiji, na primeru krake planote Haragistya, prikaemo rabo gozdni h povrin na obmoju Aggtelekega krasa v zadnji h stoletji h in obseg po sledic lovekovega vpliva. Raziskave speminjanja pokrajine v lui zgodovine rabe tal nudijo uporabne informacije za uprav ljanje v pri h odnosti. Za analizo smo uporabili integriran GIS zgodovinski h podatkov (vojake karte iz 18. in 19. stoletja, stari narti upravljanja z gozdovi, aerofoto posnetke itd.). tudijsko obmoje Haragistya lei na severnem delu Aggteleki h gora, na nadmorski viini 400 600 m. Gozdovi tega obmoja so e dolgo izpostavljeni lovekemu vplivu. Ta se je manjal v zaetku 20. soletja, odkar je obmoje pod upravo drave, pa je nepravilna in pretirana raba gozdov ustavljena. Kljune besede: upravljanje z gozdovi, Aggteleki kras, zgo dovina gozdov, Haragistya. 1 University of Szeged, Department of Climatology and Landscape Ecology, Hungary; nadragulya@geo.u-szeged.h u 2 Directorate of Aggtelek National Park, Hungary; info.anp@t-online.h u 3 University of Szeged, Department of Climatology and Landscape Ecology, Hungary; keveibar@eart h.geo.u-szeged.h u Received/Prejeto: 18.09.2006 COBISS: 1.01 ACTA CARSOLOGICA 36/3, 441-451, POSTOJNA 2007 A bstract UDC 551.44:630*6(439) Eszter Tancs, Ferenc Szmorad, Ilona Brny-Kevei: A rew iew of the forest management history and present state of the Haragistya karst plateau (Aggtelek Karst, Hungary e potential vegetation of Hungarian karsts is mainly mixed deciduous forest; consequently forest management is a key is sue in t h e sustainable management of our karsts. In t his study we attempt to describe on t h e example of t h e Haragistya karst plateau h ow t h e forests of t h e Aggtelek karst region were used in t h e last few centuries and to w h at extent t h ey were aected by ant hropogenic activity. Investigating c h ange patterns in t h e context of land use history provides useful information for fu ture management. For t h e analysis we used an integrated GIS of historical data (18t h-19t h century military maps, old forest management plans, aerial imagery etc.). Our study area, t h e Haragistya, is situated in t h e nort h ern part of Aggtelek Moun tains at 400-600 m above sea level. e forests of t his karst pla teau h ave been subject to signicant h uman impact over t h e centuries. Ant hropogenic inuence gradually decreased over t h e 20t h century and wit h t h e area being taken into state man agement t h e irregular and oen exaggerated use h as stopped. Keywords: forest management, Aggtelek Karst, forest history, Haragistya. I NTRODUCTION Due to t h eir geograp hical position t h e natural vegeta tion in Hungarian karstlands is mainly mixed deciduous forest and in our karst mountains t h ere is still consider able forest cover. If forest management is carried out in accordance wit h t h e interests of protection and tourism, it can serve as a basis of sustainable landuse in t h ese ar eas. Silvicultural activity also h as an impact on t h e w h ole karstecosystem t hroug h c h anging t h e forest and t h e site. It denes t h e state and extension of a forest for centuries.

PAGE 2

ACTA CARSOLOGICA 36/3 2007 442 Forest management is t h erefore a key issue in t h e sustain able management of our karsts. e present forest condition in Europe is consider ably modied from t h e natural state (Brads h aw 2004). As a result, t h e natural dynamics of European deciduous forests are little known, w hic h makes successful man agement dircult. In order to nd out more about t h ese processes forest reserves were designated w h ere manage ment activities were stopped entirely. Y et to understand c h anges t h at follow t h e abandonment of a forest we rst need to t h oroug h ly examine t h e history of forest man agement. e specialty of karst areas (especially open karsts) regarding forest management is t h eir hig h vulnerability. e forestry services of countries wit h larger consistent karst areas are well aware of t his t h erefore several at tempts h ave been made to work out guidelines regard ing forest management activities in suc h areas (e.g. Karst management h andbook for Britis h Columbia 2003). Fig. 1: e study area (source: www.anp.hu). Besides t h e pollution issues of increasing impor tance forest management in Hungarian karst areas h ad to face special c h allenges t hroug h out its history. e most important of t h ese are t h e s h allow karst soils c h aracter ised by an extreme water balance, w hic h aects species composition and quality. roug h t h e ages t his h ad an ambiguous eect on t h e forest management practices in Hungary`s karst areas. e forests wit h t h e most extreme conditions (e.g. t h e steep sout h ern slopes in t h e Aggtelek region) got legal protection very early in t h e interests of erosion protection and were usually not managed. Con scious management strategies favouring t h e sustainabil ity of bot h wood supply and forest environment were only applied in a few privately owned big estates, w hic h mostly encompassed hig h-quality forests. However most karstic forests did not belong in eit h er category and were t h erefore exploited in many ways. e numerous uses t h ese forests were continu ously subject to h ave c h anged t h eir structure and in many cases led to degradation. Our study area, t h e Haragistya karst plateau, is situated in t h e Aggtelek Karst (in t h e sout h ern part of t h e Gmr-Torna Karst), mainly in t h e strictly protected A zone of Aggtelek National Park (Fig.1) is plateau, similarly to ot h er Hungarian karst areas, h as been inten sively used over t h e centuries so investigating its patterns in t h e context of land use history may provide useful information for t h e future management of karsts. As t h e rst step, t his study summarizes t h e forest management history of t h e Haragistya plateau M ATERIALS AND METHODS e study area was dened considering t h e available his torical sources; we researc h t h e forests formerly belong ing to Szdvrborsa village (today called Silick Brezov) and now to Aggtelek village. Most of t h ese forests became part of t h e Haragistya-Lfej Forest Reserve in 1993. e Haragistya is part of t h e Silicka karst plateau, w hic h mainly belongs to Slovakia. Its h eig h t above sea level var ies between 400-600 meters. e bedrock mainly consists of Triassic limestone (Wetterstein Formation), dolomite and dolomitic limestone. e surface of t h e karst plateau is hig h ly varied, especially t h e surface of t h e dolomite (Fig.2) In t h e east dry valleys can be found wit h a main N-S direction w hile in t h e nort h ern and western parts series of dolines dominate t h e surface. e vegetation E SZTER T AN CS F ERENC S ZMORAD & I LONA B R N YK EVEI

PAGE 3

ACTA CARSOLOGICA 36/3 2007 443 mainly consists of sessile oak-h ornbeam forests, wit h dry oak forests on t h e ridges and beec h stands in t h e h ollows and deeper valleys. ere are also a few meadows and non-native coniferous plantations. Most of t h e forests originate from t h e period between t h e two world wars but t h e proportion of forests originating from t h e begin ning of t h e 20 t h century (1906-1926) is also signicant. We collected and integrated t h e available h istori cal data of t h e area into a GIS. It includes some military maps of t h e Habsburg Empire (called Austrian-Hungar ian Monarc h y from 1867) made in 2 periods, namely between 1763-87 and 1819-1869, w h ic h are t h e oldest sources of spatial information of t h is area. Apart from de ning t h e spatial extension of t h e forest at a certain time, t h ese h istorical maps mostly provide indirect information by s h owing t h e location of important buildings (foresters h ouse, roads etc.) or old geograp h ical names t h at mig h t be a h int of t h e c h aracteristic activities of t h at age. Forest management plans and t h e attac h ed maps are t h e most important data sources concerning a for ested area. e arc hive plans were provided by t h e Mis kolc Directorate of t h e State Forestry Service wit h t h e permission of szakerd Forestry Co., t h e present forest manager of t h e area. e rst management plan of t h e Haragistya dates back to 1934. e attac h ed map h olds lots of geograp hical names besides t h e boundaries of t h e forest sections. e temporary plan made in 1944 and t h e rst one made aer t h e 2nd World war are missing. e rst document available from aer t h e war dates back to 1961 and it is followed by plans prepared every ten years in 1973, 1983 and 1993. Fig. 2: e elevation of the study area. In t h e interests of conservation we took p h otos of t h e arc hive plans and t h e attac h ed maps instead of scanning. e data were integrated in Excel 2003 tables w hile t h e maps were georeferenced wit h t h e h elp of ERDAS Imag ine 8.6 soware. First we used polynomial transforma tion but t h e RMS error was about 30-40 meters in most cases because of t h e strong distortion. Since t h e section boundaries are oen dened according to t h e elevation (along valleys or ridges) or existing routes and t h e occa sional modications usually h appen by t h e union or divi sion of t h e already existing sections we nally decided on using t h e rubbers h eet met h od instead. e forestry maps were tted to eac h ot h er using h undreds of GCPs, mainly t h e breakpoints of t h e section boundaries. e maps were t h en digitised and t h e forestry database joined to t h em in ArcView GIS 3.3 soware. For control t h e section area data in t h e database were compared wit h t h e area of t h e corresponding ArcView polygons. e average dierence is about 0.26 0.3 h ectares. Besides t h e data referring to t h e basic c h aracteristics of a forest (production site c h aracteristics, species com position, average h eig h t, average db h diameter at breast h eig h t , etc.), forest management plans also contain a re cord of t h e work performed in t h e given period. In some cases t h ey include historical descriptions and t h e critics of earlier management as well. e content of t h e plans made in dierent periods is not uniform, t h e record ing met h ods can only be considered standardized aer 1973. e unit of forest management is t h e forest section, of w hic h t h e boundary (and consequently t h e size) can c h ange from time to time because t h e boundaries are de ned according to t h e production site c h aracteristics and t h e performed works. e data apply to t h e sections and t h us c h anging t h e boundaries makes temporal analysis rat h er complicated. is practice also means information loss in t h e case of a hig h ly varied surface (like t h at of a karst plateau) because t h e spatial variability of t h e sur face is hig h er t h an t h e resolution of t h e available data. By t h e union of t h e maps from t h e dierent plan ning periods we got polygons representing areas wit h a uniform management history. In t h e period between 1973 and 1993 most section boundaries were unc h anged so c h anges in t h e species composition can be quantita tively analysed. Y et data interpretation s h ould be carried out carefully wit h regard to t h e fact t h at t h e introduc tion of a new species in t h e range of t h e recorded species (eg. Q uercus pubescens or Larix decidua) can signicantly modify t h e data wit h out any major c h ange occurring in t h e examined period. Arc hive aerial imagery can provide useful comple mentary information. e oldest p h otos of t h e area date back to 1952 and were made available by t h e Directorate of Aggtelek National Park. A REVIEW OF THE FOREST MANAGEMENT HISTOR Y AND PRESENT STATE OF THE HARAGIST Y A KARST PLATEAU ...

PAGE 4

ACTA CARSOLOGICA 36/3 2007 444 T HE AGE OF IRREGULAR FOREST USE Before t h e 16 t h century forest use was generally h ap h az ard, wood production was carried out irregularly and only in small areas at t h e same time. People used t h e forest resources but didnt need to consider regenera tion and sustainability. From t h e 16 t h century on wood was used in new ways and production became regular. A special product of karst regions aecting forest use was lime, made in pit kilns. In t h e Gmr-Torna karst area clearcuttings were carried out in large areas in order to satisfy t h e c h arcoal claim of t h e furnaces in t h e nearby Saj valley. Meanw hile grazing, a common practice at t h at age, prevented t h e natural regeneration by seedlings. Forest was not as valuable as arable land or vineyards so w h erever possible it was replaced wit h ot h er forms of agricultural landuse. In t h e remaining forests species composition c h anged in favour of less seeked-out species wit h a better ability to tolerate disturbance and grazing, especially h ornbeam (Carpinus betulus) e impact of grazing was recognised as early as t h e 16 t h century but despite continuous banning it was a c h aracteristic way of forest use in t h e area until t h e 20 t h century (Jrsi 1997). In order to cover t h e need for timber and to h elp natural regeneration some older and larger trees were sparsely le standing as seed trees. is is t h e so-called composite forest, w hic h consists of younger coppice stands wit h a few older trees originating from seedlings. It was a com mon form of forest management in t h e karst area until t h e 20 t h century (Jrsi 1998). Some of t h e seed trees can still be found in t h e older forests. e early history of t h e Haragistya area is similar to t h e w h ole of t h e Gmr-Torna karst region. In t h e Middle Ages it formed part of t h e Szdvr estate, w hic h, accord ing to t h e custom of t h e age, oen c h anged its owners but most of t h e proprietors were private individuals until 1934 (Jrsi 1997, 1998). e earliest available map was pre pared during t h e rst military mapping of t h e Habsburg Empire (1763-87). Nagy (2003) h as processed t h e mate rial of t h e military mappings concerning t h e w h ole area of t h e Aggtelek National Park. e landuse map of t h e 1780s, prepared by him, s h ows t h at t h e Haragistya area was partly used as forest, arable land or dry meadows, but mostly as pasture. e military descriptions attac h ed to t h e maps describe t h e Szdvrborsa forests as sparse and bus h y, w hic h supports t h e evidence of t h e maps. It is also mentioned t h at oak (probably sessile oak) and beec h dominated t h ese forests at t h e time. e memories of common landuse at t h at time live on in t h e geograp hical names found on topograp hical maps and on t h e forestry map of 1934: Cabbage-peak, Ewe-h ollow, S h ep h erdvalley, etc. e traces of strongly eroded agro-terraces in some parts of t h e plateau can also be considered a sign of former arable farming along wit h t h e very s h allow (5-10 cm) soils of ridges and tops, w hic h h old clearings w h ere reforestation is very slow. T HE MANAGEMENT IN THE 19 t h CENTUR Y In t h e beginning of t h e 19 t h century t h e question of t h e sustainability of forests in Hungary h as already arisen and a forestry law was broug h t in for t h eir protection (no. XXI in 1807). But by t h at time forests h ad become t h e main and most reliable prot sources of t h eir owners and consequently t h e provinces w h o h ad t h e executive power were not eager to act in t h e interests of protection. Due to t h at and some legal loop h oles t h e law didnt h ave muc h eect. Wit h t h e construction of new railways clear cuttings started in remote areas w h ere large-scale wood production h ad not been protable before. According to t h e map of Nagy (2003) t h at describes t h e areas landuse in t h e 1850s (based on maps of t h e sec ond military mapping between 1819-1869), reforestation on t h e Haragistya plateau h as started or intensied by t h at time. is is opposite to t h e general tendency of defores tation at t h e time explained by t h e increasing value of t h e forest in relation to ot h er ways of landuse. Apart from a few clearings t h ere h as been continuous forest cover on t h e plateau ever since. In t h e 1870s t h e furnaces in t h e Saj valley switc h ed to t h e use of coal but t h e claim for a new product, t h e tanning bark intensied t h e unfavour able tendencies in t h e areas management policy. e h ar vest cycle of sessile oak forests was reduced to 15-20 years and t h ey were renewed by means of sprouting. ese cop pice stands were gradually degraded and in time h orn beam took t h eir place (Jrsi 1998). Coppicing generally resulted in a decrease of genetic diversity, even in t h e case of species wit h a good sprouting ability. e state and wood quality of sprouted individuals or t h ose cut down several times is usually worse t h an t h ose of trees originat ing from seeds and oen t h eir production rates are lower as well (Bart h a 2001). Hornbeam dominating t h e forests may h ave been unfavourable from t h e wood production point of view but t h e dense juvenile stands of h ornbeam prevented soil erosion and t h us in many cases t h e utter disappearance of t h e forest (Gencsi, Vancsura 1992). e h ig h proportion of coppice forests and h ornbeam in t h e study area in 1934 suggests t h at t h e forests of t h e Harag istya plateau were not an exception eit h er. Because of t h e general degradation of t h e native for ests, in t h e end of t h e 19 t h century coniferous forests were planted in many places; in t h e Aggtelek karst region as R ESULTS AND DISCUSSION E SZTER T AN CS F ERENC S ZMORAD & I LONA B R N YK EVEI

PAGE 5

ACTA CARSOLOGICA 36/3 2007 445 well. In t h e area black pine (P inus nigra) and Scots pine (P inus sylvestris) were preferred especially in protective forests (of w hic h t h e primary role was to prevent soil erosion). ese tries usually proved a failure in t h e long run (Jrsi 1997). According to t h e plan of 1934 t h e P inus nigra forests of t h e Haragistya, probably planted at t h e beginning of t h e century, were only present in two sec tions. F ORESTS ON THE EDGE By t h e beginning of t h e 20 t h century t h e principles of sustainable forest management were worked out in Hungary but neit h er t h e economic environment nor t h e property structure were favourable. e forestry law of 1879 didnt properly regulate private forest use even t h oug h it clearly dened protective forests and intro duced t h e compulsory use of management plans in t h e case of state-owned forests (Jrsi 1997). Information on t h e state of t h e Haragistya forests at t h e beginning of t h e 20 t h century can be deduced on t h e basis of t h e manage ment plan of 1934 and of certain c h aracteristics of t h e present age distribution and species composition. Most of t h e trees originate from coppice stools, w h ic h s h ows t h at aer t h e last h arvest cuttings t h e stands h ave not been professionally renewed; t h e present forests regen erated due to mainly spontaneous processes (sprouting). Traces of c h arcoal burning and t h e former lime kilns can still be found in t h e valleys. ese activities result ed in t h e appearance of pioneer species in t h e h ollows, like birc h (Betula pendula) and trembling aspen ( P opu lus tremula) bot h preferring acidic soils. Birc h being a s h ort-lived tree, is about to disappear from t h e area; a few can still be found standing but more oen it can be seen in t h e form of deadwood. e wars and t h e economic crisis in t h e rst h alf of t h e 20 t h century increased t h e need for timber and made it impossible to bring in proper legislation or even en force t h e existing laws. According to t h e peace treaty of Trianon in 1920 Hungary lost two-t hird of its original area and most of its forests. is h ad a subtle but impor tant long-term eect on t h e Haragistya since it resulted in t h e plateau becoming a perip h eral area. Its c h aracter istic triangle-s h aped road network can already be dis covered on t h e military maps. e fork of t h e two main roads to Szilice (in t h e nort h-east) and Szdvrborsa (in t h e nort h west) was situated in t h e sout h ern part of t h e area. Since t h e Trianon treaty bot h settlements h ave be longed to Slovakia, t h erefore t h e plateau h as become a perip h eral area and t h e roads t h at h ad been used daily before h ave been gradually degraded to forestry access roads. is may h ave contributed to t h e gradual decrease of ant hropogenic impact in t h e 20 t h century. e increasing demand for wood aer t h e 1st World War meant an increasing pressure on t h e remaining for ests, especially t h at during t h e economical crisis most forest owners h ad no ot h er income. Productions exceed ing t h e plans were more easily allowed because of t h e wood s h ortage and because t h e state h oped for more tax income (Jrsi 1997). e result of t h ese tendencies is still mirrored in t h e age distribution of t h e forests in t h e Aggtelek area. In 1934 t h e estate w h ere t h e Haragistya belonged was boug h t by t h e Treasury and became part of t h e Szin management unit (Jrsi 1997, 1998). e rst manage ment plan covering our study area was made at t h at time. e age data of t his plan are contradictory as many of t h e sections h ave two data; besides t h e ones in t h e ta bles records t h ere are some scribbled pencil notes w hic h s h ow rat h er hig h er values. We assume t h at t h ese notes s h ow t h e ages of some remnants of older forests, w hic h were cut down during t h e following war. Y et most forests were only a few years old at t h e time, especially in t h e sout h ern area, w hic h is closest to t h e nearest settlements. e species composition data only h old information on a few species (Fig. 3a); according to t his t h e forest mainly consisted of sessile oak (Quercus petraea) and h ornbeam (Carpinus betulus) wit h trembling aspen (P opulus tremu la) in some places and wit h beec h (Fagus sylvatica) in t h e sout h east. (W h en comparing t h e distribution of beec h in t h e year 1934 to t h e military descriptions, it is surprising to see h ow t h e role of t his species diminis h ed over t h e two centuries). Some ot h er species are included in t h e foot notes, for example wild service (Sorbus torminalis) seed trees are mentioned in several cases. In some sections in t h e nort h west black pine (P inus nigra) was also present, probably as t h e result of earlier plantation works. Wood production at t h at time was mainly concen trated to t h e western and nort h ern parts of t h e area and in most cases it meant clear-cutting. e increasing wood claim of t h e war preparations can be clearly seen in t h e increasing tendency of t h e production data of t h e Harag istya area from 1939 (Fig. 4) e fact t h at t h e cities of Kassa and Rozsny were returned to Hungary in 1938 wit h out t h eir forests (due to t h e First Vienna Arbitration Award) probably also played a role in t his; t h e remain ing areas h ad to cover t h e increased wood claims (Jrsi 1998). In 1942 and 1943 t h ere were no production works in t h e Haragistya area, probably due to t h e increasing in tensity of war activities and t h e following lack of labour. e data of t h e works carried out in 1944 are but pencilscribbled notes on s h eets; t h e wood was used for satisfy ing local needs (t h at of t h e sc h ool, t h e village judge, etc.). e s h are of t h e forest manager signing t h e plan was also provided from t h e Haragistya forests. e temporary plan made in 1944 and t h e rst one made aer t h e 2 nd A REVIEW OF THE FOREST MANAGEMENT HISTOR Y AND PRESENT STATE OF THE HARAGIST Y A KARST PLATEAU ...

PAGE 6

ACTA CARSOLOGICA 36/3 2007 446 World war in t h e beginning of t h e 1950s are missing so t h ere are no data available from t h e reconstruction pe riod. Y et it is known t h at by t h e end of t h e II World War and t h e reconstruction t h ere were almost no forests le suitable for production in t h e Aggtelek area. Oen t h ere were no proper nancial sources for renewal and even if it was successful, it was usually based on sprouting (Jrsi 1997, 1998). T HE STATE OF FORESTS AFTER THE II. W ORLD W AR Aer t h e war t h e most important tasks of forestry were to make up for t h e unnis h ed regeneration works, to plant fast-growing species (conifers, etc.) and to convert t h e degraded coppice forests (oen by replacing t h em wit h non-native coniferous plantations). Accordingly in t h e years aer t h e war t h ere were almost no produc tion works, forestry concentrated on renewal. C h anges in t h e property structure aer t h e war were favourable for long-term planning: during t h e agrarian reform in 1945 all forest areas larger t h an approx. 142 acres were taken into state management. is made it possible to apply t h e principles of sustainable management and to correct ear lier mistakes (Jrsi 1997). From t h e 1950s regular man agement, based on a profes sional background, gradually took over. is practically de termined t h e fate of t h e for ests in t h e Aggtelek karst re gion until t h e area was placed under protection. Despite t h e missing data, aerial p h otos from 1952 and 1956 provide some in formation on t h e state of t h e Haragistya forests in t his pe riod (Fig. 5) ese s h ow open, grove-like forests in muc h of t h e area, w hic h, as it can be seen on t h e aerial p h oto from 1997, h ave almost entirely closed in (probably due to t h e declination of grazing). is is conrmed by t h e presence of dead juniper ( J u niperus communis) in almost every forest type, w hic h is t h e sign of t h e formerly more open vegetation. On t h e slowly reforesting clearings of t h e sout h ern slopes t h ere are still a few wit h ering junipers, w hile under t h e beec h stands of t h e valleys t h e remnants of individuals wit h ered decades ago sign t h e closure of t h e vegetation. According to t h e sto ries of t h e in h abitants of Jsvaf (t h e nearest village), t h e foresters h ouse situated in t h e nort h-western part of t h e plateau was still in h ab ited at t h e time and t h e farm was ouris hing. Livestock was grazed in t h e clearings around t h e h ouse, w hic h prevented reforestation (to day t h e national park keeps up some of t h ose clearings by mowing). In t h e vicinity of t h e foresters h ouse t h ere was also rat h er more intensive management activity t h an in t h e remote parts of t h e plateau. A lot can also be learnt about t h e post-war works from t h e management plan made in 1961. ere was evi dently a signicant modication of t h e section boundar ies compared to 1934. More and smaller sections were formed, probably in order to adapt t h em to t h e c h ar acteristics of t h e forest stands and t h e production sites. is plan also h olds information on more species t h an t h e 1934 one (Fig. 3b). In t h e nort h ern and nort h-western part of t h e area coniferous plantations (mainly Scots pine, spruce and in a few places black pine) appear on t h e species composi tion map of 1961 in t h e place of native forests cut down Fig. 3: Species composition in a) 1934 b) 1961. Fig. 4: e yearly volume of wood production 1935-1992. E SZTER T AN CS F ERENC S ZMORAD & I LONA B R N YK EVEI

PAGE 7

ACTA CARSOLOGICA 36/3 2007 447 earlier. Because of t h e wood claim of t h e post-war recon struction typical management at t h e time included t h e plantation of fast-growing non-native conifers. Forest managers aimed to utilise t h e poor-quality karstic sites by introducing t h ese species. Increasing t h e proportion of conifers in t h e Aggtelek area was a proclaimed aim of management even in t h e 1970s alt h oug h t h e last P icea abies plantation (started in 1964) on t h e Haragistya pla teau was nis h ed at t h at time. It is important to note t h e relatively hig h number of sections (10) w h ere t h e stands are s h own to h ave con sisted entirely of sessile oak. e average size of t h ese sections (7.74 h a) s h ows t h at t h ese were not all smaller stands. e management plan of 1973 emp h asized t h e need to suppress h ornbeam in favour of sessile oak; t his endeavour h as probably been present in forest manage ment already before and t h ey tried to suppress h ornbeam by carrying out stand treatment in favour of t h e sessile oak. Beec h appears in an apparently larger area t h an in 1934. Besides species selection carried out in favour of t his species, t h e modication of t h e section boundaries could h ave also played a role in t his since beec h in t his area can usually be found in small patc h es in t h e valleys and on nort h ern slopes and its proportion was probably only a fraction in t h e large sections. By 1961 beec h also appeared in sections w h ere it h ad not yet been present in 1934 (alt h oug h only in t h e sout h west). Wood production in t h e 1960s was not signicant as it mainly came from t hinning works rat h er t h an clearcutting (Fig.6) Sessile oak provided t h e hig h est volume, followed by h ornbeam and beec h. e young conifers were sold for C hristmas trees. T HE LAST ACTIVE PERIOD OF MANAGEMENT In 1978 t h e Aggtelek Landscape Protection Area was formed in t h e karst area and in 1979 a small part of t h e plateau (compartment 20) was also designated UNESCO MAB Biosp h ere Reserve. Y et in t h e management plan of 1973 most of t h e Haragistya is already labelled pro tected or strictly protected as t h e surface protection zone of t h e nearby Vass Imre cave. Besides an eort to simplify data management t h is mig h t h ave been t h e cause of anot h er signicant c h ange in t h e average size of t h e forest sections, w h ic h increased from 7.72 h a to 11.23 h a). As a result, t h e species composition of t h e sections seemingly became more diverse (Fig. 7a). e spreading of spruce continued according to plan in t h e period between 1961 and 1973 and by 1973 t h e conifer ous plantations h ave appeared in most sections in t h e nort h -western h alf of t h e area. e conifers were mostly planted toget h er wit h native species, pure spruce stands can only be found in very small (0,2-0,3 h a) sections. e proportion of conifers was usually increased at t h e expense of sessile oak stands. In t h e sout h -eastern part of t h e plateau sessile oak, h ornbeam and beec h were dominant, along wit h trembling aspen in certain places. e proportion of beec h ap parently increased in t h is period. Downy oak (Quer cus pubescens) appears in t h e dataset for t h e rst time in 1973 in a few sections in t h e east. It h ad surely been t h ere even before but since it doesnt play an important role in wood production it was probably ignored. Aer 1973 t h ere were only lesser modications of t h e section boundaries so in t h e case of t h e unc h anged sections it is possible to Fig. 5: Aerial photos of the northern part of the H aragistya plateau a) 1956 b) 1997. Fig. 6: e yearly volume of wood production by species 1961-1972. A REVIEW OF THE FOREST MANAGEMENT HISTOR Y AND PRESENT STATE OF THE HARAGIST Y A KARST PLATEAU ...

PAGE 8

ACTA CARSOLOGICA 36/3 2007 448 quantify c h anges in t h e species composition. We concen trated on t h e data of sessile oak, h ornbeam and beec h since t h ese are t h e dominant native species of t h e area and t h e future c h anges in t h eir proportions are of great importance. In t h e period between 1973-1982 wood production still mainly came from t hinning works. In 1976 incre ment t hinning was carried out in a few sections; t h e trees removed were mostly sessile oak and h ornbeam (Fig. 8). ere were only lesser c h anges in t h e species composi Fig. 7: Species composition in a) 1973 b) 1983. tion at t h e time (Fig. 7b). e proportion of sessile oak in t his period slig h tly increased in a few sections (mainly due to t h e removal of ot h er spe cies) w hile t h at of h ornbeam slig h tly decreased. C h anges in t h e two species propor tion usually h ave opposite tendencies. e spreading of h ornbeam usually results in t h e wit h drawal of oak w hile forest management practices usually aim to suppress h orn beam and aid t h e spread ing of oak. e proportion of beec h in t his period only c h anged in 3 sections; it increased in all 3 cases, due to management works. In t h e period between 1983 and 1992 some major c h anges occurred. By t h e 1990s ot h er functions of t h e forests were increasingly accepted and besides t h e eco nomic interests forest management concentrated more and more on ot h er aspects, like conservation and tour ism. On 1st January 1985 t h e Haragistya area became part of t h e newly declared Aggtelek National Park. Al t h oug h some large-volume increment t hinnings took place in 1985 and 1986 nal cuttings were still not carried out. Aer 1986 major wood production works only con tinued in t h e coniferous for ests, due to restrictions from t h e conservation aut h orities (Fig. 9). In most of t h e area t h e proportion of sessile oak de creased in t his period w hile t h ere was a signicant in crease in t h e proportion of h ornbeam at t h e same time (Fig. 10). ese c h anges are complementary in most sections (Fig. 11) except t h ose w h ere beec h started to spread. Since during t h e t hinning works t h e wood produced mainly consisted of sessile oak, t h e decrease in its proportion is not surpris ing; alt h oug h it is interesting t h at t h e attempts to suppress Fig. 9: e yearly volume of wood production by species 1983-1992. Fig. 8: e yearly volume of wood production by species 1973-1982. E SZTER T AN CS F ERENC S ZMORAD & I LONA B R N YK EVEI

PAGE 9

ACTA CARSOLOGICA 36/3 2007 449 h ornbeam ceased. Nevert h eless t h e tendency is also pres ent in sections w h ere production works were not carried out at t h e time. e proportion of beec h mainly increased in areas wit h western aspect. T h e description of t h inning works carried out in spruce forests in t h is period often contains t h e note cut down for h ealt h reasons, w h ic h s h ows t h at t h ese stands are not too stable. According to t h e h ealt h data in t h e management plan of 1993, most of t h e damaged stands (10-50% of t h e w h ole section area) contained conifers. In 1993 t h e sout h-eastern part of t h e area was desig nated forest reserve and consequently all forest manage Fig. 10: Species composition in 1993. ment activity ceased in t h e core areas of t h e reserve. e directorate of t h e national park still permitted produc tion in t h e coniferous forests and some t hinning works in t h e buer zone. C HANGES IN THE AREAS OCCUPIED B Y THE DIFFERENT TREE SPECIES FROM THE FIRST HALF OF THE 20 t h CENTUR Y UNTIL 1993 Besides overviewing t h e ex periences of t h e dierent historical periods examining t h e areal c h anges of t h e tree species over t h e 6 decades of w hic h t h e data are available broug h t interesting results and new questions. In t h e study area sessile oak h as been dominant over t h e w h ole period; from 1934 its area h as been slig h tly de creasing (Fig.12.) despite t h e fact t h at in t h ese 60 years t h ere were no nal cuttings in any oak stands. e loss of area may be explained by t hinnings carried out at t h e expense of t h e oak stands al t h oug h management s h ould h ave served t h e spreading of t his species. e best example of unprofessional concen trated t hinnings are t h e pro duction works carried out in 1985 (nearly 2500 m 3 of ses sile oak was cut down); t h e impact of t h ese will probably inuence t h e look of t h e involved stands for decades. e ot h er signicant species c h aracteristic of t h e area is h ornbeam, w hic h appears as t h e rival of sessile oak. e hig h proportion of h ornbeam in t h e young stands of t h e 1930s was, over time, partly reduced by t h e articial plantation of conifers (spruce, Scots pine and black pine) and to a lesser extent by t h e spontaneous spreading of beec h. Aer t h at t h e area of t his species stagnated for 3040 years (Fig.13.). From t h e 1980s t h e area of h ornbeam h as started to increase again, mainly because of t h e t hin nings t h at resulted in t h e suppressing of sessile oak. Fig. 11: Changes in the proportion of a) sessile oak b) hornbeam between 1983-1992. A REVIEW OF THE FOREST MANAGEMENT HISTOR Y AND PRESENT STATE OF THE HARAGIST Y A KARST PLATEAU ...

PAGE 10

ACTA CARSOLOGICA 36/3 2007 450 pine was present in only two sections. Due to t h e cam paign-like planting of conifers aer t h e II. World War t h eir area grew fast. In t h e 1970s t h ey already covered 10% of t h e study area (Fig. 15). Later, due to t h e lack of space, t h is increase h as stopped w h ile t h ese days it is nature conservation t h at put an end to t h e spreading of t h ese species. Beec h s h ows a slow but constant increase of area over t h e examined period (Fig. 14.). e reason for t his is probably t h e fact t h at it was strongly suppressed in t h e past. is species, now occupying a fraction of its origi nal area, is slowly conquering its former sites again. Finally it is important to mention t h e non-native coniferous species t h at were planted on t h e plateau. eir proportion in t h e 1930s was insignicant; black Fig. 12: e area (ha) occupied by sessile oak (Quercus petraea). Fig. 13: e area (ha) occupied by hornbeam (Carpinus betulus). Fig.14: e area (ha) occupied by beech (Fagus sylvatica) Fig. 15: e area (ha) occupied by conifers (P icea abies, P inus sylvestris, P inus nigra). C ONCLUSION e forests of t h e Haragistya plateau (situated in t h e Ag gtelek karst region) serve as a good example of karstic forest management in Hungary. ey h ave been subject to signicant h uman impact over t h e centuries. Aer t h e area being used partly as agricultural land, forest cover h as been continuous on t h e plateau since t h e beginning of t h e 19t h century. Traces of t h e exaggerated use (ex treme production rates, coppicing and ot h er forest uses) E SZTER T AN CS F ERENC S ZMORAD & I LONA B R N YK EVEI

PAGE 11

ACTA CARSOLOGICA 36/3 2007 451 c h aracteristic at t h e end of t h e 19t h and t h e beginning of t h e 20t h century can still be discovered in todays spe cies composition and stand structure. Human impact gradually decreased over t h e 20t h century and wit h t h e area being taken into state management t h e irregular and oen exaggerated use h as stopped. In t h e middle of t h e century coniferous species were planted on t h e karst pla teau in t h e interests of a more intensive utilisation, as was t h e case in a lot of ot h er karst areas. e volume of wood production h as s h own a gradually decreasing tendency from t h e 1960s; at rst t h ere was a lack of forests suit able for production w hile production works were limited w h en t h e area was placed under protection. e propor tion of sessile oak and h ornbeam dominating t h e plateau in t h e rst h alf of t h e 20t h century gradually decreased, partly due to t h e coniferous plantations in t h e nort h, partly because of t h e spontaneous spreading of beec h in t h e sout h east. Finally t h e realisation of t h e importance of karst surface protection led to t h ese forests being placed under protection. us t h e proportion of old stands (more t h an 80 years old) on t h e plateau is hig h er t h an t h e national average. e relatively slow growt h of t h ese forests, due to earlier coppicing and t h e s h allow karstic soils also played a major role in t his. A CKNOWLEDGMENT e researc h was funded by t h e Hungarian Scientic Re searc h Fund (OTKA/T048356). e aut h ors express t h eir t h anks to t h e szakerd Forestry Company for t h eir per mission to use t h e data, to Sndor Bnyei, deputy direc tor of e Miskolc Directorate of t h e Forestry Service for his h elp and to Zoltn Szabados, geograp h y student at t h e University of Szeged, for his work! R EFERENCES Bart h a D. 2001: Veszlyeztetett erdtrsulsok Mag yarorszgon (Endangered forest associations in H un gary) .WWF fzetek, 18, WWF Magyarorszg, Bu dapest (h ttp://www.wwf.h u/wwuzetek.p h p) Brads h aw R. H. W. 2004: Past ant hropogenic inuence on European forests and some possible genetic con sequences. Forest Ecology and Management 197 (2004) 203 Gencsi L.,Vancsura R. 1992: Erdszeti nvnytan II. (Dendrolgia Dendrology ).Mezgazda Kiad, Budapest. Jrsi L. 1997: Erdgazdlkods Bnkttl Nagy-Milicig (Forest management from Bnkt to Nagy-M ilic). szakerd Rt., Miskolc. Jrsi L. 1998: Az erdgazdlkods mltja (e past of forestry in Aggtelek National Park). In: Baross G. (szerk.): Az Aggteleki Nemzeti Park. Mezgazda Kiad, Budapest, 395-410. Karst management h andbook for Britis h Columbia. For. B.C. Min. For., Victoria, B.C. h ttp://www.for.gov. bc.ca/hfp/fordev/karst/karstbmp.pdf Kevein Brny I. 2004: A karsztkolgiai rendszer szerkezete s mkdse (e structure and function ing of the karstecosystem). Karsztfejlds (BDF Ter mszetfldrajzi Tanszk, Szombat h ely), 9, 65-74. Nagy D. 2003: Tjtrtneti kutatsok a Gmr-Tornaikarszton I. A trtnelmi tj rekonstrukcija az ANP krnyezetben az I-III Katonai Felmrsek alapjn (Landscape history research in the Gmr-Torna karst area I. e reconstruction of the historical land scape in the vicinity of Aggtelek National P ark on the basis of the I-III military maps). ANP fzetek, 2, 107-143. Temesi G. 2002 Az erdrezervtumok fenntartsnak l talnos irnyelvei (General guidelines of forest reserve management). In: Horvt h F., Bor hidi A. (szerk.): A h azai erdrezervtum-kutats clja, stratgii s mdszerei. TermszetBvr Alaptvny Kiad, Budapest, 38-45. A REVIEW OF THE FOREST MANAGEMENT HISTOR Y AND PRESENT STATE OF THE HARAGIST Y A KARST PLATEAU ...



PAGE 1

FOSSIL VERTEBRATES AND PALEOMAGNETISM UPDATE OF ONE OF THE EARLIER STAGES OF CAVE EVOLUTION IN THE CLASSICAL KARST, SLOVENIA: PLIOCENE OF RNOTIE II SITE AND RAIKA PEINA CAVE F OSILNI VRETENARJI IN PALEOMAGNETNA POSODOBITEV ENE OD ZGODNJIH FAZ RAZVOJA JAM NA K LASINEM KRASU S LOVENIJA : PLIOCEN RNOTI II IN R AIKE PEINE Ivan HOR EK 1 Andrej MIHEVC 2 Nadja ZUPAN HAJNA 2 Petr PRUNER 3 & Pavel BOSK 2,3 Izvleek UDK 551.44:550.38(497.4 rnotie) Ivan Horek, Andrej Mihevc, Nadja Zupan Hajna, Petr Pruner & Pavel Bosk: Fosilni vretenarji in paleomagnetna posodobitev ene od zgodnjih faz razvoja jam na Klasinem krasu, Slovenija: pliocen rnoti II in Raike peine Na Klasinem krasu so paleontoloki podatki prvi omogoili uskladitev magnetostratigrafski h podatkov z geomagnetno polarizacijsko asovno skalo. Na dve h proueni h mesti h: (i) v prolu s sigami in rdeimi ilovicami v Raiki peini (Matarsko podolje) in (ii) v brezstropi jami, lokacija rnotie II (Podgor ski kras), ki je zapolnjena s uvialnimi klastinimi sedimenti prekritimi s sigo. Ta zadnja lokacija se odlikuje z bogatim na h ajaliem fosilni h cevic jamskega serpulida M arifugia ca vatica. Ostanke vretenarjev predstavljajo predvsem fragmenti sklenine glodalcev in soricomorfov. Odsotnost volu h aric brez zobni h korenin, kakor tudi taksonomska sestava sesalske favne, kaeta na pliocensko starost obe h lokacij. Favna iz (i) Raike peine (z Apodemus cf. Borsodia ) pripada srednji do pozni bioconi MN17 (ca.1.8-2.4 Ma). Fosilna zdruba (ii) iz rnoti II (z Deinsdora sp., Beremedia ssidens, Apodemus cf. atavus, Rhagapodemus cf. frequens, Glirulus sp., Cseria sp. ) je oitno precej stareja in pripada bioconi MN15MN16 (ca 3.0.1 Ma). Z upotevanjem skladnosti biostratigrafski h in paleomag netni h datacijski h podatkov in verodostojnega sedimentaci jskega okvirja vzorcev predlagamo, da se tako dobljeno datacijo uporabi tudi kot as, ko se je konala ena od stari h speleogenet ski h faz na Klasinem krasu. Kljune besede: jamski sedimenti, magnetostratigraja, sesalci, Soricomorp h a, Rodentia. 1 Department of Zoology, Faculty of Science, C h arles University, Vinin 7, 128 44 Pra h a, Czec h Republic, e-mail: h oracek@natur.cuni.cz 2 Karst Researc h Institute, SRC SASA, Titov trg 2, 6230 Postojna, Slovenia, e-mail: mi h evc@zrc-sazu.si; zupan@zrc-sazu.si 3 Institute of Geology AS CR, v.v.i., Rozvojov 269, 165 00 Pra h a 6, Czec h Republic, e-mail: pruner@gli.cas.cz; bosak@gli.cas.cz Received/Prejeto: 02.10.2007 COBISS: 1.01 ACTA CARSOLOGICA 36/3, 453-468, POSTOJNA 2007 Abstract UDC 551.44:550.38(497.4 rnotie) Ivan Horek, Andrej Mihevc, Nadja Zupan Hajna, Petr Pruner & Pavel Bosk: Fossil vertebrates and paleomagnetism update of one of the earlier stages of cave evolution in the Clas sical karst, Slovenia: Pliocene of rnotie II site and Raika peina Cave For t h e rst time in t h e Classical Karst, paleontological data enabled to matc h t h e magnetostratigrap hic record precisely wit h t h e geomagnetic polarity timescale in two studied sites: (i) a series of speleot h ems alternating wit h red clays in Raika peina Cave (Matarsko podolje), and (ii) an unroofed paleo cave of t h e rnotie II site (Podgorski kras Plateau) completely lled by uvial clastic sediments covered by speleot h ems. e later sites are also c h aracterized by a ric h appearance of fossil tubes of autoc h t h onous stygobiont serpulid M arifugia cavatica e vertebrate record is composed mostly of enamel fragments of rodents and soricomorp h s. Absence of rootless arvicolids as well as taxonomic composition of t h e mammalian fauna suggests t h e Pliocene age of bot h sites. For (i) Raika peina (wit h Apodemus cf. Borsodia ) it was estimated to middle to late MN17 (ca 1.8.4 Ma), w hile (ii) t h e assemblage from rnotie II (wit h Deinsdora sp., Beremedia ssidens, Apodemus cf. ata vus, Rhagapodemus cf. frequens, Glirulus sp., Cseria sp. ) is obvi ously quite older: MN15MN16 (ca 3.0.1 Ma). In respect to congruence of biostratigrap hic and paleomagnetic data and a reliable sedimentary setting of t h e samples we propose to apply t h e respective datum also as t h e time of one ancient speleoge netic p h ase in t h e Classical Karst. Keyword: cave sediments, magnetostratigrap h y, mammals, Soricomorp h a, Rodentia.

PAGE 2

ACTA CARSOLOGICA 36/3 2007 454 e sedimentary ll of caves (cave sediments) in t h e Clas sical Karst of SW Slovenia was expected not to be muc h older t h an 350 ka (Gospodari, 1972, 1974, 1976,1977, 1981,1984, 1985, 1988). is view was based on results of sedimentologic researc h, arc h eologic and paleon tologic nds, and on numerical dating (like 14 C, /U, ESR; a.o., Franke & Gey h, 1971; Ikeya et al., 1983; Ford & Gospodari, 1989) from many sedimentary proles in Postojna, Planina, Krina, kocjan and ot h er caves. Gospodari (1988) distinguis h es dierent deposition p h ases of cave sediments related to glacioeustatic oscil lations of t h e Adriatic Sea and t h e global paleoclimate evolution during Pleistocene. Later researc h es in several caves of Notranjska and Primorska karst regions (central and western Slovenia) indicate t h e existence of many owstones older t h an 350 ka, i.e. out of limit of t h e met h od (Zupan, 1991; Mi h evc & Lauritzen, 1997; Mi h evc, 2001). Geomorp h ologic com parative met h od s h ows t h at many accessible caves in t h e Classical Karst are at least of t h e Pliocene age (Mi h evc, 1996, 2001). Nevert h eless, paleontological nds indicate only Pleistocene age of fauna in studied cave sediments (Brodar, 1952, 1958, 1966, 1970; Rakovec, 1958; Aguillar et al., 1998) Only t h e application and interpretation of palaeo magnetic analyses and magnetostratigrap h y of cave sedi ments, bot h clastic and c h emogenic, started in t h e Clas sical Karst in 1997, indicate t h e substantial c h ange in t h e lower limit of cave ll deposition (Bosk et al., 1998, 1999, 2000a, b, 2003; 2004a, b; ebela & Sasowsky, 2000; Prun er & Bosk, 2001; Mi h evc et al., 2002; Sasowsky et al., 2003; Zupan Hajna et al., 2005, 2007). Magnetostratigra p h y data and t h e arrangement of obtained magnetozones oen indicate ll ages of more t h an 1.77 Ma wit h pos sibility of a lower limit even more t h an 5 Ma (Bosk et al., 1998). e interpretation of magnetostratigrap h y is based on t h e matc hing of interpreted results wit h t h e global geomagnetic polarity scale (GPTS; Cande & Kent, 1995) and only oers t h e correlative ages ( sensu Col man & Pierce, 2000). e proper correlation of obtained magnetostratigrap h y wit h t h e GPTS is possible only in t h e combination wit h ot h er dating met h ods (numerical or calibrate), and especially wit h paleontological dates (Bosk, 2002; Bosk et al., 2003). Nevert h eless, paleonto logical data for t h e calibration wit h t h e GPTS h ave been missing, until t h e rst nds of mammal remains in t h e Raika peina in 2003 (Bosk et al., 2004b) and subse quent discovery in t h e rnotie II prole in 2005. Here, we are using t h e term Pliocene in a tradi tional sense. e Plio-Pleistocene boundary is located at 1.8 Ma (Aguirre & Pasini, 1985), close to t h e base of Olduvai normal subc hron (C2n) at 1.77 Ma (wit hin t h e Matuyama C hron). e newly proposed Plio-Pleistocene boundary at about 2.6 Ma (close to Matuyama/Gauss boundary, base of C hron C2Ar at 2.581 Ma and at t h e base of t h e MIS 103 at 2.588 Ma; Ogg, 2007) h as not yet been approved by t h e IUGS. DESCRIPTION OF SITES INTRODUCTION e Classical Karst of W Slovenia belongs to Adriat icDinaric Carbonate Platform of t h e Outer Dinarids composed of s h allow marine Cretaceous and Paleogene carbonate rocks ( cf. e.g., Otoniar, 2007). e Eocene deep-marine siliciclastics (ysc h) encircle and cover t h e carbonate plateau. Complicated imbricate structure (t h e alternation of ysc h and limestone t hrust slices) paral lel to t h e Dinaric direction h as been formed since Oli gocene. Karst landscape of t h e region is c h aracterized by extensive leveled surfaces, blind valleys at limestone/y sc h contacts, and old caves in places dissected by younger s h as. Some caves were uncovered at t h e present surface by c h emical denudation (unroofed caves sensu Mi h evc 1996, for details see Mi h evc 2001). e origin of s h as is connected wit h substantial drop of piezometric level, w hic h now lies about 200 m below t h e surface. S h as are eit h er empty or lled wit h Pleistocene sediments (e.g., Rakovec 1958; Brodar 1958). ere are several karst plateaus in t h e area (Fig. 1). e largest is Kras t h e NW SE trending region along t h e Golf of Trieste (Adriatic Sea) from t h e Vipava Val ley in t h e NE up to FriuliVenezia Giulia lowlands and t h e Soa River in t h e NW (Fig. 1). Its central part lies at 200 to 500 m a.s.l. (for more details see Kranjc, 1997). M atarsko podolje is situated to t h e SE of t h e Kras. Its karst landscape was formed by denudation on t h e NE dipping limestone and dolomites at t h e elevations of 500 m a.s.l. Allogenic rivers at t h e contact wit h ysc h form sev eral blind valleys wit h traces of constant tectonic upli (Mi h evc, 2004). P odgorski kras P lateau in t h e S edge of t h e Kras is widely extended levelled plateau (440 m a.s.l.) at t h e foot hills of t h e Slavnik Mt. (Mi h evc, 2007; Fig. 1). I VAN HOR EK, A NDREJ MIHEVC, N ADJA ZUPAN HAJNA, P ETR PRUNER & P AVEL BOS K

PAGE 3

ACTA CARSOLOGICA 36/3 2007 455 Denudation surface cuts t h e Paleocene limestone alter nating wit h t hin slices of Eocene ysc h t h at dip nort h eastwards ( cf. Placer 1981). is levelled surface exposes numerous unroofed caves. Large karst springs are located at t h e plateau foot hills at 50 m a.s.l. RNOTIE e rnotie and rni kal quarries are situated on t h e western margin of t h e Podgorski kras. In bot h quarries during excavations numerous caves h ave been opened, most of t h em completely lled by sediments (see review in Bosk et al., 1999, 2004a; Mi h evc, 2007). Dating of cal cite cement in gravels ric h in large mammal bones yield ed age of 211 ka in one s h a of t h e rnotie Quarry (Mi h evc, 2001). Caves in t h e rni kal Quarry contained Paleolit hic tools, Pleistocene large mammals (Rakovec, 1958; Brodar, 1958), and Middle and Late Pleistocene small mammals (Aguilar et al., 1998). Carbonate cement of gravel ll of one s h a was dated to143 ka (Mi h evc, 2001). Palaeomagnetic analysis was used to date pale okarst site of rnotie I in t h e rni kal Quarry (Bosk et al., 1999), a part of an extensive cave lled wit h about 1.75 m t hick banded sandy cave stromatolite intercalated wit h red clays covering eroded surface of older massive and hig h ly recrystallized speleot h ems. Palaeomagnetic analysis indicates t h e age more t h an 1.77 Ma. e nds of problematic s h teet h are still not determined. rnotie I and II sites belong to one paleocave sys tem, w hic h started to be successively opened since 1990. is was about 150 m long relic of cave formed by an al logenic river. e extensive passage h ad t h e diameter of about 10 m wit h t h e dip in t h e NW SE direction. In t h e eastern part, t h e passage was unroofed, but in t h e west ern part t h e ceiling h as been still preserved. e passage was entirely lled by cave sediments deposited over mas sive owstones, several metres t hick. Gravels were pre served and in places mixed up wit h sand and clay. Lami nated yellow brown clays to silts contained fragments of speleot h ems. Reddis h clays composed t h e upper part of proles. Fig. 1: P osition of studied sites. FOSSIL VERTEBRATES AND PALEOMAGNETISM UPDATE OF ONE OF THE EARLIER STAGES OF CAVE EVOLUTION ...

PAGE 4

ACTA CARSOLOGICA 36/3 2007 456 e studied prole (unroofed cave, rnotie II site) was more t h an 17 m hig h and from 4 to 7.5 m wide. Lower nine metres were composed of cyclically arranged cave sediments. It was divided by an expressive erosion boundary into two parts. e lower one was built of multicoloured clays and silts overlain by upwards ning uvial cycles composed of micro conglomerates to con glomerates (clay and silt clasts), w hic h sometimes pass to sands. Individual cycles were separated by t hin interbeds of clays to silts. e upper part, above uneven erosion boundary, consisted of laminated to banded silts, clays and very ne-grained silty sands. e clastic ll was cov ered (above 9 m up to t h e surface) by speleot h ems, and collapsed speleot h em and limestone blocks wit h some terra rossalike red clays at t h e base (Fig. 2). Cave walls built of Palaeocene limestone wit h scallops were covered by tubes of fres h water stygobiont worm M arifugia cavat ica Absolon et Hrab, 1930 about 3 m above t h e quarry oor (Fig. 2; Mi h evc, 2000; Mi h evc et al., 2001, 2002). Was h ed material was obtained from t h e rig h t lower part of t h e prole near t h e cave wall (see Fig. 2). Multicoloured silt and clays deposited from relatively calm cave uvial environment wit h medium rates of deposition. e cyclic deposition resulted from repeated sudden and intensive oods eroding already deposited cave sediments. e rate of deposition is expected to be hig h. RAIKA PEINA Raika peina is situated in t h e SE part Cave of t h e Ma tarsko podolje near t h e Croatian border (Fig. 1). Cave at about 590 m a.s.l. developed in Lower Cretaceous carbon ate rocks and represents t h e relic of an old cave system, w hic h part was already unroofed (Ulica peina Cave and Ulica unroofed cave; Mi h evc, 2004). e Raika peina is paragenetic or epip hreatic simple sout h wards dipping gallery, w hic h is mostly over 10 wide, 5 m hig h and Fig. 2: P hoto of the rnotie II site with position of paleomagnetic samples (white marks), A position of sediment with fauna, B position of M arifugia cavatica on cave wall (photo by P Bosk). Fig. 3: Drawing of the Raika peina prole: F position of fauna nd, grey shading clays, inclined shading collapsed blocks of roof, black boxes paleomagnetic samples, speleothems, black circles paleomagnetic samples, clays. I VAN HOR EK, A NDREJ MIHEVC, N ADJA ZUPAN HAJNA, P ETR PRUNER & P AVEL BOS K

PAGE 5

ACTA CARSOLOGICA 36/3 2007 457 304 m long (Bosk et al., 2004b). On t h e sout h ern side, t h e cave terminates wit h t h e collapse and sedimentary plug. Clays of unknown t hickness covered by massive owstone form t h e bottom of t h e cave. ere are traces of Ursus spelaeus on t h e cave oor (Mi h evc, 2003). Re mains of Pre historic pottery were found at t h e cave en trance. e studied prole represents slig h tly more t h an 3 m hig h and about 20 m long cut in t h e cave ll (Fig. 3) uncovered w h en t h e cave was adapted as t h e military magazine. It is situated in t h e sout h ern part of t h e cave, about 200 m from t h e present entrance. e composite t hickness of sampled prole reac h es about 6.5 m. e lower part is built of t hree sequences, representing t h e growt h stages of a h uge vaulted stalag mite (lig h t brown to reddis h brown mostly corroded calcite) and containing two angular unconformities (ex pressed as t hin intercalations of red clays) wit h broken rests of stalagmites on t h eir tops. e upper part consists of sub-h orizontal laminated, mostly porous and lig h t-co loured owstones intercalated by owstone wit h gours Fig. 4: Correlation of magnetostratigraphic logs of the rnotie II site (le) and the Raika peina (right; simplied) with the GPTS (Cande & Kent 1995; center): black normal polarity, grey transient polarity, white reverse polarity, ~~~ principal hiatus. and red clays and silts. Collapsed roof blocks cover red clays wit h nds of fauna (F on Fig. 3). e topmost part, in t h e t hickness of several tens of centimetres, is com posed of massive owstone layers wit h intercalations of brown cave loams containing large bones (mostly Ursus spelaeus). e prole is covered by h uge stalagmites, w hic h were not studied. e samples for paleontological analyses were taken at 4 principal clay h orizons. All t h e described remains were obtained from t h e layer situated just at t h e base of t h e section (F on Fig. 3). Lutitic intercalations represent results of numerous and intensive oods bringing alloc h t h onous material. Clays are well-sieved and represent t h e deposited suspension of t h e cave stream load. Very ne-grained sand and silt laminas at t h e bottom or inside some layers are rare and occur only in clay beds of hig h er t hickness, and indicate ood pulses wit hin one lutitic layer. e c h aracter of allogenic lutites can indicate t h at studied site occurred far from t h e cave ponor and/or t h at allogenic stream h ad to pass t hroug h a system of sumps. MAGNETOSTRATIGRAPH Y Palaeomagnetic analysis was based on t h e progressive demagnetisation by alternating eld (Sc h onstedt GSD or LDA) or t h ermal demagnetisation (MAVACS). e natural remanent magnetization, as well as volume mag netic susceptibility was measured on JR-5A or JR-6A spinner magnetometres and KL Y or KL Y -3 kappa bridges. e separation of t h e respective remanent mag netisation components was carried out by Kirc h s h vink multi-component analysis (Kirsc h vink, 1980). e statis tics of Fis h er (1953) was applied for calculation of mean directions of pertinent remanence components derived by multicomponent analysis. For details of sampling and analytical procedure see Bosk et al. (1998, 1999, 2000ab, 2003, 2004a). RNOTIE II SITE Bot h parts of t h e prole s h ow normal polarity magneto zones. One narrow reverse polarity subzone is situated in t h e middle part of t h e prole. Anot h er narrow reverse polarity subzone, documented on one sample only, is sit uated in t h e upper part above t h e erosion surface (Fig. 4). FOSSIL VERTEBRATES AND PALEOMAGNETISM UPDATE OF ONE OF THE EARLIER STAGES OF CAVE EVOLUTION ...

PAGE 6

ACTA CARSOLOGICA 36/3 2007 458 PALEONTOLOG Y Was hing of samples taken from t h e Raika peina (about 40 kg from site marked F on Fig. 3) and rnotie II site (300 kg from site marked A on Fig. 2) provided ric h material of microscopic fragments of vertebrate fos sil remains. e appearance of t h e remains, t h eir tap h o nomical c h aracteristics and to a considerable degree also taxonomic composition was quite similar in bot h sites. As a rule t h e individual fossil rests are corroded and fragmented into quite small pieces. Fossils are covered by t hin amorp h ous fossilization crust ric h in Si (?opal), Ca and Cr (microprobe analysis) largely resistant to acid etc hing. is fact essentially complicated t h e SEM study. Altered surfaces under t h e crust are ric h in Ca, Si, Fe, Cr, but lack Mn, Mg and Al (microprobe analysis). Fos sil samples were studied on CAMECA 100 microprobe (Institute of Geology AS CR, v.v.i., Prague) and JEOL 6386 scanning electron microscope (Faculty of Science, C h arles University, Prague). MATERIAL rnotie II site Vast predominance of autoc h t h onous stygobiont serpulid tubes (Marifugia cavatica Absolon et Hrab, 1930) was detected in t h e fossil sample. is taxon, pat terns of its appearance in t h e site, and its paleoecologic and stratigrap hic meaning h as been already discussed (Mi h evc, 2000; Mi h evc et al., 2001, 2002). e vertebrate record consists of 58 items, mostly poorly preserved and corroded fragments of teet h enamel. e best preserved items are gure on Plates 1 to 3. Some of t h em allow at least a tentative identication. e list of material is as follows: Non-mammalian remains (15): ?C h ondric h t h yes indet: 15 small conical and at lanceolate toot h tips. Mammalian remains: Eu lipotyp h la (4): Deinsdora sp. (2I/1), Beremendia ssides (Petenyi, 1864): M/1, M/2; Rodentia (39): indet. fragments of incisor enamel (15 items), Glirulus a. pusillus (Heller 1936): 1 M3/, Apodemus (Sylvaemus) cf. atavus Heller, 1936: 9 fragments (2 M1/, 1M2/, 1M3/, 2M/1, 1M/2, 2M/3), Rhagapodemus cf. frequens Kretzoi, 1959 (1 M/2), Arvicolidae indet. (13 fragments of molar enamel): sp. (cf. Cseria carnutina Rabeder 1981), sp. 2 (cf. M imomys (Cseria s.s.) gracilis ). Raika peina Totally 74 items were obtained. ey were mostly very poorly preserved, fragile, and composed of particu larly small fragments of teet h enamel (max. 1 mm in size), corroded and wit h out a trace of dentine or cemen tum (Pl. 4). e list of material is as follows: Non-mammalian remains (7 pieces) were as follows: cf. P otamon (Crus tacea), 2 tips of small conical teet h (not identied until now), and 3 p h aryngeal pearl teet h of a Cyprinid s h, most probably Barbus sp. Mammalian remains were 67 in number: Rodentia (29 fragments of incisors, 2 fragments of metapodia, 35 fragments of molar enamel of arvicol ids): at least 2 spp. (incl. cf. Borsodia ), 1 M/1 (enamel): Apodemus (Sylvaemus) sp. cf. atavus Heller, 1936. e matc hing of obtained magnetozone arrangements wit h t h e GPTS indicated t h e age of sediments more t h an 1.77 Ma (base of t h e C2n Olduvai subc hron). e ll most probably belongs to t h e Gauss C hron (ca 2.6 to 3.6 Ma) or t h e ot h er normal subc hron (wit hin t h e Gilbert C hron; 4.18.29 or 4.48.62 Ma; Bosk et al., 2004a). RAIKA PEINA Clear paleontological data enabled, for t h e rst time in t h e Classical Karst, to x t h e arrangement of interpreted magnetozones wit h t h e GPTS. Fauna belonging to t h e MN17 biozone was found in red clays in t h e dept h inter val of 3.79 to 4.05 m (Fig. 3 and 4). erefore t h e base of long normal polarized magnetozone terminating wit hin t h e interval wit h fauna was identied wit h t h e base of C2n Olduvai subc hron (1.95 Ma; Bosk et al., 2004b; Fig. 4). Normal polarized magnetozone above t his boundary belongs to t h e Olduvai subc hron (1.770.950 Ma). S h ort reverse polarized magnetozone represents some part of t h e Matuyama C hron. Normal polarized top of t h e pro le belongs to Brun h es C hron (C1n, younger t h an 0.780 Ma), w hic h corresponds wit h t h e c h aracter of fauna in loam interbeds. e interpretation of magnetostratig rap h y below t h e Olduvai base is not completely clear. e geometry of obtained magnetozones is c h anged as compared wit h subc hrons on t h e GPTS due to numer ous principal breaks in deposition. Nevert h eless, we ex pect t h e correlation wit h t h e lower part of t h e Matuyama C hron (2.150.581 Ma) and t h e Gauss C hron (2.581 3.58 Ma). e comparison wit h t h e GPTS can indicate duration of individual breaks in deposition in 150 ka, in t h e lower part probably substantially more (Bosk et al., 2004b). I VAN HOR EK, A NDREJ MIHEVC, N ADJA ZUPAN HAJNA, P ETR PRUNER & P AVEL BOS K

PAGE 7

ACTA CARSOLOGICA 36/3 2007 459 COMMENTS ON PARTICULAR TA X A Non-Mammalia indet. isolated tooth tips (Pl.1:4-6) e conical s h arply tapered to at lanceolate toot h tips wit h s h arp lateral blades present a conspicuous ele ment of t h e fossil samples. e respective structures are covered by a very h ard enameloid built by tiny (less t h an 1 in diameter) and densely packed crystallites and lled wit h compact dentine, in contrast to mammalian teet h not corroded. Until now, we did not succeed in taxonomic identication of t h ese structures. Tentatively, we h ypot h esize t h at t h ey mig h t belong to juvenile s h arks (probably was h ed from Eocene marls by t h e sinking river) or to un known form of bone s h, eventually. Unfortunately, t h e extensive comparisons wit h t h e taxa available at our dis position did not elucidate t h e identication w hic h t h us remains a task for future study. Deinsdora sp. (Pl. 2:1-2) Two lower incisors belonging to a medium sized s hrew (roug h ly of a Sorex araneus size) wit h a c h aracteris tic series of rounded cusps at t h e major occusal ridge and dark colouration at tips belong undoubtedly to a member of Soricini, t h e genera of ot h er tribes (suc h as Blarinella, M aa, Z elceina, P etenyia etc.) can be excluded. Default identication wit h t h e genus Sorex is, of course, not en tirely supported for t h at t h e members of t h e genus of t h e corresponding size appearing prior to Middle Pleistocene (w h en t h e respective size was attained in t h e subgenus S orex cf. Horek & Loek, 1988) were exclusively mem bers of t h e subgenus Drepanosorex (MN17 praearaneus Kormos, 1934, Q1 margaritodon Kormos, 1934), w hic h bear an indistinct and lig h t orange teet h colouration not dark red as in t h e items under study. e full combina tion of t h e c h aracters appearing in t h e respective fossils (including size and colouration of cusp enamel) can be found in anot h er genus of Soricini, Deinsdora Heller, 1963. Also relative s h arp incision at mesial margin of t h e rst ridge cusp (particularly in t h e specimen gured in Pl. 2/2) and relatively inated occlusal basin of t h e toot h (Pl. 2/1) seem to support t h at identication. e genus was described from MN17 site of De insdorf and its separate status was conrmed by furt h er studies includ ing a monograp h by Reumer (1984) w h o synonymized its nominotypi cal species franconica Heller, 1963 was later synonymized wit h hibbardi Sulimski 1962 described (as Sorex hibbardi ) from MN15 site of We (see also Rzebik-Kowalska, 1990). Subsequently, six ot h er species were described from various sites in Hun gary, Poland and Greece ranging from MN14 (Podleice site) to early MN17 (Deinsdorf site) or MN16. e exact species identication of our specimens is not possible t h oug h a broad measure of agreement is wit h bot h hibbardii / franconica and kor dosi Reumer, 1984 (MN16 Csarnta site) for w hic h good information on incisor morp h ology is available (Re umer, 1984). Beremendia ssidens (Petenyi, 1864) (Pl. 2: 3-4) e enamel coat of t h e le M/1 and M/2 (apparently belonging to one and t h e same individual) pro vides a quite reliable morp h ological information: t h e teet h clearly s h ow t h e c h aracters distinguis hing Bere media ssidens, one of index fossils P late 1: Non-mammalian fossils of the rnotie II site: 1-3 M arifugia cavatica Absolon et H rabr, 1930; 4-6 indetermined tooth tips, supposedly a juvenile Chondrichthyes shes. FOSSIL VERTEBRATES AND PALEOMAGNETISM UPDATE OF ONE OF THE EARLIER STAGES OF CAVE EVOLUTION ...

PAGE 8

ACTA CARSOLOGICA 36/3 2007 460 of t h e Pliocene s hrew communities. It concerns t h eir large size (M/1: L 2.40, tlL 0.80, tlW 1.40, M/2: L 1.98, tlL 0.5, tlW 1.25), c h aracteristically s h aped entoconid, mod erately developed labial cingulum or a dark red coloura tion in major crown tips. e teet h are at lower margin of t h e metric variation of t h e species, distinctly smaller t h an Blarinoides mariae Sulimski, 1962 (t h e species w hic h also diers in proportions of M/2 and oen in lig h ter colouration of enamel) t h oug h t h ey are larger t h an Beremendia minor RzebikKowalska, 1976 t h e form reported from four MN14-MN16 localities. Beremendia ssidens was eu constant element of t h e Pliocene mammalian communities from Spain, France and England to Po land, Ukraine, Romania, Croatia and Greece. oug h its rst ap pearance is reported even from late MN14 (Podleice site) as a common species it appeared particularly in MN16 and MN17 (Rzebik-Kowalska in 1998 listed at least 30 sites of t h at age). It was regular but rare element even in t h e midEuropean Q1 com munities (as well as in Italy Gliozzi et al., 1997) w hile during Q2 biozone (late Bi h arian) it was in most parts of its former range already extinct (van den Hoek et al., 2005) Glirulus sp. (Pl. 3: 1) Nearly complete fragment of a crown enamel of glirid toot h (M3/ d. or P4/ d.) represents quite important item of t h e collection. e toot h is remarkable for its very small size (L 0.70 x W 0.82) and besides of four main ridges (connected wit h a com plete endolop h) it bears quite a dis tinct secondary ridge between t h em. e occlusal surface is relatively at and not deected as in extant genera Dryomys, Eliomys or M yomimus (in cluding t h e smallest form M yomimus setzeri Rossolimo, 1976 to w hic h it corresponds in size) w hic h may come in account for biogeograp hic reasons (Krytufek & Vo hralk, 2005). e above mentioned dier ences, w hic h exclude co-identica tion of t h e fossil item wit h t h ese extant taxa and t h e PlioPleistocene forms related to t h em (suc h as Dryomimus eliomyoides Kretzoi, 1959) t quite well to t h e diagnostic c h aracters of Glirulus pusillus, t h e form described as Am phidyromys pusillus Heller, 1936 from MN15/16 site of P late 2: Soricidae and M uridae of the rnotie II site: 1 Deinsdora sp. le I/1; 2 Deinsdora sp., right I/1; 3 Beremendia ssidens (P etenyi, 1864), le M/1; 4 Beremendia ssidens (P etenyi, 1864), le M/2; 5 Apodemus (Sylvaemus) cf. atavus H eller, 1936, le M1/;; 6 Apodemus (Sylvaemus) cf. atavus H eller, 1936, le M1/; 7 Apodemus (Sylvaemus) cf. atavus H eller, 1936, right M2/; 8 Apodemus (Sylvaemus) cf. atavus H eller, 1936, right M3/; 9 Apodemus (Sylvaemus) cf. atavus H eller, 1936, right M/1; 10 Apodemus (Sylvaemus) cf. atavus H eller, 1936, right M/1; 11 Apodemus (Sylvaemus) cf. atavus H eller, 1936, le M/2; 12 Apodemus (Sylvaemus) cf. atavus H eller, 1936, right M/3;13 Rhagapodemus cf. frequens Kretzoi, 1959, le M/2. I VAN HOR EK, A NDREJ MIHEVC, N ADJA ZUPAN HAJNA, P ETR PRUNER & P AVEL BOS K

PAGE 9

ACTA CARSOLOGICA 36/3 2007 461 Gunders h eim. In Europe, t h e genus Glirulus (extant in Japan) to w hic h pusillus was replaced by Kowalski (1963) was continuously represented from MN3 wit h several species t hroug h out Miocene ( cf. Kowalski, 1997, 2001), w hile its diversity got restricted to a sole species, pusillus in MN15 to MN17 (Daams & de Bruijn, 1995). In Europe, t his species probably locally survived up to t h e earliest stage of Q1 zone of t h e Early Pleistocene (Valerots, abia sites). In Italy Glirulus pusillus is recorded from MN15 Arondelli-Triversa and Glirulus sp. from Q1 Monte Peglia (Kotsakis, 2003). In t h e MN15,16 and early MN17 com munities of central and sout h ern Europe Glirulus pusil lus appears as nearly a constant element ( cf. its records in sites of Gunders h eim, Ivanovce, Frec h en, Sc h ernfeld, Osztramos 7, Kielniki 1, a.o.). Last but not least we are obliged to note t h at despite good correspondence between t h e fossil item and Glirulus pusillus in t h e essential c h aracters of t h e genus, t h e fossil items dier markedly in proportion and s h ape of t h e toot h and t h e dierences seem to fall beyond t h e variation range of pusillus Alternatively, t h e toot h can be looked upon as P4/, eventu ally (t h en t h e dierences from ot h er species of t h e genus would be even larger). Unfortunately, wit h just a single fragmentary item, t h ere is a little c h ance to compre h end t h e ac tual meaning of t h ese dierences properly and it would be muc h be yond scope of t his report to discuss t h at topic in details. Apodemus (Sylvaemus) cf. atavus Heller, 1936 (Pl. 2/5-13, Pl. 4/1) e teet h remains of wood mouse are t h e best preserved fossils in bot h studies sites. ey belong to medium-sized forms of t h e subgenus Sylvaemus and could be co-identi ed even wit h t h e most widespread extant species Apodemus sylvaticus, eventually. Nevert h eless, t h e items from rnotie II exceed t h e variation range of t h e recent species at least in two respects: in appearance of dis tal cingulum on M1/ and M2/ and posterior accessory cusp ( sensu van Dam, 1997) on M/2. Bot h t h e c h ar acters, considered plesiomorp hic states in murids, distinguis h a fos sil species Apodemus atavus Heller, 1936, described from MN16 Gunders h eim site, and con sidered to be an ancestor of t h e recent European repre sentatives of t h e subgenus Sylvaemus (Rietsc h el & Storc h, 1974; Kowalski, 2001). In t h e state of bot h t h e c h aracters t h e specimens under study seem to exhibit an intermedi ate position between t h e typical MN15 atavus (including its topotypes cf. Storc h & Fejfar, 1989; Fejfar & Storc h, 1990) and t h e recent taxon. Of course, a real taxonomic meaning of similar dierences could be estimated only aer detailed comparative studies on variation dynam ics of t h e respective c h aracters for w hic h t h e available material is apparently too scarce. Moreover, in regard to t h e recent situation of t h e genus in t h e SE Europe and EP late 3: Gliridae and Arvicolidae from the rnotie II site: 1 Glirulus sp., right M3/ or P4/; 2 Arvicolidae g.sp. Cseria gracilis/carnutina grade, ?le M1/; 3 Arvicolidae g.sp. Cseria gracilis carnutina grade. FOSSIL VERTEBRATES AND PALEOMAGNETISM UPDATE OF ONE OF THE EARLIER STAGES OF CAVE EVOLUTION ...

PAGE 10

ACTA CARSOLOGICA 36/3 2007 462 Mediterranean (extensive genotypic diversication and considerable overlap among all t h e species in morp h o metric c h aracters), t h e attempts to resolve t h e taxonomic status of fragmentary fossil samples seem to be only little promising. In any case, in metrical c h aracters t h e teet h under study are at t h e lower variation range of A. sylvati cus but seem to fall well in t h e range t h e respective spe cies: Raika peina M/1 (Pl4/1): 1.67x1.5 mm, rnotie M1/ (Pl 2/5): 1.75x1.12 mm, M1/ (Pl2/6) L 1.73 mm, M2/ (Pl2/7): 1.13x1.06 mm, M3/ (Pl2/8): 0.85x0.86 mm, M/1 (Pl2/9): L 1.75 mm, (pl2/10): W 1.00 mm, M/2 (Pl 2/11): 1.20x 1.02 mm, M/3 (Pl 2/12): 0.97x0.86. Rhagapodemus cf. frequens Kretzoi, 1959 (Pl. 2 /13) A single M/2 nearly unaected by abrasion is c h ar acterized by very large dimensions (1.72x1.35 mm), ex treme degree of h ypsodonty, nearly completely separated h ypoconid and entoconid, well developed lateral antero conid and postcingulid w hile labial cingular cusps are quite minute and arranged in a s h allow cingular crestlike ridge at t h e base of t h e crown. In all t h ese c h aracters t h e toot h exceeds variation ranges in all European extant species of Apodemus (including epimelas and mystacinus i.e. t h e larges extant W-Palearctic forms of t h e genus) but corresponds well to t h e diagnostic c h aracters of Pliocene genus Rhagapodemus Kretzoi, 1959. At t h e same time, t h e state of t h e respective c h ar acters in t h e specimen seems to exhibit a more derived conditions t h an Turolian Rhagapodemus primaevus (Hu geney et Mein, 1965) or early Ruscinian R. hautimag nensis Mein et Mic h aux, 1970 (comp. also Popov, 2004), t h e ot h er species of t h at age suc h as R. ballesioi Mein et Mic h aux, 1970 or R. debruijni (Kotlia, 1992) dier even more (t h ey are distinctly smaller, lack t h e posteriod cin gular cusps etc. comp. Fejfar & Storc h 1990; Kotlia et al., 1998; Popov, 2004). Per h aps t h e best agreement is t h us wit h t h e nominotypical species Rhagapodemus fre quens Kretzoi, 1959 described from t h e MN15 site of Csarnta and later found in more sites of a similar age (MN 15 We, Gunders h eim 4) and wit h even a more derived form, distinguis h ed at t h e subspecic level as R. frequens athensis De Bruijn et van der Meulen, 1975 re ported also from t h e early Pleistocene of Greece (locality Tourkoubounia 1: De Bruijn & van der Meulen, 1975). e actual stratigrap hic position of Tourkoubounia 1 in terms of MN zones is MN16 (Doukas, 2005). e genus Rhagapodemus is furt h er reported from t h e early and middle Pleistocene of Sardinia. Of course, particu larly under island situations, a parallel evolution of t h e Rhagapodemus p h enotypes not related to t h e Miocene and Pliocene mainland forms could also come in ac count, similarly as it was suggested by Kotlia et al. (1998) for Indian R. debruijni (Kotlia, 1992). Arvicolidae (see Pl. 3/2-3, Pl. 4/ 2-4 for t h e best pr served items) In general, Arvicolidae is t h e group of t h e h ig h est im portance for t h e biostratigrap h ic analyses of t h e late Ceno zoic terrestrial deposits for well marked gradual anagen etic trends, parallel in dierent clades, extensive cladogetic diversity and predominant representation in t h e mamma lian communities of t h at period (Fejfar et al., 1997). Cor respondingly, also in t h e present samples t h e remains of arvicolids compose an essential part of t h e total material. Unfortunately, not one complete toot h is available and a vast majority of t h e material consists of isolated, broken and corroded enamel fragments, as a rule representing just single molar synclinales. e most complete specimens are gured on Plates 3 and 4. In bot h localities, any rootless form absents, and no one piece s h ows a trace of dental ce mentum in t h e molar synclinales. e molar fragments from t h e Raika peina ex hibit a considerably hig h degree of h ypsodonty, some of t h em are wit h deep and narrow synclinales typical for Lagurini. e enamel microstructure revealed by t h e SEM analyses suggest t h at at least two dierent forms are present in t h e Raika peina: one exhibiting t h e pattern typical for advanced lagurine voles (dense radial enamel alternated wit h a t hick layer of lamellar enamel w hic h prisms are distinctly h eavier t h an t h e prisms of radial enamel, Pl. 4/3), t h e ot h er s h ows t h e pattern wit h less dis tinct dierences between radial and lamellar enamel (Pl. 4/2) corresponding to t h e situation in t h e genera M yodes or advanced forms of M imomys (Koenigswald, 1980; Fe jfar & Heinric h, 1982). In contrast, all fragments available from rnotie II site s h ow rat h er plesiomorp hic constitution particu larly in clearly lower degree of molar h ypsodonty. eir Sc hmelzmuster is quite dierent from t h at in t h e Raika peina, rst in t h at it bears t h e same enamel pattern bot h in mesial and distal walls of a toot h (Pl. 3/2, 3), second t h at t h e lamellar enamel is not developed and only radial enamel is present wit h faint anastomozing near t h e EDJ instead of prism decusation. Suc h constitution is appar ently quite plesiomorp hic and it was recorded in ancient member of M imomys/Cseria lineages suc h as Cseria car nutina Rabeder 1981 or M imomys ( Cseria s.str) gracilis i.e. t h e taxa typical of t h e MN15 stage of arvicolid ra diation (Rabeder, 1981; Fejfar & Heinric h, 1982). In any case, suc h enamel arrangement is undoubtedly connect ed wit h rat h er arc h aic grade of arvicolid dental evolution and probably does not occur in t h e European forms since t h e early MN17. BIOSTRATIGRAPHIC CONSIDERATIONS For a proper biostratigrap hic analysis, t h e available ma terial is unfortunately too scarce and too fragmentary. I VAN HOR EK, A NDREJ MIHEVC, N ADJA ZUPAN HAJNA, P ETR PRUNER & P AVEL BOS K

PAGE 11

ACTA CARSOLOGICA 36/3 2007 463 Nevert h eless, it provides a certain information enabling to draw h ypot h esis on possible maximum and minimum age of t h e communities. e absence of rootless molars in arvicolid material, bot h in t h e rnotie II site and t h e Raika peina, suggests t h at t h e communities do not come from t h e Quaternary period namely because since beginning of t h e Q1 t h e communities of small ground mammals are c h aracterized by total predominance of rootless voles (particularly M icrotus s.l. and Lagurus s.l., etc.) and its seems greatly improbable t h at no one of 48 molar fragments obtained from studied sites would ex hibit suc h a state. In Lagurini, t h e rooted forms, arranged in genus Borsodia (to w hic h we ten tatively allocate some items from t h e Raika peina) soon at t h e very beginning of Q1 disappeared in ac count of t h eir rootless descendants. At least for above-mentioned reasons, we propose MN17/Q1 boundary as t h e possible minimum age bot h for t h e rnotie II site and Raika peina. In bot h sites, t h e faunal assemblages are of t h e Plio cene age (note t h e predominance of arvicolids and absence of any el ement suggesting strictly t h e Mio cene age as well as t h e data on strati grap hic distribution of t h e recorded taxa Tab. 1). e possible maxi mum age for t h e Raika peina can be tentatively placed to MN16 or early MN17 particularly due to t h e advanced degree of h ypsodonty in arvicolid molars, t h e correspon dence of some items to Borsodia (t h e genus representing an index fossil of MN17) or to relatively advanced M i momys spp. (a grade corresponding to e.g., jota/pitymyoides sensu Rabe der, 1981). However for rnotie II site, t h e grade of arvicolid molar development would suggest a rat h er earlier age. Correspondingly, also all ot h er forms identied t h ere ( De insdora, Beremendia, Glirulus a. pusillus, Rhagapodemus frequens and Apodemus cf. atavus ) are typical more for MN15 or MN 16 t h an for MN17. e possible maximum age corresponding to t h e stratigrap hic span of t h ese taxa would t h en be even around MN14/MN15 bound ary. Considering t h e fragmentary faunal list from t h e rnotie II site, as a relevant commu nity sample, we could nd close resemblance to t h e fauna composition of t h e Ruscinian or Csarntan assemblages suc h as in Gunders h eim or Csarnta 2 sites. Summarizing t h e above-mentioned arguments, an intuitive consensual biostratigraphic allocation of t h e as semblages under study would be: (i) middle to late MN17 (ca 1.8.4 Ma) for t h e Raika peina and (ii) MN15MN16 (ca. 3.0.1 Ma) for t h e rnotie II site. e dating of t h e latter site is particullary important because t h e source deposits are directly realted to termi nation of active freatic regime. P late 4: M ammalian fossils from the Raika peina: 1 Apodemus (Sylvaemus) cf. atavus H eller, 1936, le M/1; 2 Arvicolidae g.sp. indet., fragment of a lingual? wall of M/1, cf. Borsodia spp.; 3 Arvicolidae g.sp. indet., fragment of a palatal wall of an upper molar (M1/ or M2/), cf. M imomys (Cseria) sp.; 4 Arvicolidae, g.sp. indet., lingual wall of the right M3/, cf. Borsodia sp. FOSSIL VERTEBRATES AND PALEOMAGNETISM UPDATE OF ONE OF THE EARLIER STAGES OF CAVE EVOLUTION ...

PAGE 12

ACTA CARSOLOGICA 36/3 2007 464 Table 1: Stratigraphical distribution of some taxa which come in account in context of the forms recorded in faika peina and rnotie II (in bold). M inimum number of localities for shrews (aer Rzebik-Kowalska 1998, slightly supplemented) or presence/absence (+/-) for rodents (mostly aer Kowalski 2001). For denitions and chronological setting of MN zones see M ein 1990, for Q zones see H orek & Loek (1988). P l i o c e n e Q u a t e r n a r y sp. / biozones (MN/Q) 14 15 16 17 1 2 3-R Beremendia ssidens 3 11 32 20 44 6 Beremendia minor 1 1 3 Blarinoides mariae 5 9 9 5 Blarinoides sp. 3 2 1 Deinsdora fallax 1 Deinsdora hibbardi 1 3 9 3 1 Deinsdora insperata 1 Deinsdora janossyi 1 Deinsdora kerkho 1 Deinsdora kordosi 5 1 Deinsdoa reumeri 1 Glirulus spp. + Glirulus pusillus + + + (+) Glirulus sp. + Apodemus dominans + + (?) (?) Apodemus atavus + + Apodemus cf. sylvaticus + + + + Rhagapodemus primaevius ? Rhagapodemus hautimagnesis + + Rhagapodemus vanderweerdi + Rhagapodemus frequens + + Rhagapodemus (frequens) athensis + Myodes + + + + + Borsodia (+) + (+) Lagurus s.l. + + + Cseria gracilis/minor/csarnutina grade + + Mimomys pitymyoides/pusillus grade (+) + (+) Microtus s.l. + + + Alt h oug h t h e fossil record obtained from cave sediments of t h e rnotie II site and Raika peina is rat h er poor, bot h in its taxonomic structure and preservation it pro vided a valuable biostratigrap hic information. e data enabled to specify t h e supposed time frame for dating of respective cave deposits dated by t h e previous inter pretation of paleomagnetic (magnetostratigrap hic) data. Paleontological data proved suggested MN15-16 mam malian biozone as t h e most probable date (about 3.0.1 Ma). Aguilar et al. (2002) placed t h e MN15/16 boundary in t h e period between c hron C2A2.1n (based on MN16 Seynes) and C2r.2r (based on MN15 Sete), i.e. at ca 2.5 3.0 Ma, alt h oug h Lindsay et al. (1997) or Lindsay (2001) proposed MN15/16 boundary position rat h er at t h e base of Gauss C hron. e arrangements of obtained magnetozones in t h e rnotie II site were originally interpreted as older t h an 1.770 Ma, most probably belonging to t h e Gauss C hron (2.581.580 Ma) or t h e normal subc hrons wit hin t h e Gilbert C hron (4.180.230 Ma). e long normal pa leomagnetic polarity zone in t h e lower segment of t h e ll t h erefore corresponds to basal normal polarized subc hron C2An.3n (3.330.580 Ma) wit hin t h e Gauss C hron and t h e normal polarized upper segment can be CONCLUSIONS

PAGE 13

ACTA CARSOLOGICA 36/3 2007 465 ACKNOWLEDGEMENT e study was supported by t h e Grant Agency of t h e Acad emy of Sciences of t h e Czec h Republic Nos. GAAVR IAA300130701 Paleomagnetic researc h of karst sedi ments: paleotectonic and geomorp h ologic implications (2007), and GAAVR IAA3013201 Magnetomin eralogic and magnetostratigrap hic researc h of cave and uvial sediments in t h e Central European region (2002 2005), and by t h e Institutional Researc h Plan of t h e GLI AS CR, v.v.i. No. CEZ AV0Z30130516 (to PP, PB). Par tial support was also obtained from grants of t h e Grant Agency of t h e Czec h Republic No. GACR 206/05/2334, and Ministry of Education, Sport and Y out h s of t h e Czec h Republic No. 0021620828 (to IH). e researc h was performed wit hin t h e frame of researc h programs nanced by Ministry of Science of Slovenia and Slovenian Researc h Agency No.: P6 and P0 Karst Researc h, and projects Nos.: J6 Origin and development of karst caves, and J6 Development and function of caves in dierent speleo logical settings (to AM, NZH). e researc h represents also t h e results of t h e Czec h Slovenian bilateral sci ence and tec hnology cooperation programs Nos.: ME 251(1998) Researc h of karst sediments on t h e example of t h e Classical Karst, Slovenia (1998), No. 2001/009 Evolution of karst and caves based on study of cave lls, Slovenia (2001), 28 Reconstruction of speleogenesis and karstogenesis from t h e study of cave ll, Slovenia (2003), 13 Paleomagnetic studies of sediments in karst areas of Slovenia: implica tion for paleotectonic reconstructions (2005), and 9 Paleomagnetism of sediments in karst areas of Slovenia (2007-2008) covered by Ministry of Science of Slovenia and Ministry of Education, Sport and Y out h s of t h e Czec h Republic (to AM, NZH, PP, PB). We acknowledge t h e eld assistance of Mr. Fran jo Drole, Mr. Jure Hajna, Dr. Jaroslav Kadlec, Mr. Petr Sc hnabl and Mr. Stanislav lec h ta, and t h e h elp of Mrs. Jana Rajlic h ov (Institute of Geology AS CR, v.v.i., Pra h a) and Mr. Jure Hajna (Karst Researc h Institute, ZRC SAZU, Postojna), w h o drew some of t h e gures. We are particularly obliged to Jean-Pierre Aguilar, Jacques Mi c h aux, Bernard Sig and Dragica Turnek w h o carefully reviewed t h e manuscript and t h eir suggestions improved it in an essential way. compared to some of hig h er normal subc hrons of t h e Gauss C hron (C2An.1n subc hron = 2.581.040 Ma or C2An.2n subc hron = 3.110.220 Ma). e combina tion of paleontological and paleomagnetic data indicates, t h at t h e fauna cannot be older t h an about 3.6 Ma, due to reverse polarized magnetozone at top of Gilbert C hron terminating at 4.180 Ma. is level represents approxi mately also t h e base of t h e MN15 mammalian biozone. In t h e Raika peina, t h e boundary of normal and reverse polarized magnetozone wit hin t h e layer wit h fau na (Fig. 3) is identied wit h t h e bottom of C2n Olduvai subc hron (1.770.950 Ma). e geometry of obtained magnetozones is deformed as compared wit h subc hrons on t h e GPTS due to numerous principal breaks in de position in t h e lower part of t h e prole. Break can last more t h an 250 ka. erefore, we correlate t his part wit h t h e lower part of t h e Matuyama C hron (2.150.581 Ma) and individual subc hrons of t h e Gauss C hron (2.581.58 Ma). e prole above Olduvai subc hron records s h ort part of Matuyama C hron (some of reverse polarized subc hrons C1r.3r, C1r.2r, or C1r.1r wit hin t h e time span of 1.770.780 Ma) and Brun h es C hron (C1n; younger t h an 0.780 Ma). For t h e rst time, t h e combination of vertebrate fossil records and magnetostratigrap h y proved expected antiquity of t h e cave fossilization in t h e region of t h e Classical Karst. A good agreement of biostratigrap hic and magnetostratigrap hic inferences suggests autoc h to nous synsedimentary origin of t h e faunal remains and sediments and supports strongly expected relevance of t h e dating eort and its applicability in karstogenetic reconstructions. Furt h er step in t h at direction will be publis h ed in anot h er place. Wort h mentioning is t h at t h e important paleotectonic movements recently interpreted in Dinarides (Ili & Neubauer, 2005) and Sout h ern Alps (Neubauer, 2007), w hic h could be related to t h e upli in t h e Classical Karst and rearrangements of its h ydrologi cal systems resulting in increased fossilisation rate, cor respond in age to MN15 zone. e fossilization during MN15MN17 nis h ed one of important older p h ases of speleogenesis in t h e region.

PAGE 14

ACTA CARSOLOGICA 36/3 2007 466 REFERENCES Absolon, K. & S. Hrab, 1930: ber einen neuen Sss wasser-Polyc h aten aus den H h lengewssern der Herzegowina.Zoologis h er Anzeiger, 88, 9-10, 259264, Jena. Aguilar, J.-P., Croc h et, J.-Y ., Hebrard, O., Le Strat, P., Mi c h aux, J., Pedra, S. & B.Sig, 2002: Les micromam miferes de Mas Rembault 2, gisement karstique du Pliocne suprieur du sud de la France, ge, palo climat, godynamique.Gologie de la France, 4, 17-37. Aguilar, J.-P., Croc h et, J.-Y ., Krivic K., Marandat, B., Mic h aux, J., Mi h evc, A., Sig, B. & S. ebela, 1998: Pleistocene small mammals from karstic llings of Slovenia Preliminary results.Acta Carsologica, 27/2, 141-150, Ljubljana. Aguirre, E. & G. Pasini, 1985: e Pliocene-Pleistocene boundary.Episodes, 8, 116-120, Ottawa. Bosk, P., 2002: Karst processes from t h e beginning to t h e end: h ow can t h ey be dated?In: Gabrovek, F. (Ed.): Evolution of Karst: From Prekarst to Cessa tion, Carsologica, Zaloba ZRC, Postojna-Ljubljana, 191-223. Bosk, P., Pruner, P. & N. Zupan Hajna, 1998: Paleomag netic researc h of cave sediments in SW Slovenia.Acta Carsologica, 27/2, 3, 151-179, Ljubljana. Bosk, P., Mi h evc, A., Pruner, P., Melka, K., Ven h odov, D. & A. Langrov, 1999: Cave ll in t h e rnotie Quarry, SW Slovenia: Palaeomagnetic, mineralogi cal and geoc h emical study.Acta Carsologica, 28/2, 2, 15-39, Ljubljana. Bosk, P., Pruner, P., Mi h evc, A. & N. Zupan Hajna, 2000a: Magnetostratigrap h y and unconformities in cave sediments: case study from t h e Classical Karst, SW Slovenia.Geologos, 5, 13-30, Pozna. Bosk, P., Knez, M., Otrubov, D., Pruner, P., Slabe, T. & D. Ven h odov, 2000b: Palaeomagnetic Researc h of Fossil Cave in t h e Hig h way Construction at Kozina, SW Slovenia.Acta Carsologica, 29/2, 1, 15-33, Ljubljana. Bosk, P., Pruner, P. & J. Kadlec, 2003: Magnetostratig rap h y of cave sediments: application and limits.Studia Geop h ysica et Geodaetica, 47, 2, 301-330, Pra h a. Bosk, P., Mi h evc, A. & P. Pruner P., 2004a: Geomorp h o logical evolution of t h e Podgora Karst, SW Slowe nia: Contribution of magnetostratigrap hic researc h of t h e rnotie II site wit h Marifugia sp.Acta Car sologica, 33, 1, 12, 175-204. Ljubljana. Bosk, P., Pruner, P., Mi h evc, A., Zupan Hajna, N., Horek, J., Kadlec, J., Man, O. & P. Sc hnabl, 2004b: Raika peina.12 t h International Karstological Sc h ool, Classical Karst Dating of Cave Sediments, Postojna. Guide booklet for t h e excursions and ab stracts of presentations, 23-27, Postojna. Brodar, S., 1952: Prispevek k stratigraji kraki h jam Pivke kotline, posebej Parske golobine.Geografski vestnik, 24, 43-76, Ljubljana. Brodar, S., 1958: rni kal, nova paleolitska postaja v Slov enskem Primorju.Razprave 4.razreda, 4, 271-363, Ljubljana. Brodar, S., 1966: Pleistocenski sedimenti in paleolitska najdia v Postojnski jami.Acta carsologica, 4, 57138, Ljubljana. Brodar, S., 1970: Paleolitske najdbe v jami Risovec pri Postojni.Acta carsologica, 5, 273-300, Ljubljana. Bruijn, H. De & A.J. van der Meulen, 1975: e early Pleistocene Rodents from Tourkoubounia-1 (At h ens, Greece). I.Koninklijke Nederlandse Akademie van Wetensc h appen, Proceedings B, 78: 314-338. Cande, S.C. & D.V. Kent, 1995: Revised calibration of t h e geomagnetic polarity timescale for t h e Late Creta ceous and Cenozoic. Journal of Geop h ysical Re searc h, 100, B4, 6093-6095, Was hington. Colman, S.M. & K.L. Pierce, 2000: Classication of Qua ternary geoc hronologic met h ods.In: Noller, J.S., J.M. Sowers & W.R. Lettis (Eds.): Quaternary Geo c hronology. Met h ods and Applications, American Geop h ysical Union, 2-5, Was hington. Daams, R. & H. de Bruijn., 1995: A classication of t h e Gliridae (Rodentia) on t h e basis of dental morp h ol ogy.Hystrix, 6, 3-50, Rome. Dam, J.A. van, 1997: e small mammals from t h e Up per Miocene of t h e Teruel-Alfambra region (Spain): palaeobiology and paleoclimatic reconstructions.Geologica Ultraiectina, 156, 1-204. Utrec h t. Doukas, C.S., 2005: Greece.In: Hoek Ostende, L.W. van den, Doukas, C.S. & J.W.F. Reumer, (Eds.): e Fos sil Record of t h e Eurasian Neogene Insectivores (Er inacomorp h a, Soricomorp h a, Mammalia). Part I. Scripta Geologica, Special Issue, 5, 99-112, Leiden. Fejfar, O. & W.-D. Heinric h, 1982: Zur Evolution von M imomys (Rodentia, Mammalia) in Csarntanum und Villafranc hium Europas.Eclogae Geologicae Helvetiae, 72, 779-793, Basel. Fejfar, O., Heinric h, W.-D., Pevzner, M.A. & E.A. Van gen h eim, 1997: Late Cenozoic sequence of mam malian sites in Eurasia: an updated correlation.Pa laeogeograp h y, Palaeoclimatology, Palaeoecology, 133, 259-288, Amsterdam. I VAN HOR EK, A NDREJ MIHEVC, N ADJA ZUPAN HAJNA, P ETR PRUNER & P AVEL BOS K

PAGE 15

ACTA CARSOLOGICA 36/3 2007 467 FOSSIL VERTEBRATES AND PALEOMAGNETISM UPDATE OF ONE OF THE EARLIER STAGES OF CAVE EVOLUTION ... Fejfar, O. & G. Storc h, 1990: Eine pliozne (ober-rusci nisc h e) Kleinsugerfauna aus Gunders h eim, R h ein h essen 1. Negetiere: Mammalia, Rodentia.Sencke bergiana let h aea, 71, 139-184, Frankfurt a. M. Fis h er R., 1953: Dispersion on a sp h ere.Proceedings of t h e Royal Society, A 217, 295-305, London. Ford, D. & R. Gospodari, 1989: U series dating studies of Ursus spelaeus deposits in Krina jama, Slove nia.Acta carsologica, 18, 39-51, Ljubljana. Franke, H. & M. Gey h, 1971: 14C datierungen von Kalk sinter aus slowenisc h en H h len.Der Aufsc h luss, 22, 7-8, 235-237. Gliozzi, E., Abbazzi, L., Argenti, P., Azzaroli, A., Caloi, L., Barbato, L.C., Di Stefano, G., Esu, D., Ficcarelli, G., Girotti, O., Kotsakis, T., Masini, F., Mazza, P., Mezzabotta, C., Palombo, M.R., Petronio, C., Rook, L., Sala, B., Sardella, R., Zanalda, E. & D. Torre, 1997: Bioc hronology of selected mammals, mol luscs and ostracods from t h e Middle Pliocene to t h e Late Pleistocene of Italy. e state of t h e art.Rivista Italiana di Paleontologia e Stratigraa, 10, 3, 369387, Milano.Gospodari, R., 1972: Prvi podatki o absolutni starosti sige v Postojnski jami na podlagi 14C.Nae jame, 13, 91-98, Ljubljana. Gospodari, R., 1974: Fluvialni sedimenti v Krini jami.Acta carsologica, 6, 327-366, Ljubljana. Gospodari, R., 1976: Razvoj jam med Pivko kotlino in Planinskim poljem v kvartarju.Acta carsologica, 7, 5-139, Ljubljana. Gospodari, R. 1977: Generacije sige v nekateri h kraki h jama h.Geograja krasa, Poroila 1977, 1-31, IZRK SAZU, Postojna. Gospodari, R. 1981: Generacije sig v klasinem krasu Slovenije.Acta carsologica, 18, 39-51, Ljubljana. Gospodari, R. 1984: Jamski sedimenti in speleogeneza kocjanski h jam.Acta carsologica, 12 (1983), 2748, Ljubljana. Gospodari, R. 1985: O speleogenezi Divake jame in Tr h lovce.Acta carsologica, 8 (1984), 5-36, Lju bljana. Gospodari, R. 1988: Paleoclimatic record of cave sedi ments from Postojna Karst.Annales de laq Socit gologique de Belgique, 111, 91-95, Lige. Heller, F., 1936: Eine oberpliocne Wirbeltierfauna aus R h ein h essen.Neues Ja hrbuc h der Mineralogie, Geologie und Palontologie, Beilage-Band 63, Abt. B, 76, 99-160, Stuttgart. Hoek Ostende, L.W. van den, Doukas, C.S. & J.W.F. Re umer, Eds., 2005: e Fossil Record of t h e Eurasian Neogene Insectivores (Erinacomorp h a, Soricomor p h a, Mammalia). Part I.Scripta Geologica, Special Issue, 5. Leiden. Horek, I. & V. Loek, 1988: Palaeozoology and t h e mid-European Quaternary past: scope of t h e ap proac h and selected results.Rozpravy SAV, ada matematickc h a prodnc h vd, 94, 1-106, Pra h a. Ikeya, M., Miki, T. & R. Gospodari, 1983: ESR Dat ing of Postojna Cave Stalactite.Acta carsologica, 11(1982), 117-130, Ljubljana. Ili A. & F. Neubauer, 2005: Tertiary to recent oblique convergence and wrenc hing of t h e Central Dina rides: Constraints from a palaeostress study.Tecto nop h ysics, 410, 465-484, Amsterdam. Kirsc h vink, J.L., 1980: e least-squares line and plane and t h e analysis of palaeomagnetic data.Geop h ysi cal Journal of t h e Royal Astronomical Soc., 62, 699718, Oxford. Koenigswald, W. von, 1980: Sc hmelzstruktur und Mor p h ologie in den Molaren der Arvicolidae (Roden tia).Ab h andlungen der senckenbergisc h es natur forsc h ung Gesselsc h a, 536, 1-129, Frankfurt a. M. Kotlia, B.S., Storc h G. & T. Da h lmann, 1998: R h agapode mus debruijni from t h e Pliocene of Kas h mir Basin: First record of R h agapodemus from India (Mammalia: Rodentia).Senckenbergiana let h aea, 78: 213-216. Kotsakis, T., 2003: Fossil glirids of Italy: t h e state of t h e art.Coloquios de Palentologia, Vol. Ext. 1, 335343, Barcelona. Kowalski, K., 1963: e Pliocene and Pleistocene Gliridae (Mammalia, Rodenia) from Poland.Acta zoologica Cracoviensis, 8, 533-567, Krakw. Kowalski, K., 1997: Gliridae (Mammalia: Rodentia) from t h e Miocene of Belc h atw in Poland.Acta zoolo logica Cracoviensis, 40, 173-198, Krakw. Kowalski, K., 2001: Pleistocene rodents of Europe.Folia Quaternaria, 72, 3-389, Krakw. Kranjc, A., 1997: Introduction. In: A. Kranjc et al (Eds): Kras. Slovene Classical Karst.Znanstvenorazis kovni center SAZU, Intitut za raziskovanje krasa, p. 11-17, Ljubljana. Krytufek, B. & V. Vo hralk, 2005: Mammals of Turky and Cyprus. Rodentia I: Sciuridae, Dipodidae, Gliridae, Arvicolinae.Zaloba Annales, 292 pp, Koper. Lindsay, E., 2001: Async hrony in mammalian bioc hro nology.Lynx, N.S. 32, 201-214, Pra h a. Lindsay, E., Opdyke, N.D. & O. Fejfar, 1997: Correlation of selected ate Cenozoic European mammal fauna wit h t h e magnetic polarity time scale.Palaeogeog rap h y, Palaeoclimatology, Palaeoecology, 133, 205226, Amsterdam. Mein, P., 1990: Updating of MN zones.In: Lindsay, E.H., Fa h lbus h V. & P. Mein (Eds.): European Neogene mammal c h ronology, 73-90, Plenum Press, New Y ork. Mi h evc, A., 1996: Brezstropa jama pri Povirju.Nae jame, 38, 92-101, Ljubljana.

PAGE 16

ACTA CARSOLOGICA 36/3 2007 468 Mi h evc, A., 2000: Fosilne cevke iz brezstrope jame ver jetno najstareji ostanki jamskego cevkarja Marifu gia (Annelida: Polyc h aeta).Acta Carsologica, 29/2, 261-270. Ljubljana. Mi h evc, A., 2001: Speleogeneza Divakega krasa.Zbirka ZRC, 27, 1-180, Ljubljana. Mi h evc, A., 2003: Sledovi jamskega medveda v Raiki peini.Nae jame, 45, 48-55, Ljubljana. Mi h evc, A., 2004: e unroofed Cave Ulica and Ulica peina Cave.12t h International Karstological Sc h ool, Classical Karst Dating of Cave Sediments, Postojna. Guide booklet for t h e excursions and ab stracts of presentations, 27, Postojna. Mi h evc, A., 2007: e age of karst relief in West Slove nia.Acta Carsologica, 36/1, 35-44, Ljubljana. Mi h evc, A., Bosk, P., Pruner, P. & B. Vokal, 2002: Fosilni ostanki jamske ivali Marifugica cavatica v brezstro pi jami v kamnolomu rnotie v za h odni Slovenii.Geologija, 45/2, 471-474, Ljubljana. Mi h evc, A. & Lauritzen, S.E., 1997: Absolute datations of speleot h ems and its speleomorp h ological signi cance from Divaka jama and Jazbina caves; Kras pla teau, Slovenia.Proceedings of t h e 12t h International Congress of Speleology, 1997, Switzerland, 1 Sym posium 7 P h ysical Speleology & Symposium 8 Karst Geomorp h ology, 57-59, Speleo Projects, Basel. Mi h evc, A., Sket, B., Pruner, P. & P. Bosk, 2001: Fossil remains of a cave tube worm (Polyc h aeta: Serpuli dae) in an ancient cave in Slovenia.Proc., 13t h In ternational Speleological Congress, 4t h Speleologi cal Congress of Latin America and t h e Carribean, 26t h Brazilian Congress of Speleology, Brasilia, July 15-22, 2001, 2: 20-24, Brasilia. Neubauer F., 2007: Neogene to Recent Motion of Adria, formation of t h e Friuli orocline, and deformation of Eastern Alps and nort h eastern Dinarides.Geo p h ysical Researc h Abstracts, 9, 06232 (SRef-ID: 1607-7962/gra/EGU2007-A-06323). Ogg, J., 2007: Request for IUGS Ratication to Establis h t h e Quaternary as a System/Period of t h e Cenozoic and revise t h e associated base of t h e Pleistocene Se ries.Unpubl. letter, International Commission on Stratigrap h y, IUCN, 21 pp., Oslo. Otoniar, B., 2007: Upper Cretaceous to Paleogene for bulge unconformity associated wit h foreland basin evolution (Kras, Matarsko podolje and Istria; SW Slovenia and NW Croatia).Acta Carsologica, 36/1, 101-120. Postojna. Placer, L., 1981: Geoloka zgradba jugoza h odne Sloveni je.Geologija, 24/1, 27-60, Ljubljana. Popov, V., 2004: Pliocene small mammals (Mammalia, Lipotyp h la, C hiroptera, Lagomorp h a, Rodentia) from Muselievo (Nort h Bulgaria).Geodiversitas, 26, 403-491. Pruner, P. & P. Bosk, 2001: Palaeomagnetic and mag netostratigrap hic researc h of cave sediments: t h eo retical approac h, and examples from Slovenia and Slovakia. Proc., 13t h International Speleologi cal Congress, 4t h Speleological Congress of Latin America and t h e Carribean, 26t h Brazilian Con gress of Speleology, Brasilia, July 15-22, 2001, 1, 9497, Brasilia. Rabeder, G., 1981: Die Arvicoliden (Rodentia, Mamma lia) aus dem Pliozn und dem ltesten Pleistozn von Niedersterreic h.Beitrge zur Palontologie sterreic h s, 10, 1-373, Wien. Rakovec, I., 1958: Pleistocenski sesalci iz jame pri rnem kalu.Razprave, IV, 365-434, Ljubljana. Reumer, J.W.F., 1984: Ruscinian and early Pleistocene Soricidae (Insectivora, Mammalia) from Tegelen (Net h erlands) and Hungary.Scripta Geologica, 73, 1-173, Leiden. Rietsc h el S., Storc h G., 1974: Aussergew hnlic h er h altene Waldmuse ( Apodemus atavus Heller 1936) aus dem Ober-Pliozn von Willers h ausen am Harz.Sencke bergiana let h aea, 54, 491-519, Frankfurt a. M. Rzebik-Kowalska. B., 1990: Pliocene and Pleistocene In sectivora (Mammalia) of Poland. VI. Soricidae: De insdora Heller, 1936 and Z elceina Sulimski, 1962. Acta Zoologica Cracoviensia 33: 45-77. Rzebik-Kowalska, B., 1998: Fossil History of S hrews in Europe. In: Wojcik, J.M. & M. Wolsan (Eds.): Evolu tion of S hrews. Mammal Res. Inst. PAN, pp. 23-92, Biaowiea. Sasowsky, I.D., ebela, S. & W. Harbert, 2003: Concur rent tectonism and aquifer evolution >100,000 years recorded in cave sediments, Dinaric karst, Slovenia.Environmental Geology, 44, 1, 8-13, New Y ork. ebela, S. & I.D. Sasowsky, 2000: Paleomagnetic dating of sediments in caves opened during hig h way con struction near Kozina, Slovenia.Acta Carsologica, 29/2, 23, 303-312, Ljubljana. Storc h, G. & O. Fejfar, 1989: Gunders h eim-Findling, a Ruscinian rodent fauna of Asian arnities from Germany. In: Lindsay E.H. et al (Eds.): European Neogene Mammal C hronology, Plenum Press, pp. 405-412, New Y ork. Zupan, N., 1991: Flowstone datations in Slovenia.Acta carsologica, 20, 187-204, Ljubljana. Zupan Hajna, N., Mi h evc, A., Pruner, P. & P. Bosk, 2005: Nova spoznanja o starosti jamski h sedimentov v Slo veniji.Geoloki zbornik, 18, 132-135, Ljubljana. Zupan Hajna, N., Mi h evc, A., Pruner, P. & P. Bosk, 2007: Time recorded in cave deposits 10 years of paleo magnetic researc h in Slovenian caves.Acta Carso logica, 36/1, 242, Ljubljana. I VAN HOR EK, A NDREJ MIHEVC, N ADJA ZUPAN HAJNA, P ETR PRUNER & P AVEL BOS K



PAGE 1

E ARL Y LATE P LIOCENE PALEOKARSTIC FILLINGS PREDATING THE MAJOR P LIO P LEISTOCENE EROSION OF THE QUERC Y TABLE SWF RANCE P OZNOPLIOCENSKE PALEOKRAKE ZAPOLNITVE IZ ASA PRED GLAVNO PLIO PLEISTOCENSKO EROZIJO PLANOTE QUERC Y, JUGOZAHODNA F RANCIJA JeanP IERRE A GUILAR 1 Jacques M ICHAU X 2 ierry P LISSI 3 Bernard S IG 4 Izvleek UDK 903.5:569.32(118.2)(44) Jean-Pierre Aguilar, Jacques Michaux, ierry Plissi & Ber nard Sig: Poznopliocenske paleokrake zapolnitve iz asa pred glavno plio-pleistocensko erozijo planote Quercy, jugo zahodna Francija V paleokraki h zapolnitva h Quercyja smo nali ostanke zob glodalcev iz zgodnjega dela poznega pliocena. Ostanki so dokaz o sedimentnem obdobju povezanem z visoko gladino morja pred priblino 3,5 milijoni let. Najdie odraa stanje pred regionalno erozijsko fazo in vzpostavitvijo hidrografskega sistema. Kljune besede: Glodalci, plio-pleistocenska erozija, Quercy, JZ Francija. 1 Institut des Sciences de lEvolution (UMR 5554-CNRS) Universit Montpellier 2, CC 064, Place E-Bataillon, F34095 Montpellier, cedex 5. (aguilar@isem.univ-montp2.fr) 2 Institut des Sciences de lEvolution & EPHE (UMR 5554-CNRS) Universit Montpellier 2, CC 064, Place E-Bataillon, F34095 Montpellier, cedex 5. (mic h aux@isem.univ-montp2.fr) 3 P h osp h atires du Quercy, 85 mas de Couderc, F-46260 Limogne en Quercy. (t hierry.pelissie@wanadoo.fr) 4 UMR 5125 PEPS CNRS, France; Universit Lyon 1, Campus de la Doua, Btiment Geode 69622 Villeurbanne Cedex, France. (bernard.sige@univ-lyon1.fr) Received/Prejeto: 24.08.2007 COBISS: 1.01 ACTA CARSOLOGICA 36/3, 469-473, POSTOJNA 2007 Abstract UDC 903.5:569.32(118.2)(44) Jean-Pierre Aguilar, Jacques Michaux, ierry Plissi & Ber nard Sig: Early late Pliocene paleokarstic llings predating the major Plio-Pleistocene erosion of the Quercy table, SW France Early late Pliocene rodent toot h remains, h ave been found in situ wit hin Quercy paleokarstic llings. ey provide evidence of a sedimentary episode related to a hig h Pliocene marine level ca 3.5 Ma ago, t h at illustrates t h e situation predating t h e region al erosion p h ase and setting up of t h e h ydrograp hic system. Key words: Rodents, Plio-Pleistocene erosion, Quercy, SWFrance. INTRODUCTION e underground paleokarstic network system in t h e Quercy area (Fig. 1) was revealed at t h e end of t h e19t h century, along wit h t h e extended p h osp h ate ore mining on t his Jurassic limestone plateau. A long karstic evolu tion began as soon as t h e emergence of t h e platform dur ing Late Cretaceous times. is evolution consists, rst in t h e elaboration of t h e underground net systems, t h en in a long polyp h ased pro cess of lling emptying t h e karstic voids, according to t h e up and down base level c h anges t h at occurred almost continuously during t h e rst h alf of t h e Tertiary era. e lling sediments are mostly vadose clay deposits, t h e various ages of w hic h being establis h ed from t h e study of t h eir fossil vertebrate contents (for latest accounts see Plissi & Sig 2006). en, since latest Oligocene times, t h e Quercy platform was covered wit h prograd ing lacustrine sediments of t h e Aquitaine Basin. Finally, t h e w h ole structure was strongly worn down by t h e socalled Plio-Pleistocene erosional p h ase: t h e previously deep underground system became closer to t h e surface, and was exposed bot h to erosion and widening, but also Plio-Pleistocene llings occurred as s h own by t h e fossils t h ey include (Croc h et et al 2006). Among t h e latter are rare Late Pliocene and Plio-Pleistocene toot h specimens.

PAGE 2

ACTA CARSOLOGICA 36/3 2007 470 Suc h remains result from sub-erosion, or more probably from contamination over t h e underlying Tertiary llings t h at occurred w h en t h e pits were quarried from top to bottom (ibid, p. 86-87). Since t h ey sustain our new data, t h ese fossils are reported h erein (Fig. 3). Fig. 1: Geological setting and localization of the discuted structures in the Quercy, SW France. DESCRIPTION In addition to natural erosion, various h uman works resulted in sections of t h e Jurassic limestone as well as sections of natural llings. In t h e central western part of t h e p h osp h oritic area, near Cremps village (Fig. 1), a road passing t hroug h a meandriform dry valley directed toward t h e river Lot h ad its rocky nort h ern side cut for enlargement (Fig. 2). is results in sections of several closely spaced karstic gullies wit h t h eir natural llings. About 1 m wide, some of t h em visible on t h e ground surface, t h e gullies appear as digits of a major c h annel (Fig. 2). e gullies s h are similar lling sediments made of brown and green is h sandy clay, including centimetric doll concretions. ese sediments fully lack t h e iron pisolit h s t h at are common in t h e Quercy Paleogene llings. e ne sandy residues obtained by was hing and sieving sediments con tain millimetric bone and toot h fragments, among w hic h enamel blades of rodent incisors are dominant, along wit h rare molar fragments. ese fossils occur in some w h at varied proportion wit hin four of t h e ve exposed J EAN P IERRE A GUILAR J ACQUES M ICHAU X, T HIERR Y P LISSI & B ERNARD S IG In spite of long lasted eldwork in Quercy and hig h number of recorded localities (over 150 for t h e w h ole pa leokarstic area), until now not a single in situ lling was reported to t h e Pliocene period. e present paper docu ments suc h a case of Pliocene llings wit hin t h e Quercy paleokarst, and addresses its bearing for t h e geodynamic and geomorp h ological evolution of t h e area. Fig. 3 M urid rodents molar teeth from the Quercy early Late P liocene and P lio-P leistocene localities : Vielase-P Stephanomys cf. donnezani : 1 M1 sin. (2,88 x 1,92) ; P ech Desse-P Stephanomys cf. donnezani : 2 M1 dex. (3,04 x 2,09), 3 M2 sin. (2,03 x 1,96) ; Cuzal P -Q, Apodemus sp. : 4 m2 sin. (1,27 x 1,17) ; Trifon P -Q, Apodemus cf. dominans : 5 M2 sin. (1,17 x 1,13), 6 m2 sin. (1,18 x 1,09). Cremps lling n, : locates the recognized fragment (cusp t9) of Stephanomys cf. donnezani All gures X 20, drawn by J .-P Aguilar and L. M eslin.

PAGE 3

ACTA CARSOLOGICA 36/3 2007 471 As mentioned above, not a complete toot h could be found from t h e sampled four c h annels, but only fragments. Be cause of t h e weak mineralization condition and t h e toot h type, suc h remains are dircult to assess as regards t h eir systematic position. Nevert h eless, among various frag ments, t h e Cremps 5 sample provided specimens, w hic h we condently recognized, some as strips of arvicolid rodent molars, and ot h ers as murid rodent cusps. e h eig h t and s h ape of one of t h em, a labial fragment (Fig. 3), allowed us to recognize it as t h e cusp t 9 of a Steph anomys molar t h at belongs to a species close to S. don nezani reported from localities as Ste or Mont-Hlne in France (Aguilar et al., 1985, Bac h elet 1990). Lets recall t h at t hree isolated S. donnezani molars (Fig. 3, n 1-3) and t hree Apodemus molars (Fig. 3, n 4-6) were previously found wit hin some Paleogene Quercy faunas (Vielase-P, Pec h-Desse-P, Cuzal-P-Q, Trifon-P-Q) as resulting from mining contamination (Croc h et et al., 2006). e association of Stephanomys donnezani wit h an arvicolid c h aracterizes t h e MN 15 zone (cf. de Bruijn et al 1992), t h e corresponding time interval is s h ort and extends from t h e end of t h e Lower Pliocene to t h e begin ning of t h e Late Pliocene; an age ca 3.5 Ma can be sug gested (Aguilar et al., 1999). On t h e Mediterranean bor der, t his period corresponds to t h e end of t h e Roussillon Basin lling, s h ortly followed by t h e progressive retreat of t h e Pliocene sea. Fig. 2: Cremps 1 to 5 paleokarstic llings views (photos P & B. S., improved J .-P A.). P AL E ONTOLOGY E ARL Y LATE P LIOCENE PALEOKARSTIC FILLINGS PREDATING THE MAJOR P LIO P LEISTOCENE EROSION OF THE ... lling sections. A long transportation wit hin h eavy sandy mud likely explains suc h a preservation state immediate ly prior to t h e destruction of t h e teet h. Suc h sediments muc h probably h ave lled again emptied conducts of t h e Paleogene paleokarst.

PAGE 4

ACTA CARSOLOGICA 36/3 2007 472 CONCLUSION Rodent toot h remains, early late Pliocene in age, h ave been found in situ wit hin paleokarst llings near Cremps village, in t h e median western p h osp h ate ore area of t h e Quercy. ey give evidence of a sedimentary episode re lated to a hig h Pliocene sea level predating t h e incision of t h e present regional river system. Previously suspected from specimens articially mixed to some Paleogene faunas, Pliocene deposits occur wit hin t h e Quercy pale okarst, as s h own by t h e fossil contents from some llings near Cremps. Contributions ISE-M 2007-065 (CNRS-UMR 5554) & UMR 125-07.033 Acknowledgments : t h e aut h ors t h ank Ms. Laurence Meslin for drawings improvement, M. Mic h el Legris for englis h text c h ecking and Prof. Ivan Horacek for h elpful lecture. W h erever occurring, caves as well, t h e deposition of sedi ments generally relates to hig h marine base levels. For t h e period considered h ere, t h e corresponding one is likely t h e early Piacenzian hig h level (Haq et al., 1987). In t h e Aquitaine area, t h e related Pliocene marine levels occur far away from t h e current coast, wit h t h e exception of t h e Redonian deposits of t h e Loire-Atlantique region (Alvin erie et al., 1992). Once admitting t h at t h e Cremps llings imply a Pliocene hig h sea level, we s h ould suppose t h at t h e deposits related to t his transgressive episode, w h at ever marine or continental, h ave been generally removed during t h e late Pliocene and/or t h e Pleistocene. e Cremps llings need to be considered in t h eir relation to t h e h ydrograp hic system of t h e river Lot. Ac cording to Astruc et al., (2000), t h e Lot corridor, as well as t h ose of t h e ot h er regional rivers (Cl and Aveyron) would h ave been settled at t h e very beginning of a late Pliocene to lower Pleistocene upli p h ase. If so, t h e early Late Pliocene Cremps llings would just precede t h e set ting down of t h e incipient Lot River into its present place, long before t h e limestone table was deeply dug and its surface lowered to its present level. e duration of t h e time interval between t h e Cremps llings (ca 3.5 Ma) and t h e river system setting is open to question. A long gap may separate t h ese llings from t h e beginning of t h e incision of t h e Lot River, h ypot h etized at ca 1 Ma as t h e oldest age (e.g. Simon-Coinon 1989). From our present data, t his incision, along wit h t h e planation of t h e Quercy plateau, could h ave occurred muc h earlier. In t h e course of suc h a strong erosion process, t h e Pliocene Cremps llings as well as t h e Paleogene cave llings, initially located very deep under t h e surface, get closer to it. In many cases t h eir roofs were partly or fully removed (e.g. t h e Mmerlin caon). Suc h evidence muc h extends t h e w h ole vertical original dept h of t h e Quercy p h osp h ate ore pits, down to 150 meters or even more for some of t h em (e.g. St-Jean-de-Laur, Roqueprune, Coulou). Fres h water fossils (s h vertebrae) h ave been re ported from rare Quercy llings, suc h as t h e Cremat one (Sig et al., 1991), ca 2 km nort h ward of t h e Lot valley (Fig. 1), in a small pit located at t h e average altitude of t h e Quercy plateau. According to its mammal remains, especially t h e rodents, t h e Cremat fauna is MN 3 zone, Burdigalian, early Miocene in age, ca 18 Ma. Its s h re mains indicate t h e connexion between t h e past river sys tem and t h e Cremat karst system. ese remains provide evidence of a paleo-Lot t h en owing erratically and de positing sediment 150 m over t h e present day level, in a way probably similar as regards t h e ot h er regional rivers. e Cremps llings (today at 214 m asl) testify to a hig h early Late Pliocene base level, still well above t h e present day one at ca 3.5 Ma. GEOMORPHOLOGICAL INFERENCE J EAN P IERRE A GUILAR J ACQUES M ICHAU X, T HIERR Y P LISSI & B ERNARD S IG

PAGE 5

ACTA CARSOLOGICA 36/3 2007 473 R EFERENCES Aguilar, J.-P., Calvet, M. & Mic h aux, J. 1985 : Description des rongeurs pliocnes du Mont-Hlne (Pyrnes Orientales), nouveau jalon entre les faunes de Per pignan (Serrat den Vacquer) et de Ste. P alaeoverte brata 16, 3, 124-144. Aguilar, J.-P., Legendre, S., Mic h aux, J. & Montuire, S. 1999 : Pliocene mammals and climatic reconstruc tion in t h e Western Mediterranean area; In Wrenn, J.H., Suc J.-P. & Leroy S.A.G. (Eds), e Pliocene, Time of C h ange. American Association of Strati graphic P alynologists Foundation, 109-120. Alvinerie, J., Antunes, M.T., Ca h uzac, B., Lauriat-Rage, A., Montenat, C. & Pujol, C. 1992 : Synt h etic data on t h e paleogeograp hic history of Nort h eastern At lantic and Betic-Rian basin, during t h e Neogene (from Brittany, France, to Morocco). P alaeogeogra phy, P alaeoclimatology, P alaeoecology, 95, 263-286. Astruc, J.G., Cubaynes, R., Durand-Delga, M., Legendre, S., Muratet, B., Pajot, B., Plissi, T., Rey, J. & Sig, B. 2000 : Notice explicative, Carte gol. France (1/50 000), feuille Ngrepelisse (931). Orlans : BRGM 1112. Bac h elet, B., 1990 : M uridae et Arvicolidae (Rodentia, M ammalia) du P liocne du Sud de la France: sys tmatique, volution, biochronologie. se Univ. Montpellier II, 1-199. Bruijn, H. de, Daams, R., Daxner-Hck, G., Fa h lbusc h, V., Ginsburg, L., Mein, P. & Morales, J., 1992: Report of t h e RCMNS working group on fossil mammals, Reisensburg 1990. Newsletters on Stratigraphy, 26 (2/3), 65-118. Croc h et, J.-Y ., Aguilar, J.-P., Astruc, J. G., Boulbs, N., Escarguel, G., Mic h aux, J., Montuire, S., Plissi, T., Simon-Coinon, R. & Sig, B., 2006 : Reprises plioplistocnes du palokarst quercinois. Strata 1, 13, p. 85-95. Haq, B. U., Hardenbol, J. & Vai, P.R., 1987 : C hronology of uctuating sea levels since t h e Triassic. Science 235, 1156-1167. Plissi, & Sig, B., 2006 : 30 millions dannes de bio diversit dynamique dans le palokarst du Quercy. Journes Bernard Gze. Strata (sr. 1) 13, 1-284. Sig, B., Aguilar, J.-P., Marandat, B. & Astruc, J.G., 1991: Extension au Miocne infrieur des remplissages p h osp h ats du Quercy. La faune de vertbrs de Crmat (Lot, France). Geobios 24, 4, 497-502 Simon-Coinon, R., 1989 : Le rle des paloaltrations et des paloformes dans les socles : lexemple du Rouergue (Massif Central Franais). se Doct. s Lettres, Paris I, ENSMP, Mmoire des Sciences de la Terre, n 9, 290 p. E ARL Y LATE P LIOCENE PALEOKARSTIC FILLINGS PREDATING THE MAJOR P LIO P LEISTOCENE EROSION OF THE ...



PAGE 1

F OSSIL POPULATION STRUCTURE AND MORTALIT Y OF THE CAVE BEAR FROM THE M OKRICA CAVE N ORTH S LOVENIA STRUKTURA FOSILNE POPULACIJE IN UMRLJIVOST JAMSKEGA MEDVEDA IZ M OKRIKE JAME S EVERNA S LOVENIJA Irena D EBELJAK 1 Izvleek UDK 902.035:569.74(497.4) Irena Debeljak: Struktura fosilne populacije in umrljivost jamsk ega medveda iz Mokrike jame (severna Slovenija): Struktura fosilne populacije jamskega medveda iz Mokrike jame je bila prouena z namenom, da bi dobili nove podatke o vedenju in umrljivosti te izumrle vrste. Starost ob smrti je bila ocenjena za 128 razlini h osebkov, in sicer na podlagi analize cementni h prirastnic, formiranosti korenine in obrabe krone levi h M 1 zob. Po frekvenni distribuciji primerkov v enolet ni h intervali h la h ko ocenimo trend umrljivosti v razlini h ivljenjski h obdobji h in ji h interpretiramo glede na podatke o dananji h medvedi h. V prvotni mrtvi zdrubi so po vsej ver jetnosti prevladovali mladii. Izjemno kr h ki molarji manj kot 6 mesecev stari h ivali se niso fosilno o hranili. Enoletniki so najbolj tevilno zastopana starostna skupina v fosilni popu laciji iz Mokrike jame. Umrljivost je drastino upadla, potem ko so jamski medvedi preiveli svojo prvo hibernacijo v drugi zimi. Najnija stopnja umrljivosti je bila ugotovljena za sta rostno obdobje 9-15 let, ko so bili jamski medvedi oitno na vr h uncu moi. Z analizo cementa ugotovljena najstareja sta rost znaa okoli 30 let, kar kae, da je bila najdalja ivljenjska doba podobna kot pri dananji h medvedi h. Analiza zobni h tkiv je pokazala, da je bila umrljivost v jami sezonsko omejena veina ivali je poginila pozimi in zgodaj spomladi. Spolna struktura je bila raziskana na vzorcu 750 podonikov. Izrazito veji dele samcev v skupini stareji h mladiev in napol odras li h ivali bi la h ko razloili z dejstvom, da je tudi pri dananji h medvedi h obdobje osamosvajanja bolj kritino za samce. Pri mlaji h odrasli h in zreli h odrasli h ivali h je bila umrljivost ver jetno veja pri samica h. Spolna struktura odrasli h medvedov, e posebej stareji h ivali, kae, da so Mokriko jamo kot zimski brlog veinoma zasedali samotarski samci. Kljune besede: jamski medved, umrljivost, starostna struk tura, spolna struktura, cement. 1 Ivan Rakovec Institute of Palaeontology, ZRC SAZU, Novi trg 2, Ljubljana, Slovenia, e-mail: irenade@zrc-sazu.si Received/Prejeto: 25.09.2007 COBISS: 1.01 ACTA CARSOLOGICA 36/3, 475-484, POSTOJNA 2007 Abstract UDC 902.035:569.74(497.4) Irena Debeljak: Fossil population structure and mortality of the cave bear from the Mokrica cave (North Slovenia) e fossil population structure of t h e cave bear from t h e Mokri ca cave was evaluated to provide new data concerning t h e be h aviour and mortality of t his extinct species. Age at deat h was estimated for 128 dierent individuals by analysing cementum increments, root formation and crown wear of le M 1 teet h. Aer t h e frequency distribution of specimens t hroug h oneyear intervals, t h e mortality trends can be estimated for various lifetime periods, and interpreted in accordance wit h data for present-day bears. e original deat h assemblage was presum ably juvenile-dominated. Extremely fragile molars of less t h an 6 mont h old cubs did not get preserved. Y earlings are t h e most numerous age class in t h e fossil population from t h e Mokrica cave. Mortality drastically dropped aer cave bears survived t h eir rst hibernation in t h e second winter. e lowest mortal ity rate was observed in t h e 9-15 years age group, w h en cave bears would be expected to be in t h eir prime. e oldest age re corded by cementum analysis is approximately 30 years, w hic h indicates t h at t h e maximum life span was similar to present-day bears. Study of dental tissues s h ows t h at t h e mortality in t h e cave was seasonally restricted t h e majority of deat h s in t h e cave occured during winter and in early spring. Sex structure of t h e fossil population h as been studied on t h e sample of 750 ca nines. e signicantly hig h er proportion of males in t h e group of older juveniles and subadults could be explained by t h e fact t h at t h e weaning period is more critical for males also in pres ent-day bears. In young adults and prime adults t h e mortality was presumably hig h er in females. e sex structure of adult bears, especially in t h e sample of older individuals, indicates t h at t h e Mokrica cave was used as winter den mostly by solitary males. Keywords: cave bear, mortality, age structure, sex structure, cementum.

PAGE 2

ACTA CARSOLOGICA 36/3 2007 476 Age and sex structure of t h e fossil population of t h e cave bear (Ursus spelaeus) from t h e Mokrica cave (Mokrika jama) was studied to generate new data and insig h ts con cerning t h e be h aviour and mortality of t his extinct spe cies. Mokrica cave is located in t h e Kamnik-Savinja Alps (N Slovenia) between Mokrica and Koutna, at an alti tude of 1500 metres, about 1000 metres above Kamnika Bistrica river valley. It is h orizontal, about 45 metres long and 7 to 20 metres wide h orizontal cave, formed in Tri assic limestone. Its entrance faces in Nort h east. Mokrica cave, also called Bear cave (Medvedja jama), h as been well known since t h e early 1800s for its ric h accumula tion of cave bear remains. Systematic excavations were carried out by Mitja Brodar during 1954-1956 and 1960 in t h e entrance part of t h e cave (Brodar, 1959). Ten strata (5-14) of Pleistocene sediments containing cave bear re mains h ave been recognized t h ere, approximately 6 me tres t hick. e majority of cave bear remains (isolated teet h and fragmented bones) were recovered from strata 6 and 7. Systematic palaeontological analysis of t h e mate rial was carried out by Rakovec (1967). Cave bear domi nates t h e assemblage; fossil remains of ot h er animals (wolf, cave lion and ibex) are scarce and h ave been found only in stratum 7 (Rakovec, 1967). Artefacts from strata 6 and 7 h ave been attributed to t h e Aurignacian cultural level (Brodar, 1959). Findings of M. Brodar suggest t h at cave bear remains mostly derive from t h e Middle Wrm ian period (OIS 3). Data on t h e cave bear population structure from t h e Mokrica cave can be compared to data (Debeljak, 2004) obtained from a similar study of cave bear mate rial from anot h er well-known Aurignacian station and Middle Wrmian hig h Alpine cave bear site Potoka zi jalka, w hic h is situated just 17 km away from t h e Mokrica cave. MATERIAL AND METHODS INTRODUCTION Fossil population structure of t h e cave bear from t h e Mokrica cave h as been derived from teet h t h at were col lected along wit h ot h er fossil material during Brodars excavations more t h an 50 years ago. e majority of t h e material is stored in t h e Natural History Museum in Ljubljana, but some specimens are kept in t h e palaento logical collection of t h e Department of Geology, Univer sity in Ljubljana. e sediment h as not been water-sieved during t h e excavations in t h e Mokrica cave. For t his rea son, milk teet h were not recovered and it is t h erefore not possible to study t h e mortality of cubs younger t h an 6 mont h s. I analysed fossil population age structure using rst lower permanent molars (M 1 ), bot h isolated and still at tac h ed in mandibles. Only le specimens were analysed to prevent t h e doubling of data for t h e same individu als. For 107 isolated M 1 specimens age at deat h was de termined by t h e analysis of dental tissues (dentine and cementum) in cross sections t hroug h t h e root, about 7.5 mm below t h e crown. in sections and etc h ed/coloured ground sections were prepared for all samples (for labo ratory tec hniques see Debeljak, 1996; 2000). In juvenile specimens I measured t h e t hickness of t h e root wall. e relative widt h of pulp canals and dentine increments can be used as criteria to assess individual age in juveniles and subadults (up to t h e 4 t h year). e individual age of all teet h belonging to animals older t h an 1 year was de termined by counting cementum increments. is is t h e most reliable and objective met h od for determining in dividual age of many mammals and h as long been used routinely in present-day bears as well (see Debeljak, 2000 and references t h erein). Cementum is a bone-like tissue t h at is continuously deposited on t h e surface of toot h roots t hroug h out life. Dental cementum is composed of so-called winter and summer layers, similar to growt h rings in trees (Fig. 1). Aer t h e nature of t h e outermost cementum increment, it is sometimes possible to deter mine even t h e season of deat h. Dental tissues of 21 le M 1 teet h still attac h ed in mandibles h ave not been micro scopically examined, because it would be impossible to extract t h em wit h out visible damage to t h e jawbone. For t h ese specimens individual age was roug h ly estimated by comparing t h eir crown wear patterns to t h e teet h of known individual age. Altoget h er, age at deat h h as been estimated for 128 dierent individuals. In bears, sex can be determined aer dimensions of canines. Several studies indicate t h at t h ere is no signi cant dierence between measurements of lower and upper canines (Koby, 1949; Kurtn, 1955; Rabeder, 2001). Bot h upper and lower specimens were included in t h e anal ysed sample, because it is oen impossible to distinguis h between t h em w h en t h ey are damaged or h eavily worn, especially in t h e group of older individuals. Measure ments of crown widt h of 377 undamaged specimens were I RENA D EBELJAK

PAGE 3

ACTA CARSOLOGICA 36/3 2007 477 taken at t h e crown base. e histogram of crown widt h s h ows clear bimodal distribution (Fig. 2); specimens wit h crown widt h larger t h an 18.5 mm were attributed to males and specimens wit h crown widt h smaller t h an 18 mm as belonging to females. Mean value of crown widt h is 21.3 mm for males and 15.8 mm for females. Also Kur tn (1955) determined mean size dierence of more t h an 5 mm and joint overlap of less t h an 1 per cent in t h e Mixnitz sample. Sexual dimorp hism is so well pronounced in ca nines t h at usually even t h ose specimens t h at are too dam aged or worn to be measured accurately can be sexed based on t h eir overall dimensions. Altoget h er, population sex structure (males to females ratio) h as been establis h ed on t h e sample of 750 canines (isolated and t h ose still attac h ed in jawbones) of indi viduals older t h an 18 mont h s. Canines of yearlings were omitted from t h e study because t h ey are composed only of a h ollow crown cone w hic h is usually damaged in t h e lowermost part and t h erefore dircult to measure. All t h e canines h ad been previously separated into 3 dierent age classes. Canines wit h t h e root widely open at t h e apex are classied into t h e group of older juveniles and subadults. Our studies of canines, w h ose individual age h as been determined by cementum analysis, s h ow t h at t h eir root closed at t h e age of 5-6 years. Canines wit h a closed root can be t h erefore all assigned adult status. Adult specimens were furt h er divided into two groups: specimens wit h slig h tly to moderately worn crowns, and specimens wit h markedly to h eavily worn crowns. ere is no clear distinction between t h e two groups. In ca nines, crown wear was quite erratic and h eavy wear does not necessarily indicate old age. Canines could h ave been broken and pat h ologically worn even by prime adults. However, preliminary study of dental cementum in ca nines from Divje babe cave indicates t h at t h e majority of specimens from t h e rst group belonged to 5-15-yearold individuals (youngand prime adults) and specimens from t h e second group to 15-25-year-old individuals (older adults). Fig. 1: Dental cementum in the cave bear M 1 (inv. number M. 4228) from the M okrica cave. Summer and winter increments can be seen, divided by lines of arrested growth. Dark dots are cementocyte lacunae. An arrow indicates the direction of cementum deposition. in section; plane-polarized light (le image) and cross-polarized light (right image). e eld is about 0.8 mm wide. Fig. 2: H istogram for crown width (in mm) of upper and lower canines from the M okrica cave. Bimodal distribution shows marked sexual dimorphism. RESULTS AND DISCUSSION A GE STRUCTURE : e results of t h e individual age analysis of 107 isolated le M 1 teet h are s h own in t h e frequency histogram of individuals in one-year age classes (Fig. 3). Frequencies in some of t h em are not integers because, for example, in t h e case of a specimen wit h age estimation of 14-15 years, one h alf was attributed to t h e class of 14-year-olds and one h alf to 15-year-olds. From t his prole, t h e course F OSSIL POPULATION STRUCTURE AND MORTALIT Y OF THE CAVE BEAR FROM THE M OKRICA CAVE N S LOVENIA

PAGE 4

ACTA CARSOLOGICA 36/3 2007 478 of mortality in t h e cave can be estimated for various life periods or age groups. Fig. 4 s h ows t h e proportions of dierent age classes in t h e sample of 128 le M 1 speci mens, including bot h isolated and t h ose still attac h ed in t h e jawbones. Y earlings and 2-year-olds are considered to be juveniles, 3-4-year-olds are subadults, 5-8-year-old individuals young adults, 9-20-year-olds prime adults and t h ose older t h an 20 years old adults (Craig h ead et al., 1995). e mortality in t h e main age groups will be discussed briey: Like in present-day bears, cave bear cubs were born in a winter den, most probably in December or January. Still unerupted M 1 crowns of 2-4-mont h -old cubs t h at died in t h e rst winter, before t h e onset of spring, were h ollow and fragile as an egg s h ell. ey were completely destroyed post mortem and are t h erefore not represented in t h e analysed sample (Fig. 3). In Divje babe cave, w h ere all t h e sediment h as been water-sieved during t h e excava tions, a very large number of milk teet h h ave been recov ered. e analysis of fossil population structure from Divje babe I site (Debeljak, 2002) indicated t h at t h e mortality was h ig h est for t h is youngest age group. Most likely, t h is was t h e case in t h e Mokrica cave as well, w h ic h is indicat ed by an arrow in Fig. 3. Y earlings (9-15-mont hold individuals) are t h e most numerous age group in t h e fossil population wit h a 21.1% s h are (Figs. 3, 4), sug gesting t h at t h e mortality in t h e second winter was very hig h, alt h oug h presumably lower t h an t h e mortality in t h e rst winter, as explained above. Very hig h mortality in yearlings may be at least partly attributable to dier ent predators, suc h as wolves and male bears. Cannibalistic be h aviour of adult males and t h eir aggression towards cubs is well known in present-day bears (McCulloug h, 1981; Tietje et al., 1986; LeCount, 1987; Miller, 1990; Craig h ead et al., 1995; Deroc h er & Wiig, 1999). Bears hibernate in t h eir second winter for t h e rst time (Craig h ead et al., 1995; McNamee, 1997). Dirculties in establis h ing and maintaining special metabolic processes t h at are c h aracteristic of hibernation (Nelson et al., 1983; Hellgren, 1998) could be important factor for t h e observed hig h mortality during t his period. W h en t h eir fat reserves are not adequate, bears are un able to start hibernating, h owever, only hibernation en ables t h em to survive more t h an 6 mont h s wit h out food. Cases of yearlings t h at died in t h eir den during especially long winters h ave been reported for present-day bears (Sc h oen et al., 1987; Rogers, 1981; 1987). e frequency of 2-year-old individuals is also hig h, but signicantly lower t h an in yearlings (Fig. 3), w hic h indicates a s h arp decline in mortality aer t h e second winter. e proportion of yearlings in comparison to 2year-olds was presumably even larger originally, because t h eir teet h were muc h more prone to destruction. M 1 teet h of yearlings h ave only about 1.25 mm t hick root wall and are t h erefore more easily fragmented t h an M 1 teet h of 2-year-old individuals wit h mec h anically more resistant, approximately 2.5 mm t hick root wall. e measurements of t h ickness of root wall by juve Fig. 3: M ortality prole frequency distribution of isolated le lower M 1 teeth through one-year intervals (N=107). Extremely fragile molars of less than 6 month old cubs were not preserved. e supposed highest mortality in this age group is indicated by an arrow. Fig. 4: P ie chart with proportions of le M 1 teeth; isolated and still attached in mandibles (N=128) from dierent age groups. I RENA D EBELJAK

PAGE 5

ACTA CARSOLOGICA 36/3 2007 479 nile M 1 teet h (Fig. 5) yielded anot h er conclusion, t h at t h e mortality in t h e cave was not distributed evenly t h roug h out t h e year. e rst peak in t h e distribution s h own in Fig. 5 represents yearlings t h at died during t h eir second winter or in t h e following early spring. e second peak consists of 2-year-olds t h at died in t h eir t h ird winter. e interven ing h iatus indicates t h at bears were absent from t h e cave during t h e second summer period. Specimens t h at could be attributed to about 6-8-mont h -old cubs are also missing in t h e sample. Bot h gaps in t h e mortality prole suggest t h at cave bears were not returning to t h e cave aer t h ey h ad le it in t h e spring (from April to early June). Per h aps t h ey visited it occasionally, but t h ey stayed away at least during mont h s of intensive feeding. Microscopic analysis of growt h layers in dental tissues (dentine and cementum) h as also s h own t h at great majority of deat h s in t h e Mokrica cave occured during t h e h ibernation and in a s h ort post h ibernation period. e same results from Potoka zijalka (Debeljak, 2004) contradict some older assumptions t h at h ig h Alpine caves were continuously occupied by bears in all seasons, all t h e year round (E h renberg & Sickenberg, 1929; E h renberg, 1967). It seems t h at mortality of 3-year-olds was not muc h lower t h an in 2-year-olds (Fig. 3). Alt h oug h 3-year-olds were p hisically stronger, t his was probably quite a critical age. Brown bear females usually abandon t h eir young at t h e age of about 2.5 years, sometimes one year earlier or later (Bunnel & Tait, 1981; Craig h ead et al., 1995). Promi nent decline of mortality by 4-year-olds (Fig. 3) indicates t h at t h e weaning period presumably took place already in t h e t hird year, like in present-day bears. Frequencies of M 1 teet h belonging to young adults are more or less evenly distributed t hroug h one-year in tervals (Fig. 3). At t h e age of 5-8 years brown bear females w h elp t h eir young for t h e rst time (Bunnel & Tait, 1981; Craig h ead et al ., 1995). e mortality could be somew h at hig h er in young, inexperienced cave bear females in comparison to older females. Fig. 3 s h ows t h at t h ere was a furt h er decline of mortality by prime adults, aer t h e age of 8, and became more prominent again aer t h e age of 15. As in present-day bears, t h e mortality caused by factors ot h er t h an man is extremely low in t his age group (McNamee, 1997). Rare cases of 9-15-year-old bears t h at died in t h e Mokrica cave could be explained by natural causes (injuries, disease) and per h aps even by occasional h unting by pre historic man. Rakovec (1967) assumed t h at man came to t h e cave from time to time only, above all during t h e summer mont h s w h en h e h unted cave bears. However, t h ere is no direct evidence for cave bear h unting in t h e Mokrica cave. Furt h ermore, t h e supposed h unting would probably take place during hibernation, w h en bears were in t h e most vulnerable position. It seems t h at in t h e group of older adults, mortality started to rise already at t h e age of 16 (Fig. 3). In Potoka zijalka moderate mortality increase was observed not earlier t h an around t h e age of 20. is dierence could be due to a small sample size. e proportion of individuals older t h an 20 years was most likely hig h er in t h e origi nal deat h assemblage t h an it could be inferred from t h e analysed sample (Fig. 3), because remnants of extremely worn teet h were easily broken and h ave not been recov ered during excavations. ere is a somew h at lower pro portion of older adults in t h e sample of isolated M 1 teet h (Fig. 3) t h an in t h e sample of all M 1 teet h (Figs. 4 and 5), because 12 out of 21 M 1 specimens still rmly attac h ed in t h e mandibles were attributed to individuals older t h an 15 years, based on t h eir crown wear. However, wit h out sectioning t h e root and microscopic examination of den tal cementum it was not possible to accurately determine t h eir individual age. Mortality around t h e age of 20 can be attributed to senescence-related malnutrition and disease. M 1 teet h were h eavily worn by t his age, sometimes to suc h a de gree t h at pulp canal in t h e root became exposed. Dental pat h ologies are very frequent on jawbones of older cave bears, t h at could hinder feeding to suc h degree t h at ani mal could not gain enoug h weig h t before winter. It h as been suggested in t h e past t h at t h e longevity in cave bear was severely restricted by t h e h eavy wear of teet h and t h at it was somew h at s h orter t h an in t h e brown bear, because cave bear teet h wore out more quickly. Kurtn (1958; 1976) expressed doubt t h at any cave bear lived beyond about twenty years of age. e mortality prole from t h e Mokrica cave does not support t his h ypot h esis. e oldest age recorded is approximately 30 years, w hic h indicates t h at t h e maximum life span was similar to pres ent-day bears; in t h e wild t h eir life span is 25-30 years (Bunnel & Tait, 1981; Deroc h er, 2000). Fig. 5: Number of M 1 specimens in classes with particular thickness of root wall, measured approximately 7.5 mm under the crown. Specimens (N=42) belonged to yearlings and 2-year-olds. F OSSIL POPULATION STRUCTURE AND MORTALIT Y OF THE CAVE BEAR FROM THE M OKRICA CAVE N S LOVENIA

PAGE 6

ACTA CARSOLOGICA 36/3 2007 480 e comparison of proportions of dierent age groups for t h e Mokrica cave and Potoka zijalka is given in Fig. 6. e most striking dierence between t h e two sites is in t h e proportion of individuals younger t h an 18 mont h s, w hic h is muc h hig h er in Potoka zijalka (64% vs. 21.1% in t h e Mokrica cave). Suc h a discrepancy could be at least t h eoretically explained if Potoka zijalka cave would be used primarily by females and t h eir ospring and t h e Mokrica cave on t h e ot h er h and by adult males. However, t his was obviously not t h e case. On t h e con trary, males predominate in Potoka zijalka in even hig h er proportion t h an in t h e Mokrica cave. It is also not likely t h at dierent ecological conditions would inu ence t h e survival rate of juveniles to suc h a great degree. As it h as been already pointed out, teet h of individuals younger t h an 18 mont h s are h ollow and were t h erefore muc h more prone to destruction t h an teet h of older in dividuals. In t h e analysed sample of 89 M 1 teet h from Potoka zijalka, t h ere are even 10 specimens (11.2%) of 2-4-mont h-old individuals, w hic h indicates t h e excep tionally good state of preservation of t h e fossil material. is assumption is furt h er supported by t h e very hig h ratio of M 1 teet h of yearlings to mec h anically more resis tant M 1 teet h of 2-year-olds; t h e ratio of 8.5 to 1 h as been establis h ed in t h e Potoka zijalka sample, in comparison to less t h an 1.5 to 1 ratio for t h e Mokrica cave. Teet h and mandibules from Potoka zijalka are generally in a muc h better state of preservation t h an t h ose from t h e Mokrica cave. Discrepancy in t h e proportion of less t h an 18 mont h old individuals between t h e two sites could be t h erefore explained by dierent tap h onomic conditions. Sediment protects skeletal material from destructive tap h onomic factors, suc h as scavenging, trampling and weat h ering (Lyman, 1994). e condi tions were more favourable for preservation of juvenile teet h w h en t h e material be came quickly buried in t h e sediment, w hic h was prob ably t h e case in Potoka zi jalka. Furt h ermore, t h e pro portion of h ollow and for t his reason oen fragmented juvenile teet h is also related to t h e accuracy in collecting fossil material during excava tions. In Potoka zijalka also fragmented teet h were recov ered during t h e new 19972000 excavation campaign (Pac h er et al ., 2004). On t h e ot h er h and, only more or less complete teet h were kept in t h e more t h an 50 year old collection of cave bear material from t h e Mokrica cave. is could additionally contribute to t h e underrepresen tation of juvenile specimens in t h e analysed sample. If we exclude M 1 specimens of less t h an 18 mont h old in dividuals, t h at are most sensitive to dierent tap h onomic conditions and recovery tec hniques, we get very similar proportions of t h e remaining age groups for bot h sites: t h ere are 40.1% of 2-4-year-olds, 39.6% of youngand prime adults, and 20.3% of more t h an 20 years old adults in t h e Mokrica sample (n = 101), and 37.5% of 2-4-yearolds, 35.9% of youngand prime adults, and 26.6% of old adults in t h e Potoka zijalka sample (n = 32). Aer t h e model proposed by Stiner (1990; 1994; 1998), t h e fossil population from t h e Mokrica cave exhib its t h e so-called normal non-violent attrition (NNVA) mortality pattern, w hic h is in Stiners opinion consistent wit h hibernation-related mortality, caused by non-vio lent factors, principally starvation. However, in Potoka zijalka, w h ere fossil material is better preserved, juve nile-dominated mortality pattern h as been establis h ed (Fig. 7). Considering t h e presumed underrepresentation of juveniles in t h e analysed sample, t h e original deat h assemblage was most likely distinctly juvenile-domi nated in t h e Mokrica cave as well. e predominance of juveniles h as been reported from many cave bear sites (Bc h ler, 1957; Kurtn, 1958; Rabeder, 1992; Andrews & Turner, 1992; Grandal dAnglade & Vidal Roman, 1997; Weinstock, 1999; 2000; 2001; Pac h er, 2000; Germonpr & Sablin, 2001; Debeljak, 2002; Turner, 2002). In my opin Fig. 6: Frequencies and percentages of le lower M 1 teeth in dierent age classes from M okrika jama (N= 128) and P otoka zijalka (N=89). I RENA D EBELJAK

PAGE 7

ACTA CARSOLOGICA 36/3 2007 481 Fig. 8: e proportion of female canines to male canines (upper and lower) in dierent age groups for the M okrica cave (N=750) and P otoka zijalka (N=90). Number of specimens is shown inside the bars. ion, t h e juvenile-dominated population structure is c h ar acteristic of hibernation-related mortality in typical cave bear sites, provided t h at fossil material, including fragile juvenile teet h, is well preserved and accurately collected. Mortality patterns in t h e Mokrica cave and Potoka zi jalka s h ow no evidence for intensive, non-selective cave bear h unting, w hic h s h ould, according to Stiner (1990; 1994; 1998), aect all age groups randomly and result in t h e so-called living structure (LS) mortality pattern, re sembling t h e age structure of t h e living population. S E X STRUCTURE : e results of sexing 750 canines of individuals older t h an 18 mont h s are presented in Fig. 8. In t h e group of older juveniles (2-year-olds) and subadults (3-4-year-olds), males predominate over females wit h 1.8 to 1 ratio (63.8% males, 36.2% females). is could be explained wit h hig h er mortality rate of males in t his age class, w hic h was observed also in populations of present-day bears. e weaning period is obviously somew h at more critical for males t h an females. Subadult females are better tolerated by adult females and are oen allowed to stay in t h e same h ome range aer weaning, w hile subadult males usually travel larger distancies to establis h t h eir own territory (Elowe & Dodge, 1989; Swenson et al., 1998; McLellan & Hovey, 2001). ey can be outcompeted in t h e searc h for food and can fall pray to adult males, w h o can be very aggressive toward subadult males. Intraspecic compe tition is especially en h anced by hig h population density and s h ortage of available food (McCulloug h, 1981; Bun nel & Tait, 1981; Tietje et al., 1986; McNamee, 1997). e proportion of males and females is nearly equal in t h e group of young and prime adults (~ 5-15-year-olds). Hibernation period is considerably prolonged in females wit h cubs in comparison to solitary males in present-day bears (Judd et al., 1986; Sc h oen et al., 1987; Sc h wartz et al., 1987; Miller, 1990; Van Daele et al., 1990; McNamee, 1997; Manc hi & Swenson, 2005). Furt h ermore, females can get injured w h en protecting t h eir ospring. We as sume t h at mortality was actually somew h at hig h er in adult females t h an in adult males. However, if solitary males denned in t h e Mokrica cave more frequently, t his could result in approximately equal proportion of bot h sexes in t h e fossil population. In present day bears, females rst produce cubs by t h e age of 5-8 years (Bunnel & Tait, 1981; Craig h ead et al., 1995). In t h e sample of 247 canines of prime and young adults, 42 specimens belong to t h e 5-8 years age class. Interestingly, 28 of t h em are of females and only 14 of males (2:1 ratio). It s h ould be taken into consid eration t h at large, unworn ca nines are t h e most attractive nds and t h at males could be articially underrepresented in t his age group. If not, t his Fig. 7: Tripolar graph denoting proportions of three main age categories in the samples of le M 1 teeth from the M okrica cave and P otoka zijalka. Aer Stiner (1990; 1994; 1998) dark grey eld represents NNVA (normal non-violent attrition) mortality pattern and light grey LS (living structure) pattern. F OSSIL POPULATION STRUCTURE AND MORTALIT Y OF THE CAVE BEAR FROM THE M OKRICA CAVE N S LOVENIA

PAGE 8

ACTA CARSOLOGICA 36/3 2007 482 CONCLUSIONS Mortality pattern in t h e Mokrica cave s h ows no evidence for intensive, non-selective cave bear h unting. e age structure of t h e analysed sample of t h e cave bear fossil population exhibits t h e so-called normal non-violent at trition (NNVA) mortality pattern, wit h a 52.7% s h are of juveniles and subadults (1-4-year-olds), 31.3% of youngand prime adults (5-20-year olds) and 16% of old adults (>20 years old individuals). e Mokrica cave was used by cave bears as a winter and birt hing den and was pre sumably not visited regularly during summer mont h s. e majority of deat h s occured during hibernation or in s h ort post hibernation period, mainly because of star vation and predator attacks. Mortality was presumably hig h est in t h e rst winter w h en cubs were born, h ow ever, extremely fragile crowns of M 1 teet h of individuals younger t h an 4 mont h s h ave not been preserved. Y ear lings (9-15mont h-old individuals) are t h e largest age class wit h 21.1% s h are. Hollow teet h of yearlings presum ably underwent muc h hig h er tap h onomic losses t h an mec h anically more resistant teet h of older individuals. It can be inferred t h at t h e proportion of individuals young er t h an 18 mont h s was muc h hig h er originally and t h at t h e deat h assemblage was markedly juvenile-dominated. e mortality rate dropped drastically aer cave bears successfully survived t h e second winter. e next signi cant decrease of mortality occured at t h e age of 4 years, w hic h indicates t h at females le t h eir young in t h e t hird year on avarage, like in present-day bears. e weaning age is especially critical for subadult males, w hic h could explain hig h er proportion of males (63.8% M, 36.2% F) in t h e group of older juveniles and subadults. e mortality rate in young adults (5-8-year-olds) was still muc h hig h er t h an in prime adults. From t h e age of 9 to 15 years, deat h s were very rare. Mortality was pre sumably hig h er in females, because t h ey hibernated lon ger t h an males. However, solitary males probably denned in t h e Mokrica cave more frequently, w hic h resulted in approximately equal proportion of bot h sexes (51.8% M, 48.2% F) in t h e studied sample of canines of youngand prime adults (5-15-year-olds). In t h e sample of older adults, h owever, males clearly predominate (73% M, 27% F), w hic h is anot h er indication t h at t h e Mokrica cave was mostly occupied by solitary males. Old females c h ose t h e cave for t h eir denning place less oen. Moderate increase in mortality can be observed already aer t h e age of 16. Mortality around t h e age of 20 can be attributed to se nescence-related malnutrition and disease. e oldest age recorded by cementum analysis is approximately 30 years, w hic h indicates t h at t h e maximum life span was similar as in present-day bears. could be t h e evidence t h at maternity-related mortality was increased especially in t h e group of very young, in experienced females. In t h e group of older adults (~ 15-25-year-olds) males considerably outnumber females (73% males, 27% females; 2.7:1 ratio), w hic h is anot h er indication t h at t h e Mokrica cave was mostly occupied by solitary males. Old females obviously c h ose t h e cave for t h eir denning place less oen. In Potoka zijalka, even hig h er proportion of males (83.3%) is observed in t his age class. e predomi nance of males cannot be explained by t h e assumption t h at males lived longer on average t h an females. is is not t h e case in present-day bears. On t h e contrary, some data even s h ow t h at females live longer on average t h an males (Craig h ead et al., 1995; McNamee, 1997; Wiig, 2000). ACKNOWLEDGEMENTS I t h ank Mitja Brodar and t h e Institute of Arc h aeology, ZRC SAZU, for t h e material provided by t h eir excava tions, and Katarina Krivic for h er h elp wit h t h e cave bear collection in t h e Natural History Museum. I am also very grateful to Kata Cvetko-Bari for t h e laboratory prepa ration of t hinand ground sections, to Vida Po h ar and Martina Pac h er for valuable comments and critical re view of t h e text, and to Glenn Jaecks for language-edit ing. I RENA D EBELJAK

PAGE 9

ACTA CARSOLOGICA 36/3 2007 483 Andrews, P. & Turner, A., 1992: Life and deat h of t h e Westbury bears.Ann. Zool. Fennici, 28, 139-149, Helsinki. Bc h ler, H., 1957. Die Altersgliederung der H h len brenreste im Wildkirc h li, Wildenmannlisloc h und Drac h enloc h.Quartr, 9, 131-146, Berlin. Brodar, M., 1959: Mokrika jama, nova visokoalpska aurignaka postaja v Jugoslaviji. (Mokrika jama, station nouvelle aurignacienne des Hautes-Alpes en Y ougoslavie.)Razprave IV. razr. SAZU, 5, 415-469, Ljubljana. Bunnel, F. L. & Tait, D. E. N., 1981: Population dynam ics of bears implications. In: C. W. Fowler & T. D. Smit h (Eds.): Dynamics of Large M ammal P opula tions. Jo hn Wiley & Sons, 75-98, New Y ork. Craig h ead, J. J., Sumner, J. S. & Mitc h ell, J. A., 1995: e grizzly bears of Y ellowstone: their ecology in the Y el lowstone ecosystem, 1959-1992.Island Press, p. 535, Was hington. Debeljak, I., 1996: A simple preparation tec hnique of cave bear teet h for age determination by cementum in crements.Rev. Palobiol., 15, 1, 105-108, Genve. Debeljak, I., 2000: Dental cementum in t h e cave bear; comparison of dierent preparation tec hniques.Geol. zbornik, 15, 53-66, Ljubljana. Debeljak, I., 2002: Fossil population structure of t h e cave bear from Divje babe I site, Slovenia: Preliminary results.Ab h. zur Karstund H h lenkunde, 34, 4148, Mnc h en. Debeljak, I., 2004: Fossil Population Structure of t h e Cave Bear from Potoka zijalka (Slovenia).Mitt. Komm. Quartrforsc h. sterr. Akad. Wiss., 13, 173-182, Wien. Deroc h er, A. E., 2000: Die Bren Ein berblick ber Ver h alten und kologie. In: U. Gansloer (Ed.): Die Bren. Filander-Verl., 29-67, Frt h. Deroc h er, A. E. & Wiig, O., 1999: Infanticide and canni balism of juvenile polar bears (Ursus maritimus) in Svalbard.Arctic, 5273, 307-310, Calgary. Elowe, K. D. & Dodge, W. E., 1989: Factors aecting black bear reproductive success and cub survival.J. Wildl. Manage., 53/4, 962-968, Bet h esda. E hrenberg, K., 1967: Der H h lenbr, sein Vorkommen und seine Bezie h ungen zur Umwelt.Mitt. sterr. Arbeits. Ur.-u. Fr h gesc h., 18, 34-50, Wien. E hrenberg, K. & Sickenberg, O., 1929: Eine plistozne H h lenfauna aus der Hoc h gebirgsregion der Ostal pen.Palaeobiologica, 2, 303-364, Wien, Leipzig. Germonpr, M. & Sablin, M. V., 2001. e cave bear (Ur sus spelaeus) from Goyet, Belgium. e bear den in C h amber B (bone h orizon 4).Bull. Inst. Roy. Sci. Nat. Belg. Sci. Terre, 71, 209-233, Brussel. Grandal dAnglade, A. & Vidal Roman, J. R., 1997: A population study on t h e Cave Bear ( Ursus spelaeus Ros.-Hein.) from Cova Eirs (Triacastela, Galicia, Spain).Geobios, 30, 5, 723-731, Villeurbanne. Hellgren, E. C., 1998: P h ysiology of hibernation in bears.Ursus, 10, 467-477. Judd, S. L., Knig h t, R. R. & Blanc h ard, B. M., 1986: Den ning of grizzly bears in t h e Y ellowstone National Park area.Bears eir Biology and Management., 6, 111-117. Koby, F.-Ed., 1949: Le dimorp hisme sexuel des canines d Ursus arctos et d Ursus spelaeus .Rev. Suisse Zool., 56/36, 675-687, Genve. Kurtn, B., 1955: Sex dimorp hism and size trends in t h e cave bear, Ursus spelaeus Rosenmller & Heinrot h.Acta Zool. Fennica, 90, 1-48, Helsinki. Kurtn, B., 1958: Life and deat h of t h e Pleistocene cave bear. A study in paleoecology.Acta Zool. Fennica, 95, 1-59, Helsinki. Kurtn, B., 1976: e cave bear story. Life and death of a vanished animal. Columbia Univ. Press, p. 163, New Y ork. Lyman, R. L., 1994: Vertebrate Taphonomy. Cambridge Manuals in Arc h aeology, Cambridge University Press, p. 524, Cambridge. LeCount, A. L., 1987: Causes of black bear cub mortality.Bears eir Biology and Management, 7, 75-82. Manc hi, S. & Swenson, J. E., 2005: Denning be h aviour of Scandinavian brown bears Ursus arctos .Wildlife Biology, 11, 2, 123-132, Ronde. McCulloug h, D. R., 1981. Population dynamics of Y el lowstone grizzly bear. In: C. W. Fowler & T. D. Smit h (Eds.): Dynamics of Large M ammal P opulations. Jo hn Wiley & Sons, 173-196, New Y ork. McLellan, B. M. & Hovey, F. W., 2001: Natal dispersal of grizzly bears.Can. J. Zool., 79, 838-844, Ottawa. McNamee, T., 1997: e grizzly bear. Lyons & Burford Publis h ers, p. 314, New Y ork. Miller, S. D., 1990: Denning ecology of brown bears in sout h central Alaska and comparisons wit h a sym patric black bear population.Bears eir Biology and Management, 8, 279-287. Nelson, R. A., Folk, G. E., Pfeier, E. W., Craig h ead, J. J., Jonkel, C. J. & Steiger, D. L., 1983: Be h avior, Bio c h emistry, and Hibernation in Black, Grizzly, and Polar Bears.-Bears eir Biology and Manage ment, 5, 284-290. REFERENCES F OSSIL POPULATION STRUCTURE AND MORTALIT Y OF THE CAVE BEAR FROM THE M OKRICA CAVE N S LOVENIA

PAGE 10

ACTA CARSOLOGICA 36/3 2007 484 Pac h er, M., 2000: Tap h onomisc h e Untersuc h ungen der H h lenbren-Fundstellen in der Sc h wabenreit hH h le bei Lunz am See (Niedersterreic h).Beitr. Palont., 25, 11-85, Wien. Pac h er, M., Po h ar, V. & Rabeder, G. (Eds.), 2004: P otoka zijalka P alaeontological and archaeological re sults of the excavation campaigns 1997-2000 .Mitt. Komm. Quartrforsc h. sterr. Akad. Wiss., 13, 1245, Wien. Rabeder, G., 1992: Ontogenetisc h e Stadien des H h len bren aus dem Nixloc h bei Losenstein-Ternberg (O..).Mitt. Komm. Quartrforsc h., 8, 129-131, Wien. Rabeder, G., 2001: Gesc h lec h tsdimorp hismus und Kr pergrse bei h oc h alpinen H h lenbrenpopula tionen.Beitr. Palont., 26, 117-132, Wien. Rakovec, I., 1967: Jamski medved iz Mokrike jame v Sav injski h Alpa h. (e cave bear from t h e Mokrica cave in t h e Savinja Alps (Slovenia, Y ugoslavia).)Razpr. IV. razr. SAZU, 10, 4, 123-203, Ljubljana. Rogers, L. L., 1981: A bear in its lair.Natural History, 90, 64-70, New Y ork. Rogers, L. L., 1987: Eects of food supply and kins hip on social be h avior, movements, and population growt h of black bears in nort h eastern Minnesota.Wildl. Monogr., 97, 1-72. Sc h oen, J. W., Beier, L. R., Lentfer, J. W. & Jo hnson, L. J., 1987: Denning ecology of brown bears on Admi ralty and C hic h agof Islands.Bears eir Biology and Management, 7, 293-304. Sc h wartz, C. C., Miller, S. D. & Franzmann, A. W., 1987: Denning ecology of t hree black bear populations in Alaska.Bears eir Biology and Management, 7, 281-291. Stiner, M. C., 1990: e use of mortality patterns in ar c h aeological studies of h ominid predatory adapta tions.J. Ant hrop. Arc h aeol., 9, 305-351, Albuquer que. Stiner, M. C., 1994: H onor among ieves: A Z ooarchaeo logical Study of Neandertal Ecology. Princeton Uni versity Press, p. 422, Princeton. Stiner, M. C., 1998: Mortality analysis of Pleistocene bears and its paleoant hropological relevance.J. Human Evolution, 34, 303-326, New Y ork. Swenson, J. E., Franzen, R., Segerstrom, P. & Sandegren, F., 1998: On t h e age of self-surciency in Scandina vian brown bears.Acta er., 43, 2, 213-218, Bi alowieza. Tietje, W. D., Pelc h at, B. O. & Ru, R. L., 1986: Canni balism of denned black bears.J. Mammal., 67, 762766, Provo. Turner, E., 2002: e Cave Bears from t h e Wildsc h euer Cave in Hesse, Germany.Ab h. zur Karstund H h lenkunde, 34, 17-21, Mnc h en. Van Daele, L. J., Barnes, V. G. & Smit h, R. B., 1990: Den ning c h aracteristics of brown bears on Kodiak Is land, Alaska. Bears eir Biology and Manage ment, 8, 252-267. Weinstock, J., 1999: e Upper Pleistocene mamma lian fauna from t h e Groe Grotte near Blaubeuren (sout h western Germany).Stuttgarter Beitr. Naturk. Ser. B., 277, 1-49, Stuttgart. Weinstock, J., 2000: Cave Bears from Sout h ern Germany: Sex Ratios and Age Structure. A Contribution To wards a Better Understanding of t h e Palaeobiology of Ursus spelaeus .Arc h aeofauna, 9, 165-182. Weinstock, J., 2001: Age structure and sex ratio of cave bears in t h e Zoolit h en h h le, sout h ern Germany.Cadernos Lab. X eoloxico de Laxe, 26, 289-299, Corua. WIIG, O., 2000: Der Eisbr. In: U. GANSLOER (Ed.): Die Bren. Filander-Verl., 163-197, Frt h. I RENA D EBELJAK



PAGE 1

RESULTS OF PALEOSTOMATOLOGICAL ANAL Y SIS OF MATERIAL FROM THE CAVE NEAR GORNJI VAKUF SOUTHWEST BOSNIA I ZSLEDKI PALEOSTOMASTOLOKIH RAZISKAV GRADIVA IZ JAME PRI G ORNJEM V AKUFU J UGOZAHODNA B OSNA Amila Z UKANOVI 1 Jasminko M ULAOMEROVI 2 & Damir M ARJANOVI 3 Izvleek UDK 903.5:572.7(497.6) Amila Zukanovi, Jasminko Mulaomerovi & Damir Marjanovi: Izsledki paleostomastolokih raziskav gradiva iz jame pri Gornjem Vakufu (Jugozahodna Bosna) Med vsemi raziskanimi speleolokimi objekti v bliini vasi Kru pa, je s paleokulturnega vidika najbolj zanimiva Jama t. 5. V notranjosti jame je bilo najdeno, poleg odlomkov lonevine in kovinske igle, tudi bogato antropoloko gradivo, tevilne loveke kosti in zobje. Antropoloko-morfoloke preiskave dela tega gradiva kaejo, da gre za predzgodovinsko grobnico z ostanki najmanj 13 ljudi. Analiza kosti z radiokarbonsko analizo 14 C je pokazala, da so ti ljudje iveli 2765 75 let pred sedanjostjo. Dejstvo, da antropoloko gradivo vkljuuje tudi ostanke otrok, kae znailnosti pokopavanja iz tega asa na obmoju Uskoplja. To jamsko grobie je eno najveji h s tega obmoja. Paleostomatoloke preiskave zob in odlomkov eljustnic nam jasno dokazujejo zdravstvene posege iz tisti h asov. S pomojo arkov X je bil v kosti odlomka eljustnice odkrit kovinski de lec. Domnevno je ta kovinski delec odlomljeni del instrumenta, ki so ga uporabljali za ruvanje zob. Kljune besede: jamsko grobie, kostni ostanki, 14 C, paleostomatoloka preiskava, Uskoplje, Bosna. 1 Faculty of Dentistry, University of Sarajevo, 71.000 Sarajevo, Bolnika 4a, Bosnia and Herzegovina migulin@lsinter.net 2 Speleological society Speleo dodo, 71000 Sarajevo, Branilaca Sarajeva 30, Bosnia and Herzegovina, jasminko.mulaomerovic@ b h telecom.ba 3 Institute for Genetic Engineering and Biotec hnology, University of Sarajevo, Kemalbegova 10, 71.000 Sarajevo, Bosnia and Her zegovina, damir.marjanovic@ingeb.ba Received/Prejeto: 07.02.2007 COBISS: 1.01 ACTA CARSOLOGICA 36/3, 485-492, POSTOJNA 2007 Abstract UDK 903.5:572.7(497.6) Amila Zukanovi, Jasminko Mulaomerovi & Damir Marjanovi: Results of paleostomatological analysis of material from the cave near Gornji Vakuf (South-West Bosnia) Among explored speleological objects near t h e Krupa village, close to Gornji Vakuf, t h e most interesting cave from paleocul tural aspect, is t h e cave named Cave number 5. In t h e internal part of t h at cave, besides ceramic fragments and one metal nee dle, ric h ant hropological material h ave been found, compounds of numerous h umans bones and teet h. Anato-morp h ological analysis of t h e one part of material s h owed t h at t his site is a pre historic crypt wit h remains of at least 13 people. Radiocarbon 14 C bone analysiss results s h owed t h at t h ose people lived 2765 75 before presence. e fact t h at ant hropological material in cludes some remains w hic h belonged to c hildren s h ows some c h aracteristics of burial culture of t h at period on area of Us koplje. is cave crypt is t h e one of t h e biggest on t his region. Paleostomatological analysis of teet h and jaw fragments gave us clear evidence about t h erapeutic interventions performed in t his period. X-rays analysis s h owed presence of t h e metal body inside t h e bone in one mandibular fragment. Its supposed t h at t his metal is broken part of instrument used for toot h extrac tion. Key words: cave burial, bone remains, 14 C met h od, paleostoma tological analysis, Uskoplje, Bosnia.

PAGE 2

ACTA CARSOLOGICA 36/3 2007 486 Speleologist of Bosnia and Herzegovina organized t h eir traditional camp for year 2003 near Gornji Vakuf. Major ity of exploration h as been done in speleological objects around Bistrica and Krupa villages near Gornji Vakuf town (Figs. 1 and 2). As muc h as we know, speleological explorations of t his area h ave been performed only in few caves near Krupa village, in period 1971-1973, for water supplying needs for Gornji Vakuf, and eventual touristic arrangement (Gaparovi, 1979). Head of t his researc h was Ratimir Gaparovi, but his report of t his researc h or any ot h er maps of t h e objects are unknown. Frenc h en gineer Albert Bordeaux (1987) w h o passed t hroug h t his area about 1900 year, mentioned one cave and old mines in t his part of Vranica mountain. Some speleological explorations probably also h ave been performed at t h e end of 1950s, during abounding geological researc h es of Vranica mountain, but also t h eres no evidence and data about t h at exploration. Amir Topi from Gornji Vakuf started rst explorations and recognitions of area of t h e west slope of Vranica in last few years (Topi, 2000). ere are few t hings known about traces of h umans culture in caves of t his area. Arc h eological lexicon of Bos nia and Herzegovina doesnt mention any cave from t his area. Old arc h eological site on t h e hill above source of Krunica, wit h stone wall and ceramic fragments, its dat ed in early iron age (Marijan, 1988). Franjo Martinevi (1901) noticed story of old man Matko from Bistrica w hic h sad t h at some people from village enter in to t h e cave above source of Krunica, and t h at one of t h e visitors found old ceramic pot on t h e cave s h elf, w hic h s h ouldnt take out because h e was scared. A.Topi found ceramic fragments in few caves, w hic h are t h oug h t to h ave been in ceremonial proceedings. Because of t h at, during t h e speleological camp, special attention was given to t h ose caves but also to t h e caves w h ere we could expect similar founds. During eld recognition, single pieces of ceramic were found under Baba Janja cave and Velika Peina cave. We visited 8 caves wit h intention to collect datas about traces of h umans culture. In Velika Peina cave, Peina in Guser, caves number 4, 5, 6 and 7 (marked wit h numbers cause t h ey dont h ave names), Mala Peina cave and Cave t hree entrances (Mulaomerovi, 2004), it h as been found few or less fragments of ceramic w hic h s h owed continu ity of living from bronze age till middle age. INTRODUCTION Fig. 1: Geographic position of Gornji Vakuf on the map of Bosnia and H erzegovina. Fig. 2: Krupa village near Gornji Vakuf at the bottom of the rock creep is the spring of Krunica river; in marked part there are numerous caves entrances. Fig. 3: e last part of the chanel with bones covered with cave sediments. A MILA Z UKANOVI J ASMINKO M ULAOMEROVI & D AMIR M ARJANOVI

PAGE 3

ACTA CARSOLOGICA 36/3 2007 487 Cave number 5 is of course, considering founding till now, t h e most interesting cave, rst at all because of ndings of bigger number of h uman bones.(*) Along corridor inside t h e cave, also big number of ceramic frag ments is found, w hic h can be wide dated from iron till middle age. Also, in t h e middle part of passage, 10 centi meters dept h, one metal needle is found. Bones are found at t h e end of passage, at t h e space of 10-15 meters. Some of t h em are covered and xed wit h cave sediments on t h e slope w hic h is formed under one small c himney t hroug h w hic h water easily passes. (Fig. 3). On t h e rst sig h t its evident t h at all t h ose bones belong to t h e rests of several people. ose are mostly fragments of bones of t h e h umans trunk, skulls, upper and lower extremities and some permanent and decidual teet h. All bones and teet h are mixed all around t h e cave, probably because of t h e water activity. e biggest part of t h e bones is covered wit h cave sediments and by t h at xed for t h e underlying surface, w h at unable collecting all material for analysis. e only material w hic h could be taken for identication were bones and teet h placed in t h e water or in t h e clay, mostly near t h e caves walls and bones w hic h were just partially covered wit h sediments. Material is collected from 7 marked places. MATERIALS AND METHODS Found material presents fragments of skull bones (ossa cranii) one part of le temporal bone, fragment of rig h t temporal bone, two frontal bones (one fragment and one complete), rig h t and le pelvic bone, and one part of rig h t metacarpal bone (ossa membri inferioris), rig h t forearm bone and part of le clavicular bone (ossa mem bri superioris), parts of jaw bones (4 fragments of man dibulae, one complete mandibula and one fragment of rig h t maxilla), and 6 primary teet h (dentes decidui) and 30 permanent teet h (dentes permanentes) All material rst was mec h anically cleaned wit h wa ter and so brus h. Cave sediments from teet h were re moved mec h anically, using h and instruments for dental calculus scaling. One part of t h e bones w hic h were covered wit h cave sediments was cleaned mec h anically and rest was moved wit h acetyc acid suspension. Material was analyzed using following met h ods: Anatomicmorp h ological analysis of bones and teet h: For description of osteological ndings, interna tional anatomical nomenclature Terminologia Anatomi ca, from 1998. year h as been used. For morp h ological features of teet h crowns and roots, standards of Arizona State University Dental An t hropology System ASU Dental Ant hropology System (C hristy et al 1991) h ave been used. For bones and teet h analysis, 4 X magnifying lens h as been used. Radiocarbon 14 C analysis for bone absolute age datation: Age span is given in historical ages (AD/BC), wit h probability expressed in percentages, inside 1 mistake, and h as been determined based on dendroc hronological calibrating curve according to Stuiver & Reimer (1986) and Bronk & Ramsey (1995). Metrical and odontometric analysis of teet h and bone fragments: All measurements was performed wit h millimeter measurer (Dentalschiebelehre) produced by Dentaurum company. Measures of eac h bone and eac h toot h were tak en t h ree times, in intervals of one mont h aer t h e mean of t h ose individual measures was taken as nal result. All measurements were performed by t h e same person. X -Rays analysis: Jaws fragments and teet h are radiograp hicaly re corded and analyzed. Ortopantomograp h OP100 ma c hine produced by Instrumentarium company and Uni versal Oralix mac hine produced by Dentex company h as been used for recording, using Kodak T-MAT E (150x300 mm) and Kodak Insig h t Occlusal (57x76 mm) XRay lms. DNA toot h extraction: Optimized Qiagen extraction protocol was used for DNA extraction from toot h. DNA was determined and quantied by use of QuantiBlot assay (Applied Biosys tem, Foster City, CA, USA) (1996). e PowerPlex 16 kit (Promega Corp., Madison, WI) h as been used for am plication of targeted loci. Amplication was carried out as described previously (Promega Corporation, 2001). e total volume of eac h reaction was 25l. e PCR am plication h as been carried out in PE Gene Amp PCR System ermal Cycler (ABI, Foster City, CA) according to t h e manufacturers recommendations. Electrop h oresis of t h e amplication products was preformed on an ABI PRISM 377 genetic analyser (ABI, Foster City, CA), using 5% bis-acrilamide gel (Long Ranger Single Packs). Raw data h ave been compiled and analyzed using t h e acces sory soware: ABI PRISM Data Collection Soware and Gene Scan. Numerical allele designations of t h e proles were obtained by processing wit h Powertyper16 Macro. RESULTS OF PALEOSTOMATOLOGICAL ANAL Y SIS OF MATERIAL FROM THE CAVE NEAR GORNJI VAKUF ...

PAGE 4

ACTA CARSOLOGICA 36/3 2007 488 Material analysis s h owed presence of several peoples ndings. Results of anatomical and morp h ological analy sis of teet h and bones, s h owed t h at t h ere are remains of between 13 and 20 persons. W h ereas all material from t h e site wasnt analyzed, number of people w h ose re mains are found in t h e cave can easily be more t h an 20, w h at future researc h es will s h ow. Human remains from t h e site belong to persons of dif ferent age in moment of deat h Among t h ese persons, t h ere were also few persons of c h ildren age, w h at can be seen from ndings of upper and lower jaws wit h mixed denti tion, decidual teet h and ulna bone of c h ild 1-mont h old. Radiocarbon 14 C analysis of bone sample s h owed t h at people w h ose remains were found, lived 2765 before presence (today), w h at belongs to t h e period of early iron or late bronze age. Due to absence of t h e positive colorimetric detec tion, approximated amount of DNA recovered from t h e toot h was less t h an 0,15 ng/l. Consequently, 5 l of DNA extract was amplied by usage of prolonged PCR proto col. STR prole was obtained only at Amelogenin sex determining locus. is result suggested t h at female per son was carrier of t his biological specimen. One fragment of lower jaw requires special attention (Figs. 4 and 5). is fragment is from le side of man dibula. e fracture lines of t his fragment passed: in ad vance till canines alveolus (t hroug h w hic h also fracture line passes, back till ascendant branc h of ramus mandib ulae so only retromolar area wit h ascendant part of linea obliqua and linea milo h yoidea can be seen. Anterior and posterior edges of ramus mandibulae are preserved to 25 mm hig h. Down fracture goes t hroug h canalis mandibu laris. Lingual side of fragment is intact, but fracture line passed t hroug h biggest part of bucal side. Total lengt h of fragment is 57,8 mm. Empty alveolus of canine, rst and second mandib ular premolars are signs of postmortem teet h loss. eres a rst permanent molar present. Be hind it, t h eres long retromolar area. Alveolus of second and t hird molar (if it existed) cant be seen, because new bone is created on t h at place, aer t h ose teet h are lost during t h e lifer (ante mortem teet h loss). First molar alvelus and empty alveoli of rst and second premolars s h ows bone lost of vestibular crest of alveolus, but on ot h er sides of alveolus crests t h ere are no signs of any bone loss. Bone loss values are: First molar: Vestibular alveolar crest of distal root alveolus till cementoenamel junction (CEJ): 2.6 mm Vestibular alveolar crest of mesial root alveolus till cementoenamel junction (CEJ): 3.5 mm Second premolar alveolus: Vestibular alveolar crest is 2.6 mm lower t h an ot h er alveolar crests, considering t h at complete alveolar edge is lower positioned according alveolus of rst premolar. First premolar alveolus: Vestibular alveolar crest is 4.8 mm lower t h an ot h er alveolar edges. Vestibular edges of empty alveolus of caninus, rst and second premolars and rst molar h ave bone de hiscence. Interproximal bone between canine alveolus and rst premolar are well preserved, regular structure and h eig h t but interproximal alveolar bone between second premolar and rst molar is reduced h eig h t, wit h regular structure of supercial bone lost. RESULTS Fig. 4: One fragment of the mandibular bone with rst permanent molar (external view). Fig. 5: One fragment of the mandibular bone with rst permanent molar (internal view). A MILA Z UKANOVI J ASMINKO M ULAOMEROVI & D AMIR M ARJANOVI

PAGE 5

ACTA CARSOLOGICA 36/3 2007 489 Bone around second premolar is roug h, ragged wit h many perforations. First le mandibular molar teet h description (36): e maximum mesiodistal crown lengt h 10,95 mm, maximum buccolingual breadt h ,15 mm, maxi mum crown h eig h t 6,9 mm, crown area 111,14 mm. ose data s h owed t h at t his toot h is smaller in all dimensions of average dimensions of t his toot h, w h at can be found in tables of anatomical teet h measures w hic h Konjh odi (1978) measured for recent Y ugoslav popu lation. Morp h ological features of toot h 36 according to ASU: Radical number: 2 Anterior fovea: is sign cant be determined due to present abrasion of t h ese teet h. is sign its not recom mended for classication in persons older t h an 12 years. is toot h belonged to person more t h an 14 years old so from t h at reason t his sign in t his case is not determined. Groove pattern: Even if toot h h as hig h level of abra sion, t his sign can be determined. But in our case, be cause of abrasion, we cant be sure w hic h type of occlusal morp h ology t his toot h h as. Cusp number: Its not possible to determine cusp number because of abrasion, only two lingual cusps are visible on abraded oclusal surface. Deecting wrinkle: is toot h belonged to t h e per son w hic h in moment of deat h was more t h an 14 years old. From t h at reason, and also because of t h e abrasion, t his sign is dircult to notice. Distal trigonid crest: Cant be determined w h en abrasion of rst grade is present t his molar toot h h ave second grade of abrasion. Protostylid: 0 Cusp 5 /H ipoconulid/, Cusp 6 /Entoconulid/, Cusp 7 /M etaconulid/: Cant be determined due to abrasion. Tooth wear: 2 eres wide toot h wear on occlusal surface in orovestibular direction. Abrasion erased vestibular cusps completely, also ssure system vestibular to central s sure, so on t h e place of t his morp h ological features le only slope wit h lig h t declivity to central ssure. Distoap proximal ridge is totally abraded. Mesioapproximal ridge h as full h eig h t only in lingual part. Lingual cusps are also abraded in occlusal area, specially t h eir external slopes, DISCUSSION Preliminary results of DNA analysis open slig h t possibil ity for additional molecular examination of discovered skeletal remains. However, before t h at, optimization of existing DNA isolation and amplication protocols will be required. but ssure w hic h separate mezial from distal lingual cusp is present. On mesioapproximal side of t h e toot h, t h eres contact surface on t h e place of t h e contact wit h second premolar during t h e lifetime. On t h e distoapproximal side, t h eres no abrasion on t h e place w h ere contact point or surface wit h second molar s h ould be present. On ECJ, on mezioapproximal surface of t h e toot h initial carious lesion is visible, w h at also X-rays analysis conrmed (Fig. 6). On t h e top of lingual cusps, t h ere is w hite enamel opalescence. On upper t hird of toot h crown, on central and distal part of vestibular side and, t h ere is enamel de fect. at defect h as s h arp border and rm bottom, w h at lead us to conclusion t h at t h at defect appeared as con sequence of t h e mec h anic force, and it can be described as non complicated dental fracture on enamel ec h elon (fractura dentis simplex) X-ray analysis conrms t h at second molar h as been lost during t h e lifetime (ante mortem). On t h e places w h ere alveolus of second and t hird molars s h ould be found, t h eres bone wit h normal trabecular structure. On occlusal dental x-ray lm, w h ere all mandibular bone fragments wit h rst molar can be seen, on t h e place w hic h belongs to bottom part of second molar distal root alveolus, clear s h ape of metal body is visible. Fragment analysis wit h magnifying glass s h owed t h at metal body is not on surface of t h e bone, but its placed deep in t h e bone structure. Its supposed t h at t his metal is broken part of instrument for toot h extraction used for second molar extraction. Fig. 6: X-ray of fragment of the mandibular bone with rst permanent molar. RESULTS OF PALEOSTOMATOLOGICAL ANAL Y SIS OF MATERIAL FROM THE CAVE NEAR GORNJI VAKUF ...

PAGE 6

ACTA CARSOLOGICA 36/3 2007 490 Review of paleostomatological literature from t h e former Y ugoslavia region didnt give any results w hic h designate on t h erapeutic intervention like toot h extrac tion. Miki (1981) gives catalogue data review of former Y ugoslavia area, of all pre historical periods, and men tioned t h at only on two arc h eological sites from iron age (Glasinac in Bosnia and evelija in Macedonia) are found jaws w h ere is evidenced ante mortem toot h loss (intra vitam).e same aut h or also gives interesting data, t h at on t h e bot h sites t h e ot h er teet h le h ave very few carious lesions or dont h ave it at all. at suggests t h at t h e reason w h y t h ose teet h are extracted is not due to t h e caries w hic h progrediated in painful diseases w hic h leads to t h at toot h extraction. is fact is in accordance also wit h results from our study on all 36 teet h t h eres no deep carious lesions. On just few teet h initial carious le sions was evidenced. Possible reasons for second molar extraction on t h e fragment of t h e mandibular bone can be trauma, par odontological disease or caries. Trauma: On t h e upper parts of buccal and distoap proximal side of crown of rst mandibular molar t h eres enamel defect w hic h appeared due to trauma. Its possible t h at t his crown damage appeared at t h e same time w h en trauma aected second mandibular molar, w h at caused complications followed by pain and consequent toot h extraction. But its also possible t h at enamel defect on rst molar appeared during extraction of second molar w h ere bad extraction tec hnique or loosing control un der instrument for toot h extraction leaded to rst mo lar damage. Metal body inside alveolus of second molar supports t his t h eory. is piece of metal is probably part of instrument for toot h extraction w hic h small part frac tured during toot h extraction and stayed inside postex straction wound. Parodontological reasons If t his toot h was extracted because of t h e parodontal disease, parodontal abscesses w hic h caused t h e pain were t h e most probable reason for toot h extraction. Extraction because hig h level of toot h mobility (w hic h appears in parodontal diseases) is not possible reason for extraction because canine alveolus h as hig h degree of bone resorp tion comparing wit h bone resorption of rst molar, and bot h of t h em didnt lose ante mortem. Its important to mention t h at alveolar bone resorption w hic h is visible especially on t h e vestibular part, is not necessary con sequence of parodontal disease. Its more probable t h at alveolar crest p h ysiologically migrated apical because of t h e continuous toot h eruption and abrasion. If we know t h at abrasion degree of t his toot h is 2, according to ASU system, we can easily suppose t h at t h ere was abrasion also on ot h er teet h. C h anges of cementoenamel junction al veolar crest (CEJ-AC) distance are also connected wit h continuous toot h eruption, abrasion and facial growt h, not only wit h pat h ological factors w hic h increase t h e CEJ-AC distance. Caries: Its not so possible t h at second molar extrac tion h appened because of pulpal or periodontal diseases w hic h occurred as consequence of caries progression. First molar appears in t h e mout h rst of all permanent teet h and usually its rst aected wit h caries. is rst molar on t h e bone fragment h ave just initial caries le sion on t h e cervical part, and t h eres no reason to believe t h at second molar (w hic h appears in t h e mout h 6 years aer rst one), developed caries lesion w hic h could lead to complications w hic h s h ould be t h e reason for toot h extraction. Even more, none of t h e teet h w hic h h ave been found on t his location h ad any deep carious lesions, only initial caries lesions h ave been found. No matter w hic h is t h e reason for t his toot h extrac tion, its obvious t h at its extracted wit h some instrument w h at points to t h e presence and performance of t h e t h er apeutic interventions in t his historical period. Literature review s h owed also some evidence of stomatologycal in terventions in iron age period on area of former Y ugosla via Miki described one mandibula found on Glasinac site, w h ere clear evidence of mec h anical scraping is pres ent on t h e place of second premolar and rst molar, w h ere scraping went to bottom of alveolus, w h ere also one part of external alveolar bone is moved. e aut h or supposes t h at t his is t h e trace of surgical intervention of sanation of toot h gangrene or periodontitis (Miki, 1973). Literature didnt give any data about similar found ing. Lovrinevi & Miki (1989) made radiological and p h oto documentation of osteological material wit h pat h ological c h anges on historical populations of Y ougo slavia, but t h eres no evidence of foreign object located intraossealy. e reason w h y t his toot h is extracted is unknown. e bone w hic h is present on t h e place of extracted sec ond molar h as regular structure, like t h e bone structure around, wit h out any pat h ological c h anges. It means t h at moment of deat h of t h at person h appened at least six mont h aer toot h extraction, because its known t h at p h ysiological process of bone regeneration w hic h can be radiological veried takes 8 weeks. Aer period of at least 6 mot h s, t h eres no visible dierence in toot h bone structure between new formed bone on t h e place of ex tracted toot h and old bone. On our radiological lm, t h eres no dierence between t h e bone on t h e place of extracted second molar and t h e rest bone. It means t h at new formed bone is older t h an 6 mont h, e.g. moment of deat h of t his person h appened at least six mont h s aer toot h extraction. Analysis of t his mandibular fragment s h owed t h at t his person h ad more t h an 14 years in t h e moment of A MILA Z UKANOVI J ASMINKO M ULAOMEROVI & D AMIR M ARJANOVI

PAGE 7

ACTA CARSOLOGICA 36/3 2007 491 CONCLUSION Researc h es performed in speleological objects near Bis trica and Krupa villages near Gornji Vakuf s h owed pres ence of h uman remains traces. e biggest attention of explorers engaged t h e cave named Cave number 5. In t h e internal part of t h at cave, besides ceramic fragments and one metal needle, ric h ant hropological material h ave been found, compound of numerous h umans bones and teet h. Anato-morp h ological analysis of t h e one part of material s h owed t h at t his site is pre historical crypt, w h ere are remains of at least 13 people. Radiocarbon 14C bone analysiss results s h owed t h at t h ose people lived 2765 before presence. e fact t h at ant hropological material includes some remains w hic h belonged to c hil dren, points some c h aracteristics of burial culture of t h at period on area of Uskoplje. is cave crypt is t h e one of t h e biggest on t his region. Paleostomatological analysis of teet h and jaw frag ments gave us clear evidence about t h erapeutic interven tions performed in t his period. X-rays analysis s h owed presence of t h e metal body inside t h e bone in one man dibular fragment. Its supposed t h at t his metal is broken part of instrument used for toot h extraction. It can also be sad t h at, based on t h ose modest re searc h es, t h at caves in area of Bistrica and Krupa villages, t h e traces of h uman abode in dierent history periods are found, and some of t h ose ndings very clearly and directly point on t h eir ritual usage. deat h. Second molar wasnt in function for a long time before its extracted w h at can be concluded because t h eres no contact surface on t h e distoaproximal side of rst molar. is person h ad rst and second premolars, rst and second molars, but t hird molar never existed or was also extracted, w h at is less possible. Special meaning of t his site location is t h at cave burial is c h aracteristic of earlier historical periods. For bronze and iron age its typical skeletal burial in tumulus and at necropolis as muc h as cremation. Burial in t h e cave is very rare in t his period. e evidence of cave burial is present on kocjanske Jame area, in Tomineva jama cave (Leben, 1975), w h ere in 1903 20 h allstate skeletons h ave been discovered and excavated. On t h at place also few c hild skeletons is found. Tombs also h ad been found in Ajdovska jama cave near Nemka Vas, and in t h e area of Kras, sout h west Alps and Panonic area as well (Koroec, 1981-1982; Leben, 1970; Gutin, 1977; Dombay, 1960; Ruttkay, 1970). Cave burial evidence is also found in t h e cave Bezdanjaa near Vr h ovina in Lika. at site is dated in middle and late bronze age based on ceramic and met al founding (Drec h ler-Bii, 1981). On t h at site several dozen skeletons h ad been found, w hic h ant hropological analysis made Pilari G. No matter t h at t his necropolis h as been used for a long time, on t h at location any c hil dren skeletons never found. In t h e cave site near Gornji Vakuf, its evidenced t h at few skeletal remains certainly belong to c hildren, w h at points on dierent cultures of burial. is data opens many questions for w hic h answers will give future researc h es on mountain Vranica area. A CKNOWLEDGEMENT In t h e exploration in July 2003 participated: Am ila Zukanovi, Tarik and Emir Dizdarevi, Davor Dautbegovi, Ivica Bonjak, Mirjana Palavra, Ajna and Jasminko Mulaomerovi. In t h e exploration in No vember 2003 participated: Il h an Dervovi, Jasminko Mulaomerovi, Mustafa Duratbegovi, Amir Topi, Amila Zukanovi, Tarik Dizdarevi, Edis Duvnjak i Samir Pokvi. RESULTS OF PALEOSTOMATOLOGICAL ANAL Y SIS OF MATERIAL FROM THE CAVE NEAR GORNJI VAKUF ...

PAGE 8

ACTA CARSOLOGICA 36/3 2007 492 Bronk, R.C., 1978: Radiocarbon calibration and analysis of stratigrap h y: t h e OxCal program. Radiocarbon 37: 425-430. Bordeaux, A., 1987: Narodna Bosna. In: Fojnica kroz vijekove.Skuptina optine Fojnica & SOUR Vese lin Maslea, Fojnica Sarajevo. (According to issue of PLON, Paris, 1904.). C hristy, G., Turner, I. I., C hristian, R., Nic h ol, G. & R ich ard, S., 1991: Scoring procedures for key morp h ological traits of t h e permanent dentition e Ari sona State University Dental Ant hropology System. In: Kelley M. A., Larsen C. S. (eds). Advances in Dental Ant hropology. New Y ork:Wiley-Liss, Inc., pp.13-31 Dombay, J., 1960: Die Siedlung und das Graberfeld in Zengovarkony : Beitraege zur Kultur des Aeneolit hikums in Ungarn. Arc h aeologica Hungarica Dissertationes Arc h aeologicae Musei Nationalis Hun garici, s.n 37, Budapest: Verlag der Ungarisc h en Akademie der Wissensc h aen, pp. 235. Drec h ler-Bii, R., 1981: Nekropola bronanog doba u peini Bezdanjai kod Vr h ovina. Vjesnik ar h eolokog muzeja u Zagrebu XII-XIII/1979-1980: 27-78 + I-XLVI. Gutin, M., 1977: Poroilo o izkopu kulturni h slojev v Levakovi jami. Ar h eoloki vesnik 27: 260-282. Konjh odi-Rai, H. (1978): Prosjene anatomske mjere zuba jugoslovenske populacije i postojanje seksualni h razlika u veliini zuba.Univerzitet u Sa rajevu, Sarajevo, (doktorska disertacija). Koroec, P., 1981-1982: Kult mrtvi h v lui najnoveji h ra ziskav v Ajdovski jami pri Nemki vasi. Poroilo o raziskovanju paleolita, neolita in eneolita v Sloveniji 9-10: 15-20. Leben, F., 1970: Znailnosti in pomen nekateri h ar h eoloki h jamski h najdi na podroju jugovzh od ni h Alp. In: Adriatica prae historica et Antiqua : Zbornik radova posveen Grgi Novaku, XXXIV, pp. 739, Zagreb: Sveuilite u Zagrebu : Ar h eoloki in stitut Filozofskog fakulteta. Leben, F., 1975: kocjanske jame (kocjan), Tomineva jama. In: Ar h eoloka najdia Slovenije, p. 132133, SAZU & DZS, Ljubljana,. Lovrinevi, A., Miki, ., 1989: Atlas osteopatoloki h promjena na istorijskim populacijama Jugoslavije. Svjetlost, Sarajevo. Marijan, B., 1988: Gradina, Krupa, Gornji Vakuf In: Ar h eoloki leksikon Bosne i Hercegovine, T. 2., Ze maljski muzej BiH, Sarajevo. Martinevi, F., 1901: Narodna prianja iz Bosne. Glasnik Zemaljskog muzeja, XIII: 468-471. Miki, (1981): Stanje i problemi zike antropologije u Jugoslaviji (praistorijski periodi). -Akademija nauka i umjetnosti Bosne i Hercegovine, Sarajevo. Mulaomerovi, J., 2004: Tragovi kultura u peinama oko lice Gornjeg Vakufa u Bosni. Speleozin, XII, 17: 60-63. Promega Corporation, 2001: Promega Corporation PowerPlex 16 system Tec hnical Manual. Promega Corporation, Madison (WI). (1996): QuantiBlot Human Quantication Kit Users Manual. e Perkin-Elmer Corporation, Foster City (CA). Ruttkay, E., 1970: Das jungsteinzeitlic h e Hornsteinberg werk mit Bestattung von der Antons h h e bei Mauer (Wien 23). Die Ausgrabungen Josef Bayers in den Ja hren 1929-1930. Mitteilungen der Ant hropolo gisc h en Gesellsc h a Wien 100: 70-115. Stuiver, M. & Reimer, P. J., 1986: A computer program for radiocarbon age calculation. Radiocarbon 28(2B): 1022-1030. Topi, A., 2000: Speleologija u Gornjem Vakufu (I). Speleobi h V, 5: 20. Zukanovi, A., 2004: Preliminarni izvjetaj o oseo-den talnom h umanom materijalu iz peine iznad vrela Krunice kod Gornjeg Vakufa.Na kr, VII (in print). REFERENCES A MILA Z UKANOVI J ASMINKO M ULAOMEROVI & D AMIR M ARJANOVI



PAGE 1

BIODIVERSIT Y AND ECOLOG Y OF FAUNA IN PERCOLATING WATER IN SELECTED SLOVENIAN AND ROMANIAN CAVES BIODIVERZITETA IN EKOLOGIJA FAVNE V PRENIKLI VODI IZBRANIH SLOVENSKIH IN ROMUNSKIH JAM Oana T. MOLDOVAN 1 Tanja PIPAN 2 Sanda IEPURE 1 Andrej MIHEVC 2 & Janez MULEC 2 1 Romanian Academy, Institutul de Speologie Emil Racovitza, Department of Cluj, Clinicilor 5, 400006 Cluj-Napoca, Romania, E-mail: oanamol@h asdeu.ubbcluj.ro 2 Karst Researc h Institute, Scientic Researc h Centre of t h e Slovenian Academy of Sciences and Arts, Titov trg 2, p.p. 59, SI-6230, Postojna, Slovenia, E-mail: pipan@zrc-sazu.si Received/Prejeto: 08.10.2007 COBISS: 1.01 ACTA CARSOLOGICA 36/3, 493, POSTOJNA 2007 Abstract UDC 574.5:551.44(497.4+498) Oana T. Moldovan, Tanja Pipan, Sanda Iepure, Andrej Mi hevc & Janez Mulec: Biodiversity and ecology of fauna in per colating water in selected Slovenian and Romanian caves Biodiversity and ecology of fauna in percolating water from Slo venian and Romanian caves was studied. Researc h focused on unravelling t h e community structure of epikarst fauna, w hic h is carried away by t h e trickles of percolating water from t h e epikarst and vadose zones. e major part of t h e fauna found in percolating water is represented by copepods. is fauna, originating in t h e epikarst, was analysed and by means of t h e systematic sampling and observation t h e same groups of ani mals were found in Slovenian and Romanian caves. Dierences among caves and sampling points indicate t h at epikarst is a h et erogenous h abitat. Relations hip between faunal ric hness and t h e p h ysical c h aracteristics of t h e water was found. Correlation between surface geomorp h ology and fauna in percolating wa ter was statistically signicant in t h e Postojna cave system. Key words: karst, epikarst, percolating water, speleobiology, fauna, Copepoda. Izvleek UDK 574.5:551.44(497.4+498) Oana T. Moldovan, Tanja Pipan, Sanda Iepure, Andrej Mi hevc & Janez Mulec: Biodiverziteta in ekologija favne v pre nikli vodi izbranih slovenskih in romunskih jam V prenikli vodi iz slovenski h in romunski h jam smo prouevali biodiverziteto in ekologijo favne. V raziskavi smo se osredotoili na prouevanje zdrube epikrake favne, ki jo odnaajo curki prenikle vode iz epikrake in nezasiene cone. Najveji dele organizmov v prenikli vodi zavzemajo cepononi raki. Na pod lagi sistematinega vzorenja in opravljeni h analiz smo v jama h slovenskega in romunskega krasa ugotovili prisotnost enaki h skupin organizmov. Razlike med jamami in posameznimi vzornimi mesti nakazujejo, da je epikras h eterogeni h abitat. Ugotovili smo statistino znailno povezavo med prisotnostjo doloeni h skupin organizmov in zikalnimi parametri vode. Korelacija med povrinsko geomorfologijo Postojnskega jam skega sistema in tevilom cepononi h rakov v prenikli vodi je bila statistino znailna. Kljune besede: kras, epikras, prenikajoa voda, speleobiologi ja, favna, Copepoda. I NTRODUCTION e upper layer of t h e unsaturated zone called epikarst is t h e interface zone between soil and rock in karst land scape and is c h aracterized by small fractures and solu tion pockets t h at may or may not be lled wit h water (Williams 1983). It is an important zone for t h e trans port of pollutants, an important water storage site (Petri 2002, Trek 2003), and a h abitat of hig h diversity of in vertebrates (Pipan 2005). ere is a complex and diverse aquatic community of microinvertebrates in t h e epikarst and vadose zone, including a ric h crustacean fauna, es pecially copepods (Pipan 2005). By contrast wit h many subsurface h abitats wit h large voids, t h e epikarst h abitat

PAGE 2

ACTA CARSOLOGICA 36/3 2007 494 is poorly integrated; catc hment areas extend only a h un dred or so meters (Pipan & Culver 2007). In t he last t h ree years t h roug h t he joint researc h project between Slovenia and Romania our basic re searc h focused on study of biodiversity and ecology of fauna in trickles of percolation water. Researc h was done in selected caves in Slovenian Classical Karst and Romanian Western Carpat h ians. It is wort h mentioning t h at bot h researc h groups beneted from t he interesting results t h at came from t he collaboration. e new ap proac h of investigation of cave-dwelling fauna broug h t also a new and improved understanding of t he biodiver sity of t he groundwater and opened new horizon of t he investigation w h ic h will be upgraded in t he future. Due to inaccessibility of t he unsaturated zone it is impos sible to examine directly t he biological processes wit h in t he epikarst and vadose zone. On t he ot her h and it is relatively easy to study t hem indirectly by taking sam ples of water w h ic h drips down t he cave walls or seeps from t he cave ceiling. Trickles represent t he transition between two h abitats: t he inaccessible epikarst toget her wit h vadose zone and pools of water occurring in cave passages. e rst investigations on fauna of ssures were done by Rouc h (1968), Delay (1968), and many ot h er aut h ors as mentioned in Sket et al. (2003), but t h e rst large scale study of fauna from epikarst and vadose zone started in 2000 in Slovenia. By t h e end of year 2005, 51 copepod species and subspecies were recorded in six caves; out of 32 are stygobionts (Pipan 2005). Many of t h ese species are endemic to Slovenia and frequently even to a single site. Study of t h e fauna in epikarst water was later done in a cave system in West Virginia, U.S.A. (Pipan & Cul ver 2005, Pipan et al. 2006) and in Spain (Camac h o et al. 2006). In Romania a similar study started in 2005 wit h t h e investigation of fauna in percolating water in caves of t h e Western Carpat hians. Due to t h e numerous newly discovered species, and little knowledge about t h e biodiversity and ecology of fauna from percolating water we continued wit h sam pling and investigations. In t his paper we report on t h e results of fauna from percolation water collected in two Slovenian and t hree Romanian caves. Copepod commu nity composition, population dynamics of fauna in per colating water and its relation to t h e surface geomorp h ol ogy and c h emistry of water are discussed. MATERIAL AND METHODS A one-year eld study was conducted in two caves in Slovenia Planinska jama and Postojnska jama and in t hree caves in Romania Petera Ungurului, Petera Vntului and Petera Vadu-Criului (Fig. 1). Postojna and Planina caves are part of t h e Posto jna-Planina cave system w hic h consists of approximately 23 km of surveyed passages (17 in Postojna cave and 6 in Planina cave connected by 2 km of ooded passage), developed in Upper Cretaceous carbonate rocks (ebela 1995). e cave system is t h e most diverse cave in t h e world (Culver & Sket 2000). e Pivka River sinks near t h e Postojna cave entrance, and ows towards t h e Plani na cave w h ere joins wit h t h e ot h er stream Rak and exits at Planina cave entrance. e land over t h e cave is for ested, and t h e Pivka River drains land wit h a variety of uses, including forest and agriculture, as well as several small towns. e area over t h e cave system itself is ap proximately 20 km 2 Samples were taken from 5 drips in Planina cave and from 10 drips in Postojna cave (see Pipan [2005] for locations in Postojna cave). In Planina cave 5 sampled drips were at a distance of 18 m between t h e rst and t h e last drip, w hile in Postojna cave t h e rst t hree sampled drips were wit hin 125 m and t h e distance between t h e t hird and t h e tent h drip was 1 km. In Posto jna cave sampling was done mont h ly from April to Oc tober 2000, and in Planina cave from April 2004 to De cember 2005. Fig. 1: Geographical location of studied caves (black dots on the right) in Slovenia and Romania. Ungurului, Vntului and Vadu-Criului caves belong to t h e Criul Repede basin, in nort h-western Romania. Ungurului and Vntului caves are developed in Ladinian (middle Triassic) limestone, w hile Vadu-Criului cave is in Barremian (Cretaceous) limestone. e t hree caves are near to eac h ot h er, covered by soil and vegetation repre O ANA T. MOLDOVAN, T ANJA PIPAN, S ANDA IEPURE, A NDREJ MIHEVC & JANEZ MULEC

PAGE 3

ACTA CARSOLOGICA 36/3 2007 495 RESULTS e groups of fauna found in t h e studied caves are pre sented in Table 1, and t h e identied species in Table 2. ese groups are not in a taxonomic hierarc h y equal, but t h ey are common for many subterranean h abitats. In all ve caves we found 12 groups of invertebrates (Table 1). is is remarkable level of biodiversity in un derground h abitats, and can be explained wit h t h e great diversity of micro h abitats in t h e unsaturated layers. Un gurului cave in Romania was t h e ric h est in groups of taxa (11) and individuals (355) among t h e all sampled caves and also seasonally. Planina cave (Fig. 2 a) was t h e second ric h est cave wit h 10 groups of animals (Table 1). A total of 345 invertebrates were found in drips. Except larvae of Insecta t h at were presumably was h ed from t h e surface, Harpacticoida was t h e most abounded group, represent ing more t h an 30% of t h e total number of individuals. In Postojna cave (Fig. 2 b) 8 animal groups were found, including terrestrial species of Acarina, Collembola and Insect larvae (Diptera). Among aquatic species stygobi onts of Copepoda prevailed (Tables 1 and 2), followed by Oligoc h aeta, Turbellaria and Nematoda. For Romanian samples (Fig. 2 c-e) t h e results were dierent among sampling stations and among caves. In Ungurului cave 11 groups of animals were identied: Nematoda, Planaria, Gastropoda, Oligoc h aeta, Araneae, Acarina, Cyclopoida, Harpacticoida, Amp hipoda, Col lembola, Insect larvae. Of t h e total of 355 individuals most of t h em were h arpacticoids (145) and oligoc h ets (70). Six fauna groups h ave representatives in Vntului samples: Nematoda, Cyclopoida, Harpacticoida, Acarina, Collembola, Insect larvae. Best represented, in one sta tion and one sampling were nematodes. In Vadu-Criului cave representatives of 5 fauna groups were identied: Gastropoda, Cyclopoida, Harpacticoida, Amp hipoda, Collembola, poorly represented. Most of t h e individuals were h arpacticoids. e same fauna groups are represented in percola tion water of bot h countries, except Ostracoda, w hic h sented by deciduous forests and pastures. e area was intensively exploited for reclay and accidental leakage can h appen in t h e rst two caves. Vntului is t h e longest cave in Romania (about 50 km) and t h e two ot h ers are among t h e best known s h ow caves of Transylvania. Sam pling stations were establis h ed near t h e entrance, on t h e active passage in t h e Vntului cave and along t h eir entire lengt h s for t h e Ungurului and t h e Vadu-Criului caves. All t hree caves h ave subterranean water courses, more important in Vntului and Vadu-Criului. Sampling points were establis h ed w h ere percolation is permanent or quasi-permanent along t h e year and collecting of ac cumulated material was done according to t h e activity of trickles, and accessibility in one case (Vntului cave). In Ungurului cave, wit h a development of approximate 1.5 km and a large entrance (32 / 22 m), samples were taken from 4 sites, during t h e period Marc h 2006 June 2007. In Vntului cave samples were taken at t h e base level, near t h e entrance in 2 sites (April 2005 May 2006) and in Vadu-Criului cave percolation water was sampled in 3 sites along t h e main gallery of 1 km, in 2007 (January June). Sampling of percolation water fauna in bot h coun tries was done in t h e same way. Drips were collected di rectly and simple sampling devices developed by Pipan (2003, 2005) were used to collect water and invertebrates. Dripping water was directed into a plastic container and ltered t hroug h a net mes h size 60 m. Drip rates were measured only in Slovenian caves by collecting water under drips in a graduated cylinder for timed intervals. Samples were sorted in a laboratory using a dissecting microscope and only Copepoda specimens were identi ed to species from Postojna cave and Cyclopoida species from Romanian caves. Temperature (C), conductivity (Scm -1 ) and pH were measured in situ by a conductiv ity meter (LF 91, WTW), pH meter (323, WTW), and Combo multi-parameter Hanna Instruments. Canonical correpondence analysis was used to cor relate data on abundance of dierent fauna groups in ve studied caves and measured p h ysico-c h emical param eters, as was used in t h e previous studies (Pipan 2003, 2005). Analysis was run by t h e program XLSTAT 2007. Spearman correlation test was performed to determine t h e relations hip between copepod abundance versus drip rate in Slovenian caves. To establis h if a relations hip between t h e density of t h e dolinas above t h e Postojna cave system and fauna abundance in percolation water exists, t h e outer edge of dolinas was dened according to t h e 1:5000 maps and personal observations of one of us (A. M.). Six frames of 0.12 km 2 were virtually dened on t h e surface projec tion of t h e Postojna system. e total surface of dolinas in eac h frame was digitized and measured wit h Didger 3.05 soware. Except one of t h e frames all ot h ers in cluded sampling stations for percolating water. e sum of dolinas surface and t h e fauna abundance of t h e cor responding sampling station(s) of eac h frame were t h an correlated. BIODIVERSIT Y AND ECOLOG Y OF FAUNA IN PERCOLATING WATER IN SELECTED SLOVENIAN AND ROMANIAN CAVES

PAGE 4

ACTA CARSOLOGICA 36/3 2007 496 was identied in one Romanian cave (Ungurului), in a single sample. A rat h er rare nding in t his h abitat is one amp hipod stygobiont in Vadu-Criului cave ( Niphargus sp.) and in Planina cave. For Planina cave it is known t h at an endemic stygobiont Niphargobates orophobata (Sket 1981) was found in dripping water. Alt h oug h it is not clear yet w h at causes t h e dierences among drips in fau na composition, t his suggests t h at t h e unsaturated zone can represent a unique ecological nic h e. Non-stygobiotic fauna may accidentally reac h subterranean h abitats and oen cannot survive t h ere for a long time, but may be an important source of nutrients in a cave. Results on subterranean aquatic fauna from Ungu rului and Vntului caves are publis h ed for t h e rst time, alt h oug h studies were done in t h e last decades (Plea & Iepure unpublis h ed). e groups of animals we found in percolating water are generally adapted to t h e interstitial h abitats, already known from ot h er subterranean h abitats in some of t h e studied caves (Plea 1969, Plea & Racovi 1973, Plea et al. 1996). Two strictly stygobiotic cyclopoid species h ave been found in trickles in t hree Romanian caves: Speocyclops troglodytes (C h appuis 1923) in Vadu Criului and Vntului caves and Acanthocyclops reductus (C h appuis 1925) in Ungurului cave. In general, trickles provide a lower number of individuals t h an t h e associ ated pools, of bot h cyclopoids and ot h er invertebrates. Wit h t h e exception of Speocyclops troglodytes from Vn tului cave sampled in trickles near t h e entrance, t h e spe cies of t h e ot h er two caves were reported in t h e sampled drips located between 300 500 m away from t h e en trance. Bot h mentioned species are rare in t h e karst area of t h e Western Carpat hians, being recorded until now from pools and deep p hreatic aquifers (C h appuis 1925, Plesa 1965, Plesa et al. 1996). Alt h oug h, Speocyclops trog lodytes is a cosmopolitan stygobiont widespread in vari ous groundwater h abitats t hrog h out Europe, in Romania t his species h as been recorded by now from only t hree caves of Western Carp h atians. A. reductus h as been re ported from ot h er two caves in Western Carpat hians. Table 2: e identied species of Copepoda Cyclopoida (C) and H arpacticoida (H) in the studied caves Cave Species Postojna Speocyclops infernus (C) Parastenocaris n. sp. (H) Bryocamptus balcanicus (H) Moraria poppei (H) Nitocrella n. sp. (H) Ungurului Acanthocyclops reductus (C) Vntului Speocyclops troglodytes (C) Vadu-Criului Speocyclops troglodytes (C) On t h e ot h er h and, some dierences were observed for eac h of t h e studied area, among caves and among stations in eac h cave, in terms of dierent copepod spe cies and of dierent fauna assemblages. Seasonal or even mont h ly dierences were also observed. Alt h oug h in eac h country t h e studied caves are very close to eac h ot h er, and no apparent dierences at t h e surface geomor p h ology can be observed, dierences in fauna assem blages indicate t h e h eterogeneity of t h e s h allow subter ranean h abitats (including epikarst), already mentioned for ot h er subterranean h abitats. Table 1: List of taxa and abundance of individuals collected in ve caves in Slovenia and Romania Animal group/Cave Slovenia Romania Total (%) Postojna Planina Ungurului Vntului Vadu-Criului Turbellaria 3 19 2.8 Nematoda 2 48 6 25 10.3 Gastropoda 8 5 1 1.8 Oligochaeta 4 28 70 12.9 Araneae 1 1 0.3 Acarina 2 7 18 2 3.7 Ostracoda 1 0.1 Cyclopoida 6 16 28 7 2 7.5 Harpacticoida 5 115 145 1 7 34.6 Amphipoda 1 1 0.3 Collembola 9 2 10 1 2 3.0 Insect larvae 7 119 52 2 22.8 O ANA T. MOLDOVAN, T ANJA PIPAN, S ANDA IEPURE, A NDREJ MIHEVC & JANEZ MULEC

PAGE 5

ACTA CARSOLOGICA 36/3 2007 497 e measured p h ysico-c h emical parameters (tem perature, pH, conductivity) s h own in Table 3 are emp h a sized by t h e results of t h e canonical correpondence anal ysis in Figure 3. e measured parameters are dened mostly by t h e F1 axis (62% signicance). Ordination of t h e fauna is related to t h e tem perature and pH in Postojna, Planina, Vadu-Criului and Un gurului caves, and to t h e con ductivity in t h e Vntului cave, w hic h is negatively correlated to t h e two before mentioned Fig. 2: Relative abundance of fauna groups in percolation water of the studied caves: (a) P lanina, (b) P ostojna, (c) Ungurului, (d) Vntului, (e) Vadu-Criului (samples without fauna and periods of dryness are not presented). Fig. 3: Canonical correspondence analysis based on fauna groups (green dots) and physico-chemical parameters (red romboids) of the percolating water in ve studied caves (blue squares) (see also Tables 1 and 3). BIODIVERSIT Y AND ECOLOG Y OF FAUNA IN PERCOLATING WATER IN SELECTED SLOVENIAN AND ROMANIAN CAVES

PAGE 6

ACTA CARSOLOGICA 36/3 2007 498 Table 3: Basic environmental parameters of sampled drips in Slovenia and Romania Parameter / Cave Postojna Planina Ungurului Vntului Vadu-Criului Temperature (C) mean 9.02 8.99 9.20 7.70 11.40 SD 0.45 3.24 2.59 1.31 0.60 min 7.30 2.40 4.40 5.40 10.20 ph max 10.70 12.80 15.20 8.60 12.50 mean 7.80 7.68 7.80 7.00 7.80 SD 0.18 0.53 0.82 0.80 0.23 min 7.40 6.53 6.40 6.30 7.40 max 8.70 8.22 8.30 8.30 8.10 Conductivity (Scm -1 ) mean 360.40 343.54 321 484 423 SD 56.09 65.17 116.15 157.50 65.70 min 215.00 214.00 127 286 303 max 465.00 464.00 568 630 574 Drip rate (mlmin -1 ) mean 7.03 12.79 SD 14.16 23.07 min 0.50 0.20 max 156.00 88.00 Fig. 4: e projection of P ostojna cave system (in red) and the frames (1-6) dened for the calculation of dolinas surface. Black dots represent sampling points for percolation water fauna inside the system. parameters (temperature, pH). e presence of nematodes can be also related to t h e hig h con ductivity and is probably due to pollution from t h e surface. Correlation of abundance among sampling sites and drip rates (mlmin-1) was slig h tly sig nicant in t h e Planina cave (r s = 0.5; p<0,002) w hile not in t h e Postojna cave. For t h e Postojna cave system we obtained relatively signicant correlation (0.71) between t h e surface area of dolinas (Fig. 4) and t h e abun dance of fauna in percolation water in corre sponding station(s). It can be explained by t h e importance of dolinas for water accumulation and continuous release underground, w hic h en sures t h e permanent dri of animals inside t h e caves. O ANA T. MOLDOVAN, T ANJA PIPAN, S ANDA IEPURE, A NDREJ MIHEVC & JANEZ MULEC

PAGE 7

ACTA CARSOLOGICA 36/3 2007 499 is study provides insig h ts in t h e knowledge about fauna in percolation water from ve Slovenian and Romanian caves. Romanian long-term investigations of t his h abitat are publis h ed for t h e rst time and t h us, new ecological information was obtained for t h e studied area. Alt h oug h belonging to dierent karst regions (Slovenian Classical Karst and Western Romanian Carpat hians), wit h dier ent c h aracteristics and fauna composition, some simi larities were observed. e same fauna groups can be ob served in t his h abitat. Suc h groups h ave representatives adapted to life in small volumes, as t h e interstices in t h e unsaturated zone of t h e caves. Dierences between caves and sampling points can h ave various explanations, suc h as degree of frac turisation, cavernosity, epikarst development, land use, surface (soil) and subsurface (unsaturated zone) capac ity for water storage, presence/absence of pollutants etc. e importance of t h e surface landforms, as dolinas, can be determinant in t h e case of Postojna cave system. For t h e moment we can not sustain one or anot h er of t h ese h ypot h eses, because furt h er and additional information are needed. New directions for furt h er researc h es must be proposed, and beside biological studies information about c h emistry of t h e water, geomorp h ology, geology, h ydrology are necessary to complete our view about epi karst and vadose zone fauna. Culver (2003) emp h asizes t h e importance of sub surface h abitats, including epikarst, in cave colonization. Fauna from unsaturated layers, as epikarst and vadose zone, are introduced by percolating water in subterra nean basins, lakes and rivers. is h as been proven, but w h at we want to stress on is t h e problem of active colo nization of fauna in percolating water, in t h e frame of t h e active colonization model proposed by Rouc h & Danielo pol (1987). Active colonization is possible t hroug h active dris t h at can be dened as t h e be h aviourally controlled downstream displacement of animals as an adaptation to escape predation, or to colonize new sections of a river. e concept of dri was rstly introduced by Need h am (1928), but dri attracted more attention aer t h e 1950s. Excepting stygop hiles and stygoxenes, w hic h can be in deed accidentally was h ed inside t h e caves, stygobiont h orizontal and vertical movements can be actively deter minated by factors as population growt h or even preda tion. Periods of intense percolation are not necessarily correlated wit h hig h abundance and hig h biodiversity, as s h own in t h e correlation between drip rates and num ber of individuals. Dri could be t h e best way of deep underground colonization of aquatic fauna especially in areas w h ere t h ere is no organised inow from t h e sur face. ere is a strong need of taxonomic work for all t h e collected material in order to sustain our assumption of an active colonization of cave h abitats by fauna from un saturated layers. e epikarst and vadose h abitats are hig h ly vulner able. Bot h h old a considerable reservoir of subterranean water and h arbor exceptionally ric h copepod fauna wit h a hig h frequency of endemism t h at particularly increas es conservation value of t h e epikarst and karst area as a w h ole (Pipan 2005). Study of t h e unsaturated zone rep resents a powerful tool for understanding environmental eects on groundwater fauna. e study of t h e fauna in percolation water of selected caves in Slovenia and Ro mania can serve as a model for subterranean fauna in general. Results of t his work increase our understanding of ecology in t h e epikarst and deeper unsaturated zone. In particular, t his researc h contribute to t h e c h aracteriza tion of geomorp h ologic and biologic processes in t h e karst aquifers and knowledge t h at is valuable for future pollution control eorts not just in Classical Karst in Slo venia and Western Carpat hians in Romania but in karst terrains worldwide. From a biological perspective, very little work h as been done investigating t h e unsaturated zone fauna worldwide and our results contribute to t h e state of current knowledge about t his fauna. In a future perspective we will use our results to support t h e design and implementation of studies at larger scales, wit h t h e ultimate goal to understand better h ow small scale pro cesses in t h e unsaturated layers are responsible for t h e quality of water in a karst aquifer. DISCUSSION AND CONCLUSIONS ACKNOWLEDGEMENT Aut h ors are grateful to David C. Culver and Dan L. Dan ielopol for revision of t h e Englis h text and h elpful com ments on an earlier version of t h e manuscript. Romanian aut h ors are grateful to Andreea Oarg and Graiela Stur za for h elp during sampling, Bogdan Onac for geological information, and t h e custodians of t h e Vntului cave and t h e administrations of t h e Ungurului (Speleological Club CSA) and t h e Vadu-Criului (t h e rii Criurilor Mu BIODIVERSIT Y AND ECOLOG Y OF FAUNA IN PERCOLATING WATER IN SELECTED SLOVENIAN AND ROMANIAN CAVES

PAGE 8

ACTA CARSOLOGICA 36/3 2007 500 REFERENCES Camac h o, A. I., A. G. Valdecasas, J. Rodrguez, S. Cuez va, J. Lario, & S. Snc h ez-Moral, 2006: Habitat con straints in epikarstic waters of an Iberian Peninsula system cave.Annales de Limnologie-International Journal of Limnology, 42, 2, 127-140. C h appuis, P.A., 1925: Sur les Coppodes et les Syncarides des eaux souterraines de Cluj et des Monts Bi h ar.Bulletin de la Socit de Sciences de Cluj, 2, 157182. Culver, D. C. & Sket B., 2000: Hotspots of subterranean biodiversity in caves and wells.-Journal of Cave and Karst Studies, 62, 11-17. Culver, D. C., 2003: Epikarst from an ecological and evolutionary perspective: suggestions for future re searc h. In: Jones, W. K., D. C. Cuver, & J. S. Herman (eds.) P roceedings of the Epikarst Symposium, Oc tober 1-4, 2003, Sheperdstown, West Virginia, USA, Karst Waters Institute Special Publication 9, pp. 127-131. Delay, B., 1968: Donnes sur le peuplement de la zone de percolation temporaire.Annales de splologie, 23, 4, 705-716. Need h am, P. R., 1928: A net capture of stream dri or ganisms.Ecology, 9, 339-342. Petri, M., 2002: C h aracteristics of rec h arge-disc h arge relations in karst aquifer.Intitut za raziskovanje krasa ZRC SAZU, Zaloba ZRC, p. 154, Postojna Ljubljana. Pipan, T., 2003: Ecology of copepods (Crustacea: Co pepoda) in percolation water of t h e selected karst caves.Doctoral Dissertation (in Slovene wit h Eng lis h abstract and summary), University of Ljubljana, Dept. of Biology, p. 130, Ljubljana. Pipan, T., 2005: Epikarst A Promising Habitat. Cope pod fauna, its diversity and ecology: a case study from Slovenia (Europe).Karst Researc h Institute at ZRC SAZU, ZRC Publis hing, p. 101, Postojna Lju bljana. Pipan, T., & Culver, D. C., 2005: Estimating biodiversity in t h e epikarstic zone of a West Virginia cave.Jour nal of Cave and Karst Studies, 67, 2, 103-109. Pipan, T., M. C. C hristman, & D. C. Culver, 2006: Dy namics of epikarst communities: microgeograp hic pattern and environmental determinants of epikarst copepods in Organ Cave, West Virginia.American Midland Naturalist, 156, 75-87. Pipan, T., & Culver, D. C., 2007: Epikarst communities: biodiversity h otspots and potential water tracers.Environmental Geology, 53, 265-269. Plea, C., 1969: Researc h es on reproductive periodicity at some cave troglobiontic Crustacea (in Romanian in original). -Doctoral t h esis, Traian Svulescu Biological Institute of t h e Romanian Academy, Bucureti, 180 pp. Plea, C., O. Moldovan, & A. Munteanu, 1996: Aperu biospologique sur la Grotte de Vadu-Criului, Monts Pdurea Craiului (Transylvanie, Roumanie).Travaux de lInstitut de Spologie Emil Racovitza, 35, 115-142. Plea, C., M. PinteaAlb, M. Sljan, & F. Wild, 1965: Date noi asupra ecologiei i rspndirii geograce a ciclopidelor (Crustacea, Copepoda) n Romnia.Lucrrile Institutului de Speologie Emil Racovi, 4, 141. Plea, C., & G. Racovi, 1973: Rec h erc h es sur les gours, en tant que milieux aquatiques souterrains. Note I. In: Livre du cinquantenaire de LInstitute de Spolo gie Emile Racovitza, Academia R.S.R, Bucureti, pp. 487-502. Rouc h, R., 1968: Contribution la connaissance des h arpacticides h ypogs (Crustacs Coppodes).Annales de splologie, 23, 1, 5-167. Rouc h, R., & D. L. Danielopol, 1987: Lorigine de la faune aquatique souterraine, entre le paradigme du refuge et le modle de la colonisation active.Stygologia, 3 (4), 345-372. Sket, B., 1981: Niphargobates orophobta n.g., n.sp. (Am p hipoda, Gammaridae s.l.) from cave waters in Slovenia (NW Y ugoslavia).Bioloki vestnik, 29, 1, 105-118. seum) caves. is study was funded by t h e Ministry of Hig h er Education, Science and Tec hnology of t h e Repub lic of Slovenia, t h e Slovenian Researc h Agency (project no. BI-RO/05-06/001), and by t h e Ministry of Educa tion and Researc h, Department of European Researc h Programs in Romania (MCT project no. C18446/2006) and grant 1150/2007 of t h e National University Researc h Council. O ANA T. MOLDOVAN, T ANJA PIPAN, S ANDA IEPURE, A NDREJ MIHEVC & JANEZ MULEC

PAGE 9

ACTA CARSOLOGICA 36/3 2007 501 Sket, B., P. Trontelj, & C. agar, 2003: Speleobiological c h aracterization of t h e epikarst and its h ydrologi cal neig h bor h ood: its role in dispersion of biota, its ecology and vulnerability. In: Jones, W. K., D. C. Cuver, & J. S. Herman (eds.) P roceedings of the Epi karst Symposium, October 1-4, 2003, Sheperdstown, West Virginia, USA, Karst Waters Institute Special Publication 9, pp. 104-113. ebela, S., 1995: Geoloke osnove oblikovanja najveje podorne dvorane v Postojnski jami-Velike Gore.Annales, 7, 95, 111-116, Koper. Trek, B., 2003: Epikarst zone and t h e karst aquifer be h aviour. A case study of t h e Hubelj catc hment, Slo venia.Geoloki zavod Slovenije, p. 100, Ljubljana. Williams, P. W., 1983: e role of t h e subcutaneous zone in karst h ydrology.Journal of Hydrology, 61, 4567. BIODIVERSIT Y AND ECOLOG Y OF FAUNA IN PERCOLATING WATER IN SELECTED SLOVENIAN AND ROMANIAN CAVES



PAGE 1

SUBTERRANEAN WATERS DESCRIBED IN CARNIOLAN LETTERS MAILED TO A. KIRCHER 1601; 1680 PODZEMNE VODE V KRANJSKIH PISMIH POSLANIH A. KIRCHERJU 1601; 1680 Stanislav J UNI 1 Izvleek UDK 551.44(091):929 Stanislav Juni: Podzemne vode v kranjskih pismih poslanih A. Kircherju (* 1601; ) Prvi so obravnavana pisma, ki so ji h Kirc h erju poiljali jezuiti in plemii povezani z obmoji sedanje Slovenije. Kirc h er je z nji h ovo pomojo sestavil eno prvi h te h tni h teorij Cerknikega jezera in objavil tevilna druga dognanja o krasu. Ugotovljeni so zbiralci Kirc h erjevi h podatkov o Cerknikem jezeru in idri jskem rudniku, ki ju osebno ni obiskal. Pregledana so poroila jezuita Wilpen h oerja, ki je skrbel za irjenje Kirc h erjevi h del na Kranjskem. Opisana so pisma, ki ji h je Kirc h erju poiljal Janez Vajkard Turjaki, mecen Kirc h erjevi h knjig, samosvoj pi onir raziskovanja rib v globina h Postojnske jame in Valvasorjev vzornik. Kljune besede: At h anasius Kirc h er, Kritof Wilpen h oer, Janez Vajkard Turjaki, Janez Vajkard Valvasor, zgodovina ra ziskovanja krasa, Cerkniko jezero, jezuiti. 1 Stanizlav Juni, Fara 2, 1336 Vas, juznic@h otmail.com Received/Prejeto: 04.09.2007 COBISS: 1.01 ACTA CARSOLOGICA 36/3, 503-510, POSTOJNA 2007 Abstract UDC 551.44(091):929 Stanislav Juni: Subterranean waters described in Carniolan letters mailed to A. Kircher (* 1601; ) Kirc h ers letters connected wit h t h e area of todays Slovenia were analyzed. His Jesuit informer Wilpen h oers reports on t h e Cerknica Lake and Idrija Mine were put forward. He also h elped distribution of Kirc h ers books among Auerspergs and ot h er Ljubljana nobles. Janez Vajkard Auerspergs letters as an example of hig h nobility correspondent patronage were put at t h e limelig h t in connection wit h Janez and his admirer Valva sors own researc h of t h e Postojna Cave ora and fauna. Keywords: At h anasius Kirc h er, C hristop h orus Wilpen h oer, Janez Vajkard Auersperg, Janez Vajkard Valvasor, History of Karst Researc h, Cerknica Lake, Jesuits. I NTRODUCTION e baroque letters are important source for t h e history of science and karst in particular. In fact, t h ey are some times t h e only baroque sources available because at t h ose times t h e printed publications were not suc h necessary part of t h e researc h as t h ey are nowadays. Kirc h ers cor respondence kept at Roman Arc hive of t h e Ponticia Universit Gregoriana (APUG) contains 2291 letters of 763 writers distributed at 14 folios. Almost one t hird of t h e correspondents were Jesuits. We knew some of t h e letters contained information about Cerknica Lake w hic h Kirc h er never visited personally but nevert h eless described as one of its rst scientic researc h ers. We sus pected t h at h e h ad Carniolan informers and t h e guess proved to be a true one. We examined all letters mailed to Kirc h er from Carniola or Gorica, and also t h e letters of people connected wit h t h ose areas. We are h appy to report t h at at least some success was obtained.

PAGE 2

ACTA CARSOLOGICA 36/3 2007 504 KOBAV Tab. 1: Latin Kobavs letters mailed from Vienna to Kirchers Rome (Gramatowski & Rebernik, 2001, 63). Date APUG January 1, 1640 567 20r-21v February 20, 1640 567 f. 205rv Our rst suspect was Cerknica native Andreas Ko bav (* November 7 or 11, 1593 Cerknica; joined Societatis J esu (SJ) October 22, 1610 Brno at Moravia; February 12, 1654 Trieste), one of t h e most talented Paul Guldins (Habakkuk, 1577; 1643) students at Graz University. Aer his yout h at native Notranjska, Kobav joined t h e Franciscans but soon c h anged his mind and became a Jesuit. At least in 1652 h e was a confessor at Ljubljana College. Just few mont h s before deat h h e moved to Tri este College, probably stopping at Cerknica on his way. As t h e native of Cerknica h e was certainly an expert for t h e Cerknica Lake p h enomena, but h e did not describe t h em at his known letters mailed to Kirc h er. He rat h er examined t h e properties of new comet paving t h e way for his fort h coming book. CUZZIO Tab. 2: Italian letters of Giulio Cuzzio (S J; 1688) mailed from Gorica to Kirchers Rome (Gramatowski & Rebernik, 2001, 37, 131) Date APUG September 6, 1675 565 f. 184rv November 1, 1675 565 f. 263rv Besides Ljubljana also Jesuits from Gorica or Trieste College could h ave informed Kirc h er about t h e nearby Notranjska limestone area p h enomena. Just two let ters from Gorica mailed to Kirc h er are kept at APUG Cuzzio wrote bot h of t h em in Italian language wit h Latin appendix of his college rector. Among ot h er reports h e described Jesuit visit to t h e Count Torrismondo de la Torres (urn-Valsassina) castle Devin (Duino, Tybein) He described t h e administrative curiosities and did not take care of t h e limestone region Timav River near Devin or Cerknica Lake. WILPENHOFFERS CERKNICA LAKE REPORT Tab. 3: Wilpenhoers Latin letters mailed to Kircher aer Wilpenhoers settlement at Ljubljana in 1653 Date APUG October 24, 1658 567 f. 111rv October 23, 1669 559 f. 95rv November 28, 1669 559 f. 50rv Ljubljana Jesuit C hristop h orus Wilpen h oer (Kritof Wilpen h ofer, Wilpen h ofer, Vilpen h oer, Bilpen h ofer, February 1597 Radstadt on Enns at Salzburg; SJ Marc h 19, 1614 Brno; Marc h 30, 1671 Ljubljana) stayed at Ljubljana between 1640-1641 and 1644-1648. At t h at time h e did not write to Kirc h er, w h o was just beginning his ascent to fame. During his t hird and nal stay at Lju bljana Wilpen h oer mailed several letters to Kirc h ers Rome (Gramatowski & Rebernik, 2001, 113, 136; Do linar, 1976, 186). ose documents are of greatest inter est for t his study. At least one of Wilpen h oers Ljubljana letters went t hroug h Venice and got a mark franc hise of Venice. Fig. 1: Wilpenhoers October 24, 1658 letter with Kirchers address and a mark of franchise (APUG 567 folio 111v). (Courtesy of Stanford University and Glen Worthey). STANISLAV J UNI

PAGE 3

ACTA CARSOLOGICA 36/3 2007 505 Fig. 2: Wilpenhoer wrote to Kircher about Volfs copy of M undus, Idria M ine, and Cerknica Lake on October 24, 1658 (APUG 567 folio 111r). (Courtesy of Stanford University and Glen Worthey) Wilpen h oer was a novice at Brno. He studied p hi losop h y and t h eology at Graz between 1616 and 1625. On June 29, 1635 h e gave his fourt h mobility vows, but h e never le his h omeland for missions and stayed at Habsburg domains. In 1628 h e taug h t logic at Graz and became t h e Viennese College confessor (1629-1631). In 1640 and 1641 h e h ad similar function at Ljubljana Col lege. Later h e became superior at Varadin (1642-1643), prefect at Ljubljana (1644-1648), and superior at Sopron (1649). In 1651 h e was at Vienna and next year h e be came t h e confessor of Jaurin (Raab) bis h op. He informed Kirc h er about Cerknica Lake p h enomena during his stay at Ljubljana College. He was an administrator, scribe (scriptor) counsellor, and prefect of t h e Ljubljana Jesuit c h urc h and stayed at Ljubljana almost for two decades between 1653 and 1671 (Lukcs, 1982, 789-790). Wilpen h oer listed t h e Carniolan local natural cu riosities and Kirc h er in fact used his report for his own description of Carniola. Wilpen h oer took care for t h e distribution of his Jesuit-fellow Kirc h ers books among Carniolans, mostly on demand of t h e Governor General Count Volf Engelbert Auersperg. Immediately aer t h e M undus Subterranei printing (1657) h e reported on Oc tober 24, 1658 t h at t h e Count, our Governor wis h ed to get t h e item. In fact, Volfs librarian and personal friend Janez Ludvik Sc h nleben (* 1618 Ljubljana; SJ October 28, 1635 Vienna-1654; 1681) marked Volfs bookplate at his copy of t h e famous Kirc h ers M undus Subterranei during t h e same year 1658. Somew h at later Sc h nleben became rat h er prolic researc h er of Cerknica Lake and publis h ed his opinions just before h e died (1680-1681). He collected t h e library of his own and bequeat h ed his books to t h e Ljubljana Jesuits at his last will. e readers also expected similar Kirc h ers work M undus M editerranes because Kirc h er announced h e would publis h it. It was never put at t h e limelig h t, at least not under t h at title. In fact, t h e M undus M editerranes case seems to be a part of t h e very clever politics. Kirc h er paved his road to fame by adding index of his publis h ed and would-be publis h ed books at t h e end of his early works. Several of t h em were never printed, but readers Fig. 3: Wilpenhoer writes to Kircher on October 23, 1669 (APUG 567 folio 95r). (Courtesy of Stanford University and Glen Worthey) SUBTERRANEAN WATERS DESCRIBED IN CARNIOLAN LETTERS MAILED TO A. KIRCHER 1601; 1680

PAGE 4

ACTA CARSOLOGICA 36/3 2007 506 expected t h em anyway. Just by reading t h e titles and de scriptions of his planned works everybody was won dering h ow clever and erudite man Kirc h er must be. e Baroque aut h ors actually knew very well h ow to sell t h eir books and we could learn a lot from t h eir examples. In addition Wilpen h oer broug h t previously un known facts about karst to Kirc h ers attention. In M undus Kirc h er did not mention Idrija Mine or Cerknica Lake yet, alt h oug h at t h e paragrap h Alpine hydraulic h e de scribed waters at Alps region wit h silver and ot h er min erals beneat h t h e mountains (Kirc h er, 1657, 185-186). He carefully described Danube River (Kirc h er, 1657, 162) and properties of quicksilver (Kirc h er, 1657, 202) but failed to mention famous Idria mercury mine. Wilpen h oer reported on t h e mercury mine wit h out mentioning t h e city of Idrija. He also described t h e vein of fossil silver and cited t h e Jesuit Casatis work connected wit h mining. Wilpen h oer put special con cern on t h e lake near t h e town named Cerknica. e lake altoget h er sank to enable ploug hing of t h e dried elds and even h unting. Aer a w hile t h e lake reappeared again and ooded t h e elds wit h t h e water and s h, ac cording to Wilpen h oers report. Eleven years later on October 23, 1669 Wilpen h of fer again wrote from Ljubljana to Kirc h ers Rome wit h out mentioning any karst p h enomena. A week later h e wrote a new letter (November 28, 1669) wit h last lines some w h at compressed to t t h e single page. He mentioned two parts of Kirc h ers Ars M agna sciendi seu combinatori ca under t h e s h ortened title Ars combinatorica. at book again pleased very muc h t h e Count his Eminence Land Governor Volf Auersperg. In t h e meanw hile Volf was promoted to a hig h er status of emperors secret counsel lor (1660) and consequently got t h e prestigious title his Eminence (Mol, 1937, 50). Volf certainly got Kirc h ers book Ars combinatorica too, and Sc h nleben s h elved it at t h e class of linguistics. Certainly, t h e words of t h e baroque times did not always h ave t h e meaning we as cribe to t h em today. At t h e front page bookplates of t hree ot h er Kirc h ers books Sc h nleben wrote a note: aut h ors (Kirc h ers) present. Ars combinatorica and M undus Sub terranei were not among t h em. Volf certainly h ave to buy some of his books as h ad all ordinary mortals, even if h e was t h e Carniolan Governor. Kirc h er continued Frisc h lins researc h of Carniola subterranean waters (Habe, Kranjc, 1981, 20-21; Koroec, 1967, 12) wit h out ever seeing t h e p h enomena h e de scribed. Kirc h ers museum at Rome was in fact t h e Jesuit collection of all world curiosities and Carniolan ones just t t h ere. Already before Wilpen h oers notes Kirc h er received some ot h er information about Carniolan pe culiarities focusing on Idrija Mercury Mine. Kirc h ers informer was very young Klagenfurt Jesuit Sigismond Siserius (* May 1, 1636 Klagenfurt; SJ; December 29, 1693 Vienna). Kirc h er publis h ed Siserius letters as h e was used to do wit h ot h er fellow Jesuits reports, alt h oug h h e eventually did not care to publis h Wilpen h oers con tributions. Siserius work was put at t h e limelig h t as t h e appendix of rst edition of Kirc h ers subterranean book under t h e title Epistola ad R.P Kircherum, in qua hydria sive fodina argenti vivi in Carniola describitur (Kirc h er, 1657, 2: 173). Wilpen h oer certainly read t h e item at Au erspergs Ljubljana library and decided to improve Kirc h ers understanding of t h e p h enomena wit h his own Idrija Mine and Cerknica Lake descriptions. Kirc h er eventually used t h e new data acquired from Wilpen h oers letter for his M undus Subterraneus (Kirc h er, 1678, book 9, part 2, paragrap h 7). From t h e early little booklet later publis h ed as t h e addenda to Kirc h ers outer space trip, t h e t hird printing of M undus (1678) wit h Cerknica Lake description h ave grown up to nearly t h ousand pages of t h e great folio format. Its weig h t is almost 5 kg! e book was not just Fig.4: Wilpenhoer reports to Kircher about Volfs excitement while reading his Ars combinatorica on November 28, 1669 (APUG 559 folio 50r). (Courtesy of Stanford University and Glen Worthey) STANISLAV J UNI

PAGE 5

ACTA CARSOLOGICA 36/3 2007 507 h eavy, it was considered to be t h e most important caves and underground waters baroque researc h. Kirc h er put at t h e limelig h t early description of t h e caves Antiparos at t h e middle of Greek Kiklads (Kyklads) Islands sout h east of At h ens, G h ar Kebir of Tunis, Drac h en h h le near Mixnitz sout h east of Styrian Bruck on Mur River, and, least but not last, t h e Cerknica Lake. He founded t h e modern h ydrology of karst and connected t h e subterra nean springs wit h seas and precipitations. In 1669 Englis hman Edward Brown (* 1642; 1708) continued Kirc h ers Cerknica Lake researc h on be h alf of t h e London Royal Society. Browns mistakes stimulated t h e development of Valvasors own opinions and ideas crowned wit h Valvasors ambitious ellection at t h e same Royal Society as its rst and so far t h e only fellow from Carniola on December 14, 1687. Former Jesuit Sc h nle ben and Valvasor modied Kirc h ers explanation of udds (Valvasor, 1689, 1: 626, 630), and Valvasor boug h t almost all Kirc h ers works including M undus During Kirc h ers lifetime E.G. Kappelius (1672, 1685) described Cerknica Lake s h and elds at Hamburg Journal (Koroec, 1967, 13) and paved t h e way for t h e researc h of local aut h ori ties, namely t h e Prince Janez Vajkard Auersperg. JANEZ VAJKARD AUERSPERG Tab. 4: Count, later P rince J anez Vajkard Auerspergs letters mailed to Kircher Date Place APUG May 30, 1651 Vienna 556 214r, 215v February 9, 1654 Regensburg 556 212r, 213v October 27, 1655 Vienna 556 204r, 205v May 3, 1661 Vienna 562 26rv November 12, 1665 Vienna 555 212r1,v1, 213v1 April 23, 1671 Ljubljana 565 70rv, 70av Janez Vajkard Auersperg, his older brot h er Volf En gelbert, and t h eir protg Valvasor were among t h e most erudite Carniolans. Six Janez Vajkard Auerspergs letters to Kirc h er were preserved at APUG In rst letters Janez ensured his patronage and eventually also payment of Kirc h ers usually very expensive books. Kirc h er dedicated a part of his Oedipus to Janez, w h o was just promoted to t h e rank of Prince (Kirc h er, 1654, 139). Janez mailed t h e last preserved letter to Kirc h er from Ljubljana on April 23, 1671, more t h an a year aer h e was dismissed from his hig h est positions at Viennese Court on December 10, 1669. On July 3, 1669 t h e Prince Janez Vajkard Auersperg boug h t t h e Prince Janez Anton Eggenbergs Notranjs ka manors Snenik and Lo and also took over Posto jna manor. Probably h e expected t h e Viennese troubles Fig. 5: Kircher dedicates 110 pages of the Oedipus second part fourth section to Ljubljana P rince J anez Vajkard Auersperg (Kircher, 1654, 139). Fig. 6: P rince J anez Vajkard Auersperg writes to Kircher from Ljubljana on April 23, 1671, rst page (APUG 556 folio 70r (courtesy of Stanford University and Glen Worthey)). SUBTERRANEAN WATERS DESCRIBED IN CARNIOLAN LETTERS MAILED TO A. KIRCHER 1601; 1680

PAGE 6

ACTA CARSOLOGICA 36/3 2007 508 would force him to return to his native Carniola few mont h s later. e Prince Janez Vajkard Auersperg was aware of t h e Leyden Professor P hilippus Cluverius (Cluverij, 1580 Danzig; 1622) description of Postojna Cave (1623) as a big cave wit h noisy river at h ollow hill near Ljubljana. Janez brot h er Volf h ad at least four Cluverius items at rst or early editions (1619, 1624, 1631, and 1653) bound into t hree volumes. Five Cluverius works were on sale at Ljubljana and Valvasor eventually boug h t all of t h em for his Bogenperg library (Sot h ebys, 1982, 30-31; Mayr, 1678, 71; Valvasor & Magi, 1995, 219-220). Soon aer his last letter mailed to Kirc h er, t h e Prince Janez Vajkard Auersperg used ropes to descend one of his peasants-s h ermen beneat h one foot t hick Postojna Cave natural bridge (1673). Janez owners hip of t h e Postojna area stimulated his curiosity w hic h culminated aer reading Culverius and Kirc h ers books h e in h erited from his recently deceased brot h er Volf. Janez became one of t h e rst systematic researc h ers of Postojna Cave subter ranean ora and fauna. Kirc h er would certainly love to read his results, but no Janez letters eventually written aer 1671 were preserved. Just before Kirc h er died, Valvasor (1679) tried to upgrade t h e discoveries of his model Prince Janez, but h e was eventually not quite successful. Valvasor discovered t h e Auerspergs subject-s h erman w h o formerly broug h t some s h from subterranean of Postojna Cave. e poor old fellow was still alive and in pretty good h ealt h. But h e was certainly not very cooperative being still somew h at frig h tened by t h e g h ost h e seems to h ave met at t h e dark Postojna Cave underground (Reisp, 1983, 144; Kova Ar temis, 2005, 72; Valvasor, 1689, 1/4: 532). Fig. 7: P rince J anez Vajkard Auerspergs subject-sherman examines P ostojna Cave subterranean ora and fauna on the rst planned expedition of its kind in 1673 (Valvasor, 1689, 1/4: 532). CONCLUSION e Jesuits from Carniola informed Kirc h er about t h e Cerknica Lake karst p h enomena. Kirc h er was considered an expert in several elds, but h e did not widely travel aer his young years. His books were in fact t h e com pilation of letters h e received from his fellow Jesuits all over t h e world, and wit h t h at in mind we claimed t h at t h e Carniolian Jesuits were partly t h e aut h ors of Kirc h ers books describing Carniola local subterranean limestone regions. Kirc h er became well aware of t h e local peculi arities, corresponded wit h t h e hig h est Carniolan nobility, and eventually stimulated t h e Auerspergs subterranean karst fauna researc h programme. ACKNOWLEDGEMENTS is researc h was carried out in t h e frame of t h e pro gramme e P h ysics Books and Manuscripts in Car niolan Libraries before Napoleonic Wars founded by t h e Mellon fellows hip of t h e Okla h oma University History of Science Department in 2007. We t h ank acad. Dr. Andrej Kranjc for constructive advices. STANISLAV J UNI

PAGE 7

ACTA CARSOLOGICA 36/3 2007 509 Baraga, France (ur.), 2003: Letopis Ljubljanskega kolegija Drube Jezusove (1596-1691). Ljubljana: Druina. Dolinar, F.-M., 1976: Das Jesuitenkolleg in Laibac h und die Residenz Pleterje 1597-1704, Dissertatio ad Doctoratum in Facultate Historiae Ecclesiasticae Ponticiae Universitatis Gregorianae. Tiskarna Ljubljana, Ljubljana. Gramatowski, W. & Rebernik, M. (ur.), 2001: Epistolae Kirc h erianae index alp h abet hicus index geograp hi cus. Institutum historicum S.I., Roma. Habe, F. & Kranjc, A., 1981: Dele Slovencev v speleologi ji. Zbornik za zgodovino naravoslovja in te hnike. 5-6: 13-93. Kirc h er, A., 1653: At h anasii Kirc h eri e Soc. Jesu Oedi pus Aegyptiacus. Hoc est Universalis Hieroglyp hi cae Veterum Doctrinae temporum iniuria abolitae instauratio. Opus ex omni Orientalium doctrina et sapientia conditum, nec non viginti diuersarum lin guarum aut h oritate stabilitum, Felicibus Auspicijs Ferdinandi III. Austriaci Sapientissimi et Inuictis simi Romanorum Imperatoris semper Augusti tenebris erutum, Atque Bono Reipublicae Literariae consecratum. Tomus II. Mascardi, Romae. Kirc h er, A., 1657: At h anasii Kirc h eri e Soc. Jesu Iter ex taticum II. Qui et Mundi Subterranei Prodromus dicitur. Quo Geocosmi opicium sive Terrestris Globi Structura, un cum abditis in ea constitutis arcanioris Naturae Reconditoriis, per cti raptus integumentum exponitur ad veritatem. In III. Dia logos distinctum (Structura globis terrestris). Ad Serenissimum Leopoldum Ignatium Hungariae, et Bo h emiae Regem. Mascardi, Romae. Reprint: 1659: Frideric Gall, Turnaviae. Kirc h er, A., 1665: Mundus Subterraneus, In XII Libros digestus Joann Jansson, Elizeum Weyerstrat, Amsterdam. ird edition: 1678 : Joann Jansson, Amsterdam. Koroec, B., 1967: Beseda dve o Steinbergovem in drugi h opisi h Cerknikega jezera. Kronika. 15: 11-22. Kova Artemis, T., 2005: Kako je nastala knjiga o deeli Kranjski. Samozaloba, Ljubljana. Lukcs, L., 1982: Catalogi presonarum et orciorum Pro vinciae Austriae S.I. II (1601-1640). Institutum Historicum S.I., Romae. Mol, R., 1937: K zgodovini knejega dvorca. Kronika. 4/4: 48-50. Reisp, B., 1983: Janez Vajkard Valvasor. Mladinska knji ga, Ljubljana. Sot h ebys, Bloomeld Place, New Bond Street, London WIA 2AA, 1982: e Catalogue (A Collection) of Valuable Printed Books and Atlases of t h e Fieent h to t h e Seventeent h Century Formed in t h e Seven teent h Century by a Continental Nobleman and Now t h e Property of Sen h or German Mail h os and Sen h ora Jo h anna Auersperg de Mail h os from Uru guay. Days of Sale Monday, 14t h June 1982 Lots 1223. Tuesday, 15t h June, 1982 Lots 223-440 at eleven oclock precisely eac h day. Sot h ebys, London. Valvasor, J. V. 1689 : Die E hre dess Hertzogt h ums Crain. Wolfgang Moritz Endter, Laybac h-Nrnberg. Valvasor, J. V. & Magi, V. (editor), 1995: Bibliot h eca Val vasoriana. Katalog Knjinice Janeza Vajkarda Valva sorja. Valvasorjev odbor pri SAZU, Ljubljana/Za greb. LITERATURE SUBTERRANEAN WATERS DESCRIBED IN CARNIOLAN LETTERS MAILED TO A. KIRCHER 1601; 1680

PAGE 8

ACTA CARSOLOGICA 36/3 2007 510 P OVZETEK Pisma, ki so ji h Kirc h erju poiljali jezuiti in plemii z obmoja sedanje Slovenije, so prvorazredni zgodovinski dokumenti. Znano je, da Kirc h er nai h krajev osebno ni obiskal; prav logino se zdi, da je moral imeti v Cer knici obveevalce, ki so mu pripovedovali o notranjski h kraki h pojavi h. e dolgo nas zanima, kdo bi utegnili biti ti ljudje, kateri h pisma je Kirc h er vtkal v svoje opise krakega podzemlja. V potev je kazalo vzeti predvsem ljubljanske, gorike in trake jezuite. In res, prav z ljubljanskega kolegija je Kirc h er dobival potrebne po datke o notranjskem krasu. Kirc h er je s pomojo Kranjski h dopisnikov sest a vil prvo upotevanja vredno teorijo Cerknikega jezera po N. Frisc h linovi h pesnitva h in objavil tevilna druga dognanja. Med najbolj uglednimi zbiralci podatkov za Kirc h erja je bil njegov sobrat Wilpen h oer, tisti as pomemben jezuit ljubljanskega kolegija. Salzburan Wilpenh oer se je dodobra seznanil z nenavadnimi krakimi pojavi o Cerknikem jezeru, zanimal pa ga je tudi rudnik ivega srebra v Idriji kot tedaj najdonosneje kranjsko podjetje. Zato smo pregledali poroila in ivljenjsko pot je zuita Wilpen h oerja, ki je obenem skrbel e za irjenje Kirc h erjevi h del na Kranjskem, predvsem po elja h deelnega glavarja Volfa Engelberta Turjakega. Volfov brat Janez Vajkard Turjaki, mecen Kirc h erjevi h knjig in zelo samosvoj pionir raziskovanja rib v glo bina h Postojnske jame, je poiljal Kirc h erju pomembna pisma. Med prvimi je organiziral ekspedicijo s potrebno opremo, da je la h ko dal na plan privlei ribe iz podzem ni h voda; le-te so se resda izkazale za nekoliko mrave in komajda vredne knejega obeda. Zaenkrat e ne vemo na kaken nain je o tem obvestil svojega dopisovalca Kirc h erja. Janezova postojnska raziskovanja je nadaljeval Valvasor. Kupil je domala vsa Kirc h erjeva dela, ki ji h je zaradi denarni h teav pozneje prodal v Zagreb. PODZEMNE VODE V KRANJSKIH PISMIH POSLANIH A. KIRCHERJU 1601; 1680 STANISLAV J UNI



PAGE 1

SOUND PROPERTIES OF PLANINSKO POLJE SLOVENIA ZVONE ZNAILNOSTI PLANINSKEGA POLJA SLOVENIJA Josip K OROEC 1 Botjan P EROVEK 2 Dejan V ONINA 3 1 J. Koroec, Institut za konservatorstvo in restavratorstvo pri Zavodu za varstvo kulturne dediine Slovenije 2 B. Perovek, SaetaZavod za kulturne in promocijske aktivnosti 3 D. Vonina, Notranjski muzej, Postojna Received/Prejeto: .29.10.2007 COBISS: 1.04 ACTA CARSOLOGICA 36/3, 511-519, POSTOJNA 2007 e researc h project Dening t h e Heritage P h enom enon h as been organized wit hin t h e programme of t h e Institute for Conservation and Restoration of t h e Insti tute for t h e Protection of Cultural Heritage of Slovenia (ZVKDS) and t h e Notranjska Museum in Postojna since 2005. Its direct cause was t h e constant discoveries of new qualities and ingredients of t h e notion h eritage itself and t h e increasing complexity of t h e conditionality of its ex istence, meaning and protection. In t h e processes of sus tained c h anges in existing relations, it is also s h own as a most sensitive area, w h ose properties are t hreatened by various, oen interrelated causes. Conservation practice warns t h at t h eir state depends on a large number of fac tors. us t h e denition of t h e h eritage p h enomena and t h e met h ods of recognizing, evaluating and dening it also brings a gradual implementation of new met h ods of conservation and adaptations of establis h ed ones to its core meaning. However, h eritage represents a p h enomenon only if it is constantly upgraded as a notion and if it uses its h umanization to encourage creativity and to nd mean ing for t h e future in t h e past. It also denes t h e under standing of t h e dimensions of space-time for a period, t h e dimensions in w hic h its substance exists and gives it its civilizational-cultural c h aracter, and also t h e dimen sions creating t h e balance of relations wit hin it or even conditions t h e deconstruction of its ingredients. e fact is t h at t h e notion of h eritage c h anges drastically depend ing on t h e meaning of its completeness. Some types or protected groups 1 are only now entering public aware ness and receiving expert treatment, surely because of t h e recognition of t h e value of suc h goods, and also because of t h eir actual involvement and essentiality as t h e core of t h e active space. e realization t h at some of t h em, suc h as cultural landscapes, need a more direct protection, be cause t h ey are increasingly endangered, also requires a dierent estimate of causes and consequences of t h e ac tual states of individual types or goods. Planinsko polje was c h osen as t h e rst researc h area because it was estimated t h at it could provide t h e most answers to t h e initial essential questions; t h e purpose is to use t h e gat h ered data to contribute to t h e valorization of t h e eectiveness of individual met h ods and of t h e system of protection itself especially in t h e areas of prevention and intervention. Considering t h e size, it is a manage 1 ey are individually dened by t h e Law on Cultural Heritage Protection, especially articles 5 and 6. Abstract Josip Koroec, Botjan Perovek, Dejan Vonina: Sound properties of Planinsko polje (Slovenia) Geograp hically, t h e Planinsko Polje eld is one of t h e most pre served cultural regions of t h e Karstic landscape of inner Slo venia and as suc h a protected area of national importance. It can be recognized by its exceptional features and landmarks of material and non-material h eritage and by its hig h-quality sym biosis of all ingredients in its space. e researc h of sound in t h e area of Planinsko Polje s h ows t h at it is an important, preserved constituent of natural and cultural h eritage. Its manifestations are also interesting as indicators of its actual endangerment. Key words: sound, non-material h eritage, cultural region, pro tected area. Izvleek Josip Koroec, Botjan Perovek, Dejan Vonina: Zvone znailnosti Planinskega polja (Slovenija) Planinsko polje kot geografski pojem sodi med najbolj o hran jene kulturne krajine krake krajine notranje Slovenije in s tem med varovana obmoja dravnega pomena. Prepoznaven je po izjemni h dobrina h in znamenitosti h snovne in nesnovne dediine ter h krati po kakovostnem soitju vse h sestavin v svo jem prostoru. Raziskave zvoka v obmoju Planinskega polja so pokazale, da le ta predstavlja pomembno o hranjeno sestavino naravne in kulturne dediine. V svoji pojavnosti pa je zanimiv tudi kot pokazatelj njene dejanske ogroenosti. Kljune besede: zvok, nesnovna dediina, kulturna krajina, varovano obmoje.

PAGE 2

ACTA CARSOLOGICA 36/3 2007 512 able area wit h a steady low activity of present factors, an area containing a condensed version of h eritage proper ties similar to t h ose in more distinctive areas and w h ose complete preservation points to a creative co h abitation between its natural and cultural parts and t h e intertwin ing wit h ot h er values. e articulation of met h ods and perfection of tools oers among ot h er t hings a dierent perception and treatment of t h at part of h eritage w hic h was in t h e past protected only because of its exclusive existence. One of suc h types is also sound as a c h aracteristic element of expression of a certain time, as a eld of presence and creativity of a range of activities in a certain space or as an indicator of development and autonomy of its carriers in overcoming t h e endangerment presented by various factors. at is w h y t h e Institute for Preservation and Res toration of ZVKDS and t h e Notranjska Museum used t h e programme Dening t h e Heritage P h enomenon to present its state in Planinsko Polje; it joins t h e c h aracter istics of a cultural landscape wit h regional creation of a natural area into a w h ole, w hic h in time became a quality of national importance. Paradoxically, we realize t h at a sound exists only af ter it stops. Sounds t h at surround us, or h ad surrounded us, are not given enoug h attention, so t h ey are quickly for gotten. e disappearance of sounds is just as severe a loss for society as t h e disappearance of t h e items t h ese sounds belong to. Some sounds will never be h eard again, and some only coincidentally: manual grass scyt hing, manual iron forging, rattling of t h e mills Suc h t hings s h ould be arc hived. In some places, recording h as already been undertaken. In Slovenia t h e lost sounds only appear in individual, mostly museum exhibitions and in fragments in some sound arc hives of individual institutions. Not only items, even past periods h ad t h eir own sound. History books record w h at was h appening in a certain era, allowing us to imagine t h e sound of t h e times. Sounds of eac h era are in tune wit h t h e periods internal life beat, w hic h can be completely dierent even between villages and towns. Regions h ave t h eir own dis tinct sounds as well. e Karst sounds dierent from t h e Littoral and t h e Pannonian plain; because of dierent ac tivities of t h e in h abitants and of course because of t h e natural c h aracteristics of t h e environment. Sound, an ingredient of t h e objective world, wit h its presence denes space as suc h and is a c h aracteris tic of individual areas. It is all t h at makes sound; carriers or creators of sound can be classied as: direct nature wit h its p h enomena and h umans wit h t h eir activities and indirect activities, w h ere sound is only a negligible side eect. Sound is a category understood and perceived as an independent p h enomenon or as a constituent of Sound recording locations, photo: Google Veliki podkovnjak (Rhinolophus ferrumeqinum) P laninska J ama, photo: Slavko P olak J OSIP K OROEC B OTJAN P EROVEK & D EJAN V ONINA

PAGE 3

ACTA CARSOLOGICA 36/3 2007 513 ot h er processes. Simultaneously, sound is an intangible substance, limited in time and variously endangered. Preserving, protecting and saving its values demands a suitable integration into a system of non-material, intan gible h eritage 2 1 w hic h can transport us into sound-active, long-gone spaces. Notionally, t h e eld of sound as a range of sound oscillation is divided into four essentially dierent types: tone, timbre, noise and bang. All four are in t h e so-called sound eld, w hic h denes t h e space in w hic h sound p h e nomena occur. us, t h e area of capturing and editing t h e sounds must be in tune or connected to t h e space 3 ; t h e 2 topic of researc h s h ould be elements of natural and urbane environments because t h at is t h e only way to obtain correct realizations about t h e origins, creation, transformation and disappearance or perseverance of sound in a specic area. e research encompasses: General sound phenomena: are not directly con nected to a specic p h enomenon and are not created only by h umans wit h t h eir activity and are not only a product of natural p h enomena. Example: slamming: wind slam ming into windows (natural p h enomena) or doors or windows slamming (h uman p h enomena) Natural sound phenomena: are directly connected to certain natural p h enomena caused by natural factors (wind, rain, snow, water, re, rock crumbling ) or are caused by animals 4 3 2 Convention for t h e Safeguarding of Intangible Cultural Heritage, UNESCO 2003 3 Miroslav Adlei, Svet zvoka in glasbe (2. knjiga), Svet ive zike (World of Sound and Music (2nd book), World of Liv ing P h ysics), Mladinska knjiga, Ljubljana 1964, p.p.657 4 M. Adlei, pp. 356-362 Human sound phenomena: are directly connected to certain natural p h enomena caused by h uman activity. We can divide t h em into sound p h enomena caused by h uman tools or activities developed t hroug h develop ment of civilization (treating materials wit h manual or mac hine tools, electronics, appliances, hig h tec hnolo gies), current state in t h e eld of culture (artistic and scientic activities, system of verbal expression in inter personal contacts, oral h eritage ) and sound eects specic to h umans as living beings (breat hing, h eartbeat, blood circulation, spatial movement ). During t h e gat h ering, sorting and preservation of sound materials, special attention s h ould be paid to rec ognition of sound equivalents. ese are sound eects from various sources and categories, w hic h are analysed as similar or comparable. In t his case it can be said t h ey are sound associations or similar sounds 5 4 Sound is an element of a certain space and time, not only because of its presence, but also as a reection of activity and c h angeability of relations inside and outside of t h e monitored area; a part of sound enters, invades t h e monitored area from outside. Sound exists also as a cocreator of properties and co-designer of c h aracteristics and specics of a researc h ed area, w h ic h is in our case Planinsko Polje. It was recorded at t h e locations around t h e castle Haasberg, near t h e cave spring kratovka, in t h e h amlet Malni, w h ere t h ey used to h ave mills, near t h e spring of Unica, in Planinska Jama cave, Vranja Jama cave, Jama Makovica cave, near t h e Ravbarjev Stolp tower, on Planinska Gora h ill, around Jakovica, on Pla ninsko Polje itself and in t h e village Planina. Connecting t h e locations h elps create t h e geometry of space for t h e spread of sound and its intertwining wit h ot h er ingre dients. 5 M. Adlei, pp. 710-718 Sound sceneries endangering the space of P laninsko P olje, photo: Google SOUND PROPERTIES OF FIELD PLANINSKO POLJE SLOVENIA

PAGE 4

ACTA CARSOLOGICA 36/3 2007 514 e recording of t h e materials mentioned in t his ar ticle 6 1 was carried out in one year. So t h e recordings are in a c hronological order: from t h e sunny and warm Sep tember 15 t h 2005 to t h e relatively cold and rainy Septem ber 15 t h 2006, wit h t h e sound of a roebuck troating in t h e nig h t of September 20 t h 2006 7 2 appended. It is of course a very condensed representation of a relatively big sound space, intended more to familiarize people wit h t h e sound situation t h an for intense study and understanding. For a better sound image it would be necessary to integrate t h e sound into space wit h a multi-c h annel system, wit h a so-called surround sound tec hnique. And it would make sense to continuously inform listeners of t h e location and t h e season w h en t h e material was recorded. ese are our goals for t h e future 8 3 roug h time, sound actually c h anges t h e core un derstanding and purpose of space. For example, in t h e cemetery near Planina, it is obvious t h at w h en it was created, t h e space was perfect for its purpose if sound is considered; it was an area of relative quiet. Now t h ere is a well travelled road rig h t next to it, a source of aggres sive noise; it can be seen as a disturbance in space and a degradation of t h e original purpose 9 4 Of course we can not treat all sounds introduced by civilization as degra dation, but also as t h e accompanying element of h uman presence and our activity. We can also speak about degra dation w h en t h e quantity of sound presence of an activity becomes t h e prevalent sound element and is as suc h seen as sound pollution; t h e sound balance is disrupted. As an example of sound pollution and sound transforma tion we can list sound events located on t h e CD Celoletni zvoki Planinskega polja / Sound Properties of Planinsko Polje, appended to t h e article. First, t h e ruins of castle Haasberg, w h ic h were t h e starting point for collecting sound materials. Be cause of t h e absence of windows, doors and walls, t h e presence of exterior sounds is normal; and access into t h e building (alt h oug h dangerous) is possible. On 6 attac h ed CD, Celoletni zvoki Planinskega polja, zvoni posnetek t. 1 / Sound Properties of Planinsko Polje, sound track nr. 1 7 e w h ole sound arc hive consists of 7 h ours and 12 minutes of sound at 16 bit/48kHz. Most of t h e material is in stereo. e following equipment was used for recording: professional sol id state recorder Marantz PMD 660 and microp h ones S h ure VP88, Sony ECM-672 and t.bone EM9900. Sorting and edit ing materials was carried out on computers Apple wit h t h e MOTU 828 mk2, RME Fireface 800 interface and soware packages Cubase 3SX, Cubase 4 and Logic Pro 7.2.3. e CD master was prepared wit h t h e program WaveBurner. 8 e task could be executed as a multimedia presentation, w h ose integral part could be t h e recording on DVD. 9 Appended CD, recording nr. 12 recording 10 5 no. 2 a German s h ep h erd, can be h eard freely walking around inside. Because t h e image of t h e w h ole area around t h e castle c h anged, domestic animals live t h ere now. We can h ear a donkey braying on recording no. 3. T h ere is an asp h alt road running past t h e castle, on w h ic h footsteps (recording no. 4) make a completely different sound t h an t h ey used to on a dirt road. Below t h e castle, t h e river Unica flows, s h aring its sound space wit h traffic on recording no. 5 and wit h t h e ringing of remote bells (recording no. 6). At t h e castle, we can also h ear t h e sounds of t h e road t h roug h t h e village Planina accompanied by birdsong (recording no. 7). In t h e h amlet Malni, w h ere t h ey used to h ave windmills, t h e sound of t h eir w h eels h as been replaced by t h e sound of water in t h e remain ing mill troug h s (recording no. 8). Around Ravbarjev Stolp, t h e pleasant sound of snowmelt dripping from t h e roof (recording no. 9) is joined by t h e disturbing buzz of a near-by power plant (recording no. 10). In t h e village Planina, t h e sound of carts was replaced by t h e h eavy traffic of motor ve h icles (recording no. 11) on t h e road w h ic h also leads past t h e cemetery near t h e c h urc h (recording no. 12). T h e evening peace during a sunset at t h e castle Haasberg is s h attered by a remote motorcyclist and a motor plane (recording no. 13); in t h e middle of Planinsko Polje, t h e frog concert is joined by a h ig h -flying passenger plane (re cording no. 14). Listening to t h e roebucks near t h e c h urc h above Jakovica, we are constantly reminded of t h e near-by h ig h way (recording no. 15) and it seems t h at t h e primordial sounds can only be found in t h e embrace of caves. In t h e Planinska Jama cave we are welcomed by t h e overw h elming sound of t h e conflu ence of t h e rivers Pivka and Raka (recording no. 16), not far into Vranja Jama we can h ear t h e crowing of crows toget h er wit h t h e romantic dripping of water (recording no. 17). Researc h of sound demands interdisciplinarity; de spite its non-material nature it is an intertwining of bot h areas of h eritage: material and non-material. Since it is necessary to encapsulate t h e p h enomena in t h e w h ole of t h e researc h ed space, including natural p h enomena and h uman activity, we need to originate from t h e values t h e space possesses, into w hic h we can plant t h e gat h ered materials and establis h criteria to give certain types of sounds t h e status of h eritage. It will also be necessary to dene t h e core terminology and gat h er t h oug h ts and data on sound and its c hronological order. In t h e later process of presenting t h e collected materials it is t h e duty 10 All subsequent recordings are located on t h e CD Celoletni zvoki Planinskega polja / Sound Properties of Planinsko Polje, appended to t h e article. J OSIP K OROEC B OTJAN P EROVEK & D EJAN V ONINA

PAGE 5

ACTA CARSOLOGICA 36/3 2007 515 past eras t h us be come a basis or at least a part of a modern art work, w h ic h as suc h represents an ele ment of modern information. us we prove t h e con nection of t h e past wit h t h e present. e disappearance of sounds is only one of t h e as pects of t h e sound image of our lives. Sounds are con stantly c h anging, as t h e environment is c h anging as well. We need to become aware of t h e eeting nature of moments we experience. Sounds accompanying an ex of t h e relevant institution to present and interpret results. e presentation can be h eld in a museum in t h e tradi tional style or by creating a virtual space, oering a spa tial sound and image presentation. Reasons for t his activity, w hic h at rst glance could not be classied as economically benecial, s h ould be listed or pointed out as well. Usefulness of arc hiving it self can be seen in its connections wit h researc h in t h e areas of history, art history, et hnology even art. e understanding of items and t h e w h ole environment itself is done wit h all t h e senses. us, information about a tool or activity, i.e. at a museum exhibition is not complete wit h out sound. How to explain t h e workings of a tool or mac hine and t h e disrupting impact of its sound on t h e environment, if t h e item displayed is small; people oen connect p h ysical size wit h t h e (non)noisiness of t h e item. at is very h ard to do wit h out a sound illustration, wit h only tec hnical data. However, arc h ive sounds are not only a means to turn back t h e clock. Ordering t h em in a h armonious series we can create sound-noise compositions, coming alive as new aest h etic forms; old sounds as carriers of information of Detail from the cave (P laninska J ama), photo: S. P olak P lane photo of P laninsko P olje, photo: T. M eze M eadow squill (scilla litardierei), photo: A. Selikar Kapnika mokrica, Niphargus stygius, photo: S. P olak P laninsko P olje, photo: S. P olak M eandering of river Unica, photo: T. M eze Recording sound in P lanina, photo: J Koroec Roe Deer (Capreolus capreolus), photo: S. P olak SOUND PROPERTIES OF FIELD PLANINSKO POLJE SLOVENIA

PAGE 6

ACTA CARSOLOGICA 36/3 2007 516 perience are gone rst. As a society and state we could prevent t h at. e project Sounds properties of t h e Karst eld Planinsko Polje is t h us only a stone in t h e mosaic of suc h endeavours. Experts consider Planinsko Polje to be a typical ex ample of a polje, nature lovers see it as a pearl of nature, w h ere t h ey return again and again, s h ermen see it as a y s hing paradise, natives see it as land feeding t h eir livestock, a region t h at silently, almost unnoticeably lives wit h t h em. A eld wit h at least four faces: spring, sum mer, autumn and winter; and at t h e same time, a eld wit h numerous mysterious sounds. Planinsko Polje is our most typical polje and seen by various sciences (geology, geomorp h ology, geog rap h y, h ydrology, regional arc h itecture, h eritage pres ervation etc.) as an exceptional natural p h enomenon of not only national wort h but also natural h eritage of all h umanity. Experts in general see Planinsko Polje as a globally known example of an ideal polje wit h a h ig h density of classical karstic p h enomena in t h e bottom, on t h e fringes and in t h e broader h interlands. Considering t h e centuries of t h e h istory of karst researc h in Slovenia, from Valvasor onwards, it could be said t h at it is a locus classicus for some karstic p h enomena. Due to natural peculiarities and a unique activity of natural processes and t h e endeavours of many generations of natives to co-naturally manage t h e space, it is also an exception al cultural region. Elsew h ere in t h e world, most poljes h ave been ruined because of anti-ooding measures. Even t h oug h it was partially drained, Planinsko Polje as a w h ole luckily remained relatively preserved and is in t h is way unique as well. As suc h it is a rarity in t h e w h ole of t h e Dinaric world, w h ere suc h poljes are rare and muc h degraded. us, we support t h e endeavours to preserve Planinsko Polje wit h its karstic fringe and caves as a natural monument of state or national importance. Planinsko Polje is a part of t h e karstic river basin of Ljubljanica and t h e nort h -westernmost in a series of poljes of t h e rivers system. Its altitude is 447 metres above sea level. e area of t h e polje is 11 km 2 and t h e longer axis, 6 km long, is in line wit h t h e Dinaric direc tion. is area is t h e place of t h e conuence of t h e river Pivka and water from Cerkniko Polje. e rivers join in Planinska Jama and are seen as a unique example of underground river conuences. Waters of t h ese rivers, w h ic h burst out of t h e eart h under a stunning gable in front of Planinska Jama, are h encefort h known as river Unica. e eld h as a prominent inux side wit h beautiful springs in front of t h e cave itself and strong inuxes Malenica and torrential kratovka. roug h millennia, t h e river wit h its sediments sculpted t h e Pla ninska Kotlina Basin into a picturesque and very typi cal polje. Because its river bed was created in t h is way, t h e river Unica meanders in great turns, so its ow is t h ree times longer (18 km) t h an Planinsko Polje. Documentary drawings, global threats: litho bio atmosphere, J Koroec Lime blossom, J Koroec Natural stone pattern, J Koroec Corncrake, J Koroec J OSIP K OROEC B OTJAN P EROVEK & D EJAN V ONINA Climate c h anges Demograp hic growt h Civilization development

PAGE 7

ACTA CARSOLOGICA 36/3 2007 517 Especially prominent are t h e sout h western, and nort h -western fringes of t h e polje wit h springs, and eastern and nort h -eastern wit h a bunc h of sink h oles. e polje does not h ave surface tributaries, receiving t h e water from real karstic springs. Floods on Planinsko Polje h ave become a part of its in h abitants lives. W h en t h e ow of Unica near castle Haasberg rises above 60 m 3 /s, t h e sink h oles are unable to swallow all t h e water. Unica breac h es t h e banks and spreads across t h e eld. Floods occur several times a year and are most common in late autumn and early spring and last for a few days. Because of t h e yearly oods t h e bottom of t h e eld is a more or less natural grass eld mostly of t h e swamp eld type wit h some vegetative peculiarities: meadow squill, blue Siberian iris, red medicinal burnets, purple Illyrian gladioli, to mention only t h ose t h at give colour to w h ole meadows. ey are oen h osts to numerous birds, butteries and even to 110 species of snails and s h ells. In numerous caves in Planinsko Polje, t h ere is a ric h collection of underground fauna. e most famous case is t h e h uman s h (Croteus anguinus). e symbiosis between cultural and natural h eri tage, co-created by planned or spontaneous, organized or coincidental processes and p h enomena, h as c h anged into a quality recognizable in regularly preserved, ex ceptionally or generally recognized goods and attrac tions 11 1 ey inform us of t h e h istoric events, develop ment processes and h abits. Relatively small and wit h a distinctive natural c h aracter, Planinsko Polje is dened to suc h an extent t h at it would be h ard to introduce into it somet h ing not subject to existing factors. Precisely because of t h e size it represented a geograp h ically, stra tegically useful and active point, interesting to t h e out side world, and not an area in w h ic h dierent interests and ambitions could be realized. Probably, t h at is also t h e cause w h y even now t h e in h abitants can decide on suc h ways of being in Planinsko Polje t h at s h ow t h e un derstanding and t h e ad h erence to its properties gained t h roug h time. On t h e basis of conservatorial and ot h er scientic results and suggestions in international recommenda tions and legal acts 12 2 t h e researc h, w hile gradually dis 11 Ivan Marui, Margita Jani, Krake krajine notranje Slovenije (Regionalna razdelitev krajinski h tipov Slovenije, 4) (Karstic Regions of Interior Slovenia (Regional Division of Landscape Types of Slovenia, 4), Ministrstvo za okolje, prostor RS, Urad RS za prostorsko planiranje (Ministry of t h e Environment and Spatial Planning RS, Orce for Spatial Planning), Ljubljana 1998 12 Expert meeting on community involvement in safeguard ing intangible cultural h eritage: Towards t h e implemen tation of t h e 2003 convention (13-15 Marc h 2006, To kyo), Intangible Heritage Section, UNESCO 2006 covering individual facts, is mostly focused on t h e area of connections between t h e material and non-material and between t h e natural and t h e cultural. It is a basic quality and a moderator of existing processes. Most of t h e prom inent carriers of suc h properties of t h e protected area are in t h e sout h-western part of Planinsko Polje, w h ere t h e morp h ology of t h e relief and t h e strategy of work caused t h e activities to condense; now t h ey are spreading to t h e ot h er areas in accordance wit h t h e regional image, allow ing natural p h enomena to be even more independent. us, t h e dialogue between t h e market Planina, w hic h is urbanistically organized in two parallels along t h e regional road, and castle Haasberg, t h e histori cally and economically based complex 13 3 gave birt h to a functional and imaginary triangle, into w h ose frame all t h e key factors of bot h agglomerations t. e pres ence of t h e river Unica, w hic h functions as a belt, cre ated mostly by nature and its p h enomena, is a border between t h em and at t h e same time an equal ingredient of t h e 14 4 Morp h ologically, Planina, called Alben in t h e Middle Ages, was created by transport, t h e core activity of t h e ancient road connection of Ljubljana and Postojna and t h e wider sout h western area. Before t h e demanding mountain pass Postojna Gate it split into t h e old pre sumably Roman route towards Predjama, Nanos and t h e Littoral; and t h e one t h at goes from Haasberg castle to wards Unec, Rakek and Cerknica. e nal point of t h e triangle is t h e preserved, round Ravbarjev Stolp tower, t h e remains of t h e medi eval, well-fortied Mali Grad castle. It was placed above t h e entrance to Planinska Jama cave and aer several owners it at last became property of t h e Haasberg gentry. Culturally, it is t h e most prominent ingredient of Planin sko Polje. Today more of a trace of t h e previous subject, t h e castle started its developmental journey, in accordance wit h t h e dangers of medieval times and relevance of pre vious experience, on t h e top of a hill as a fortied castle. Because of t h e c h anges in t h e way of life and wit h t h e factors w hic h are a constant in forming t h e regional im Miroslav Adlei, Svet zvoka in glasbe (2. knjiga), svet ive zike, zbirka Priroda in ljudje (World of Sound and Music (2nd book), World of Living P h ysics, collection Nature and People), Mladinska knjiga, Ljubljana 1963 13 Nataa tupar umi, Dvorec Planina (Haasberg) pri Rakeku, Varstvo spomenikov 28 (Manor Planina (Haasberg) near Rakek), Zavod RS za varstvo naravne in kulturne dediine (Institute for t h e Protection of Cultural Heritage of Slovenia), Ljubljana 1986, pp. 89-100 14 Vodno bogastvo Slovenije (Water Wealt h of Slovenia), Ministrstvo za okolje, prostor in energijo, Agencija RS za okolje (Ministry of t h e Environment, Spatial Planning and Energy, Environment Agency), Ljubljana 2003 SOUND PROPERTIES OF FIELD PLANINSKO POLJE SLOVENIA

PAGE 8

ACTA CARSOLOGICA 36/3 2007 518 points into an organized system. Bot h components, t h e natural and t h e articial, became h umanized into a h ar monic w h ole in t his system despite t h e extravagance of t h e arc hitecture. us t h e roof of t h e castle building could age, it obtained its nal s h ape as a representative building in t h e beginning of t h e 18 t h century, w h en t h e estate was owned by t h e family Cobencel, it was build on an axis, precisely pointing across t h e terraced descent into t h e valley to Ravbar Tower and entrance to Planinska Jama cave. An actual support for building and park arc hitec ture, t h e tower was t h e essential connection of remote Wet-canvas drawing, by: Franz Kurz, H aasberg through time, photo: M. Koroec, H aasberg today detail, photo: M. Koroec Aerial photo of the SE Art of P laninsko P olje, photo: T. M eze Castle H aasberg, photo: M. Koroec Castle H aasberg, photo: B. P erovek Buzzard, photo: L. Kebe Y ellowhammer, photo: L. Kebe Coenonympha arcania, photo: S. P olak M orimus funereus, photo: S. P olak J OSIP K OROEC B OTJAN P EROVEK & D EJAN V ONINA

PAGE 9

ACTA CARSOLOGICA 36/3 2007 519 protect all t h e peculiarities of aristocratic life w hic h was seen as generally accepted by t h e environment, mostly because of t h e t h oug h tful and continuously upgraded at titude towards nature, i.e. water, t h e crucial ingredient of developed ambiences, besides its picturesque looks en h anced by t h e karstic p h enomena, moisture and sound also supplied economic structures, suc h as mills, saws and suc h wit h energy. e t hings produced by t h e cre ativity of gardeners en h anced wit h imagination, suc h as t h e variety of gardens or individual rock formations, were already naturally supplied by t h e environment. We s h ould not forget t h e environment, not only bordering but also entering t h e castle area wit h its vegetation and ora, s h owing its existence in non-material terms as well. e p h ysical connection among t h e establis h ed objects and Trg Planina market was provided by t h e road per pendicular to t h e axis and leading towards t h e settlement and t h e one following t h e edges of t h e valley towards for mer work buildings and Malni and so on. Less geometrically strict, but no less welcoming are t h e rural settlements Laze, Grarevec and Jakovica and ot h er h amlets and individual farms, placed on t h e rises and t h e terraced fringe. Described, some only partially preserved values, forming a cultural ingredient of h eritage, do not t hreaten nature and its existence wit h t h eir presence. Furt h ermore, researc h s h ows t h at in Planinsko Polje bot h are equal and t h us prove t h e importance of h armony in suc h a space as a quality t h at needs protection. Similarly, all t h ose values included in t h e notion of non-material h eritage s h ould be protected as well, even t h oug h t h e eld is only now re ceiving a more complex scientic treatment in Slovenia. us, t h e sound recordings in Planinsko Polje start an analysis of its basic presence, about t h e types of its mani festations and t h e role it h as as an indicator of relations between individual carriers and t h e degree of endanger ment, especially t h at caused by t h e environment outside t h e studied area. Undoubtedly, t h e systematic recording of sound is a way of documenting certain h eritage types and a form of preserving its non-material properties. Simultane ously, it is a tool w hic h can more prominently contribute to prevention and h elp wit h protection measures. Fur t h ermore, t h e existing and anticipated relations dictate a more complete use of t h e dimensions of sound for gen eral prevention purposes. L ITERATURE Tanja igon, 1997: Planinsko Polje, publis hing h ouse Tuma d.o.o., Ljubljana, Rado Smerdu, 1979: Planinsko polje kot del najpo membneje dediine Slovenije (Planinsko Polje as part of Slovenias most important h eritages), Varstvo narave t. 12 (Nature Protection, issue 12), Ljubljana, Slavko Polak, 2006: Opazovanje ptic, Prironik za opa zovanje ptic na izbrani h obmoji h Notranjske, krasa in obale (Manual for Bird Watc hing on Selected Ar eas of Notranjska, t h e Karst and t h e Littoral), Posto jnska jama, turizem d.d., Postojna, Drago Kolenc, 2006: Dober dan, Krpanova deela (Good Morning Land of Krpan), Notranjsko kraki region alni zavod za turizem Postojna in Postojnska jama, turizem d.d., Postojna, Pobuda za zaito Planinskega polja Slovenskega geolokega drutva in Geomorfolokega drutva Slovenije (Initiative to Protect Planinsko Polje by t h e Slovenian Geological Society and t h e Geomorp h ological Society of Slovenia) SOUND PROPERTIES OF FIELD PLANINSKO POLJE SLOVENIA


printinsert_linkshareget_appmore_horiz

Download Options

close
No images are available for this item.
Cite this item close

APA

Cras ut cursus ante, a fringilla nunc. Mauris lorem nunc, cursus sit amet enim ac, vehicula vestibulum mi. Mauris viverra nisl vel enim faucibus porta. Praesent sit amet ornare diam, non finibus nulla.

MLA

Cras efficitur magna et sapien varius, luctus ullamcorper dolor convallis. Orci varius natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Fusce sit amet justo ut erat laoreet congue sed a ante.

CHICAGO

Phasellus ornare in augue eu imperdiet. Donec malesuada sapien ante, at vehicula orci tempor molestie. Proin vitae urna elit. Pellentesque vitae nisi et diam euismod malesuada aliquet non erat.

WIKIPEDIA

Nunc fringilla dolor ut dictum placerat. Proin ac neque rutrum, consectetur ligula id, laoreet ligula. Nulla lorem massa, consectetur vitae consequat in, lobortis at dolor. Nunc sed leo odio.