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. 37, no. 1 (2008)

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:
K26-00159 ( USFLDC DOI )
k26.159 ( USFLDC Handle )
5220 ( 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:

Layla Lakes, Saudi Arabia: The World-Wide Largest Lacustrine Gypsum Tufas / Stephan Kempe - Heiko Dirks ( .pdf )

Flank Margin Cave Development In Telogenetic Limestones Of New Zealand / John E. Mylroie - Joan R. Mylroie - Campbell S. Nelson ( .pdf )

Cave Turbidites / R. Armstrong L. Osborne ( .pdf )

Broken Speleothems As Indicators Of Tectonic Movements / Stanka Å ebela ( .pdf )

Cave Sediments From The Postojnska-Planinska Cave System (Slovenia): Evidence Of Multi-Phase Evolution In Epiphreatic Zone / Nadja Zupan Hajna - Petr Pruner - Andrej Mihevc - Petr Schnabl - Pavel Bosák ( .pdf )

Karst In France And Unesco World Heritage / Patric Cabrol - Alain Mangin ( .pdf )

The Flow Rate Of The Fonte Pliniana (Como, Italy): Two Thousands Years Of Data / Arrigo A. Cigna ( .pdf )

Hydrologic Connections And Dynamics Of Water Movement In The Classical Karst (Kras) Aquifer: Evidence From Frequent Chemical And Stable Isotope Sampling / Daniel H. Doctor ( .pdf )

Groundwater Flow In Crystalline Carbonates (Jeseniky Mts., Czech Rep.): Using Stream Thermometry And Groundwater Balance For Catchment Delineation / Viola Altová ( .pdf )

Investigation Of Structure Of Various Surface Karst Formations In Limestone And Dolomite Bedrock With Application Of The Electrical Resistivity Imaging / Uroš Stepišnik - Andrej Mihevc ( .pdf )

Studies Of The Fauna Of Percolation Water Of Huda Luknja, A Cave In Isolated Karst In Northeast Slovenia / Tanja Pipan - Vesna Navodnik - Franc Janžekovic - Tone Novak ( .pdf )

Microorganisms In Hypogeon: Examples From Slovenian Karst Caves / Janez Mulec ( .pdf )


Full Text

PAGE 1

L A Y LA L AKES S AUDI A RABIA : T HE W ORLD W IDE LARGEST LACUSTRINE G Y PSUM TUFAS J EZERA L A Y LA V S AUDSKI A RABIJI : NAJOBSE NEJE PODRO JE JEZERSKEGA SADRINEGA TUFA NA SVETU Step h an K EMPE 1 & Heiko D IRKS 2 Izvleek UDK 911.2(532):553.556 Stephan Kempe & Heiko Dirks: Jezera Layla v Saudski Arabi ji: najobseneje podroje jezerskega sadrinega tufa na svetu Skozi vso osrednjo Saudsko Arabijo, od severa proti jugu, je mogoe slediti zgornjejurski Heet h formaciji. Lokalno jo prekinjajo hipogene vrtae. Najpomembneja so nekdanja jezera Layla na 22,170 severne zemljepisne irine in 46,700 vzh odne zemljepisne doline. Jezera (prvotni h 17) so bila v 80-ti h leti h prejnjega stoletja izpraznjena in pokazalo se je 19 vrta. Nekat ere sestavlja ve lokalni h podroij pogrezanja. Najveja vrtaa je 1,1 km dolga, 0,4 km iroka in okoli 40 m globoka. Ostale imajo manj kot 10 m premera in so precej mlaje. Dna nekdan ji h jezer in uravnave okoli nji h sestavljajo debele plasti drob nozrnate jezerske krede. Pri te h vrtaa h najbolj preseneajo ve metrov debele plasti oborine sadre, ki prekrivajo stene vrta in so nastale pod vodo. V spodnjem delu je v obliki debeli h go moljev, ki v srednjem delu dobivajo konino obliko, proti vr h u, to je proti nekdanji jezerski gladini, pa pre h ajajo v ponvaste, lebaste in lopataste oblike. Mineralogija in morfologija te h oborin sta edinstveni v svetovnem merilu. Kljune besede: Layla jezera, Saudska Arabija, Heet h formaci ja, hipogeni kras, vrtae, le hnjak, sadra, sadrna oborina, nek danja jezera. 1 Institut fr Angewandte Geowissensc h aen, Tec hnisc h e Universitt Darmstadt, Sc hnittspa hnstrae 9, D-64287 Darmstadt, kempe@geo.tu-darmstadt.de 2 GTZ International Services, Riyad h, h eiko.dirks@gtzdco-ksa.com Received/Prejeto: 21.03.2008 COBISS: 1.01 ACTA CARSOLOGICA 37/1, 7-14, POSTOJNA 2008 Abstract UDC 911.2(532):553.556 Stephan Kempe & Heiko Dirks: Layla Lakes, Saudi Arabia: e world-wide largest lacustrine gypsum tufas roug h out t h e center of Saudi Arabia t h e an h ydrite upper Jurassic Heet h formation can be followed N to S. Locally it is punctured by h ypogene karst sink h oles. e most prominent are t h e former Layla Lakes at 22.17N 46.70E. e lakes (17 originally) h ave been drained in t h e late 1980s, revealing 19 sink h oles, some of t h em composites of several subsidence centers. e largest is 1.1 km long, 0.4 km wide and about 40 m deep. Ot h ers are less t h an 10 m across and rat h er recent. e bottom of t h e former lakes and t h e ats around t h em are composed of t hick layers of ne-grained lake c h alks. e most striking feature of t h ese sink h oles is t h e several meters t hick tufa covering t h e vertical walls of t h e sink h oles. It formed subaqueous and is entirely composed of gypsum. Morp h ologically t h e tufa displays t hick bulbous forms at t h e bottom c h anging to conical forms at middle dept h to gour-, gutter-, or s h ovellike forms near to t h e former lake surface. e mineralogy and morp h ology of t his tufa appear to be singular world-wide. Key words: Layla Lakes, Saudi Arabia, Heet h Formation, h ypo gene karst, sink h oles, tufa, gypsum tufa, paleolakes. I NTRODUCTION e Arabian plate is composed of crystalline basement in t h e west and of eastward dipping P h anerozoic strata in t h e east. In t h e center of t h e Kingdom of Saudi Arabia (KSA), a band of Jurassic and Cretaceous sediments crop out running nearly N-S t hroug h out t h e entire country. Prominent escarpments are formed by middle Jurassic and lower Cretaceous limestone. e N-S extending plain in between t h e escarpments is partly formed by t h e upper Jurassic Heet h (or Hit h) formation composed of a >150 m t hick sequence of laminated and autobrecci ated an h ydrite forms a N-S extending plain. On it most of t h e inner-KSA cities are situated, including Riyad h, t h e

PAGE 2

ACTA CARSOLOGICA 37/1 2008 8 capital of KSA wit h four million in h abitants. is band of an h ydrite can be followed on satellite images, suc h as t h ose provided by Google Eart h, as a string of brig h t, al most w hite areas. Ground inspection s h owed t h at t h ese areas are marked by gypsum calic h e, presumably formed by ascending waters t h at le gypsum upon evaporation. e Heet h Formation is an aquitard, below w hic h fossil groundwater is trapped t h at is extensively used for domestic and agricultural purposes. Consequently t h e ground water level h as dropped in many areas by >100 m. is is best illustrated by t h e history of t h e deepest cave in KSA, Ain Heet h, a >160 m deep cave at t h e type location of t h e Heet h Formation near t h e town of Al-K h arj ca. 60 km sout h of Riyad h. e cave apparently formed by upward solution of t h e groundwater body in a h ypogene setting sensu Klimc h ouk (2007, Fig. 16). In t h e 1930s t h e cave was a spring, allowing t h e deep groundwater to ow out freely. In t h e 1980s t h e cave formed a pool, of ten visited by locals for picnics. en a pump h ouse was installed and t h e water was used locally. In addition deep wells in t h e surrounding area tapped t h e underlying aquifer. In 2002 t h e lake h ad receded to a dept h of 137 m (pers. comm. Greg Gregory) and divers explored a large c h amber and a h orizontal slowly, descending passage at its bottom. During t h e visits of t h e aut h ors on February 19 t h 2008, t h e large c h amber (up to 70 m long and 20 m wide) at a dept h of ca. 145 m was accessible, wit h t h e groundwater surface forming a lake at its bottom. e h orizontal passages apparently were not yet free of water. e cave walls s h ow t h e morp h ology typical of a convec tive cave formation in a p hreatic setting in gypsum (e.g. Kempe, 2008). As one descends steeply over t h e boulders of t h e cave oor, one passes t hroug h almost all of t h e Heet h Formation, t h us making it t h e only outcrop w h ere it can be studied in detail. Above t h e entrance of t h e cave t h e transgressive contact of lower Cretaceous marl and platy limestone is well displayed. Similar h ypogene karstication appears in ot h er places as well and h as led to sink h oles. One of t h e areas is around t h e town of As Sulayyil, 500 km sout h of Riyad h, were several sink h oles h ave opened up. At least two h ave recently been lled by t h e farmers, but one rat h er recent one (at Umm Sulaim; N20.42414 E45.66311), 47 m long and 27 m wide and about 1 m deep, was even vent ing h ot and h umid air t hroug h fres h circumferential and radial cracks, apparently rising from t h e deeper underly ing aquifer. Similar sink h oles are also reported from t h e area nort h of Riyad h. L A Y LA L AKES S INKHOLES e largest series of sink h oles is, h owever, found sout h east of t h e town of Layla, 300 km sout h of Riyad h. Most of t h em were lled by lakes until t h e mid 1980s (Fig. 1). A h otel wit h bungalows and restaurant was build over looking t h e largest lake in t h e sout h (Fig. 2). But it appar ently never went into operation because h ere t h e same h appened as in Ain Heet h: water was abstracted not only by pumping it directly from t h e nort h ern end of t h e lake but also from t h e deep aquifer below. In t h e early 1990s t h e lakes dried up, revealing a series of text-book sink h oles (Fig. 3). Google Eart h provides a hig h resolution view of t h e nort h ern sink h oles, w hile t h ose in t h e sout h are barely perceptible once t h e ground situation is known (Fig. 4). Along t h e eastern side a prominent escarpment is vis ible, w hic h is as muc h as 10 m hig h in places. Ground inspection s h ows, t h at it is accompanied by small gra ben structures wit h open cracks and small tectonic caves. Inspection also reveals t h at t h e rocks displaced are un consolidated marls and not Mesozoic rocks. All in all (Table 1) 23 sink h oles can be listed of w hic h some h ave several subsidence centers (Nos. 2, 4, 18 and 22). W e were able to visit most of t h e sink h oles on February 21 st and 25 t h in order to inspect t h em for tufa occurrence. e largest sink h ole is No. 4 wit h an N-S axis of 1.1 km and a widt h of 0.4 km and a dept h of up to 40 m. e small est ones h ave openings of less t h an 10 m across, but are bellowing out below, forming real sink h ole caves. Two of t h e sink h oles (18 and 19) are connected by a natural bridge. For a few sink h oles, we were not able to esti mate t h e dept h, because treir bottom was not visible. In F ig. 1: Two of the smaller Layla Lakes still lled with water in the beginning of the 1980ies (M inistry of Agriculture and Water, 1984). STEPHAN K EMPE & H EIKO D IRKS

PAGE 3

ACTA CARSOLOGICA 37/1 2008 9 general t h e small, over h anging sink h oles are extremely dangerous because t h eir margins are composed of loose material (lake marls). Many appear to h ave opened rat h er recently because t h ey lack signs of tufa. All in all 19 of t h e structures (including t h e sub-sink h oles) h ave tufa, i.e. t h ey are older and h ad standing water in t h em. e Google Eart h images also revealed t h at t h e lakes served as natural outlets of t h e deeper aquifer in as muc h as a series of parallel c h annels and qanats (Arabic for subterraneous water c h annels) conducted water over about 5 km to farms at Al Say h in t h e nort h. is usage of t h e lake water was sustainable, since t h e qanats did only allow for gravitational outow, t h us only water t h at natu rally owed out of t h e aquifer was consumed. Apparently most of t h e sink h oles were originally covered by water and t h us not known at all. e water At Table 1: List of sinkholes near Layla. Sizes according to Google Earth and own eld inspection. No North East Size (m) Depth (m) Type Remarks (T = Tufa) Qanat 1 22.1689 46.7166 200*50 ca. 3-10 elongated T, fresh circular cracks no 2 22.1626 46.7034 400*100 ca. 5-30 elongated T, fresh cracks; pit caves no 2a 22.1628 46.7024 10*10 >20 narrow pit T no 2b 22.1626 46.7045 50*50 ca. 10 circular T no 2c 22.1632 46.4043 30*30 ca. 10 circular T, fresh cracks no 3 22.1634 46.7019 60*60 ca. 15 circular post-lake no 4 22.1681 46.7087 1100*400 ca. 30 elongated T, former main lake, fresh cracks, terraces channel 4a 22.1682 46.7077 50*50 ca. 10 circular T, sand dune no 4b 22.1686 46.7073 15*15 ca. 10 circular T, sand dune at bottom no 5 22.1706 46.7087 50*40 >20 circular T ?, bottom not visible no 6 22.1715 46.7087 60*60 ca. 30 circular T, undercut, caves? no 7 22.1744 46.7068 60*60 ? circular not visited no 8 22.1774 46.7133 70*70 ca. 40 circular T no 9 22.1771 46.7140 9*8 ca. 10 circular overhanging, with sand pile no 10 22.1769 46.7139 15*13 6 circular overhanging, with sand pile no 11 22.1776 46.7142 7*10 >25 circular overhanging 12 22.1785 46.7143 15*15 5 circular half lled by dunes no 13 22.1786 46.7146 16*14 6 circular half lled by dunes no 14 22.1807 46.7157 105*90 ca.40 irregular T, sand dunes yes 15 22.1820 46.7178 50*50 ca.50 circular T yes 16 22.1840 46.7256 100*110 35 irregular T, collapse blocks yes to 15 (80 m) 17 22.1853 46.7241 15*15 30 circular overhanging no 18 22.1865 46.7257 315*110 up to 25 elongated T, composite of 3 sinkholes channel to 17 (130 m) 18a 22.1857 46.7254 95*70 ca. 15 elongated T, terraces 18b 22.1870 46.7257 112*85 ca. 15 irregular T, terraces 18c 22.1872 46.7268 95*80 ca. 25 irregular T, collapse blocks 19 22.1869 46.7273 51*25 ca. 20 elongated T, natural bridge with 15c yes 20 22.1884 46.7280 38*38 ca. 20 irregular T, pit channel to 19 (160 m) 21 22.1880 46.7287 26*22 ? elongated not visited yes 22 22.1897 46.7294 60*35 ? elongated not visited channel 22a 22.1897 46.7292 20*18 ? circular not visited 23 22.2097 46.7345 48*38 ? circular not visited, half lled by dunes yes (?) L A Y LA L AKES S AUDI A RABIA : T HE W ORLD W IDE LARGEST LACUSTRINE G Y PSUM TUFAS

PAGE 4

ACTA CARSOLOGICA 37/1 2008 10 las of Saudi Arabia lists 17 lakes (Ministry of Agriculture and W ater, 1984). It also lists t h e electrical conductivities of t h e lake water t h at ranged from 2510 to 8600 S/cm, i.e. t h e values are hig h er t h an expected for carbonate satu rated waters. Some of t h e values appear to h ave been even hig h er t h an for gypsum (or an h ydrite) saturated waters, but t his may be due to ongoing evaporative concentra tion in s h allow lakes. us t h e lake water most probably was hig h in sulp h ate as well as carbonate. is may ex plain w h y we only found two species of gastropods, quite a small number of species of molluscs for a fres h water h abitat. e s h ells of one of t h e species are ubiquitous (Radix cf. natalensis, Krauss, 1848; Neubert, 1998; pers. com. H.-J. Nieder h fer, Stuttgart; also known as Lym F ig. 2: e Layla Lake Hotel overlooking the dried-up lakes in the southern section of the area. F ig. 4: Google Earth view of the Layla Lakes area and annotated map of sinkholes. STEPHAN K EMPE & H EIKO D IRKS

PAGE 5

ACTA CARSOLOGICA 37/1 2008 11 very ne-grained marls or lake c h alk (seekreide). Since t h e draining of t h e lakes, t h ese deposits h ave dried out and s hrank. Numerous meter-deep s hrinkage cracks crisscross t h e lake bottoms and surround t h e sink h oles. ey can not be dierentiated from cracks caused by on going subsidence. e s hrinkage of t h e lake bottom sedi ments h as caused t h eir subsidence; in places s hrinkage amounted to more t h an a meter (Fig. 5) and opened over 10 m deep s hrinkage caves between t h e sediment and t h e solid tufa (for example at t h e sout h ern end of Sink h ole 2). naea arabica; a very common fres h water snail) w hile t h e second one, (t h e cerit hiid Melanoides cf. tuberculata, O. F. Mller, 1774; Krupp & Sc hneider, 1988; Neubert, 1998; pers. com. H.-J. Nieder h fer, Stuttgart; t h e species is salt tolerant and lives in warm waters), was found only in one place (h alf-way down in Sink h ole 2). e s h allow lake bottoms (around t h e sink h oles) was occupied by reeds, t h e roots of w hic h are still noticed everyw h ere. e bottom of t h ose sink h oles t h at contained lakes as well as t h e ats around t h e sink h oles is composed of F ig. 3: Sinkhole No. 6, note tufa at bottom of lake and layered lake chalk prole near the top (car for scale). F ig. 5: Shrinkage of lake-bottom chalk deposits has caused their subsidence by over one meter. e former contact of the sediment with the solid tufa is marked by a small ledge (Sinkhole 2, eastern side). L A Y LA L AKES GY PSUM T UFAS e most important discovery, h owever, made in t h e Layla Lakes is t h e magnicent tufa t h at covers t h e verti cal walls of t h e lake sink h oles (Fig. 6). Repeated test wit h 2nHCl, ngernail scratc hing and macroscopic inspection s h owed t h at t h e tufa is composed entirely of gypsum. In it numerous gastropod s h ells can be found immured at places. Even t h oug h t h e tufa surface is covered wit h negrained, cream-colored calcareous dust, t h e tufa itself is a sparitic selenite or a massive, ne-grained gypsum. Since no suc h site h as to our knowledge been previously described in t h e literature, t h e term tufa is used even t h oug h it h as so far exclusively been used for calcareous deposits (compare Ford & Pedley, 1996). e term tra vertine is avoided because it is more commonly used for h ydrot h ermal and hig h p CO 2 sources of calcareous deposits. e expectation to nd calcareous tufa or stro matolitic microbialites, as is common in CaCO 3 -super saturated lakes suc h as Plitvice (Kempe & Emeis, 1985) or Mono Lake, Pyramid Lake, W alker Lake and Searles Lake in t h e western USA (compare Kempe & Kazmierc zak, 2008), Lake Van in eastern Anatolia (Kempe et al., 1991), or in crater lakes suc h as Vai La hi on Niuafoou (Kazmierczak & Kempe, 2006), was not met. An over view over calcareous tufa and travertine deposition in low temperature environments was given by Ford & Ped ley (1996). Apart from t h e singular mineralogy is t h e morp h ol ogy of t h e tufa t h e most striking feature: t h e 20 to 30 m hig h walls of t h e sink h oles are covered wit h a several me ter t hick crust of tufa. ere is a distinct morp h ological L A Y LA L AKES S AUDI A RABIA : T HE W ORLD W IDE LARGEST LACUSTRINE G Y PSUM TUFAS

PAGE 6

ACTA CARSOLOGICA 37/1 2008 12 F ig. 7: P anorama view of the eastern wall of Sinkhole 2, showing clear vertical change in the morphology of the tufa. F ig. 6: V iew of the northern section of Sinkhole 2 from under neath the solid tufa showing the vertical tufa walls. F ig. 8: Upward oriented, shovel-like cups of tufa, typical for the forms near to the former lake surface (Sinkhole 2, western wall, ca. 7,4 m). F ig. 9: Shell-like basin reminding of the large sea-shell in B oticellis V enus (le, with one of the authors standing in it) and cup-like tufa (Sinkhole 5, southern wall). Note the terraces le by the sink ing water level in the foreground. F ig. 11: Tufa gours at the northern wall of Sinkhole 4b suggesting origin of tufa by water cascading down from the up-slope side of the sinkhole. STEPHAN K EMPE & H EIKO D IRKS

PAGE 7

ACTA CARSOLOGICA 37/1 2008 13 tially lled wit h loose sediment. In a sense, t h ey remind of speleot h em gours. In places w h ere t h e sink h ole does not display vertical walls, suc h as in t h e center of Sink h ole 4, we nd rat h er regular rows of cup-like structures forming steps (Fig. 10). A specically regular tufa display was found at t h e nort h ern wall of Sink h ole 4b (Fig. 11). Overall t his triple division of t h e gross morp h ology of t h e tufa is found in all of t h e sink h oles. c h ange from bottom to top (example Sink h ole 2, eastern wall; Fig. 7): At t h e bottom t h e crust forms a solid, over h anging panel wit h only s h allow vertical groves. Above, t h e crust is more segmented into m-sized bulbous, in verted cones and furt h er up (le on Fig. 7) t h e tufa forms protruding, upward-oriented, s h ovel-like bowls or cups (Fig. 8). Some of t h e cups are more delicate t h an in ot h er places and can form large s h ell-like basins (Fig. 9). e rims are oen less t h an 10 cm t hick and t h e cups are par L A Y LA L AKES S AUDI A RABIA : T HE W ORLD W IDE LARGEST LACUSTRINE G Y PSUM TUFAS F ig. 10: P anorama of a step-like assemblage of gour-like tufa forms in the northern part of Sinkhole 4. Since the lakes dried up, sand started to dri in. Note the ripple marks on the dune slopes accentuated by white gypsum dust (length of view ca. 50 m). It is clear, t h at we will never be able to fully understand t h e genesis of t his special form of tufa because t h e lakes and t h eir p h ysico-c h emistry in w hic h t h ey grew are gone forever. W e t h us can only speculate on t h ese conditions. e observation t h at t h e conductivities were hig h, i.e. in or even above t h e range of gypsum-saturation, and from t h e fact t h at t h ese tufa forms exist, we must con clude, t h at t h e Layla Lakes were saturated wit h respect to gypsum. In contrast to calcite, gypsum cannot be hig h ly supersaturated but precipitates at saturation. Apart from evaporation two more processes govern gypsum satura tion, best explained by W igley (1973) and exemplied for a gypsum karst setting by Brandt et al. (1976). ese processes are: co-dissolution or co-precipitation of calcite and temperature alteration. us several processes can be discussed t h at may h ave caused t h e gypsum precipitation along t h e walls of t h e lakes: a) Concentration by evaporation. It is conceivable t h at on t h e wide ats t h at surrounded t h e deeper sink h ole centers of t h e lakes evaporation would concentrate t h e water muc h faster t h an over t h e deep sink h ole sec tions of t h e lakes. Hig h er concentrated solutions may h ave run down along t h e oor and cascaded underwa ter over t h e lips of t h e walls of t h e sink h ole, bringing saturated solutions into contact wit h t h e crystals already growing on t h e walls (precipitation upon reac h ing nu cleation sites). b) Calcite co-dissolution. Concentration on t h e ats by evaporation and degassing of CO 2 causes a precipi tation of CaCO 3 (calcite and/or aragonite). Downward cascading solutions bring t h e water into deeper water layers wit h a hig h er p CO 2 ere, t h e co-transported ne-grained lake c h alk is being dissolved, pus hing (be cause of t h e common ion eect of t h e increasing calcium activity) t h e gypsum over t h e saturation limit and caus ing its crystallisation along t h e walls. c) Temperature change. Ascending water from t h e deep aquifer, delivering gypsum-saturated solutions warmer t h an t h e lake temperatures, cools in t h e lake and gypsum crystallises along t h e walls. e possibility exists t h at all t hree of t h e processes may be applicable, eac h in a dierent season of t h e year. In any case is t h e continued delivery from t h e underly ing aquifer of nearly gypsum saturated water t h e key condition. ese solutions may eit h er h ave risen slowly t hroug h t h e sediments of t h e lake bottoms or t hroug h distinct vents. One of suc h vents may h ave been adjacent to Sink h ole 2. Anot h er vent may h ave been located at t h e NW -end of Sink h ole 4 and in t h e western wall of Sink h ole 20. e situation seen today is t h e product of a long geological history. At t his point one can only guess w h at t his history is. One of t h e clues is t h e fact t h at all of t h e sink h oles seem to occur in unconsolidated lake c h alks. G ENETIC C ONSIDERATIONS

PAGE 8

ACTA CARSOLOGICA 37/1 2008 14 A CKNO W LEDGEMENTS e aut h ors are grateful to Dr. Randolf Rausc h, Tec hnical Director for GTZ International Services at Riyad h, w h o not only advised us to visit Layla Lakes but also nanced t h e stay of t h e rst aut h or in KSA in February 2008. e aut h ors t h ank also Paolo Forti, Bologna, for h elpful re marks and H.-J. Nieder h fer, Stuttgart, for his determi nation of t h e gastropod s h ells. STEPHAN K EMPE & H EIKO D IRKS us t h e outow of deep aquifer water over long peri ods rells any subsidence h oles wit h c h alk. During times of wetter climate t h e groundwater is rec h arged and t h e lakes expand. During dry climate periods, t h e ground water ows out only sparingly and t h e lakes retreat to t h e immediate vicinity of t h e sink h oles. Wh ere t h e gypsum tufa h as collapsed from sink h ole walls, we nd over 10 m hig h proles of laminated or layered lacustrine c h alks. Similar, albeit older, deposits are exposed along t h e east ern escarpment. ese oer t h e possibility to do paleo climate researc h, an aim we will try to pursue next. e apparently dozens of meter t hick calcareous c h alks (lime muds) s h ould be listed as an own Pleistocene formation: t h e Layla Lakes Marls. Brandt, A., Kempe, S., Seeger, M., & Vladi, F., 1975: Geoc h emie, Hydrograp hie und Morp h ogen ese des Gipskarstgebietes von Dna/Sd h arz. Sc h weizerbartsc h e Verlagsbuc hh andlung, Stuttgart, Geol. Ja hrbuc h, Rei h e C, 45: 55 pp. Ford, T.D. & Pedley, H.M., 1996: A review of tufa and travertine deposits of t h e world. Eart h Science Re views 41: 117-175. Kazmierczak, J. & Kempe, S., 2006: Modern analogues of Precambrian stromatolites from caldera lakes of Niuafoou Island, Tonga. Naturwissensc h aen 93: 119-126. Kempe, S., 2008: Vom Urkanal zur unterirdisc h en Kat h edrale, H h lenformen und i hre Entste h ung. In: Kempe, S. & Rosenda h l, W (eds.), H h len: Verbor gene W elten; W issensc h alic h e Buc h gesellsc h a Darmstadt, Darmstadt: 54-64. Kempe, S. & Emeis, K., 1985: Carbonate c h emistry and t h e formation of Plitvice Lakes. In: Degens, E.T., Kempe, S. & Herrera, R. (eds.) Transport of Car bon and Minerals in Major W orld Rivers, Pt. 3, Mitt. Geol.-Palont. Inst. Univ. Hamburg, SCOPE/ UNEP Sonderband 58: 351-383. Kempe, S., Kazmierczak, J., Landmann, G., Konuk, T., Reimer, A. & Lipp, A., 1991: Largest known micro bialites discovered in Lake Van, Turkey. Nature 349: 605-608. Kempe, S., & Kazmierczak., J., 2008: Soda lakes. In: Re itner, J., and iel, V. (eds.) Encyclopedia of Geobi ology, Springer, submitted. Klimc h ouk, A., 2007: Hypogene Speleogenesis, Hydro geological and Morp h ogenetic Perspective. Nat. Cave and Karst Res. Inst. Spec. Pap. No 1, 106 pp. Krupp, F. & Sc hneider, W ., 1988: Die Swasserfauna des Vorderen Orients. Natur und Museum 118: 193213. Ministry of Agriculture and W ater of t h e Kingdom of Saudi Arabia, 1984: W ater Atlas of Saudi Arabia. Saudi Arabian Printing Company, 112 pp. Neubert, E., 1998: Annotated c h ecklist of t h e terrestrial and fres h water molluscs of t h e Arabian Peninsula wit h descriptions of new species. Fauna of Arabica 17: 333-461; Basle, Switzerland. W igley, T.M.L., 1973: C h emical solution of t h e system calcite-gypsum-water. Canadian J. Eart h Sci. 10: 306-315. R EFERENCES



PAGE 1

F LANK M ARGIN C AVE D EVELOPMENT IN T ELOGENETIC L IMESTONES OF N E W Z EALAND J AME TIPA F LANK MARGIN V TELOGENETSKIH APNENCIH N OVE Z ELANDIJE Jo hn E. MY LROIE 1 Joan R. MY LROIE 1 & Campbell S. N ELSON 2 Izvleek UDK 552.54:551.3.051(931) John E. Mylroie, Joan R. Mylroie & Campbell S. Nelson: Jame tipa Flank margin v telogenetskih apnencih Nove Zelandije Z namenom, da odkrijemo prisotnost jam nastali h v prio balnem pasu meanja slane in sladke vode (jame tipa Flank margin), smo raziskovali diagenetsko zrele apnence Sever nega (pristanie Raglan in Kaw hia, drage Napier in W aipu) in Junega otoka (Po h ara, reka Paturau, Piunakaiki, Kakanui in Kaikoura) Nove Zelandije. Na obalni h podroji h so jame v karbonati h la h ko rezultat razlini h psevdokraki h (npr. erozija valovanja) in kraki h procesov, ki niso povezani z meanjem slane in sladke vode (npr. epikrake oblike in jame nastale z me teorno krako drenao). Jame tipa Flank margin smo uspeno loili od ostali h tipov jam z opazovanjem naslednji h znailnosti: morfologija freatini h skalni h oblik v razlini h merili h, odsot nost skalni h oblik in sedimentov, ki so rezultat hitrega turbu lentnega toka, nepovezanost sosednji h jam z zvezno vodno potjo. Zaradi aktivne tektonike je nivo morske gladine na Novi Zelandiji zelo spremenljiv, obdobja stabilnega vodnega nivoja so kratka, zato so jame tipa Flank margin na Novi Zelandiji manje v primerjavi s tistimi v tektonsko stabilneji h okolji h, kot so Ba h ami, kjer nivo morske gladine uravnava zgolj gla cioevstazija. Poudarjen vertikalni razvoj jam tipa Flank mar gin kae na poasno in enakomerno tektonsko dvigovanje, raziritve in kanali v diskretni h h orizonti h, ki sekajo strukturo, pa kaejo na obasna (nezvezna) tektonska dogajanja. Jame tipa Flank margin nosijo neodvisen zapis o trajanju in hitrosti dvigovanja obmoij na Novi Zelandiji in tako nudijo monost umeritve ostali h metod vrednotenja tektonski h dogodkov. Kljune besede: jame tipa Flank margin, Nova Zelandija, telogenetski karbonati, priobalne jame. 1 Department of Geosciences, Mississippi State University, Mississippi State, MS 39762, USA Fax: (662) 325 9423, Email: mylroie@geosci.msstate.edu 2 Department of Eart h and Ocean Sciences, e University of W aikato, Private Bag 3105, Hamilton, New Zealand Fax: 0064-7-856 0115, Email: c.nelson@waikato.ac.nz Received/Prejeto: 30.01.2008 COBISS: 1.01 ACTA CARSOLOGICA 37/1, 15-40, POSTOJNA 2008 Abstract UDC 552.54:551.3.051(931) John E. Mylroie, Joan R. Mylroie & Campbell S. Nelson: Flank Margin Cave Development in Telogenetic Limestones of New Zealand Coastal limestone outcrops, typically wit h advanced levels of diagenetic maturity (i.e., are telogenetic carbonates), were ex amined on Nort h Island (Raglan Harbour, Kaw hia Harbour, Napier, and W aipu Cove) and Sout h Island (Po h ara, Paturau River, Punakaiki, Kakanui, and Kaikoura), New Zealand, to de termine if ank margin caves, produced by mixing dissolution, were present. In coastal settings, caves in carbonate rock can be t h e outcome of pseudokarst process, primarily wave erosion, as well as karst processes not associated wit h fres h and sea-water mixing suc h as epikarst features and conduit-ow stream caves. Flank margin caves were successfully dierentiated from ot h er cave types by t h e following criteria: p hreatic dissolutional mor p h ologies at t h e wall rock and c h amber scales; absence of hig hvelocity, turbulent-ow wall sculpture and sediment deposits; and lack of integration of adjacent caves into a continuous ow pat h. e active tectonics of New Zealand creates a variable sealevel situation. e relatively s h ort time of sea-level stability lim its t h e size of t h e New Zealand ank margin caves compared to tectonically-stable environments, suc h as t h e Ba h amas, w h ere glacioeustasy alone controls sea-level stability. Upli events can be identied as slow and steady w h en t h e ank margin caves are uniformly elongated in t h e vertical direction, and episodic w h en t h e ank margin caves s h ow widening and tube develop ment at discrete h orizons t h at cut across rock structure. New Zealand ank margin caves contain information on upli dura tion and rates independent of ot h er commonly used measures, and t h erefore can provide a calibration to ot h er met h ods. Key Words: ank margin caves, New Zealand, telogenetic car bonates, coastal caves.

PAGE 2

ACTA CARSOLOGICA 37/1 2008 16 It h as been recognized for many years t h at t h e mixing of marine and fres h waters in coastal settings produces unique dissolutional features on bot h t h e surface (e.g. Folk et al. 1973; Taboroi et al., 2004) and in t h e sub surface (e.g. Back et al., 1986; Mylroie & Carew, 1995) of carbonate rocks. One particular dissolutional feature, t h e ank margin cave, h as been identied from coastal carbonates in a wide variety of settings, including t h e Ba h amas (Mylroie & Carew, 1990), Isla de Mona in Puerto Rico (Frank et al., 1998), Y ucatan (Kelley et al., 2006), t h e Mariana Islands (Jenson et al., 2006), and Australia (White et al., 2007). A common factor in all of t h ese lo cations is t h at t h e ank margin caves h ave developed in young carbonates t h at are diagenetically immature, or eogenetic ( sensu C h oquette & Pray, 1970; see also Vac h er & Mylroie, 2002). Very little work h as been done on ank margin caves in diagenetically mature, or telogenetic, carbonate rocks. One possible example is at Berry Head in Devon, UK (Proctor, 1988), w h ere mixing dissolution in Paleozoic carbonates at a coastal setting may h ave re sulted in cave development. e paucity of study of telo genetic mixing dissolution caves can be in part attributed to t h e relative scarcity of telogenetic carbonate rocks in coastal settings, and in part attributed to t h e inappropri ate application of traditional continental conduit-ow stream cave dissolution models in coastal settings by ear lier workers. Eogenetic carbonate rocks are generally found close to t h eir environment of deposition, so suc h carbon ate rocks are usually found in coastal settings around t h e world. Carbonate islands in particular h ave oered a natural laboratory for study of mixing dissolution, as t h eir rocks are commonly very young, and t h e islands are commonly very small. Suc h constraints on time and space limit t h e possible interpretations of t h e dissolution features found on and in t h ose rocks. e eogenetic na ture of t h e carbonate rock results in very hig h primary porosity, 30% being a common value (Vac h er & Mylroie, 2002). is setting facilitates diuse ow in t h e carbon ate aquifer, and allows mixing of fres h and marine waters over a broad area. As establis h ed for vadose and p hre atic fres h water by Bgli (1980), and for saline and fres h water by Plummer (1975), mixing of waters t h at h ave saturated wit h respect to CaCO 3 can result in more dis solution, if t h e waters became saturated at dierent initial conditions (Dreybrodt 2000). In carbonate islands wit h a fres h-water lens, t h e maximum dissolution occurs in t h e distal margin of t h e lens, under t h e ank of t h e enclosing landmass. e caves t h at form in t his location are t h ere fore called ank margin caves (Mylroie & Carew, 1990; Neundorf et al., 2005, p. 241). e top of t h e fres h-water lens is a site of mixing of vadose fres h water and p hreatic fres h water. e bottom of t h e lens is a site of marine wa ter and fres h water mixing. Bot h mixing environments are capable of creating renewed dissolutional aggressiv ity, and at t h e lens margin t h ey are superimposed upon eac h ot h er to create an additive dissolutional eect. Fur t h er dissolutional potential is created by organic material t h at collects at t h e density contrasts provided by bot h t h e top and bottom of t h e lens. ese organics can decay, re leasing CO 2 to promote furt h er CaCO 3 dissolution, and w h en present in abundance, t h ese organics may lead to anoxic conditions and H 2 S-mediated dissolution (Bot trell et al., 1993). Finally, t h e t hinning of t h e lens at its distal margin results in a decrease in lens cross-sectional area, and a consequent increase in water ow velocities (Raeisi & Mylroie, 1995; Moore et al., 2007). Reactants ow in, and products ow out, faster at t h e distal margin t h an elsew h ere in t h e lens. e ultimate outcome of all t h ese geoc h emical and h ydrologic parameters is t h e de velopment of large dissolutional voids in a s h ort amount of time at a specic location: t h e ank margin cave. Flank margin caves in eogenetic carbonate islands h ave a c h ar acteristic morp h ology of irregular c h ambers, maze areas, dead-end passages, and a h orizontal dimension many times t h eir vertical dimension (Fig. 1). e initial work on ank margin caves was conduct ed in t h e Ba h amas, w h ere t h e h ost eogenetic carbonate rock was mid to late Quaternary eolianites. is rock is almost an ideal, simplistic carbonate material, as its de position as a terrestrial limestone means it lacks marine cements, h as never been buried or loaded, is well sorted, and is free of secondary structures suc h as faults, folds or joints. e rst ank margin caves described were above modern sea level. In a tectonically-stable environ ment suc h as t h e Ba h amas, elevation of t h e fres h-water lens above modern sea level in t h ese eolianites could only h ave been accomplis h ed by glacioeustasy during interglacials. e rate of Ba h amian platform subsidence is 1 to 2 m per 100 ka (Carew & Mylroie, 1995; McNeill, 2005). In t h e Atlantic Basin, only t h e last interglacial or Marine Isotope Stage (MIS) 5e, from about 131 to 119 ka (C h en et al., 1991), could h ave placed sea level, and t h e fres h-water lens supported by t h at sea level, at t h e 6 m elevation required by existing ank margin caves. Ear lier interglacials were eit h er not hig h enoug h (MIS 7), or too far back in time, given isostatic subsidence rates (MIS 9 or 11), to h ave created t h e necessary elevated fres h-water lens position and caves as observed today (Carew & Mylroie, 1995). e time window in w hic h to form ank margin caves, w hic h commonly h ave volumes in excess of 25,000 m 3 was only t h e 12,000 years dura I NTRODUCTION J OHN E. MY LROIE J OAN R. MY LROIE & C AMPBELL S. N ELSON

PAGE 3

ACTA CARSOLOGICA 37/1 2008 17 tion of t h e last interglacial. During t h e last interglacial, w h en sea level was up to 6 m hig h er t h an at present in t h e Ba h amas, t h e land area was muc h smaller t h an today, and consisted mostly of eolian ridges acting as small, in dependent islands. Flank margin caves in t h e Ba h amas t h erefore developed under severe constraints of time and space: in 12,000 years in fres h-water lenses of areal extent as small as a few h ectares. Researc h into ank margin cave development sub sequently expanded into ot h er carbonate island settings w h ere t h e geology was more complex, suc h as Isla de Mona in Puerto Rico (Frank et al., 1998) and t h e Mari ana Islands (Jenson et al., 2006). In t h ese cases t h e car bonate units, w hile older t h an in t h e Ba h amas, were still eogenetic, but t h e structural situation was more complex. Tectonics overprinted glacioeustasy, and faults and gentle folds were present, along wit h abundant jointing. e car bonate rocks were marine, diagenetically variable, and in t h e Marianas, juxtaposed wit h non-carbonate rocks. De spite t h ese additional complications, t h e general pattern of ank margin cave development as seen in t h e Ba h amas F ig. 1: A M ap of Salt P ond Cave on Long Island, B ahamas, a typical B ahamian ank margin cave. Note the attened, globular nature of the cave chambers, their irregular walls, and maze-like conguration. A note on map symbols for cave maps in this paper: rectilinear blocks represent breakdown (collapse material); lines with hachures indicate a vertical surface with hachures on the down side; three diverging lines indicate a slope, downward in the direction of divergence; black triangles indicate stalactites (apex down) and stalagmites (apex up); cross sections designated by labeled lines (A to A, etc.), or by lines extending from, but not touching, cave walls. B F lank margin cave passage in Cueva del Agua Sardinera, Isla de M ona, P uerto Rico. e cave passage is wider than it is high, and has wall cusps, bedrock pillars, and dissolutional morphology consistent with low-velocity, phreatic-ow conditions such as exist in ank margin caves. C F lank margin caves exposed in the Suicide Clis, Tinian, Commonwealth of the Northern M ariana Islands. Cli retreat has exposed a series of ank margin caves at a single horizon, which represents a stable position of the fresh-water lens prior to tectonic upli. Upper part of telephone pole in foreground (circled) for scale. persisted. Passage elongation parallel to major joints was one signicant dierence from t h e Ba h amian condition, as joints provided preferential ow pat h s in t h e fres hwater lens, and t h erefore mixing environments, in t h ese porous and permeable eogenetic carbonate rocks. e question of interest t h en became t h e outcome of mixing zone dissolution in dense, diagenetically ma ture, telogenetic carbonate rocks. Suc h rocks h ave little or no matrix permeability. In interior continental settings suc h rocks create t h e classic uvial karst wit h hig h-veloc ity, turbulent conduit ow; t h e epigenic caves of Palmer (1991). e permeability is almost entirely along bedding planes, joints, and faults, and conduits preferentially fol low t h ese fundamental ow pat h ways. It was reasonable to consider t h at ow in a fres h-water lens in telogenetic carbonate rocks in a coastal setting would be similarly constrained to bedding planes, joints, and faults. To test t h e reaction of mature telogenetic carbonate rocks to mixing dissolution in t h e distal margin of a fres h-water lens required a location w h ere t h ese rocks would be pres ent, preferably in many varied settings. F LANK M ARGIN C AVE D EVELOPMENT IN T ELOGENETIC L IMESTONES OF N E W Z EALAND

PAGE 4

ACTA CARSOLOGICA 37/1 2008 18 New Zealand, bot h Nort h and Sout h Island, contain a va riety of limestones cropping out in coastal locations in a multitude of settings. While t h ese rocks are geologically young, Cenozoic in age, t h ey are associated wit h a very active tectonic environment and h ave been subjected to a variety of diagenetic conditions, leading to a variety of diagenetic maturities between eogenetic and mainly te logenetic (Nelson, 1978; Nelson et al., 1988; Caron et al., 2006). e caves described in t his paper are entirely from telogenetic carbonate rocks. e active tectonic setting also means t h at bot h glacioeustasy and tectonic move ment h ave controlled t h e placement of sea level wit h re spect to t h e coastal carbonate outcrops. e key factor in t his regard is t h e duration of a stable sea-level position, w hic h in turn controls t h e time stability of t h e fres h-wa ter lens. Data from t h e Ba h amas (Mylroie & Mylroie, 2007a) and from Isla de Mona (Frank et al., 1998) indi cate t h at a constantly moving sea level, and a constantly moving fres h-water lens, allow too little time for mac roscopic dissolution to develop ank margin caves. On t h e ot h er h and, t h e Ba h amian data also s h ow t h at even limited lens-stability time can create observable ank margin caves. e tectonic setting of New Zealand oers an opportunity to determine w h at minimum sea-level stability time would be necessary for a fres h-water lens to produce a macroscopic ank margin cave by mixingzone dissolution. C OASTAL C ARBONATES IN N E W Z EALAND M ETHODOLOG Y e development of ank margin caves, at t h e distal margin of t h e fres h -water lens and under t h e ank of t h e enclosing landmass, places t h eir site of origin in a coastal environment. ese dissolution voids develop mere meters from t h e s h oreline and so t h ey are vulner able to breac h ing and destruction by marine erosional processes ( W alker, 2006). In t h e stable Ba h amas, sea level today is 6 m below t h at of t h e last interglacial w h en t h e ank margin caves formed. In most cases, t h ese ank margin caves are separated today by a considerable lat eral distance from modern wave action. In some cases, h owever, modern wave erosion is cling some eolianite ridges and breac h ing t h e ank margin caves contained t h erein. In rugged and steep-sloping carbonate outcrops, as exist in muc h of coastal New Zealand, past and cur rent glacioeustatic sea-level h ig h stands place wave en ergy in approximately t h e same position. Flank margin caves are t h erefore vulnerable to erosional removal, and t h e searc h for ank margin caves must take into account t h eir survivability at any given coastal location. It is also important t h at sea caves, produced by mec h anical ero sive action of waves (Fig. 2), can be dierentiated from true ank margin caves produced by dissolution and subsequently breac h ed by wave action. In t h e Ba h amas ( W aterstrat, 2007), and in Puerto Rico (Lace, 2008), cri teria for doing suc h a dierentiation h ave recently been establis h ed. To overcome survivability issues for ank margin caves, protected coastal locations, suc h as h ar bours and inlets, were given priority for examination. Suc h locations would h ave a coastal fres h -water lens, but t h e rocks would be protected from t h e h ig h -energy wave attack of t h e open coast t h at mig h t remove existing ank margin caves (Mylroie & Mylroie, 2007b). At eac h location, coastal outcrops were searc h ed for p hreatic dissolutional voids. ese were surveyed wit h a berglass tape, Suunto compass, and Suunto inclinom eter, following t h e guidelines of Das h er (1994). Eac h sur vey was linked to t h e hig h tide mark. Eac h karst feature F ig. 2: Cave at Waipu Cove, North Island, New Z ealand, in aggy Oligocene Whangarei Limestone. is small cave has been cre ated by wave energy moving a block of limestone, separated by joints, laterally along bedding planes, to the le in the gure. While minor dissolution may have enlarged the bedding planes and joints, making the block vulnerable to wave energy, produc tion of the large void was a mechanical process. J OHN E. MY LROIE J OAN R. MY LROIE & C AMPBELL S. N ELSON

PAGE 5

ACTA CARSOLOGICA 37/1 2008 19 was p h oto-documented. Wh en karst features were in close proximity, t h ey were linked by a surface survey to establis h relative position wit hin 10 cm. e survey data were processed to create a plot of t h e survey stations, around w hic h t h e cave walls were drawn using sketc h notes from t h e eld site. e caves were examined for evidence of p hreatic dissolution: bedrock pendants and columns, curvilinear bedrock surfaces, bell h oles (ceiling pockets), and wall and ceiling cusps. e caves were also examined for evi dence t h at would indicate a hig h-velocity, turbulent-ow conduit origin: turbulent ow wall markings suc h as scal lops, vadose canyons, and stream-laid clastic sediments. e continuity of t h e cave c h ambers was also examined, to see if t h ey were isolated bedrock pockets, or truncat ed portions of larger, integrated conduit caves. Caves in coastal settings wit h abundant p hreatic dissolutional fea tures, but no evidence of hig h-velocity turbulent stream ow, and w hic h are present as isolated and unconnected c h ambers, would be interpreted as ank margin caves. e presence of speleot h ems, especially in caves now open and well-ventilated, suggests t h at t h e caves were once sealed c h ambers, w hic h allowed speleot h em growt h in h umid (non-evaporative) conditions by CO 2 diusion. e caves were later breac h ed to create an entrance and subsequent ventilation. Dense, well-crystallized calcite speleot h ems are an excellent indicator of a sealed cave environment (Taborosi et al., 2006). Speleot h em criteria allow sea caves, w hic h form in an always-open condition and do not precipitate dense calcite speleot h ems, to be dierentiated from breac h ed ank margin caves. Wh en a ank margin cave is breac h ed and undergoes wave at tack, t h e cave can be stripped of its speleot h ems and its delicate p hreatic dissolutional morp h ology. In suc h cases, dierentiation of t h e two cave types is dicult. R ESULTS N ORTH I SLAND Coastal limestone outcrops were examined in t hree dif ferent settings on Nort h Island: Raglan and Kaw hia Har bours on t h e west coast, t h e Napier area on t h e sout h east coast, and W aipu Cove on t h e nort h east coast (Fig. 3). Raglan Harbour e entrance to Raglan Harbour is 40 km west of t h e city of Hamilton (Fig. 3). e h ar bour extends eastward for approximately 10 km. Raglan towns hip is on t h e sout h s h ore of t h e entrance c h annel to t h e h arbour, and directly across t h e c h annel to t h e nort h is Marotaka Point, w h ere a t hin band of Oligocene Rag lan Limestone is exposed in a coastal setting. e Raglan Limestone is a telogenetic, lig h t grey, aggy bioclastic limestone wit h about 80% CaCO 3 e local exposed t hickness is between 10 and 15 m. e coastal outcrops are a series of blocks and tow ers of limestone, t h at grade inland into solid terrain cut by deep, dissolutionally-enlarged joints, before t h e lime stone disappears under overlying units (Fig. 4). e tidal range is up to about 3 m, and t h e coastal outcrops are cyclicly inundated and drained of tidal water. e coastal outcrops s h ow a distinct intertidal notc h. e deep disso lutional ssures, w hic h are s h allow and c h oked wit h soil and debris inland, become progressively deeper and more open towards t h e coast. Some of t h e en h anced expression of t h ese features is t h e result of sapping by tidal waters, suc h t h at t h e inll material h as been removed allowing t h e full dept h of t h ese epikarst features to be observed. However, t h e ssures widen dramatically in t h e coastal zone, especially in and slig h tly above t h e intertidal zone. is widening h as been sucient enoug h to isolate indi vidual blocks of limestone as towers and pedestals. Some of t h ese towers h ave toppled (Fig. 4A). e dissolutional sculpturing of t h e limestone h as c h anged from an epi karst, vadose-ow environment inland to a coastal, cyclic marine-erosion environment at t h e s h oreline. Subsoil epikarst dissolution, w h ere soil water and descending vadose water is h eld against t h e rock like wa ter in a sponge, creates smoot h, curvilinear dissolutional surfaces t h at can h ave a similar appearance to p hreatic dissolutional surfaces. ese dissolutional surfaces ex tend from t h e top, inland portion of t h e limestone out crop to t h e coastal outcrops in t h e intertidal zone. e epikarst subsoil dissolutional surfaces rst merge wit h, and t h en are replaced by, marine intertidal erosion sur faces, w hic h are also smoot h and curvilinear (Fig. 4B). Numerous oval, tubular passages up to 1 m in diameter and larger, are present in t h e limestone wit hin t h e inter tidal zone, commonly wit h oysters attac h ed to t h e entire tube wall (Fig. 4C and 4D). Some of t h ese passages cut t hroug h a tower or outcrop, some end in sediment ll, and some end in a solid bedrock wall. All follow a joint or bedding plane or bot h, but t h e oval passage s h ape can make t h e initiating ow pat h obscure. ese caves are s h ort in lengt h, rarely exceeding 10 m in linear dimen sion. e caves are consistent wit h a ank margin origin. Kaw hia Harbour Kaw hia Harbour is also on t h e west-central coast of Nort h Island, about 25 km sout h of Raglan Harbour (Fig. 3). It extends inland 12 km, and along t h e sout h east s h ore of t h e h arbour are coastal out F LANK M ARGIN C AVE D EVELOPMENT IN T ELOGENETIC L IMESTONES OF N E W Z EALAND

PAGE 6

ACTA CARSOLOGICA 37/1 2008 20 F ig. 3: M ap of North and South Island, New Z ealand, showing limestone outcrop locations investigated for this study. e inset maps are numbered in order of the discussion of those sites in the text. J OHN E. MY LROIE J OAN R. MY LROIE & C AMPBELL S. N ELSON

PAGE 7

ACTA CARSOLOGICA 37/1 2008 21 crops of telogenetic Oligocene Ora hiri Limestone, a san dy, bioclastic limestone wit h oysters. e outcrop h ere is at least 30 m t hick. ere are two main areas of study at F ig. 4: Raglan Harbour (Late Oligocene Raglan Limestone). A Isolated towers of Raglan Lime stone, M arotaka P oint. Note that the le tower has toppled. e right tower is 7 m high. B Joints descending from the epikarst into the intertidal zone, where marine erosion has notched the lime stone. C Coastal outcrop showing tubular passages and coastal notching. D Limestone tube approximately 1 m in diameter, covered with scars of modern (not fossil) oysters. t his site: outcrops on t h e cur rent coastal area and isolated coastal hills reac hing an ele vation of up to 40 m (Fig. 5). e coastal outcrops s h ow t h e enlarged epikarst joints t h at were common at Raglan Harbour, wit h some coastal notc hing. A series of small p hreatic tubes, averag ing 10 to 30 cm in diameter, are present at a common h o rizon slig h tly above t h e hig h tide mark along t h e west ern margin of t h e outcrop (Fig. 5A). ese tubes may represent mixing dissolution prior to recent upli on t his tectonically-active coast (Pil lans, 1986). Larger tubes, in t h e 1 to 3 m diameter range, are present in several loca tions (Fig. 5B, 5C and 5D). e two hills h ave a se ries of small caves and p hre atic tubes. On t h e lower hill, t h e tubes almost entirely en circle t h e hill at 15 m eleva tion (Fig. 5B and 5C, Fig. 6A and 6B). On t h e hig h er hill, at approximately 25 m elevation, a series of small p hre atic caves are found in a continuous line on one side of F ig. 5: Kawhia Harbour (Late Olig o cene Orahiri Limestone). A Southeastern shore of Kawhia Harbour, with a line of dissolu tion tubes trending continuously along a coastal outcrop of Orahari Limestone at just above the high tide mark, on the le side of the hill shown in (5C). P erson at le for scale. B Line of dissolution tubes, marked by black arrows, at an elevation of 15 m above sea level. is face is on the far side of the hill shown in (5C). C Isolat ed hill of Orahari Limestone. All sides of this hill have dissolution features at about 15 m above sea level (arrow), as described in the text. D Large dissolution tubes in the side of a hill (the hill from which the F ig. 5C photograph was taken) at 25 m elevation, marked by black arrows. F LANK M ARGIN C AVE D EVELOPMENT IN T ELOGENETIC L IMESTONES OF N E W Z EALAND

PAGE 8

ACTA CARSOLOGICA 37/1 2008 22 t h e hill (Fig. 5D). Some of t h e caves penetrate t h e hillside for a distance of 8 to 10 m as irregular c h ambers up to 2 m hig h and 4 m wide (Fig. 7D), w hile ot h ers are s h ort tubes only a few meters long. Many of t h e smaller tubes are associated wit h dissolutionally-widened joints, and h ave an appearance similar to tubes found in typical epi karst situations. However, t h e alignment of t h ese tubes along a h orizontal datum suggests a base-level control, w hic h in t his setting could be a pre-upli sea level and concurrent mixed-water dis solution. One 15 m long tube network is associated wit h Kaw hia Oyster Tube Cave, a very large cave c h amber open to t h e seaward side (Fig 7 and Fig 8). Given t h e low wave energy conditions of t h e h ar bour interior, and t h e lack of ot h er major wave-erosion features along t his coast, t his cave may well be a breac h ed ank margin cave. Napier Nort h of Na pier (Fig. 3), at t h e sout h ern end of Hawke Bay, are coast al outcrops of coarse-grained bioclastic Pliocene limestone assigned to various forma tions (Nelson et al. 2003) stretc h ing from W aipatiki Beac h sout h to Bay View, just nort h of Napier. Coastal reconnaissance failed to nd any dissolutional caves ex cept at a small coastal out crop of W aipatiki Limestone to t h e nort h of t h e Wh irina ki community and just sout h of Te Ngaru Stream. In t h is small, low h ill, t h e seaward face h as been clied by wave action, and a series of small caves revealed (Fig. 9A). e caves are a collection of small tubes, small c h am bers, and vertical ssures (Fig. 9B and 9C). e top of t h e outcrop is capped by a siliciclastic mudstone unit, w h ic h isolates t h e carbonate unit below from any epikarst development. e caves all s h ow p h reatic morp h ology, and are somew h at regularly spaced along t h e clied h illside (Fig. 10). ey are dis tributed over a vertical range of 4 m. e Napier area is famous for tectonic activity, and in 1931, a magnitude 7.8 eart h quake resulted in coastal upli in t h e vicinity of Napier of up to 2.4 m (Hull, 1990). It is quite likely F ig. 6: Kawhia Harbour (Late Oligocene Orahiri Limestone). A V iew of tube segments visible on the face of the cli highlighted by the black arrow in F ig. 5C. B M ap of the tube segments visible in F ig. 6A. J OHN E. MY LROIE J OAN R. MY LROIE & C AMPBELL S. N ELSON

PAGE 9

ACTA CARSOLOGICA 37/1 2008 23 tion hig h on an inland cli face, along t h e sout h bank of t h e Ponui Stream t h at ows to t h e h amlet of Kairakau Beac h, just west of t h at town. Between Kairakau Beac h and Hastings, and inland 15 km, is 399 m hig h Te Mata Peak, formed on t h e Middle Pliocene Awapapa Limestone (Fig. 3). While no longer in a coastal setting, t h e dramatic east-facing scarp contains several large p hreatic pockets t h at may be relict ank mar gin caves (Fig. 11). A smaller pair of t h ese p hreatic pockets can be reac h ed at t h e base of a small cli t h at is above t h e road just nort h of t h e main Te Mata summit. ese pockets are rounded cavities in t h e limestone, wit h a mor p h ology typical of small ank margin caves (Fig. 11D). e presence of suc h relict fea tures is a measure of t h e de gree of hillside retreat. In t his case, retreat would be rela tively minor as t h e caves are still present. Rapid upli would explain t h eir survival at suc h a hig h elevation. e location of t h e caves along a single, but not level, h orizon is evidence t h at ank mar gin cave remnants can survive upli and tectonic rota tion (Fig. 11C). W aipu Cove W aipu Cove is 30 km sout h east of Wh angarei (Fig. 3). From W aipu Cove sout h to Langs Beac h, telogenetic Oligocene Wh angarei Limestone crops out in a coastal setting. At least 15 vertical meters of limestone are exposed. e outcrop is at t h e sout h ern end of Bream Bay, a relatively exposed situation as regards wave energy (Fig. 2). e limestone rests unconformably on Permo-Triassic basement greywackes and argillites. is contact is above sea level on t h e W aipu Cove end of t h e outcrop. As previously described at Raglan Harbour, large vertical ssures cut t hroug h t h e sequence, initiat ing inland and upwards as epikarst features, t h at grade downward and seaward in t h e coastal notc hing zone to create isolated stacks and towers (Fig. 12A). e walls of t h ese enlarged ssures contain numerous p hreatic tubes and small caves (Fig. 12B t hroug h Fig. 12E). In ssures slig h tly inland from t h e active wave base, delicate disso lutional morp h ology is visible, indicating t h at wave ener t h at t h e ank margin caves seen at Wh irinaki are Ho locene in age. In Napier itself, coastal cli outcrops up to 30 m h ig h of t h e Late Pliocene Scinde Island Limestone front t h e h ar bour, but contain no signicant ank margin caves. Nu merous small subaerial weat h ering pockets, w h ic h appear like tafoni, mottle t h e outcrop surface. Dierentiation of tafoni from ank margin caves on carbonate clis h as been establis h ed from work in t h e Ba h amas (Owen, 2007). Sout h of Napier, and sout h of Hastings and Cape Kidnappers, t h e Early Pliocene Kairakau limestone crops out nort h along t h e coast for several kilometers from Kai rakau Beac h (Fig. 3). No dissolution caves were found on t h e seaward facing clis of t h e unit. A stream valley 1.5 km nort h of Kairakau Beac h, owing west to east, bisects t h e coastal cli, creating protected limestone clis on its nort h and sout h banks. Despite a detailed searc h, no dis solution caves were found in t his locality. e site could be considered ideal, as at past hig h er sea levels it would h ave contained marine water, but its protected setting would h ave prevented signicant cli retreat and ank margin cave removal. e absence of ank margin caves h ere is unexplained. A small pit cave was noted in sec F ig. 7: Kawhia Harbour (Late Oligocene Orahiri Limestone). A Large cave chamber with its outer wall breached. At the back end of this chamber, passages shown in F ig. 7 B and 7C lead through the outcrop and back to the coast. P erson at le for scale. B Tubular passage leading from the coast into the cli shown at the far end of F ig. 7A. P erson for scale; note the old owstone on ceiling and oor. C P assage into the large breached chamber. P erson in distance for scale; note modern oyster scars on cave walls and old owstone. D Abundant speleothems in a small cave at the 15 m level of the hill shown in F ig. 5C. P erson seated against right wall for scale, white arrow points to persons head. F LANK M ARGIN C AVE D EVELOPMENT IN T ELOGENETIC L IMESTONES OF N E W Z EALAND

PAGE 10

ACTA CARSOLOGICA 37/1 2008 24 gy alone is not responsible for muc h of t h e observed cave morp h ology. P hreatic tubes and enlarged ssures pen etrate for more t h an 10 m in numerous locations, wit h lateral enlargements t h at h old a h orizontal trend across t h e dipping beds (Fig. 12D). Followed into t h e outcrop, t h ese dissolutional passages undulate in a classic p hreatic pattern (Fig. 12C), but contain no indicators of hig h-ve locity turbulent ow. S OUTH I SLAND Coastal limestone outcrops were examined at ve dier ent locations in Sout h Island: Po h ara on t h e nort h coast, Paturau River on t h e nort h west coast, Punakaiki on t h e west-central coast, Kakanui on t h e east-central coast, and Kaikoura on t h e nort h east coast (Fig. 3). Po h ara Telogenetic Oligocene Takaka Limestone crops out in t h e vicinity of Po h ara and Tarako h e on t h e sout h east s h ore of Golden Bay (Fig. 3). e coastline in t h is area is fairly rugged, and wave energies are h ig h At t h is locality t h e Takaka Limestone overlies t h e mec h ni cally-weak Eocene Brunner Coal Measures and as a result is quickly undermined by wave action in coastal areas to form a limestone mega-rubble embankment at sea level (Fig. 13A). Muc h of t h e coastal outcrop is t h us obscured, but at t h e Abel Tasman monument just nort h of Tarako h e, a large, clied blu of Takaka Limestone is preserved west and above t h e road. is outcrop h as a series of dissolu tional voids t h at cut across t h e existing rock structure (Fig. 13B). ese voids are up to 2 to 3 m h ig h and wide, and extend inward for 3 to 4 m (Fig. 13C). ey are regularly spaced along t h e outcrop and t h eir pattern and morp h olo gy are consistent wit h mixedwater dissolution. Nort h west 80 m along t h e road and t h e exposed cli, in a road cut on t h e west side of t h e road, is a single c h amber cave at t h e same elevation as t h ose to t h e sout h e cave is 7.5 m long and 4 m wide, wit h a vertical range of 3 m, trending sout h west (Fig. 13D). It contains muc h collapse material, and secondary calcite precipitates (speleot h ems), so t h e origi nal dissolution surfaces of t h e cave are obscured, and its mode of speleogenesis is not obvious. F ig. 8: M ap of cave shown in F ig. 7A, 7B and 7C. e cave is an isolated chamber. F ig. 9: Napier district (Late P liocene Waipatiki Limestone). A e small beachside hill north of Whirinaki contains a series of short ank margin caves, which are exposed on the seaward side of the hill, where wave action has breached into them. P erson standing at cli base in right center, circled, for scale. B Some of the caves at Whirinaki. Note that the limestone is covered by silici clastic mudstone, so no epikarst development has occurred. C Isolated tubes and chambers at Whirinaki. e vertical extension of these caves, as seen in F ig. 9B and the separation into stacked tubes, as seen in F ig. 9C, are evidence of rapid upli. J OHN E. MY LROIE J OAN R. MY LROIE & C AMPBELL S. N ELSON

PAGE 11

ACTA CARSOLOGICA 37/1 2008 25 e clis west of t h e Abel Tasman monument con tain a number of isolated p hreatic c h ambers t h at h ave been breac h ed by cli retreat, to reveal t h e calcite spe F LANK M ARGIN C AVE D EVELOPMENT IN T ELOGENETIC L IMESTONES OF N E W Z EALAND leot h ems t h at formed inside t h e c h ambers w h en t h ey were still sealed (Fig. 14A and 14B). eir position and s h ape is consistent wit h a ank margin origin. Just east of t h e natural bridge at t h e small h arbour east of Po h ara, a large talus block rests on t h e coastal side of t h e road. Two readily acces sible cave entrances are in t his block, just centimetres from t h e roadbed (Fig. 14C). e block contains wit hin it 30 m of cave passage formed along joints and bedding planes, consistent wit h t h e cave passages forming prior to t h e block tumbling fro t h e cli face across t h e road (Fig. 14D). P hreatic dissolutional surfaces are abundant in t h e cave, consistent wit h mixing-zone dissolution and speleogenesis; h owever, its relations hip wit h t h e origi nal h ost environment of t h e talus block is unknown, and a ank margin designation cannot not made. Paturau River Paturau River is located on t h e nort h east coast of Sout h Island, 30 km sout h west of t h e Golden Bay community of Pakawau (Fig. 3). Telogenetic Takaka Limestone is extensive in t his area, and forms a coast al outcrop trending sout h west from t h e community of Paturau River (Fig. 15). e outcrop consists of a low benc h 2 to 3 m hig h t h at be comes up to 8 m hig h to t h e sout h west (Fig. 15B). A num ber of small streams and riv ulets drain to t h e coast in t h e low benc h, cutting canyons t hroug h t h e outcrop t h at are enlarged into oval tubular cross sections in t h e inter tidal zone (Fig. 15A). Far t h er sout h west, t h e limestone benc h broadens, and a small ank margin cave, modied F ig. 10: M ap of the Whirinaki caves. e caves open southward on to beach seen in the foreground of F ig. 9. F ig. 11: Napier district (M iddle P liocene Awapapa Limestone). A Te M ata P eak formed in wellbedded Awapapa Limestone, here about 120 m thick. Arrows marked B and D identify the loca tions of the features shown in F ig. 11B and 11D, respectively. B Close-up of F ig. 11A, showing a large phreatic pocket on right, and smaller ones to the le. C Te M ata P eak, with dashed line to show the majority of the observed phreatic pockets follow a line that would have been at sea level prior to upli and tilting. Letters and arrows as in F ig. 11A. D Small ank margin cave next to person on le, breached phreatic pocket to the right.

PAGE 12

ACTA CARSOLOGICA 37/1 2008 26 by wave erosion, is present (Fig. 16). e cave consists of a c h amber 9 m long by 5 m wide, breac h ed on two sides, wit h a large ceiling collapse (Fig. 16A). e presence of bell h oles in t h e ceiling, and ot h er p hreatic dissolution features in t h e cave walls, in dicate a dissolutional origin for t h e void (Fig. 16B). Given t h e setting and a lack of a conduit continuing inland, t h e cave is most likely a ank margin cave. Inland on t h e benc h a few meters, are sev eral ot h er smaller p hreatic voids t h at contain delicate bedrock pillars and spans, in dicating dissolutional origin (Fig. 16C and 16D). ese voids are isolated and not in tegrated, w hic h agrees wit h a ank margin mode of origin. Fart h er to t h e sout h west, t h e limestone exposure t hickens into a sea cli at t h e back of t h e beac h. e cli h as a step in it, and at t h e back of t his step are numerous small notc h es along a single h orizontal datum (Fig. 15B, 15C, 15D). ese notc h es could be intertidal bioero F ig. 12: Waipu Cove district (Oli gocene Whangarei Limestone). A 10 m-tall towers of Whan gaeri Limestone at Waipu Cove, isolated by epikarst and marine processes. B Dissolutional en largement along a ssure passage; ashlight 15 cm long. C Elon gated and undulating tube, which dead ends within the limestone outcrop. D Joints and passages enlarged by dissolution cut across tilted bedding. E P hreatic pas sages within the limestone mass are being destroyed as the out crop retreats in response to wave attack; ashlight 15 cm long. J OHN E. MY LROIE J OAN R. MY LROIE & C AMPBELL S. N ELSON F ig. 13: P ohara-Tarakohe (Oligocene Takaka Limestone). A Limestone clis at P ohara, showing large mega-talus obscuring the actual face of the coastal outcrop. e talus block containing the cave shown in F ig. 14C and F ig. 14D is in the large tree-covered block on right below the arrow seen in F ig. 13A. P erson (circled) on le for scale. B Series of phreatic pockets and ank margin caves following a level horizon that cuts across structure just south of the Abel Tasman M onument. P erson (circled) in center of photograph for scale. C One of the small ank margin caves seen in F ig. 13B Note the wall pockets and eroded bedrock column (le side of photo), typical of phreatic dissolution. D B reakdown and owstone-choked cave just north of F ig. 13B Soda straws in upper right are 5 cm long for scale. Note that some of the stalagmites on the block on the right background of the image are not vertical, indicating slumping of the block aer stalagmite deposition.

PAGE 13

ACTA CARSOLOGICA 37/1 2008 27 sion notc h es from an earlier sea level, except t h at some contain delicate p hreatic dis solutional features and small p hreatic tubes (Fig. 15E), suggesting an origin as ank margin caves wit hin t h e bed rock mass, and subsequently breac h ed by cli retreat. Punakaiki Teloge netic Oligocene W aitakere Limestone trends in a band parallel to t h e west coast of Sout h Island from C h arles ton, 20 km sout h of Cape Foulwind, sout h to Punak aiki (Fig. 3). e outcrop is mostly inland, but reac h es t h e coast sout h of Fox Creek at an area called Limestone Creek, and again at t h e clas sic Pancake Rocks outcrop at Punakaiki. e Pancake Rocks outcrops are in a na tional park and t h e sea clis are not routinely accessible. Observation from t h e tourist trails indicates t h at t h e clis h ave numerous openings into s h allow caves (Fig. 17), but t h e origin of t h ese caves could not be determined. At t h e nort h end of t h e park, on t h e east side of t h e road, is t h e entrance to Punakaiki Cav ern, a self-guided tour. e F LANK M ARGIN C AVE D EVELOPMENT IN T ELOGENETIC L IMESTONES OF N E W Z EALAND F ig. 14: P ohara-Tarakohe (Oligo cene Takaka Limestone). A Cli with a breached phreatic pocket on the right (approximately 5 m in vertical height), and exposed cave owstone on the le. B B reached phreatic pocket with cave stalactites hanging from the remnant ceiling; the large ow stone column to the far right is approximately 3 m long. C Two entrances to the dissolution cave within the large talus block noted in F ig. 13A. D P assage inside the talus block that connects the two entrances, showing dissolu tional morphology. F ig. 15: P aturau River (Oligocene Takaka Limestone). A Oval passage in the intertidal zone, which carries turbulent stream water from the overlying clastic units through the limestone to the sea. B Coastal Takaka Limestone outcrop at P aturau River. P erson is circled on le for scale. Regular pockets and notches look like a degraded marine-erosion notch, but the presence of dis solutional features indicates a ank margin cave origin for many of these pockets. C Cli, 10 m high, showing large pockets. B oxed section labeled E shown expanded as F ig. 15E. D Large, un dulating notch in cli shown in F ig. 15B P erson circled for scale. E Close up of the pocket shown in F ig. 15C. e delicate rock fretwork shown here is a product of phreatic dissolution in a cave chamber, and not from wave erosion in an open environment.

PAGE 14

ACTA CARSOLOGICA 37/1 2008 28 F ig. 16: P aturau River (Oligocene Takaka Limestone). A A breached ank margin cave at P aturau River. B B ell holes in the ceiling of the cave chamber shown in F ig. 16A. Note pencil 15 cm long for scale stuck in the le bell hole. C Small phreatic cave chamber breached by wave erosion, 10 m inland of the cave shown in F ig. 16A. e cave is in the early stages of destruction, as the thin bedrock pillar is still intact. F lashlight 15 cm long for scale. D Another small phreatic cave, inland of the cave shown in F ig. 16A. B edrock pillar on the le was partially dissolved from its ceiling con nection prior to the cave being breached. e angled bedrock pillar to the right is fragile but still intact. F lashlight 15 cm long for scale. cave is a large c h amber tens of meters across and deep, muc h modied by collapse, wit h ssure passages leading inland and terminating. It is dicult to determine if t h e cave is a fragment of a conduit system, or t h e result of mixing dissolution. At Limestone Creek, a few kilometres nort h of Pan cake Rocks (Fig. 3), coastal outcrops are accessible from t h e beac h. A number of small caves exist at t h e sout h end of t h e beac h next to Limestone Creek. Proceeding nort h, two very important outcrops are reac h ed. Lime stone Creek roug h Cave is a p hreatic passage devel oped along joints t h at passes for 30 m t hroug h a small limestone h eadland (Fig. 18). e passage averages sev eral metres hig h and wide, but becomes smaller at eac h end. e cave h as no indications of hig h-velocity, turbu lent conduit ow, suc h as scalloping of passage walls. e main cave passage splits, and a large side passage trends into t h e seaward side of t h e h eadland and ends in a blank bedrock wall, a h allmark of ank margin cave dissolution (Fig. 19). Fart h er nort h along t h e beac h, is an area named Tube City. In t his outcrop, numerous p hreatic tubes exist on a h orizontal datum t h at cuts across t h e bedding and jointing (Fig. 20). e dip of t h e bedding in t his outcrop is 20 o to t h e SSE (Fig. 21). e jointing is closely spaced, dipping at 65o to t h e NNE. e caves in t h e outcrop face, as s h own by t h e wave-cut limestone benc h in front of t h em (Fig. 20B), were once enclosed. e p hreatic tubes trend inland up to 5 m, before closing down to a single joint (Fig. 20C). e tubes contain layers of owstone on t h eir walls, indicating t h ey were once closed c h ambers and h ave been breac h ed by mod ern wave activity (Fig. 20C). A total of 7 p hreatic tubes are found in an outcrop span of 44 m, all tube oors being between 3 and 5 m above sea level. Dead-end p hreatic tubes wit h out turbulent ow markings, suc h as scallops, at a constant elevation cutting across structure, are some of t h e indicators of a ank mar gin origin (Fig. 20D). Kakanui On t h e east coast of Sout h Island, approximately 12 km sout h of Oamaru, at Kakanui Point at t h e nort h end of All Day Bay, t h e Oligocene Ototara Limestone, and t h e overly ing Otekaike Limestone, crop out in a coastal setting (Fig. 3). Bot h limestone units are topped by an uncon formity displayed as a well-developed paleokarst surface. e paleokarst surface of t h e Ototara extends seaward and eastward as a low limestone benc h from t h e sout h side of Kakanui Point (Fig. 22A and 22B). In t h e blu be hind t h at beac h are a series of small, lenticular voids up to 2 m wide and 0.5 m hig h, t h at cut across t h e primary limestone structure in t h e outcrop (Fig. 22C and 22D). ese voids h ave bedrock pendants and residual columns typical of p hreatic dissolution, but lack any evidence of conduit ow, and exist as isolated c h ambers. ey meet t h e eld criteria for ank margin caves. Kaikoura Telogenetic Paleocene Amuri Lime stone and overlying Oligocene Grey Marl crop out in a coastal setting on t h e Kaikoura Peninsula in nort h eastern Sout h Island (Fig. 3). Compared to t h e medium to very coarse-grained bioclastic carbonates c h aracter izing all t h e ot h er New Zealand limestones studied, t h e Amuri Limestone is an indurated very ne-grained J OHN E. MY LROIE J OAN R. MY LROIE & C AMPBELL S. N ELSON

PAGE 15

ACTA CARSOLOGICA 37/1 2008 29 limestone, or micrite. e Kaikoura area is geologically complex, wit h folding and faulting leading to a com plex outcrop pattern. e tectonics h as created a very fractured and jointed lime stone (Fig. 23A and 23B). e initial feature is a small cave on t h e nort h side of t h e peninsula, on t h e outskirts of t h e main town, just west of t h e road leading to t h e w h arf. In a low vertical face trending N-S is an entrance to a small cave 6 m deep and 4 m wide, 1 to 1.5 m h ig h (Fig. 23C and 23D). Several small dissolution pockets are nearby on t h e cli face. e interior of t h e cave is somew h at s h attered, but p h reatic dissolution pockets are still discernible in t h e roof and walls of t h e cave (Fig. 23D). It appears to be an isolated c h amber and, given its setting, t h erefore a small ank margin cave. Moving east along t h e nort h coast of t h e peninsula, t h e Amuri Limestone outcrop disappears at Avoca Point, to be replaced by t h e Grey Marl. Going around Point Kean, and continuing sout h west, t h e limestone outcrop again appears, and a small cave, 2 m hig h, 0.5 m wide, and 3 m deep opens at t h e base of a steep hill. e cave walls are s h attered and irregular, and t h e origin of t h e cave is dicult to determine. It could be a sea cave or a dissolution cave. No caves were located for about t h e next kilometre going sout h west, but around East Head, on its sout h west side, be hind trees at t h e base of t h e cli, is Kaikoura Point Sea Cave (it is marked as a sea cave on a tourist map), a large dissolution cave (Fig. 24). e entrance is 5 m wide and 5 m hig h, wit h a triangular cross section (Fig. 25A and 25B). e passage leads nort h east, as an oval tube 3 to 5 m wide, and 3 m hig h, wit h an additional disso lutional slot in t h e ceiling a metre wide (Fig. 25B). e passage continues for 30 m to a terminal c h amber 14 m F LANK M ARGIN C AVE D EVELOPMENT IN T ELOGENETIC L IMESTONES OF N E W Z EALAND F ig. 17: P unakaiki (Oligocene Waitakere Limestone). A Coastal outcrop at the north end of the self-guided tour at P unak aiki, showing cave entrances. B Close-up of the middle entrance shown in F ig. 17A. C Natural bridge at P unakaiki. A horizon of dissolution pockets and voids is visible halfway up the bridge, in box. P eople on horizon (circled) for scale. D Close-up of the pockets and voids noted in the box of F ig. 17C. F ig. 18: M ap of Limestone Creek rough Cave.

PAGE 16

ACTA CARSOLOGICA 37/1 2008 30 F ig. 19: P unakaiki (Oligocene Waitakere Limestone). A P oint containing Limestone Creek rough Cave. Note that the bedding dips from le (north) to right (south). P erson on le for scale. B e main passage in Limestone Creek rough Cave, formed by dissolution at the intersection of a bed ding plane and joint. P erson sitting at end of the passage for scale. C P assage developed along a joint, with a dissolutional widening that is horizontal despite steep dip of the bedding. is passage leads to the chamber seen in F ig. 19D. P erson sitting low in the distance for scale. D Terminal chamber, with passage pinching down along the guiding joint seen in F ig. 19C. F ig. 20: P unakaiki (Oligocene Waitakere Limestone). A Tube City outcrop on the Limestone Creek coast. Seven individual tubes, all within 1 or 2 m of the same elevation, are in the outcrop. e bedding dips from le (north) to right (south), and the jointing is very closely spaced. P erson at le for scale. B Tube City outcrop from a distance, showing the wavecut platform in front of the caves, indicating signicant shoreline re treat in this location. C V iew in ward in the large phreatic tube seen on the right in F ig. 20A. Note how the bedrock walls pinch down to the guiding joint, and the horizon tal owstone horizons on the right wall, and in the back. F lashlight is 15 cm long for scale. D Looking out the same tube as in F ig. 19C, across the wave-cut platform. ets, evident despite t h e s h at tered and irregular nature of t h e cave walls (Fig. 25D). A smaller cave exists 5 m to t h e rig h t (east) of t h e main cave, and is 4 m long (Fig. 24 and Fig. 25A). ese caves appear to matc h t h e conditions for ank margin cave develop ment. W est and sout h west from East Head, and ap proac hing Atia Point, t h e limestone outcrop again ends, to be replaced by t h e Gray Marl. A few tens of meters east of t his contact, a small limestone point contains two relatively large caves, col lectively named Kaikoura Penguin Cave (Fig. 26). e caves run nort h east to sout h west, parallel to one anot h er, one inland of t h e ot h er, bot h wit h entrances on eac h side of t h e point. e more inland cave is entered by ascend ing and descending a large collapse pile, w hic h leads to a voluminous entrance c h amber (Fig. 27A). Trending nort h west from t h e entrance c h amber is a passage 16 m wide and 4 m hig h (Fig. 25C). e cave ends in bedrock walls. e cave walls contain numerous dissolution pock J OHN E. MY LROIE J OAN R. MY LROIE & C AMPBELL S. N ELSON

PAGE 17

ACTA CARSOLOGICA 37/1 2008 31 long, 5 m wide and 3 m hig h. is passage h as well devel oped dissolution pockets and small c h ambers developed along its lengt h, and it terminates in bedrock walls (Fig. 27B). Sout h west from t h e en trance c h amber, a hig h, nar row passage h eads 28 m di rectly t hroug h t h e limestone outcrop to a small entrance on t h e far side (Fig. 27C). e passage starts out 10 m hig h and 3 m wide, sloping upward to t h e sout h west so t h at at t h e far entrance, t h e cave is only 1 m wide and 0.5 m hig h. e cave oor is mud and sediment, and t h e nature and position of t h e bedrock oor cannot be determined. is passage h as a number of dissolution pockets in t h e side walls and ceiling, consis tent wit h a p hreatic origin. Seaward of t h e rst cave is a second cave t h at also runs from t h e nort h east to t h e sout h west, t hroug h t h e limestone point (Fig. 28A). ere are two additional entrances t h at open to t h e sout h on t h e face of t h e limestone point (Fig. 28B). e west ern of t h ese two entrances is at an elevation low enoug h to h ave back-beac h cobble rubble on t h e oor (Fig 28B). e sout h western entrance on t h e far side of t h e lime stone point h as some collapse debris blockage (Fig. 28C). As wit h t h e inland cave, t h e nort h eastern entrance is dominated by a large pile of cli collapse (Fig. 28D). e two entrances to t h e sout h are unobstructed. e main pas sage t hroug h t h e limestone point is 40 m long, and up to 5 m hig h and wide. S h ort side passages trend nort h west to wards t h e rst cave but end in bedrock walls. e passages leading sout h to t h e sout h ern entrances are 10 m long. e passage to t h e nort h east is 5 m hig h and 3 m wide, w hile t h e passage to t h e sout h west is 2 m hig h and wide. e two caves bot h s h ow p hreatic dissolutional F LANK M ARGIN C AVE D EVELOPMENT IN T ELOGENETIC L IMESTONES OF N E W Z EALAND F ig. 21: M ap of the Tube City outcrop, plan view and composite section. Compare with F ig. 6B F ig. 22: Kakanui (Oligocene Ototara Limestone). A Outcrop of Ototara Limestone with paleo karst surface exposed. B Close-up of the outcrop in F ig. 22A, showing paleokarst inllings. P encil 15 cm long for scale. C F lank margin cave in the cli face. D Close-up of the cave entrance seen in F ig. 22C.

PAGE 18

ACTA CARSOLOGICA 37/1 2008 32 F ig. 23: Kaikoura (P aleocene Amuri Limestone). A F olded beds of the Amuri Limestone distorted to a vertical position, north side of Kaikoura P oint. B Contact of the Amuri Limestone, le, with the Grey M arl, right. Note in foreground that faulting has displaced the contact. C Entrance to a small ank margin cave, north side of Kaikoura P oint. e small phreatic pocket above and to the le does not connect with the cave. D Single chamber that forms the cave. Note the shattered nature of the limestone walls and ceiling, and the dissolution pocket at the back of the chamber, to the right of the person. F ig. 24: M ap of Kaikoura P oint Sea Cave. surfaces and passage com plexity, but are not cur rently connected. ey pass t hroug h t h e limestone point containing t h em, and ap pear to h ave been opened by cli retreat, demonstrated by t h e broad wave-cut benc h in front of t h e point (Fig. 28A). e walls contain no turbu lent ow markings. e caves seem to t t h e criteria for ank margin development. A few kilometers sout h of Kaikoura Peninsula, just west of Hig h way 1, is Maori Leap Cave, a s h ow or com mercial cave. e cave was opened from above by lime stone quarrying, but rubble inll was removed to open t h e cave h orizontally on t h e Amuri Limestone cli face. e cave is a single passage, about 100 m long, up to 8 m hig h and 6 m wide, perpen dicular to t h e cli face. ere is muc h collapse material in t h e cave, and all original dis J OHN E. MY LROIE J OAN R. MY LROIE & C AMPBELL S. N ELSON F ig. 26: M ap of Kaikoura P enguin Cave, which is actually two separate caves.

PAGE 19

ACTA CARSOLOGICA 37/1 2008 33 solutional surfaces (if origi nally present) are gone. e cave is presented on t h e tour as a fossil sea cave, but t h ere is no evidence for eit h er t h at origin or a dissolutional spe leogenesis. e cave is very long relative to its widt h, and propagation of wave energy to suc h a penetration seems unlikely. e walls of t h e cave display many folds and small faults, as well as c h ert, and t h e limestone is hig h ly frac tured (Fig. 27D). e collapse material is made up mostly of blocks less t h an 1 meter in maximum dimension, and commonly only 20 cm or so in maximum dimension. F LANK M ARGIN C AVE D EVELOPMENT IN T ELOGENETIC L IMESTONES OF N E W Z EALAND F ig. 25: Kaikoura (P aleocene Amuri Limestone) Kaikoura P oint Sea Cave. A Entrance area of Sea Cave; main entrance is in the background, secondary cave entrance is in the foreground. B M ain passage in Sea Cave. Note ceiling slot. B eds dip towards the camera. C Terminal room in Sea Cave, looking west. Note steeply-dipping (up to 55 o ) beds. D P hreatic pocket in wall of the main passage of Sea Cave. e pocket is directly behind the person in F ig. 25B F ig. 27: Kaikoura (P aleocene Amuri Limestone) Kaikoura P enguin and M aori Leap Caves. A Looking south into the en trance chamber of the inland of the two Kaikoura P enguin Caves. B Side pocket in le wall of the passage shown in F ig. 27A, dis playing smooth, phreatic origin. C V iew northeast in the passage that cuts through the point in the inland cave. Note the smooth, curved ceiling of this passage, indicating a phreatic origin. D F olded and distorted wall rock in M aori Leap Cave.

PAGE 20

ACTA CARSOLOGICA 37/1 2008 34 F ig. 28: Kaikoura (P aleocene Amuri Limestone) Kaikoura P enguin Caves. A Limestone point containing the two Kaikoura P enguin Caves. Note that the caves have been breached by erosion and planation of the bedrock outcrop in front of the point. B e two southern entrances into the seaward of the two Kaikoura P enguin Caves. Note the cobbles in the oor area of the western (le) entrance. C M ain passage of the seaward cave. P enguin for scale is 1 m high. To the le is the southwestern entrance, passages ahead of the penguin end in bedrock walls before reaching the inland cave. D Looking northeast to the northeastern entrance of the seaward cave. e person is illuminated by light from the easterly (right) of the two entrances seen in F ig. 28B e wall behind the person shows dissolutional smooth, curvilinear morphology. J OHN E. MY LROIE J OAN R. MY LROIE & C AMPBELL S. N ELSON Distinguis hing cave origins Caves located in limestones in coastal settings can be of a variety of origins. ey can be pseudokarst caves, formed by processes ot h er t h an dissolution. Examples would be talus caves formed by cli collapse, tafoni by surcial weat h ering, or sea caves formed by wave erosion. e caves can also be true karst caves of dissolutional origin. Talus caves are self evident, alt h oug h as s h own by t h e example at Po h ara, talus blocks may contain dissolutional caves present in t h e rock prior to cli collapse (Fig. 14C and 14D). Tafoni caves can be dierentiated from dissolutional caves by morp h omet rics (Owen, 2007). If t h e original bedrock surfaces from t h e time of cave genesis are preserved, dierentiating pseudokarst sea caves from dissolutional karst caves is relatively straig h t forward. Dissolution of limestones to create a cave leaves be hind a variety of bedrock forms and s h apes t h at are dierent from t h ose produced by me c h anical erosion by wind, or by waves and wave-entrained debris. Wh en a dissolution cave opens onto a coastal en vironment, wave action may enter t h e cave and obscure t h e original dissolutional surfaces, creating confusion as to origin. is alteration is of particular importance for ank margin caves, as small ones, once breac h ed, com monly may not extend beyond t h e inuence of wave ac tion, and t h eir entire interior may be modied. Even in t his case, use of morp h ometrics h as allowed sea caves to be dierentiated from ank margin caves in t h e Ba h amas (W aterstrat, 2007) and in Puerto Rico (Lace, 2008) based on s h ape parameters suc h as area to perimeter ratio, and entrance widt h to maximum interior widt h ratio. Dense, well crystallized secondary calcite speleot h ems stalac tites, stalagmites, owstone, etc., s h ould not form in sea caves, even w h en subsequently abandoned by sea level, as suc h speleot h ems do not develop in an open atmosp h ere environment. Given t h at calcareous tufa, w hic h can crudely mimic cave speleot h ems, can form in abandoned D ISCUSSION

PAGE 21

ACTA CARSOLOGICA 37/1 2008 35 sea caves, criteria for dierentiating t h e two s h ould be utilized, as presented in Taboro i et al. (2006). e situation becomes more complex w h en attempt ing to discriminate between mixing zone dissolution and ank margin cave formation wit h in a fres h -water lens, and h ig h -velocity, turbulent conduit ow in a cave stream passage. In a setting suc h as t h e Bah amas, w h ere conduitow stream caves do not exist today, suc h discrimination is not necessary (Mylroie et al. 1995). In more complex carbonate island settings, suc h as t h e Mariana Islands, stream conduit caves form at t h e contact between carbon ates and underlying volcanics, w h ile in adjacent solely carbonate coastal areas, ank margin caves h ave formed (Jenson et al. 2006). If a conduit cave is still active, it may be possible to enter t h e cave and see t h e turbulent ow in situ. Flank margin caves form by dissolution wit h in t h e bedrock mass; reactants enter, and products leave, by diuse ow. e caves develop wit h out entrances and t h erefore are dicult to view in situ. Flank margin caves are observed aer sea level c h ange (eit h er eustatic or tec tonic) h as placed t h e cave above sea level, and erosion h as breac h ed into t h e cave, making it accessible. Conduit-ow stream caves, w h en t h ey disc h arge t h eir water back to t h e surface environment, commonly form a distributary pattern analogous to t h e distributary pattern of deltas. A drop in base level as a result of sealevel c h ange can lead to abandonment of a set of distribu tary passages. Furt h er erosion may incise and segment suc h a set of cave passages, to produce a local abundance of small caves. Additional erosion may obscure t h e ge netic relations hip of t h ese cave fragments to eac h ot h er. e observer coming to a coastal environment may see a series of small caves t h at contain evidence of dissolution al development. e consistent elevation of t h ese caves, and t h eir location at a sea coast may lead to a conclusion of cave development in t h e fres h-water lens during a past, hig h er sea level. e caves may be incorrectly interpreted as ank margin caves. Caves in telogenetic carbonates In eogenetic car bonate rocks, wit h hig h primary porosity, fres h water and seawater mix over a broad front, and diuse ow of reac tants and products occurs over a large area. e end re sults are broad, globular c h ambers and mazes t h at inter sect to form caves of some complexity, but are separated from nearby caves as t h ere is no integration or conduit ow in suc h an environment (Fig. 1). ese caves are eas ily dierentiated from conduit caves based on cave mor p h ology, and t h e absence of hig h-velocity turbulent ow indicators. In dense, telogenetic carbonate rocks present in New Zealand, w h ere ow permeability is along bed ding planes, joints, and fractures, and matrix porosity is minimal, mixing dissolution is concentrated along t h ose ow pat h s, and not in t h e matrix. As a result, mixing dis solution caves suc h as ank margin caves would be ex pected to h ave a passage pattern t h at reected bedding, joint, and fault orientations and trends. To dierentiate ank margin caves formed in telo genetic rocks from stream conduit caves formed in telo genetic rocks t h erefore requires close observation. Two main observational dierences s h ould remain true: 1) telogenetic ank margin caves s h ould s h ow no evidence of hig h-velocity turbulent or conduit ow, in bot h wall rock dissolutional morp h ology, and in interior sediments; 2) passage relations hips in telogenetic ank margin caves s h ould s h ow no dendritic or organized links wit h adjacent caves, and s h ould contain passages and c h am bers t h at end in blank bedrock walls. Again, morp h omet rics h ave successfully been used to distinguis h ank mar gin caves from conduit-ow stream caves (Rot h, 2004; Rot h et al. 2006). Flank margin cave size is a result of t h e geoc h em istry at t h e distal margin of t h e fres h-water lens, t h e na ture of t h e h ost carbonate rock, and t h e duration of time t h at t h e fres h-water lens is at a single h orizon. Given t h at sea water and fres h water c h emistries tend to fall wit hin standard limits, t h e main geoc h emical dierence is most likely related to t h e degree of organic loading of t h e den sity boundaries at t h e top and bottom of t h e lens. ere is no eective way to judge past organic loading, except t h at w h en it is excessive and leads to anoxic conditions, it can leave a signature in t h e cave wall rock (Bottrell et al. 1993). e h ost carbonate rock inuences ank margin cave size based on two parameters: t h e amount and distribution of rock porosity and permeability, and t h e amount, c h emistry, and purity of t h e carbonate ma terial. ese properties can be determined from outcrop sampling. e duration of time t h at t h e geoc h emical en vironment of t h e fres h-water lens remains in t h e same section of carbonate rock is tied to sea-level stability. e longer t h e lens is in a stable position, t h e more geoc h emi cal dissolutional work t h at can be done and t h e bigger t h e ank margin cave produced (Mylroie & Mylroie, 2007). Sea-level histories for coastal areas in tectonically-active environments suc h as coastal New Zealand can be dif cult to determine (Pillans 1986). e New Zealand situation Despite t h e tectoni cally-active environment of coastal New Zealand, it can be reasonably assumed t h at during t h e late Pleistocene, many of t h e carbonate outcrops observed for t his study were well above sea level as a result of glacioueustasy and in a w h olly-vadose environment, free of direct marine inuence. In suc h a setting, exposed limestone outcrops can develop an extensive and mature epikarst. In dense, telogenetic limestones, suc h an epikarst consists of dis solutionally-enlarged joints, bare bedrock surfaces t h at F LANK M ARGIN C AVE D EVELOPMENT IN T ELOGENETIC L IMESTONES OF N E W Z EALAND

PAGE 22

ACTA CARSOLOGICA 37/1 2008 36 develop rillenkarren and related dissolutional fretwork, and soil-covered bedrock w h ere water is h eld against t h e rock as if in a sponge and smoot h, curvilinear dis solutional forms are produced. Wh en sea level rises, and t h ese outcrops are partially inundated by marine waters, t h ese epikarst features are overprinted by marine ero sion, especially bioerosion in t h e intertidal zone. Soil and ot h er inll material in t h e dissolutional ssures and on t h e bedrock surface can be stripped away by wave action, exposing previously buried dissolutional forms. Any mixing dissolution t h at occurs in t his environment may be dicult to separate from earlier vadose dissolutional processes in t h e former subsoil epikarst, and modern marine erosion processes. e outcrop at Raglan Har bour s h ows t his superposition of forms extremely well (Fig 4). Despite all t h e overprinting at Raglan, p hreatic tubes and tube segments were abundant and well placed to h ave formed by mixing dissolution in a coastal setting. But ank margin caves larger t h an tubes a few meters in lengt h were not observed. Outcrops suc h as at Whirinaki (Fig 9 and 10), w h ere t h e carbonate unit is overlain by a ne-grained clastic unit, and no epikarst h as h ad a c h ance to develop, t h e results of ank margin cave development are more easily observed. At W aipu Cove (Fig. 12) t h e epikarst h as been overprinted by coastal processes, but tube development and widening of joints at a specic h orizon on some outcrops presents t h e opportunity t h at t h ese voids are small ank margin caves. e ank margin caves observed in t h e various Nort h and Sout h Island outcrops came in a variety of forms and sizes. e largest ank margin caves were found at Kaikoura (Fig. 23 to 28). Wh ile t h is may rep resent a h ig h er degree of sea-level stability t h an else w h ere, t h e h ig h ly-fractured nature of t h e carbonate rock created a porous and permeable environment. e telogenetic Amuri Limestone was able to function more like a porous and permeable eogenetic rock be cause of t h e very large number of ow pat h ways suc h fracturing produced. e location of t h e Kaikoura caves on a peninsula, in isolated limestone points jut ting outward from t h at peninsula, argue against t h e caves being truncated fragments of past conduit ow caves. No rec h arge area exists for conduit ow caves in t h is setting. e Kaikoura Peninsula and limestone points are of a size similar to t h ose of t h e eolian ridges of t h e Ba h amas, w h ic h h ost ank margin caves. Despite t h e multiple ow pat h s t h e fractured Amuri Limestone presents, t h e main cave trends still follow master joint sets wit h in t h e bedrock. A similar pattern of cave de velopment occurred at Limestone Creek roug h Cave on t h e west coast of Sout h Island, w h ere one ank mar gin cave also transects a small limestone point (Fig. 18 and 19). At bot h Limestone Creek roug h Cave and Kaikoura, t h e caves contain well-developed p h reatic dissolutional features, no h ig h -velocity turbulent ow indicators, and numerous blind or dead-end passages and c h ambers. e Tube City outcrop at Limestone Creek (Fig. 20 and 21) demonstrates one of t h e main tenants of ank margin cave development, w hic h is t h e h eadward mi gration of t h e mixing front up t h e available ow pat h. At Tube City, p hreatic tubes of 1 to 2 m diameter follow joints inward and t h en abruptly taper and end, wit h t h e joint trace continuing inland. is abrupt c h ange from tube to joint is w h ere t h e mixing front was at t h e time sea level s hied position relative to t h e outcrop, and t h e tube was drained and mixing dissolution ceased. e presence of owstone material in t h ese tubes indicates t h at aer sea level fell away and t h e caves drained, t h ey were sealed c h ambers capable of forming h ard, dense calcite spe leot h ems. Cli retreat caused by wave energy (Fig. 20B) h as breac h ed t h ese small ank margin caves and opened t h em for inspection. e regular spacing of t h ese tubes along t h e outcrop face is anot h er indication of diuse ow of fres h water in a lens, suc h t h at mixing dissolu tion occurred across t h e face of t h e original rock outcrop w h erever a favorable joint allowed sucient fres h-water input to t h e lens margin. Regular spacing of p hreatic caves is also evident at Po h ara and Paturau River on Sout h Island (Fig. 13 to 16), and at Whirinaki and Kaw hia Harbour on Nort h Island (Fig. 6 to 9). Kaw hia Harbour is especially compelling, as small dissolution tubes and one large ank margin cave exist in t h e current interand supra-tidal environment, w hile t h e hill above h as a literal ring of small caves at approximately similar elevations near t h e crest of t h e hill. is hill, at a hig h er sea-level position, would h ave been an island. To explain t h ese Kaw hia Harbour caves as truncated fragments of a distributary conduit ow cave system would place a very hig h density of unconnected caves in a very small area. However, t h e presence of ig nimbrite in karst depressions adjacent to t his hill speaks to a long time of karst activity in t his area, and truncation of an ancient complex conduit-ow cave system cannot be discounted. e presence of relatively delicate p hreatic disso lutional features in voids found in coastal carbonates is a good indicator of ank margin cave development. At Paturau River, ank margin caves h ave been breac h ed by modern coastal processes. Preserved features suc h as bell h oles, small bedrock pillars, and dissolutional fretwork indicate t h at p hreatic voids h ave been breac h ed by wave erosion. e presence of t h ese voids as discrete c h ambers, isolated from one anot h er, and t h e absence of hig h-veloc ity turbulent ow markings, indicate t h at t h ese Paturau River voids are breac h ed ank margin caves. J OHN E. MY LROIE J OAN R. MY LROIE & C AMPBELL S. N ELSON

PAGE 23

ACTA CARSOLOGICA 37/1 2008 37 Indicators of tectonic movement Many of t h e ank margin caves described, suc h as at Kaikoura and Whiri naki, are elongated in t h e vertical direction. Classic ank margin caves from t h e Ba h amas are noted for t h eir wide h orizontal extent, but t h eir small vertical extent, an out come of development in t h e t hin, distal margin of a stable fres h-water lens in a very permeable limestone. e key h ere is lens stability. If t h e lens is migrating as a result of sea-level c h ange, t h en t h e dissolutional environment migrates vertically as well. Too fast a lens migration, and no macroscopic voids develop, as during Pleistocene glacioesutatic c h anges in t h e Ba h amas, w h ere caves are found at t h e elevations of t h e sea-level hig h stands and lowstands only. But if lens migration is slow, t h e dissolu tional environment also migrates slowly, leaving a verti cally extended void be hind. Because t h e lens is resident at any elevation for only a s h ort time period, t h e widt h of t h e void is small. is void widt h migrates as t h e lens migrates. Jumps and discontinuities of t his vertical en largement, especially if mimicked by many voids in an outcrop, are a measure of a sudden jump in lens position. e small size of most observed ank margin caves, and t h e vertically-extended nature of caves in some outcrops, speaks to t h e rapid tectonic movement of t h ese carbon ate outcrops in New Zealand. e regular, circular tubes found in ot h er outcrops, indicate stable sea-level condi tions followed by sea level c h ange too rapid to create a macroscopic void. In t h ese outcrops, suc h as Tube City at Limestone Creek (Fig. 20 and 21), a single sea-level pause is recorded. Tectonics in New Zealand is compli cated (e.g. W illiams, 2004), and every site discussed h as h ad a dierent tectonic history. Some caves, suc h as at Whirinaki, could be Holocene in origin; ot h ers, suc h as at Kaikoura, appear to be Pleistocene in origin. Does New Zealand h ave ank margin caves? If it is assumed t h at fres h water is disc h arged diusely from t h e land to t h e sea in coastal locations, t h en it can also be assumed t h at mixing of fres h water and sea water will occur. If t h e rock units present in t h e coastal location are carbonate rocks, it is reasonable to also assume t h at mixing dissolution will occur to some extent. In eoge netic carbonate rocks, diagenetic immaturity results in a h ig h ly porous and permeable be h avior across t h e bulk of t h e rock material. e results are caves wit h globular morp h ologies, and caves t h at act independently even of close neig hbors. Almost wit h out exception, eld study of coastal eogenetic carbonate rocks around t h e world h as described ank margin caves wit h morp h ologies and c h aracteristics very distinct from conduit-ow stream caves. Conduit-ow stream caves, w h et h er in eogenetic or telogenetic carbonate rocks, tend to follow linear ow pat h s. In telogenetic coastal carbonate rocks, low matrix permeability and h ig h bedding, joint, and fault permeability restrict diuse ow to linear pat h ways. erefore, ank margin caves in telogenetic carbonate rocks could be expected to h ave a muc h more linear pattern t h an ank margin caves in eogenetic carbonate rocks. As a result, in telogenetic carbonate rocks, dif ferentiating between senescent ank margin caves, and fragmented and degraded conduit-ow stream caves, can be dicult. Overprinting by epikarst processes, and by marine p h ysical processes, can add to t h e dif culty. Despite t h ese diculties, is t h ere any reason to assume t h at mixing dissolution will not occur in coastal telogenetic carbonate rocks? In t h e absence of a reason to disbelieve t h e existence of telogenetic ank margin caves, t h e question becomes h ow are t h ey congured, and t h en, h ow are t h ey identied? e observations presented h ere are an attempt to answer t h at question. It is t h e opinion of t h e aut h ors t h at ank margin caves h ave developed in t h e coastal carbonate rocks of New Zealand, and t h at t h e examples presented demonstrate t h ey can be dierentiated in t h e eld from ot h er cave types, bot h karst and pseudokarst. S UMMAR Y Analysis of nine coastal carbonate locations on Nort h and Sout h Island, New Zealand, indicate t h at flank margin caves do exist on t h ese two islands. T h e caves are muc h smaller t h an many Ba h amian flank mar gin caves, w h ic h are known to h ave formed in 12,000 years or less. T h is small size is t h e result of denser and less permeable telogenetic carbonate h ost rocks, and t h e rapid c h anges in sea level t h at occur in t h is environment, w h ere glacioeustasy is overprinted by very active tectonics. T h e fres h -water lenses in New Zealand were in a single position for an amount of time muc h less t h an t h at available in t h e tectonicallystable Ba h amas. Most of t h e flank margin caves de scribed h ave passage patterns t h at reflect t h e controls of dissolution provided in dense teleogenetic rocks by bedding planes, joints, and faults. T h e largest flank margin caves occur w h ere tectonics h as caused rock fracturing at t h e centimetre scale, creating a very po rous and permeable rock condition for t h e fres h -water lens. Epikarst development in teleogenetic carbonate rocks creates a dissolutional environment, t h at w h en overprinted by marine processes, makes it difficult to F LANK M ARGIN C AVE D EVELOPMENT IN T ELOGENETIC L IMESTONES OF N E W Z EALAND

PAGE 24

ACTA CARSOLOGICA 37/1 2008 38 A CKNO W LEDGMENTS e aut h ors t h ank Mississippi State University for pro viding Jo hn Mylroie wit h a sabbatical, and Joan Mylroie wit h a leave of absence, so t h at t his researc h could be done. e University of W aikato in Hamilton, New Zea land, provided logistical support for t h e project. Step h anie Nyman provided essential ground transport on Nort h Island. e Department of Geosciences, Missis sippi State University, assisted wit h travel costs. Leif Myl roie and Max Oulton are t h anked for map production. Paul W illiams is especially t h anked for leading eld trips, providing insig h t on New Zealand geology and caves, exceptional h ospitality, and overall enjoyable discussions t h at h elped guide t his researc h. J OHN E. MY LROIE J OAN R. MY LROIE & C AMPBELL S. N ELSON assess t h e role of mixing dissolution in creation of t h e observed caves. e presence of ank margin caves, and t h eir s h ape and conguration, provide useful information about sea-level position stability. While t h e absence of ank margin caves in a carbonate outcrop proves little, verti cally-elongated ank margin caves, or multiple but dis crete h orizons of ank margin caves, indicate sea-level c h ange t h at was rapid but not too rapid. If t his t hres h old of sea-level c h ange rate could be quantied, t h en ank margin caves would oer a dierent measure of tectonic activity. e mere presence of ank margin caves in a carbonate outcrop is an indication of subaerial exposure of t h at carbonate outcrop, suc h t h at it received meteoric rec h arge and built a fres h-water lens. e presence of hig h-elevation (375 m) ank margin cave remnants on Te Mata Peak, Nort h Island, indicates t h at t h e carbonate unit was subaerially exposed w hile in a coastal setting. e survival of ank margin caves on Te Mata, and at t h e ot h er described outcrops, is a measure of t h e rapid rate of upli and/or slope or cli retreat in t h at area. Be cause ank margin caves develop in t h e distal margin of t h e fres h-water lens, t h ey are vulnerable to exposure and subsequent complete destruction by erosion. e exis tence of senile ank margin caves, especially in currently interior locations, puts a boundary condition on denuda tion rates in t h at area. e documentation of ank margin caves in New Zealand for t h e rst time is not just an exercise in stamp collecting. e s h ape, size, abundance, and mere exis tence of ank margin caves provide geoscientists wit h information about a variety of tectonic, surcial, and h y drological processes. Back, W ., B.B. Hans h aw, J.S. Herman, & J.N. Van Driel, 1986: Dierential dissolution of a Pleistocene reef in t h e ground-water mixing zone of coastal Y ucatan, Mexico: Geology, 14, 137-140. Bgli, A., 1980: Karst Hydrology and P h ysical Speleol ogy, p. 284, Springer Verlag, New Y ork. Bottrell, S.H., J.L. Carew, & J.E. Mylroie, 1993: Bacterial sulp h ate reduction in ank margin environments: Evidence from sulp h ur isotopes. In: White, B. ed., Proceedings of t h e Sixt h Symposium on t h e Geol ogy of t h e Ba h amas, Port C h arlotte, Florida, Ba h a mian Field Station, 17-21. Carew, J.L., & J.E. Mylroie, 1995: Quaternary tectonic Stability of t h e Ba h amian Arc hipelago: Evidence from fossil coral reefs and ank margin caves: Qua ternary Science Reviews, 14, 144-153. Caron, V., C.S. Nelson, & P.J.J. Kamp, 2006: Microstratig rap h y of calcite cements in Pliocene cool-water limestones, New Zealand: relations hip to sea-level, burial and exh umation events. In Pedley, H.M., and Carannante, G., eds., Cool-W ater Carbonates: Depositional Systems and Palaeoenvironmental Controls. Geological Society, London, Special Pub lications 255, 337-365. C h en, J.H., H.A. Curran, B. White, & G.J. W asserburg, 1991: Precise c hronology of t h e last interglacial pe riod: 234 U-230 data from fossil coral reefs in t h e Ba h amas: Geological Society of America Bulletin, 103, 82-97. C h oquette, P.W ., & L.C. Pray, 1970: Geologic nomencla ture and classication of porosity in sedimentary carbonates: American Association Petroleum Ge ologists Bulletin, 54, 2, 207-250. R EFERENCES

PAGE 25

ACTA CARSOLOGICA 37/1 2008 39 F LANK M ARGIN C AVE D EVELOPMENT IN T ELOGENETIC L IMESTONES OF N E W Z EALAND Das h er, G.R., 1994: On Station: A Complete Handbook for Surveying and Mapping Caves, National Speleo logical Society, p 242, Huntsville, Alabama. Dreybrodt, W ., 2000: Equilibrium c h emistry of karst wa ter in limestone terranes, In Klimc h ouk, A.B., Ford, D.C., Palmer, A.N. & Dreybrodt, W ., eds., Spelogen esis Evolution of karst aquifers: National Speleo logical Society, Huntsville, Alabama, 126-135. Folk, R.L., H.H. Roberts, & C.H. Moore, 1973: Black p h ytokarst from Hell, Cayman Islands, Britis h W est Indies: Geological Society of America Bulletin, 84, 2351-2360. Frank, E.F., J.E. Mylroie, J., Troester, E.C. Alexander, & J.L. Carew, 1998: Karst development and speleolo gensis, Isla de Mona, Puerto Rico: Journal of Cave and Karst Studies, 60, 2, 73-83. Hull, A.G., 1990: Tectonics of t h e 1931 Hawkes Bay eart h quake. New Zealand Journal of Geology and Geop h ysics 33, 309-322. Jenson, J. W ., T.M. Keel, J.R. Mylroie, J.E. Mylroie, K.W Staord, D. Taboroi, C. W exel, 2006: Karst of t h e Mariana Islands: e interaction of tectonics, gla cioeustasy and fres h-water/sea-water mixing in island carbonates: Geological Society of America Special Paper 404, 129-138. Kelley, K., J.E. Mylroie, J.R. Mylroie, C. Moore, P.J. Moore, L. Collins, V. Ersek, I. Lascu, M. Rot h, R. Passion, & C. S h aw, 2006: Eolianites and Karst Develop ment in t h e Mayan Riviera, Mexico. In Davis, R.L., & Gamble, D.W ., eds., Proceedings of t h e Twel h Symposium on t h e Geology of t h e Ba h amas and Ot h er Carbonate Regions, Gerace Researc h Center, San Salvador, Ba h amas, 88-99. Lace, M. J., 2008: Coastal caves of Puerto Rico: Journal of Coastal Researc h, 24, 508-518. McNeill, D.F., 2005: Accumulation rates from well-dat ed late Neogene carbonate platforms and margins: Sedimentary Geology, 175, 73-87. Moore, P.J., J.H. Martin, & J.E. Mylroie, 2007 (abstract): Rapid development of secondary porosity wit hin fres h water lenses of carbonate islands: Geological Society of America, Abstracts wit h Programs, 39, 6, 467. Mylroie, J.E. & J.L. Carew, 1990: e Flank Margin Mod el for Dissolution Cave Development in Carbonate Platforms: Eart h Surface Processes and Landforms, 15, 413-424. Mylroie, J.E., & J.L. Carew, 1995: C h apter 3, Karst devel opment on carbonate islands. In Budd, D.A, Harris, P.M. & Saller, A., eds., Unconformities and Porosity in Carbonate Strata: American Association of Petro leum Geologists Memoir 63, 55-76. Mylroie, J.E., J.L. Carew, & H.L. Vac h er, 1995: Karst de velopment in t h e Ba h amas and Bermuda. In Cur ran, H.A. and White, B., eds., Geological Society of America Special Paper 300, Terrestrial and S h allow Marine Geology of t h e Ba h amas and Bermuda, 251267. Mylroie, J.E. & J.R. Mylroie, 2007a, Development of t h e Carbonate Island Karst Model: Journal of Cave and Karst Studies, 69, 59-75. Mylroie, J.E., & J.R. Mylroie, 2007b (abstract), Quaterna ry geologic interpretations from ank margin caves, Kangaroo Island, Australia: Geological Society of America, Abstracts wit h Programs, 39, 6, 78. Nelson, C.S., 1978: Temperate s h elf carbonate sediments in t h e Cenozoic of New Zealand. Sedimentology 25, 737-771. Nelson, C.S., G.J. Harris, & H.R. Y oung, 1988: Burialdominated cementation in non-tropical carbonates of t h e Oligocene Te Kuiti Group, New Zealand. Sed imentary Geology 60, 233-250. Nelson, C.S., P.R. W ineeld, S.D. Hood, V. Caron, A. Pallentin, & P.J.J. Kamp, 2003: Pliocene Te Aute limestones, New Zealand: Expanding concepts for cool-water s h elf carbonates. New Zealand Journal of Geology and Geop h ysics 46, 407-424. Neuendorf, K.K.E., J.P. Me h l, & J.A. Jackson, 2005: Glos sary of Geology, Fi h Edition, American Geological Institute, p. 779. Owen, A.M., 2007: Tafoni caves in Quaternary carbon ate eolianites: Examples from e Ba h amas. Masters t h esis, Mississippi State University, p.187. Palmer, A.N., 1991: Origin and morp h ology of limestone caves: Geological Society of America Bulletin, 103, 1-25. Plummer. L.N., 1975: Mixing of sea water wit h calcium carbonate ground water. In E. H. T. Whitten, ed., Quantitative studies in geological sciences: Geologi cal Society of America Memoir 142, 219-236. Pillans, B., 1986: A late Quaternary upli map for Nort h Island, New Zealand. In W .I. Reilly & B. E. Harford, eds., Recent crustal movements of t h e Pacic region. Royal Society of New Zealand Bulletin 24, 409-417. Proctor, C.J., 1988: Sea-level related caves on Berry Head, Sout h Devon: Cave Science, 15, 2, 39-49. Raeisi, E., & J.E. Mylroie, 1995: Hydrodynamic be h avior of caves formed in t h e fres h-water lens of carbonate islands: Carbonates and Evaporites, 10, 2, 207-214. Rot h, M. J., 2004: Inventory and geometric analysis of ank margin caves of e Ba h amas. Masters t h esis, Mississippi State University, p. 117.

PAGE 26

ACTA CARSOLOGICA 37/1 2008 40 Rot h, M. J., J.E. Mylroie, J.R. Mylroie, V. Ersek, C.C. Ersek, & J.L. Carew, 2006: Flank Margin Cave In ventory of t h e Ba h amas. In Davis, R.L., and Gamble, D.W ., eds., Proceedings of t h e Twel h Symposium on t h e Geology of t h e Ba h amas and Ot h er Carbon ate Regions, Gerace Researc h Center, San Salvador, Ba h amas, 153-161. Taboroi, D., J.W Jenson, & J.E. Mylroie, 2004: Karren features in island karst: Guam, Mariana Islands: Zeitsc hri fur Geomorp h ologie. N.F. 48, 369-389. Taboroi, D., J.E. Mylroie, & K. Kirakawa, 2006: Stalac tites on tropical clis: Remnants of breac h ed caves or subaerial tufa deposits? Zeitsc hri fur Geomor p h ologie, 50, 117-139. Vac h er, H.L.& J.E. Mylroie, 2002: Eogenetic karst from t h e perspective of an equivalent porous medium: Carbonates and Evaporites: 17, 2, 182-196. J OHN E. MY LROIE J OAN R. MY LROIE & C AMPBELL S. N ELSON W alker, L.N., 2006: e caves, karst and geology of Abaco Island, Ba h amas. Masters t h esis, Mississippi State University, 241 p. h ttp://sun.library.msstate.edu/ ETD-db/t h eses/available/etd 03292006 153441/un restricted/L W alker_Geology_Abaco.pdf W aterstrat, W J., 2007: Morp h ometric dierentiation of ank margin caves and littoral, or sea caves. Masters t h esis, Mississippi State University, 201 p. h ttp://li brary.msstate.edu/etd/s h ow.asp?etd=etd 04052007 150907 White, S.Q., K. Grimes, & J.E. Mylroie, 2007: (abstract) e earliest time of karst cave formation. Time In Karst Symposium, Postojna, Slovenia, CD. W illiams, P.W ., 2004: e evolution of t h e mountains of New Zealand. Oxford University Press, New Y ork, 89-106.



PAGE 1

CAVE TURBIDITES J AMSKI TURBIDITI R. Armstrong L. OSBORNE 1 Izvleek UDK 551.3.051:551.44 R. Armstrong L. Osborne: Jamski turbiditi Sedimenti odloeni iz turbiditni h tokov so v recentni h jama h relativno redki, bolj pogosto pa se pojavljajo kot paleokraki sedimenti. Tako med jamskimi kot paleokrakimi turbiditi so najpogosteji kajmaniti, ki so nastali s presedimentacijo iz vorno morskega karbonatnega sedimenta. Tako v recenti h ja ma h kot v paleokraki h sekvenca h najdemo tudi nekarbonatne morske in karbonatne ter nekarbonatne sladkovodne turbidite. Ena najbolj kompleksni h sekvenc jamski h turbiditiov se na h aja v sklopu fosfatnega rudnika W ellinton Caves v Avstraliji. Sedimentacijo jamski h turbiditov, ki se odlagajo v sigasti h ponvica h, la h ko sproijo poplavni dogodki ali mona deevja. Kajmaniti se zelo verjetno odlagajo med transgresijo morja, sproijo pa ji h la h ko tsunamiji. Sladkovodni jamski turbiditi se najpogosteje odlagajo v poplavljeni h hipogeni h jama h trop ski h obmoij s sezonski h vlanim obdobjem in v obmoji h z obasnimi monimi nalivi. Kjune besede: jamski sedimenti, turbiditi, paleokras, kajmaniti. 1 Faculty of Education and Social W ork, A35, University of Sydney, NS W 2006, Australia; a.osborne@edfac.usyd.edu.au Received/Prejeto: 06.12.2007 COBISS: 1.01 ACTA CARSOLOGICA 37/1, 41-50, POSTOJNA 2008 Abstract UDC 551.3.051:551.44 R. Armstrong L. Osborne: Cave turbidites Turbidites are uncommon in caves, but are more common as palaeokarst deposits. Marine carbonate turbidites, called cay manites, are t h e most common cave and palaeokarst turbidites, but marine non-carbonate turbidites, fres h water carbonate turbidites and fres h water non-carbonate turbidites are also de posited in caves and preserved in palaeokarst sequences. One of t h e most complex sequences of cave turbidites occurs in t h e W ellington Caves P h osp h ate Mine in Australia. Cave turbidites form in ponded water in caves and may be triggered by oods and hig h intensity rain events. While caymanites are most likely to form during marine transgressions, t h ey can be emplaced by tsunami. Fres h water cave turbidites are most likely to form in ooded h ypogene caves located in t h e seasonally wet tropics and in areas wit h irregular hig h intensity rainfall events. Key words: cave sediments, turbidites, palaeokarst, cayman ites. I NTRODUCTION Turbidites can be deposited in still bodies of water in caves. ese range from large p hreatic lakes to small rim pools developed on t h e side of stalagmites positioned hig h above t h e water table. While ne graded-bedded sands and muds are t h e most common turbidites found in caves, large deposits resulting from t h e slumping of entrance facies and bio genic talus cones into p hreatic lakes can result in depo sition of bot h turbidite conglomerates (W alker, 1975) and all of t h e elements of t h e Bouma sequence (Bouma, 1962).

PAGE 2

ACTA CARSOLOGICA 37/1 2008 42 ere are few reports of cave turbidites in t h e literature. Most accounts de scribe palaeokarst deposits, wit h ma rine palaeokarst turbidites (cayman ites), receiving t h e greatest attention (see Korps, 2002). ere are fewer reports of non-carbonate turbidites lling karst cavities; t h e most striking is t h at of Marsc h alko and Mello (1993) w h o described a 38 m t hick sequence of graded sandstones and mudstones lling karst cavities exposed in Gom basek Quarry in eastern Slovakia. By far t h e most abundant occur rences of cavity-lling turbidites are t h ose found in surface exposures of palaeo karst deposits. Good examples of t h ese are t h e laminated micrites of Otoniar (1997). Cavity-lling turbidite deposits are frequently quite small, but stand out from t h e enclosing bedrock because of t h eir well-developed and oen colour ful laminations (Fig. 1A and 1D). Most palaeokarst turbidites ex posed in caves are caymanites, de scribed furt h er below. ere are very few reported examples of non-marine palaeokarst turbidites exposed in caves. Some of t h e best examples are t h e ex posures in t h e walls of Okno Cave in Slovakia, described by Osborne (2007; Fig. 1B). Fres h water palaeokarst car bonate turbidites occur in a quarry in sout h western Slovenia and an un roofed cave near Trieste, Italy. e W ellington Caves P h osp h ate Mine in central western NSW Austra lia, exposes a remarkable range of tur bidite facies exposed in a former cave (Fig. 1C). DISTRIBUTION F ig. 1: P alaeokarst turbidite exposures. A Laminated palaeokarst microturbidite with high initial dip at Douglas River, Northern Territory, Australia. Host rock is Cambrian Tindall Limestone. e age of palaeokarst is unknown. Coin diameter is 31 mm; B Ex posure of sequence of lithied palaeokarst sediments including turbidites, Okno Cave, Slovakia. Scale is 1 m. Host rock is Triassic Gutenstein Limestone. Turbidite is probably P alaeocene in age; C Sequence of graded-bedded sandstones exposed in the Wellington Caves P hosphate M ine, Wellington Caves, NSW, Australia. Distinct lines across strata are mud crack horizons. Host rock is Devonian limestone. Turbidite is P liocene in age; D Laminated microturbidite, in cutting on road between B asovizza and P adricano, Trieste Carso, Italy. Host rock is Late Cretaceous (Santonian) limestone. Turbidite is probably marine and Late Cretaceous (M astricjtian) in age. ENVIRONMENTS AND MECHANISMS OF DEPOSITION While turbidites from deep marine environments are best known to geologists, turbidites were rst described from Lake Mead, t h e water impounded by t h e Hoover Dam on t h e Colorado River in t h e USA by Grover & Howard (1938). W alker (1969) pointed out t h at turbidites are in dicative of a process of deposition in quiet water and not indicative of water dept h. e essential requirements for turbidite deposition are a still body of water and a rapid input of sediment or sudden disturbance of a sediment body wit hin t h e wa R. A RMSTRONG L. OSBORNE

PAGE 3

ACTA CARSOLOGICA 37/1 2008 43 ter. Gravity and t h e density dierence between t h e turbid and clean water does t h e rest. In t h e case of caves, t h e im portant components are a cave or small cavity, still water and a sediment source. A number of processes h ave been responsible for ponding in caves w h ere turbidites occur. Old h ypogene caves, poorly connected to t h e modern h ydrological sys tem, can easily ood, as t h ere is no simple pat h for water to ow out of t h e cave. e ponds in w hic h t h e W elling ton Caves P h osp h ate Mine turbidites formed probably resulted from sediment blockage of poor h ydrological connections. Local ponding and sediment disturbance due to ooding h as been implicated in turbidite deposi tion in McEac h erns Deat h trap Cave, in Sout h Australia by Kos (2001). Sediment disturbance due to ooding is also involved in t h e case of active cave turbidite depo sition reported by Hosie & Smit h (2005). Turbidites in Ola h ola Cave, Norway formed in an ice-dammed lake (Valen et al., 1995). While marine transgression and slumping of sedi ment cones inside caves appear to be t h e most common mec h anisms for cave turbidite deposition, more exotic origins h ave been suggested. Burney et al. (2001) in t h eir study of t h e Mr h rulepf Caves and Sink h ole on Kauai, Hawaiian Islands, described a unit consisting of alloc h t h onous stones and fractured bedrock, 1 m t hick on t h e seaward facing wall of t h e cave and extending t hroug h out t h e cave as turbidite fans and gravel beds. Burney et al. proposed t h at t his unit resulted from a marine overwas h probably due to a tsunami inundating t h e site. SOME SPECIAL FEATURES OF CAVE TURBIDITES T HE I NFLUENCE OF CAVE GEOMETR Y Turbidites mostly occur in marine and lacustrine envi ronments w h ere t h e depositional front is able to spread out, resulting in t h e unconstrained deposition of tur bidite fans. is will occur in caves if t h e lake or pond is large relative to t h e size of t h e current, as in t h e case of Mr h rulepf Caves and Sink h ole described by Burney et al. (2001). Wh en turbidites are deposited in narrow or network caves, we s h ould expect t h e cave walls to c h annel t h e progress of t h e turbidity current. is interaction needs to be studied using p h ysical models in umes. In t h e ab sence of suc h studies, we mig h t expect a large current in a narrow cave to be h ave like a current in a submarine canyon (see Fig. 10 of W alker 1975). e current would deposit conglomerate facies along t h e main c h annel, wit h fan lobes extending into side passages depositing gradedbedded sands and distal muds (Fig. 2). H IGH INITIAL DIP Many palaeokarst microturbidites h ave steeply-dipping laminae (Fig. 1A). is can lead to t h e false assumption t h at t h ese sediments are deformed. It is more likely t h at t h ese are s h owing initial dip, resulting from t h e tendency for ne sediments in caves and microcaves to stick to t h e walls of t h e cavity. Bull (1977) s h owed t h at ne-grained sediments in caves could remain stable on slopes in ex cess of 70 degrees and may ad h ere to vertical surfaces. F ig. 2: Turbidites in a ooded network cave. A Sketch of a ood ed network cave with a window at point X in the cave ceiling. Sediment entering the cave through the window can form a talus cone. e talus cone can slump into the still ponded water, ini tiating turbidity current; B Hypothetical distribution of facies deposited by turbidity current propagated from point X. CAVE TURBIDITES

PAGE 4

ACTA CARSOLOGICA 37/1 2008 44 Wh ere t h e bedrock is h orizontally bedded, it is clear t h at t h e sloping laminae are a reection of t h e initial dip of t h e turbidite deposit. If t h e bedrock is dipping and t h e lled cavity developed along dip t h en it becomes possible to confuse disconformable palaeokarst wit h bedrock (see Fig. 1 of Osborne, 2000). ACTIVE TRUBIDITE DEPOSITION IN CAVES Unlike uvial cave sediments, w h ose transport and deposition is easily observed in accessible caves, eyewit ness accounts of turbidity currents depositing sediment in caves are rare. ere are two reasons for t his; rstly, most of t h e palaeokarst microturbidite deposits were de posited in environments too small for h uman access and secondly cave divers are only just beginning to explore caves w h ere turbidity currents are likely. R EPORT OF AN ACTIVE CAVE TURBIDIT Y CURRENT Hosie & Smit h (2005) reported a potentially fatal cave diving incident t h at took place in June 2005 in a cave near Kununurra in seasonally wet tropical nort h ern W estern Australia. eir account concerns a dive into a cave w h ere visibility h ad c h anged overnig h t from rela tively clear to absolutely atrocious (Hosie & Smit h, 2005, p. 24). Paul Hosie followed t h e dive line into t h e cave, but on his return found t h e way blocked and t h e dive line buried under sand. A turbidity current h ad apparently owed, blocking a constriction in t h e cave be hind him. Despite almost losing consciousness, Paul Hosie was able to dig his way out and return safely to t h e surface. e report is h arrowing in detail, but Hosie and Smit h s initial conclusion is of great interest: Turbid ow causing silt slump and cave blockage is a very real risk in Australias top end caves. Annual wet season deluges will most likely reset t h ese traps (Hosie & Smit h, 2005, p. 27). EXAMPLES OF PALAEOKARST AND RELICT TURBIDITE DEPOSITS T HE WELLINGTON C AVES P HOSPHATE M INE e Ph osp h ate Mine at W ellington Caves is a system of mine s h as and drives excavated t h roug h an almost completely sediment-lled cave. e mine intersects and exposes a sequence of lit h ied and partly lit h ied cave sediments ranging in age from Neogene to Late Pleistocene (Osborne 1982, 1997, 2001). e Pliocene formations, w h ic h include t h e main p h osp h ate deposit, contain a range of facies indicating deposition by turbid ity currents. e sequence contains a number of h ig h ly unusual rock types, produced from guano piles, ric h in bone fragments excreted by carnivorous bats, slumping into still ponds. Osseous Sandstones (osseous grainstones) Osseous sandstones (Osborne, 1982) are clast-support ed, graded-bedded sandstones and ne conglomerates in w hic h t h e sand fraction is almost entirely composed of bone fragments. Osseous sandstones occasionally occur as t hick beds up to 170 mm, but more commonly as sequences of graded layers 50 mm t hick, oen terminating in t hin mud crack h orizons (Fig. 1C). Figures 3A and 3B s h ow Bouma divisions A and B developed in t h e osseous sand stones. Coarse blocky spar h as strongly recemented t h ese rocks (Fig. 4A and 4B). e initial cement is oen p h os p h ate, or acicular carbonate, but later percolating water frequently replaces t his wit h equant spar (Fig. 4C). Sparcemented osseous sandstones are strong rocks, w hic h can resemble quartz arenites to t h e naked eye, and may be confused wit h bedrock. Laminates Ripple laminations, Bouma division C, do occur, but are relatively uncommon in t h e P h osp h ate Mine turbidites (Fig. 3C). e brown layers in t his specimen are com posed of clay w hile t h e lig h ter-coloured layers in t h e ripples are made of p h osp h atic mudstone wit h scattered bone fragments. Secondary w hite apatite lls irregular cavities below t h e ripples. Phosphatic mudstone e main p h osp h ate deposit at W ellington consists of approximately four metres of poorly-bedded p h osp h at ic mudstone, representing units D and E of t h e Bouma sequence. In t hin section, it is seen to consist largely of p h osp h atic mud wit h a few scattered grains of vein quartz and small bone fragments (Fig. 4F). R. A RMSTRONG L. OSBORNE

PAGE 5

ACTA CARSOLOGICA 37/1 2008 45 C OARSE FACIES e most important consideration wit h coarse cave tur bidite facies is t h e origin of t h e coarse clasts. While sur face-derived clasts are t h e main coarse components of slump and mudow breccias in caves, t h ese materials are generally too dense for transport by turbidity currents in caves. For t his reason, rip up clasts of older lit hied cave deposits are t h e dominant large clasts in coarse cave turbidites. Two types of coarse turbidite facies occur in t h e W ellington Caves P h osp h ate Mine se quence. I described t h ese as disorganised conglomerate and organised conglomerate following W alker (1975). Disorganised Conglomerate ere is one relatively small deposit of dis organized conglomerate in t h e P h osp h ate Mine. It consists of subangular clasts of laminated brown and grey p h osp h atic mud in a brown matrix (Fig. 3D). ese large clasts are all redeposited fragments of older cave sediment. In t hin section, t h e matrix appears as a poorly-sorted mixture of clay pellets, quartz grains (some coated wit h clay) and p h osp h atic mud wit h patc h es of acicular p h osp h atic cement (Fig. 4D). Organised Conglomerate e organized conglomerate exposed in t h e W ellington Caves P h osp h ate Mine is called c h ocolate c hip rock, as it consists of aligned clasts of dark brown mudstone set in an o-w hite matrix (Fig. 3E). In t hin section, t h e matrix appears as osseous sandstone wit h secondary spar cement (ow hite colour) almost completely replacing t h e original muddy matrix (Fig. 4E). In ad dition to t h e brown mudstone clasts, t h ere are also smaller grey p h osp h atic mudstone lit h oclasts. Some sand-sized quartz grains are present in t h e osseous sandstone matrix. As wit h t h e disorganized conglomerate, all t h e large clasts in t h e organized conglomer ate are re-worked cave sediments. F RESH W ATER CARBONATE TURBIDITES Carbonate muds and sands, can be deposited by turbidity currents in t h e p hreatic zone or as micro-turbidites in perc h ed water bodies in t h e vadose zone. Fine-grained fres h water turbidites, deposited in h olokarsts are fre quently dicult to distinguis h from marine carbonate mudstones. While t h ey lack obvi ous fossils so do many marine lime mudstones. e carbonate sand may be derived from decom posing speleot h ems, disaggregating coarse marble bed rock or from precipitation in carbonate-saturated wa ter. e ne lime mud may be precipitated in saturated water, t h e cave equivalent of w hiting formed in tropical seas (see Bat h urst, 1975, p. 137) or may be a product of incomplete solution as described by Zupan Hajna (2003) F ig. 3: Turbidite facies from the Wellington Caves P hosphate M ine. A Lacquered sawn block. Lower part is osseous sandstone and upper part is laminated clay representing B ouma divisions A and B; B Lacquered sawn block, consisting of a thick graded bed, a laminated bed, a thin graded bed and at the top a thin lami nated clay, representing B ouma divisions A, B A, B; C Lacquered sawn block showing ripple laminations, representing B ouma division C; D Lacquered sawn block of disorganized conglomerate. Note that the clasts are composed of older cave sediment, not bedrock or surcial deposits; E Exposure in wall of P hosphate M ine showing organized conglomerate. Note aligned chocolate chips (lithied cave sediment lithoclasts) in upper right of image. CAVE TURBIDITES

PAGE 6

ACTA CARSOLOGICA 37/1 2008 46 or result from biogenic weat h ering (micritization) of t h e bedrock. In t h e absence of convincing eld or microfossil evidence, c h emical and isotopic analyses may be t h e only means of distinguis hing between lit hied fres h water car bonate cave muds and t h eir marine counterparts. A good example of a fres h water carbonate turbi dite comes from an unroofed cave in t h e surface of t h e karst plateau adjacent to Trieste, Italy. It consists of nely F ig. 4: Wellington P hosphate M ine thin sections. A & B in sections of osseous sandstone, A is under plane-polarized light and B is under crossed nicols. In plane-polarised light, the bone fragments are honey coloured, while their low birefringence makes them almost black under crossed nicols. Note spar cement; C in section of osseous sandstone, showing detail of honeyyellow coloured bone fragments. Note brown remnants of origi nal matrix attached to some bone fragments and secondary spar cement; D in section of disorganized conglomerate crossed nicols. Note yellow-grey phosphatic mudstone lithoclasts behind gure number and scale and elongate brown mudstone lithoclast in lower frame; E in section of organized conglomerate in plane-polarized light. Note aligned small grains and dark (brown mudstone) lithoclast in upper half of frame; F in section 10X crossed nicols of phosphatic mudstone from main phosphate de posit. Grey background is phosphatic mud, which is resolved at high power (500x) as radiating crystals. Small light grains are vein quartz clasts. F ig. 5: F reshwater carbonate turbidites and caymanites. A in section of lime mudstone from an unroofed cave in the Trieste Karst, Italy; B in section of graded-bedded carbonate from a lled cave intersected by rnotie Quarry southwestern Slove nia. Graded laminae are approximately 4 mm thick; C P alaeo karst caymanite microturbidite from P odgrad, M atarsko P odolje, southwestern Slovenia. B edrock is Upper Cretaceous (Stantonian) limestone; caymanite is probably Late P alaeocene in age. Image is by B Otoniar; D in section, showing details of caymanite illustrated in C. Note laminae at base overlain by thick graded bed. Image eld of view is approximately 20 mm long. Image is by B Otoniar. laminated lime mudstone recrystallized to microspar, in w hic h a few large idiomorp hic calcite crystals h ave grown (Fig. 5A). Anot h er example of a probably fres h water gradedbedded carbonate comes from a lled cave intersected by rnotie Quarry in sout h west Slovenia. In t hin section, a series of graded lamina, approximately 4 mm t hick are visible (Fig. 5B). e laminae grade from ne lig h t co loured microspar at t h e base to brown micrite at t h e top. R. A RMSTRONG L. OSBORNE

PAGE 7

ACTA CARSOLOGICA 37/1 2008 47 CAVE TURBIDITES Marine carbonate sands and muds, deposited in caves ooded by t h e sea, frequently resemble normal lime stones. Jones (1992) described graded-bedded limestones deposited in caves of t h e Cayman Islands during peri ods of elevated sea level in t h e Cainozoic as caymanites. Similar graded-bedded marine palaeokarst sediments of probable Carboniferous age were recognised in east ern Australia by Osborne (1991, 1993). Korps (1998) and Korps et al. (1999) described multiple generations of caymanites from t h e Buda Hills of Hungary. Korps (2002) listed 32 caymanite deposits internationally, rang ing in age from Middle Cambrian to Quaternary. Caymanites can occur as microturbidite deposits (Figs. 5C and 6C) or more rarely as deposits lling caves large enoug h for h umans to enter (Fig. 6B). It is t h e out crop/exposure pattern and stratigrap hic relations hips of t h ese rocks, not t h eir petrograp h y, w hic h allows t h em to be recognised as lit hied cave deposits. Korps (pers. comm.) divides caymanites into con formable and unconformable, t h e idea being t h at con formable caymanites are likely to be syndepositional and h ence t h e oldest, w hile unconformable caymanites will be post-tectonic and t h us t h e youngest. While t his is a good rule of t h umb, it is not always reliable. e cayman ites at Ida Bay in Tasmania, Australia, described by Os borne & Cooper (2001) and s h own in Figure 6A are dis conformable rat h er t h an conformable or unconformable. is tells us little about t h eir age, as t h e enclosing Ordo vician limestone is essentially undeformed. Similarly, t h e steeply-dipping laminations in Figure 1A are denitely unconformable wit h t h e bedrock, but t h e bedrock h ere is almost h orizontally bedded Cambrian limestone. While caymanites h ave similar sedimentology, t h at is t h ey range in size from mud to gravel and include graded-bedded sands, laminated muds and ot h er units of t h e Bouma sequence, t h ey vary considerably in com position. is is because coarse grains in caymanites are a sample of t h e carbonate debris, i.e. t h e biota wit h carbon ate skeletons present in t h e ocean w h en t h e karst cavities were inundated. is similarity and diversity can be seen in Figures 5, 6 and 7. e caymanites from Ida Bay s h own in Figures 6A and 7A and from t h e underground ex posure at Jenolan Caves s h own in Figures 6B and 7B are bot h probably of Early Car boniferous age, yet t h ey are very similar in sedimentology to t h e Late Palaeocene cay manite from Slovenia s h own in Figure 5D. e large clasts in 7A and 7B are fragments of crinoid ossicles seen more clearly in Fig ure 7C, a t hin section of a coarser unit from a related deposit at Colong Caves, NSW CA Y MANITES: MARINE CARBONATE TURBIDITES F ig. 6: Caymanite outcrops. A Caymanite ex posed in wall of Lune River Quarry, Ida B ay, Tas mania, Australia. Caymanite is light-coloured and laminations are visible. Host rock, dark grey is Ordovician Gordon Limestone. e cayman ite is probably Early Carboniferous in age. Hat is approximately 320 mm in diameter; B Cay manite exposed in wall of River Cave, Jenolan Caves, NSW, Australia. Note bedding with gentle apparent dip to the le. Host rock is Silurian Jenolan Caves Limestone. e caymanite is prob ably Early Carboniferous in age; C Cayman ite exposed in F enygyngye Quarry, B udapest, Hungary. Host rock is Upper Eocene Szpvlgy Limestone. e palaeokarst is Late Eocene, see Korps (1998); D Caymanite exposed in wall of Cathedral Cave, Wellington Caves, NSW, Australia. Host rock is Devonian Garra F orma tion. e age of the caymanite is unknown. B lack squares on scale = 10 mm.

PAGE 8

ACTA CARSOLOGICA 37/1 2008 48 R. A RMSTRONG L. OSBORNE ese eastern Australian caymanites resemble Carbon iferous and Permian cold water limestones t h at formed w h en Australia was part of Gondwana and located at hig h latitudes. e coarse grains in Figure 5D are fora minifera indicating a quite dierent marine environment. Two quite dierent caymanites are s h ow in Figures 7D and 7E. Figure 7D is a pelletal caymanite wit h poor grading. It is overlain by t h e lime mudstone wit h ostra cod fragments s h own in Figure 7E. ese are unconformable caymanites ling cavi ties in Silurian limestone at Borenore Caves in central western NSW Australia. eir age remains undetermined. At bot h t h eir original locality in t h e Cayman Islands (Jones, 1992, 1994) and in Hungary (Korps 1998; Korps et al., 1999), caymanites h ave provided insig h ts into past marine transgressions and sea levels. Cay manites elsew h ere h ave great potential to reveal information about marine transgres sions t h at occurred in t h e past for w hic h no ot h er evidence is preserved. F ig. 7: Caymanite thin sections. A in section 3.2x crossed nicols from Lune River Quarry, Ida B ay, Tasmania, Australia, showing top of one graded bed and the base of another; B P art of a sequence of graded beds and lime mudstones 3.2x crossed nicols from River Cave, Jenolan Caves, NSW, Australia. Note erosion at base of upper graded bed; C Coarse crinoidal grainstone 6.4x crossed nicols from exposure in Lannigans Cave, Colong Caves, NSW, Austra lia; D Graded-bedded pelletal caymanite 6.4x plane-polarized light from surface outcrop at B or enore Caves, NSW, Australia; E Lime mudstone with ostracod bioclasts 6.4x crossed nicols from same caymanite sequence as D, B orenore Caves, NSW, Australia. W HERE ARE TRUBIDITES FORMING IN CAVES TODA Y AND W HERE SHOULD W E EXPECT TO FIND OLDER ONES? ere is no problem wit h nding modern analogues for microturbidite and large-scale caymanites, but modern examples of non-marine turbidite deposition, particu larly in h uman-scale caves remain rare. So w h ere s h ould we look for active cave turbidite deposition and for relict cave turbidites? Firstly, we s h ould look for turbidites in caves t h at contain still bodies of water. ese could be caves blocked by ice or rockfalls and caves wit h poor drainage or h y drological connections. Hypogene caves s h ould be good targets. Flooding and extreme rainfall are factors in bot h Hosie and Smit h s example of active turbidite deposition

PAGE 9

ACTA CARSOLOGICA 37/1 2008 49 CAVE TURBIDITES and Kos example of a cave turbidite from Sout h Austra lia. Flooding and extreme rainfall events are c h aracter istics of bot h t h e Australian seasonally wet tropics and of t h e Australian continent in general wit h its droug h ts and ooding rains (Mackellar, 1911). On reection, t h e W ellington Caves P h osp h ate Mine turbidites are unsur prising, as t h ey formed in an Australian h ypogene cave. Hypogene caves in seasonally wet tropical regions, or regions wit h extremely variable rainfall are likely lo cations for cave turbidites in all continents. In t h e case of palaeokarst, t his will include locations t h at h ave h ad a seasonally wet tropical climate or extremely variable rainfall in t h e past. While caymanites usually result from sea level c h anges, Burney et al. (2001) suggested anot h er mec h anism for caymanite deposition, tsunamis. As well as looking for caymanites on carbonate coasts w h ere sea levels were hig h er in t h e past, we s h ould also be looking for caymanites along limestone coasts w h ere tsunamis occur. is includes carbonate coasts of t h e Indonesian arc hipelago. ACKNO W LEDGEMENTS I would like to t h ank Bojan Otoniar for long discussions over t h e years about palaeokarst and petrology bot h in t h e eld and looking down t h e microscope and also for t h e images in Figures 5C and 5D. I am also t h ankful to Ls zl Korps for his generosity in t h e eld and for h elpful discussions and correspondence. Andrej Mi h evc assisted wit h eld visits to rnotie Quarry and t h e Trieste Karst. I also wis h to record my t h anks to W ellington Council and t h e sta at W ellington Caves for t h eir support for my researc h for over t hirty years. Penney Osborne h elped wit h reading and correcting t h e dras. REFERENCES Bat h urst, R.G., 1975: Carbonate Sediments and their Dia genesis, Second Enlarged Edition. Developments in Sedimentology, 12, 658 pp., Elsevier, Amsterdam. Bouma, A.H., 1962: Sedimentology of some F lysch Depos its. Elsevier, 168 pp., Amsterdam. Bull, P.A., 1977: Laminations or varves? Processes and mec h anisms of ne grain sediment deposition in caves. Proceedings of t h e 7t h International Speleo logical Congress, S h eeld, 86-89. Burney, D.A., James, H.F., Burney, L.P., Olson, S.L., Kiku c hi, W ., W agner, W l., Burney, M., McCloskey, D., Kikuc hi, D., Grady, F.V., Gage, R. II & Nis h ek, R., 2001: Fossil evidence for a diverse biota from Kauai and its transformation since h uman arrival. Ecolog ical Monograp h s, 71/4, 615-641. Grover, N.C. & Howard, C.S., 1938: e passage of tur bid water t hroug h Lake Mead. Transactions of t h e American Civil Engineering Society, 103, 720-790. Hosie, P. & Smit h, K., 2005: Cave Diving Deat h Trap. Caves Australia, 166-167, 24-27. Jones, B., 1992: Caymanite, a cavity-lling deposit in t h e Oligocene-Miocene Blu Formation of t h e Cayman Islands. Canadian Journal of Eart h Science, 29, 720735 Jones, B., 1994: Void-lling deposits in karst terrains of isolated oceanic islands: a case study from Tertiary carbonates of t h e Cayman Islands. Sedimentology, 39, 857-876 Korps, L., 1998: P alaeokarst Studies in Hungary. Oc casional Paper, 195, 139 p., Geological Institute of Hungary, Budapest. Korps, L., Lantos, M & Nagymarosy, A., 1999: Timing and genesis of early marine caymanites in t h e h y drot h ermal palaeokarst system of Buda Hills, Hun gary. Sedimentary Geology, 123, 9-29 Korps, L., 2002: Are palaeokarst systems marine in ori gin? Caymanites in geological past. In: F. Gabrovsek (Ed.): Evolution of Karst from P rekarst to Cessation. Zalozba ZRC, PostojnaLjubljana, 415-424 Kos, A.M., 2001: Stratigrap h y, sedimentary development and palaeoenvironmental context of a naturally ac cumulated pitfall cave deposit from sout h eastern Australia. Australian Journal of Eart h Sciences, 48/5, 621-632. Mackellar, D., 1911: M y Country, e Closed Door. Aus tralasian Aut h ors Agency, 9-11, Melbourne.

PAGE 10

ACTA CARSOLOGICA 37/1 2008 50 Marsc h alko, R. & Mello, J., 1993: Turbidites as llings of cavities in Triassic limestones of t h e Silica Nappe (W estern Carpat hians, Pleivec Karst Plateau). Geo logica Carpat hica, 44/1, 35-42 Otoniar, B., 1997: Macroscopic paleokarstic features in Upper Cretaceous limestones of t h e Adriatic-Di naric carbonate platform (SW Slovenia). Proceed ings. 12t h International Congress of Speleology, Speleo Projects, La C h aux-de-Fonds, Switzerland, 417-420. Osborne, R.A.L., 1982: Cainozoic stratigrap h y at W el lington Caves, New Sout h W ales. Proceedings of t h e Linnean Society of New Sout h W ales, 107/2, 131147 Osborne, R.A.L., 1991: Palaeokarst deposits at Jenolan Caves, NSW Journal and Proceedings of t h e Royal Society of New Sout h W ales, 123/3-4, 59-73. Osborne, R.A.L., 1993: Geological Note: Cave formation by exh umation of Palaeozoic palaeokarst deposits at Jenolan Caves, New Sout h W ales. Australian Jour nal of Eart h Sciences, 40, 591-593. Osborne, R.A.L., 1997: Re h abilitation of t h e W ellington Caves P h osp h ate Mine: implications for Cainozoic stratigrap h y. Proceedings of t h e Linnean Society of New Sout h W ales, 117, 175-180. Osborne, R.A.L, 2000: Paleokarst and its Signicance for Speleogenesis. In: A.B. Klimc h ouk, D.C. Ford, A.N. Palmer &W Dreybrodt (Eds.): Speleogenesis, Evolu tion of Karst Aquifers. Huntsville, National Speleo logical Society, p. 113-123. Osborne, R.A.L., 2001: Karst geology of W ellington Caves, a review. Helictite, 37/1, 3-12. Osborne, R.A.L., 2007: Intensely lit hied palaeokarst deposits in Okno Cave, Demnovsk Valley (Slova kia). Geologica Carpat hica, 58/6, 565-578. Osborne, R.A.L. & Cooper, I.B., 2001: Sulde-bearing palaeokarst deposits at Lune River Quarry, Ida Bay, Tasmania. Australian Journal of Eart h Sciences, 48, 409-416. Valen, V., Larsen, E. & Mangerud, J., 1995: Hig h reso lution paleomagnetic correlation of Middle W eic h selian ice-dammed lake sediments in two coastal caves, western Norway. Boreas, 42, 141-153. W alker, R.G., 1969: e juxtaposition of turbidite and s h allow-water sediments: study of a regressive se quence in t h e Pennsylvanian of nort h Devon, Eng land. Journal of Geology, 77, 125-143 W alker, R.G., 1975: Generalised facies models for resedi mented conglomerates of turbidite association. Bul letin of t h e Geological Society of America, 86, 737748. Zupan Hajna, N., 2003: Incomplete Solution: Weather ing of cave walls and the production, transport and deposition of carbonate nes. Zalozba ZRC, 167 pp., Ljubljana. R. A RMSTRONG L. OSBORNE



PAGE 1

BROKEN SPELEOTHEMS AS INDICATORS OF TECTONIC MOVEMENTS PRETRTA SIGA KOT POKAZATELJ TEKTONSKIH PREMIKOV Stanka EBELA 1 Izvleek UDK 552.545:551.248 Stanka ebela: Pretrta siga kot pokazatelj tektonskih premikov V kraki h jama h pogosto najdemo pretrto in nepravilno sigo. Veinoma nastane zaradi nestabilnosti tal, in sicer zaradi sestave tal (nesprijet pesek in ilovica). Prisotnost dananje krio turbacije v nekateri h jama h v viji h nadmorski h viina h nas napeljuje na trditev, da je nekatere stare razpoke na siga h la h ko povzroila prisotnost ledu v jama h. Imamo pa tudi nekaj do bri h dokazov za dananjo in staro tektonsko aktivnost v kraki h jama h. V veini primerov je teko doloiti vzrok za pretrto sigo, ker se med seboj la h ko prekriva ve vzrokov. V Postojnski jami smo raziskovali nekaj primerov vezani h na tektoniko. Kljune besede: pretrta siga, tektonika, Postojnska jama, Slo venija. 1 Karst Researc h Institute ZRC SAZU, Titov trg 2, 6230 Postojna, Slovenia, sebela@zrc-sazu.si Received/Prejeto: 08.11.2007 COBISS: 1.01 ACTA CARSOLOGICA 37/1, 51-62, POSTOJNA 2008 Abstract UDC 552.545:551.248 Stanka ebela: Broken speleothems as indicators of tectonic movements In karst caves broken and non-ideal speleot h ems are very wide spread. Mostly t h ey develop because of t h e instability of t h e ground due to its composition (loose sand or loam). e pres ence of recent cryoturbation in some caves in hig h er altitudes suggests t h at some ancient breaks of t h e speleot h eme can be caused by ice in t h e cave. And we also h ave some good proofs for recent and past tectonic activity in karst caves. In many cases it is very dicult to determine t h e real reason for broken speleot h ems, because several reasons could be interacting. In Postojna Cave some examples related to tectonics were stud ied. Key words: broken speleot h ems, tectonics, Postojna Cave, Slo venia. I NTRODUCTION In Slovenia but also in ot h er countries t h e inuence of ac tive tectonics in t h e formation of karst caves is being dis cussed a lot. Aer an ent h usiastic period w h en researc h ers from Slovenia (Gospodari 1977), Belgium (Quinif & Genty 1998; Delaby 2001; Vandycke & Quinif 2001), France (Gilli 1986; Gilli 1992, Gilli 1999a etc.), Italy (Postpisc h l et al. 1991; Bini et al. 1992) related many bro ken speleot h ems wit h eart h quakes or at least wit h older tectonic events, in recent years some researc h ers dened ice (Kempe 2004; Gilli 2004) to be one of t h e principal causes for broken speleot h ems in t h e caves. Becker et al. (2005) determined caves to be a dif cult arc hive, refering to t h e complexity of t h e processes. is means t h at in t h e study of t h e cave as a complex en vironment we cannot select only one cause for broken speleot h ems, but need to consider more reasons t h at can interact. Tectonics in Slovenia is active today, and karst caves are a useful place to searc h for evidence of deformation. in stalactites, stalagmites, owstone, and h elictites may all be easily damaged, and t h erefore serve as indicators of paleoseismicity. Observations during and aer eart h quakes, modeling and laboratory experiments indicate t h at, except for some slender speleot h ems, cave forma tions usually do not break during an eart h quake. Most stalactites remain undamaged aer an eart h quake because t h eir natural frequencies are hig h er t h an t h e maximum for seismic frequencies (w hic h are 0,1

PAGE 2

ACTA CARSOLOGICA 37/1 2008 52 to 30 Hz). But, t hin and long stalactites may be broken (Lacave et al. 2001). Regarding ice as a cause for broken speleot h ems I t hink we need representative proofs. Delaby (2001) stat ed t h at it seems very unlikely t h at t h e Hotton cave was aected by frost or permafrost. is is conrmed by t h e total absence of cryo-clastic features in t h e surrounding limestone h ost rock. Frost would probably cause t h e sta lagmite to crumble w hile t h e observed breaks are very clean and occur in t h e lower h alf of t h e stalagmite (De laby 2001). Ice is a kind of cave lling t h at leaves almost no traces (Becker et al. 2006) e entrance of Snena Jama (1556 m above sea level) in Slovenia, is lled wit h ice and in t h e cave t h e ef fect of t h e ice on t h e speleot h ems is obvious (Gilli 2004). Regarding our experiences t h e presence and t h e eect of ice in Postojna Cave described by Kempe (2004) and Gilli (2004) is more doubtful. e ice was probably not so widespread in t h e cave as is described by Kempe (2004). W it h t his article I want to s h ow good proofs for broken speleot h ems related to tectonic deformations. An overview of selected literature will h opefully provoke new prospective studies. BROKEN SPELEOTHEMS IN KARST CAVES e principle causes recognized for speleot h em break age/disruption in karst caves are: 1) Instability of t h e ground due to its composition (loose sand or clay). 2) Removal of grounding due to water ow 3) Collapse of cave oor (for example into underly ing passage). 4) Gravitational deformation close to valleys 5) Ice 6) C h anges in drip water resulting in dissolutional loosening of ceiling deposits (for example t h e falling s hield in kocjanske Jame (Kranjc 1999)). 7) Eart h quakes (Bini et al. 1992) 8) Ant hropogenic impacts (accidental, mining) 9) Faunal impacts (cave bears, bats etc.) A BOUT BROKEN FLO W STONE IN S LOVENE CAVES In 1830, in Postojna Cave, Ho h enwart rst recognized t h at fallen stalagmites and stalactites required unusual conditions, and t h at t h ey mig h t represent geologically important information. He did not, h owever, determine t h e specic causes for broken speleot h ems. In Postojna Cave it was found t h at stalactites were broken due to ceiling collapse caused by eart h quakes, and also due to t h e removal of alloc h t h onous sediments (Gospodari 1968). e same aut h or examined t h e ori entation of broken stalactites and stalagmites in Rov za Veliko Goro passage of Postojna Cave. e passage is lled wit h ysc h sediments and covered by owstone and collapse blocks. Gospodari (1968) concluded t h at some stalagmites fell down from t h e ceiling toget h er wit h pieces of t h e limestone bedrock w h en aggressive percola tion water widened t h e ssures. Gospodari (1968) did not, h owever, nd any good evidence of eart h quake damage from t h e periods of owstone deposition t h at were from t h e Atlantic pe riod (about 6.000-3.000 BC). His directional analysis of t h e broken stalagmites from Rov za Veliko Goro did not reveal any signicant particular direction. He also noted t h at local ground s h aking could be initiated not only t hroug h tectonic processes, but also t hroug h ceiling collapses or t hroug h t h e development of collapse dolines nearby. However, suc h local events would not be of suf cient magnitude to break t h e owstone (w hic h was 2 metres t hick). In a subsequent study, Gospodari (1977) investi gated a broken speleot h em dated at 10.000 BC. In t his particular case t h e speleot h em was broken due to consol idation of t h e underlying loamy sedimentary oor. It was later overturned due to eart h quake activity. Compression of loamy sediments is also responsible for consolidation of t h e cave oor, and consequently t h e breakage of ow stones in Pisani Rov. is process is a long term one, and t h erefore cannot be caused by eart h quakes. Gams (2003) mentioned broken stalagmites in Zgornji Tartarus of Postojna Cave. He also described t h e consolidation of clastic sediments as a cause for broken stalagmites. Perko (1910) believed t h at a massive stalag mite called Zvrnjeni steber in Postojna Cave (Figure 1) is 150.000 years old, and fell down more t h an 67.000 years BC. He obtained t h ese age results by calculating and ex trapolating from measured deposition rates for owstone in t h e cave. On t h e Zvrnjeni steber some glass plates (6x3 cm) were installed about 30 years ago. Today some of t h em are broken and t h ere is also one t h at is still cemented but curved. It looks as if t h e crack t hroug h t h e stalagmite is still opening. e movements s h ould be connected wit h ground instability and not wit h tectonics. STANKA EBELA

PAGE 3

ACTA CARSOLOGICA 37/1 2008 53 Brodar (1966) t h oug h t t h at frost was t h e cause of bro ken owstone between clastic sediments in t h e entrance portion of Postojna Cave. Kranjc, in 1999 reports t h at be tween Marc h 2 and 3, 1999, a s h ield style stalactite fell 70 m from t h e ceiling of Velika Dvorana in kocjanske Jame. e weig h t of t h e speleot h em was estimated at 2.500 ki los. e speleot h em falling was attributable to c h anges in drip water c h emistry. Mih evc (2001) used Udating to determine t h at breakage of a owstone deposit in Divaka Jama occurred 176.000 years BC (wit h in t h e Mindel/Riss interglacial). He did not h ypot h esize any cause for t h e breakage. In anot h er cave, Jazbina v Rovnjah owstone deformed by cryoturbation was found to be 241 Ka old (Mih evc 2001). In t h e example of Snena Jama t h e ow stone in t h e entrance area of t h e cave was broken by frost action. In addition, owstones deeper in t h e cave were de formed by cryoturbation (Bosak et al. 2002). e observations (Gilli 2004) in Postojna Cave and Snena Jama na Radu hi reveal mec h anisms as t h e creep ing of ice or clay lling and in t his way explain most of t h e speleot h eme breaks. e most promising cause for natural speleot h em breakage in general in Postojna Cave is, according to Kempe (2004), cave ice. It must h ave formed in caves dur ing glacial maxima w h en permafrost spread t hroug h out nort h ern, eastern and central Europe. e ice could be t h e most prominent factor in explaining non-recent spe leot h em damage. In t h e side passages of t h e cave (Pisani Rov and Brezimenski Rov) t h ere are masses of broken stalagmites and speleot h em fragments in precarious po sitions. Cave ice oers an overall process to explain t h ese observations. It is suggested t h at all or parts of t h e Posto jna Cave were lled wit h ice during t h e last and earlier Glacials (Kempe 2004). T ECTONIC MOVEMENTS DURING THE FORMATION OF KARST CAVES Gams (2002) reported on tectonic deformation in t hree Slovene caves (Postojna Cave, Planinska Jama, Lukenjska Jama), occurring aer t h e passages h ad been formed. Gospodari (1964) identied tectonic deformations on t h e ysc h sediments in t h e articial passage in Postojna Cave. Maurin (1953) and W ojcik & Zwolinski (1959) found evidence of Riss/W rm tectonic movements due to s hied cross-sections in caves. In Frassino Cave (Italy), a p hreatic passage is oset along a bedding plane reverse fault. Dating of owstone s h ows t h at t his movement is older t h an 350 Ka (Bini et al. 1992). roug h researc h in Poloka Cave (Figure 2), Habi (1971) found t h at t h e passages in one part of t h e cave are developed close to t h e crossing of two faults. He in dicated t h at t h e passages are developed along N-S fault zones. However, h e did not mention tectonically oset passages. e fact is t h at Poloka Cave, w hic h is 10.800 m long, is located just about 300 m sout h from Ravne Fault, w hic h was active in t h e eart h quake of April 12, 1998 (Mw=5.6). In Kamnika Jama one of t h e passages ends wit h a tectonic fault. Urbanc (1982) described t h at t h e fault o sets t h e principal passage into two parts. In connection wit h t h at, I h ave to mention Predjama Cave (ebela 1996), w h ere t h e nort h ern wall of t h e passage at t h e beginning of Vzh odni Rov h as a generally east-west direction, and is basically limited wit h a fault zone (Figure 3). In t h at example, t h e cave is developed along t h e mentioned fault F ig. 1: Z vrnjeni steber in P ostojna Cave (photo J. Hajna). F ig. 2: P oloka Cave, NW Slovenia (photo S. ebela). BROKEN SPELEOTHEMS AS INDICATORS OF TECTONIC MOVEMENTS

PAGE 4

ACTA CARSOLOGICA 37/1 2008 54 zone, w hic h was active before t h e nal formation of t h e passage. ar et al. (2002) described t h at active tectonics along t h e nort h ern part of Cerknica Polje is very strong. ey compare t h e position of t h e systems of Postojna and Planina Caves, wit h Krina Cave, because between t h ose t h ere is an oset of 12 km, w hic h is due to dextral move ment along t h e Idrija Fault. During t h e exploration of t h e 1,2 km long and 173 m deep Hirsc h gruben h h le, abundant scratc h ed, s h eared and broken speleot h ems were found. Because Quaterna ry mass movement and ice-movement in t h e cave, w hic h are known to cause deformations on speleot h ems, can be excluded, t h ese deformations are presumably sismot h ems, i.e. speleot h ems t h at were broken or deformed by fault slip associated wit h seismic events. e cave is lo cated 8 km sout h from t h e master fault of t h e le-lateral Salzac h tal-Ennstal-Mariazell-Puc h berg-Line (SEMP). First results suggest t h at a seismic event wit h at least 20 cm oset (sinistral strike-slip) took place between 11-86 Ka (Plan et al. 2005). Unstuck parts of t h e vault, broken stalactites, and h orizontally s h eared columns s h ow t h at t h e W intimdou ine Cave in Marocco was and is still modeled during t h e time of vertical and h orizontal s his t h at correspond to t h e contemporary rising up of Hig h Atlas and its t hrust ing to t h e sout h. is dynamic is sprinkled by brief seis mo-tectonic events, suc h as are recorded in speleot h ems (Angelova et al. 2005). In t h e Roc h efort Cave (Belgium) t h e recent tectonic features are bedding planes reactivated as normal faults. e faults are still active because fres h scaling and fallen blocks are observable (Vandycke & Quinif 2001). In t h e s h a of Avent Abel and Calernaum (France), t h e recent movement of a t hrust fault h as cause t h e break age of many speleot h ems (Gilli 1992 and 1999 e). Gilli & Delange (2001) examined t h e active move ments of t hrust faults along w hic h karst caves are devel oped in France. ey found proof of displaced faults and broken owstone. E VIDENCES OF SEISMIC EVENTS IN THE CAVES Among t h e older scientists and relation between ow stone and eart h quakes, t h ere is a work by Becker (1929) about t h e Bing Cave in Germany and Han-sur-Lesse (Belgium). Postpisc h l et al. (1991) recognized t h at karst caves h ave great potential for t h e study of tectonic, especially paleoseismic, events. ey recognized t h at deviations from vertical growt h for stalagmites could be caused by local factors (suc h as movements of supporting blocks), or t h ey could be due to tectonic events and eart h quakes. e evidence for eart h quakes in karst caves, can be seen in: deviation of growt h axes (from t h e vertical), dier ences in owstone growt h, and colour dierences in spe leot h em bands, possibly resulting from c h anges in t h e p h ysio-c h emistry of percolation water input. Postpisc h l et al. (1991) examined t h e growt h of owstone in two dierent caves (Buco dei Buio and Spi pola near Bologna, Italy), particularly wit h respect to a strong eart h quake t h at occurred in t h e region on January 3, 1117. e epicenter was 15 km far from t h ese caves. ey found t h at t h e observed anomalies in t h e stalagmite growt h s are always related to eart h quakes or at least to tectonic events. In sout h ern Italy, t h e same researc h ers (Postpisc h l et al. 1991) examined Grotta Grande del Cervo. ey analyzed 25 stalagmites using 14-C and Udisequi librium met h ods. ey were able to connect some mor p h ological deviations in t h e speleot h ems wit h times of known seismic events. e eart h quake of December 1456 was particularly obvious. In t h e last 350 Ka t h ey believe t h at t h ere were 4 strong eart h quakes, wit h t h e 1456 one being t h e most recent. at one, as indicated by arc h aeo F ig. 3: e fault zone in Vzhodni Rov of P redjama (photo J. Haj na). STANKA EBELA

PAGE 5

ACTA CARSOLOGICA 37/1 2008 55 logical studies, closed t h e entrance to t h e cave. So, t h e eart h quakes did not only overturn t h e speleot h ems, t h ey also caused ceiling collapses and movements of walls. Some stalagmites did not continue to grow aerwards, because t h e stalactites over t h em (supplying drip water) h ad been displaced. Quinif & Genty (1998) spoke about t h e met h ods by w hic h deformed owstone can be used for t h e period of t h e Holocene and Upper and Middle Pleistocene. And eart h quakes from t h e Holocene can be studied by lami nations from t h e owstone. In any case it is necessary to do a correlation of t h e events from dierent caves. Quinif et al. (1994) also observed t h at displacements along bed ding planes could cause t h e same stalactite to feed dier ent stalagmites. Dating of disturbed stalagmites from Grotta Gigan te, Italy (Cucc hi & Forti 1989) s h owed t h at disruption in growt h (reorientation of t h e growt h axis) occurred at 25 Ka, 20 Ka, and 15 12 Ka. ey believe t h at t h e causes are seismic events. Aer t h e eart h quake of February 18, 1996 (magni tude 5.2) in t h e area of t h e eastern Pyrenees, France, 8 caves were examined (Gilli 1999 a). ey ranged in dis tance from 2 to 10 km from t h e epicenter. e h ypocen ter was at a dept h of 5 to 10 km. On t h e cave oor t h ey found t hin stalactites, w hic h h ad fallen from t h e ceiling. One of t h e caves, Barrenc du Paradet, is located directly on t h e active fault. e prevalent orientation of t h e fallen soda-straw stalactites was concordant wit h t h e orienta tion of t h e fault, and also probably accordant wit h t h e direction of maximum ground acceleration of t h e eart h quake. In t h e same cave some older broken soda-straws, w hic h are evidence of t h e eart h quake of 1922, could be found. ese are now covered wit h newer owstone de posits. It does not appear t h at t h e eart h quakes caused t h e breakage of all soda straws t h at were present, but rat h er only t h ose wit h structural anomalies. On May 3, 1887 t h ere was a 7.2 magnitude eart h quake in Sonora, Mexico. It probably inuenced t h e owstone about 100 km away at a cave at Sut h erland Peak, Arizona (Gilli 1999 b). Broken soda straws in t h e cave in Monaco (Grotte de lObservatoire) are related to t h e year 1887, w h en an eart h quake of magnitude 6 to 6.5 occurred. e cave is developed inside a t hrust fault on t h e sout h and between a vertical fold on t h e nort h. e lower layer of t h e broken owstone and soda straws in t h e cave is related to t h e eart h quake from t h e year 1564 (Gilli 1999 c). And t h ere is a collapse material wit hin Nimte Cave, Bulgaria (Angelova et al. 2003), w hic h h as been correlated to t h e 1928 eart h quake. In a study from t h e sout h-central massif in France, Camus et al (2001) made an investigation of owstone in Garrel Cave, w hic h is cut by a fault t h at, according to some aut h ors, was active in t h e upper Pleistocene. ey found, h owever, by analysis of unbroken owstone in stratigrap hic context, t h at t his fault h as not been active for 466 Ka. In t h e Belgian karst (Delaby & Quinif 2001) t h ey found proof for previously unknown eart h quakes. In Hotton Cave t h ey found t h at 23-55% of all stalactites are t h e broken t hin stalactites in dierent parts of t h e cave. ey explained t his as caused by an undocumented eart h quake wit h an epicenter somew h ere near t h e cave. Probably t h e strongest recorded eart h quake in eastern Belgium (September 18, 1692) did not leave any break age evidence in t h e caves. ere are many broken stalagmites present in t h e Milandre cave (Swiss Jura), situated about 40 km far from 1356 Basel eart h quake (Mw=6.9.5). Only most of t h e long and slender stalactites are expected to break during a reasonably strong eart h quake, wit h 0.3 g
PAGE 6

ACTA CARSOLOGICA 37/1 2008 56 fault were analysed by Udating and a hig h resolution 18 O prole. W it hin t h e 185 Ka, 38 seismites were sam pled. ese stem from 13-18 eart h quakes wit h a mean recurrence interval of 10-14 Ka. e study s h owed t h at t h e deformational events dated in t h e caves complement independent near-fault paleoseismic records by temporal correlation wit h t h e eart h quakes recorded t h erein (Ka gan et al. 2005). Most of t h e Dead Seas Scrolls, dating from about 200 B.C., were found buried under rubble. Amos Nur contens t h at eart h quakes, including a devastating one re corded in 31 B.C., may h ave caused t h e roofs of caves to collapse on top of t h e scrolls (h ttp://www.stanford.edu/ dept/news/pr/02/scrolls124.h tml). R EPORTS ABOUT EARTHQUAKES FELT IN KARST CAVES e newspaper Edinost (8 t h January 1926) publis h ed t h at during t h e eart h quakes in t h e area of Postojna between 1-7 t h January 1926 a big stalagmite (1 m in diameter) in Postojna Cave collapsed (Anon 1926). is was t h e peri od of Cerknica eart h quakes wit h t h e magnitude Mm=5.2 (Ribari 1982). umer (1996) described t h e eart h quake in Dimnice Cave (Slovenia) on May 22, 1995. ere were two eart h quakes, of 4 and 4.2 magnitude and an intensity of V to VI. e epicenter was in t h e vicinity of t h e town of Ilirska Bistrica 25 km away from t h e cave. e source was 18 km below t h e surface. e eart h quake did not damage t h e COUNTRY CAVE OR KARST AREA EARTHQUAKE OR SEISMIC EVENT REFERENCE Belgium Pere Noel (Han-sur-Lesse) Quinif & Genty 1998 Bulgaria Douhlata Angelova et al. 2005 Bulgaria Lepenitsa Shanov et al. 2001 Bulgaria Troana cave possible coseismic origin in deformations of speleothems Kostov 2002 Costa Rica Gilli 1997 China Furong Dong show cave during the Febr. 2003 earthquake M=3.5 a large number of speleothems were broken http://www.hongmeigui. net/~hmg/news.php?page=4 France La grotte de Villars, Perigord Quinif & Genty 1998 France Grotte de Deux Gourdes, eastern Pirenee neotectonics and recent movement of the fault Gilli 1986 France Aven de la Portalerie, Causse du Larzac detected 2 earthquakes in the period 36.800-4.500 let BP Bruxelles et al. 1998 Germany Gaislochhole probable coseismic origin of speleothem breaks Moser & Geyer 1979 Israel caves in the Dead sea region archaeological records show evidence of several earthquakes every 500 y., including quake of 31 B.C. http://www.geotimes.org/feb03/ feature_stories.html Jordan Khirbet Rufeis cave complex damaged by earthquakes in the 840s and 50s A.D. http://www.casa.arizona.edu/ MPP/rufeis_cave/rufeis.html Portugal Zambujal prooves for 2 seismic events Crispim 1999 Romania cave system Humpleu, Bihor Mt. detected 2 tectonic events in the last 250.000 years Onac et al. 1998 Switzerland area north from the lakeThoune Jeannin 1990 Switzerland Dieboldslochli Basel earthquake in 1356 with Mw= 6.9 Lemeille et al. 1999 Turkey Tilkiler Gilli 1997 USA Coronado cave, Kartchner caverns and Colossal Cave 1887 earthquake http://www.thoughtsandplaces. org/Caverns/Coronado3.html Table 1: Examples of seismic activities in karst environments. STANKA EBELA

PAGE 7

ACTA CARSOLOGICA 37/1 2008 57 cave, but t h e cavers noticed stronger wind ow, sound and splas hing water in t h e underground river. On t h e 12 t h July 2004 t h e eart h quake (Mw=5.2) wit h t h e epicenter 70 km away was felt in Postojna Cave (Glaar 2006). ey did not feel t h e eart h s h aking, but t h e noise as if t h e train was coming down from Velika Gora towards t h e Koncertna Dvorana. While taking t h e data from TM 71 extensometer in Postojna Cave on January 14, 2005 we h eard a noise as if t h e train was coming and later realized it was an eart h quake situated 45 km NW from Postojna (Cerkno, ML=3.8). In W ind Cave National Park (USA) on Marc h 24, 1964, a tremor caused small rocks to fall in t h e cave. is was not connected wit h t h e great Alaska eart h quake ( h ttp://neic.gov/neis/states/sout h /sout h _dakota/sout h dakota_history.h tml). During t h e very strong C hirpan eart h quake in Bul garia in 1928 (Ms=7.0) two men from t h e village were in t h e cave and witnessed t h e breakdowns. e S h erpan cave is about 55 km to t h e SW of t h e epicenter of t h e eart h quake (Kostov 2002). During t h e New Madrid eart h quakes (1811-1812) rocks cracked in Mammot h cave and large rocks fell in some parts of t h e cave. On June 10, 1987, an eart h quake centre in Laurnenceville, Illinois, wit h a magnitude of 5.0, caused rocks to fall in Audubon Avenue (h ttp://www. nps.gov/arc hive/maca/rockfall.pdf). Cavers report an eart h quake (May 1995) wit h epi center in t h e vicinity of Jenolan Caves. In W iburds Lake Cave t h ey noticed a w hite rock 1x0.5 m t h at must h ave fallen from t h e roof, probably during t h e eart h quake ( h ttp://www.ee.usyd.edu.au/suss/Bulls/35(2)eart h quake. h tml). In Table 1 some additional karst caves and karst ar eas are collected wit h described eart h quake or seismic event. TECTONICALL Y BROKEN FLO W STONE IN POSTOJNA CAVE Experiences of almost 20 years of detailed tectonic-lit h o logical mapping of karst caves in Slovenia (ebela & ar 1991; ebela 1998) disclosed many examples of t h e inu ence of geological structural elements in t h e formation of karst caves. In most of t h e cases we h ave older tectonic deformations, but in some t h e still active tectonic struc tures in cave passages can be detected. In t h e nort h ern part of Pisani Rov in Postojna Cave t h e cross-Dinaric oriented (NE-SW ) fault cuts t hroug h t h e ceiling (Figure 4). Below t h e NW end of t h e fault fres h cracks are visible on t h e owstone. Even if t h e nort h ern end of Pisani Rov is situated just some metres away from t h e bottom of Velika Jeranova doline (ebela & ar 2000) we do not believe t h at t h e fault is connected wit h relaxation processes around t h e doline edge. In suc h case more joints wit h fres h cracks would be present, but in our case we h ave only one fault t h at looks to be still tectonically active. One of t h e best-expressed fault zones in Postojna Cave runs t h roug h t h e collapse c h amber of Pisani Rov, continuous on t h e nort h ern edge of t h e biggest collapse c h amber Velika Gora and narrows in t h e small side pas sage of Lepe Jame (Figure 5). Along t h e Dinaric oriented (N W -SE) fault zone we h ave t h e traces of sinistral and dex tral h orizontal movements, as well as reverse movement in Velika Gora. Regarding results obtained by extensom eter TM 71 it looks t h at t h e fault zone is still tectonically active wit h an average 0,02 mm/year dextral h orizontal movement (Gosar et al. 2007). W e h ave t h e example of slow movement along non-seismogenic fault zone. Along t h e same fault zone we can nd some broken owstone t h at mig h t be t h e result of tectonic deforma tions along t h e fault (ebela 2005) and not t h e result of instability of t h e collapse blocks. e ssure cuts a small owstone column and it looks as if t h e lower part of t h e column h as slid, w hic h means t h e opening of t h e fault. Following t h e same fault zone t hroug h t h e cave pas sages for a distance about 800 m one can realize t h e posi tion of collapse c h ambers t h at are developed inside t h e fault zone. e traces of older tectonic movements are dierent t hroug h out t h e fault zone. Todays tectonic ac tivity produces micro-deformations, w hic h can be seen as fres h cracks on owstone. F ig. 4: Cross-Dinaric oriented fault in P isani Rov of P ostojna Cave (photo J. Hajna). BROKEN SPELEOTHEMS AS INDICATORS OF TECTONIC MOVEMENTS

PAGE 8

ACTA CARSOLOGICA 37/1 2008 58 F ig. 5: Collapse chambers in P ostojna Cave are aligned along a Dinaric oriented fault zone. 1-underground river, 2-dextral horizontal slow-movement tectonic activity monitored with T M 71 extensometers from the year 2004, 3-P ostojna anticline, 4-horizontal move ments, 5-vertical movements, 6-strike and dip direction of geological elements. STANKA EBELA

PAGE 9

ACTA CARSOLOGICA 37/1 2008 59 e traces of neotectonics and active tectonics in Posto jna Cave are well documented (Gospodari 1964; Sas owsky et al. 2003; ebela 2004). e only description of owstone column collapse, during t h e eart h quake in Pos tojna Cave, dates back to t h e 1926 Cerknica eart h quake (Anon 1926). Postojna does not represent a seismically hig h-risk area but slow tectonic micro-deformations measured wit h TM 71 extensometers can in 10.000 years of t h e cave development represent an important factor for speleogenesis. e average 0,02 mm/year of dextral h orizontal movement (Gosar et al. 2007) in 10.000 years, wit h constant tectonic deformation, represents 2 m and in 100.000 years 20 m of t h e movement. ese are not negligible data in t h e speleogenetic time period. Eart h quakes can be one possible cause for t h e col lapse of speleot h ems, but mostly it is dicult to prove t h at owstone deformation is due to t h e eart h quakes. First, we must eliminate all ot h er possibilities. W e h ave to determine t h e age of deformation of t h e owstone, and statistical analyses of data from a variety of caves (Forti 1997). Some researc h ers (Becker et al. 2005; Lacave et al. 2003; Gilli 1999 d) t hink t h at to use broken speleot h ems preferentially stalagmites to indicate past strong eart h quake s h ocks is becoming increasingly questionable and is now generally accepted only for t h e most fragile spe leot h ems, i.e. soda straws. For cave systems in strong t hickly bedded slig h tly jointed limestones at distance of per h aps 150 m below t h e ground surface, damage of walls and roofs are un likely to take place except in cases of a very large s h allow eart h quake not far away. During seismic motion most speleot h ems would not experience dynamic amplica tion, but would move wit h t h eir base as a rigid structure (Becker et al. 2006). Flowstone being broken due to eart h quakes is di cult to verify, because more reasons could be interacting. But from a speleogenetical point of view t h e eect of slow non-seismogenic movements along fault planes is im portant. is is t h e subject of t h e study in Postojna Cave w h ere t h e Dinaric-oriented fault is being monitored wit h TM 71 extensometers from t h e year 2004 (Gosar et al. 2007). I believe t h at t h e study of past and recent tectonic deformations in karst caves is and will remain one of t h e important views in speleogenesis. It surely demands ot h er parallel studies t h at can eliminate or prove ot h er causes for speleot h em damage. It is important to detect and date old speleot h em damage due to paleotectonic deformations and it is also necessary to nd karst caves close to t h e recent strongly tectonic active areas. I t hink t h at detailed studies s h ould be oriented into t h e karst ar eas wit h active tectonics as Indonesia, C hina, Andes, etc., and s h ould be accompanied by ot h er met h ods (GPS, Ra don and Hg emanations, 3D extensometer monitoring, precise dating of dierent tectonic and ot h er events, etc.). In t his sense we can compare old speleot h em damage due to paleotectonics wit h recent speleot h em breaks due to t h e eart h quakes or slow tectonic non-seismogenic move ments. In a natural cave in NE Italy a station for t h e joint monitoring of h orizontal deformation, tilt variations and Radon emanation from t h e soil operates since 1994 (Ga ravaglia et al. 1998). Long-term monitoring wit h various observations is t h e best answer for understanding t h e complex karst environment. ACKNO W LEDGEMENTS CONCLUSIONS is study was done wit h t h e nancial support of t h e Slovenian Researc h Agency (programme P6-0119 and project J6-7022), project COST 625 and Slovenia-Czec h Republic cooperation in science and tec hnology (projects BI-CZ/06-07-011 and BI-CZ/08-09-015). I am grateful to Dr. Trevor S h aw and Dr. Ira D. Sasowsky for editing t h e Englis h text and to dr. P. Forti for constructive review. BROKEN SPELEOTHEMS AS INDICATORS OF TECTONIC MOVEMENTS

PAGE 10

ACTA CARSOLOGICA 37/1 2008 60 REFERENCES Angelova, D., Belfoul, M.A., Bouzid, S., Fila hi, M. & Faik, F., 2003: Paleoseismic p h enomena in karst terrains in Bulgaria and Marocco.Acta carsologica 31/1, 101-120, Ljubljana. Angelova, D., Belfoul, M.A., Bouzid, S. & Faik, F., 2005: Karst and cave systems in Bosnek region (Vitos h a mountain, Bulgaria) and in W in-Timdouine (Hig h Atlas mountain, Marocco).Acta Carsologica 34/1, 87-111, Ljubljana. Anonym, 1926: Iz trake pokrajine, Postojna, sedem dni potresa.Edinost, dne 8.januar 1926, Trst. Becker, H.K., 1929: H h le und Erdbeben: Mitt. ber H h len-u.Karstf. no. 1-4, p. 130-133. Becker, A., Ferry, M., Monecke, K., Sc hnellmann, M. & Giardini, D., 2005: Multiarc hive paleoseismic record of late Pleistocene and Holocene strong eart h quakes in Switzerland.Tectonop h ysics 400, 153-177. Becker, A., Davenport, C., Eic h enberger, U., Gilli, E., Jeannin, P-Y & Lacave, C. 2006: Speleoseismology: A critical perspective.Journal of Seismology, vol. 10, no. 3, 371-388. Bini, A., Quinif, Y ., Sules, O. & Uggeri, A., 1992: Les mouvements tectoniques rcents dans les grottes du Mont Campo dei Fiori (Lombardie, Italie).Karsto logia 19, 1, 23-30. Bosak, P., Hercman, H., Mi h evc, A. & Pruner, P., 2002: Hig h-resoultion magnetostratigrapfy of speleo t h ems from Snena jama, Kamnik-Savinja Alps, Slovenia.Acta carsologica 31/3, 15-32, Ljubljana. Brodar, S. 1966: Pleistocenski sedimenti in paleolitska najdia v Postojnski jami.Acta carsologica 4, 55139, Ljubljana. Bruxelles, L., Guendon, J.L. & Quinif, Y ., 1998: Indices de la palosismicit des grands Causses dans les cavi ties karstiques: lexemple de laven de la portalerie (larzac, Aveyron).Speleoc hronos h ors-srie, Livre des contributions au colloque Karst & Tectonics, Han-sur-Lesse, 9-12 mars 1998, 19-22, Mons. Camus, H., Seranne, M. & Quinif, Y ., 2001: Activit tec toniques rcente enregistre par les splot h mes: un contre exemple sur la faille des Cvennes (sud du Massif Central, Hrault, France).RIVIERA 2000, Tectonique active et gomorp h ologie, Villefranc h esur-Mer, Revue dAnalyse Spatial-No. Spcial 2001, 53-54, Nice. Crispim, J. A., 1999: Seismotectonic versus man madeinduced morp h ological c h anges in a cave on t h e Ar rbida c h ain (Portugal).-Geodinamica Acta, vol 12, 3-4, Elsevier, 135-142, Paris. Cucc hi, F. & Forti, P., 1989: e rst absolute datation of a speleot h en from Trieste Karst.Acta carsologica 18, 53-64, Ljubljana. ar, J., uteri, F. & Vrabec, M., 2002: Geomorfologija in aktivna tektonika ob severnem robu Cerknikega polja.-1.slovenski geoloki kongres, rna na Korokem, 9-11.oktober 2002, Knjiga povzetkov, 17, Ljubljana. Delaby, S., 2001: Palaeoseismic investigations in Belgian caves.Net h erlands Journal of Geosciences/Geolo gie en Mijnbouw 80 (3-4):323-332. Delaby, S. & Quinif, Y ., 2001: Palaeoseismic investigation in Belgian caves.-Riviera 2000, Tectonique active et gomorp h ologie, Villefranc h e-sur-Mer, Revue dAnalyse Spatial-No. Spcial 2001, 67-71, Nice. Forti, P., 1997: Speleot h ems and Eart h quakes.284-285, In: Hill, C. & Forti, P., Cave Minerals of t h e W orld. Second edition, 463 p., NSS, Huntsville. Gams, I., 2002: O kruenju, premikanju skalni h blokov in tektonski h dislokacija h v jama h v Sloveniji.Nae jame 44, 14-24, Ljubljana. Gams, I., 2003: Kras v Sloveniji v prostoru in asu.Zaloba ZRC, 516 pp., Ljubljana. Garavaglia, M., Braitenberg, C. & Zadro, M., 1998: Radon Monitoring in a Cave of Nort h-Eastern Italy.P h ys. C h em. Eart h, vol. 23, no. 9-10, 949-952. Gilli, E., 1986: Notectonique dans les massifs karstiques. Un exemple dans les Pralpes de Nice: la Grotte des Deux Gordes.Karstologia 8, 51-52, Bull F.F.S., Lyon. Gilli, E. 1992: Au cur dun c h evauc h ement dans les goures du Calernaum et des Baoudillouns.Kar stologia 19, 39-48. Gilli, E., 1997: Enregistrement de mouvements recent par lendokarst.Arc h ologie et Sismicit, dition AP DACA, 133-156. Gilli, E., 1999 a: Researc h on t h e February 18, 1996 eart h quake in t h e caves of Saint-Paul-de-Fenouillet area, (eastern Pyrenees, France).Geodinamica Acta 12, 3-4, 143-158, Paris. Gilli, E., 1999 b: Evidence of palaeoseismicity in t h e caves of Arizona and New Mexico (USA).Surface Geo sciences, C.R. Acad. Sci. Sciences de la terre et des planets, 329, 31-37, Paris. Gilli, E., 1999 c: Breaking of speleot h ems by creeping of a karstic lling. e example of t h e Ribire cave (Bouc h es-du-R h ne).Surface Geosciences, C.R. Acad. Sci. Sciences de la terre et des planets, 329, 807-813, Paris. STANKA EBELA

PAGE 11

ACTA CARSOLOGICA 37/1 2008 61 Gilli, E., 1999 d: Evidence of palaeoseismicity in a ow stone of t h e Observatoire cave (Monaco).Geodin amica Acta 12, 3-4, 159-168, Paris. Gilli, E., 1999 e: Recent, slow and aseismic movement of an overt hrust observed in t h e Abel sink h ole (St Vallier de iey, 06, France).Geodinamica Acta 12, 3-4, 169-177, Paris. Gilli, E., 2004: Glacial causes of damage and diculties to use speleot h ems as palaeoseismic indicators.Geo dinamica Acta, 17(3), 229-240. Gilli, E. & Delange, P. 2001: Utilisation des splot h mes comme indicateurs de notectonique ou de palo sismicit.RIVIERA 2000, Tectonique active et gomorp h ologie, Villefranc h e-sur-Mer, Revue dAnalyse Spatial-No. Spcial 2001, 79-90, Nice. Glaar, S., 2006: Strela in potres v Postojnski jami.Nae jame 46, 118-119, Ljubljana. Gosar, A., ebela, S., Kotk, B. & Stemberk, J. 2007: Mi cro-deformation monitoring of active tectonic st cructures in W Slovenia.Acta Geodyn. Geomater., vol. 4, no. 1, 87-98. Gospodari, R., 1964: Sledovi tektonski h premikov iz ledene dobe v Postojnski jami.Nae jame 5 (1963), 5-11, Ljubljana. Gospodari, R., 1968: Podrti kapniki v Postojnski jami.Nae jame 9 (1-2), 15-31, Ljubljana. Gospodari, R., 1977: e collapse of speleot h ems in t h e Postojna cave system.Proceedings of t h e 7t h inter national speleological congress S h eeld 1977, Haw t h ornes of Notting h am Limited, 444 p., England. Habi, P., 1971: Poloka jama-najgloblja v Jugoslaviji.Nae jame 12 (1970), 23-34, Ljubljana. Hill, C.A. & Forti, P., 1997: Cave Minerals of t h e W orld.National Speleological Society, 463 pp., Huntsville. Ho h enwart, F.J.G., 1830: W egweiser fr die W anderer in der ber hmten Adelsbergerund Kronprinz Ferdinands-Grotte bey Adelsberg in Krain, h er ausgegeben von Franz Grafen von Ho h enwart. Als Erklrung der von Hern Aloys Sc h aenrat h, k.k. Kreis-Ingenieur in Adelsberg gezeic hneten Ansic h ten dieser Grotte.I. Helf, 1-16, W ien. h ttp://neic.gov/neis/states/sout h /sout h _dakota/sout h dakota_history.h tml h ttp://www.casa.arizona.edu/MPP/rufeis_cave/rufeis. h tml h ttp://www.ee.usyd.edu.au/suss/Bulls/35(2)eart h quake. h tml h ttp://www.geotimes.org/feb03/feature_stories.h tml h ttp://www.h ongmeigui.net/~hmg/news.p h p?page=4 h ttp://www.nps.gov/arc hive/maca/rockfall.pdf h ttp://www.stanford.edu/dept/news/pr/02/scrolls124. h tml h ttp://www.t h oug h tsandplaces.org/Caverns/Coronado3. h tml Jeannin, P.-Y .,1990: Neotectonique dans le karst du Nord du Lac de oune (Suisse).Karstologia 15, 41-54, Lyon. Kagan, E.J., Agnon, A., Bar-Matt h ews, M. & Ayalon, A., 2005: Dating large infrequent eart h quakes by dam aged cave deposits.Geology, v. 33, no. 4, 261-264. Kempe, S., 2004: Natural speleot h em damage in Posto jnska jama (Slovenia), caused by glacial cave ice? A rst assessment.Acta carsologica 33/1, 265-289, Ljubljana. Kostov, K., 2002: Speleot h ems as paleoseismic indica tors: example from Bulgaria.Environmental Ca tastrop h es and Recoveries in t h e Holocene, August 29September 2, 2002, Department of Geograp h y & Eart h Sciences, Brunel University, UK, selected abstracts. Kranjc, A., 1999: O odpadanju sige (primer odpadlega stalaktita v kocjanski h jama h).-Acta Carsologica 28/1, 201-214, Ljubljana. Lacave, C., Koller, M. & Levret A., 2001: Measurement of natural frequencies and damping of speleot h ems.RIVIERA 2000, Tectonique active et gomorp h olo gie, Villefranc h e-sur-Mer, Revue dAnalyse SpatialNo. Spcial 2001, 99-104, Nice. Lacave, C., Egozcue, J.J. & Koller, M.G. 2003: Can bro ken-and unbroken-speleot h ems tell us somet hing about seismic history?12t h European Conference on Eart h quake Engineering, paper 349, Elsevier. Lemeille F., Cus hing, M., Carbon, D., Grellet, B., Bitterli, T., Fle h oc, C. & Innocent, C., 1999: Co-seismic rup tures and deformations recorded by speleot h ems in t h e epicentral zone of t h e Basel eart h quake.Geo dinamica Acta 12, 3-4, Elsevier, 179-191, Paris. Quinif, Y ., Genty D. & Maire, R, 1994: Les splot h mes: un performant pour les tudes paloclimatiques.Bull. Soc. Gol. France, t. 165, no. 6, 603-612, Paris. Quinif, Y ., & Genty, D., 1998: Sedimentary recording and dating of sismo-tectonic events by t h e speleot h ems.Sploc hronos no 9, 27-32, Mons, Belgique. Maurin, V., 1953: ber jngste Bewegungen im Grazer Palozoikum.Ver h. D. Geol. BD., H. 4, W ien. Menic h etti, M., 1998: Central Italy eart h quakes of au tumn 1997 and t h e underground karst features of t h e area.Speleoc hronos h ors-srie, Livre des con tributions au colloque Karst & Tectonics, Han-surLesse, 9-12 mars 1998, 121 p., Mons. Mi h evc, A., 2001: Speleogeneza Divakega krasa.Zaloba ZRC 27, 180 pp., Ljubljana. BROKEN SPELEOTHEMS AS INDICATORS OF TECTONIC MOVEMENTS

PAGE 12

ACTA CARSOLOGICA 37/1 2008 62 Mocc hiutti, A. & Valent, M., 1998: Evidences mor p h ologiques de mouvements tectoniques rcents dans les grottes du Friuli (Nord Est Italie), avec r frence particulire aux mouvements postrieurs au tremblement de terre de 1976.Speleoc hronos h orssrie, Livre des contributions au colloque Karst & Tectonics, Han-sur-Lesse, 9-12 mars 1998, 123-125, Mons. Mocc hiutti, A. & DAndrea, A., 2002: Evidenze morfo logic h e di movimenti tettonici recenti ed attuali, stazioni di monitoraggio in alcune grotte delle Pre alpi friulane (Nord-Est Italia).Mem. Soc. Geol. It., 57, 487-491. Moser, M. & Geyer, M., 1979: Seismospelologie-Erd bebenzerstrungen in H h len am Beispiel des Gailoc h es bei Oberfellendorf (Oberfranken, Bay ern).Die H h le, v. 30-4, 89-102. Onac, B.P., Papiu, F., Aanei, M., Bujor, O., Irimie, D. & Fi lotie, O., 1998: Neotectonic features in t h e Humpleu cave system (Bi h or Mountains, Romania).Spele oc hronos h ors-srie, Livre des contributions au col loque Karst & Tectonics, Han-sur-Lesse, 9-12 mars 1998, 135-136, Mons. Perko, I.A., 1910: Die Adelsberger Grotte in W ort und Bild.p. 78, Adelsberg. Plan, L., Sptl, C h., Grasemann, B., Decker, K., Oen bec h er, K.H. & W iesmayr, G., 2005: Seismot h ems caused by neotectonic activity in t h e Eastern Alps.14t h International Congress of Speleology, 21-28 August 2005, Final Programme & Abstract Book, 117-118. Postpisc h l, D., Agostini, S., Forti, P. & Quinif, Y ., 1991: Palaeoseismicity from karst sediments: t h e Grotta del Cervo cave case study (central Italy).In: Stuc c hi, M., Postpisc h l, D. & Slejko, D., (Editors), Inves tigation of Historical Eart h quakes in Europe. Tecto nop h ysics, 193 (1991), 33-44, Amsterdam. Ribari, V., 1982: Seismicity of Slovenia, Katalog potresov (792-1981).649 p., Seizmoloki zavod SR Slovenije, Ljubljana. Sasowsky, I.D., ebela, S. & Harbert, W ., 2003: Concurent tectonism and aquifer evolution >100,000 years re corded in cave sediments, Dinaric karst, Slovenia.Environmental Geology 44, 8-13. S h anov, S., Kourtev K., Kostov, K., Nikolov, G., Boykova, A. & Benderev, A., 2001: Palaeoseismological traces in t h e Lepenitsa Cave, Velingrad district, Sout h Bul garia.Tectonique active et gomorp h ologie, Riviera 2000, Villefranc h e-sur-Mer 18-22 oct. 2000, Revue dAnalyse Spatial N spcial 2001, 151-154, Nice. ebela, S., 1996: e inuence of tectonic zones on cross section formations in t h e Predjama cave, Slovenia.Kras i speleologia, 8 (XVII), 72-76, Uniwersytet Slaski, Katowice. ebela, S. 1998: Tectonic structure of Postojnska jama cave system.Zbirka ZRC 18, 112 p., Ljubljana. ebela, S., 2004: Potresi v kraki h jama h.Raziskave s podroja geodezije in geozike, 9.strokovno sreanje Slovenskega zdruenja za geodezijo in geoziko, Ljubljana, 15.januar 2004, 5-14, Ljubljana. ebela, S., 2005: Monitoring of active tectonic structuresProject COST 625.Acta carsologica 34/2, 471-488, Ljubljana. ebela, S. & ar, J., 1991: Geological setting of collapsed c h ambers in Vzh odni rov in Predjama cave.Acta carsologica 20, 205-222, Ljubljana. ebela, S. & ar, J. 2000: Velika Jeranova doline a for mer collapse doline.Acta carsologica 29/2, 201212, Ljubljana. Urbanc, J., 1982: Kamnika jama.Nae jame 23/24 (1981/1982), 25-34, Ljubljana. Vandycke, S. & Quinif, Y ., 2001: Recent active faults in Belgian Ardenne revealed in Roc h efort Karstic net work (Namur Province, Belgium).Net h erlands Journal of Geosciences 80 (3-4), 297-304. W ojcik Z. & Zwolinski, S., 1959: Mlode przesuniecia tek toniczne w jaskiniac h tatrzanskic h.Acta Geol. Pol., vol. IX, W arszawa. umer, J. 1996: Potres v Dimnica h.Nae jame 38 (1997), 152-154. Ljubljana. STANKA EBELA



PAGE 1

CAVE SEDIMENTS FROM THE POSTOJNSKAPLANINSKA CAVE S Y STEM SLOVENIA: EVIDENCE OF MULTIPHASE EVOLUTION IN EPIPHREATIC ZONE JAMSKI SEDIMENTI IZ POSTOJNSKOPLANINSKEGA JAMSKEGA SISTEMA SLOVENIJA: PRI A VE FAZNEGA RAZVOJA V EPIFREATI NI CONI Nadja ZUPAN HAJNA 1 Petr PRUNER 2 Andrej MIHEVC 1 Petr SCHNABL 2 & Pavel BOSK 2,1 1 Karst Researc h Institute, Scientic Researc h Centre, Slovenian Academy of Sciences and Arts, Titov trg 2, 6230 Postojna, Slovenia, e-mail: zupan@zrc-sazu.si, mi h evc@zrc-sazu.si 2 Institute of Geology of t h e Academy of Sciences of t h e Czec h Republic, v.v.i., Rozvojov 269, 165 00 Pra h a 6, Czec h Republic, e-mail: bosak@gli.cas.cz, pruner@gli.cas.cz, sc hnabl@gli.cas.cz Received/Prejeto: 20.04.2008 COBISS: 1.01 ACTA CARSOLOGICA 37/1, 63-86, POSTOJNA 2008 Abstract UDC 551.352.551.44(497.4) Nadja Zupan Hajna & Petr Pruner & Andrej Mihevc & Petr Schnabl & Pavel Bosk: Cave sediments from PostojnskaPla ninska cave system (Slovenia): Evidence of multi-phase evolu tion in epiphreatic zone e Postojnska jamaPlaninska jama cave system and number of smaller adjacent caves are developed in t h e Postojnski kras. ese caves are located between two dextral strike-slip fault zones oriented in t h e Dinaric direction. e caves contain lit h ologically diversied cave ll, ranging from speleot h ems to allo genic uvial sediments. e allogenic clastic material is derived from a single source, Eocene siliciclastics of t h e Pivka Basin. Small dierences in mineral/petrologic composition between t h e sediments can be attributed to dierent degrees of weat h er ing in t h e catc hment area and h omogenization of source sedi ments. ick sequences of ne-grained laminated sediments, deposited from suspension are common. e depositional en vironment was mostly calm, but not completely stagnant. Suc h a sedimentary environment can be described as cave lacustrine, wit h deposition from pulsed ow. e h omogeneity of t h e pa laeomagnetic data suggests rapid deposition by a number of s h ort-lived single-ood events over a few t h ousand years. is depositional style was favourable for recording of s h ort-lived excursions in t h e palaeomagnetic eld. e sediments were originally not expected to be older t h an Middle Quaternary in age (i.e. about 0.4 Ma). Later numerical dating (/U and ESR) indicated ages older t h an 0.53 ka. New palaeomagnetic data from selected sedimentary proles wit hin t h e cave system detected normal polarization in muc h of t h e proles studied. Reverse polarized magnetozones, interpreted mostly as s h ortlived excursions of magnetic eld, were detected in only a few places. erefore, we interpreted most of t h e sediments as be ing younger t h an 0.78 Ma, belonging to dierent depositional p h ases wit hin t h e Brun h es c hron. Palaeomagnetic properties Izvleek UDK 551.352.551.44(497.4) Nadja Zupan Hajna & Petr Pruner & Andrej Mihevc & Petr Schnabl & Pavel Bosk: Jamski sedimenti iz postojnskopla ninskega jamskega sistema (Slovenija): pria vefaznega raz voja v epifreatini coni Postojnsko-planinski jamski splet in tevilne manje jame med jamama so razvite v Postojnskem krasu, ki lei med dvema desnozminima prelomnima conama v Dinarski smeri. Jame veinoma vsebujejo obsene in litoloko raznovrstne jamske zapolnitve, od sige do aligeni h uvialni h jamski h sedimentov. Alogeni klastini sedimenti izvirajo iz enega vira (eocenski siliciklasti iz Pivke kotline). Majhne razlike v mineralni/ petroloki sestavi med sedimenti, la h ko pripisujemo razlini stopnji preperevanja v povodju in h omogenizaciji izvornega sedimenta. Najpogosteji so drobnozrnati sedimenti, pogosto v debeli h sekvenca h. Sedimentirani so bili iz suspenzije. Sedi mentacijsko okolje je bilo veinoma mirno vendar ne popol noma stojee. Taka sedimentacijska okolja la h ko opiemo kot jamsko jezersko ali kot usedanje iz pulzni h tokov. Homogenost palaeomagnetni h podatkov la h ko nakazuje hitro sedimentacijo (nekaj tiso let) med kratkotrajnimi poplavnimi dogodki. Tak nain sedimentacije je dobro odrazil kratkotrajajne eksurzije magnetnega polja. Zapolnitve jam naj bi bile po prvotni h priakovanji h stareje od srednega kvartarja (okrog 400 tiso let). Kasneja numerina datiranja (/U in ESR) nakazujejo starosti na ve kot 530 tiso let. Novi palaeomagnetni rezul tati, pridobljeni iz izbrani h sedimentni h prolov v jamskem sistemu, so odkrili normalno pol a rizacijo v veini preiskoavni h prolov. Reverzno usm erjene magnetocone, interpretirane veinoma kot kratkotrajne ekskurzije magnetnega polja, so bile zaznane samo na nekateri h mesti h. Zato smo veino tudirani h sedimentov interpretirali za mlaje od 780 tiso let. Pripadajo razlinim sedimentacijski fazam v Brunes razdobju. Palaeomagnetne lastnosti v dve h proli h, v jama h presekani h

PAGE 2

ACTA CARSOLOGICA 37/1 2008 64 of two proles in caves intersected by t h e articial tunnel be tween Postojnska jama and rna jama h ad reverse polarized magnetozones and of sediments in Zguba jama, may indicate an age muc h greater t h an 0.78 Ma. e cave system h as evolved over a long period of time, governed by t h e functioning of Pla ninsko polje in t h e relation to t h e evolution of t h e resurgence area in Ljubljana Moor furt h er to t h e east. General stabilization of t h e h ydrological system wit h low h ydraulic h ead led to t h e evolution of caves in epip hreatic and paragenetic conditions over a long time-span. Individual cave segments or passages were completely lled and exh umed several times during t h e evolution of t h e cave. Alternation of depositional and erosional p h ases may be connected wit h c h anging conditions wit hin t h e cave system, t h e functioning of t h e resurgence area, collapse, climatic c h ange, tectonic movement and t h e intrinsic mec h a nisms of contact karst. Key words: dating, palaeomagnetism, magnetostratigrap h y, karst geomorp h ology, Classical Karst. z umetnim tunelom med Postojnsko in rno jamo, z reverzno usmerjeno magnetocono, in sedimenti v Zgubi jami, so la h ko precej stareji od 780 tiso let. Jamski sistem kae na dolgotra jen razvoj pogojen z delovanjem Planinskega polja in odnosom do razvoja izvirov na Ljublanskem barju proti vzh odu. Splona stabilizacija hidrolokega sistema z majhnim gradientom je v zelo dolgem asovnem obdobju vodila do razvoja jam v epifreatini h in paragenetski h pogoji h. Posamezni deli jam ter nji h ovi h rovov, so bili za asa nji h ovega razvoja la h ko vekrat popolnoma zapolnjeni ali izpraznjeni. Izmenjava sedimentaci jski h in erozijski h faz je bila la h ko povezana z menjavo pogojev v jamskem sistemu, delovanjem izvirov, s podori, klimatskimi spremembami, tektonskimi premiki in me h anizmi, povezanimi s kontaktnim krasom. Kljune besede: datacije, palaeomagnetizem, magnetostrati graja, kraka geomorfologija, klasini kras. INTRODUCTION e Postojnska jama (Postojna Cave) Planinska jama (Planina Cave) cave system is t h e largest cave system in t h e Notranjski kras (Karst of Notranjska; Fig. 1), w h ere several ot h er large cave systems also occur. W aters from t h e Pivka kotlina (Pivka Basin) are drained to Planinsko polje (Planina Polje) t hroug h t h e Postojnski kras (Karst of Postojna). e Postojnska jama Planinska jama cave system is t h e main drainage route, wit h a lengt h of about 27 km. Only 2,200 m of unexplored water-lled passages separate t h e two caves. e cave system is developed in t h e Upper Ceno manian and Turonian to Senonian limestone sequence, w hic h is about 800 m t hick (Buser, Grad & Pleniar 1967). e limestone sequence is deformed into a complicated t hrust structure (nappes). ese formed aer t h e termi nation of Eocene ysc h deposition, as a consequence of post-collision processes between African and European Plates. e area is situated between two dextral strike-slip fault zones, t h e Idrija and Predjama faults (Placer 1996), oriented in t h e Dinaric direction (NW SE). ese faults resulted from reactivation in t h e latest tectonic p h ase by counter clockwise rotation of t h e Adria Microplate, w hic h started before ca 6 Ma (Vrabec & Fodor 2006). Some tectonic blocks along t h e Idrija fault were drowned and levelled by corrosion at t h e karst water table into karst poljes. is fault zone is well expressed in t h e karst relief. ar & Gospodari (1984) distinguis h ed dierent generations of fault zones accompanied by crus h ed, brec ciated and ssured zones. e cave system developed in t h e Postojnski kras geomorp h ological unit. e geomorp h ology of t h e area was described by Gams (1965), Gospodari (1979), Habi (1982), and Mulec et al. (2005). e Postojnski kras is situated between t h e two hig h karst plateaus, Hruica and Javorniki, wit h surfaces above 800 m a.s.l. and by two lower relief units, t h e Pivka kotlina (511 m a.s.l.) on t h e west and Planinsko polje (446 m a.s.l.) on t h e nort h east. Karst depressions h ave directly inuenced t h e evo lution of bot h t h e surface and t h e caves. e Postojnski kras surface, wit h dolines as t h e dominant relief forms, h as an elevation of about 600 m wit h only some parts above 700 m a.s.l. (Fig. 9). Karst relief s h ows strong tec tonic control, wit h structure guiding t h e general position of collapse dolines bot h on t h e surface and in t h e caves (Habi 1982). e lower part of Pivka kotlina (511 m a.s.l.) is developed on Eocene ysc h rocks, w hile t h e upper part is partly developed on limestone. Basinal waters con centrate on t h e ysc h and sink in system of blind valleys at its boundary wit h t h e limestone. Rivers in t h e basin eroded t h e rocky substratum and transported sediments into t h e caves. is resulted in t h e erosion of more t h an 100 m of ysc h during basin evolution. e morp h ol ogy of t h e contact karst at t h e ponors s h ows t h at ponor altitudes and h ydraulic h ead wit hin t h e Postojnski karst controlled t h e evolution of Pivka kotlina (Mi h evc 1990). e recent evolution of Pivka kotlina is also inuenced by t h e relative subsidence of t h e Vipava valley on t h e west, but its inuence on erosion of Pivka kotlina h as been negligible. N ADJA ZUPAN HAJNA, P ETR PRUNER, A NDREJ MIHEVC, P ETR SCHNABL & P AVEL BOSK

PAGE 3

ACTA CARSOLOGICA 37/1 2008 65 F ig. 1: Notranjski kras. Planinsko polje, formed along t h e Idria fault zone, is about 6 km long and 2 km wide. Several springs of t h e Unica River are situated on t h e sout h ern edge of t h e polje. e river crosses t h e polje and sinks on its nort h ern side. e levelled polje oor resulted from regular ooding ac companied by lateral corrosion. e caves in t h e Postojnski kras are rat h er s h allow because t h eir dept h is limited by t h e elevations of Pivka kotlina and Planinsko polje. During t h e evolution of t h e caves Pivka kotlina was levelled and adjusted to t h e alti tude of t h e ponors, w h ere large blind valleys developed. Postojnska jama h as presumably served as t h e main out ow from t h e Pivka kotlina for a very long time but only about 20 m of entrenc hment h as occurred. Distinct trac es of paragenetic development can be found in numerous parts of t h e accessible cave. Clastic sediments h ave an important role in de cip h ering geological history and processes in t h e caves (Sasowsky 2007). Cave ll, in respect to its composition, position and fossil remains, can oer number of useful information on evolution of surface geology, morp h olo gy, and cave itself, c h aracter and environment of deposi tion or precipitation, age, palaeoclimatic and palaeogeo grap hic conditions during t h eir sedimentation. PREVIOUS W ORK e large caves h ave attracted researc h ers for many years (see S h aw 1992). Some focused on t h e study of sedi ments and fossils. Cave sediments in t h is large system were studied at rst by arc h aeologists and palaeontolo CAVE SEDIMENTS FROM THE POSTOJNSKAPLANINSKA CAVE S Y STEM SLOVENIA: EVIDENCE OF MULTIPHASE ...

PAGE 4

ACTA CARSOLOGICA 37/1 2008 66 gists (Rakovec 1954; Brodar 1966, 1969). Older sedi ments were systematically studied by Gospodari (1972, 1976, 1977a, b, 1981, 1988), w h o suspected t h at t h e old est sediments from Postojnska jama and ot h er caves in t h e region (Otoka, Risovec, Krina, Planinska a.o.) were not older t h an Middle Quaternary (i.e. about 400 ka). He compiled a stratigrap h y of t h e cave sediments, from t h e oldest to t h e youngest, as follows: (1) gravel wit h col oured c h ert, (2) older laminated loam (Middle Quater nary), w h ic h is missing in Postojnska jama (Gospodari 1976, p. 100), (3) rubble and gravel wit h w h ite c h ert (Riss), (4) speleot h ems and red loam (Riss W rm), (5) younger laminated loam (lower W rm) and ysc h sand/ ood loam (upper W rm), (6) speleot h ems (post-Gla cial), (7) breakdown rocks, ood loam and sand, (8) spe leot h ems (Holocene). Alt h oug h t h e old laminated loam cannot be developed in Postojnska jama, some sediments from t h e Otoka jama and Male jame were misinterpret ed as belonging to t h is lit h ostratigrap h ic unit (ebela & Sasowsky 1999). He concluded t h at t h e oldest sediments from t h e cave system were no older t h an Middle Quater nary (i.e. about 400 ka) and t h at cave sediment and spe leot h em deposition was related to Quaternary climatic oscillations (e.g., glacial cycles; Gospodari 1988). Some of h is studies (Gospodari 1977a, b, 1981) indicate t h at h e started to doubt h is c h ronostratigrap h y, especially w h en muc h greater age for speleot h ems was obtained by numerical dating (Gospodari 1981). Unfortunately all h is attempts at palaeomagnetic dating produced by no results. A number of samples h as been analysed by 14 C, / U and ESR met h ods in Postojnska jama (Franke & Gey h 1971, 1976; Gospodari 1972, 1977b, 1981; Ikeya et al. 1983; Zupan 1991; Zupan Hajna 1996; Mi h evc 2002, 2004, 2008). Samples dated by t h e radiocarbon met h od range from about 7.5 to 42.9 ka, mostly clustering at about 7.5.2, around 20 and over 37 ka (Gospodari 1972, 1977b). ESR ages are 125, 280 and 530 ka (Gospodari 1981, p. 93) and 190 ka (Ikeya et al. 1983). Samples analysed by /U met h od are dated from 4 to more t h an 350 ka (see Mi h evc 2008). Two samples from Podorna dvorana (Collapsed Dome) in Pisani rov were dated by t h e /U met h od (Zupan 1991): two samples were older t h an 350 ka, and also dates of 269.4 +130.3/-80 and 76 +24.9/-21.9 ka were obtained from one sample. e base of a stalactite under a collapse block yielded an age of about 75 ka and t h e base of a stalagmite, w hic h grew on t h e block, was estab lis h ed to be 19.9 +25.2/-24.7 ka old. Radiocarbon data (Franke & Gey h, 1971; Gospodari 1977, 1981) indicated t h e same ages for young stalagmite growt h (13.5.6 ka). Some collapsed stalagmites are overgrown by new speleot h em 8.4.8 ka old, similarly to ot h er sites in t h e system. Mi h evc (2002, 2004, 2008) recorded two periods of owstone growt h at Velika gora (Great Mount) and arobni vrt (Enc h anting Garden). e oldest owstone was dated at t h e foot of t h e collapse at t h e railway sta tion, w h ere owstone was deposited (152 40 ka) above collapse boulders. A owstone dome at t h e top of Velika gora was dated to 70 26 ka. e stalagmites were grow ing on clays (41 3 ka, 43 10 ka), on rubble (47 7 ka) and on collapse boulders (37 7 ka). e youngest p h ase of owstone deposition is recorded in samples of grey crystalline owstone and stalagmite (12 5 ka and 6 4 ka) covering all collapse blocks. Franke & Gey h (1971) recorded stalagmite growt h between ca 12.1 and 14.4 ka from Kongresna dvorana (Congress Hall). Radiocarbon dated collapse of big sta lagmite in arobni vrt to about 10 ka. It is overgrown by stalagmites dated to about 7.5 ka (Franke & Gey h, 1971, 1976). Overgrowt h on fallen older stalagmites dated to about 7.4.5 ka was recorded also from ot h er sites (Gospodari 1977b). More recent researc h on sediments and speleot h ems (Zupan 1991 a.o.), and t h e evolution of karst morp h olo gy (Mi h evc 1996, 2001, 2002, 2004, 2007) suggested, t h at t h e sediments are older t h an previously t h oug h t. Zupan Hajna et al. (2008a) described in detail 36 sedimentary proles in caves, unroofed caves and surface sediments studied in Slovenia since 1997. Sedimentary proles in Postojnska, Zguba and Planinska jama were sampled and analyzed between 2003 and 2007 (Tab. 1; Figs. 2, 4 and 8). e names of caves and geograp hic features in t h e text are given in t h e original Slovene form: jama means cave dolina means doline, kotlina means basin kras means karst and spodmol means overhang Englis h tran scription of geograp hic names is used in accordance wit h t h e recommendations of t h e Slovene Commission on Geograp hic Names. e following abbreviations are used in t h e text: AF = alternating eld (demagnetization); CRM = c h emical remnant magnetism/magnetization; D = declination; HFC = hig h-eld component; I = inclination; IG AS CR = Institute of Geology of t h e Academy of Science of t h e Czec h Republic, v.v.i.; IZRK = Karst Researc h Institute; JZS = Speleological Association of Slovenia; LFC = loweld component; MAVACS = Magnetic Vacuum Control System; MS = magnetic susceptibility; N = normal polar ization/polarity; NRM = natural remnant magnetization; R = reverse polarization/polarity; RM = remnant magne tization; TD = t h ermal demagnetization; ZRC SAZU = Scientic Researc h Centre of t h e Slovenian Academy of Sciences and Arts. N ADJA ZUPAN HAJNA, P ETR PRUNER, A NDREJ MIHEVC, P ETR SCHNABL & P AVEL BOSK

PAGE 5

ACTA CARSOLOGICA 37/1 2008 67 Due to t h e limitations of dierent numerical dating met h ods (see Bosk 2002), t h e palaeomagnetic met h od was used on carefully selected proles of cave sediments. e mineralogy of t h e sediments was investigated using x-ray diraction; w hile speleot h em was dated using t h e /U disequilibrium met h od. Palaeomagnetic analyses were completed in t h e Laboratory of Palaeomagnetism, IG AS CR in Pra h a Prn h onice. All specimens were oriented in situ prior to removal. Unconsolidated sediments were sampled in boxes (non-magnetic plastic, size = 20 x 20 x 20 mm, ap proximate volume 6.7 cm 3 ; Natsu h ara Giken Co., Ltd., Ja pan). Samples from consolidated rocks and speleot h ems were collected from t h e prole in large pieces, w hic h were cut in t h e laboratory into cubes 20 x 20 x 20 mm. Sam ples were demagnetized by AF (all samples) and or/TD (consolidated samples). Palaeomagnetic procedures were selected to allow t h e separation of t h e respective compo nents of t h e RM and t h e determination of t h eir origin. Sc h onstedt GSD or LDA apparatus was employed for t h e AF demagnetization and MAVACS (Pt h oda et al. 1989) for t h e TD demagnetization. e RM was measured on JR or JRA spinner magnetometers (Jelnek 1966). AF demagnetization was carried out up to a eld of 100 mT in 12 steps. e NRM of specimens is identied by t h e symbol J n t h e corresponding RM moment by t h e symbol M. e MS values were measured on KL Y (or KL Y ) kappa-bridges and a KLFA Automatic Mag netic Susceptibility Meter (Jelnek 1966, 1973). e multi-component analysis tec hnique of Kirsc h vink (1980) was applied to separate t h e respective RM components. Fis h er statistics (1953) were employed for t h e calculation of mean directions of t h e CRM compo nents derived by t h e multi-component analysis. Results of palaeomagnetic analyses, including val ues and mean values of t h e MS, NMR, I, D, discussion of primary data, palaeomagnetic proles were summarized by Zupan Hajna et al. (2008a, b). Data in t his article rep resent summary comments, only. X-ray powder diraction analyses were performed (1) in t h e Laboratory of P h ysical Met h ods, IG AS CR in Pra h a (analysts Dr. Karel Melka, Dr. Roman Skla, Mr. Jit Dobrovoln), and (2) at t h e Geological Institute of Faculty of Natural Sciences and Engineering in Ljubljana (analyst Dr. Meta Dobnikar). In t h e Czec h laboratory all powder patterns were collected wit h a P hilips XPert APD diractometer (Cu and Co radiation, grap hite monoc hromator, 40 kV, 32 or 40 mA). For eac h specimen four individual sets of patterns were acquired for: ran domly-oriented material, oriented specimens, glycolated specimens, and samples h eated to 400 C under ambi ent atmosp h ere. e qualitative mineral composition of samples in Ljubljana was determined by X-ray powder diraction wit h a P hilips diractometer (anode CuK b 40 kV, 30 mA and Ni lter). e concentrations of minerals were determined from t h e h eig h t of t h e main reection of eac h particular mineral in t h e X-ray record. /U (U-Series) analyses were performed (1) in Geoc hronology Laboratory, Institute of Geological Sci ences, Polis h Academy of Sciences in W arsaw (analysts Prof. Dr. Helena Hercman, Dr. Tomasz Nowicki), and (2) in t h e Uranium Series Laboratory, Department of Geol ogy, University of Bergen (Head Prof. Dr. Stein Erik Lau ritzen; analysts Dr. Andrej Mi h evc). Standard c h emical procedures for uranium and t h orium separation from carbonate samples were used (Ivanovic h & Harmon 1992). In W arsaw activity was measured by b-spectrom etry, using an ORTEC OCTETE PC. Spectral analyses and age calculations were made wit h URANOTHOR 2.5 soware, w hic h is t h e standard soware in t h e Geoc hro nology Laboratory in W arsaw (Gorka & Hercman 2002). e quoted errors are one standard deviation. Measure ments in Bergen also used b-spectrometry, but t h e data were processed using t h e Age4U2U program (Lauritzen 1993). METHODS POSTOJNSKA JAMA Postojnska jama (Reg. No. 747; 45.79N; 14.18E; 511 m a.s.l.; Fig. 2) is developed in t h e Postojnski kras (Fig. 1). Its entrance is situated near t h e contact of Eocene ysc h wit h Upper Cretaceous lime stone (Buser, Grad & Pleniar 1967). e limestones are folded into t h e NW SE trending Postojna Anticline (Gospodari 1976). Cave passages h ave a general NS trend, oblique to t h e anticlinal axis. e Pivka River formed t h e cave. Its modern ponor h as an elevation of 511 m a.s.l., w hile t h e terminal sump in Pivka jama h as an elevation of 477 m a.s.l. Currently explored passages occur at two main levels wit h a total CAVE SEDIMENTS FROM THE POSTOJNSKAPLANINSKA CAVE S Y STEM SLOVENIA: EVIDENCE OF MULTIPHASE ...

PAGE 6

ACTA CARSOLOGICA 37/1 2008 68 lengt h of 20.5 km. e upper cave level ranges in altitude between 529 m a.s.l. at t h e main entrance and 520 m a.s.l. in rna jama. is level, mostly formed by epip hreatic speleogenesis, is composed of large passages, generally up to 10 m hig h and wide wit h rounded proles and traces of paragenesis. All studied sediment proles were situat ed in t his level between 533 and 520 m a.s.l. (Tab. 1). e lower level, w h ere t h e modern underground Pivka river ows, is about 18 m below t h e upper one. e riverbed h as a low gradient. e active river passages are smaller t h an t h e hig h er ones. e water level can rise substan tially, by more t h an 10 m during oods. e cave contains remnants of several kinds of al luvial deposits c h aracteristic of internal cave facies cov ered by and/or intercalated wit h speleot h ems. Entrance facies deposits consisting of slope-derived debris mixed wit h t h e uvial deposits are found in Biospeleoloka postaja (Biospeleological station). Samples from a total of eig h t proles (Fig. 2; Tab. 1) were collected in Posto jnska jama. SPODNJI TARTARUS Spodnji Tartarus (Lower Tartarus) passage connects t h e upper cave level to t h e lower active level. e oor of t h e passage is covered wit h grey uvial lutites and occasionally oods. ere is a small h ig h er-level side passage, Rov koa licije (Coalition Passage), in t h e sout h ern part of Spodnji Tartarus. It is a canyon-like passage about 50 m long, 2 m wide and up to 17 m h ig h Inclined and undulating notc h es on t h e walls s h ow t h at it evolved in p h reatic and paragenetic conditions. is passage is almost lled wit h sediments, leaving only a 2 m h ig h space below t h e at paragenetic ceiling (Pl. I/A, B). e sediments were later partly was h ed out, leav ing a 13 m long pile of sedi ments in t h e central part of t h e passage, terminated on t h e nort h ern and sout h ern ends by two vertical proles (Fig. 3): t h e Spodnji Tartarus Nort h prole (at t h e entrance to Rov koalicje; Pl. I/A) and Spodnji Tartarus Sout h pro le (Pl. I/B). L ITHOLOG Y Bot h proles in Spodnji Tar tarus are composed of two units diering in colour. e lower unit is rat h er red w hile t h e upper unit is rat h er yellow. A s h arp erosional boundary wit h a distinct colour c h ange occurs in t h e F ig. 2: Location of the proles in P ostojnska jama and Zguba jama (aer Cave Register of IZRK ZRC SAZU and JZS). N ADJA ZUPAN HAJNA, P ETR PRUNER, A NDREJ MIHEVC, P ETR SCHNABL & P AVEL BOSK

PAGE 7

ACTA CARSOLOGICA 37/1 2008 69 nort h ern prole w hile t h e colour c h ange in t h e sout h ern prole is not distinct. e lower part (Fig. 3; Pl. I/A, B) consists of red to reddis h brown clays overlying collapsed limestone blocks wit h weat h ered surfaces. e clays are silty, commonly laminated, and wit h sandy admixture. Small erosional features at layer boundaries are common and locally associated wit h slig h t ferruginization. Small clasts of corroded limestone occur close to t h e top of t h e nort h ern prole. e upper part of t h e prole (Pl. I/A, B) is composed of yellow, yellowis h brown and lig h t brown Tab. 1: Coordinates of the bases and altitudes of the bases and tops of studied proles. Prole Co-ordinates Altitude [m a.s.l.] N E base top Spodnji Tartarus North 45.16 14.72 526 533 Spodnji Tartarus South 45.58 14.19 522 533 Umetni tunel I 45.20 14.22 520 521 Umetni tunel II 45.82 14.79 521 522 Biospeleoloka postaja 45.97 14.80 530 531 Male jame 45.49 14.89 520 521 Stara jama 45.26 14.81 528 530 Pisani rov 45.90 14.66 529 531 Zguba jama 45.10 14.26 563 564 Rudolfov rov (PL)* 45.49 14.27 468 470 PL = Planinska jama, all ot h er proles are located wit hin t h e Postojnska jama. laminated and banded clays. Some laminae and bands are reddis h brown. Mineral composition is summarized in Table 2. in calcite crusts, sometimes wit h candle stick stalagmites, cover t h e sout h ern prole. Numerous scratc h marks made by t h e cave bear ( Ursus spelaeus ), were discovered at t h e top of t h e sediments and on t h e passage walls. N UMERICAL DATING A small stalagmite on an in clined surface below t h e pro le (sample No. 3) was taken radiometric dating from t h e Spodnji Tartarus Nort h pro le (Tab. 3). Four samples of speleot h ems were taken from t h e sout h ern prole: No. 1, a stalagmite about 25 cm hig h wit h t hin calcite crust at t h e base (wit h base at about 344 cm above t h e base of prole); No. 2, a small stalagmite on a rotated, collapsed limestone block (Pl. I/ C), Nos. 4 and 5, t hin calcite crusts covering t h e prole. Unfortunately, except for two analyses, t h e samples con tained too hig h a proportion of detrital contamination e new /U data placed t h e erosion of t h e pro le in Spodnji Tartarus before 169 ka and t h e collapse took place aer t h e erosion, during t h e Eemian (around 108 ka). Most probably, collapse oc curred aer t h e part of sedi mentary prole was eroded away. Distinct and perfectly preserved scratc h es of Ursus spelaeus on t h e prole (top and incline) indicates t h at t h ere h ave been no substan tial c h anges to t h e morp h ol ogy of t h e prole since t h e cave bear was climbing it. P ALAEOMAGNETIC ANAL Y SIS e upper part (3.85 to 4.84 m) of t h e red prole h as no dened polarity because it F ig. 3: Cross-section of the Rov koalicije passage showing the location of the Spodnji Tartarus North (AA), Spodnji Tartarus South (BB ) proles and white sandstone (modied from Z upan Hajna et al. 2008a). CAVE SEDIMENTS FROM THE POSTOJNSKAPLANINSKA CAVE S Y STEM SLOVENIA: EVIDENCE OF MULTIPHASE ...

PAGE 8

ACTA CARSOLOGICA 37/1 2008 70 prole s h ow a s h ort R polarized excursion in t h e mid dle part of t h e section. It cannot be excluded, t h at t h e R excursion in prole Nort h yellow (4.85.95 m) prob ably correlates wit h t h e R excursion in t h e Sout h prole (9.45.55 m). us we may assume t h at deposition of t h e entire nort h ern prole occurred wit hin t h e Brun h es c hron (<0.78 Ma), but we cannot be certain wit h out com parative data (e.g. wit h fossils). Principal palaeomagnetic parameters and t h eir c h aracteristics for all proles are presented in Table 4. UMETNI TUNEL I e Umetni Tunel I prole samples a passage lled wit h uvial sediments, intersected by t h e articial tunnel (Umetni tunel) between Postojnska Jama and rna jama. contains coarse magnetic grains. e petromagnetic re sults from t h e upper part (4.5 to 7.8 m) of t h e yellow pro le (Spodnji Tartarus Nort h ), s h ow c h anged RM and MS values w h en compared wit h bot h values of lower part of t h e prole. Detailed comparison based on palaeomagnetic inclination and t h e NRM and MS s h ows t h e possibility of correlating t h e following segments: (a) Nort h red 2.554.83 m wit h Sout h 2.454.73 m and (b) Nort h yellow 3.054.80 m wit h Sout h 7.669.40 m. e correlated segments h ave identical depositional rates. e basal parts of t h e Nort h and Sout h proles cannot be correlated at all. Palaeomagnetic results from t h e Spodnji Tartarus Nort h yellow prole s h ow two s h ort R polarized excur sions in t h e middle of t h e section. Palaeomagnetic results from t h e rest of t h e red prole s h ow only N polarization. Palaeomagnetic results from t h e Spodnji Tartarus Sout h Tab. 2: M ineralogical composition of selected cave sediments (mostly aer Z upan Hajna 1998). Mineral/site Spodnji Tartarus Umetni tunel I Umetni tunel II Rudolfov rov Red clay Yellow clay Brown clay Quartz D D + D D D Chlorite + + + + Tr + Kaolinite + + Tr Tr Muscovite + + + + + + Montmorillonite + Tr Feldspars + + + + Microcline + + Plagioclase + + Calcite + + + Goethite Tr + Note: D = dominant; Tr = traces. Tab. 3: /U dating results using -spectrometry, Spodnji Tartarus. Sample Lab. No. U content [ppm] 234 U/ 238 U 230 Th/ 234 U 230 Th/ 232 Th Age [ka] Tartarus 3/2 W 1624 0.0638.0040 1.0419.0886 0.1611.0259 2.8.8 Tartarus 3/1 W 1637 0.0862.0044 1.0339.0691 0.0943.0157 12 Tartarus 5 W 1639 0.0249.0020 0.9732.1045 0.4016.0587 1.2 + 26 Tartarus 4 W 1638 0.0489.0032 1.0552.0934 0.7397.0701 >1,000 + 143 22 Tartarus 2/2 W 1623 0.2988.0094 1.2166.0461 0.9452.0345 2,9.2 + 19 Tartarus 2/1 W 1636 0.0486.0031 0.9054.0825 0.6219.0664 25 108 17 Tartarus 1/2 W 1625 0.0523.0030 0.8276.0670 0.4304.0472 3.4.7 Tartarus 1/1 W 1635 0.0367.0028 0.7570.0838 0.5312.0764 4.5.2 N ADJA ZUPAN HAJNA, P ETR PRUNER, A NDREJ MIHEVC, P ETR SCHNABL & P AVEL BOSK

PAGE 9

ACTA CARSOLOGICA 37/1 2008 71 nor Sasowsky et al. (2003) detected t his N polarized magnetozone, due to t h eir smaller number of samples. e presence of t h e N polarized magnetozone may s hi t h e age of sediments in t his prole beyond t h e expected limit of 0.99 Ma. e N polarized zone could represent t h e Jaramillo subc hron (0.99.07 Ma) as well as ot h er N polarized subc hrons wit hin t h e Matuyama (e.g., Olduvai: 1.77.95 Ma, Reunion: 2.14.15 Ma) or older c hrons (Pruner et al. 2004). e sediments in t h e Umetni Tunel I prole are t h e oldest so far discovered in t h e Postojnska jama-Planinska jama system. UMETNI TUNEL II is prole is located near t h e entrance to t h e tunnel from Postojnska jama (Fig. 2; Tab. 1). Lithology e section is composed of yellowis h brown clays, laminated to banded wit h sandy admixture, wit h yellow is h brown neto medium-grained sands at t h e base. Cross bedding can be observed in t h e middle part of t h e section. e upper part consists of h omogeneous lig h t yellowis h brown coloured silty clays, laminated at t h e top and wit h an indistinct erosional base. Mineral composi tion is summarized in Table 2. Palaeomagnetic analysis Only two components were isolated aer AF de magnetization. e c h aracteristic component is not stable and cannot be isolated in AF. e samples are not suitable for palaeomagnetic investigation. It is located approximately h alf way along t h e tunnel (Fig. 2; Tab. 1; Pl. I/D). e lled passage is cut by t h e NESW trending cross-Dinaric fault. Bot h Gospodari (1964) and Sasowsky et al. (2003) concluded t h at t h e fault was active aer t h e sediments were deposited. Lithology e Umetni tunel I prole is 130 cm hig h, termi nated by t h e limestone ceiling and t h e excavated oor of t h e tunnel. e lower h alf of t h e prole is composed of alternating medium to coarse-grained, slig h tly arcosic and cross-bedded sands. e upper h alf (Pl. I/D) consists of alternating laminated brown to grey clays and medi um-grained, slig h tly arcosic sands. A t hin ferruginized h orizon (ferricrete) is developed h ere, accompanied by weat h ering of t h e clay surface. Mineral composition is summarized in Table 2. e sediments are disturbed by slickensides and t h e middle clay layer is anticlinally banded. Palaeomagnetic analysis ebela & Sasowsky (1999) and Sasowsky et al. (2003) studied t his prole previously, but t h eir sampling was widely spaced and t h eir results s h owed distinct scatter. ey interpreted t h e age of prole as being between 0.99 and 0.78 Ma (top of t h e Matuyama c hron) and also calcu lated palaeo-rotations, but t h e distribution of directions from so few samples cannot be used for Fis h er statistics (Fis h er, 1953). Our hig h-resolution magnetostratigrap h y of t h e ne-grained deposits indicated a N polarized zone wit h in t h e R polarized. Neit h er ebela & Sasowsky (1999) Tab. 4: Review and characteristics of principal magnetic values. Prole/parameter Interval [m]* J n [mA.m -1 ] k n x 10 -6 [SI] Character of J n and k n magnetic values Scatter J n and k n n* Spodnji Tartarus North, red 4.835.03 7 228,180 Low high Great 101 Spodnji Tartarus North, yellow 7.445.05 1 89 Low high Great 88 Spodnji Tartarus South 9.41.02 0.6 80,610 Very low high Extreme 176 Umetni tunel I 0.02.42 0.6 123 Very low intermediate Great 59 Umetni tunel II 0.54.04 0.2 50 Very low Minor 14 Biospeleoloka postaja 0.31.04 97 1,9602,720 Intermediate high Minor 15 Male jame 0.95.03 0.8 96 Very low intermediate Great 33 Stara jama 1.35.01 2.7.6 225 Low intermediate Minor 26 Pisani rov 1.92.19 6.8 27 Low intermediate Moderate 66 Zguba jama, prole I 1.04.01 2 114,119 Low intermediate Moderate 37 Zguba jama, prole II 0.21.03 1 114 Low intermediate Great 8 Rudolfov rov 0.02.23 0.68.94 82 Low very low Moderate 85 Note: All listed samples are demagnetized CAVE SEDIMENTS FROM THE POSTOJNSKAPLANINSKA CAVE S Y STEM SLOVENIA: EVIDENCE OF MULTIPHASE ...

PAGE 10

ACTA CARSOLOGICA 37/1 2008 72 BIOSPELEOLOKA POSTAJA e prole in t h e Biospeleoloka postaja (Fig. 2; Pl. I/E) is situated at t h e entrance to Rov novi h podpisov (Pas sage of New Signatures). is passage is one of t h e former entrances to t h e cave. Lithology Two sediment proles eac h approximately 7 m t hick occur in t h e Biospeleoloka postaja section of t h e cave. One is exposed in t h e (articial) railway tunnel and t h e ot h er is exposed along t h e tourist pat h. e prole along t h e tourist pat h was investigated. It contains four types of sediment. e top layer consists of reddis h loam covered wit h owstone (Pl. I/E). e upper part of t his layer is composed of limestone clasts. Brodar (1969) found Palaeolit hic tools (Mousterian, about 40 ka old) h ere. Below t h em t h e prole consists of gravels composed of angular clasts of owstone and limestone. e base of t h e exposed prole is composed of silts, sands and gravels wit h clastic components derived from ysc h source rocks. Palaeomagnetic analysis W e were aware t h at t h e sediments in t his prole are very young but wanted to test t h e met h od in t his kind of deposit. e prole s h owed only N polarized magneti zation. W e assume t h at deposition occurred wit hin t h e Brun h es c hron (<0.78 Ma), w hic h is in accordance wit h arc h aeological nds. MALE JAME Male jame (Lesser Cave) is a smaller epip hreatic passage. It connects Stara jama passage to Zgornji Tartarus (Upper Tartarus) passage. e prole is located in t h e nort h ern part of t h e passage (Fig. 2; Pl. II/A). ebela & Sasowsky (1999) sampled t his prole previously. Lithology T h e top 60 cm of t h e profile consists of green is h grey, clayey-silty-sandy gravels wit h fine-grained laminae and bands. T h ey are underlain by 120 cm of quite well lit h ified, brownis h grey to reddis h brown and red, fine to medium-grained, h ig h ly-micaceous sands wit h a clayey-silty matrix. T h ese sands contain well-rounded pebbles of flysc h sandstone, small clasts of decomposed w h ite c h erts and angular clasts coated by ferruginous crusts, wit h sc h lieren enric h ed in Mn compounds. T h e sandstone pebbles are strongly de composed in situ and covered wit h Mn-enric h ed coat ings. T h e lower part of t h e profile (90 cm) is composed of soft yellowis h brown clays, wit h brown bands and laminae in its upper part, w h ile t h e lower part is h o mogeneous (Pl. II/A, B). Palaeomagnetic analysis All samples from t his prole h ave N polarization suggesting, as far as can be determined, deposition wit h in t h e Brun h es c hron (<0.78 Ma). ebela & Sasowsky (1999) reac h ed t h e same conclusion. Nevert h eless, t h e in situ weat h ering of t h e ysc h-derived sandstone pebbles and strong lit hication could indicate a substantial age for t h ese deposits. According to t h e stratigrap h y of Gospodari (1976, p. 100), t h e so and slig h tly lit hied clays s h ould be t h e youngest sediment in t h e passage (<0.78 Ma) overlying, deposited on or in a cavity wit hin t h e already eroded sandy-gravely unit wit h w hite c h erts (see Osborne 1984). e sandy-gravely unit could easily be muc h older (even >0.78 Ma). STARA JAMA Stara jama (Old Cave) is t h e main passage in t h e upper level of t h e cave. e passage slopes gently to t h e nort h. ere are traces of paragenesis on t h e passage ceiling. Relicts of sedimentary ll are preserved in some parts of t h e passage w h ere t h ey h ave been protected by owstone or collapse (Pl. II/C). e Stara jama prole is located at Kri (Old Cave at Cross) opposite to t h e entrance to Kristalni rov (Crystal Passage; Fig. 2). Lithology is prole is approximately 220 cm hig h. e lower h alf of t h e prole consists of grey clayey-sandy gravels wit h at clasts to pebbles and some bands of ne-grained sand constitute. Laminated greyis h lig h t brown silty-san dy clays overlie t h e gravels. Fragments of bones (prob ably Ursus spelaeus ; det. I. Horek, 2007) were found in t h e rig h t part of t h e outcrop. e clays are succeeded by greyis h lig h t brown, clayey, ne-grained sands wit h darkcoloured sc h lieren/laminae and clay clasts in t h e upper part. ese are overlain by yellowis h to lig h t brown and silty clays, micaceous, indistinctly laminated, wit h sandy admixture increasing to t h e top, at t h e top up to unsorted sands wit h single at pebbles. e top of t h e prole con sists of lig h t brown silty clays wit h aky disintegration and small clasts of limestone near t h e top. e owstone at t h e top is about 20 cm t hick (Pl. II/C, D). Palaeomagnetic analysis Since only N polarization was detected, t h e palaeo magnetic results indicate t h e sediments are younger t h an 0.78 Ma. Pleistocene bear bones occurring on t h e top of t h e prole support t his interpretation. N ADJA ZUPAN HAJNA, P ETR PRUNER, A NDREJ MIHEVC, P ETR SCHNABL & P AVEL BOSK

PAGE 11

ACTA CARSOLOGICA 37/1 2008 73 PISANI ROV Pisani rov (Coloured Passage) deviates to t h e nort h from t h e main passage of Stara jama. It terminates below t h e slopes of t h e collapse doline of Velika Jeranova w h ere t h e bottom is lled by sediments. e prole is situated at t h e end of t h e Pisani rov (Fig. 2; Pl. II/E). Lithology is prole is roug h ly 145 cm t hick and is com posed of yellowis h brown silts to clays wit h dark stains and sc h lieren, wit h an olive-green h orizon in t h e lower t hird, nely laminated in t h e lower part. Cubic to colum nar disintegration wit h Fe stains on t h e fractures occur at t h e top of t h e silts and clays. e prole is covered by two layers of owstone wit h broken stalagmites (Pl. II/E, F). Numerical dating Gospodari (1981, p. 93) dated a stalactite from Pisani rov using ESR. Red owstone in t h e nucleus was older t h an 530 ka; t h e next two rings, separated by lami na of ood loam, were dated to about 280 ka and to about 125 ka, respectively. Anot h er stalactite core was dated to 190 ka (Ikeya et al. 1983). Palaeomagnetic analysis e prole s h owed only N polarized magnetization. is suggests t h at deposition occurred wit hin t h e Brun h es c hron. Numerical dates from speleot h ems limit t h e time span of deposition to between >0.53 and <0.78 Ma. ZGUBA JAMA Zguba jama (Reg. No. 6290; 45.50N; 14.57E; 561 m a.s.l.; Fig. 2) is a small blind cave formed in Upper Cretaceous limestones (Buser, Grad & Pleniar 1967). e cave located at an elevation above Postojnska jama. Pisani rov, t h e closest part of Postojn ska jama, ends 265 m h orizontally and 34 m below Zgu ba jama (Fig. 2). e cave entrance opens in t h e eastern slope of Mala Jeranova dolina. Zguba jama is 122 m long and consists of a simple SW NE-trending c h annel, no more t h an 2 m wide, 2 m hig h and 4 m deep. It was once completely lled wit h allogenic uvial sediments (Pl. II/G). ebela (1994) in terpreted t h e cave as probably being a hig h level fossil p hreatic c h annel. Lithology e Zguba jama prole is located inside t h e cave close to its end, w h ere a small excavation on t h e NE side of t h e c h annel was sampled to a dept h of about 121 cm. e upper part (Pl. II/G) consists of reddis h brown silt to clay, bioturbated at t h e top (worms). e lower part is very ne-grained sand, wit h akes of muscovite and small irregular concretions. e boundary wit h t h e un derlying layer is marked by a carbonate-cemented h ori zon about 3 cm t hick. e lower 70 cm is multicoloured banded silty-clayey ne-grained sand wit h akes of mus covite and some clay laminae near t h e base. e basal part (about 20 cm) contains desiccation cracks, lled in places by red silt to clay. Palaeomagnetic analysis Nearly all samples in t h e prole s h owed N polar ity, except for some very s h ort R or transient excursions (NR) in upper prole (I). e upper and lower parts of t h e prole dier in principal palaeomagnetic parameters (Tab. 4). Alt h oug h data could support quite young ages, D and I parameters indicate t h at t h e N polarization could belong to an older c hron/subc hron t h an Brun h es and t h at t h e sediments could be substantially older t h an 0.78 Ma. is is supported by t h e alteration due to a deposi tional hiatus in t h e lower part of t h e prole. PLANINSKA JAMA Planinska jama (Reg. No. 748; 45.62N; 14.39E; 453 m a.s.l.; Fig. 2) is situated on t h e sout h ern edge of Planinsko polje, wit h its entrance at t h e end of a large pocket valley. e cave is developed in Up per Cretaceous limestones and dolomites (Buser, Grad & Pleniar 1967). Planinska jama is a 6,656 m long outow cave (Fig. 4) and is t h e t main spring of t h e Unica River (Fig. 1). Pas sages in Planinska jama are large, about 15 m wide and CAVE SEDIMENTS FROM THE POSTOJNSKAPLANINSKA CAVE S Y STEM SLOVENIA: EVIDENCE OF MULTIPHASE ...

PAGE 12

ACTA CARSOLOGICA 37/1 2008 74 hig h. ere are some collapse c h ambers. e conuence of t h e Pivka River, from Postojnska jama and t h e Rak River from Rakov kocjan occurs in t h e cave. e two main passages of t h e cave were named aer t h eir tribu taries: Pivki rokav (Pivka Branc h) and Rakov rokav (Rak Branc h). ere are a few inactive small side passages at hig h er elevations, suc h as Rudolfov rov (Passage of Ru dolf). Rudolfov rov, a sout h-side passage of Rakov rokav, is approximately 200 m long and is located at an elevation of 460 m a.s.l., about 16 m hig h er t h an t h e main pas sage oor. A small stream ows out of Rudolfov rov. e water in t his stream comes from t h e Javorniki Mountains and Cerkniko polje (Cerknica Polje) and uvial sedi ments once completely lled t h e passage. e Planinska jama prole sampled t h e same sediments, old laminat ed loam, as Gospodari (1976) and ebela & Sasowsky (1999), but in a dierent location, at an elevation of about 460 m a.s.l. (Fig. 4; Pl. II/H). Lithology Gospodari (1976) compiled a stratigrap hic succes sion for sediments from Planinska jama. He considered t h at erosion of t h e sediments was contemporaneous wit h t h e last p h ase of owstone formation (post-glacial). He reported t h at Rudolfov rov was lled wit h t h e older laminated (varved) loam, resting on t h e erosional bed of t h e c h annel. Like all t h e oldest sediments in ot h er caves of t h e region (Posto jnska, Otoka, Risovec & Krina caves), Gospodari dated t h e older laminated loam to about 350 ka (Gospodari 1976, 1981, 1988). Flowstone on old laminated loam was dated by t h e /U met h od to 77.8 ka by Gospodari (1977b). e mineral composition of t h ese sediments indicates t h at t h eir provenance is from t h e Eo cene ysc h of Pivka kotlina (Gospodari 1976; Zupan Hajna 1992, 1998). e prole is approxi mately 2.2 m t hick and is a sequence of yellowis h brown and yellow silty clays in terlaminated by yellowis h brown to w hitis h grey silts and ne-sandy silts to sands. Cross bedding wit h ferruginous laminae are locally de veloped in t h e lower part of t h e prole (Pl. II/H). Miner alogical composition is summarized in Table 2. Numerical dating Gospodari (1976) dated speleot h ems from a num ber of localities in Planinska jama using 14 C and /U met h ods. e range of dates is from 3.6 to 44.2 ka by t h e rst met h od. Seven samples were older t h an 30.7 to 49.9 ka. One sample, wit h age >45.6 ka was dated by / U met h od to 79.7 1.6 ka. Flowstone above t h e older laminated loams was dated to 77.8 8.4 ka (Gospodari 1976). Comparison of /U and radiocarbon ages s h ows t h at 14 C dates older t h an about 30 ka are not reliable due to t h e state of t h e radiometric met h od at t h e time Gospodari was working. Palaeomagnetic analysis ebela & Sasowsky (1999) publis h ed results from a prole consisting of approximately 4 m of yellowis h brown laminated loams. Palaeomagnetic properties were studied on 12 samples (6 duplicates). Normal polariza tion of all samples was interpreted as indicating an age F ig. 4: Location of the Rudolfov rov prole; P laninska jama (aer Cave Register of IZRK ZRC SAZU and JZS). N ADJA ZUPAN HAJNA, P ETR PRUNER, A NDREJ MIHEVC, P ETR SCHNABL & P AVEL BOSK

PAGE 13

ACTA CARSOLOGICA 37/1 2008 75 younger t h an 0.73 Ma. ey suggested t h at t h e palaeo magnetic results are in good accordance wit h Mindel age (0.35.59 Ma) proposed by Gospodari (1981). Our re sults conrmed only t h e N polarization. It appears t h at t h e prole can be placed wit hin t h e Brun h es c hron, i.e. t h e sediments are most probably younger t h an 0.78 Ma but older t h an ca 80 ka. DISCUSSION SEDIMENTOLOG Y AND MINERALOG Y T h e most common clastic sediments in t h e profiles are fine-grained laminated sediments (laminites), often occurring in quite t h ick sequences. Grain size often fines upwards wit h in t h e laminae, creating a colour c h ange. Lamination and fining upwards are normal and indicate deposition from suspension of waning floodwaters or ot h er pulsed flow or from ponding as consequence of outflow routes becoming blocked by collapse, tectonic movements etc. ( cf. Ford & W illiams 2007). T h e abundance of lutitic clastic components indicates t h at t h e coarse-grained load was previously deposited or sieved (e.g., in sumps and semi-sumps) as suggested by Gospodari (1976) and Zupan Hajna (1992). e depositional environment was mostly calm but not completely stagnant. Suc h an environment can be de scribed as cave lacustrine wit h deposition from pulsed ow. Homogeneity of t h e palaeomagnetic data indicates fast continuous deposition during s h ort-lasting single ood events. e mineral composition of t h e uvial sediments indicates t h at t h e catc h ment area for allogenic streams was situated for a very long time on weat h ered ysc h rocks in t h e Pivka kotlina (Gospodari 1976; Zupan Hajna 1992, 1998). e uniformity of mineral composi tion results from h omogenization of t h e source material. According to Zupan Hajna et al. (2008a), t h e principal reasons for h omogenization are rstly t h at t h e inuent streams ow perpendicular to strata of varying compo sition in t h e catc h ment area, secondly weat h ering and pedogenesis on t h e surface aected by t h e alternation of climates in post-Oligocene times, t h irdly multiple reworking and re-deposition wit h in t h e catc h ment area and subsequently in t h e subsurface, and fourt h ly un roong of caves. e lit h ological c h aracter of t h e upper parts of t h e proles in Male jame and Zguba jama, prole II, indicates in situ weat h ering of previously deposited sediments. In situ weat h ering inside caves requires a long residence time in t h e cave and may indicate a dierent external cli mate t h an currently exists (Zupan Hajna et al. 2008a). NUMERICAL DATING A number of samples from Postojnska jama were ana lysed by radiometric dating ( 14 C, /U and ESR) and pal aeontology. Numerical dates identied speleot h em older t h an 350 ka (Gospodari, 1981; Zupan, 1991) wit h some older p h ases dated to about 270 ka, 190 ka, 150 ka and around 70 ka (Ikeya et al. 1983, Zupan 1991; Mi h evc 2002). Suc h dates indicate a substantial age for t h e clastic sediments underlying t h e speleot h em. Layers of ood loam inside t h e stalagmites dated ood events during Riss, Riss/W rm and W rm (Ikeya et al. 1983; Zupan 1991). Y ounger /U dates yielded ages from 47 7 ka to 12 5 ka and 6 4 ka (Mi h evc 2002). /U and radiocarbon dates younger t h an about 20 ka are in very good agreement and s h ow a p h ase of stalagmite col lapse at about 10 ka (Franke & Gey h 1971, 1976). e sediments at several sites in Postojnska jama and t h e dat ing of owstone, even if t h e errors are large, indicate t h at some clear p h ases of collapse alternating wit h p h ases of owstone deposition occurred aer t h e upper cave level was drained. MAGNETOSTRATIGRAPH Y ebela & Sasowsky (1999) did t h e rst palaeomagnetic researc h on uvial sediments from Postojnska jama (Male jame, Otoka jama & Partizanski rov = Umetni tunel). Unfortunately t h e number and sample density of t h eir sampling was not sucient for detailed palaeomag netic analysis, so t h eir and magnetostratigrap hic frame work was very roug h. All samples s h owed N orientation. ey concluded t h at all t h e sediments are younger t h an 0.73 Ma. eir palaeomagnetic measurements from t h e natu ral cavity in Umetni tunel s h owed R polarity and t h ey proposed t h ese sediments were at least 0.73.90 Ma in age. ey deduced t h at reddis h-brown sandy loams in Umetni tunel are older t h an bot h t h e coloured c h ert grav el and older laminated loam ( sensu Gospodari 1976). e Umetni tunel sediments were disturbed by a recent movement on a cross-Dinaric fault situated in t his part of t h e cave (Gospodari 1964; Sasowsky et al. 2003). ebela (2008) interpreted t his movement as being younger t h an CAVE SEDIMENTS FROM THE POSTOJNSKAPLANINSKA CAVE S Y STEM SLOVENIA: EVIDENCE OF MULTIPHASE ...

PAGE 14

ACTA CARSOLOGICA 37/1 2008 76 0.78 Ma, w hic h conicts wit h our magnetostratigrap hic results from prole Umetni tunel I, previously presented by (Pruner et al. 2004). e multi-component analysis tec hnique of Kirsc h vink (1980) s h ows t h at all our samples, except t h ose from prole Umetni tunel II, display t hree-component RM. e A component in all samples is undoubtedly of vis cous origin. It can be demagnetized in AF and at a tem perature range of 0 to 5 (up to 10) mT. e B compo nent (mostly LFC) is also of secondary origin, but s h ows h arder magnetic properties. e B components represent t h e transition between t h e A and C components and oc cur in many samples in t h e AF range of 2 to 15 mT. ey are c h aracterized by a great scatter of directions (see Tab. 5). For example, t h e B component is clearly visible in Figures 5 and 6. e C component (mostly HFC) is t h e most stable and can be demagnetized in t h e AF or at temperature range of ca (10) 15 to 100 mT or 320 to 520 (560) C. Magnetomineralogical analyses and un blocking temperatures (520 to 560 C) determined indi cate t h at magnetite is t h e carrier of t h e RM for all samples analysed, w hic h is in accordance wit h publis h ed data (for summary see Bosk, Pruner & Kadlec 2003). Our data partly conrmed t h e ndings of ebela & Sasowsky (1999), produced some new dates, but indi cated a dierent age interpretation t h an was oered by Sasowsky et al. (2003) and ebela (2008). Samples from most proles were N polarized. ree s h ort R magne F ig. 5: Results of AF demagnetization with R palaeomagnetic polarity (ad). Top le: stereographic projections of the remnant mag netization of a sample in the natural state (crossed circle) and aer AF demagnetization. Top right: Z ijderveld diagram solid circles represent projection on the horizontal plane (X, Y), open circles represent projections on the northsouth vertical plane (X, Z). B ottom le: a graph of normalized values of the remnant magnetic moments versus AF demagnetizing elds. N ADJA ZUPAN HAJNA, P ETR PRUNER, A NDREJ MIHEVC, P ETR SCHNABL & P AVEL BOSK

PAGE 15

ACTA CARSOLOGICA 37/1 2008 77 tozones (excursions) were detected in only a few places (Spodnji Tartarus). e present magnetic directions for study sites are: D = 2.46 (2.46.47) [ o ] and I = 62.18 (62.17.20) [ o ]. Palaeomagnetic directions for t h e proles are summa rized in Table 6 and Figure 7. ese data indicate t h at t h e Spodnji Tartarus Nort h, Pisani rov and Biospeleoloka postaja proles s h ow D and I directions close to t h e pres ent, wit hin t h e limits of statistical error. e Rudolfov rov, Spodnji Tartarus Sout h, Umetni tunel I, Male jame and Zguba jama proles s h ow slig h t to distinct coun ter-clockwise rotation. Palaeomagnetic directions of t h e Stara jama prole indicate clockwise rotation since t h e prole was deposited. e inclination in N polarized samples from Spodnji Tartarus Sout h and Umetni tunel I proles is anomalously low. W e h ave interpreted most of t h e sediments as being younger t h an 0.78 Ma, belonging to dierent depositional phases wit hin t h e Brun h es c hron. Nevert h eless, t h e N po larization in some proles may be linked wit h ot h er N polarized subc hrons muc h older t h an 0.78 Ma (e.g., t h e Umetni tunel 1 site and Zguba jama; Fig. 8). e lit h o logical situation in Male jame is questionable. e dominant N polarization of t h e sediments and t h e lack of more common R polarized excursions also indicate t h at deposition of most proles was relatively rapid, occurring in s h ort time-spans. Some proles may t h erefore represent single ood events or deposition last ing at most a few t h ousand years. Homogeneity of pa laeomagnetic data and lit h ological c h aracter, especially in t h e laminated sequences, may indicate continuous deposition, w hic h is generally favourable for recording F ig. 6: Results of AF demagnetization with R (a) and N (bd) palaeomagnetic polarity. F or explanations see F igure 5. CAVE SEDIMENTS FROM THE POSTOJNSKAPLANINSKA CAVE S Y STEM SLOVENIA: EVIDENCE OF MULTIPHASE ...

PAGE 16

ACTA CARSOLOGICA 37/1 2008 78 s h ort-lived palaeomagnetic excursions. Despite t h ese fa vourable conditions, excursions of t h e palaeomagnetic eld were not detected in t h e proles from t h e Postojnska jamaPlaninska jama cave system. Our team h as, h owever detected t h em in lu titic sediments in some ot h er caves (Bella et al. 2007; Zu pan Hajna et al. 2008a). AGE e interpretation of t h e age of t h e proles in Table 7 is based on all available nu merical and correlated-ages, but t h e primary source is t h e results of our magne tostratigrap h y of t h e cave sediments. Maximum ages, except w h ere substantiated by ot h er met h ods, represent roug h estimates only and could be far from reality. Furt h ermore, at most sites t h e sedimentary prole was not excavated completely to t h e rock cave oor, so t h e be ginning of deposition may be substantially older, particu larly w h ere ages are designed as over 0.78 Ma. Zupan Hajna et al. (2008a) recognised t hree pe riods of cave ll deposition in Slovenia, wit h t h e oldest sediment dated as Pliocene by Horek et al. (2007). De position in t h e studied cave system can be placed in two periods (Tab. 7). Sediments dated from about 0.78 M a (palaeomag netic age) up to more than 4.0 M a (palaeomagnetic age; i.e. P leistocene max. to P lio cene): t his group includes a succession of detected ages across t h e w h ole Kras. e base of most of t h ese proles is probably not muc h older t h an 3.58 Ma, i.e. t h e datum adjusted by palaeontologi cal nds at t h e rnotie II and Raika peina sites in t h e Classical Karst (Horek et al. 2007). Some p h ases can be distinguis h ed from t h e Tab. 5: Directions of B and C components in selected samples. Prole Polarity of C comp. B component C component D [ o ] I [ o ] D [ o ] I [ o ] Spodnji Tartarus N Yellow Sample P 220 R 216 71 185 -49 Spodnji Tartarus South Sample P 556 R 292 5 278 -46 Umetni tunel I Sample U 11 R 314 63 171 -56 Umetni tunel I Sample P 71 R 82 -7 165 -71 Zguba jama, prole I Sample PC 010 R 209 47 206 -40 Zguba jama, prole I Sample PC 083 N 219 61 27 72 Pisani rov Sample PA 072 N 355 60 10 49 Rudolfov rov (PL) Sample PL 16 N 27 52 358 56 F ig. 7: Directions of C-components of remanence of samples with N and R polarities, P ostojnska, Zguba and P laninska jama. Stereographic projection, open (full) small circles represent projection onto the lower (upper) hemisphere. e mean direction calculated according to F isher (1953) is marked by a crossed circle; the condence circle at the 95% probability level is circumscribed about the mean direction. N ADJA ZUPAN HAJNA, P ETR PRUNER, A NDREJ MIHEVC, P ETR SCHNABL & P AVEL BOSK

PAGE 17

ACTA CARSOLOGICA 37/1 2008 79 Tab. 6: M ean palaeomagnetic directions, P laninska (P L), P ostojnska and Zguba jama. Prole Polarity Mean palaeomagnetic directions 95 [ o ] k n* D [ o ] I [ o ] Spodnji Tartarus N Red N 3.11 65.17 3.52 20.55 77 Spodnji Tartarus N Yellow N 355.18 63.05 4.35 16.41 63 R 190.13 -52.25 14.23 24.19 4 Spodnji Tartarus South N 349.34 49.72 3.2 11.49 167 R 263.48 -30.63 2 Umetni tunel I N 312.23 48.63 10.22 7.22 26 R 140.33 -57.38 12.26 5.01 26 Biospeleoloka postaja N 358.34 54.47 3.86 87.52 15 Male jame N 323.93 71.67 4.37 32.08 32 Stara jama N 19.19 64.84 4.59 51.61 18 Pisani rov N 6.19 63.79 3.31 27.16 66 Zguba jama, prole I N 340.38 69.93 4.62 27.81 33 R 217.04 -31.14 2 Zguba jama, prole II N 19.27 75.25 12.99 19.37 6 Rudolfov rov (PL) N 349.73 61.33 2.16 50.2 84 Note: only samples suitable for statistical evaluation are listed. dates: (a) more t h an 0.78 up to about 4.2 Ma (palaeomag netic ages e.g., Umetni tunel I & Zguba jama), and (b) less t h an 0.78 to about 2 Ma (palaeomagnetic ages), i.e. somet hing between Brun h es/ Matuyama boundary (and somew h at younger) and base of Jaramillo and/or Olduvai subc hrons (and somew h at older). Dates from Male jame and Zguba jama do not allow for more detailed age correla tion. Sediments younger than 0.78 M a (i.e. P leistocene): caves containing sedimenta ry ll younger t h an t h e Brun h es/Matuyama boundary h ave one common feature a part of t h e cave is still h ydro logically active, wit h one or more streams owing in t h e lower levels (e.g., Postojnska, Planinska). EVOLUTION OF THE CAVE S Y STEM Data obtained from sedimen tological analyses, palaeonto logical and quantitative dat ing indicates t h at t h e Pivka kotlina Postojnska jama Planinska jama Planinsko polje system evolved over a long period of time during relatively stable h ydrologi cal conditions. e develop ment of t h e w h ole system h as been governed by t h e level of Planinsko polje in relation to t h at of t h e resurgence area in Ljubljansko barje (Ljubljana Moor). Ljubljansko barje is a tectonic basin t h at h as been slowly subsiding during t h e Quaternary. e evolution of Pivka kotlina h as been gov erned by t h e elevation of t h e ponors of Pivka River t h at drain into Postojnska jama. F ig. 8: Correlation of magnetostratigraphy proles with the GPTS (on both sides; modied from Cande & Kent 1995). P osition of individual study proles does not reect mutual correlations and/ or altitude; top of proles is located on the same line. Numbers of proles from P ostojnska jama re fer to F igure 2. 1a Spodnji Tartarus North, red; 1b Spodnji Tartarus North, yellow; 1c Spodnji Tartarus South; 2 Umetni tunel I; 4 B iospeleoloka postaja; 5 M ale jame; 6 Stara jama; 7 P isani rov; 8a Zguba jama, prole I; 8b Zguba jama, prole II; 9 Rudolfov rov. Right scale bar is valid for proles Nos. 2 and 8b, le scale bar is valid for the rest of proles. CAVE SEDIMENTS FROM THE POSTOJNSKAPLANINSKA CAVE S Y STEM SLOVENIA: EVIDENCE OF MULTIPHASE ...

PAGE 18

ACTA CARSOLOGICA 37/1 2008 80 Recent palaeomagnetic data from t h e lower part of a prole in Markov spodmol indicate t h at Pivka kotlina h as been active for a long period of time. Declination and inclination parameters in t his prole indicate t h at t h e N polarization could belong h ere to an older c hron/ subc hron t h an t h e Brun h es. is suggests t h at t h e sedi ments may be substantially older t h an 0.78 Ma. Several segments in t his prole are separated by distinct uncon formities, and at least one of t h em is accompanied by in Tab. 7: Ages of cave sediments interpreted on studied sites (modied from Z upan Hajna et al., 2008a). Name of site Name of prole Age [Ma] Age of cave ll Min. Max. Planinska jama Rudolfov rov >0.08 <0.78 Pleistocene (Mindel) to Holocene Postojnska jama Stara jama ? <0.78 Postojnska jama Tartarus South >0.122 <0.78 Postojnska jama Pisani rov >0.53 <0.78 Postojnska jama Tartarus North ? <0.78 Zguba jama I+II <0.78 >0.78 Pliocene* to Pleistocene (Gnz/Mindel) Postojnska jama Male jame ? >0.78 Postojnska jama Umetni tunel 1 <0.99 >2.15 Explanations: bold numbers = /U data; Pliocene mentioned in t h e traditional sense (1.8 Ma). situ underground weat h er ing, indicative of a substan tial time break (Zupan Hajna et al. 2008a). Continued stability of t h e Pivka River h ydrologi cal system over a long period of time led to t h e formation of a long and complex cave system wit h t hree currently accessible cave levels (Zguba jama + upper and lower lev els of Postojnska jama; Fig. 9). Vertical separation be tween t h e levels is not great (about 30 and 18 m, from top down). e Zguba jama level is a single p hreatic c h annel abandoned wit h out epip hre atic or vadose modication; t h e period of h ydrological stability represented by t his level was apparently rela tively s h ort. e upper level of Postojnska jama s h ows traces of original p hreatic c h annels developed along geological structures. ey were substantially modied in t h e epi p hreatic zone by paragenesis, entrenc hment and bypass es during a long period of stable h ydrologic and speleo genetic conditions wit h a low h ydraulic h ead. Processes in epip hreatic zone involved multiple periods of sediment lling and erosion causing t h e present cave s h ape to evolve by paragenesis. A two-p h ase evolution is also expressed in t h e relief of Pivka kotlina (see Mulec et al. 2005). In Planinska jama, located on t h e lower (outow) side of t h e system, sedimentation was less important and t h e dominant process in cave de velopment was and is vadose entrenc hment. uteri and his col leagues (a.o. uteri, uter i & Stepinik 2003) in a number of nearly identi cal papers, proposed a very young evolutionary history for Planinsko and Cerkniko poljes and related cave sys tems, wit h cave development F ig. 9: Schematic cross-section of P ostojna cave system showing the location of sediment proles. Legend: 1 Spodnji Tartarus proles, 2 Umetni tunel I prole, 3 Umetni tunel 2 prole, 4 B iospeleoloka postaja prole, 5 M ale jame prole, 6 Stara jama prole, 7 P isani rov prole and Zguba jama prole. N ADJA ZUPAN HAJNA, P ETR PRUNER, A NDREJ MIHEVC, P ETR SCHNABL & P AVEL BOSK

PAGE 19

ACTA CARSOLOGICA 37/1 2008 81 occurring wit hin t h e last 30 ka. is h ypot h esis s h ould now be abandoned, as it is not in accordance wit h t h e data summarized h ere. eir model was based on reinterpretation of t h e original stratigrap h y of Gospodari and his evolution p h ases (1976, 1977a, b). e succession of events uteri and ot h ers propose may be correct, but t h eir timeframe for t h e inlling and erosional p h ases is unrealistic and far too s h ort. e interpretation given h ere is based on (1) all available numerical and correlative ages (including ar c h aeology) referred to or described in t his paper: nu merical dates up to 530 ka from speleot h ems and cor related-ages over 1.1 Ma (magnetostratigrap h y); (2) t h e morp h ology of Planinsko and Cerkniko poljes: t h e pres ent geograp hical limits of t h e at polje oors are far from t h e faults limiting t h e Idrija fault zone. If poljes along t h e Idrija fault zone really h ave developed as pull-apart ba sins (Vrabec 1994), t h en t h eir enlargement by lateral cor rosion to t h e present extent requires muc h more t h an 100 ka (uteri, uteri & Stepinik 2003), wit h a stable karst water table, and (3) knowledge of t h e dynamics of depositional and erosional p h ases in t h e Postojnska and Planinska jama system (e.g., Gospodari 1976; Pruner et al. 2004; Zupan Hajna et al. 2008a, b). e dating presented above and eld observations bot h indicate a long and complex h istory of alternating depositional and erosional p h ases in t h e Postojnska Planinska cave system. Individual cave segments or pas sages were fully lled and ex h umed several times during cave evolution, as indicated, for example, by paragenetic features and remains of cemented sediments on walls and ceilings in t h e main passage of Stara jama and in ot h er places, and by dierent palaeomagnetic param eters (D, I). As previously noted by Gospodari (1976), depo sition was not uniform t hroug h out t h e entire cave at any one time. ere was erosion in one part of t h e cave and deposition in anot h er. Repeated reworking and redeposition of t h e same sedimentary material is typical for long, voluminous and complicated cave systems like PostojnskaPlaninska system. Some passages underwent repeated ooding and deposition (e.g., Pisani rov). is complicated depositional history was caused by t h e prolonged evolution of t h e drainage system in epip hreatic conditions, particularly in t h e upper level of Postojnska jama. e alternation of depositional and erosional p h ases may be connected wit h (1) low h ydrau lic h ead in a stable h ydrological situation and (2) oscilla tion of t h e karst water table t hroug h unblocked connec tion passages between t h e lower and upper cave levels as a result of c h anging conditions wit hin t h e cave system (reecting t h e function of t h e resurgence area, climatic c h anges, tectonic movements, collapse and t h e intrinsic mec h anisms of contact karst). Unfortunately, most of t h ese processes h ave not yet been properly dated; but cave sediments can provide a useful tool and t h e rst attempt of dating t h ese processes is presented h ere. Nevert h eless, sediments from dierent proles from caves developed in dierent geomorp hic units s h owing t h e same palaeomagnetic polarity do not necessarily represent t h e same depositional conditions/ times in dierent segments of t h e cave system. Sediment dating does not give t h e time of speleogenesis itself, but only t h e age of t h e last preserved cave ll (see Bosk 2002). CONCLUSIONS e Postojnska jama Planinska jama cave system and number of ot h er smaller caves contain ric h and lit h ologically diverse cave ll, ranging from autogenic speleot h ems to allogenic uvial sediments. e most common clastic sediments are ne-grained (lutitic) laminated sediments (laminites). ey were deposited from suspension from waning oodwaters or ot h er pulsed ow or as a result of ponding due to t h e block age of outow routes. e prevalence of a lutitic clastic component indicates low h ead in t h e catc h ment area and/or fact t h at t h e coarse-grained load was already de posited. e deposition of ne-grained material was due to t h e regular ooding, c h aracteristic for t h e sinking riv ers. Homogeneity of palaeomagnetic data can indicate fast and continuous deposition during s h ort-lasted (few t h ousand years) single-ood events. Depositional style was favourable for record of s h ort-lived excursions of t h e palaeomagnetic eld, w h ic h is rarely reported from cave deposits. e mineral composition of t h e uvial sediments in dicates t h at t h e catc hment area of allogenic streams was situated on weat h ered ysc h rocks in t h e Pivka kotlina for a very long time. e uniformity of mineral and pet rologic composition found in all proles resulted from h omogenization before t h e sediments were deposited in t h e caves and from multiple reworking and re-deposition in t h e subsurface. In situ weat h ering of grains inside t h e cave was also detected. CAVE SEDIMENTS FROM THE POSTOJNSKAPLANINSKA CAVE S Y STEM SLOVENIA: EVIDENCE OF MULTIPHASE ...

PAGE 20

ACTA CARSOLOGICA 37/1 2008 82 Numerical dating identied many p h ases of spe leot h em growt h. Layers of loam inside stalagmites indi cate repeating ood events in some parts of cave system. Some speleot h em dates clearly indicate a substantial age for underlying cave sediments. e sediments, especially in several sites of t h e Postojnska jama, and dating of spe leot h ems, even if t h e errors are large, s h ow some clear p h ases of erosion and collapse in alternating wit h sedi ment and owstone deposition. Multi-component analysis s h ows t h at sediments mostly display a t hree-component remnant magnetiza tion. Magnetomineralogical analyses and unblocking temperatures (520 to 560 C) indicate t h at magnetite is t h e carrier of t h e remnant magnetization in t h e analysed samples. Palaeomagnetic and magnetostratigrap h y data re ported h ere partly conrmed previous results, but also indicated dierent age interpretations. Samples from most proles were N polarized. ree s h ort R magnetozones (excursions) were detected only in a few places (Spodnji Tartarus). W it h in t h e limits of statistical error, t h e Spod nji Tartarus Nort h Pisani rov and Biospeleoloka postaja proles s h ow declination and inclination directions close to t h e present. e Rudolfov rov, Spodnji Tartarus Sout h Umetni tunel 1, Male jame and Zguba jama proles must be older due to slig h t to distinct counter-clockwise rota tion. Palaeomagnetic directions of t h e Stara jama prole indicate clockwise rotation since prole was deposited. e inclination in N polarized samples from Spodnji Tartarus Sout h and Umetni tunel I proles is anomalously low. W e t h erefore interpreted most of t h e sediments as being younger t h an 0.78 Ma, belonging to dierent depo sitional events wit hin t h e Brun h es c hron. Nevert h eless, t h e N polarization in some proles can be linked wit h N polarized subc hrons older t h an 0.78 Ma, as in t h e Umet ni tunel 1 site and Zguba jama. e lit h ological situation in Male jame is questionable. Sediments in Umetni tunel 1 are t h e oldest in t h e system (below t h e gravel wit h co loured c h ert) and were not included in older stratigrap hic sc h emes. ey may be correlated wit h Olduvai, Reunion or even older c hrons (i.e. from 1.77 to over 2.15 Ma). Data from t h e Kras as a w h ole, suggests t h at t h e cave ll in t h e region is unlikely to be muc h older t h an t h e Plio cene (in t h e traditional sense). Deposition in t h e Posto jnska jama Planinska jama cave system can be placed in two principal deposition periods: (1) from about 0.78 Ma (palaeomagnetic age) to more t h an 4.0 Ma (palaeo magnetic age) Pliocene to Pleistocene (Gnz/Mindel) t his group contains a suc cession of detected ages: (a) more t h an 0.78 up to about 3.58 Ma (palaeomagnetic ages), and (b) less t h an 0.78 to about 2 Ma (palaeomag netic ages), and (2) from 0.78 Ma Pleistocene (Mindel) to Holocene. e dating suggests a prolonged evolution of t h e Pivka kotlina Postojn ska jama Planinska jama Planinsko polje system in relatively stable h ydrologi cal conditions related to t h e function of Planinsko polje. Hydraulic stability and a low h ydraulic h ead a for a long period of time led to t h e for mation of a long and complex cave system wit h t hree cur rently accessible cave levels in t h e ponor area, t h e middle level being t h e most evolved, mostly in epip hreatic zone. P hoto plate I: Examples of uvial sediments and speleothems sampled for dating; part 1 P ostojn ska jama. A Spodnji Tartarus North prole. e lower part consists of red clay with some yellow laminae; the upper part consists of yellow clay and some sandy layers; B Spodnji Tartarus South prole; C Small stalagmite on rotated collapse block was dated by /U method, Spodnji Tartarus South; D Umetni tunel I prole; E Red clay from the upper part of the B iospeleoloka postaja prole. N ADJA ZUPAN HAJNA, P ETR PRUNER, A NDREJ MIHEVC, P ETR SCHNABL & P AVEL BOSK

PAGE 21

ACTA CARSOLOGICA 37/1 2008 83 P hoto plate II: Examples of uvial sediments and speleothems sampled for dating; part 2 P osto jnska and P laninska jama. A, B M ale jame prole; C, D Stara jama prole; E, F P role at the end of P isani rov; G Zguba jama prole; H P role in Rudolfov rov, P laninska jama. e proposed model of prolonged evolution of t h e cave system is based on (1) all numericaland correla tive-ages (dates over 530 ka from speleot h ems and up to 3.58 Ma from cave sediments); (2) t h e morp h ology of Planinsko and Cerkniko poljes, w hic h present limits are far be hind marginal faults of t h e Idrija fault zone; t h eir lateral enlargement and bot tom planation by corrosion to t h e present extent needed prolonged stabilization of t h an karst water table t h an earlier proposed 30 to 100 ka, and (3) dynamics of ll ing and erosion p h ases in t h e system, w h ere several depositional and erosion events/p h ases alternated es pecially in epip hreatic evolu tion p h ase. Individual cave segments or passages of t h e system were fully lled and exh umed several times dur ing t h e cave evolution. e deposition was not uniform t hroug h out t h e entire cave at t h e same time. ere was erosion in one part of t h e cave and deposition in an ot h er. e alternation of dep ositional and erosion p h ases may be connected wit h c h anging conditions wit hin t h e cave system, functions bot h of t h e catc hment ba sin and t h e resurgence area, climatic c h anges, tectonic movements, collapses, and t h e intrinsic mec h anisms of t h e contact karst. e petrologic, pal aeomagnetic and numerical data do not allow t h e con struction of facies time-de pendent models because (1) t h ere are very few sequences of sediment suitable for pal aeomagnetic analysis in t h e cave system and most h ave been examined h ere, (2) most of t h e system was in epip hreatic conditions wit h a low h ydraulic h ead for a long period of time causing multiple re-deposition and reworking and (3) t h ere is little dierence in elevation between t h e proles mak ing stratigrap hic correlation extremely dicult wit h out additional numerical or palaeontological dating. CAVE SEDIMENTS FROM THE POSTOJNSKAPLANINSKA CAVE S Y STEM SLOVENIA: EVIDENCE OF MULTIPHASE ...

PAGE 22

ACTA CARSOLOGICA 37/1 2008 84 W e are particularly grateful to t h e management of Pos tojnska jama, tourism, d.d. Turizem Kras, destinacijski management, d.d. (Postojna, Slovenia), especially to Mr. Ivan Vekar, Mr. Tomo e h, Mr. Matja Beron, Mr. Pe ter ten and C hief Guides: Mr. Alojz ernigoj and Mr. Bogdan Debevc for permission to work in t h e cave sys tem. W e acknowledge eld assistance of t h e tec hnical sta of t h e IZRK ZRC SAZU (especially Jurij Hajna and Franjo Drole). Palaeomagnetic analyses were performed by Dr. Daniela Ven h odov, Mrs. Jana Dra h otov, Dr. Martin C h adima, Mgr. Stanislav lec h ta and Mr. Jit Petrek (IG AS CR). Soware for evaluation of palaeo magnetic measurements was prepared by Dr. Ota Man and Dr. Martin C h adima (IG AS CR). Prof. Dr. Ivan Horek (Department of Zoology, Faculty of Science, C h arles University, Pra h a, Czec h Republic) carried out palaeontological determinations. Some of t h e draw ings were produced by Mrs. Jana Rajlic h ov (IG AS CR) and Mr. Jurij Hajna (IZRK ZRC SAZU). e aut h ors ac knowledge critical reviews by one anonymous reviewer, Dr. Gregory S. Springer and DR. R.A.L. Osborne. Travel costs were covered wit hin t h e frame of t h e program KONTAKT of t h e Ministry of Education, Y out h and Sports of t h e Czec h Republic and Slovenian Ministry of Education, Sport and Science Nos. 13-2005-06 (2005 2006) and 9-06-19 (2007). Analyses, processing and interpretation in t h e Czec h Republic (PP, PB) were carried out wit hin Researc h Plan of t h e IG AS CR No. CEZ AV0Z30130516, and Grant of t h e Grant Agency of t h e Academy of Sciences of t h e Czec h Republic No. IAA30013001. Researc h activities in Slovenia (NZH, AM) were covered by researc h programs of t h e Ministry of Science of Slovenia and Slovenian Researc h Agency Nos. P6 and P0, and project No. J6 6345 (2004). ACKNO W LEDGEMENTS Bella P., Bosk P., Pruner P., Hoc hmut h Z. & Hercman H., 2007: Magnetostratigraa jaskynnc h sedimen tov a speleogenza Moldavskej jaskyne a spodnc h ast Jasovskej jaskyne.Slovensk kras, XLV, 15-42. Liptovsk Mikul. 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 Cessation, Carsologica, 191-223, Zaloba ZRC, PostojnaLju bljana. Bosk, P., Pruner, P. & Kadlec, J., 2003: Magnetostratig rap h y of Cave Sediments: Application and Limits.Studia Geop h ysica et Geodaetica, 47, 2, 301-330. Brodar, S., 1966: Pleistocenski sedimenti in palaeolitska najdia v Postojnski jami.Acta carsologica, 4, 57138, Ljubljana. Brodar, M., 1969: Nove palaeolitske najdbe v Postojnski jami.Ar h eoloki vestnik, 20, 141-144, Ljubljana. Buser, S., Grad, K. & Pleniar, M., 1967: Osnovna geoloka karta SFRJ, list Postojna, 1 : 100 000.Zvez ni geoloki zavod Beograd, Beograd. Cande, S.C. & Kent, D.V., 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. ar, J. & Gospodari, R., 1984: O geologiji krasa med Postojno, Planino in Cerknico.Acta carsologica, 12(1983/1984), 91-105, Ljubljana. Fis h er, R., 1953: Dispersion on a sp h ere.Proceedings of t h e Royal Society, A 217, 295-305. Ford, D.C. & W illiams, P.W ., 2007: Karst Hydrology and Geomorp h ology.W iley, 562 pp., C hic h ester. Franke, H. & Gey h, M., 1971: 14 C datierungen von Kalk sinter aus slowenisc h en H h len.Der Aufsc h luss, 22, 7-8, 235-237. Franke, H. & Gey h, M., 1976: Zur Datierung vor Ver sturzereignissesn.Proceedeng of t h e 6t h Interna tional Congress of Speleology, OlomoucSSR, III, 95-100, Pra h a. Gams, I., 1965: K kvartarni geomorfogenezi ozemlja med Postojnskim, Planinskim in Cerknikim poljem.Geografski vestnik, 37, 61-101, Ljubljana. Gorka, P. & Hercman, H., 2002: URANOTHOR v. 2.5. Delp hi Code of calculation program and user guide.Arc hive, Quaternary Geology Department, Institute of Geological Sciences, Polis h Academy of Sciences, W arsaw. Gospodari, R., 1963: K poznavanju Postojnske jame Pisani rov.Nae jame, 4 (1962), 9-16. Gospodari, R., 1964: Sledovi tektonski h premikov iz ledene dobe v Postojnski jami.Nae jame, 5 (1963), 5-11, Ljubljana. REFERENCES N ADJA ZUPAN HAJNA, P ETR PRUNER, A NDREJ MIHEVC, P ETR SCHNABL & P AVEL BOSK

PAGE 23

ACTA CARSOLOGICA 37/1 2008 85 Gospodari, R., 1972: Prvi podatki o absolutni starosti sige v Postojnski jami na podlagi 14 C.Nae jame, 13, 91-98, Ljubljana. Gospodari, R., 1976: Razvoj jam med Pivko kotlino in Planinskim poljem v kvartarju.Acta carsologica, 7, 5-139, Ljubljana. Gospodari, R. 1977a: Generacije sig in jamski h sedi mentov v kraki h jama h Slovenije.Unpublis h ed researc h report, 1-37, 6 appendices, IZRK SAZU, Postojna. Gospodari, R. 1977b: Generacije sige v nekateri h kraki h jama h Slovenije.Unpublis h ed researc h report, 131, 4 appendices, IZRK SAZU, Postojna. Gospodari, R., 1981: Generacje sig v klasinem krasu Slovenije.Acta carsologica, 9 (1980), 90-110, Lju bljana. Gospodari, R., 1988: Palaeoclimatic record of cave sedi ments from Postojna karst.Annales de la Socit Gologique de Belgique, 111, 91-95. Habi, P., 1982: Kraki relief in tektonika.Acta carso logica, 4, 23-43, Ljubljana. Horek, I., Mi h evc, A., Zupan Hajna, N., Pruner, P. & Bosk, P., 2007: Fossil vertebrates and palaeomag netism update one of t h e earlier stages of cave evo lution in t h e Classical Karst, Slovenia: Pliocene of rnotie II site and Raika peina.Acta carsologica, 37/3, 451-466, Ljubljana. Ikeya, M., Miki, T. & Gospodari, R., 1983: ESR Dating of Postojna Cave Stalactite.Acta carsologica, 11 (1982), 117-130, Ljubljana. Ivanovic h, M. & Harmon, R.S. (Eds.), 1992: Uranium se ries disequilibrium. Applications to environmental problems. 2 nd Ed.Clarendon, 910 pp., Oxford. Jelnek, V., 1966: A hig h sensitivity spinner magnetom eter.Studia Geop h ysica et Geodaetica, 10, 58-78. Jelnek, V., 1973: Precision A.C. bridge set for measuring magnetic susceptibility and its anisotropy.Studia Geop h ysica et Geodaetica, 17, 36-48. 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 Society, 62, 699-718. Lauritzen, S-E., 1993: Age4U2U. Program for reading ADCAM energy spectra, integration, peak-correc tion and calculation of 230 / 234 U ages.Arc hive, Computer program, Bergen University. Mi h evc, A., 1990: Nekatere morfoloke znailnosti kon taktnega krasa : Pivka kotlina.In: Natek K. (Ed.): Geomorfologija in geoekologija. Zbornik referatov 5. znanstvenega posvetovanja geomorfologov Jugo slavije, Krko. Znanstvenoraziskovalni center SAZU, 253-258, Ljubljana. Mi h evc, A., 1996: Brezstropa jama pri Povirju.Nae jame, 38, 92-101, Ljubljana. Mi h evc, A., 2001: Speleogeneza Divakega krasa. Zbirka ZRC, 27, 180 pp., Ljubljana. Mi h evc, A., 2002: Postojnska jama cave system, U/ datation of t h e collapse processes on Velika Gora (Point 4).In: Gabrovek F. (Ed.): Programme and guide booklet for t h e excursions: Evolution of Karst: from Prekarst to Cessation, September, 17t hst, 2002. Postojna, 14-15, Karst Researc h Institute ZRC SAZU, Postojna. Mi h evc, A., 2004: Dating of t h e cave sediments wit h a relative geomorp hic dating met h od case studies in Slovenia.12 t h International Karstological Sc h ool Classical Karst Dating of cave sediments, Guidebooklet for t h e excursions and abstracts of presenta tions, 57-58, Postojna. Mi h evc A., 2007: e age of karst relief in W est Slovenia.Acta carsologica, 36/1, 35-44, Ljubljana. Mi h evc, A., 2008: O starosti kapnikov in podoru Kalvarije v Postojnski jami.Kras, april 2008, no. 22, 52-54. Mulec, J., Mi h evc, A. & Pipan, T., 2005: Intermittent lakes in t h e Pivka Basin.Acta carsologica, 34/3, 543-565, Ljubljana. Osborne, R.A.L., 1984: Lateral facies c h anges, unconfor mities and stratigrap hic reversals: t h eir signicance for cave sediment stratigrap h y.Cave Science, 11, 3, 175-184. Placer, L., 1981: Geoloka zgradba jugoza h odne Sloveni je.Geologija, 24/1, 27-60, Ljubljana. Placer, L., 1996: O zgradbi Sovia nad Postojno.Ge ologija, 37/38 (1994/95), 551-560, Ljubljana. 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(1998), 223-255, Ljubljana. Pt h oda, K., Krs, M., Peina, B. & Bl h a, J., 1989: MAVACS a new system of creating a non-magnetic environ ment for palaeomagnetic studies.Cuadernos de Geologa Ibrica, 12, 223-250. Pruner, P., Bosk, P., Zupan Hajna, N., Mi h evc, A., Man, O. & Sc hnabl, P., 2004: Palaeomagnetic researc h of cave sediments in Krina jama, Planinska jama and Postojnska jama: preliminary results.12 t h Interna tional Karstological Sc h ool, Classical Karst Dat ing of Cave Sediments, Postojna. Guide booklet for t h e excursions and abstracts of presentations, 61-62, Postojna. Rakovec, I., 1954: Podvodni konj iz Pivke kotline.Raz prave 4. razreda SAZU, 2, 297-317, Ljubljana. Rinar, I., 1997: Geology of Postojna area.MSc. esis, NTF, University of Ljubljana, 78 pp., Ljubljana. CAVE SEDIMENTS FROM THE POSTOJNSKAPLANINSKA CAVE S Y STEM SLOVENIA: EVIDENCE OF MULTIPHASE ...

PAGE 24

ACTA CARSOLOGICA 37/1 2008 86 Sasowsky, I.D., 2007: Clastic sediments in caves imper fect recorders of processes in karst.Acta carsologi ca, 36/1, 143-149, Ljubljana. Sasowsky, I.D., ebela, S. & Harbert, W ., 2003: Concur rent tectonism and aquifer evolution >100,000 years recorded in cave sediments, Dinaric karst, Slove nia.Environmental Geology, 44, 1, 8-13. S h aw, T., 1992: e History of Cave Science, t h e Explora tion and Study of Limestone Caves, to 1900. Sydney Speleological Soc., 338 pp., Broadway. Stepinik, U., 2004: e origin of sediments inside t h e collapse dolines of Postojna karst (Slovenia).Acta carsologica, 33/1, 237-244, Ljubljana.. ebela, S., 1994: e caves Jama na poti in Zguba jama.In: Kranjc A. (Ed.): Proceedings of 1 st International Karstological Sc h ool Classical Karst, Lipica, Sep tember 20, 1993 and La table ronde internatio nale E. A. Martel et le karst Slovene (1893), Postojna, 12 Novembre, 1993, IZRC ZRC SAZU, 234-243, Postojna. ebela, S., 2008: Tektonska zgradba sistema Postojnski h jam.Kras, april 2008, no. 22, 44-45. ebela, S. & Sasowsky, I., 1999: Age and magnetism of cave sediments from Postojnska jama cave system and Planinska jama Cave, Slovenia.Acta carsologi ca, 28/2, 18, 293-305, Ljubljana. uteri, F., uteri, S. & Stepinik, U., 2003: e Late Quaternary dynamics of Planinska jama, sout h-cen tral Slovenia.Cave and Karst Science, 30, 2, 89-96. Vrabec, M., 1994: Some t h oug h on t h e pull-apart pri gin of karst poljes along t h e Idrija strike-slip fault zone in Slovenia.Acta carsologica, 23, 158-168, Ljubljana. Zupan, N., 1991: Flowstone datations in Slovenia.Acta carsologica, 20, 187-204, Ljubljana. Zupan Hajna, N., 1992: Mineralna sestava me h anki h sedimentov in nekateri h delov slovenskega krasa.Acta carsologica, 21, 115-130, Ljubljana. Zupan Hajna, N., 1998: Mineral composition of clastic cave sediments and determination of t h eir origin.Kras i speleologia, 9(XVIII), 169-178, Katowice. Zupan Hajna, N., 1996: e valuation of absolute spe leot h em dating from Slovenia.In: Lauritzen, S.-E. (Ed.). Climate c h ange: t h e Karst record: extended abstracts of a conference h eld at Department of ge ology University of Bergen, Norway, August 1-4t h 1996, (Special Publication, 2). C h arles Town: Karst W aters Institute, 185-188. Zupan Hajna, N., 1998: Mineral composition of clastic cave sediments and determination of t h eir origin.Kras i speleologia, 9(XVIII), 169-178, Katowice. Zupan Hajna, N., Mi h evc, A., Pruner, P. & Bosk, P., 2008: Palaeomagnetism and Magnetostratigrap h y of Karst Sediments in Slovenia.Carsologica 8, Zaloba ZRC, 266 pp., Ljubljana. Zupan Hajna, N., Mi h evc, A., Pruner, P. & Bosk, P., 2008b: Palaeomagnetne datacije uvialni h sedi mentov iz Postojnske jame, Zgube jame in Planin ske jame.Kras, april 2008, no. 22, 46-51. N ADJA ZUPAN HAJNA, P ETR PRUNER, A NDREJ MIHEVC, P ETR SCHNABL & P AVEL BOSK



PAGE 1

KARST IN FRANCE AND UNESCO W ORLD HERITAGE KRAS V FRANCIJI IN SVETOVNA NARAVNA DEDI INA UNESCO Patric CABROL 1 & Alain MANGIN 2 Izvleek UDK 551.44(44) Patric Cabrol & Alain Mangin: Kras v Franciji in svetovna naravna dediina U NESCO Kras pokriva tretjino ozemlja Francije. Ker so karbonatne kamnine, ki so bile odloene od kambrija dalje, danes kot pla note ali pa tektonsko razgibana ozemlja, je francoski kras znan po svoji pestrosti. Znane so obsene krake pokrajine, kot so Les Grands Causses ali soteski rek Tarn in Verdon. Prav tako so znani kraki izviri, npr. Fontaine de Vaucluse ali izvir reke Lez. Kraki sistemi so la h ko zelo veliki: est hidroloki h siste mov Pierre Saint-Martin meri 400 km, v gori Arbas pa 104 km. Najveje globine, Pierre Saint-Martin, jami Berger in Jean Bernard dosegajo rekorde, nedavno tega je bila v jami Mirolda doseena globina 1733 m. Raziskave v zadnji h sto leti h so od krile pomembno jamsko favno. Ni samo ve kot 160 jam s pa leolitsko jamsko umetnostjo najpomembneja dediina fran coskega krasa, ampak tudi tevilne zasigane jame. Poudariti je treba tudi redek pojav: zaganjalka Fontestorbes v departmaju Arige je najpopolneje in najbolj zvezno opazovan tak izvir na svetu. Veliko tevilo izredni h pojavov v jama h za h teva pripravo preuevanja in programov za ustrezno o hranjanje te dediine bodoim rodovom. Kljune besede: kras, naravna dediina, Francija. 1 DIREN Midi-Pyrnes, 1 rue Delpec h, Immeuble Couderc, 31 000 Toulouse, France, patrick.cabrol@developpement-durable.gouv.fr 2 32, Lotissement des Noyers, F 09200 Montjoie en Couserans, France Received/Prejeto: 04.09.2007 COBISS: 1.02 ACTA CARSOLOGICA 37/1, 87-93, POSTOJNA 2008 Abstract UDC 551.44(44) Patric Cabrol & Alain Mangin: Karst in France and U NESCO World Heritage Frenc h Karst covers t h e t hird of t h e territory. Because calcare ous deposits date back to t h e Cambrian and are now located in tabular region as well as in tectonic areas, Frenc h karst is famous for its diversity. Large karst landscapes are well known, suc h as Les Grands Causses t h e Gorges du Tarn and t h e Gorges du V erdon Resurgences suc h as t h e Vaucluse Fountain and t h e spring of t h e Lez are very famous too. Karstic units can s h ow very large dimensions: 400 km for t h e six h ydrological net works of t h e Pierre Saint-Martin, 104 km for t h e Arbas Moun tain. Maximum dept h s can reac h hig h records, at Pierre Saint Martin, Berger Cave, Jean Bernard Cave and recently Mirolda Cave was acknowledged to reac h 1733-meter dept h. Studies proceeded during t h e last century h ave revealed an im portant cave fauna. Considering t h e h eritage value of Frenc h karst, t h e key point is not only t h e presence of more t h an 160 decorated caves from Palaeolit hic period but also t h e number of caves s h owing concretions. Moreover, a rare p h enomenon h as to be underlined: in Ariege, t h e Intermittent Fontestorbes Spring is t h e most regularly and completely studied resurgence in t h e world. is amount of exceptional features for t h ese caves commands to organise studies and programmes to estab lis h conditions of a good conservation of t his h eritage for t h e next generations. Key words: karst, natural h eritage, France.

PAGE 2

ACTA CARSOLOGICA 37/1 2008 88 In France, 30% of t h e land is covered wit h calcareous for mations more or less evolved to karst formations. About 100 000 caves are explored by about 20 000 speleologists; 7000 to 8000 of t h em belong to t h e F dration F ranaise de Splologie and about 800 to t h e Club Alpin F ranais e distribution of t h ese caves is very unequal. Some ar eas are very well supplied like Hrault for example wit h 4000 caves w h ereas ot h ers like Ctes dArmor (Brittany) s h ow no cave at all. It h as been a tradition for Frenc h scientists to be involved in t h e eld of t h e underground studies. From t h e middle of t h e XIX t h century up to now, t his researc h h as been very active, and it is strongly organised t hroug h many laboratories linked or belonging to t h e CNRS. e rst laboratory in Europe devoted to karst studies was t h e Frenc h Laboratoire Souterrain de M oulis created in Arige. Many national and international working groups h ave been active in various elds related to karst stud ies: cave formation, c h aracteristics of karstic landscapes, h ydrology in calcareous regions, cave sediments studies, concretions studies, biospeleology, geoc h emistry, and so on Famous scientists w h o le strong marks on mod ern knowledge of karst evolution made t his long history. Among t h em, Delarouzee is famous for h e initiated bio speleology in t h e XIXt h century. Consecutive progress were made during t h e same century and t h e beginning of t h e XXt h, wit h Lucante, A. Vir, E Racovitza, Jeannel and essentially E.A. Martel w h ose name is undoubtedly linked to modern knowledge of karst. Among explorers we rst h ave to mention Martel but also Fournier, G. de Lavaur, R. de Joly, F Trombe and Norbert Casteret. Even if t h e list is not exh austive, we h ave to mention A Cavaill, J Corbel, B. Gze, J. Nicod, P. Renault, A. Van del, F. Trombe, as scientists, but it is essentially aer t h e Second W orld W ar t h at t h e Frenc h researc h was t h e most ecient, due to t h e results produced in t h e Laboratoire Souterrain de M oulis and t h e team of geograp h ers from Aix en Provence. Today, various researc h programs are in progress t hroug h many teams from t h e CNRS and universities. THE ECONOMIC IMPORTANCE AND THE VARIET Y OF THE FRENCH KARST. INTRODUCTION Frenc h karst is hig h ly important regarding economic, patrimonial, educative and sportive aspects. E CONOMIC INTEREST OF THE F RENCH KARST 107 caves are open for tourism and 4.000.000 of visi tors are registered every year. (Padirac, Bt h aram, Orgnac,Clamouse, Armand, Demoiselles, C h oranc h e...). Many large cities only use water provided by karst: Montpellier wit h t h e Spring in Lez, t h e F ontaine de Nmes Besanon and so on Some ores suc h as lead or zinc are trapped in karst: see for instance t h e Malines Mine (Gard). T HE VARIET Y OF THE F RENCH KARST e Frenc h karst is extremely varied depending on t h e nature of t h e calcareous formations w h ere it h as devel oped: pure lime stones but also dolomites and dolomitic formations, argillaceous lime stones and marls, and so on. ey appear as t hin layers or as massive beds. Some w h ere, t h ey can be tabular but somew h ere else, t h ey can be strongly tectonised. Some of t h em are enric h ed wit h various metals suc h as lead, zinc, copper, barium and nickel, as for t h e sout h ern edge of t h e M ontagne Noire e Frenc h karst also develops under dierent cli matic conditions. It varies from Mediterranean to Atlan tic climate, and from h umid Atlantic to Continental and even up to semi-arid Mediterranean climate. Most of t h e Frenc h karst developed during t h e Quaternary Period but in some areas suc h as Quercy p h osp h ates, t h e karst formation dates back to t h e Tertiary Period. A beautiful tertiary fossil fauna is very well preserved in t his Quercy formation. A LARGE PATTERN OF POTENTIAL VARIETIES Because of t h e extreme diversity of its environment, t h e Frenc h karst s h ows a large pattern of potential varieties. 1) Large karstic landscapes: ese large karstic landscapes made t his country P ATRIC CABROL & A LAIN MANGIN

PAGE 3

ACTA CARSOLOGICA 37/1 2008 89 very famous. First, we h ave to mention t h e Grands Causs es (Causse du Larzac, Causse Noir, Causse Mjean), t h e Gorges du Tarn and la Jonte t h e Gorges du V erdon t h e Gorges de lArdche t h e Dsert du P lat les Arres dAnnie, la P ierre Saint M artin, t h e M assif du V ercors, t h e Jura Mountains wit h famous recules suc h as la Loue, le M as sif de la Chartreuse and so on 2) Important resurgences: e most famous in t h e world is probably t h e F on taine de V aucluse but ot h er ones s h ow very hig h ows too suc h as t h e Lez Spring in Montpellier, la F ontaine de Nmes t h e Loue Spring (Doubs), t h e Girardes in Dvoluy Area, t h e Touvre in Angoulme, t h e C h artreux in Ca h ors, t h e B uge in t h e Sranne Port-Miou w hic h ows down into t h e Mediterranean Sea in t h e Calanques de Cassis near Marseille... 3) Large hydrological systems: Exceptional developments of galeries can be seen at t h e Pierre Martin, w h ere six h ydrological systems h ave a total of 400-km long gallery, t h e Massif dArbas wit h 104 km of galeries and wells, 45 entrances and 1004 m of dierence in level, t h e Alpe system wit h 60 km, t h e Granier system (Isre) wit h 55 km, t h e Dent de Crolles wit h 50 km, t h e Verneau system wit h more t h an 33 km of galeries, Saint Marcel dArdc h e wit h 27 km, t h e Andr Lac h ambre system in t h e Pyrnes Atlantiques s h ows 25 km of galeries wit h very beautiful mineralization of cal cite, aragonite and h ydromagnesite. e Cabrespine Cave covers 18 km and so on. To give a comparison, t h e most developed system in t h e world is t h e Mammot h Cave System (Kentucky, USA), w hic h s h ows about 557 km of galleries. 4) Large Depths: In France, large calcareous formations exist at very deep levels. is h as initiated t h e researc h of a world re cord of dept h: Mirolda Cave (Haute Savoie) is 1733 me tre-deep, Jean Bernard System (Haute Savoie) is 1602 m, Pierre Saint Martin (Pyrnes Atlantiques) is 1342 m, and Berger Cave (Isre) is 1271m. 5) Important fauna: e long history of t h e Frenc h karstic fauna h as in terested a lot of biologists. ey discovered and identied 639 invertebrate species. Some caves located around t h e Laboratoire Souterrain de M oulis h ave been intensively investigated (Sainte Cat h erine Cave, Liqu Cave, etc). Hautecourt Cave (Ain) may also be mentioned for t h e study of underground fauna and its relation wit h envi ronment. For about twenty years, t h e studies for h ydro logical system fauna h as been developing: we can men tion t h e Baget System in Arige (Raymond Rouc h from t h e laboratoire de Moulis), t h e Lez Resurgence in Hrault (Florian Malard from t h e university of Lyon), t h e Sauve Spring in Gard (C. Jubert hie et R Rouc h), t h e Anglars System in Lot (C. Bou, Danielopol and R Rouc h), t h e Torcien System in Ain (G. Gibert), t h e Arbas System in Haute Garonne (F Lesc h er-Moutone and N Gourbault). Because of t h e size of t h e systems still to observe, t h ese studies are just beginning. 6) e French karst particularities: Palaeolithic age artefacts, concretions and Quercy phosphate ores. An exceptional phenomenon: Fontestorbes, an intermit tent spring Large karstic formations can be found everyw h ere in t h e world (C hina, Europe, Madagascar, USA). In some countries, karst is particularly developed as it is in Slove nia for instance. Extreme dept h s are also reac h ed in Cau casus Karst, but t h e feature t h at makes t h e Frenc h karst exceptional, is t h e abundance and value of mixed arc h ae ological, mineralogical and paleontological remains, and last but not least, t h e exceptional natural p h enomenon of t h e F ontestorbes disturbed spring (literal translation of t h e name of t his intermittent spring) in Ariege. 6.1 Archaeological heritage: More t h an 160 caves are ornamented wit h engrav ings, paintings or drawings from t h e Palaeolit hic period. From t his point of view, France is in possession of an exceptional h eritage, contemporaneous of t h e origins of Art, wit h fabulous caves suc h as: C h auvet, Lascaux, Cos quer, Niaux, Pec h Merle. 6.2 M ineralogical heritage: Frenc h caves s h ow magnicent concretions of great interest. Large caves opened for tourism like Clamouse Cave, C h oranc h e Cave, Demoiselles Cave, Esparros Abyss, Aven Armand and Aven dOrgnac s h ow a great di versity of colours and s h apes t h at make t h em very attrac tive to tourists. Some caves closed to visitors contain rare if not unique concretions, suc h as mus hrooms from Lauzinas Cave, pastis distributor of Cabrespine Cave, blue needles of aragonite of Asperge Cave, cymbals of TM 71 Cave. In some cases, t h e abundance of mono type concretions is quite exceptional suc h as aragonite and h ydromagnesite from t h e Rseau Lachambre soda straws of Amlineau Cave, and t h e aragonite c h ande lier of Limousis Cave. Some concretions are exceptional for t h eir size: it is t h e case at Aven dOrgnac, at Demoi selles Cave, at Clamouse Cave. Quality, diversity and abundance of t h ese concre tions in Frenc h caves h ave justied studies on t h eir ori gins and also on t h e way we can protect t h em (Patrick KARST IN FRANCE AND UNESCO W ORLD HERITAGE

PAGE 4

ACTA CARSOLOGICA 37/1 2008 90 Cabrol and Alain Mangin essentially). ese studies h ave s h own t h at beyond t h eir est h etical aspect, concre tions are pure scientic objects, w hic h allow to study t h e mec h anism of inltrating zone in karst areas. Moreover, concretions are paleoenvironmental arc hives of great precision for t h e past h alf million years. Concretions are better arc hives t h an polar ice, because of t h e c hronology precision allowed by /U disequilibria dating met h od. 6.3 P aleontological heritage: Bones from animals of Quaternary period are very abundant in France, sometimes exceptionally numerous as in Aldne Cave (Hrault) w h ere more t h an 200 000 cave bears are fossilised and make an important p h os p h ate bed w hic h was exploited 50 years ago. Elsew h ere, we h ave to mention t h e variety of species, suc h as in C h a tillon-Saint-Jean (Isre) for instance. W e especially h ave to mention t h e p h osp h ate ores from Quercy. e w h ole evolution of vertebrate fauna during t h e 30 millions of years of t h e Tertiary Period is gat h ered t h ere on a pretty small area (20 x 50 km). 60 million years ago, calcareous formations were present in t h ese regions of Lot and Tarn-et-Garonne. Karstication produced many caves w h ere animals fell down (Avens, Abysses). eir remains are fossilised in very t hin grain sediments caused by dissolution of soils formed under tropical climate of t his period. During 30 million years caves played t his function of reservoir for t h ese fossils wit h very good preservation conditions. Scientic results obtained on t his fauna deal wit h various researc h elds. More t h an 250 papers and 15 P hD were publis h ed since 1960. Beyond t h e geological interest, Quercy p h osp h ate ores s h ow a great interest for mine exploitation tec h niques and are a witness of h uman activity from 1870 to 1905. Quercy p h osp h ate ores s h ow a scientic and h eri tage interest t h at is unique in t h e world for t his period. A question can be asked: Must karstic sites known for pre historical arc hives be always excluded or not? (Ar c h aeological relics, ornamented caves etc)? 6.4 An exceptional phenomenon: intermittent F onstestorbes Spring in Arige Fontestorbes intermittent resurgence is located near Belesta (Arige). It s h ows a rare cyclic variation of ow during low water periods. Only t hirty cases are described in t h e w h ole world but t his one is truly exceptional be cause of t h e large ow variation wit h a minima of 20 li tres per second up to a maxima of 1800 litres per second following a very regular periodicity of about an h our. Located near a road it makes an attractive site for tourism in t h e area. Speleological expeditions h ave led to discover on t h e upstream, t h e sub ground river t h at provides water to t h e intermittent p h enomenon. Two places in t h e cave s h ow t h e cyclic ascending and decreas ing movements of water, wit h amplitude of 4.5 meters and 5.5 meters. is resurgence is already protected by registration in t h e national site book (october t h e 10t h 1921) and is included in t h e project of a protected area for Ariege. 7 Hydrological systems and caves very much studied: Frenc h karst is so ric h and various t h at it h as justi ed numerous fundamental researc h es and it is now dif cult to present an exh austive list of t h ese studies. Among recent studies, we can mention experimental studies of karstic Baget System and t h e one of t h e F ontaine de F ontestorbes in Arige, t h e Lez Spring in Montpellier, t h e F ontaine de V aucluse in Vaucluse, t h e Causse du Larzac and Causse de Sauveterre in Aveyron and Hrault, t h e C h amplive Basin in t h e Jura Mountains, t h e Touvre in C h arente. Regarding formation and dating of concretions and sediments in t h e caves, we h ave to talk about t h e stud ies made in Clamouse Cave, t h e Devze, Pont de Ratz (Hrault), C h oranc h e (Isre), Orgnac Aven (Ardc h e), Niaux Cave (Arige), Villars Cave (Dordogne) and C h auvet Cave (Ardc h e). e rst environmental study before opening a cave for tourism was performed in 1987 at Esparros Cave in Hautes-Pyrnes and t his operation s h ows h ow neces sary it would be to extend t his kind of study. e sad ex perience of Lascaux Cave reminds us t h at h eavy mistakes can h appen. is is t h e reason w h y we h ave made copies of t his cave (Lascaux) t h at t h e public can visit and we are about to do so wit h t h e C h auvet cave and ot h er ones in t h e future. Some environmental studies were carried out on caves t h at are now used as examples to follow: it is t h e case of t h e famous C h auvet Cave, Esparros Cave w h ere studies h ave been going on for ten years. Ot h ers cases to be mentioned are Niaux Cave, Cabrespine Cave, Gar gas Cave, Pec h Merle Cave, Orgnac Aven, Proumeyssac Cave, Bedeil h ac Cave and so on. ese works h ave led to build a database, considered as a key for t h e understand ing of natural p h enomenon and t h e preservation of t h e h eritage in t h ese caves. P ATRIC CABROL & A LAIN MANGIN

PAGE 5

ACTA CARSOLOGICA 37/1 2008 91 KARST IN FRANCE AND UNESCO W ORLD HERITAGE

PAGE 6

ACTA CARSOLOGICA 37/1 2008 92 Regarding t h e importance of karst in France, for t h e sur face concerned as well as for t h e h eritage value, it is nec essary to establis h and publis h for public community an inventory of major karstic sites in France. Pierre C h auve organised t h e work of a team wit h dierent specialists of t h e Frenc h karst. ey produced a rst inventory and a collective book is sc h eduled for 2005. THE FIRST INVENTOR Y OF THE IMPORTANT KARSTIC SITES IN FRANCE. SELECTION CRITERIA FOR SITES W ITH A HIGH PATRIMONIAL VALUE: Aer t his rst inventory, it is necessary to t hink about criteria t h at could select sites t h at oer t h e best h eritage value in order to register t h em on t h e W orld Heritage list of t h e UNESCO, on t h e Ramsar Sites list, MAB, or ot h er records t h at provide sites wit h an international acknowl edgement. Among the criteria we can have: Geology. e meaning of t h e word h as to include karstic feature but also geomorp h ology, h ydrology, min eralogy Inventory of knowledge. Sites h ave to be well known t hroug h scientic studies. If t h e knowledge of a site is limited, t h en it is dicult to select it because it is impossible to argue in favour of its scientic interest. is important criterion, of course does not imply, t h at a well-known site h as to be selected as an international valuable site. Environmental state. If polluted or degraded by h eavy use, a site cannot be selected. Attractiveness for tourism. If accessibility is di cult or limited to specialists, and t h en prevents access to c hildren, students and people, t h ere is a problem. If so, it would be necessary to nd a way to make t h e site acces sible wit h a reinforced protection t h at allows its use by a large public. Interest of selected sites. selected sites h ave to s h ow a remarkable interest and gat h er a full knowledge of a karst feature. e number of aspects presented can be limited. e selection of a few sites t h at s h ow a particular aspect of karst features can be of a great interest if various sites complete eac h ot h er. Management of selected sites. If t h e sites are regis tered on an international reference list, t h en t h eir man agement control will be under t h e National or t h e local aut h orities, t h at can oer important means of conser vation, w hic h is precisely one of t h e main targets of t h e UNESCO W orld Heritage. MANAGEMENT AND PROTECTION OF GEOLOGICAL KARSTIC HERITAGE W ITH AN INTERNATIONAL INTEREST. Several texts of an international level are used for t h e protection of t h e geological h eritage. e International Recommendations on Karst Preservation. e International Union for Nature Preserva tion h as publis h ed in 1997 t h e Guidelines for Cave and Karst Protection w hic h proposes 31 recommendations to protect and manage t his h eritage. e European Community h as produced on May, 5, 2004 a text t h at indicates h ow to protect and manage geological sites. P ATRIC CABROL & A LAIN MANGIN

PAGE 7

ACTA CARSOLOGICA 37/1 2008 93 In France, Natural and cultural h eritage is protected wit h laws according to t h e kind of h eritage interest t h e sites h ave. 1) e law of 1913 deals wit h historical monuments and allows to take in consideration pre historic, historic and artistic remains. 2) e law of 1930 (Article L 341-1 to L 341-17 code of t h e environment) allows to protect natural monu ments, or sites if t h ey present c h aracteristics from an ar tistic, historical, scientic, legendary or picturesque point of view. ere are now more t h an 2700 sites of t his type. is law protects many sites t h at s h ow geomorp h o logic interest suc h as valleys, clis, glaciers and big rocks. It is also t h e case of more t h an 100 sand cave fossils sites (vertebrate fossils from t h e Tertiary period of t h e Cam pan Site at Sansan in Gers Department.). 3) e Law of 1976 dedicated to Nature Conserva tion (Article L.242-1 of t h e code rural) allows to take in account e preservation of biotopes and noticeable geomorp h ologic or speleologic geological formation, and the preservation of sites which have an exceptional interest for studies on life evolution and rst human activi ties is s h ort list s h ows t h at it is possible to protect un derground h eritage wit h a combination of several laws. According to t h e local context, one law or anot h er one will be used. e common c h aracteristic of t h ese laws is t h at ev ery new building construction t h at t hreaten to c h ange somet hing in t h e site are submitted to a special aut h ori sation. ese constraints are made very h eavy in order to protect t h ese valuable sites. ey form a quality label acknowledged by t h e Frenc h government for t h ese ex ceptional sites. e major dierence between t h e laws of 1930 and 1976 is t h e constitution of a management committee for natural reserves t h at was not required for registered sites wit h t h e law of 1930. is new disposition was decided w h en t h e Espar ros Site (Hautes-Pyrnes) was classied in 1987 and now it applies to la Cigalre Cave, Lauzinas Cave, R seau Lac h ambre, Clamouse Cave, Aven Armand, Aven dOrgnac, Aven des Perles, and mines Cave. is disposition will be used for ot h er classication proposals t h at are under progress at present. ese laws may also protect National Parks, consid ering t h e h eritage t h ey represent. MANAGEMENT AND PROTECTION OF GEOLOGICAL KARSTIC HERITAGE IN FRENCH LEGISLATION LA W S OF 1913, 1930 AND 1976 CONCLUSION e h eritage present in t h e Frenc h karst (geomorp h ol ogy, mineralogy, arc h aeology, palaeontology, biology) is outstandingly ric h and varied. Numerous sites are still to discover and a lot or researc h es will be necessary in order to study and protect t h em. Pierre C h auve carried out t h e rst inventory of t his h eritage, w hic h will be pub lis h ed t his current year, as part of his book Des Grottes et des Sources (Caves and Springs), Pour la Science edition. It will be t h e starting point of a national inven tory of t h e h eritage present in t h e karst. e knowledge and protection of t his h eritage are improving. About two h undred and y caves are already protected by diverse regulations and so are some gorges (Gorges of t h e Tarn, Gorges of t h e Verdon etc) and limestone plateaux. e national inventory of t h e geological h eritage started in 2000 in an experimental region, Brittany, and it will be extended to t h e entire territory as early as t his current year. It is possible to t hink t h at, in t h e few next years, a deeper knowledge of t his h eritage will be available and t h at nation-wide programs of study, protection and val orisation will be implemented. KARST IN FRANCE AND UNESCO W ORLD HERITAGE



PAGE 1

THE FLO W RATE OF THE FONTE PLINIANA COMO, ITAL Y : T W O THOUSANDS Y EARS OF DATA P RETOK KRAKEGA IZVIRA F ONTE P LINIANA C OMO I TALIJA : DVE TISO LETJI PODATKOV Arrigo A. CIGNA 1 Izvleek UDC 556.556(450.253)(091) Arrigo A. Cigna: Pretok krakega izvira Fonte Pliniana (Como, Italija): dve tisoletji podatkov Izvir Fonte Pliniana je zbujal pozornost obiskovalcev e od rim ski h asov dalje. Prispevek podaja nekaj opazovanj, zabeleeni h v skoraj dve h tisoletji h. Meritve, opravljene v 20. stoletju, omogoajo oceno hidravlini h znailnosti tega izvira. Kljune besede: Fonte Pliniana, intermitentni izvir, zgodovina, jezero Como, Italija. 1 Fraz. Tuo, I-14023 Cocconato (Asti) Italy, E-mail: arrigocigna@tiscali.it Received/Prejeto: 15.01.2008 COBISS: 1.01 ACTA CARSOLOGICA 37/1, 95-100, POSTOJNA 2008 Abstract UDC 556.556(450.253)(091) Arrigo A. Cigna: e ow rate of Fonte Pliniana (Como, Italy): two thousands years of data. Since t h e Roman time t h e Fonte Pliniana attracted t h e inter est of many visitors. Here some observations recorded in nearly 2000 years are reported. e measurements made during t h e XX century allowed also t h e evaluation of some h ydraulic pa rameters of t h e spring system. Keywords: Fonte Pliniana, ebb and ow, history, Como Lake. I NTRODUCTION e Fonte Pliniana is an ebb and ow spring. Its period of ow occurs at rat h er irregular intervals. It is a typi cal karstic spring in Lias grey limestone very close to t h e border of Como Lake (Cigna & Rondina, 1959) (Fig. 1). Some historical data on t h e spring are h ere reported as well as an evaluation of its main h ydraulic parameters. F ig. 1 Location of the F onte P liniana, near Como

PAGE 2

ACTA CARSOLOGICA 37/1 2008 96 e earliest reference to t his spring apparently is in Naturalis Historiae, a monumental work by Plinius t h e Elder (77 A.D.) publis h ed in t h e First Century A.D.: in Comensi iuxta Larium lacum fons largus horis singu lis semper intumescit et residit [in t h e vicinity of Como, close to t h e lake Larium 1 t h ere is a large spring t h at every h our ows and ebbs]. Also his nep h ew, Plinius t h e Y ounger (96-109), de scribes t h e spring: F ons oritur in monte, per saxa decur rit, excipitur cenatiuncula manu facta; ibi paulum retentus in Larium lacum decidit. Huius mira natura: ter in die sta tis auctibus ac diminutionibus crescit decrescitque. [= A spring issues in t h e mountain, ows t hroug h t h e rocks is kept in an small articial room and s h ortly aer reac h es t h e lake Larium. It h as a wonderful c h aracteristic: every day its ow increases and decreases t hree times]. He supposed t his eect to be due to t h e air, as it h appens w h en a bottle is emptied, or to winds inside t h e mountain, w hic h inuence t h e water ow. Suc h quota tions by bot h Plinies le t h eir name to t h e spring and consequently to a near by building, Villa Pliniana, as it is called presently (Fig. 2). is villa was built in 1570 by Count Anguissola as a refuge to escape vengeance being involved in t h e murder of Duke Farnese. 1 Larium is the Latin name of the Lake of Como, and it is still used presently. Nearly 1500 years later anot h er learned man, Leon ardo da Vinci, in t h e rst decade of t h e XVI century visited t h e region and was intrigued by t h e unusual be h aviour of t h e spring. In Codex Leicester (Leonardo da Vinci, 15061513), a veritable treatise on water, h e included few lines on t h e spring: Come in molti lochi sitrova vene dacqua che sei ore crescono e sei ore calano e io per me no veduta una in su lago di como ditta fonte pliniana la qual fa il predetto crescere e diminuire in modo che quando versa macina pi mulina e quando manca cali si cegli come guardarsi lacqua in un profondo pozo [= In many places t h ere are streams of water w hic h swell for six h ours and ebb for six h ours and I for my part h ave seen one above t h e lake of Como called Fonte Pliniana w hic h increases and ebbs as I h ave said in suc h a way as to turn at a giddy speed; and w h en it fails it falls so low t h at it is like looking at water in a deep pit]. In Fig. 3 t h e passage by Leonardo excerpted from t h e Codex Leicester, leaf 11, v. is reproduced mirrored for an easier reading. A couple of centuries later additional and more ac curate measurements are available. In fact t h e owner of t h e Villa Pliniana reported to G h ezzi (1742) some news about t h e ow rate of t h e spring observed from June 11 t h to July 2 nd 1741. Bot h t h e time interval between ebbs and t h e amount of t h e ow rate were not constant. A rst se ries of data supplied by a very kind gentleman owner of Villa Pliniana are summarised as reported in Table 1. But G h ezzi was muc h intrigued by t h e be h aviour of t h e spring and asked his friend to provide more accurate measurements. is gentleman measured t h e h eig h t of t h e water in t h e pool by means of a pole subdivided in Paris feet, inc h es and lines. e diagram referring to a couple of days is reported in Fig. 4 and 5. It was possible to identify t h e aut h or of t h ese data as Giuseppe Canarisi, since his family owned t h e Villa from 1676 to 1831 (Vas coni, 2007). e average period of t h e ebbs was 99 min as measured in t h e mont h of June and 85 min in t h e mont h of July, but it may be assumed a rounded value of 1h30m. HISTOR Y F ig. 2 V illa P liniana in a print of 1885 (B oniforti L., 1885) F ig. 3 e passage by Leonardo on the F onte P liniana, excerpted from the Codex Leicester, leaf 11, v. A RRIGO A. CIGNA

PAGE 3

ACTA CARSOLOGICA 37/1 2008 97 time it ows so rapidly, that(the outlet being small) the ba sin lls up to the brim and even overows its edges. Again the gush of water from the mountain gradually diminishes, and that in the basin as gradually acquires its lowest level. is alternation is repeated three times a day, though not with any great regularity. (W alter 1928) In modern time furt h er observations were carried on in t h e years aer t h e 2 nd W orld W ar w h en speleology developed again aer suc h a dicult period. Carlo Mar iani from t h e Gruppo Grotte Desio visited t h e Fonte Pliniana and in t h e mont h s of January and February 1949 h e made a series of measurements. On January 16, 1949 t h e average ow rate was around 3300 L/minute and t h e Table 1 First series of t h e results obtained in t h e Villa Pliniana in 1741 Date Time interval No. of ebbs Period (min) June 11 12 22:50 14 96 June 14 5:30 6 55 June 15 24 12 120 June 16 24 16 90 June 17 24 13 108 Average ( SD) 99 In 1828 t h e Rev. W eever W alter, M.A., of St. Jo hns, College Cambridge, publis h ed a series of letters describ ing t h e places and t h e life aer a travel. Wh en in Como in his Letter XXVIII dated May 1827 a description of t h e lake and t h e Fonte Pliniana is reported: the intermit tent fountain at V illa P liniana: at the foot of the mountain, about 30 feet above the level of the lake, in a natural basin hollowed out of the rock, having only a narrow outlet: a stream of water, beautifully clear, rushes out of the moun tain and empties itself into this basin; for a certain length of F ig. 4 F low and ebb of the F onte P liniana. e time scale starts at 9:05 of July 18, 1741. On the vertical axis the height of the water pool is reported. F ig. 5 F low and ebb of the F onte P liniana. e time scale starts at 12:49 of July 19, 1741. On the vertical axis the height of the water pool is reported. e interval between 341 and 468 minutes corresponds to the dinner time (19 to 21h), therefore during such an interval the measurement is missing. F ig. 6 F low and ebb of the F onte P liniana. e time scale starts at 13:02 of January 16, 1949. F ig. 7 F low and ebb of the F onte P liniana. e time scale starts at 9:50 of F ebruary 27, 1949 THE FLO W RATE OF THE FONTE PLINIANA COMO, ITAL Y : T W O THOUSANDS Y EARS OF DATA

PAGE 4

ACTA CARSOLOGICA 37/1 2008 98 F ig. 9 e hydraulic system of a ow and ebb spring. F rom Kircher (1668, p. 306) period was 76 minutes; on February 27, 1949 t h e av erage ow rate amounted to 1400 L/minute and t h e pe riod increased to 160 minutes; on bot h days t h e water temperature was 9.5C (Mariani, 1949). In Fig. 6 and 7 suc h data are plotted. Successively anot h er series of measurements made on Marc h 17, 1957, in t h e framework of a researc h on t h e springs in t h e province of Como (Cigna & Rondina, 1959). e average ow rate was estimated around 6000 L/minute and t h e water temperature was 8.4C. In Fig. 8 t h e results of t h e measurements are reported. It s h ould be noted t h at t h e variation of water level observed in t his last time, is of t h e order of tens of mil limetres w hile in t h e past was of tens of centimetres. e dierence is probably due to a c h ange of t h e size of t h e outside water pool. F ig. 8 F low and ebb of the F onte P liniana. e time scale starts at 11:28 of M arch 17, 1957 QUANTITATIVE EVALUATION OF THE PHENOMENON Plinius t h e Y oung rstly advanced a h ypot h esis on t h e mec h anism leading to t h e ow and ebb, as it was re ported above. But a more realistic proposal was reported by Kirc h er (1668). is aut h or did not referred directly to t h e case of t h e Fonte Pliniana, but h e described fait h fully its mec h anism. erefore t h e very original gure by Kirc h er may be used h ere. In Fig. 9 some sources G feed water to t h e ponds F; t h en water reac h es t h e point E in a cave in t h e moun tain. A sip h on B-A-I is t h e outlet of t h e reservoir inside t h e cave. Wh en t h e waterfall D lls t h e void to a level hig h er t h an t h e top of t h e sip h on (A) it disc h arges t h e water outside until t h e level inside t h e cave is lower t h an t h e inlet of t h e sip h on (B) and t h e ow is stopped. In t h e case h ere described I is t h e Fonte Pliniana and L is t h e Lake of Como. But h ow a sip h on works in nature will be discussed later. e periods of t h e ow and ebb as ob served since Plinys time to present are summarised in Table 2. In antiquity time was divided to h ave 12 h ours during t h e day and 12 h ours during t h e nig h t and since it is not known t h e date of t h e observations made by t h e Plinies and Leonardo, t h e periods corresponding to t h eir observations are reported wit h an approximation of 30%, corresponding to t h e maximum dierence of lengt h of t h e h ours between solstice and equinox at t h e latitude of Como Lake. Suc h periods span between 60 and 720 min utes and also t h e ow rate is rat h er irregular. Mangin (1973) and Bonacci & Bojanic (1991) de scribed t h e modelling of non-constant karst springs. Bot h papers reported rat h er accurate evaluations of t h e w h ole system t h anks to a large number of data of t h e ow rate in function of time. Unfortunately t his is not t h e case considered h ere, w h ere t h e records spread over two t h ousands years but t h e accuracy of t h e measurements is somew h at uncertain. Anyway it is possible to carry on some tentative evaluation of t h e factors involved in t h e h ydraulic system of t h e Fonte Pliniana. A RRIGO A. CIGNA

PAGE 5

ACTA CARSOLOGICA 37/1 2008 99 Table 2 P eriods of the F onte P liniana Reference Year Period (minutes) Plinius the Elder, 77 77 60 Plinius the Young, 96-109 96-109 240 Leonardo da Vinci, 1506-13 ~1510 720 Ghezzi, 1742 1741 90 Walter, 1828 1827 240 Mariani, 1949 1949 76 to 160 Cigna & Rondina, 1959 1957 >210 Let V be t h e volume of t h e void between B and A, Q in t h e inow from E and Q out t h e outow from I. In addition let q be a fraction of Q out w hic h ows at t h e minimum of t h e period. en t h e following equations can be written: q = Q out wit h 0<<1 (1) V= (Q out Q in )T ou (2) en t h e volume V may be calculated: V= (Qi n q) T in (3) V=(Q out + q Q in ) Ti n (4) Finally also Q in can be calculated Q in =Q out (1+ )T )T out + T in T out + T in (5) By substituting t h e results of t h e measurements made in t h e last century and by assuming t h at is small enoug h to be neglected, a roug h evaluation of t h e inow and t h e volume of t h e reservoir inside t h e mountain is obtained and reported in Table 3. e mean yearly rainfall in t h e region is around 1000 mm/year. By assuming an average value for Q out of 3000 L/min t h e annual ow is around 1.5-2 million of m 3 W it h a runo around 50% and a rainfall as reported above, a basin of about some km 2 would supply enoug h water to t h e spring. Since t h e average annual temperature of t h e mountains above t h e spring is of 7C and t h e dif ference of altitude is nearly 1000 m, a gradient of 0.234 C for 100 m of fall due to t h e transformation of work into h eat would increase t h e water temperature of t h e basin from 7 to a nal value of 9.3C, very close to t h e 9.7C measured wit h a calibrated t h ermometer (Cigna & Ron dina, 1959). is result supports t h e connection between a rec h arge basin in t h e mountains above t h e spring wit h t h e spring itself. But in nature it is necessary to eliminate t h e air in t h e upper part in order to prime t h e sip h on. Wh en t h e water level in t h e reservoir V rises above t h e upper part of t h e sip h on t h e water will start to ow in t h e descend ing branc h wit h a simple overow above t h e edge. But if t h e water entering t h e reservoir will reac h a level hig h enoug h, suc h an overow will increase to a point of drag ging t h e air in t h e descending branc h. Due to t h e irregular s h ape of t h e conduit and t h e irregularities of t h e inow t h e drag of t h e w h ole air will not take place always at t h e same conditions. If t h e water reac h es a level in t h e reservoir hig h er t h an usually before t h e sip h on is triggered, t h e outow of t h e spring will be occasionally more relevant, as it h as been recorded, e.g., by G h ezzi (1742). Table 3 Hydraulic parameters of the F onte P liniana Reference T out minutes T in minutes Q out L/minutes Q in L/minutes (calculated) V m 3 (calculated) Mariani, Jan.16,1949 43 44 3300 1600 150 Mariani, Feb.27,1949 92 40 1400 1000 60 Mariani, Feb.27,1949 70 50 1400 800 70 Cigna & Rondina, Mar. 17, 1959 219 29 6000 5000 170 THE FLO W RATE OF THE FONTE PLINIANA COMO, ITAL Y : T W O THOUSANDS Y EARS OF DATA

PAGE 6

ACTA CARSOLOGICA 37/1 2008 100 e spring was known since t h e rst observations for its irregular be h aviour bot h of t h e ow rate and t h e period. e ow rate is strongly aected by t h e amount of rain and snow melting in t h e basin above t h e spring. e ow and ebb is due to t h e existence of a sip h on at t h e outlet of a reservoir. But t his simplied model accounts only for t h e ow and ebb, since t h e real feature of t h e h ydraulic system could be complicated, e.g., by existence of more reservoirs connected eac h ot h er directly or t hroug h si p h ons. Suc h a h ypot h esis is supported by t h e irregular ow rate reported in Fig. 7, w hic h was detected by rat h er fre quent measurements of t h e water level in t h e outside ba sin. e data recorded in previous investigations aimed to establis h t h e maximum and minimum levels only, and t h erefore it was not possible to study t h e details of t h e ow rate. In addition suc h a complex network could also be modied during t h e centuries by erosion and lling of passages by bot h rock and organic debris. In order to understand fully t h e mec h anisms in volved in t his historic spring it would be desirable to car ry out a systematic monitoring of its water ow by means of modern automatic devices, w hic h provide a detailed and extended record rat h er easily. ACKNO W LEDGEMENTS CONCLUSIONS e aut h or is very grateful to dr. Saba DellOca and t h e Vice-Mayor of Torno, prof. Marcella Vasconi w h o kindly provided t h e information on t h e owner of Villa Pliniana at t h e time of t h e measurements reported by N. G h ezzi. e extremely fruitful discussion wit h prof. G. Badino claried many issues and improved greatly t h e w h ole pa per. e relevant suggestions provided by dr. O.Bonacci and dr. A.Kranjc, and t h e copy of some references trans mitted by dr. F.Gabrovsek are also gratefully acknowl edged. REFERENCES Bonacci O. & D. Bojanic, 1991: R h yt hmic karst springs. Hydrol. Sci. J. 36,1-2: 35-47, W allingford, Oxford s hire Boniforti L., 1885: Per Lag hi e Monti. Premiata Guida de scrittiva, storica, artistica, pratica. Dumolard Roux, Milano-Torino. Cigna A. & G. Rondina, 1959: Sullidrologia carsica epi gea nel territorio della provincia di Como (Lombar dia). Atti Soc. It. Sc. Nat., 98,1, 84-120, Milano G h ezzi N., 1742: Dellorigine delle fontane. Presso Sim one Occ hi, 273-275, Venezia Kirc h er A., 1668: Mundus Subterraneus in XII libros di gestus. J. Janssonium W aesberge & Filios, 190-192, Amstelodami. Leonardo da Vinci, 1506-13: Codex Leicester, leaf 11, v. Mangin A., 1973: Sur les transferts deau au niveau du karst noy a partir de travaux sur la source de Fon testorbes. Ann. Splologie, 28, 1, 21-40 Mariani C, 1949: Pliniana. Gruppo Grotte Desio, Manu script unpublis h ed. Plinius C. S., 77: Naturalis Historia. Book. 2, cap. XCV. Giardini, 1984, Como. Plinius G.C.S., 96-109: Liber IV, Epistula 30 Vasconi M., 2007: Personal communication by t h e ViceMayor of t h e Community of Torno. W alter W ., 1828: Letters from t h e continent. W Black wood Edinburg h & T. Cadell, Strand: 203-206, Lon don. A RRIGO A. CIGNA



PAGE 1

H Y DROLOGIC CONNECTIONS AND D Y NAMICS OF W ATER MOVEMENT IN THE CLASSICAL KARST KRAS AQUIFER: EVIDENCE FROM FREQUENT CHEMICAL AND STABLE ISOTOPE SAMPLING HIDROGEOLOKE POVEZAVE IN DINAMIKA PRETAKANJA VOD V VODONOSNIKU KLASI NEGA KRASA: IZSLEDKI POGOSTIH KEMI NIH IN IZOTOPSKIH ANALIZ Daniel H. D OCTOR 1 1 U.S. Geological Survey, 12201 Sunrise Valley Drive, Reston, VA 20192, USA, d h doctor@usgs.gov Received/Prejeto: 19.09.2007 COBISS: 1.01 ACTA CARSOLOGICA 37/1, 101-123, POSTOJNA 2008 Abstract UDC 556.34.044 Daniel H. Doctor: Hydrologic connections and dynamics of water movement in the classical Karst (Kras) aquifer: Evi dence from frequement chemical and stable isotope samoling A review of past researc h on t h e h ydrogeology of t h e Classical Karst (Kras) region and new information obtained from a twoyear study using environmental tracers are presented in t his paper. e main problems addressed are 1) t h e sources of water to t h e Kras aquifer resurgence zoneincluding t h e famous Ti mavo springsunder c h anging ow regimes; 2) a quantica tion of t h e storage volumes of t h e karst massif corresponding to ow regimes dened by h ydrograp h recessions of t h e Timavo springs; and 3) c h anging dynamics between deep p hreatic con duit ow and s h allow p hreatic and epip hreatic storage wit hin t h e aquifer resurgence zone as determined t hroug h c h anges in c h emical and isotopic composition at springs and wells. Partic ular focus was placed on addressing t h e long-standing question of t h e inuence of t h e Soa River on t h e ground waters of t h e aquifer resurgence zone. e results indicate t h at t h e alluvial aquifer supplied by t h e sinking of t h e Soa River on t h e nort h western edge of t h e massif contributes approximately 75% of t h e mean annual outow to t h e smaller springs of t h e aquifer resur gence zone, and as muc h as 53% to t h e mean annual outow of t h e Timavo springs. As a w h ole, t h e Soa River is estimated to contribute 56% of t h e average outow of t h e Kras aquifer resurgence. e proportions of Soa River water increase under drier conditions, and decrease under wetter conditions. Time series analysis of oxygen stable isotope records indicate t h at t h e transit time of Soa River water to t h e Timavo springs, Sardos spring, and well B-4 is on t h e order of 1-2 mont h s, depending on h ydrological conditions. e total baseow storage of t h e Ti mavo springs is estimated to be 518 million m3, and represents 88.5% of t h e storage capacity estimated for all ow regimes of t h e springs. e ratio of baseow storage volume to t h e average annual volume disc h arged at t h e Timavo springs is 0.54. e Reka River sinking in Slovenia supplies substantial allogenic rec h arge to t h e aquifer; h owever, its inuence on t h e nort h west Izvleek UDK 556.34.044 Daniel H. Doctor: Hidrogeoloke povezave in dinamika pre takanja vod v vodonosniku klasinega Krasa: izsledki pogostih keminih in izotopskih analiz V lanku predstavim pregled dosedanji h hidrogeoloki h razis kav klasinega Krasa in nova spoznanja, ki temeljijo na dvo letni tudiji naravni h sledil. Osredotoim se na tri probleme: 1) izvor voda v izviri h vodonosnika Krasa ob razlini h vo dostaji h, vkljuno z izvirom Timave, 2) ocena volumnov zalog krakega masiva v odvisnosti od pogojev toka, ki ji h razberemo iz recesijske krivulje izvira Timave; 3) uporaba analize spre memb izotopske sestave vode na izviri h in v vrtina h pri tudiji spremenljive dinamike med dotokom iz prevodnikov globoke freatine cone ter iz zalog plitve freatine in epikrake cone v izvirnem obmoju vodonosnika. Podrobno obravnavam e dalj asa aktualno vpraanje vpliva reke Soe na podzemne vode iz virnega obmoja. Izsledki kaejo, da aluvijalni vodonosnik, ki ga napaja Soa, prispeva priblino 75% povprenega letnega iz toka manji h izvirov izvirnega obmoja in do 53% povprenega letnega iztoka izvira Timave. Skupaj Soa prispeva kar 56% celotnega iztoka izvirnega obmoja Krasa. Relativni prispevek Soe je veji v sunem obdobju. Iz analize asovni h vrst vred nosti meritev stabilni h izotopov kisika smo ocenili, da je poto valni as vode med reko Soo in izviroma Timavo in Sardoem (in tudi vrtino B-4) 1-2 meseca, odvisno od hidroloki h razmer. Celotni volumen zalog baznega toka Timave ocenjujemo na 518 milijonov kubini h metrov, kar predstavlja 88,5% celotni h zalog, ocenjeni h za vse pogoje toka na izviri h. Razmerje med volumnom zalog baznega toka in volumnom povprenega let nega iztoka Timave je 0.54. Reka Reka, ki ponira v Sloveniji, je pomemben vir alogenega napajanja vodonosnika, a je njen vpliv na severoza h odnem delu izvirnega obmoja omejen predvsem na izvire Timave. Reka predstavlja pomembno komponento pretoka na izviru zgolj ob poplavni h dogodki h, ki trajajo od nekaj dni do nekaj tednov. Trajnostno upravljanje ezmejnega vodonosnika Krasa je mogoe zgolj z zagotavljanjem visoke ka kovosti vode reke Soe, prav tako pa bo potrebno izvesti e ve natanni h sledilni h poskusov v epikraki coni vodonosnika, s

PAGE 2

ACTA CARSOLOGICA 37/1 2008 102 resurgence zone is limited to t h e Timavo springs, and is only a signicant component of t h e spring disc h arge under ood con ditions for relatively brief periods (several days to weeks). Sus tainability of t h e trans-boundary aquifer of t h e Kras will benet from maintaining hig h water quality in t h e Soa River, as well as focused water tracing experiments wit hin t h e epip hreatic zone of t h e aquifer to better delineate t h e rec h arge zone and to identify sources of potential contamination to t h e Brestovica water supply well. Keywords : Classical Karst, Kras, aquifer, c h emistry, stable iso topes, h ydrograp h recession, Slovenia, Italy. INTRODUCTION e Classical Karst of sout h west Slovenia and nort h of Trieste, Italy, is a hig h limestone plateau t h at stands above t h e Adriatic Sea at t h e Gulf of Trieste. e region is t h e type locality of karst terrains, h aving been rst described by Valvasor in 1689 (Kranjc, 1997). In t h e Slovene us age, t h e word kras is a place-name for t his region, as well as a term w hic h once implied a landscape t h at is barren, rocky, and wit h out available water (Gams, 1993). For centuries, t h e term was a tting description of t h e Classi cal Karst, and t h e region still bears t h e place-name, Kras 1 e Kras region is noted for its lack of readily available water resources in spite of an abundance of precipitation. Roug h ly 1400 mm of precipitation fall annually on t h e 440 km 2 region, yet t h e in h abitants of t his sparsely popu lated region h ave historically relied upon elaborate sys tems designed to collect rainfall from rooops and store t h e water in articial cisterns and ponds (Kranjc, 1997; Ravbar, 2004). e reason for t h e absence of surface runo is t h e intense karstication of t h e uplied block of Cretaceous carbonate rocks, exacerbated by centuriesold deforestation and soil loss, w hic h permits rapid inter nal drainage of precipitation (Gams, 1993). Several large springs are located along t h e Adriatic coast in Italy on t h e low nort h western edge of t h e Kras region (Fig. 1). Col lectively, t h ese springs h ave an average annual disc h arge of 35 m 3 /sec (Gemiti, 1984a). Alt h oug h t h e springs of t h e aquifer resurgence rise in Italy, t h ey are supplied to great degree by precipitation and allogenic runo originating in Slovenia. us, t h e aquifer of t h e Kras is also a classical example of a trans-boundary aquifer. is paper is presented in two parts: t h e rst part is an overview of previous researc h on t h e h ydrogeology of t h e Kras region. e second part is a summary of t h e results of a two-year study employing natural c h emical and isotopic environmental tracers to better understand t h e dynamics of water movement wit hin t his complex carbonate aquifer system. I. PREVIOUS H Y DROGEOLOGICAL RESEARCH ON THE KRAS AQUIFER G EOLOGIC SETTING OF THE K RAS e Kras region is a hig h plateau overlooking t h e Gulf of Trieste at t h e nort h ernmost part of t h e Adriatic sea (Fig. 1). e Kras is 40 km long, up to 13 km wide, and covers approximately 440 km 2 (Krancj, 1997). e stratigrap hic sequence consists of carbonate rocks (limestones and do lomites) overlain by turbiditic argillaceous sandstone and s h ale (ysc h). e sediments were deposited from early Cretaceous to Eocene, w h en a large carbonate platform was deposited, t h en drowned and capped by t h e ysc h sediments during active tectonic upli (Cucc hi et al. 1 is region is t h e Classical Karst in Englis h, Kras in t h e Slovene language, and il Carso in t h e Italian language. In t his paper, t h e region will h ereaer be referred to by its Slovene name, Kras. Ot h er place names and p h ysical geograp hic features are referred to by t h eir respective Slovene and Italian names w h enever appropriately situated wit hin t h e respective country. Two exceptions are t h e Soa and Vipava rivers, w hic h originate in Slovenia before crossing t h e international border into Italy, and are t h erefore referred to by t h eir Slovene names t hroug h out t his paper, regardless of w h et h er t h e point along t h e river t h at is being referred to lies wit hin Slovenia or Italy. D ANIEL H. D OCTOR katerimi bi doloili izvor onesnaenja vode v vrtini vodnega vira Brestovica Kljune besede: Klasini Kras, vodonosnik, kemija, stabilni izotopi, recesija hidrograma, Slovenija, Italija.

PAGE 3

ACTA CARSOLOGICA 37/1 2008 103 1987). e entire sequence was later uplied and strongly folded and faulted. e Kras region lies on t h e AdriaticDinaric tectonic plate, w hic h collided wit h t h e Euroasian plate to t h e nort h in t h e early Eocene (roug h ly 50 million years ago). is tectonic collision caused t h e upliing of t h e Julian Alps, strong faulting and folding of t h e rocks between t h e Julian Alps and t h e Adriatic Sea, and t h e drowning of t h e carbonate platform (ebela, 1997). is process of orogenesis h as continued from t h e Eocene to t h e present, and is still active today. e plateau of t h e Kras itself is formed by t h e large overturned Trieste-Komen anticline t h at plunges to t h e F ig. 1: M ap of the study area. e resurgence zone of the Kras aquifer is enlarged. Geographic features are named according to the language of the country in which they are located, or in the case of rivers, where they originate. H Y DROLOGIC CONNECTIONS AND D Y NAMICS OF W ATER MOVEMENT IN THE CLASSICAL KARST KRAS AQUIFER ...

PAGE 4

ACTA CARSOLOGICA 37/1 2008 104 nort h west, extending along t h e Adriatic coast. Erosion h as stripped t h e ysc h rocks o of t h e surface of t h e pla teau, exposing t h e carbonates wit h in t h e core of t h e anti cline. is anticline is bounded by t h e Trieste fault along t h e Adriatic coast to t h e sout h west, and to t h e nort h east by t h e Raa fault. Bot h of t h ese large t h rust faults trend N W -SE, and form h ydrogeologic boundaries along t h e edge of t h e Kras by bringing up t h e siliceous, relatively impermeable ysc h layers into contact wit h t h e karsti ed limestones, w h ile a t h ird fault runs down t h e axis of t h e anticline and may be a more permeable pat h way for ground water ow (Fig. 2). e result is t h at t h e h ig h Kras plateau is eectively bound by less permeable rocks on its nort h ern and sout h ern borders, as well as at dept h On t h e sout h east border, t h e Reka River ows on ysc h terrain from its eastern source in Croatia un til it encounters t h e limestone bedrock, and t h en sinks. e Reka River enters t h is block of rock, sinking into t h e Kras at t h e base of a 100 m h ig h limestone cli at t h e kocjan Caves. F ig. 2: Geology of the Kras and summary of some water tracing results On t h e nort h western side of t h e Kras, t h e plateau surface abruptly slopes down to sea level and opens into t h e plain of t h e Soa River. Here, t h e ground-water of t h e karst aquifer resurges in a narrow zone of artesian springs on t h e sout h western edge, as a result of t h e pre vailing dip of t h e rock layers being west-sout h west. No surface streams exist on t h e plateau, and runo is neg ligible. Given t h e abundant annual precipitation, hig h ly permeable land surface, and lack of surface water runo, precipitation on t h e Kras surface is a major component of rec h arge to t h e underlying aquifer. Geologic structure h as been demonstrated to be a primary controlling factor on t h e development of t h e Kras aquifer and its compart mentalization of drainage (Cucc hi et al., 2001). M UNICIPAL W ATER SUPPLIES e Timavo springs were once t h e primary water sup ply for t h e city of Trieste. However, Trieste and t h e ot h er Italian towns along t h e Adriatic coast in t h e region cur D ANIEL H. D OCTOR

PAGE 5

ACTA CARSOLOGICA 37/1 2008 105 rently receive t h eir water supply from a series of pump ing wells drilled into t h e alluvial aquifer of t h e Soa River plain, west of t h e river and sout h of t h e city of Gorizia. e wells were constructed aer a c h emical spill into t h e Reka River in 1972 jeopardized t h e continued use of t h e Timavo springs as t h e primary water supply for Trieste. Sardos spring is still currently utilized for water supply of Trieste (A.C.E.G.A. Trieste, 1988). e water supply of t h e Slovenian Kras region comes from a pumping station t h at taps into t h e karst aquifer on t h e nort h western edge of t h e Kras, very near t h e inter national border between Italy and Slovenia. Several ex ploratory bore h oles were drilled, ten of w hic h are active piezometers for monitoring water level uctuations and for water sampling (Krivic, 1981). e c h osen site is t h e Klarii pumping station, w hic h is located in t h e Brestovi ca Valley, 2 km west of t h e town of Brestovica. Hereaer t his sampling point s h all be referred to as well B-4, in keeping wit h t h e notation of Krivic (1982a). W ell B-4 is screened from 14 m to 68 m dept h and more t h an 1 m of t h e bore h ole intersects an open karst conduit. Bot h t h e production wells and t h e monitoring well B-4 intersect t h e same owing karst conduit, and t h erefore draw t h e same water (samples for t his study were taken from t h e monitoring well at t h e pumping station, located 3 meters distant from t h e main pumping wells). e site is situ ated at an altitude of 16 meters above sea level (m.a.s.l.) approximately 4 km nort h of t h e Adriatic coast, and lies wit hin a karst doline located along t h e international border between Italy and Slovenia. e average water table elevation at t h e site is 2.0 m.a.s.l. A pumping test conducted wit h simultaneous sampling of c h emical and isotopic parameters in mid-August of 1995 s h owed t h at t h e Klarii source can produce more t h an 250 L/s wit h a drawdown of only 0.5 m (Urbanc and Kristan, 1998). No intrusion of seawater occurred during t h e pumping test, indicating t h at fres h ground-water exists to dept h s greater t h an 50 meters below sea level. In addition to well B-4, anot h er of t h ese bore h oles, well B-3, was repeatedly sampled during t his study. W ell B-3 is located approximately 1.5 km to t h e sout h east of t h e Klarii pumping station. Details on t h e construc tion of t h ese two wells is provided in Table 1. Bot h wells extend more t h an 40 m below sea level (m.b.s.l.), t h us intersecting a large portion of bot h t h e unsaturated and p hreatic zones of t h e aquifer. At well B-3, dept h to water during sampling usually ranged between +2 m to -10 m wit h respect to sea level, h owever during a sampling trip in August of 1999, t h e bailer hit sediment at t h e bottom of t h e bore h ole at a dept h of -10 m, and t h e well was dry. us, t h e well is completed wit hin t h e epip hreatic zone of uctuation of t h e local water table, and h as become par tially lled wit h sediment. Since t his well normally h eld water, it was of particular interest and was frequently sampled in t h e present study. Usually, t h e water level in t h e well at t h e time of sampling was between +2.0 and +3.0 m.a.s.l. On two occasions, in t h e late summer of bot h 1999 and 2000, t h e well water h ad a distinct smell of h ydrogen sulde gas, indicating reducing conditions in t h e ground-water. e coecient of correlation between p hreatic wa ter level measurements monitored for t h e period between April 1977 to January 1980 in well B-4 and well B-3 was 0.99 (Krivic, 1982a). us, t h e bore h oles are h ydraulically connected; h owever well B-4 is wit hin bedrock w hic h is more intensively karstied t h an B-3, as evidenced by t h e measured transmissivities at bot h sites (Table 1). T RACING EXPERIMENTS IN THE R EKA T IMAVO S Y STEM Past h ydrogeological researc h on t h e Kras focused mainly on t h e source of water of t h e Timavo springs and on t h e dynamics of t h e subterranean course of t h e Reka River (Kranjc, 1997). e Timavo springs are t h e largest natu ral source of ground-water in t h e region, wit h a mean an nual disc h arge of 30.2 m 3 /sec representing approximately PARAMETER Monitoring Well B-4 Monitoring Well B-3 Elevation (top of casing) 18.0 m.a.s.l. 40.7 m.a.s.l Depth ( below sea level) 52.5 m.b.s.l. (open fracture from 41.7-43.0 m.b.s.l.) 44.6 m.b.s.l. Screen interval 14 m to 68 m depth Unknown Water table elevation (range) 0.5 3.5 m.a.s.l. <0.0 3.0 m.a.s.l. Transmissivity (T) >1 x 10 -1 m 2 /s 2 x 10 -6 m 2 /s Hydraulic conductivity (K) >8 x 10 -4 m/s 4 x 10 -7 m/s Geologic material 0-0.5 m (soil) 0.5-52.5 m Cretaceous limestone 0-2 m Soil 2.3 m Cretaceous limestone Tab. 1: Well log and pump test data for Well B-4 and Well B-3 H Y DROLOGIC CONNECTIONS AND D Y NAMICS OF W ATER MOVEMENT IN THE CLASSICAL KARST KRAS AQUIFER ...

PAGE 6

ACTA CARSOLOGICA 37/1 2008 106 85% of t h e average measured outow of t h e aquifer re surgence (Civita et al., 1995). A compre h ensive histori cal account of t h e history of t h e h ydrogeologic researc h conducted on t h e springs is found in t h e book Timavo: Esplorazione e Studi by Galli (1999). It was generally recognized t h at two ot h er nearby rivers could also inuence t h e Kras aquifer resurgence: t h e Vipava River to t h e nort h, and t h e muc h larger Soa River to t h e west. In an experiment using lit hium-c h lo ride and strontium-c h loride salts, a connection was discovered between t h ese ot h er karst springs and t h e Vipava River t h at was seemingly independent of t h e Ti mavo springs (Timeus, 1928) (Fig. 2). In addition, an 80 m deep bore h ole located near Miren w h ere t h e Vipava River crosses t h e international border lost drilling wa ter during construction between dept h s of 64 m to 80 m, indicating karstic conduits at dept h (Bidovec, 1967). e disc h arge of t h e Soa River to t h e west, on t h e ot h er h and, is muc h larger; h owever, a direct articial trace h as never been conducted between t h e Soa River and t h e springs of t h e Kras resurgence zone. Conducting a water trace is complicated by t h e fact t h at t h e Soa River ows on a t hick accumulation of alluvium, and not on bed rock. is means t h at a clearly identiable sinking point cannot be easily located (Bidovec, 1967). However, losses from t h e Soa River h ave been gauged along 12 km of t h e rivers course between t h e towns of Gorizia and Gradisca. Estimates of ow loss are on t h e order of 20-25 m 3 /sec, w hic h reects approximately 10% of t h e rivers total dis c h arge upstream at Gorizia during average ow condi tions (Mosetti and DAmbrosi, 1963; A.C.E.G.A. Trieste, 1988). Measurements of piezometric water levels in t h e alluvial aquifer surrounding t h e Soa River indicate an area of focused ground-water loss between t h e river and t h e elevated Kras plateau (Mosetti and DAmbrosi, 1963). ese water table troug h s are elongated in t h e same di rection as faults trending nort h east-sout h west t h at h ave been determined t hroug h geop h ysical exploration (Mor gante et al., 1966). It was not until 1962 t h at t h e second signicant trac ing experiment conducted on t h e Reka-Timavo system took place. In t his study, tritium was used as t h e primary tracer (Eriksson et al., 1963; Mosetti, 1965). e experi ment was initiated during low ow, wit h t h e h ope t h at t h e baseow contribution of t h e Reka River to t h e Tima vo spring disc h arge could be quantitatively deduced. On July 3 rd 1962, 200 Curies of tritium and 100 kg of uores cein were injected at t h e entrance of t h e kocjan Caves, w h ere t h e sinking ow of t h e Reka River was gauged to be 0.5 m 3 /sec (Gemiti, 1984b). e ow at t h e Timavo springs, despite not being quantitatively gauged, was es timated at t h e time of tracer injection to be 10 m 3 /sec. However, t h e trace was interrupted by a large storm event on t h e day of injection, suc h t h at t h e Reka ow increased to 30 m 3 /sec and t h e Timavo disc h arge increased to an es timated 62 m 3 /sec two days aer tracer injection (Gemiti, 1984a). e initially reported recovery of tritium at t h e Timavo springs during t h e tracing experiment was only 50% of t h e injected tracer mass (Mosetti, 1965). e low amount of reported tracer recovery immediately lead to new ideas and t h eories about t h e fate of t h e underground Reka, and t h e source of t h e majority of t h e water issuing from t h e Timavo springs. However, at least two aut h ors h ave since pointed out t h at t h e interpretations of t his experiment were based upon faulty h ydrologic data and t h at t h e results s h ould be reconsidered (Bidovec, 1967; Gemiti, 1984a; Gemiti, 1996). e Reka River in fact begins to sink a few kilo meters upstream of t h e tritium injection point, near t h e Cerkvenik Mill. In earlier experiments, water disappear ing at t his sinking point h ad been traced to connect wit h t h e ow in t h e kocjan Caves. At t h e time of injection of t h e tritium, t h e actual ow of t h e Reka measured up stream of t h e injection point at t h e Cerkvenik Mill was 1.2 m 3 /sec, a factor of 2.4 times greater disc h arge t h an t h at used in t h e calculations to determine tracer recovery at t h e Timavo Springs (Bidovec, 1967). Moreover, Gemiti (1984a) concluded t h at t h e disc h arge at t h e Timavo springs during t h e trace was actually twice as great as initially supposed, and argues t h at instead of 50% tracer recovery reported in t h e results of Eriksson et al. (1963), practically all of t h e tritium t h at was injected was dis c h arged from t h e Timavo springs and in small part from t h e springs of Aurisina. Due to uncertainties in t h e ow estimations, h owever, t h e interpretation of t his test is le ambiguous. Nonet h eless, Gemiti (1984a) estimated t h e ratio of t h e total Reka River ow to t h e Timavo springs disc h arge across t h e transit time (i.e., time elapsed be tween tracer injection at t h e input and its initial detec tion at t h e outlet) to be 1:4.4, corresponding to 23% of Reka River water disc h arged at t h e springs, assuming all of t h e Reka river ow during t h at time went to Timavo. Very few controlled tracer experiments h ave been conducted on t h e Reka-Timavo system aer 1962. In 1972, an accidental h ydrocarbon spill occurred into t h e Reka River w h en a truck carrying several tons of fuel oil overturned on t h e road alongside t h e river, approxi mately 18 km upstream of t h e rivers sinking point at t h e kocjan Caves (Gemiti, 1998). At t h at time t h e Timavo springs were still in use as a drinking water supply to t h e city of Trieste, and t h e Slovenian aut h orities immediately informed t h e A.C.E.G.A. Trieste utility. e fast response enabled a close monitoring of t h e water c h emistry of t h e Reka River at kocjan Caves, t h e Timavo springs, and ot h er springs of t h e aquifer resurgence, t h us allowing t h e use of t h e pollutants as a tracers. e frequency of water D ANIEL H. D OCTOR

PAGE 7

ACTA CARSOLOGICA 37/1 2008 107 sampling was every h our at t h e Timavo springs, every two h ours at Aurisina, and twice per day at Sardos. e contaminant spill occurred under hig h ow conditions of t h e Reka River and Timavo springs (median ows of 13.5 m 3 /sec and 42.5 m 3 /sec, respectively). During t h ese hig h ow conditions t h e transit time between t h e Reka River and Timavo springs was 5 days and Gemiti (1998) estimated t h e ratio of river disc h arge to spring disc h arge across t h e transit time was 1:1.6, corresponding to 63% of Reka River water disc h arged at t h e Timavo springs, again assuming all t h e Reka ow went to Timavo. How ever, t h e initial detection of t h e pollutants was observed rst at Aurisina spring, and 12 h ours later at t h e Timavo springs. e peak of t h e breakt hroug h curves occurred two days later, rst again at Aurisina and t hree h ours later at Timavo. No pollutants were detected at Sardos spring, even w h en t h e hig h est concentrations of t h e pollutants were observed at Timavo and Aurisina springs. e pol lutant concentrations decreased to below detection limit in a matter of h ours at Aurisina, w hile t h e recession curve of t h e pollutant concentrations at Timavo extended for 3 days aer t h e peak (Gemiti, 1998). Gemiti (1998) also discussed t h ree traces t h at were conducted between t h e deep karst s h a located at Trebi ciano (Fig. 2) and t h e Timavo springs: one wit h tritium in 1962, and two wit h carbon tetrac h loride (CCl 4 ) in 1982 (note: tritium should not be used as a tracer due to its radioactivity, and carbon tetrachloride should not be used as a tracer due to its toxicity; several non-toxic uorescent dyes are commonly available as alternatives). Bot h t h e tritium trace and t h e rst CCl 4 trace were con ducted under moderate ow (t h e median ows of Ti mavo and Reka from time of injection to peak detection of t h e CCl 4 tracer were 7.1 and 32 m 3 /sec, respectively), w h ile t h e second CCl4 trace was conducted during a large ood event (median daily ows of Timavo and Reka were 96 and 50 m3/sec, respectively) (Gemiti, 1984b). e transit times of t h e 1962 and rst 1982 trace from Trebiciano to Timavo were 10.5 days and 10.0 days respectively, w h ile t h e transit time of t h e second 1982 trace conducted during ood ow was only 2.7 days. Estimates of t h e percentage of Reka River disc h arged at t h e Timavo springs during t h e trace of 1962 across t h e transit time from Trebiciano to Timavo was 18%; h owever, t h is value again assumes all of t h e Reka ow went to Timavo, and is not corrected for tracer recovery due to a lack of reliable ow data. e estimated values of percentage of Reka River disc h arge exiting at Timavo for t h e 1982 traces from Trebiciano conducted wit h CCl 4 were 19% for t h e moderate ow trace, w h ile for t h e h ig h ow trace t h e estimated percentage was 48%. However, t h e amount of CCl 4 recovered in t h e rst 1982 trace was 78%, w h ile only 32% was recovered during t h e h ig h ow experiment. In addition, bot h t h e 1982 traces conducted wit h CCl 4 h ave narrow, peaked tracer breakt h roug h curves s h owing total passage of tracer af ter 3 days for t h e rst trace and 2 days for t h e second trace, as opposed to t h e broader curve of t h e Timavo spring disc h arge w h ic h remained at elevated ow for 4 and 9 days, respectively. e results presented by Gemiti (1998 and 1984b) are summarized in Fig. 3. Note t h at w h en t h e calculated percentages of Reka River water at Timavo are adjusted for tracer recovery t his results in a lower percentage of Reka water expressed at Timavo t h an previously esti mated, averaging around 15% for t h e two traces from t h e s h a Trebiciano conducted wit h CCl 4 in 1982, one (1982b) h aving occurred under very hig h ow condi tions. e percentage of Reka water estimated from t h e 1972 trace is questioned given t h e lack of control over t h e experiment as t h e injection was due to an accidental h ydrocarbon spill. ese results hig h lig h t t h e necessity of controlled quantitative tracer recovery for accurately determining t h e proportion of t h e Reka River expressed at Timavo. In spite of t h e need for more quantitative data, t h ese traces do conrm t h e existence of a conduit con nection of large dimension between t h e Reka River and t h e Timavo springs, wit h a travel time of on t h e order of 600 m/hr (Gemiti, 1984b & 1998). Aer t h e course of t h e underground river passes t h e s h a at Trebiciano, t h e main conduit is p hreatic and extends below sea level (Boegan, 1938). e ratio between t h e total disc h arge of t h e Reka River and t h at of t h e springs during t h e tran sit time of t h e tracer is an important comparison, for it provides an estimate of t h e maximum volume of water emptied from t h e aquifer system t h at could result from t h e river input in t h at time period. However, t h ese per centages are overestimates since t h ey assume all of t h e Reka River ow measured during t h e underground pas sage of t h e tracer exits at t h e Timavo springs; in fact, it h as already been establis h ed t h at some portion of t h e subsurface Reka ow comes out at Aurisina spring along t h e coast, and also at Bagnoli spring furt h er inland under hig h ows. erefore, a major fraction of t h e Reka River ow sinking at t h e kocjan caves exits from t h e Timavo springs, but some ows elsew h ere. e complex h ydrau lic p h enomena in t h e aquifer provide additional diculty in determining h ow muc h of t h e Reka River ows out of t h e Timavo springs. More recent researc h on t h e function of t h e RekaTimavo system h as focused on continuous monitoring of water levels and p h ysico-c h emical parameters suc h as temperature and electrical conductivity between t h e sinking point of t h e Reka River and its underground course w h ere it is accessible via deep s h as (Cucc hi et al., 1997; Cucc hi et al., 2000; Cucc hi et al., 2002; Gabrovek H Y DROLOGIC CONNECTIONS AND D Y NAMICS OF W ATER MOVEMENT IN THE CLASSICAL KARST KRAS AQUIFER ...

PAGE 8

ACTA CARSOLOGICA 37/1 2008 108 and Peric, 2006). ese studies h ave provided valuable information on t h e h ydraulics of t h e system. e most recent dye tracing results on t h e Reka-Timavo system were reported by Peric et al. (2007). ey released 5 kg of uranine tracer dye into t h e Reka River on Sept. 4, 2006 at t h e sinking point on t h e contact between t h e ysc h and limestone. Dye was detected in low concentrations at t h e Timavo springs 32 days aer release under relatively low ow conditions; a single 35 mm rain event occurred dur ing t h e trace, w hic h enabled a more rapid transport of dye t h an prior to t h e event. Furt h er analysis of t h ese data is ongoing (Peric and Gabrovek, personal communica tion, 2007). T RACING EXPERIMENTS FROM THE EASTERN K RAS Additional water tracing tests were conducted in t h e late 1980s by Slovenian researc h ers from t h e Karst Researc h Institute in Postojna (Habi, 1989). ese were conduct ed from small sinks in t h e beds of ep h ermal streams lo cated on t h e eastern edge of t h e Kras, most notably from t h e Raa Lakovnik (Raa River sink), t h e Sajevki potok (brook), and wit hin t h e basin of Senoee-Dolenja Vas (Fig. 2). All of t h ese traces were conducted under low ow conditions. R h odamine injected into t h e Raa La kovnik was detected at t h e Timavo springs aer 60 days, at Aurisina aer 78 days, and at Sardos aer 105 days. e biological tracers injected at Senoee were detect ed at Aurisina aer 75 days, and a connection wit h t h e Sablici spring was also proven. Fluorescein injected into t h e Sajevki Potok was de tected rst at t h e water sup ply well B-4 at Klarii 40 days aer injection, at t h e Timavo springs aer 70 days, and at bot h t h e Sardos and Aurisina springs 80-90 days aer in jection (Galli, 1999). ese results proved t h e existence of additional sources of allo genic rec h arge to t h e aquifer from t h e nort h eastern edge of t h e Kras, but quantitative estimates of rec h arge were not made possible from t h e results, partly due to t h ese all being ep h emeral sources. In spite of t h ese stud ies, a question still remained: from w h ere originates t h e unaccounted proportion of t h e disc h arge at t h e Timavo springs? is is an important question, for t his water source provides t h e majority of t h e spring ow under low ow conditions. C HEMICAL AND STABLE OX Y GEN ISOTOPIC COMPOSITIONS OF THE K RAS GROUND W ATERS On t h e basis of t h e variations in water c h emistry and temperature measured by Gemiti and Liciardello (1977) and Cancian (1987), t h e ground-water of t h e aquifer re surgence can be divided into four separate groups of wa ters: 1) t h e water of t h e Timavo springs, 2) t h e water of Sardos and Mosc h enizze Sout h springs, 3) t h e water of Doberd Lake, Sablici springs and of Mosc h enizze Nort h spring, and 4) t h e water of Aurisina spring, or t h e socalled karst water (Gemiti & Licciardello, 1977; Gemiti, 1994). ese groupings were substantiated by later stable isotope (oxygen and carbon) analyses of t h e spring wa ters (Flora and Longinelli, 1989; Doctor et al., 2000). In addition, it was observed t h at t h e ground waters of t h e region all exhibit strong isotopic inversion in t h eir 18 O values on an annual cycle (lower values in warmer spring and summer mont h s, hig h er values in cooler fall and winter mont h s) (see Fig. 6.) is observation is opposite to t h e expected isotopic composition if t h e springs were fed primarily by runo from precipitation, given t h e in uence of seasonal uctuations in air temperature on t h e isotopic composition of rainfall (Dansgaard, 1964). e interpretation of Flora and Longinelli (1989) was t h at F ig. 3. Estimated maximum percentage of Reka River water exiting at the Timavo springs plotted against the ratio of median ow of Timavo to median ow of Reka during each trace event. e date and injection point of the tracer is indicated in the legend (data from Gemiti, 1984b and Gemiti, 1998). D ANIEL H. D OCTOR

PAGE 9

ACTA CARSOLOGICA 37/1 2008 109 t h e karst springs receive most of t h eir rec h arge from lo cal precipitation on t h e Kras region in t h e winter, and from meteoric water falling at hig h er elevations (800-900 m.a.s.l. as opposed to t h e 400 m elevation of t h e Kras pla teau) in t h e spring and summer mont h s. e overall result is a general s hi in t h e average spring water 18 O to isotopic compositions t h at are lower t h an t h e weig h ted mean annual isotopic composition of meteoric precipitation on t h e Kras region (DAmelio et al., 1994). e Kras region receives t h e majority of its precipitation in t h e fall and winter mont h s (SeptemberFebruary). Alt h oug h t h e winter rainfall on t h e Kras is generally more isotopically depleted in 18 O t h an t h e sum mer rainfall, t h e rain and snow precipitation in t h e Slo venian hig h lands is depleted even furt h er still, a result of t h e altitude eect w hic h oen results in an isotopic depletion of 0.2/100 m (Eriksson, 1983). By compar ing t h e isotopic composition of t h e spring waters to t h at of t h e Soa River, Vipava River, and local precipitation, Flora and Longinelli (1989) concluded t h at only during t h e drier summer mont h s w h en t h e local p hreatic level is lowered do t h e Timavo and ot h er springs of t h e resur gence receive rec h arge from a more isotopically depleted source. eir h ypot h esis was t h at t his source would be derived from t h e inner Slovenian hig h lands in t h e region of Postojna and of Cerknica, to t h e far nort h east of t h e resurgence zone and in fact outside of t h e previously de lineated rec h arge zone of t h e Kras aquifer. is h ypot h esis was contrary to t h at of Gemiti & Liciardello (1977) w h o attested to a more prevalent Soa River source, based upon t h e spring water c h emistry data. Later, Urbanc and Kristan (1998) estimated t h at ap proximately 50% of t h e water produced during a pump ing test at well B-4 was derived from t h e Soa River based on concurrent oxygen stable isotopic measurements of t h e river and well water. is surface water component is manifested not only at t h e Klarii pumping well, but in t h e disc h arge of all of t h e major springs in t h e surround ing area, as demonstrated by studies of t h e c h emistry and isotopic composition of t h e springs in comparison to t h at of t h e Soa River (Cancian, 1987; Longinelli, 1988; Flora and Longinelli, 1989; Krokos, 1998). I NORGANIC CARBON ISOTOPES AS TRACERS e stable carbon isotopic composition ( 13 C) of dis solved inorganic carbon (DIC) h as been proven useful for distinguis hing between t h e rec h arge areas for dier ent karstic waters along wit h oxygen and h ydrogen iso topes of water (Deines et al., 1974; Pezdi e t al., 1986; Urbanc et al., 1997). In t h e Kras system, carbon isotopes can be used to distinguis h between waters inuenced by t h e Soa River versus waters inuenced by more local rec h arge (Doctor et al., 2006). e annual range in t h e 13 C DIC composition of t h e Soa River is between -6.5 and -11.0, wit h regular seasonal oscillations in t h e carbon isotopic composition of t h e water (Urbanc et al., 1997). e most positive values occur in t h e early spring, and most negative values in t h e late fall. In t h e spring, t h e snowmelt from t h e mountainous regions of t h e Soa ba sin contributes greatly to t h e river disc h arge. e spring snowmelt brings water w hic h h as gained most of its DIC from atmosp h eric CO 2 ( 13 C CO2 = .0), from disso lution of carbonate rocks in t h e streambed ( 13 C calcite = +2.0), and from ground-water disc h arge into t h e river via springs w hic h empty into t h e river along its course ( 13 C DIC between and ). In t h e late fall, t h e river receives most of its disc h arge from spring ows, and less from surface runo. e springs issue water w hic h is h eavily inuenced by soil CO 2 and t h us t h e carbon iso topic composition of t h e water is more negative. As t h e spring water is carried downstream in t h e river, it again begins to equilibrate wit h atmosp h eric CO 2 and t h e carbon isotopic composition s his toward more posi tive values as CO 2 is lost and exc h anged. S his in 13 C as muc h as +0.4/km h ave been recorded (Urbanc et al., 1997). erefore, by t h e time t h e Soa River reac h es con tact wit h t h e western edge of t h e Kras and begins to sink just sout h of Gorizia, it h as 13 C DIC values between .7 and .0, wit h an annual mean of .3. In contrast to t h e river, local karst springs receive a greater portion of t h eir rec h arge primarily inltration of local precipitation into t h e karstied bedrock and typi cally s h ow more depleted 13 C DIC ranges between .0 and .0. e more negative values are t h e result of a greater inuence of soil CO 2 on t h e DIC in t h e water. e good separation between t h e carbon isotopic composi tion of t h e local ground-water and t h e Soa River water allows carbon isotopes to be useful for distinguis hing between t h ese two sources in t h e resurgent ground-wa ters of t h e Kras. Using bot h carbon and oxygen isotopes, Pezdi et al. (1986) estimated t h at as muc h as 50% of t h e water pumped from t h e well at Klarii during dry periods comes from t h e Soa River. us, t h e carbon isotopes are a useful c h eck on t h e interpretation of t h e ot h er c h emical and isotopic data obtained. WATER BALANCE OF THE K RAS REGION e boundaries of t h e Kras h ydrogeologic system are well dened. ere exists an impermeable ysc h layer (>400 m in places) beneat h t h e limestone bedrock. e limestone block is overt hrust upon t h e ysc h, w hic h acts as a less permeable h ydrogeologic barrier at dept h. To t h e nort h west, t h e limestone plateau slopes down to t h e plain of t h e Soa River. Here t h e karstied bedrock H Y DROLOGIC CONNECTIONS AND D Y NAMICS OF W ATER MOVEMENT IN THE CLASSICAL KARST KRAS AQUIFER ...

PAGE 10

ACTA CARSOLOGICA 37/1 2008 110 lies beneat h a t hick accumulation of alluvium, primarily sand and gravel-sized sediments. To t h e sout h east, t h e Reka River sinks completely, and is t h e major allogenic source of ground-water. e limestone rocks are again contacting ysc h bedrock on t h e sout h east edge. To t h e nort h east, t h e valley of t h e Vipava River forms anot h er boundary. is river ows also on ysc h bedrock, h owever it loses a portion of its ow near t h e conuence wit h t h e Soa. To t h e sout h east, t h e bedrock units of ysc h and limestone are bounded by t h e Adriatic Sea. e primary inputs of water into t h e system are: 1. precipitation on t h e Kras 2. t h e sinking of t h e Reka River 3. t h e Soa River sinking into its alluvial plain 4. t h e Vipava River sinking partially into its bed 5. t h e Raa River and ot h er small sinks on t h e east ern edge of t h e Kras (ep h emeral sources). e t hirty-year mean annual precipitation (19611990) values measured at several meterological stations across t h e Kras region is approximately 1400 mm/yr (data from t h e Hydro-meteorological Survey of Slove nia). e surface area of t h e h ydrologic basin delineated in Fig. 1 is approximately 440 km 2 not including t h e al logenic drainage basin of t h e Reka River 3 Muc h of t h e primary precipitation inltrates directly into t h e aquifer due to t hin soils and sparse vegetation (Krancj, 1997), h owever a mean value for annual evapotranspiration h as been estimated to be as hig h as 45% of precipitation (Polli, 1971). e ow of t h e Reka River h as a minimum disc h arge of 0.16 m 3 /sec, a maximum disc h arge of 387 m 3 /sec, and an average disc h arge of 9.0 m 3 /sec. e Reka River may contribute at most less t h an one-t hird of t h e Timavo springs total annual ow, and wit h an average baseow of 1.0 m 3 /sec, t h e Reka can contribute a maximum of 11% of t h e Timavo springs baseow (Rojek, 1996). Losses of t h e Soa River to t h e Kras aquifer h ave been estimated at 20-25 m 3 /sec during hig h ow, and between 6-10 m 3 /sec during low ow (Morgante et al., 1966). Losses of t h e Vipava h ave been gaged at an aver age of 1.0 m 3 /sec. Muc h of t his water presumably mixes wit h t h e water of t h e alluvial aquifer supplied by t h e Soa River, and ultimately resurges at t h e major springs. e primary outputs of t h e system are: 1. t h e Timavo spring group 2. Sardos spring 3. Mosc h enizze Nort h and Sout h springs 4. Lisert and Sablici springs 5. t h e springs of Doberd polje 5. Aurisina spring 6. submarine springs along t h e Adriatic coastline (disc h arge unknown) 7. evapotranspiration 8. pumping ground-water for municipal use All of t h e springs are artesian and perennial. e samples for t h e new data presented in t his study were collected at t h e following locations: Timavo: at t h e rise of t h e 1 st branc h Sardos: at t h e small canal across from t h e Roman ru ins on t h e property of t h e Trieste W ater and Gas Utility (A.C.E.G.A., Trieste) Sablici : at Sablici canal w h ere it passes beneat h t h e hig h way Doberdo: at t h e large blue-h ole rise located on t h e nort h western end of t h e polje. Moschenizze North: at t h e rise located on t h e property of A.C.E.G.A. Trieste. Moschenizze South: along t h e spring run just below t h e source Aurisina spring was not sampled during t h e study due to diculty gaining access. e water from t h e springs of Doberd polje drains t hroug h t his polje, across anot h er smaller polje (Pietrarossa polje) and eventually into Sablici canal. us t h e disc h arge of Sablici canal repre sents t h e nal outlet of t h e upstream system t h at includes t h e springs of Doberd and Pietrarossa poljes, as well as Sablici springs. Hereaer t his ow will be collectively re ferred to as Sablici springs. Aer t h e construction of a new marina along t h e coast, t h e Lisert springs are now submerged, and t h eir sampling is no longer possible. Gemiti (1984a) estimated t h e average disc h arge of t h e major springs named above as follows: Sablici + Lisert = 2.2 m 3 /sec Mosc h enizze + Sardos = 2.4 m 3 /sec Timavo = 30.2 m 3 /sec Aurisina = 0.3 m 3 /sec Total outow = 35.1 m 3 /sec 3 Ot h er publis h ed estimates of t h e Kras aquifer rec h arge area vary between 763 km 2 (Civita et al., 1995) to 1000 km 2 (Mosetti, 1966). ese include areas of bot h autogenic and allogenic rec h arge in t h e estimate. e total allogenic rec h arge area h as not yet been welldened, t h erefore only t h e surface area for autogenic rec h arge is reported h ere, and t h e estimate of 440 km 2 from Kranjc (1997) is used. D ANIEL H. D OCTOR

PAGE 11

ACTA CARSOLOGICA 37/1 2008 111 Using t h ese values, Civita et al. (1995) presented an accounting of t h e water balance of t h e region and calcu lated a 10% dierence between inputs and outputs, wit h more rec h arge t h an disc h arge. ese aut h ors concluded t h at a satisfactory water balance h ad been attained. How ever, t h eir estimate did not account for 1) losses due to evapotranspiration, 2) allogenic contributions from t h e Soa River, 3) contributions from t h e ep h emeral Raa River and ot h er nearby sinking ep h emeral streams along t h e nort h eastern border of t h e Kras, 4) outow of t h e submarine springs along t h e coastline, and 5) pumping by water supply wells. us, t h e water balance needs to be re-evaluated, and s h ould include measured or quantied estimates of t h e missing quantities. e water tracing work and t h e h ydrologic mass balance researc h previously conducted on t h e region in dicate t h e following: 1) e boundaries of t h e system h ave for t h e most part been well-dened, h owever a complete water bal ance h as yet to be constructed, mainly due to ungaged submarine spring outow. 2) e Reka River signicantly contributes to t h e Timavo disc h arge only under ood conditions, and for brief durations (several days to a few weeks). 3) e majority of t h e baseow of t h e Timavo springs is supplied by a source h ypot h esized to be t h e Soa River. In t h e conclusion of his book, Timavo: Esplorazioni e Studi Galli (1999) wrote: Aer ve hundred years of study, the hydrogeologic system of the Timavo is suciently known in its general characteristics (coexistence and interactions of conduit drainages and diuse drainages), and in the complex re lations of water fed by various catchment basinsboth karstic and uvialthat are discharged through its resur gence zone. Important unknowns, however, still await an answer, such as the quantitative determination of dier ent discharge contributions through alternating seasonal hydrological cycles and during the baseow and ood regimes, the determination of the storage capacity of the karst massif, the delineation of the recharge basin and of mixing between various reservoirs, and the determination of the course of principle drainage. e original work described in part II of t his paper was initiated prior to Gallis book being publis h ed; h ow ever, t h e goals of t his researc h were aimed toward provid ing answers to some of his unknowns, specically: 1) To test t h e h ypot h esis t h at t h e Soa-Vipava sinking zone contributes a large proportion of water to t h e disc h arge of t h e karst resurgence zone during dry periods, w hile during wet periods t h e resurgence is fed more by local precipitation. 2) To quantify volumes of storage in t h e karst aquifer based on recession analysis of t h e Timavo spring disc h arge, and to relate c h emistry wit h ow regimes in order to distinguis h t h e sources of water contributing to t h e disc h arge in t h ose ow regimes. 3) To estimate t h e transit time of t h e Soa River water to t h e springs and wells of t h e Kras resurgence zone based on measured isotopic tracers. II. RESULTS OF A T W O Y EAR STUD Y OF THE KRAS H Y DROGEOLOG Y USING NATU RAL CHEMISTR Y AND ISOTOPES AS TRACERS W ater samples from t h e major rec h arge and disc h arge points of t h e Kras aquifer were collected twice per mont h during t h e years 1999-2000, and more frequently (at least daily) during large storm events in October, 2000. e approac h taken in t his work was to rst dene t h e ow regimes of t h e aquifer resurgence, t h en to examine t h e c h emical variability wit hin eac h of t h e ow regimes. Fo cused sampling was conducted during storm events to examine s h ort-term c h anges in t h e c h emical composi tion of t h e aquifer disc h arge. Details of t h e design, met h ods, and selected results of t h e study are presented in Doctor (2002), Doctor and Alexander (2005), and Doc tor et al. (2006). Here, a summary of t h e major ndings is presented. D ETERMINATION OF AQUIFER FLO W REGIMES AND STORAGE VOLUMES OF THE T IMAVO SPRINGS In order to quantitatively dene t h e ow regimes consid ered h ere, a h ydrograp h analysis of t h e Timavo springs disc h arge was performed. Six years of disc h arge records were available, from 1995-2000. From t h e six-year record, six of t h e larger recession periods were c h osen. e re H Y DROLOGIC CONNECTIONS AND D Y NAMICS OF W ATER MOVEMENT IN THE CLASSICAL KARST KRAS AQUIFER ...

PAGE 12

ACTA CARSOLOGICA 37/1 2008 112 cession ows at t h e Timavo springs were t by a series of linear segments t hroug h ordinary linear regression in semi-log space (Fig 4a). e simple exponential decay re lation (Maillet, 1905) provides an adequate model for t h e analysis of all disc h arge regimes at t h e springs. From t h e h ydrograp h recession analysis, a Master Recession Curve (MRC) was constructed by averaging t h e recession coef cients (b) and initial disc h arge (q o ) of t h e longest in dividual event recessions from t h e long-term disc h arge records of t h e Timavo springs. Four distinct segments to t h e Timavo MRC were identied, eac h correspond ing to a c h aracteristic ow regime. e breaks in slope dene t h e disc h arge limits of eac h ow regime. e in tersections between t h ese recession segments were used to delimit four ow regimes of t h e Kras aquifer outow: (1) ood ow, (2) hig h ow (3) moderate ow, and (4) baseow. e ood ow regime exists at disc h arges of t h e Timavo above 50 m 3 /s ( m 3 /sec), hig h ow is between 30 and 50 m 3 /s, moderate ow is between 15 and 30 m 3 /s, and baseow disc h arge is below 15 m 3 /s (Fig.4b). e total recession volume of t h e Timavo springs was estimated by integrating across t h e entire MRC, from ood ow to baseow (up to 10,000 days), and is approx imately 585 million m 3 (Doctor and Alexander, 2005). Storage volumes of individual recession segments were also estimated by integrating across eac h segment and subtracting t h e volumes of t h e lower ow regimes across t h e same time interval. From t his analysis, 88.5% is esti mated to be from t h e baseow regime, about 518 million m 3 (Table 2). For comparison, Civita e t al. (1995) esti mated t h e average annual outow of t h e Timavo springs to be 952.4 million m 3 /yr, or a volume 1.8 times as great as t h e baseow storage volume estimated by t h e h ydro grap h recession. According to t h e karst aquifer classica tion proposed by El-Hakim and Bakalowicz (2007), t h e Timavo springs would fall into category 4, h aving a large regulating capacity k of 0.54. is category of karst aqui fer is described as h aving a well-karstied inltration zone and an extended conduit network ending in ooded p hreatic zone (El-Hakim and Bakalowicz, 2007). is is a tting description of t h e Kras aquifer. Alt h oug h t h e Timavo springs drain a distinct ow system from t h e ot h er springs in t h e resurgence zone, it is assumed t h at t h e ow regimes at t h e Timavo springs serve as an adequate proxy for t h e ow regimes of t h e local aquifer resurgence zone as a w h ole due to t h e pre dominant h ydraulic connection t hroug h t h e p hreatic ar tesian aquifer. G EOCHEMICAL AND ISOTOPIC CHARACTERIZATION OF THE K RAS AQUIFER RESURGENCE ZONE Figure 5a s h ows t h e oxygen and dissolved inorganic car bon (DIC) isotope compositions of Sardos spring, Ti mavo spring (1s t branc h) and W ell B-4, wit h t h e possible end-member sources of t h e Soa River, Reka River, and well B-3 also plotted. Most of t h e spring and well samples fall wit hin t h e range of t h e Soa River and well B-3 endmembers, w hile t h e Reka River is not a distinct end-mem ber source for most of t h e samples; h owever, t h e carbon isotopes are not perfectly conservative tracers, t h erefore (a) (b) F ig.4: (a)Representative recession segments from long-term recession in 1999, and (b) master recession curve compiled from 6 long term recession curves of the Timavo springs(1995-2000). D ANIEL H. D OCTOR

PAGE 13

ACTA CARSOLOGICA 37/1 2008 113 t h e interpretation is limited. Figure 5b s h ows t h e average stable oxygen and h ydrogen isotopic compositions of t h e waters of t h e resurgence zone plotted along wit h t h e Global Meteoric W ater Line (GMWL). e mean oxy gen isotope composition of t his well (-6.5) is nearly equivalent to t h e value of weig h ted mean annual rain fall (-6.4) measured at t h e Sela na Krasu meterological station during t h e study pe riod. us, well B-3 serves as a good indicator of t h e inte grated autogenic rec h arge on t h e karst. Wh en viewed in times series, t h e results of t h e iso topic sampling s h ow a repeti tive pattern: during rec h arge events, t h e 18 O composi tions s hi toward hig h er val Recession segment Flow regime Discharge range (m 3 /sec) (day -1 ) q o (m 3 /s) Recession Period (days) Storage volume (m 3 ) % of total storage 1 Flood ow > 50 1.64 x 10-1 101 0 (peak Q) 4 0.06 x 108 1.0% 2 High ow 30 to 50 4.10 x 10-2 62 4 17 0.13 x 108 2.2% 3 Moderate ow 15 to 30 1.70 x 10-2 41 17 58 0.48 x 108 8.2% 4 Baseow <15 3.00 x 10-3 18 58 >10,000 5.18 x 108 88.5% Total: 5.85 x 108 100% Tab. 2. Storage volumes estimated from recession analysis of the Timavo springs F ig. 5. P anel A: 18 O versus 13 C DIC isotope values of selected waters. P anel B: Average oxy gen and hydrogen stable isotopic compositions of all Kras waters sampled. Ranges of the measured compositions are shown with the vertical and horizontal bars. A linear least-squares regression through the mean ground water data is provided, and the Global M eteoric Water Line (GMWL) is shown for reference. H Y DROLOGIC CONNECTIONS AND D Y NAMICS OF W ATER MOVEMENT IN THE CLASSICAL KARST KRAS AQUIFER ...

PAGE 14

ACTA CARSOLOGICA 37/1 2008 114 ues w hile t h e 13 C-DIC compositions s hi toward lower values. e Soa River and well B-3 values bracket t h e compositions of t h e major springs and well B-4 t hroug h out t h e study period. Taking t h e Soa River and well B3 compositions as end-member sources, t h e source of aquifer outow s his from water derived from t h e Soa River toward water derived from autogenic rec h arge on t h e Kras plateau during rec h arge events (Fig. 6). In spite of t h e known connection between t h e Reka River and t h e Timavo springs, t h e data do not indicate t h at t h e Reka River controls t h e long-term seasonal variability in t h e composition of t h e Timavo springs; h owever, t h e Reka River may signicantly inuence t h e Timavo springs during ood ow events. e Timavo variability is sim ilar to t h at of t h e ot h er springs of t h e resurgence zone w hic h do not h ave a con nection to t h e Reka River, t h erefore is it likely t h at t h e seasonal variability of all t h e springs is controlled by mix ing between similar sources but in dierent proportion for eac h spring. For example, Mosc h enizze Nort h spring s h ows t h e lowest 18 O values t hroug h out t h e year, indicat ing a greater proportion of water derived from t h e Soa River t h an t h e ot h er springs. In contrast, Timavo s h ows t h e lowest 13 C DIC values t hroug h out t h e year, indicat ing a greater overall contri bution from t h e autogenic rec h arge component t h an t h e ot h er springs (represented by well B-3). e frequency of sampling during most of t h e study period was too coarse (mont h ly to bi-weekly) to closely evaluate all individ ual storm events; h owever, samples collected during t h e storm events of October 2000 were collected at a hig h er fre quency. ese results will be discussed later. In addition to t h e oxy gen and carbon isotopes, Ca 2+ Mg 2+ Cl temperature, pH, total DIC, and specic conductance were measured on t h ese samples (Doctor, 2002). Alt h oug h t h e results are not s h own h ere, a twocomponent mixing model is not sucient to explain all of t h e variability. However, t h e data can in most cases be completely circumscribed by a model employing t h e fol lowing t hree end-members: 1. Well B-3(karst water) 2. Soa River water 3. a high Cl component According to t h e c h emistry data, t h e most distin guis hing factor of t h e t hird end-member is a hig h er c h loride concentration. is end-member is inferred to be a sub-component of t h e karst water end-member t h at is c h emically conditioned by a source of contamina F ig. 6: Oxygen ( 18 O) and carbon ( 13 C DIC ) isotope record of selected sampling points during the study period. Note that the compositions of the well B-3 and the Soa River bracket the composi tions of the major springs and the well B-4. Discharge of the Timavo Springs is provided for refer ence. e dashed line indicates the limit of the ood ow regime at 50 m 3 /sec. D ANIEL H. D OCTOR

PAGE 15

ACTA CARSOLOGICA 37/1 2008 115 tion. W ater from nearby caves in t h e region (t h e Cave East of t h e Train Station in Monfalcone and t h e Timavo Fis h pond Cave) exhibit hig h er c h loride concentrations t h an t h e ot h er ground-waters, and may be indicative of t h e c h loride source. Despite t h e proximity of t h ese caves to t h e sea, t h e water c h emistry is not consistent wit h a seawater source. e hig h-c h loride concentrations are likely to be t h e result of t h e impact of ant hropogenic activities on t h e ground-water. It is known t h at caves in t h e region near to t h e Klarii pumping station were once used for military storage, and may be a possible source of t h e contamination (F. uteri, personal communica tion, 2007). Alt h oug h t h e hig h-c h loride source water was never directly sampled in t his study, t h e water sample wit h t h e hig h est measured c h loride was t h at collected from t h e Timavo Fis h pond Cave (Gemiti, 1994), t h us its composition was presumed to be indicative of t h e t hird end-member. e c h emical and isotopic data for eac h spring and well were grouped according to t h e ow regime in w hic h t h ey were collected across t h e two years of sampling, and principal components analysis (PCA) was performed on t h e data. Hig h ow and ood ow were combined into a single ow regime (Hig h ow) since very few data were collected during ood ow conditions. Subsequently, end-member mixing analysis (EMMA) was performed using t h e compositions of t h e end-members listed above (see Doctor et al., 2006 for details on t h e met h od). e results are presented in Table 3. e Timavo springs were treated dierently in t his analysis because t h e Reka Riv er was not considered as an end-member in t h e model. erefore, any c h emical and isotopic data obtained dur ing periods of ood ow (greater t h an 50 m 3 /sec) at Ti mavo were excluded to remove periods during w hic h t h e Reka River may h ave h ad a dominant inuence. Contri butions to Timavo from t h e Reka River are in general dif cult to separate using t his approac h given t h e lack of a distinct end-member composition to t h e Reka River (e.g., see Fig. 5). erefore, under lower ow regimes it is assumed t h at contributions from t h e Reka River are contained wit hin t h e proportion of t h e karst water component estimated for Timavo. is is likely given t h at t h e subterranean owpat h enters t h e p hreatic por tion of t h e aquifer downstream of Trebiciano s h a, and t h e composition of t h e Reka River measured upstream would likely be masked by mixing wit h water in p hreatic storage. Considering t h e results of previous quantitative tracer tests, t h e Reka contribution would probably not exceed 20% under t h e majority of ow conditions, except ood ow (Fig. 3). According to t his analysis, water from t h e Soa River provides t h e largest component of ow to t h e re surgence zone springs under all ow regimes. e ot h er major proportion is from t h e karst water end-mem Spring Flow Regime Calculated End-member proportions Soa River Karst water Anthropogenic Component Sardos Low ow 72% 20% 8% Mean ow 76% 16% 8% High ow 53% 35% 12% Mosc. South Low ow 72% 20% 8% Mean ow 63% 26% 11% High ow 51% 34% 15% Mosc. North Low ow 88% 3% 10% Mean ow 77% 10% 13% High ow 74% 12% 14% Doberdo Low ow 85% 12% 2% Mean ow 81% 11% 7% High ow 76% 16% 8% Sablici Low ow 88% 9% 4% Mean ow 82% 8% 10% High ow 78% 8% 15% Timavo Low ow 57% 38% 5% Mean ow 53% 40% 6% High ow* 49% 44% 7% *Excludes data from the ood ow periods, i.e. discharge greater than 50 m 3 /sec Tab. 3: Estimated mixing proportions of end-members in Kras ground-waters from the PCA/EMMA model using Mg, Ca, Cl, SO 4 18 OH 2 O, 13 C DIC H Y DROLOGIC CONNECTIONS AND D Y NAMICS OF W ATER MOVEMENT IN THE CLASSICAL KARST KRAS AQUIFER ...

PAGE 16

ACTA CARSOLOGICA 37/1 2008 116 ber represented by well B-3, wit h relatively minor con tributions from t h e hig h Clcomponent. e greatest c h ange in proportions of end-members occurs at hig h ow, w h en t h e contributions from well B-3 and from t h e hig h Clend-members increase as t h e Soa proportion decreases. e Sardos/Mosc h enizze Sout h and t h e Sa blici/Doberd/Mosc h enizze Nort h spring groups s h ow similar proportions of eac h end-member for eac h ow regime. e proportion of Soa River water is greatest in t h e Sablici/Doberd/Mosc h enizze Nort h group, ranging from 74% at hig h ow to 88% at low ow. W it hin t his group, only Mosc h enizze Nort h and Sablici s h ow sig nicant impact from t h e ant hropogenic component endmember (up to 15%), w hile Doberd s h ows a maximum proportion of ant hropogenic component water of 8% at hig h ow. e Sardos/Mosc h enizze Sout h springs s h ow a muc h hig h er proportion of karst water across all ow regimes t h an t h e Sablici/Doberd/Mosc h enizze Nort h group. is indicates t h at more of t h e Sardos/Mosc h eniz ze ow is supplied by autogenic rec h arge water, w h ereas t h e Sablici/Doberd/Mosc h enizze Nort h springs are fed primarily by water from t h e Soa River owing t hroug h t h e p hreatic zone. W ell B-4 stands apart from all of t h e ot h er groundwaters. It exhibits t h e greatest variation among propor tions of Soa River and karst water wit h c h anging ow regime. At low ow, nearly 75% of t h e water is supplied by t h e Soa River, wit h t h e remaining 25% split evenly between t h e hig h Cl component and karst water endmembers. is proportion c h anges rapidly during storm events, wit h t h e Soa River component replaced by an increasing proportion of karst water wit h rising h ead level, followed by subsequent replacement of t h e karst wa ter by t h e hig h Cl component during water level recession (Fig. 7). is progression re peats wit h additional rainfall, h owever t h e hig h Cl compo nent eventually dominates t h e well c h emistry at t h e hig h est water levels (Doctor et al., 2006). Using t h e results of t h e PCA/EMMA analysis, it was possible to calculate t h e pro portions of eac h of t h e t h ree end-members in t h e average gaged outow of t h e resur gence zone by multiplying t h e average proportion of eac h component estimated for eac h of t h e springs by t h eir average disc h arge and sum ming (Table 4). In summary, t h e Soa River provides approximately 56.3% of t h e total gaged outow of t h e Kras aquifer resurgence under average ow conditions; t h is represents a disc h arge of 19.8 m 3 /sec, a value t h at is nearly equivalent to t h e losses of t h e Soa River (20 m 3 / sec) gaged downstream of t h e city of Gorizia (Mosetti & DAmbrosi, 1963). e proportion of Soa River wa ter in t h e outow at t h e major resurgent springs slig h tly increases at low ow and decreases at h ig h ow. e unimpaired karst water, or stored autogenic rec h arge, provides 36.3% of t h e outow under average ow con ditions. e impaired autogenic rec h arge, or ant h ropo genic component, contributes approximately 7.4% of t h e outow water. In lig h t of t h ese results, t h e previous water balance calculation for t h e Kras aquifer by Civita et al. (1995) seemed to h ave underestimated t h e contribution of t h e Soa River component. Alt h oug h t h e Soa River con tribution to t h e Kras aquifer outow h as been quantita tively estimated in t his study, bot h evapotranspiration as well as t h e combined disc h arge of t h e submarine springs along t h e Adriatic coast h ave yet to be adequately esti mated and probably represent signicant components of t h e overall water balance. erefore, a complete water balance for t h e Kras aquifer awaits quantication of t h ese two components. F ig. 7: Changing proportions of aquifer components at the water supply well B-4 during storm events of October 2000 (from Doctor et al., 2006). D ANIEL H. D OCTOR

PAGE 17

ACTA CARSOLOGICA 37/1 2008 117 DY NAMIC CHANGES IN AQUIFER OUTFLO W DURING STORM EVENTS A detailed analysis of c h anging source contributions to t h e water supply well B-4 at Klarii during storm events of October 2000 was presented by Doctor et al. (2006). Here, t h e results of t h e Timavo springs and Sar dos springs sampled during t h e same storm events are presented. Sampling at t h e Timavo and Sardos springs was not as frequent as at t h e well B-4, t h us t h e records at t h ese springs are not as detailed. Nonet h eless, impor tant trends can be identied in t h e data. e results of t h e storm event sampling at Timavo are presented grap hical ly in Fig. 8 and Fig. 9. Two distinct peaks in t h e Timavo springs disc h arge are evident; t h e rst occurred aer t h e rain event of Oct 1-2, and t h e second resulted from t h e rain t h at fell from Oct 10-12. Based upon t h e 18 O values, t h e waters of t h e Tima vo springs, Sardos spring, and t h e well B-4 s h are a similar source under low ow, but diverge during rec h arge events (Fig. 8). Sardos spring and well B-4 respond more rapidly to storm pulses t h an t h e Timavo springs, and likely gain t h e majority of t h eir ow from t h e s h allow epip hreatic aquifer, w hile t h e Timavo springs are supplied by deeper, p hreatic ow pat h s and respond more gradually to rain fall events. A dominant inuence of t h e Reka River was not ob viously observed at t h e Timavo springs during t h e storm events of October 2000 (Fig. 9). However, t h e trends in t h e data obtained s h ow t h at t h e overall 18 O composition of t h e Timavo springs s hied toward t h at of t h e well B-3, or karst water during t h e rst storm event (from Oct. 2 to Oct. 12), t h en gradually s hied to lower values t h at may h ave approac h ed t h e composition of t h e Reka Riv er later in t h e recession period. Alternatively, t h e s his could h ave been related to mixing between a Soa River component and t h e well B-3 component. Based upon prior tracing tests, t h e transit time of Reka River water to t h e Timavo springs for a ood of Water Source: Mean Flow (m3/sec): Estimated proportion of Endmembers in the outow: Quantity of discharge apportioned to each end-member (m 3 /sec): Soa River Karst Water Anthropogenic Component Soa River Karst Water Anthropogenic Component Timavo 30 53% 40% 7% 15.9 12.0 2.1 Mosc. South + Sardos 2.0 70% 21% 9% 1.4 0.42 0.18 Sablici + Mosc. North + Lisert 3.0 80% 10% 10% 2.4 0.30 0.30 Well B-4 0.1 60% 12% 28% 0.06 .012 .028 TOTAL: 35.1 19.8 12.7 2.6 TOTAL PROPORTION: 56.3% 36.3% 7.4% Tab. 4: Quantication of end-member proportions in Kras aquifer outow under average ow conditions. t his magnitude would likely h ave been on t h e order of several days. It is interesting t h at as t h e Reka River 18 O composition abruptly decreased, t h e Timavo composi tion continued a steady increasing trend (Fig. 9). e longer t h an expected lag time in t h e isotopic response may be t h e result of t h e storm events h aving occurred aer a prolonged dry period. is indicates t h at t h e Reka River, alt h oug h constantly sinking into t h e aquifer of t h e Kras, does not contribute signicantly to t h e ow of t h e Timavo springs in t h e lower ow regimes, in spite of t h e rapid transit time of t h e river water to t h e springs during oods. T RANSIT TIME ESTIMATION OF S O A R IVER W ATER TO THE K RAS RESURGENCE ZONE Cross-correlation time series analysis was performed on t h e mont h ly 18 O values of t h e Soa River and t h ose of t h e Timavo springs, Sardos spring, and t h e pumping well at Klarii from 1984-1988 and from 1998-2000 (Fig.10). e requirements for cross correlation of time series are t h at: (1) t h e series must be a continuous set of data taken in equal time increments (daily, weekly, mont h ly, etc.); (2) no missing values are allowable for any given portion of t h e series; (3) trends in t h e data set may be removed; and (4) t h e data must be standardized. e rst require ment was dicult to meet wit h t h e raw 18 O data t h at were available. First, two separate time series existt h e data publis h ed by Flora and Longinelli (1989) and Pezdi et al. (1986) collected between 1985-1988, and t h e data generated between 1998-2000 by Doctor (2002). us, eac h data set was analysed separately, and t h e results compared. e most complete set of data for bot h periods ex ists at t h e mont hly time interval, t h erefore t h is analysis was conducted using mont hly data. However, due to t h e vagaries of eld work water samples were not all collected at equal time intervals (number of days), nor at all places on t h e same day in any given mont h us, in order to H Y DROLOGIC CONNECTIONS AND D Y NAMICS OF W ATER MOVEMENT IN THE CLASSICAL KARST KRAS AQUIFER ...

PAGE 18

ACTA CARSOLOGICA 37/1 2008 118 complete t h e time series, mont hly means were computed to account for multiple data collected during t h e same mont h in dierent years (i.e., all 18 O measurements of samples taken during any single mont h were averaged and t h e result used as t h e value for t h at mont h ). For a se ries t h at was lacking in data over two or more successive mont h s, t h at series was terminated at t h e point w h ere t h e interruption occurred. For a series t h at lacked a mont h s wort h of data between data from a preceding and follow ing mont h t h e missing value was estimated by taking t h e average of t h e two values immediately preceding and fol lowing t h at mont h In t h is way, t h e overall structure of t h e data was preserved. e Reka River is apparently of signicance in aecting t h e Timavo springs only during ood events, and even t h en only for a relatively s h ort du ration (from several days up to t h ree weeks). e met h od h ere was examining mont hly trends over years; t h erefore, t h e Reka River time series was not included in t h is analysis. For t h e 18 O isotopic time series examined h ere, bot h linear and periodic trends are evident, h owever only t h e linear trends were removed. is was accom plis h ed t hroug h isolation of t h e trend by a least-squares regression, and subsequent subtraction of t h e trend from eac h time series. e data were standardized according to t h e met h od of conversion to z-scores by subtracting t h e sample mean and dividing by t h e sample standard devia tion, suc h t h at t h e mean of t h e z-score series is 0 and t h e standard deviation is 1.0. e gaged losses of t h e Soa River (~20 m 3 /sec) are signicant enoug h to s h i t h e isotopic compositions of t h e springs, w h ose average combined disc h arge is ap proximately 32 m 3 /sec. e cross-correlations of t h e 18 O time series h ave s h own t h at t h e transit time of t h e Soa River water to t h e Timavo and Sardos springs and well B-4 is on t h e order of 1-2 mont h s. e relative transit time of t h e river water to eac h of t h e resurgence points is on t h e order of 1 mont h to t h e Timavo springs, 1-2 mont h s to Sardos, and 2 mont h s to t h e well at Klarii. us, Ti mavo is aected by Soa River water faster t h an t h e ot h er two sampling points. is would indicate t h at t h ere ex ists a deep, well integrated conduit network t h at extends beneat h t h e Sardos and W ell B-4 conduits, but is none t h eless connected to t h em by less-transmissive pat h ways. is is in accordance wit h t h e aquatic speleological ex ploration of t h e Timavo springs. e main spring branc h h as been dived to a dept h of 82 m below sea level, and t h e divers discovered a very large room, t h e dimensions of w h ic h were never completely determined (Guglia, 1994). F ig. 8: Oxygen ( 18 O) and carbon ( 13 C DIC ) isotope record of selected sampling points during storm events,October 2000. Discharge of the Timavo springs is provided for reference. D ANIEL H. D OCTOR

PAGE 19

ACTA CARSOLOGICA 37/1 2008 119 Judging from t h e publis h ed map of t h e exploration, it is evident t h at t h e main conduits t h at feed t h e Timavo h ave a widt h greater t h an 10 m in some sections. Travel times on t h e order of 1-2 mont h s are not un reasonable. e straig h t-line distance from t h e sinking reac h of t h e Soa to t h e springs is approximately 12 km. An estimated transit time of 2 mont h s is t h us equivalent to a mean ow velocity of 0.0023 m/sec, or approximate ly 200 m/day. is is a typical value for karstic conduits (Ford & W illiams, 1989). In t h e zone w h ere t h e Soa and Vipava Rivers begin to lose water underground, t h e t hickness of t h e alluvial ll directly beneat h t h e conu ence of t h e two rivers var ies between 50 m and 70 m (Morgante et al., 1966). us, t his is a minimum distance t h at t h e sinking surface water must ow before it encoun ters t h e karstied limestone bedrock. Taking t h e mini mum t hickness of 50 m as t h e distance t h e water must traverse in 2 mont h s before encountering fast ow in large karst conduits, t h e ow velocity t hroug h t h e alluvial aquifer is t h us on t h e order of 0.83 m/day. W ell logs from t h e alluvial aquifer of t h e Soa indicate t h at t h e sedi ments vary between coarse conglomerates and gravels to ner sand and clay mix tures (Morgante et al., 1966). erefore, it is possible t h at t h e river water sinks rela tively slowly t hroug h t h e al luvial sediments before mov ing quickly t hroug h conduits in t h e limestone bedrock to t h e springs. Assuming t h at once t h e water h as entered t h e conduit system it is able to reac h t h e outlet at a spring in matter of days, t h en t h e transit time is mostly a func tion of t h e movement of t h e water t hroug h t h e alluvial aquifer in t h e river plain. It is evident t h at t h e Soa River h as a persistent, widespread inuence on t h e ground-wa ter resurgence of t h e Kras over a s h ort time scale. Direct tracing of t h e Soa River from w h ere it begins to sink downstream of Gorizia to t h ese resurgence points could verify t h e transit time estimated by t h e time series analy sis. is could be accomplis h ed using articial tracers injected directly into bore h oles t h at bypass t h e alluvium and t h at intersect deep voids wit hin t h e karstied bed rock (see Bidovec, 1967). F ig. 9: 18 O, temperature, and conductivity of the Timavo springs (1 st branch), Reka River, Sardos spring, Soa River and well B-3 during storm events, 2000. Total discharge of the Timavo springs is provided for reference. H Y DROLOGIC CONNECTIONS AND D Y NAMICS OF W ATER MOVEMENT IN THE CLASSICAL KARST KRAS AQUIFER ...

PAGE 20

ACTA CARSOLOGICA 37/1 2008 120 F ig. 10: Standardized, de-trended time series of mean monthly 18 O data from the Soa River, Timavo springs, Sardos spring, and Well B-4. P anel A shows data from 1984-1988. P anel B shows data from 1998-2000. CONCLUSION In t h e case study presented h ere, t h e natural tracers present in t h e water yielded t h e information t h at led to identication and quantication of t h e primary sources to t h e Kras aquifer resurgence. Alt h oug h decades of wa ter tracing wit h articial tracers h ave proven t h e con nection between t h e Timavo springs and t h e Reka River, t h ose eorts drew attention away from t h e major source of water to t h e springs: t h e Soa River. It is estimated h ere t h at t h e Soa River supplies 75% of t h e ow to t h e smaller springs and water supply wells of t h e resurgence zone and 56% of t h e ow at t h e Timavo springs under average ow conditions. e Reka River apparently dom inates t h e ow at t h e Timavo Springs for only brief peri ods (several days) during ood events. In eect, t h e Reka River punctuates t h e longer-term trend in water qual ity of t h e Timavo springs, and h as little or no eect on t h e ot h er ground waters of t h e aquifer resurgence zone. Alt h oug h t h e Reka River vanis h es totally at t h e kocjan Caves, its ow comprises a fraction (less t h an 15%) of t h e Timavo spring ow under average ow conditions. Some of t h e Reka ow emerges at springs furt h er sout h east along t h e Adriatic coast, as proven by previous trac ing tests. It is possible t h at a signicant proportion of t h e Reka ow is lost to submarine spring disc h arge under average and low ow conditions. is h ypot h esis awaits testing t hroug h furt h er study. e Kras aquifer contains a water component t h at appears to be impacted by ant hropogenic activities as evidenced by hig h Cl and SO 4 2concentrations from a non-seawater source. e Brestovica water supply (well B-4) is most aected by t his component, especially dur ing hig h ow conditions w h en Cl concentrations reac h over 50 ppm and t h e estimated ant hropogenic propor tion is on t h e order of 30%. e source is apparently linked to t h e local s h allow circulation in and around t h e nort h west resurgence zone; h owever, t h e exact source is unknown. Furt h er measurements of t h e springs, wells, and caves in t h e vicinity of Monfalcone is encouraged to better c h aracterize and delineate t h e extent of t his hig hCl component, and to develop well-h ead protection strategies for t h e supply well. Experimental dye tracing from local caves intersecting t h e water table to t h e well may s h ed lig h t on t h e pat h ways of overow to t his well during rec h arge events. For regional aquifers, tracing wit h dyes or ot h er ar ticial tracers may not be sucient to construct a water balance or c h aracterize t h e storage capacity of t h e aqui fer. e advantages of tracing wit h dyes or ot h er articial tracers are t h at a successful trace 1) identies connec tions between t h e points of tracer input and t h e points of tracer recovery, 2) provides information on t h e veloc ity of ow between t h ose points, and 3) provides infor mation on t h e dilution or loss of tracer during transit. However, dye tracing describes only t h e fast-ow conduit portions of a karst aquifer, w hic h are in fact t h e zones of least storage potential. erefore, dye tracing alone is not an eective means of carrying out a regional-scale water resource investigation t h at requires a quantitative esti mation of t h e proportion of outow derived from more diuse aquifer storage wit hin karst systems c h aracterized by a large p hreatic component suc h as t h at of t h e Kras region. Dye traces s h ould be combined wit h detailed D ANIEL H. D OCTOR

PAGE 21

ACTA CARSOLOGICA 37/1 2008 121 c h emical and isotopic analyses of frequently collected water samples w h enever possible to observe t h e natural variation in environmental c h emistry. In conclusion, it is h oped t h at t h e results from t his investigation may pro vide a useful case study to aid in answering questions for future sustainable management of t h e water resources of ot h er karst aquifers. ACKNO W LEDGMENTS Several people were instrumental in providing t h e sup port and guidance for t his researc h, particularly E. Calvin Alexander, Jr. of t h e University of Minnesota, Minneap olis; Milena Horvat and Sonja Lojen of t h e Joef Stefan Institute, Ljubljana; Janko Urbanc from t h e Geological Survey of Slovenia, Ljubljana; Antonio Longinelli of t h e University of Parma; and W illibald Stic h ler of t h e Insti tute for Hydrology, Neu h erberg. I must also acknowl edge t h e aid of several key persons and agencies wit h out w h om I would not h ave accomplis h ed t his researc h. ey are:, Fabio Gemiti from A.C.E.G.A. Trieste, Franco Cuc c hi, Onelio Flora, and Laura Genoni from t h e Univer sity of Trieste, Metka Petri and Janja Kogovek from t h e Karst Researc h Institute, Postojna, t h e Kraki Vodovod, Seana, Francesca Serra and Reno Semararo from Ge okarst Engineering s.r.l., Trieste, and Agnes omen from t h e Healt h Institute of Koper. Funding was provided by t h e U.S. Fulbrig h t and David L. Boren fellows hip pro grams, and t h e University of Minnesota, Dept. of Geol ogy & Geop h ysics. REFERENCES A.C.E.G.A. Trieste (1988) Il problema dellacqua nella provincia di Trieste. Trieste: Arti Grac h e Smolars S.p.A, 28 pp. Bidovec, F., 1967: e h ydrosystem of karstic springs in t h e Timavo basin. In: Hydrology of F ractured Rocks, vol 1. AIHS: London, 263-274. Boegan, E., 1938. Il Timavo Studio sullidrograa car sica subaerea e sotterranea. Mem. Ist. Ital. Di Spel., serie Geol. E Geogr., Mem. II., Trieste. Cancian, G., 1987: Lidrologia del Carso goriziano-tri estino tra lIsonzo e le risorgive del Timavo. Studi Trentini di Scienze Naturali, 64, 77-98. Cancian, G., 1988: Signicato idrologico della concent razione di ossigeno e anidride carbonica nelle acque sotteranee tra il lago Doberd e le risorgive del Ti mavo. Mondo Sottorraneo, Udine, 12(1/2), 11-29. Civita, M., Cucc hi, F., Eusebio, A., Garavoglia, S., Maran zana, F. & Vigna, B., 1995: e Timavo h ydrogeo logic system: an important reservoir of supplemen tary water resources to be reclaimed and protected. Acta Carsologica, 24, 169-186. Cucc hi, F., Pirini Radrizzani, C., & Pugliese, N., 1987: e carbonate stratigrap hic sequence of t h e Karst of Trieste (Italy). Mem. Soc. Geol. It. (Proc. Int. Symp. Evolution of Karstic Carbonate Platform: Relation wit h ot h er Peridadriatic Carbonate Platforms), 40, 35-44. Cucc hi, F., Giorgetti, F., Marinetti, E., Kranjc, A., 1997: Experiences in monitoring Timavo River (Classical Karst). In: Tracer Hydrology 97 (A. Kranjc, ed.) A.A. Balkema: Rotterdam, 167-172. Cucc hi, F. Forti, P., Marinetti, E., and Zini, L., 2000: Re cent developments in knowledge of t h e h ydrogeol ogy of t h e Classical Karst. Acta Carsologica, 29, 55-78. Cucc hi, F., Marinetti, E., Potleca, M., and Zini, L., 2001. Inuence of geostructural conditions on t h e speleo genesis of t h e Trieste karst (Italy). Geologica Bel gica, 4(3-4), 241-250. Cucc hi, F., and Zini, L., 2002: Underground Timavo river monitoring (Classical Karst). Acta Carsologica, 31(1), 75-84. DAmelio, L., Flora, O., Longinelli, A. (1994) Environ mental isotope data: oxygen isotope concentration in precipitation in N-E Italy (Friuli-Venezia Giulia). Miner. Petrogr. Acta, vol. 37, pp. 113-124. Dansgaard, W ., 1964: Stable isotopes in precipitation. Tellus, 16: 436-468. Deines, P., Langmuir, D., and Harmon, R. (1974) Stable carbon isotope ratios and t h e existence of a gas p h ase in t h e evolution of carbonate ground waters. Geoc himica et Cosmoc himica Acta, vol. 38, p.11471164. H Y DROLOGIC CONNECTIONS AND D Y NAMICS OF W ATER MOVEMENT IN THE CLASSICAL KARST KRAS AQUIFER ...

PAGE 22

ACTA CARSOLOGICA 37/1 2008 122 Doctor, D.H., 2002: e Hydrogeology of the Classical Karst (Kras) Aquifer of Southwestern Slovenia. P h.D. dissertation, University of Minnesota, Minneapolis, 212 p. Doctor, D.H., Lojen, S. and Horvat, M., 2000: A stable isotope investigation of t h e Classical Karst aquifer: Evaluating karst ground-water components for wa ter quality preservation. Acta Carsologica, 29 (1), 79-92. Doctor, D.H., Alexander, E.C., Jr., Petri, M., Kogovek, J., Urbanc, J., Lojen, S., and Stic h ler, W ., 2006: Quan tication of karst aquifer disc h arge components t hroug h end-member mixing analysis using natural c h emistry and stable isotopes as tracers. Hydrogeol ogy Journal, 14, 1171-1191. Doctor, D.H. and Alexander, E. C., Jr., 2005: Interpreta tion of water c h emistry and stable isotope data from a karst aquifer according to ow regimes identied t hroug h h ydrograp h recession analysis. U.S. Geo logical Survey Karst Interest Group Proceedings, Rapid City, Sout h Dakota, September 12-15, 2005 (E. L. Kuniansky, ed.). USGS Scientic Investigations Report 2005-5160, p. 82-92. El-Hakim, M., and Bakalowicz, M., 2007: Signicance and origin of very large regulating power of some karst aquifers in t h e Middle East: Implication on karst aquifer classication. Journal of Hydrology, 333, 329-339. Eriksson, E., Mosetti, F., Hodoscek, K., Ostanek, L., 1963: Some new results on t h e karstic h ydrology wit h t h e employ of tritiated water as tracer. Boll. Geof. Teor. Appl., 5(17), 18-32. Eriksson, E., 1983: Stable isotopes and tritium in pre cipitation. In: Guidebook on Nuclear Tec hniques in Hydrology. IAEA Tec hnical Reports Series no. 91, Vienna: International Atomic energy Agency. pp. 19-27. Flora, O. and Longinelli, A., 1989: Stable isotope h ydrol ogy of a classical karst area, Trieste, Italy. In: Isotope Techniques in the Study of F ractured and F issured Rocks, International Atomic Energy Agency (IAEA) : Vienna, 306 pp. Ford, D.C. and W illiams, P.W ., 1989: Karst Geomorp h ol ogy and Hydrology. New Y ork: C h apman & Hall, 601 pp. Gabrovek, F. and Peric, B, 2006: Monitoring t h e ood pulses in t h e epip hreatic zone of karst aquifers: t h e case of Reka river system, karst plateau, SW Slove nia, Acta Carsologica, 35(1), 35-45. Galli, M., 1999: Timavo: Esplorazione e studi. Supple mento no. 23 di Atti e Memorie della Commissione Grotte Eugenio Boegan, Trieste, 195 p. Gams, I., 1993: Origin of t h e term karst, and t h e trans formation of t h e Classical Karst (Kras). Environ mental Geology, 21, 110-114. Gemiti, F., and Licciardello, M., 1977: Indagini sui rap porti di alimentazione delle acque del Carso tries tino e goriziano mediante lutilizzo di alcuni trac cianti naturali. Annali Gruppo Grotte Ass. XXX Ott., sez. C.A.I. Trieste, 6, 43-61. Gemiti, F., 1984a: La portata del Timavo alle risorgive di S. Giovanni di Duino. Annali Gruppo Grotte Ass. 30Ott., Trieste, 7, 23-41. Gemiti, F., 1984b: Nuova ed originale prova di marcatura delle acque del Timavo. Annali Gruppo Grotte Ass. 30Ott., sez. C.A.I. Trieste, 7, 43-62. Gemiti, F., 1994: Indagini idroc h emic h e alle risorgive del Timavo. Atti e Memorie della Commissione Grotte E. Boegan, 30, 73-83. Gemiti, F., 1996: Portata liquida e portata solida del Ti mavo alle risorgive di S. Giovanni di Duino. Hy droresAnnuario 1995, Trieste, 13, 74-88. Gemiti, F., 1998: Marcatura delle acque del Timavo a seguito di un versamento di idrocarburi nella valle della Recca e interpretazione delevento mediante lutizzo di dati meteorologici, idrologici, idroc hi mici. Annali Gruppo Grotte Ass. XXX Ott, 10, 930104. Guglia, P., 1994: Risultati esplorativi del Progetto Timavo (1990-1993). Atti e Memorie della Commissione Grotta E. Boegan, 31/1992-93, 25-48. Habi, P., 1989: Kraka bifurkacija Pivke na jadransko rnomorskem razvodju (Pivka karst bifurcation on Adriatic-Black Sea waters h ed), Acta Carsologica, 18, 233-164. Kranjc, A. (ed.), 1997: Slovene Classical Karst-Kras. Postojna: Institut za raziskovanja krasa ZRC SAZU, 254 pp. Krivic, P., 1981: Etude hydrodynamique dun aquifre karstique ctiere: le Kras de Slovenie, Y ougoslavie. Accadmie Montpellier, Univ. Sc. Tec hn. Langued oc, se de Docteur-Ingnieur Universit Mont pellier II: 108 pp. Krivic, P., 1982a: Variations naturelles de niveau pizo mtrique dun aquifre kartsique. Geologija, 25(1), 129-150. Krivic, P., 1982b: Transmission des ondes de mare travers laquifre ctier de Kras. Geologija, 25(2), 309-325. Krivic, P., 1983: Interprtation des essais par pompage raliss dans un aquifre karstique. Geologija, 26, 149-186. D ANIEL H. D OCTOR

PAGE 23

ACTA CARSOLOGICA 37/1 2008 123 Krokos, A., 1998: Ulteriori studi geochimico-isotopici su alcune sorgenti carsiche costiere dellarea triestina: considerazioni idrologico-ambientali. Bac h elors esis in Geological Sciences, University of Trieste, Italy. Maillet, E., 1905. Essais dHydraulique Souterraine et Fluviale. Herman, Paris, France, 218 pp. Morgante, S., Mosetti, F., Tongiorgi, E.,1966: Moderne in dagini idrologic h e nella zona di Gorizia. Bollettino di Geosica Teorica ed Applicata, 8(30), 114-137. Mosetti, F., 1965: Nuova interpretazione di un esperi mento di marcatura radioattiva del Timavo. Bollet tino di Geosica Teorica ed Applicata, 2(8). Mosetti, F., 1966: Lo stato delle attuali conoscenze sullidrologia carsica e relative ripercussioni sul problema dellalimentazione idrica di Trieste. Atti del Museo Civico di Storia Naturale, Trieste, 25(4): 73-105. Mosetti, F. & DAmbrosi, C., 1963: Alcune ricerc h e pre liminari in merito a supposti legami di alimentazi one fra il Timavo e lIsonzo. Bollettino di Geosica Teorica ed Applicata., n. 17. Polli, S., 1971: Quattro anni di meteorologia etc. Atti e memorie della Commissione Grotta E. Boegan, vol 10, Trieste. Pezdi, J., Dolenec, T., Krivic, P., Urbanc, J., 1986: Envi ronmental isotope studies related to ground-water ow in t h e central Slovenian karst region, Y ugosla via. 5 t h International Symposium on Underground W ater Tracing (SUW T), At h ens, 91-100. Ravbar, N., 2004: Drinking water supply from karst water resources (e example of t h e Kras plateau, SW Slo venia). Acta Carsologica, 33(1), 73-84. Rojek, D., 1996: Velika VodaReka, a karst river. Acta Carsologica, 25, 193-206. Timeus, G., 1928: Nei misteri del mondo sotterraneo. Risultati delle ricerc h e idrogeologic h e sul Timavo 1895-1914, 1918-1927. Alpi Giulie, vol. 29, 1. In: Atti e M em. Comm. Grotte E. B oegan, 22, 117-133. Urbanc, J., and Kristan, S., 1998: Isotope investigation of t h e Brestovica water source during an intensive pumping test. RMZ Materials and Geoenviron ment, 45(1-2), 187-191. H Y DROLOGIC CONNECTIONS AND D Y NAMICS OF W ATER MOVEMENT IN THE CLASSICAL KARST KRAS AQUIFER ...



PAGE 1

GROUND W ATER FLO W IN CR Y STALLINE CARBONATES JESENIK Y MTS., CZECH REP.: USING STREAM THERMOMETR Y AND GROUND W ATER BALANCE FOR CATCHMENT DELINEATION P ODZEMNI TOK V KRISTALINSKIH KARBONATIH POGORJE J ESENIK Y, EKA R EP .: U PORABA TOKOVNE TERMOMETRIJE IN BILANCE PODZEMNE VODE ZA OMEJNITEV ZALEDIJ Jan K UKA KA 1,3 & Viola A LTOV 2 & Jit B RUTHANS 3 & Ondtej Z EMAN 3, 4 Izvleek UDK 556.3(437.1) Jan Kukaka, Viola Altov, Ji Bruthans & Ondej Ze man: Podzemni tok v kristalinskih karbonatih (pogorje Jese niky, eka Rep.): Uporaba tokovne termometrije in bilance podzemne vode za omejnitev zaledij Ozki pasovi metamorfozorani h karbonatni h kamnin na obmoju kontaktnega krasa pogorja Jeseniky na ekem de lujejo kot vodonosnik, ki drenira obseno obmoje kristalin ski h kamnin, predvsem litov. V karbonatni h kamnina h so pomembne zaloge podzemne vode, ki ji h deloma izkoriajo za vodooskrbo. Podrobne meritve temperature in prevodnosti predstavljajo skupaj z merjenjem pretokov vzdol vse h tokov znotraj obmoja relativno hitro metodo za doloitev poloaja praktino vse h pomembni h iztokov podzemne vode iz kar bonatov. Merjenja pretokov tokov, ki prekajo ozke karbonatne pasove, so nam omogoila, da smo doloili poloaj in ocenili izdatnost ponorov in odsekov povrinski h tokov z izgubami ob razlini h hidroloki h pogoji h. Za h valjujo natannem pozna vanju odsekov toka z izgubami in dotoki smo la h ko izbrali ustrezne prole za loitev zaledij z razlinimi hidrolokimi bi lancami (uravnoteena, s preseki, z izgubami). Smeri pre takanja v karbonati h ter obmoja napajanja in praznjenja so bila doloena na osnovi primerjave specini h pretokov posa mezni h prispevni h zaledij. Ugotovljene smeri toka se ujemajo z rezultati sledilni h poskusov na tem obmoju. Predstavljeno metodo je mono na razlini h obmoji h uporabiti za doloitev poloaja skriti h dotokov v tokove in oceno toka med posa meznimi manjimi prispevnimi zaledji, vsaj deloma pa la h ko nadomestijo sledilne poskuse, saj omogoajo oceno smeri toka na osnovi hidroloke bilance in geometrije kamnin. Kljune besede: kras, karbonat, apnenec, hidrogeologija, podzemna voda. 1 Dekonta a.s., Volutov 2523, 158 00 Prague 5, Czec h Republic, jan.kukacka@centrum.cz 2 C h arles University in Prague Faculty of Science, Department of P h ysical Geograp h y and Geoecology, Albertov 6, 128 43 Prague 2, Czec h Republic, viola.altova@centrum.cz 3 C h arles University in Prague, Faculty of Science, Department of Hydrogeology, Engineering Geology and Applied Geop h ysics, Albertov 6, 128 43 Prague 2, Czec h Republic, brut h ans@natur.cuni.cz 4 PROGEO s.r.o., Tic h dol 113, 252 63 Roztoky, Czec h Republic, ondrejzeman@seznam.cz Received/Prejeto: 21.11.2007 COBISS: 1.01 ACTA CARSOLOGICA 37/1, 125-131, POSTOJNA 2008 Abstract UDC 556.3(437.1) Jan Kukaka, Viola Altov, Ji Bruthans & Ondej Zeman: Groundwater ow in crystalline carbonates (Jeseniky mts., Chech Rep.): Using stream thermometry and groundwater balance for catchment delineation Strips of metamorp h osed carbonate rocks in a contact-karst area in t h e Jeseniky Mts, Czec h Republic, act as aquifers, drain ing broad areas of crystalline rocks, mostly p h yllites. Signicant groundwater resources t h at are partly used as a water supply are in carbonate rocks. Detailed temperature and conductivity measurements coupled wit h disc h arge measurements along all streams in t h e area demonstrate a relatively quick met h od to lo cate virtually all important groundwater outows from carbon ates. Disc h arge measurements of streams crossing carbonate strips enabled us to locate and quantify t h e capacity of ponors and losing parts of streams in various water stages. anks to a detailed knowledge of losing and gaining parts of streams, we were able to select appropriate proles to separate catc hments wit h diering h ydrologic balances (balanced, gaining, losing). Flow directions in carbonates and rec h arge and disc h arge ar eas were delineated by comparing t h e specic disc h arges of individual catc hments. Resulting ow directions agree wit h tracer tests in t h e area. Our outlined approac h can be used in many ot h er areas to locate hidden inows into streams and to estimate ow between individual small catc hments, and it may partly compensate for tracer tests as it allows ow directions to be estimated from h ydrological balance and rock geometry. Keywords: karst, carbonate, limestone, h ydrogeology, ground water.

PAGE 2

ACTA CARSOLOGICA 37/1 2008 126 is study focuses on t h e h ydrogeology of a contactkarst area developed in metamorp h osed carbonates in a mountainous terrain (Jeseniky Mts, Czec h Republic). Signicant water resources in t h e carbonate strips were suggested by eznek (1990). e purpose of t his study was to locate all springs and outows from carbonates, to estimate t h e ow directions in carbonate rocks, and to delineate rec h arge and disc h arge areas. ermometry, a simple but hig h ly eective met h od, was used to locate and quantify concentrated disc h arge. Flow directions were estimated using h ydrology balance calculations rat h er t h an tracer tests due to hig h numbers of rec h arge and disc h arge points and t h e relatively simple geometry of t h e carbonates. G EOGRAPH Y AND G EOLOG Y OF STUD Y A REA e study area is located in t h e nort h western part of t h e Jesenky Mountains in t h e nort h eastern Czec h Repub lic (Fig. 1) at altitudes of 400 1,000 m a.s.l. e mean annual air temperature (3 to 7C) depends on altitude. Annual precipitation ranges from 880 mm at lower eleva tions to 1,109 mm on mountain summits. Mean regional runo, based on data from t h e Bl River at t h e Mikulov ice gauging station (located ca. 14 km from t h e conu ence of t h e Stat and t h e Bl River; gure not s h own), is 18.5 ls -1 km -2 Net groundwater rec h arge (mean annual groundwater runo) is estimated at 5 to 10 ls -1 km -2 in lowand hig h-elevation terrains, respectively (Krsn et al., 1982). e study area is composed of various rocks (p h yllite, marble, quartzite, crystalline sc hist) of t h e Brann Group, w hic h underwent regional low-grade metamorp hism during t h e Variscan orogeny. Indi vidual lit h ologies form strips and lenses, w hic h are elongated in t h e NNE-SSW direction (Fig. 1). Rock strips are interrupted by many NW -SE-trending transverse faults (eznek, 1990). Carbonates h ave lower amounts of impurities; 60% of samples h ave CaCO 3 contents >90%. Carbonates are neto medium-grained (grain size 0.1 0.2 mm). e porosity of exploited carbonates ranges be tween 0.7 1.1% (Toul and Augusta, 1972). e carbonates are karstied. Supercial karst p h enomena are scarce and restricted to ponors and several dolines, w h ile underground karst is more evolved. Sixty caves are known in t h is area, wit h a total lengt h of passages of about 2 km. e longest one, t h e Na Pomez Cave ( 1,000 m), is open to t h e public (Krl, 1958; Pano, 1961). e h ydrogeological setting can be c h aracterized as h ard rock wit h two main aquifers: I NTRODUCTION J AN K UKA KA & V IOLA A LTOV & J I B RUTHANS & O NDEJ Z EMAN F ig. 1: e study area and carbonates of the B rann Group (geology simplied af ter V oclka, 1973).

PAGE 3

ACTA CARSOLOGICA 37/1 2008 127 GROUND W ATER FLO W IN CR Y STALLINE CARBONATES METHODOLOG Y Scattered karsts of t h e Brann Group and t h eir specic conditions demand special approac h es. First, a great deal of information about geology, h ydrology, h ydrogeology, and t h e terrain h ad to be collected and studied before eld work could be done. GPS met h ods and topograp hic maps (1: 10,000 scale) were used for eld measurements. Results were analyzed using GIS soware. D ISCHARGE MEASUREMENT Tortuous ow trajectories in mountainous streams (boulders in riverbeds) preclude using current-velocity meters to measure stream disc h arge. W e used dissolved NaCl as a tracer to measure disc h arge (e.g., Kilpatric and Cobb, 1985). e essence of t his met h od is injection of a tracer of known mass (in our case dissolved NaCl) into a stream and measuring t h e concentration downstream. e concentration of added NaCl was measured by elec trical conductivity (Cond 340i, W T W co, Germany, wit h automatic measurement period 5 sec and automatic com pensation to 25C). For eac h injection prole, a calibra tion was performed relating injected NaCl and c h anges in conductivity. ree measurements were typically done for eac h prole and t h e average value was used (for large dierences t h e measurement was discarded). C OUPLED TEMPERATURE CONDUCTIVIT Y AND DISCHARGE MEASUREMENTS OF STREAMS AND INFLO W S It is known t h at most drainage from karst aquifers is via concentrated disc h arge at springs (Atkinson, 1977). Muc h of t h e karst disc h arge drains directly into streams and t h us cannot be measured directly at springs. Suc h disc h arges and hidden springs can aect t h e tempera tures and conductivities of streams; t h us, t h e disc h arges can be located and t h eir yields quantied by measuring stream temperature and electrical conductivity c h ang es. e spacing of measurements was c h osen accord ing to stream ow (faster temperature c h anges in small streams). At eac h location, bot h temperature and con ductivity were measured at at least t hree points, in t h e middle and on bot h sides of t h e stream. At springs and ot h er observable inows, t h e inow conductivity and temperature were also measured. Measurements were taken in winter and summer, w h en t h e largest contrasts between stream and spring temperatures exist. A consid erable advantage of winter measurements is t h e constant temperature of t h e streams (0oC) and t h e presence of ice crystals in parts of t h e stream indicating t h e absence of inows, t h us visually enabling faster progress (Brut h ans and Zeman, 2001). e yield of hidden inow can be calculated from a formula based on t h e law of conservation of mass and energy (Brut h ans and Zeman, 2001): Q 1 = l s -1 ] (TV 2 x Q 2 )(TV tot x Q 2 ) TV tot TV 1 (1) Explanation: Q 1 unknown yield of hidden inow (ls1 ) TV 1 temperature of hidden inow (C) Q 2 stream disc h arge above hidden inow (ls -1 ) TV 2 stream temperature above hidden inow (C) TV tot stream temperature below hidden inow (C) 1) karstied carbonate strips wit h dept h s exceeding 100 m (according to well geop h ysics), and 2) s h allow crystalline rock systems signicantly per meable to only a few tens of meters below t h e ground surface (bgs). Carbonate strips are continuous for distances of several kilometers and are dissected by important trans verse faults in places (eznek, 1990). Longitudinal faults s h ow low permeability. Major NW SE-trend ing transverse faults, w hic h divide carbonate strips into separate groundwater basins, are h ydrodynamically sig nicant. Some transverse faults drain surrounding crys talline rocks into carbonates (eznek, 1990). Based on well pumping tests, transmissivities of t h e carbonates are hig h ly variable (1.10 -6 1.10 -1 m2s -1 ; Brut h ans, 2006). e detailed karst h ydrology of t h e area was stud ied by Pano (1961, 1962), w h o performed several tracer tests in t h e karst strips of t h e Bran Group. Zeman et al. (2003) performed 3 tracer tests in t h e karst strips in t h e nort h ern part of t h e study area using sodium c h loride dye met h ods. e maximum ow velocity in karst con duits is typically 0.1 to 6 km day -1 (Zeman et al., 2003) based on t h e tracer tests. e test results suggest relatively large ow volumes and cross-sectional areas of t h e karst conduits. e mean residence time of groundwater in t h e crystalline rocks is several years based on a study of two springs and one stream using t h e environmental tracers 3 H, 18 O, CFCs, and SF6 (Brut h ans, 2006).

PAGE 4

Conductivity can be substituted in Equation 1 in stead of temperature if t h e conservation of conductivity is valid (c h anges of conductivity are due only to mix ing and c h emical reactions are not signicant). In ideal conditions, t h e accuracy of disc h arge measurement by t his met h od can be % (Brut h ans and Zeman, 2001) depending on t h e accuracy of temperature and stream disc h arge measurements. It is important to note t h at, in many cases, combining disc h arge and mixing approac h es yields better estimates of inows t h an using disc h arge measurements alone. Simply subtracting t h e measured upstream disc h arge from t h e measured downstream dis c h arge can produce larger errors for hidden spring in ows t h an using t h e met h ods described above (Brut h ans and Zeman, 2001). E STIMATING GROUND W ATER FLO W GEOMETR Y IN KARST AQUIFERS e study area was divided into 54 catc hments on t h e basis of stream temperature, conductivity, and disc h arge measurements. ArcGIS 8.3 was used to facilitate data processing. Topograp hic maps (1:10,000), geologic maps (1:25,000), and GPS measurements were used as a base. Individual catc hments were delineated in GIS and can be generally classied as one of t h e following types: I) Balanced catchments, w h ere no signicant groundwater ow to and from ot h er catc h ments occurs. Only catc h ments wit h out carbonates were selected for t h is group. In noncarbonate rocks, t h e surface catc h ment is believed to coincide wit h t h e groundwater basin as t h e water table is s h allow and generally follows sur face topograp h y. Disc h arge measurement proles of t h e F ig. 2: Estimation of ground water ow direction in karst aquifers. basin stream were made upstream of t h e rst carbonate strip. II) Losing catchments, w h ere groundwater losses prevail. Suc h areas included ponors and/or losing stream segments. In some cases, h owever, t h e stream does not lose water, but part of t h e groundwater does not arrive at t h e stream, instead being diverted to ot h er catc hments via carbonate strips. Areas were delineated suc h t h at t h ey do not include signicant groundwater inows from ot h er catc hments. III) Gaining catchments, w h ere groundwater gains prevail. Suc h areas included karst springs and gaining parts of streams. Areas were delineated suc h t h at t h ey do not include groundwater outows to ot h er catc hments. Disc h arge was measured over s h ort time periods (a few days) during stable water stages (no rain events) on tens of t h e proles, w hic h separate individual catc h ments. Measurements were made repeatedly during dif ferent water stages (low and medium water stages). Figure 2 demonstrates h ow groundwater balance for individual catc hments was calculated from disc h arge measurements. Balanced catc hments A and D (Type I noncarbonate rocks) were used to calculate specic run o for t h e measurement period. Runo was divided by catc hment area and averaged. In a second step, balanced specic runo from A and D were used to calculate ex pected runo from ot h er catc hments (B and C) by mul tiplying t h e balanced specic runo by catc hment area. For C, t h e measured runo is about 29 ls1 w hic h is less t h an t h e calculated balanced runo. Hence, about 25 ls -1 of water is lost to ot h er catc hments via carbonates (Type II). For B, t h e measured runo is about 80 ls -1 w hic h is greater t h an t h e balanced runo. Hence, about 29 ls -1 of water probably arrives via carbonates from ot h er catc h ments (Type III). Taking into account t h e geometry of car bonate strips, t h e water prob ably ows from catc hment C to catc hment B. Because runo is also aected by altitude, t h e aver age altitude was determined for eac h catc hment and t h e balanced runo was calcu lated based on data from bal anced catc hments situated at dierent altitudes. ACTA CARSOLOGICA 37/1 2008 128 J AN K UKA KA & V IOLA A LTOV & J I B RUTHANS & O NDEJ Z EMAN

PAGE 5

F ig. 3: Temperature and conductivity measurements on Ramzovsk B rook. Many newly found sink h oles and ponors suggest a larger lateral extent of carbonate rocks t h an currently s h own on existing geological maps. Detailed temperature and conductivity measure ments coupled wit h disc h arge measurements along all streams h elped us locate and quantify concentrated groundwater inows t h at were large enoug h to c h ange stream temperature and/or conductivity. Hidden inows (groundwater owing directly to streams) represent tens of percents of total groundwater inows. Moreover, sev eral major springs were located by t h ermometry t h at were unknown or known only to local residents. Many of t h e springs are partly drained directly to streams. is dem onstrates t h e advantages of t h ermometry for locating im portant groundwater outows from carbonates relatively quickly. For example, t h e abrupt c h anges of temperature and conductivity of Ramzovsk Brook clearly demon strates t h e presence of inows (Fig. 3). Disc h arge measurements of streams above and be low carbonate strips h elped us locate and quantify t h e ca pacities of ponors and losing parts of streams at various water stages. Losing parts of streams are generally located at minor tributaries and h eadwaters. In some cases, t h e ponors c h ange to springs during hig h ow (Lesn tvr catc hment). Detailed knowledge about losing and gaining parts of streams determined during previous investigations al lowed us to select appropriate proles to separate catc h ments wit h dierent h ydrologic balances (balanced, gaining, losing, see Fig. 4). About 90 measurements of disc h arge were made over 10 periods during various wa ter stages. It was not possible to measure groundwater run o from carbonates alone. W e t h erefore used t h e same groundwater runo rates for carbonates as for ot h er rocks. is is normally a source of error in balance cal culations, as carbonates tend to h ave hig h er groundwater runo rates relative to non-carbonate rocks. Because t h e surface area of carbonates is small relative to ot h er rocks in most of t h e catc hments (less t h an 10% of catc hment area), we t hink t h at t his simplication is reasonable and does not result in signicant error. e main areas of groundwater rec h arge and dis c h arge were located, and t h eir total yield was measured. Some of t h e spring groups h ave mean yield around 100 ls1 e largest ponors h ave capacities around 70 ls -1 (Fig. 4). Flow directions in carbonates and regional rec h arge and disc h arge areas were delineated by comparing specif ic disc h arge of individual catc hments wit h karst strip ge ometries (Fig. 4). A regional h ydrogeological waters h ed was located near Ramzov Saddle based on additional data from well water levels. Flow directions estimated from catc hment balance calculations agree well wit h results of tracer tests in areas w h ere data from bot h met h ods are available. RESULTS AND DISCUSSION ACTA CARSOLOGICA 37/1 2008 129 GROUND W ATER FLO W IN CR Y STALLINE CARBONATES

PAGE 6

ACTA CARSOLOGICA 37/1 2008 130 F ig. 4: e groundwater ow direction of study area, the most important springs and ponors and their ow rate and estima tion of the catchment character (loosing, balanced, gaining). CONCLUSIONS Detailed temperature and conductivity measurements coupled wit h disc h arge measurements along all streams in t h e area represent a suitable met h od for locating im portant groundwater outows from carbonates relatively quickly. Measuring disc h arge of streams crossing carbon ate strips enabled us to locate and quantify t h e capacities of ponors and losing parts of streams in various water stages. anks to detailed knowledge about losing and gaining stream reac h es, we were able to select appropriate proles to separate catc hments wit h dierent h ydrologic balances (balanced, gaining, losing). Flow directions in carbonates and rec h arge and disc h arge areas were de lineated by comparing specic disc h arges of individual catc hments. Resulting ow directions agree wit h tracer tests done in t h e area. is outlined approac h can be used in many ot h er areas to locate hidden inows into streams and estimate ow between individual small catc hments. is met h od may partly compensate for tracer tests as it allows ow directions to be estimated from h ydrologi cal balance and rock geometry alone if t h e local situation permits. J AN K UKA KA & V IOLA A LTOV & J I B RUTHANS & O NDEJ Z EMAN

PAGE 7

ACTA CARSOLOGICA 37/1 2008 131 Many t h anks to Vpenn village for nancial support, to CHKO Jesenky and to all persons w h o h elped us in t h e eld. is study was partly supported by t h e researc h project MSM 002 162 0855. W e would also like to t h ank Dr. Alan Mayo, Brig h am Y oung University, USA for im proving t h e Englis h of t h e manuscript. ACKNO W LEDGMENTS REFERENCES Atkinson, T.C., 1977: Diuse ow and conduit ow in limestone terrain in the M endip Hills, Somerset (Great B ritain). J. Hydrol., 35: 93-110 Brut h ans, J., 2006: V yuit pirozench stopova ( 18 O; 3 H; FREONY ; S F 6 ) a dalch metod pro zhodnocen doby zdren vod a charakteru proudrn v krasovch oblastech R.. Disertan prce, Ptrodovdeck fakulta, Karlova univerzita, 203 str., Pra h a, Brut h ans, J., Zeman, O., 2001: Regional permeable zones and stream thermometry in karst areas. 31. IAH Congress: vol. 2: 697-700. Munic h. Kilpatric F.A., Cobb E.D., 1985: M easurement of discharge using tracers.Tec hniques of W ater-Resources In vestigations of t h e United States Geological Survey, Book 3. C h apter A16. 52 p. Krsn J., Knek M., ubov A., Dakov H., Matuka M., Hanzel V., 1982: M apa odtoku podzemn vody SSR.esk h ydrometeorologick stav, 52str., Pra h a Krl, V., 1958: Kras a jeskynr vchodnch Sudet.Acta Univ. Carolinae, Geologica 2., 105-159., Pra h a. Pano V., 1961: Z u den karsthydrographischen P roble men der kleinen Kalksteingebiete in Nordmhren und Schlesien. Mittelungen der Oesterreic hisc h en Geograp hisc h en Gesellsc h a, 103, II: 158-177. W ien. Pano, V., 1962: V sledky koloranch experiment a po zorovn krasovch vod v Severomoravskm kraji. Sbornk vlastivdn h o muzea v Olomouci, oddl A ptrodn vdy 5, 13-60, Ostrava. eznek, V., 1990: Ramzovsk nasunut-krystalinikum Regionln HG przkum I fze, pitn voda. Unpubl. report, Geotest Brno 185 s. Toul, P., Augusta, L., 1972: Z vrren zprva z przkumu a vpotu zsob na loisku vpenc Horn Lipov Pod trat.Unpubl. Report, RD Jesenk n.p., 73 str., Jesenk. Voclka, M., 1973: V sledky geologickho mapovn a souhrn geologickch pomrr v srii B rann s geo logickou mapou v mrtku 1:10 000. Autore fert disertace. Universita J.E.Purkyn, Fakulta ptrodovdeck, 21 str., Brno. Zeman, O., Brut h ans, J, Vojtc h ovsk, A., 2003: Stopo vac zkouky v krasu skupiny B rann v Rychleb skch horch.Geologick vzkumy na Morav a ve Slezsku v r. 2002, 102 105, Brno GROUND W ATER FLO W IN CR Y STALLINE CARBONATES



PAGE 1

INVESTIGATION OF STRUCTURE OF VARIOUS SURFACE KARST FORMATIONS IN LIMESTONE AND DOLOMITE BEDROCK W ITH APPLICATION OF THE ELECTRICAL RESISTIVIT Y IMAGING P ROU EVANJE ZGRADBE RAZLI NIH KRAKIH OBLIK V APNENCIH IN DOLOMITIH Z UPORABO METODE ELEKTRI NE UPORNOSTI TAL Uro STEPINIK 1 Andrej M IHEVC 2 Izvleek UDK 551.435:552.54:537.31 Uro Stepinik, Andrej Mihevc: Prouevanje zgradbe razlinih krakih oblik v apnencih in dolomitih z uporabo metode elektrine upornosti tal Povrinske krake oblike so razvite v razlini h vrsta h kamnine in imajo razlino zgradbo. Avtorja predstavljata rezultate mer itev prolov elektrine upornosti tal vrta in udornic v Slo veniji, ki se na h ajajo v apnenci h in dolomiti h. Izbrane oblike so primeri povrinski h kraki h oblik, ki ji h je mogoe morfoloko opredeliti kot tipine. Za meritve elektrine upornosti tal je bil uporabljen merilec elektrine upornosti SuperSting R1/IP. Metoda elektrine metode je bila uporabljena za ugotavljanje lastnosti materiala pod povrjem v povrinski h kraki h oblika h. Materiali z vejim deleem vode, kot so pretrta kamnina ali ilovica, imajo manjo elektrino upornost kot nepretrta kamn ina. Uporaba metode elektrine upornosti tal se je izkazala kot primerna za podrobno prouevanje povrinski h kraki h oblik. Kljune besede: elektrina upornost tal, kras, geomorfologija, epikras, vrtae, udornice. 1 University of Ljubljana, Department of Geograp h y, Akereva 2, SI-1000 Ljubljana, Slovenia 2 Karst researc h institute ZRC SAZU, Titov trg 2, SI Postojna, Slovenia Received/Prejeto: 30.01.2008 COBISS: 1.01 ACTA CARSOLOGICA 37/1, 133-140, POSTOJNA 2008 Abstact UDC 551.435:552.54:537.31 Uro Stepinik, Andrej Mihevc: Investigation of structure of various surface karst formations in limestone and dolomite bedrock with application of the Electrical resistivity imaging Karst landforms are developed in dierent kind of rocks and in dierent structural settings. Aut h ors are presenting some results of prole measurement of Electrical resistivity imaging in some dolines and collapse dolines t h at developed in lime stone and dolomite in Slovenia. Selected features are examples of surface karst landforms w hic h could be morp h ologically de ned as typical of its kind. For electrical resistivity imaging data collection was used SuperSting R1/IP eart h resistivity meter. Electrical resistivity imaging met h od was applied for researc h of material c h aracteristics beneat h surface of karst landforms. Materials wit h hig h er portion of water as fractured bedrock or clay are less resistant to electrical current t h an unfractured bed rock. Application of t h e ERI met h od h as turned out as appro priate for detailed investigation of surface karst landforms. Keywords: electrical resistivity imaging, ERI, karst, geomor p h ology, epikarst, doline, collapse doline. I NTRODUCTION Karst surface in Slovenia covers area about 44% or 9000 km 2 (Novak, 1993). It developed mainly in limestone and dolomite bedrock. Dolomite covers t h e area of 12% or 2500 km 2 (Gabrovec, 1994) and karst in limestone covers t h e area about 32% or 6500 km 2 Some of t h e karst land forms developed bot h on dolomite and limestone and are covered wit h soil or detritic material. Dierent c h emical and mec h anical composition of limestone and dolomite results in dierent processes and dierent dynamics of processes on t h em. e morp h ology of karst surface and structure of t h e subsurface is essential for understanding of processes on karst. It is possible to image t h e subsur face quickly and inexpensively t hroug h t h e application of various nonintrusive surface geop h ysical met h ods. is paper presents results of Electrical resistivity imaging in some dolines and collapse dolines in lime stone and dolomite on Slovenian karst and interpretation of data. Selected karst landforms are examples of surface

PAGE 2

ACTA CARSOLOGICA 37/1 2008 134 karst landforms w hic h could be morp h ologically dened as typical of its kind. Subsurface structures of various dolines were revealed t hroug h construction works on karst w hic h corresponds to t h e data obtained wit h met h od of electrical conductivity imaging. On t h e ot h er h and we h ave no data about subsurface structures of collapse dolines, so subsurface data of dolines were used to inter pret subsurface data of collapse dolines revealed t hroug h application of electrical resistivity imaging. e met h od turned out as appropriate for a robust visualization of epikarst structure and subsurface structure of karst fea tures. METHODS OF ELECTRICAL RESISTIVIT Y Electrical resistivity imaging h as been successfully uti lized for c h aracterizing t h e subsurface for many years but it h as certain limitations. e met h od was labour intensive, interpretation of data was time intensive and t h e met h od based on individual subjective interpretation (Roman, 1951; Zh ou et al. 2002). Development of computer controlled multi elec trode resistivity survey systems and t h e development of resistivity modelling soware (Locke & Barker, 1996) h ave allowed more cost eective resistivity surveys and better interpretation of t h e subsurface. ese surveys are usually referred as Electrical resistivity imaging (ERI) or Electrical resistivity tomograp h y (ERT) (Zh ou et al., 2002). ese facts allow data to be collected and pro cessed quickly so t h e electrical resistivity imaging sur veys become a valuable tool in subsurface investigations (Zh ou et al. 2000). Electrical resistivity imaging surveys are typically conducted to determine t h e resistivity of t h e subsurface. Resistivity data can be used to determine t h e location of various geologic and soil strata, bedrock fractures, faults and voids. Fundamental to all resistivity met h ods is t h e concept t h at current is impressed into t h e ground and t h e eect of t his current wit hin t h e ground can be measured. e eect of potential or dierences of potential, ratio of potential dierences, or some ot h er parameter t h at is di rectly related to t h ese variables are t h e most commonly measured eect of t h e impressed current. e principal dierences among various met h ods of Electrical resistiv ity lie in number and spacing of t h e current and poten tial electrodes t h e variable quantity determined and t h e manner of presenting t h e results (Eart himager 2003). Carbonate rock in general h as a signicantly hig h er resistivity t h an clayey soil because it h as muc h smaller primary porosity and fewer interconnected pore spaces. Its resistivity value is about 1000 o hm-m (Telford et al. 1990). Clayey materials can h old more moisture and h ave hig h er concentration of ion to conduct electricity, t h ere fore, h ave resistivity values less t h an 100 o hm-m (Telford et al. 1990). e hig h contrast in resistivity values be tween carbonate rock and clayey material favours t h e use of Electrical resistivity met h od for determine t h e bound ary between bedrock and overburden (Zh ou et al. 2000). A frequently occurring problem wit h Eart h resistiv ity imaging is to determine w hic h electrode congura tion will respond best to t h e material c h anges in karst features. Eac h array h as distinctive advantages and disad vantages in terms of sensitivity to t h e material variations, dept h of investigation and signal strengt h. e most typical arrays are dipole-dipole array, W enner array and Sc h lumberger array. e dipole-dipole array gives good h orizontal resolution of data w hile W enner and Sc h lum berger arrays are more directed in vertical resolution. In application on karst surveys t h e dipole-dipole array pro vided hig h est precision of ground c h anges sensitivity and h as greatest sensitivity to vertical resistivity boundaries (Zh ou et al. 2002). STUD Y AREAS AND INTERPRETATION OF DATA Electrical resistivity data were collected in four dierent sites. ose sites include two collapse dolines and two dolines. One collapse doline and one doline are situated in limestone w hile t h e ot h er two are situated in dolomite. Selected karst landforms are typical examples of surface karst landforms w hic h occur on karst in Slovenia. e SuperSting R1/IP eart h resistivity meter de veloped by Advanced Geosciences, Inc. was used for data collection. Survey was conducted wit h dipole-di pole array. e distance between two electrode pairs was between 3 and 5 m. e data was processed to generate two-dimensional resistivity models using Eart hImager 2D resistivity inversion soware developed by Advanced U RO STEPINIK A NDREJ M IHEVC

PAGE 3

ACTA CARSOLOGICA 37/1 2008 135 Geosciences, Inc. e Root-Mean-Square (RMS) error quanties t h e dierence between t h e measured resistiv ity values and t h ose calculated from t h e true resistivity model. A small RMS error value indicates small dier ences. e minimum RMS error in t h e survey was 2.72%; t h e maximum error was 5.8%. e reliable data of two-dimensional resistivity im ages decrease in lateral extent towards greater dept h is means t h at all data in t h e rectangular raster images (Figs. 2,4,6,8) are not relevant. Only data in trapezoidal s h ape is valid, not outer lower sections of t h e rectangular images. D OLINE IN DOLOMITE IN B EGUNJSKI RAVNIK AREA Begunjski ravnik is levelled karst area nort h of Cerkniko polje in central Slovenia. On sout h ern and eastern part of t h e area is Triassic and Jurassic dolomite wit h dip of beds from 15 to 20 degrees towards west. It c h anges to Jurassic limestone towards west (Buser 1963). Hydrologically t h e area is still mainly unexplored. ere are no caves wit h access to epip hreatic or p hreatic zone, w hic h is presuma bly 40 m below t h e surface. Presumably t h e underground ow is in area directed from t h e stream Cerkniica and Cerkniko polje towards nort h to springs of t h e Ljubljanica near Vr hnika. e doline is positioned in lower Jurassic grey sparry dolomite 2 kilo metres nort h east of Begunje in t h e at karst area densely covered by dolines. e doline h as longer diameter 35 m, s h orter diameter 20 m and dept h 7 m, approximate volume of t h e doline is 1100 m 3 e slopes are covered wit h rocky soil. Dip of t h e slopes of t h e doline 40 degrees wit h hig h est dip on western slope wit h 50 degrees. e oor of t h e doline is at and covered wit h clayey soil. Results of Electrical resistivity imaging prole t h roug h t h e doline wit h distance between two electrode pairs 3 metres s h ow t h at under a t h in layer of less resis tant clayey soil and weat h ered rock on t h e slopes of t h e doline wit h resistivity value up to 1000 o h m-m t h ere is more resistant dolomite bedrock wit h resistivity values h ig h er t h an 1000 o h m-m. e oor of t h e doline is lled wit h 10 metres t h ick layer of clayey soil wit h resistiv ity value up to 250 o h m-m, overlaying unfractured do lomite bedrock. Under t h e deepest eastern part of t h e doline oor is a less resistant vertical structure wit h re sistivity value up to 250 o h m-m w h ic h mig h t be fault or vertical karst void lled wit h weat h ered bedrock or clayey material. D OLINE IN LIMESTONE IN L OGAKI RAVNIK AREA Logaki ravnik is a relatively at karst area nort h west of Begunjski ravnik and nort h east of Planinsko polje. Bed rock is mainly lower Cretaceous limestone wit h beds of dolomites wit h dip of beds about 20 degrees toward west. (Pleniar et al. 1963). Underground waters ow from t h e area of Cerkniko polje and Planinsko polje to nort h towards t h e area of t h e Ljubljanica River springs near Vr hnika. According to Habi (1985) piezometric level in t h e area is at t h e on altitude between 350 and 400 metres. Recent explorations revealed t h at in t h e area pi ezometric level is on altitude 420 metres being about 100 metres below t h e surface of Logaki ravnik (Kataster jam JZS 2006). F ig. 1: Location of the investigated sites. F ig. 2: ERI prole of the doline in dolomite in B egunjski ravnik area (ERI prole direction 88 de grees; RMS error 3.6%). INVESTIGATION OF STRUCTURE OF VARIOUS SURFACE KARST FORMATIONS IN LIMESTONE AND DOLOMITE BEDROCK ...

PAGE 4

ACTA CARSOLOGICA 37/1 2008 136 Investigated doline is situated in t h e nort h ern part of Logaki ravnik in Skalen kamen area. e doline was dened by Jovan Cviji (1893) as a typical example of a solution doline in lime stone. e doline is situated in Lower Jurassic compact limestone wit h dip of beds 35 degrees towards sout h west (Buser, 1963). e doline h as longer diameter 40 m, s h ort er diameter 35 m and dept h 7 m, approximate volume of t h e doline is 1600 m 3 e slopes are covered wit h t hin soil. Dip of t h e slopes of t h e doline are approximately 21 degrees wit h hig h est dip on western slope wit h 32 de grees w h ere are bedrock out crops. e oor of t h e doline is deepest in western part and covered wit h clayey soil. Results of Electrical resistivity imaging wit h dis tance between two electrode pairs 3 metres s h ows, t h at inside t h e doline t h ere is an extensive pocket of less re sistant clayey material wit h resistivity value up to 250 o hm-m, w hic h outcrops in eastern slope of t h e doline. Under t h e deepest part of t h e doline is limestone bedrock wit h some less resistant verti cal structures wit h resistivity values between 250 and 1000 o hm-m w hic h mig h t func tion as rainwater and solution runo. Ot h er slopes of t h e doline are covered wit h t hin layer of soil and weat h ered rock wit h resistivity values between up to 1000 o hm-m overlying limestone bedrock wit h resistivity value hig h er t h an 1000 o hm-m. C OLLAPSE DOLINE IN DOLOMITE SOUTH OF K O EVSKO POLJE e area is situated in t h e sout h ern part of Slovenia. Sout h of Koevsko polje is situated in levelled lowland in elevation of about 470 m 490 m. It consists mainly F ig. 3: DEM of the doline in dolomite in B egunjski ravnik area with the ERI prole direction. F ig. 4: ERI prole of the doline in limestone in Logaki ravnik area (ERI prole direction 286 de grees; RMS error 5.8%). F ig. 6: ERI prole of the collapse doline in dolomite south of Koevsko polje (ERI prole direction 97 degrees; RMS error 2.72%). U RO STEPINIK A NDREJ M IHEVC

PAGE 5

ACTA CARSOLOGICA 37/1 2008 137 of limestone and dolomite of Jurassic and Triassic age wit h dip of beds towards sout h and sout h west (Savi & F ig. 5: DEM of the doline in limestone in Logaki ravnik area with the ERI prole direction. Dozet, 1983). e direction of underground water ow is directed from ponors t h at are in elevation of 470 m in sout h ern part of Koevsko polje towards sout h to springs t h at are situated in t h e canyon of t h e Kolpa Riv er (Habi, 1988) in elevation of 200 m. e collapse doline Globoka jama is 8 km sout h east of Koevje above un derground water ow of t h e Rina River. It is situated in upper Triassic dolomite wit h dip of beds 20 degrees toward sout h west (Savi, Dozet, 1983). e collapse doline h as longer diameter 220 m, s h orter diameter 150 m and average dept h of 33 m, ap proximate volume of t h e collapse doline is 0.56 Mm 3 e slopes are covered wit h rocky soil. Dip of t h e slopes of t h e collapse doline is ap proximately 38 degrees wit h hig h est dip on western slope wit h 38 degrees w h ere bed rock is exposed. e oor of t h e collapse doline is at and covered wit h clayey soil. Survey of t h e collapse doline wit h distance be tween two electrode pairs 5 metres revealed t h at bed rock slopes of t h e collapse dolines are covered wit h t h in layer of less resistant soil and weat h ered rock wit h resistivity values up to 1000 o h m-m. ickness of weat h ered material is h ig h er in t h e lower part of t h e slope. Floor of collapse doline is lled wit h clayey soil wit h resistivity values up to 250 o h m-m. e dept h of more conducting material in t h e oor of t h e collapse doline is more t h an 20 metres. Dolomite bedrock un derlying slope material h as resistivity value h ig h er t h an 1000 o h m-m. C OLLAPSE DOLINE IN LIMESTONE IN R AKOV KOCJAN AREA e area of Rakov kocjan is situated between Cerkniko and Planinsko polje w h ere t h e Rak river surface ow emerges. e area mainly consists of lower Cretaceous limestone wit h dip of beds between 10 to 30 degrees to wards west (Buser 1963). Underground water ow is di rected from Cerkniko polje towards NW to Planinsko polje (Habi, Gospodari 1979; Kranjc 1996). F ig. 7: DEM of the collapse doline in dolomite south of Koevsko polje with the ERI prole direction. INVESTIGATION OF STRUCTURE OF VARIOUS SURFACE KARST FORMATIONS IN LIMESTONE AND DOLOMITE BEDROCK ...

PAGE 6

ACTA CARSOLOGICA 37/1 2008 138 F ig. 8: ERI prole of the collapse doline in limestone in Rakov kocjan area (ERI prole direction 254 degrees; RMS error 5.25%). In t h e eastern part of Rakov kocjan area near to t h e entrance of cave sys tem Zelke jame, collapse doline Globoak (Dvojni Globoak) is situated. e bedrock is lower Cretaceous well bedded limestone wit h dip of beds 25 degrees to wards west (Buser 1963). e longer diameter of t h e collapse doline is 250 m, s h orter diameter is 150 m and average dept h is 25 m. Approximate volume of t h e doline is 0.4Mm 3 Upper parts of t h e slopes are verti cal rocky walls; lower parts of t h e slopes are covered wit h scree and rocky soil. Dip of t h e slopes covered wit h scree is approximately 30 degrees. e oor of t h e doline is at and covered wit h clayey soil. Survey was carried out wit h distance between two electrode pairs 5 metres. Resistivity values of lime stone bedrock and scree on t h e slopes is more t h an 250 o hm-m. e oor of t h e col lapse doline is lled wit h less resistant clayey sediment and weat h ered bedrock wit h t hickness more t h an 10 me tres. It h as resistivity values below 250 o hm-m. e t hick ness of clayey sediment lling t h e oor of collapse doline is more t h an 10 metres. F ig. 9: DEM of the collapse doline in limestone in Rakov kocjan area with the ERI prole direc tion. CONCLUSIONS e proles made wit h eart h resistivity imaging clearly s h ow t h e important dierences of t h e analyzed dolines. ey can be interpreted by ot h er knowledge of t h e doline morp h ology and sediment proles w hic h were exposed in similar types of rock in quarries or road cuts. Application of t h e met h od in collapse dolines and dolines in limestone and dolomite bedrock revealed sub surface structure of t his surface karst features. e ap plication revealed t h at resistivity value for carbonate rock is more t h an 1000 o hm-m, in a case of collapse doline in limestone bedrock t h e value is hig h er t h an 250 o hmm. For soil and weat h ered bedrock t h e resistivity values are between 250 and 1000 o hm-m. Clayey material h as resistivity values lower t h an 250 o hm-m. In reality t h ese U RO STEPINIK A NDREJ M IHEVC

PAGE 7

ACTA CARSOLOGICA 37/1 2008 139 t hres h old values may c h ange spatially and temporally due to dierence of subsurface structure and water con tent in bedrock and sediment. Some areas of hig h resis tivity zones wit h resistivity values hig h er t h an 1000 o hmm can be even subsurface open cave-like features. But data obtained by electrical resistivity imaging presented in t his article corresponds wit h structures of various dolines t h at were revealed troug h construction works on karst and wit h data obtained by investigation of col lapse dolines on ot h er parts of Slovenian karst (Stepinik, 2006). Accuracy of subsurface images also decreases wit h hig h er distance between two electrode pairs t h erefore for smaller karst features distance between two electrode pairs s h ould be smaller. ere is a clear dierence between t h e smaller so lution dolines and collapsed dolines. e results of ERI proling in solution dolines s h ow zones wit h important dierences in electric properties of t h e rock. e bottoms of dolines s h ow smaller resistivity. is can be explained as eect of retention of t h e water by sediment. ere are no signs of empty spaces like s h as at t h e bottom and t his bot h means t h at t h e bottoms of dolines are not favourable for vertical transport of sediments, even if t h ere is a hig h vertical gradient in karst. On t h e slopes of t h ese dolines t h ere are zones w h ere t h ere is more resistant rock. ere is no sediment and t h ere are also some empty ssures as bedding planes and s h as possible, but t h ey were not de tected wit h t his ERI survey. ere are also small zones of less resistive rock, possibly cavities or fractured zones wit h clay or sediments containing some pore water. e collapse dolines are dened as suc h because of muc h larger dimensions and we can explain t h eir origin wit h t h e similar features t h at were evidently formed by transformation of underground cavities to surface de pression and were later signicantly modied by slope processes. In bot h collapse dolines we can clearly see t h e dierence between slopes and bottoms. e slopes s h ow hig h ly resistive rock, t h at is limestone or scree wit h out any ner sediment. At t h e bottom t h ere are t hick sedi ments and possibly collapse rubble mixed wit h ne grained sediments wit h larger water storage capacity. e origins of t h at sediment are not clear. ere can be residual soils and clays was h ed to t h e bottom by slope processes or t h ey are remnants of some cave inll t h at appeared at surface aer or during t h e collapse dolines were formed. e investigation presented in t his article s h ow t h at t h e met h od of Electrical resistivity imaging is useful for investigation of structure of various surface karst forms. But due to t h e lack of bore h ole data t h e presented ERI images are still one of many possible interpretations w hic h roug h ly reect subsurface structure of t h e surface karst features. REFERENCES Buser, S., 1963: Tolma osnovne geoloke karte za list Ribnica. Geoloki zavod Ljubljana, pp. 68. Cviji, J., 1893: Das Karstp h aenomen. Versuc h einer morp h ologisc h en Monograp hie. Geogr. Ab h. 5, 3, pp. 113. Eart hImager. 2003: 2D Resistivity and IP Inversion So ware Instruction Manual. Version 1.2.0, Advanced Geosciences Inc. Austin. Gabrovec, M., 1994: Relief in raba tal na dolomitni h obmoji h Slovenije. Doktorska disertacija, pp. 123. Habi, P. & Gospodari R., 1979: Karst p h enomena of Cerkniko polje. Acta carstologica 8, 1, 11-162. Habi, P., 1985: Vodna gladina v notranjskem in primor skem krasu Slovenije. Acta carsologica, 13, 1, 37 74. Habi, P., 1988: Tektonska pogojenost krakega reliefa za h odne Su h e krajine. Acta carsologica, 17, str. 33 64. Kataster jam JZS 2006. Jamarska zveza Slovenije. Kranjc, A., 1986: Cerkniko jezero in njegove poplave. Geografski zbornik, 25, 75 123. Loke, M. H., Barker, R. D., 1996: Rapid least-square in version of apparent resistivity pseudosections by a quasi-Newton met h od. Geop h ys Prospect, 44, 131152. Novak, D., 1993: Hydrogeological researc h of t h e Slove nian karst. Nae jame, 35, 1, 15-20. Pleniar, M., 1963: Tolma osnovne geoloke karte za list Postojna. Geoloki zavod Ljubljana, pp. 53. Roman, I., 1951: Resistivity reconnaissance in American society of testing and materials symposium on surface and subsurface reconnaissance. American society of testing materials special tec hnical publi cation, 122, 121-226. Savi, D., Dozet, S., 1983: Tolma osnovne geoloke karte za list Delnice. Geoloki zavod Ljubljana, pp. 62. INVESTIGATION OF STRUCTURE OF VARIOUS SURFACE KARST FORMATIONS IN LIMESTONE AND DOLOMITE BEDROCK ...

PAGE 8

ACTA CARSOLOGICA 37/1 2008 140 Stepinik, U., 2006: Loamy sediment lls in collapse dolines near t h e Ljubljanica River springs, Dinaric karst, Slovenia. Cave and Karst Science, 33, 3, 105110. Telford W M., Geldart L. P., & R. E. S h eri, 1990: Ap plied geop h ysics (2. edition). New Y ork, Cambridge University press. Zh ou, W ., Beck, B. F. & J. B. Step h enson, 2000: Reliability of dipole-dipole electrical resistivity tomograp h y for dening dept h to bedrock in covered karst terrains. Environmental Geology, 39, 760 766. Zh ou, W ., Beck, B. F., Adams, A.C., 2002: Eective elec trode array in maping karst h azards in electrical resistivity tomograp h y. Environmental Geology, 42, 922 928. U RO STEPINIK A NDREJ M IHEVC



PAGE 1

STUDIES OF THE FAUNA OF PERCOLATION W ATER OF HUDA LUKNJA, A CAVE IN ISOLATED KARST IN NORTHEAST SLOVENIA RAZISKAVE FAVNE PRENIKLE VODE: PRIMER JAME HUDA LUKNJA NA PODRO JU OSAMELEGA KRASA V SEVEROVZHODNI SLOVENIJI Tanja PIPAN 1 Vesna NAVODNIK 2 Franc JANEKOVI 3 Tone NOVAK 3 Izvleek UDK 556.34:59(497.4-18) Tanja Pipan, Vesna Navodnik, Franc Janekovi & Tone No vak: Raziskave favne prenikle vode: Primer jame Huda luknja na podroju osamelega krasa v severovzhodni Sloveniji Z mesenimi vzorevanji smo eno leto prouevali favno in ekologijo zdrub v 12 curki h prenikle vode v Medvedjem rovu Hude luknje pri Doliu. To je najveji jamski sistem tako imenovanega osamelega izoliranega krasa, ki sestoji iz manj kot 1 km 2 do ve kot 10 km 2 veliki h apnenasti h zaplat. Huda luknja je razvita v eni od te h krp, ki meri okrog 1 km 2 imenovani Paki kras, v triasni h apnenci h. Osredotoili smo se na raziskave epikrake favne. Spremljali smo letno dinamiko vrednosti tem perature, prevodnosti, trdote in koncentracije razlini h ionov. V te h zikalni h in kemijski h parametri h so se curki znailno razlikovali med seboj, prav tako voda isti h curkov med letom. Med zdrubami in parametri ni bilo korelacije. V curki h za jeti osebki so bili nakljuno razporejeni in so pripadali tako vodnim kot kopenskim povrinskim, epikrakim in podzemeljskim taksonom. Zastopane so bile iste skupine organizmov kot v jama h Dinarskega krasa, vendar so prevladovali maloetinci, ne ceponoci. Na splono je vrstna pestrost organizmov upada la od prvega vzorevalnega mesta proti notranjosti. Epikraka vodna favna je v Pakem krasu slabo zastopana, vendar ni jasno, ali je to znailnost te krake krpe ali celotnega osamelega krasa, zato so potrebne dodatne raziskave na osamelem krasu. Kljune besede: osameli kras, epikras, prenikajoa voda, bio speleologija, favna. 1 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 2 Florjan 210a, SI-3325, otanj, Slovenia 3 Department of Biology, Faculty of Natural Sciences and Mat h ematics, University of Maribor, Koroka 160, SI-2000, Maribor, Slo venia, E-mail: tone.novak@uni-mb.si, franc.janzekovic@uni-mb.si Received/Prejeto: 01.11.2007 COBISS: 1.01 ACTA CARSOLOGICA 37/1, 141-151, POSTOJNA 2008 Abstract UDC 556.34:59(497.4-18) Tanja Pipan, Vesna Navodnik, Franc Janekovi & Tone No vak: Studies of the fauna of percolation water of Huda luknja, a cave in isolated karst in northeast Slovenia e fauna and community ecology of percolation water was studied using mont h ly samples of 12 drips in t h e Medvedji rov in t h e cave Huda luknja. is is t h e largest cave system in t h e so-called isolated karst w hic h consists of limestone patc h es of <1 to 10 km 2 in size in central and nort h eastern Slovenia. Huda luknja is developed in one of t h ese patc h es measuring about 1 km 2 in t h e Triassic limestonest h e Paka karst. e researc h focused on t h e investigation of t h e epikarst fauna in NE Slo venia. Temperature, conductivity, h ardness and concentrations of various ions in water were measured. Considerable spatial and temporal variation in parameters existed among t h e drips. However, t h ere was no correlation between t h e community structure and t h e parameters. ere are six aquatic species and 19 terrestrial species in t h e epikarst of t h e Paka isolated karst. In t h e drips, individuals of epigean, epikarstic and h ypogean aquatic taxa as well as terrestrial taxa belonging to t h e same groups as t h ose in caves in t h e Dinaric karst were found. Unlike t h e Dinaric karst, t h e most abundant group was oligoc h aetes, not copepods. In general, t h e biotic diversity diminis h ed from t h e entrance to deep in t h e cave. More investigation in t h e iso lated karst is required to decide eit h er t his is t h e specic c h arac teristic of t his karst patc h or a general p h enomenon of isolated karst. Key words: isolated karst, epikarst, percolating water, biospel eology, fauna.

PAGE 2

ACTA CARSOLOGICA 37/1 2008 142 In Slovenia, karst comprises 44% of its land surface (Gams 2004) wit h more t h an 9000 caves registered to date (Cave register IZRK ZRC SAZU, Speleological Association of Slovenia). According to geological, h ydrological and spe leological c h aracteristics t h e karst in Slovenia is divided into t hree groups (Habi 1969): (1) Alpine karst wit h hig h mountain karst, (2) Dinaric karst, divided in hig h and low karst, and (3) intermediate Dinaric-Alpine and isolated karst. e isolated karst is represented by s h allow patc h es of limestones and dolomites of dierent ages and is furt h er divided into several isolated subunits. Eac h of t h em diers from t h e ot h ers in its size, structure of car bonate rocks and h ydrology (Habi 1969, Habe 1972). e evolution of particular karst areas can be, be side t h eir geological, geomorp h ological and h ydrological c h aracteristics, reected also in specic distributional patterns and diversity of t h e h ypogean fauna. Some basic biogeograp hical studies h ave been made to explain t h e processes occurring in t h e Slovenian karst, mainly on Gastropoda and Crustacea (Bole 1972, 1977, 1985, Sket 1993, 1999a, b, 2002, Sket et al. 2004a, Culver et al. 2003, Culver & Pipan 2007) and on hig h ly endemic attemsiid diplopods (cf. Mri et al. 1996). Recently, p h ylogeo grap hic approac h es and molecular tec hniques were used to explain t h e evolution of troglobionts in t h e Dinaric karst in connection wit h t h e historical events wit hin t h e area (Trontelj et al. 2007). In t h e isolated karst, similar studies h ave not been carried out, alt h oug h many biolog ical investigations h ave been made in caves in nort h ern and central Slovenia (Bole 1977, Novak 2005) but t h ey h ave not been summarized. ey are of considerable in terest because t h ey are analogous to islands, w hic h play a very prominent role in biogeograp h y in general (Culver & Pipan in press). In t h e last ten years, especially since 2000, t h e sty gobiotic fauna from percolation water h as been system atically investigated. It was found out t h at t his h abitat, called epikarst, is an biologically important h abitat in its own rig h t wit h a diverse, specialized fauna (Pipan 2005), as well as a transition zone between surface and cave water. is specialized fauna, represented mainly by co pepods, h as been intensively studied in Slovenia (Pipan 2005), and recently preliminary studies h ave been car ried out in Romania (Moldovan et al. 2007), Spain (Ca mac h o et al. 2006) and W est Virginia, U.S.A. (Pipan & Culver 2005, Pipan et al. 2006b, Fong et al. 2007). e epikarst fauna is also of special interest because it is easier to obtain quantitative samples and estimates of total spe cies ric hness (Pipan and Culver 2007). Epikarst, t h e skin of karst (Bakalowicz 2004), is t h e entry for most organic matter originating in soils into caves. In t his way t h e percolating water entering into un derground h abitat is crucial for t h e h ypogean organisms as a source of nutrients and as a reservoir of pollutants as well. ere h as been little direct investigation of t h e ux of nutrient resources in caves (Simon et al. 2007). e isolated karst h as not been investigated for t h e fauna of water trickles. One of its most nort h ern patc h es is t h e so-called Paka isolated karst, named for t h e Paka River, w h ic h originates in t h e Eastern Alps. is is a o ristically and faunistically interesting area in h abited by some disjunctively dispersed species, like Cortusa mat thioli (Primulaceae), Waldsteinia ternata subsp. trifo lia (Rosaceae), and endemics, like t h e h ypogean beetle Aphaobiella tisnicensis Pretner 1949, t h e spider Troglo phyphantes diabolicus Deeleman-Rein h old 1971 and ot h ers (Pretner 1949, Bati et al. 1980, Novak & Kutor 1982, Sket 1994, Presetnik & Hudoklin 2005, Zagma jster & Kov 2006). Tisnik h ill (786 m a.s.l.) in t h e Paka isolated karst is wit h its 17 described caves among t h e most karstied areas in Slovenia; many of t h ese caves are documented also as arc h aeological and/or Palaeolit h ic sites (Kocbek 1895, Polsc h er 1917, Brodar 1938, Ravljen 1986, 1989, Brodar 1993, Gams 2004). All t h ese make t h is region of interest also for t h e investigations of t h e epikarst biodiversity. In t his paper we report on t h e rst results of t h e bi otic diversity and t h e ecology of fauna from percolation water in t h e isolated karst of NE Slovenia. W e compare it to a previous study of t h e epikarst fauna in t h e karst h eartland of Slovenia (Pipan 2005), and report on t h e p h ysical and c h emical c h aracteristics of t h e percolating water in Huda luknja. INTRODUCTION RESEARCH AREA AND METHODS e investigations were carried out in t h e cave Huda luknja pri Gornjem Doliu (Cadastre Number 413, Cave register IZRK ZRC SAZU, Speleological Association of Slovenia) situated in t h e hill Tisnik. Its main, lower entrance opens at 503 m a.s.l. at t h e bottom of t h e hill w hile its upper entrance t h e swallet Ponor Ponikve at T ANJA PIPAN, V ESNA NAVODNIK, F RANC JAN EKOVI T ONE NOVAK

PAGE 3

ACTA CARSOLOGICA 37/1 2008 143 622 m a.s.l. is h ardly ever accessible more t h an 35 m inside. e cave begins at t h e contact of non-karstic Mi ocene sandstone, conglomerate and marl, wit h middle and upper Triassic limestones w h ere-in it is developed (Gospodari 1980, Gams 2004). e cave consists of two main passages: t h e lower active and t h e upper dry 402 m long Medvedji rov. Above t h e swallet t h ere is t h e Lisinica cave, w hic h was in t h e late 1980s connected wit h t h e Medvedji rov by pumping out water from a si p h on in-between (Fig. 1) providing an accessible cave system t hroug h t h e hill. W it h its total lengt h of 2339 m, t his is t h e longest cave system in NE Slovenia. Ponikva brook runs t hroug h t h e water passage, ot h erwise t h e karst aquifer in t h e hill Tisnik is rec h arged by diuse inltration of precipitation. e passage Medvedji rov was c h osen for t h e sampling of percolation water (Fig. 2) because it provides dierent types of drips: t h ose wit h a permanent current, suc h providing dropping water and t h ose immediately reacting to precipitation. Twelve drips in a linear distance of ca. 190 m between t h e rst and t h e last one were sampled mont h ly from November 2005 till October 2006. Sampling of percolation water fauna was done using t h e met h od described in Pipan (2003, 2005). W ater and animals were collected via a funnel w hic h emptied into a plastic container, h aving aside overow h oles covered wit h a net (mes h size 60 m) to retain t h e animals. Tem perature (C), conductivity (Scm -1 ) and pH were mea sured using conductivity meter (ISKRA MA 5964) and pH meter (ISKRA MA 5740). W ater samples for c h emi F ig. 1: Geographical location and the longitudinal section of the Huda luknja cave system; acc. to Ravljen (1986, 1989), completed by R. B rai and M. P odpean. FIRST STUDIES OF THE FAUNA OF PERCOLATION W ATER IN HUDA LUKNJA, A CAVE IN ISOLATED KARST IN SLOVENIA

PAGE 4

ACTA CARSOLOGICA 37/1 2008 144 (S J ) (3) was calculated separately for aquatic and terres trial taxa to evaluate similarities of t h ese communities in dierent drips (Legendre & Legendre 2003): D =1 ) 2 n i N (1) H = ln n i N n i N (2) S J = a a+b+c (3) w h ere n i is t h e number of individuals belonging to spe cies i, N t h e total number of individuals, a t h e number of common species in bot h drips (joint occurrence), b t h e number of species in one drip but not in t h e ot h er, and c t h e number of species in t h e second drip but not in t h e rst one. e Jaccard index values were compared among eac h ot h er using t h e Mantel t-test (program MANTEL for W indows), and t h e comparison of t h e index values between water and terrestrial taxa was performed as well. e program SPSS 11.0 was used in t h e statistical proce dures. cal analyses were transported in t h e laboratory w h ere concentrations of cations (sodium, potassium, calcium and magnesium) and anions (c h loride, sulp h ate and ni trite) as well as calcium and total h ardness were deter mined using Standard Met h ods for t h e Examination of W ater and W astewater (1989). e samples of fauna were in situ xed wit h t h e formalde h yde until t h e nal solu tion of 2%. Organisms were extracted and identied using a microscope (Olympus CH30) and stored in 70% et h anol. Descriptive statistics was used to describe p h ysical and c h emical c h aracteristics of t h e drips. e correlation between fauna assemblages and p h ysical and c h emical parameters of percolation water was performed calculat ing Pearson coecient. One-way ANOVA was used in testing dierences between t h e drips and t h e sampling dates in eac h p h ysical and c h emical parameter. In testing communities, t h e Simpson index of dominance (D) (1) and t h e S h annon-W eaver diversity index (H) (2) were calculated separately for eac h drip in t h e w h ole year. e Pearson correlation between D and H, and t h e distance from t h e entrance and t h e surface, respectively, was cal culated for all 12 drips. e Jaccard similarity coecient F ig. 2: Ground plan of the passage M edvedji rov in the Huda luknja cave system showing sampling sites; acc. to Ravljen (1986, 1989), completed by R. B rai and M. P odpean. F ig. RESULTS e descriptive statistics of t h e p h ysical and c h emi cal parameters is presented in Table 1. Considering t h e w h ole year, t h ere were signicant dierences between sampling sites in eac h p h ysical and c h emical param eter (F 11, 266 = 1.90, p = 0.039 for pH and F 11, 266 = 2.41, p = 0.007 for temperatures, and F 11, 266 = 5.1071.73, p <0.001 for t h e ot h ers) as well as in eac h parameter be tween dates (F 11, 266 = 1.95, p = 0.033 for total h ardness, and F 11, 266 = 3.09360.13, p <0.001 for t h e ot h ers). T ANJA PIPAN, V ESNA NAVODNIK, F RANC JAN EKOVI T ONE NOVAK

PAGE 5

ACTA CARSOLOGICA 37/1 2008 145 Table 3: Number of copepod species per cave, and abundance of copepods, other aquatic species, and terrestrial species per sample. Number of species per cave and number of species per sample have a complex connection (P ipan & Culver 2007), and cannot be determined simply by dividing the two. Data from present study (Huda luknja) and P ipan (2005, all other caves). Huda luknja Postojnska jama Pivka jama rna jama kocjanske jame Dimnice upanova jama Number of copepod species 2 5 11 8 9 8 14 Number of copepods/sample 0.4 1.1 120.2 58.2 47.6 12.2 63 Number of other aquatics/sample 3.5 1.6 38.4 8.8 57.4 19.4 29.4 Number of terrestrials/sample 6.5 1.1 2.2 3.8 1.6 1.6 1.4 Table 1: B asic environmental parameters of twelve sampled drips in Huda luknja (NE Slovenia), monthly sampling from November 2005 till October 2006. Drip number / Parameter 1 2 3 4 5 6 7 8 9 10 11 12 Temperature (C) 7.81.0 (5.59.9) 7.41.4 (4.39.5) 8.01.1 (4.79.5) 6.91.5 (4.59.4) 7.61.8 (5.410.3) 7.20.9 (5.68.2) 7.11.0 (5.28.2) 7.20.7 (6.28.1) 7.21.2 (3.78.1) 7.50.3 (7.07.9) 7.70.2 (7.08.1) 7.60.4 (7.0 8.2) pH 8.10.3 (7.58.3) 7.90.3 (7.68.4) 8.00.3 (7.48.4) 8.10.2 (7.78.3) 8.10.2 (7.78.3) 7.90.1 (7.78.2) 8.10.2 (7.88.4) 8.10.2 (7.88.4) 8.10.2 (7.78.3) 7.90.1 (7.68.2) 8.00.1 (7.98.3) 8.00.2 (7.78.3) Conductivity (Scm -1 ) 247.038.7 (128.0294.0) 225.153.8 (135.0319.0) 284.155.3 (188.0371.0) 209.349.9 (119.0276.0) 181.937.2 (120.7226.0) 192.344.4 (127.6245.0) 228.953.3 (149.0304.0) 252.460.9 (145.2330.0) 209.449.7 (123.4295.0) 280.965.6 (176.6345.0) 209.535.7 (144.2318.0) 221.446.9 (152.0284.0) Ca hardness (mEql -1 ) 2.60.1 (2.42.9) 3.60.3 (2.33.8) 2.10.1 (2.02.2) 2.10.1 (2.02.3) 2.30.2 (2.02.8) 2.70.3 (2.43.2) 2.70.5 (2.03.3) 2.30.2 (1.82.6) 2.50.8 (1.63.7) 2.70.2 (2.13.4) 2.70.3 (2.03.2) Total hardness (mEql -1 ) 2.90.2 (2.53.3) 2.90.2 (2.73.5) 3.90.2 (3.34.2) 2.40.1 (2.22.6) 2.40.1 (2.22.7) 2.50.2 (2.23.0) 2.90.3 (2.73.5) 2.90.5 (2.23.6) 2.70.2 (2.23.0) 3.10.4 (2.73.8) 3.10.2 (2.43.6) 2.90.3 (2.63.5) Na + (mgl -1 ) 0.70.1 (0.61.0) 1.00.2 (0.61.3) 0.80.1 (0.60.9) 0.50.1 (0.40.8) 0.60.1 (0.40.8) 0.80.1 (0.61.0) 0.80.2 (0.51.4) 0.70.3 (0.51.5) 0.90.1 (0.71.2) 0.90.1 (0.81.2) 1.00.1 (0.71.2) 0.90.1 (0.81.2) K + (mgl -1 ) 1.20.3 (0.32.0) 0.50.1 (0.40.7) 0.70.5 (0.42.1) 0.40.1 (0.30.6) 0.30.1 (0.20.5) 1.00.3 (0.41.8) 0.60.2 (0.31.1) 0.50.4 (0.32.0) 0.50.2 (0.40.9) 0.30.2 (0.20.7) 0.20.1 (0.10.9) 0.40.3 (0.31.4) Ca + (mgl -1 ) 39.114.3 (2.252.1) 19.913.9 (2.352.1) 30.518.3 (3.276.2) 26.513.1 (1.744.1) 20.98.8 (1.844.1) 24.011.7 (1.848.1) 26.59.6 (2.450.1) 26.811.4 (2.554.1) 24.210.1 (2.548.1) 30.513.6 (2.874.2) 9.512.6 (2.352.1) 25.210.8 (2.452.1) Mg + (mgl -1 ) 0.20.1 (0.00.6) 0.30.1 (0.10.4) 0.30.3 (0.11.0) 0.30.1 (0.10.4) 0.30.1 (0.10.6) 0.20.2 (0.00.5) 0.30.2 (0.00.7) 0.30.2 (0.00.6) 0.40.2 (0.01.0) 0.60.5 (0.01.3) 0.30.1 (0.00.4) 0.30.2 (0.00.6) NO 3 (mgl -1 ) 20.57.2 (3.130.9) 13.34.4 (4.921.5) 16.518.1 (3.565.5) 12.63.8 (0.015.8) 2.90.2 (2.53.3) 14.74.8 (1.319.2) 3.91.6 (1.55.9) 2.71.7 (0.54.8) 2.42.5 (0.810.3) 1.10.2 (0.81.6) 0.90.6 (0.44.3) 5.85.9 (0.023.9) SO 4 2(mgl -1 ) 13.92.7 (8.621.5) 19.14.2 (10.123.5) 16.72.9 (8.620.6) 9.63.3 (0.013.8) 11.11.7 (8.413.5) 12.93.1 (8.519.4) 14.54.0 (7.920.0) 13.24.8 (6.920.5) 13.92.9 (9.817.6) 12.93.3 (9.318.2) 15.22.7 (7.816.6) 17.54.0 (10.122.5) Cl (mgl -1 ) 1.00.3 (0.91.4) 1.20.2 (0.81.6) 1.40.3 (0.81.9) 0.90.3 (0.31.6) 0.90.1 (0.91.3) 0.90.2 (0.81.3) 1.20.2 (0.91.6) 1.20.3 (0.81.5) 1.30.3 (0.91.9) 1.20.2 (0.91.6) 1.30.2 (0.81.4) 1.20.2 (0.91.6) FIRST STUDIES OF THE FAUNA OF PERCOLATION W ATER IN HUDA LUKNJA, A CAVE IN ISOLATED KARST IN SLOVENIA

PAGE 6

ACTA CARSOLOGICA 37/1 2008 146 In all t h e twelve drips we found 132 individuals and 67 eggs, respectively, belonging to 6 aquatic, 1 amp hibi ous and 18 terrestrial invertebrate taxa (Table 2). Among t h ese taxa, t h ere are species w hic h in h abit t h e h ypogean, soil and epigean environments. Number of individuals was in signicant positive correlation wit h K + ( r = 0.64, p = 0.026) and NO 3( r = 0.69, p = 0.013). ere were no correlation between t h e numbers of individuals, D H and t h e distance from t h e entrance and t h e surface (p>0.05) in t h e dataset of all 12 drips. e 11 t h drip is t h e output of a passage more or less directly connected wit h t h e surface. W it h respect to t h e presence t h ere were no specic dispersion patterns neit h er in aquatic not in t h e terrestrial taxa t h erefore t h e Jaccard index dendrograms Table 2: List of taxa found in twelve drips during a one year sampling from November 2005 till October 2006 in Huda luknja (NE Slovenia)*. HIGHER GROUP CLASS ORDO FAMILY GENUS and/or SPECIES A amphibic, T terrestrial, AQ aquatic taxon NEMATODA RHABDITIDA unidentied, AQ ANNELIDA CLITELLATA OLIGOCHAETA Enchytraeidae unidentied, AQ Lumbricidae cf. Dendrobaena sp., A ARTHROPODA ARACHNIDA PALPIGRADI Eukoeneniidae Eukoenenia cf. austriaca T (a tail fragment) ARANEAE Linyphiidae Troglohyphantes diabolicus T ACARINA Ixodidae Ixodes vespertilionis T ** MESOSTIGMATA unidentied, T ORIBATIDA unidentied, T AMPHIPODA Niphargidae Niphargus scopicauda AQ OSTRACODA unidentied, AQ COPEPODA HARPACTICOIDA Parastenocarididae Parastenocaris nolli alpina AQ Canthocamptidae Bryocamptus balcanicus AQ CHILOPODA unidentied, T DIPLOPODA ACHEROSOMATIDA unidentied, T Attemsiidae Polyphematia moniliformis T INSECTA ENTOGNATHA COLLEMBOLA Poduridae unidentied, T Sminthuridae unidentied, T Entomobryidae unidentied, T DIPLURA Campodeidae Plusiocampa sp. T PTERYGOTA COLEOPTERA Staphylinidae cf. Atheta sp., T Carabidae Laemostenus ( Antisphodrus ) schreibersi, T Curculionidae Otiorhynchus ( Troglorhynchus ) anophthalmus T DIPTERA Sciaridae unidentied, T Trichoceridae unidentied, T MUSCOMORPHA unidentied larva, T *Eggs, coccons and rests of undetermined taxa not included. **Parasitic. are not presented h ere. ere were also no signicant correlation between t h e values of t h e Jaccard index and it was also t h e case in t h e comparison of t h e aquatic against t h e terrestrial taxa ( r = -0.08; Mantel t-test, t = -0.54, p = 0.294). While t h e ot h er taxa appeared in t h e drips more or less randomly, t h e only exception is t h e lumbricid spe cies, present solely in drip No. 3. e most abundant taxa represented by ten or more individuals were oligoc h aetes from t h e families Enc h y traeidae and Lumbricidae (Fig. 3), and eggs of undeter mined organisms. Bot h t h e aquatic Enc h ytraeidae and amp hibious Lumbricidae were t h e most abundant in t h e warm period of t h e year, w hile t h e newly-h atc h ed enc h y traeids were caug h t in January, aer h eavy raining. e T ANJA PIPAN, V ESNA NAVODNIK, F RANC JAN EKOVI T ONE NOVAK

PAGE 7

ACTA CARSOLOGICA 37/1 2008 147 lumbricid species, recently found also in h alf ripe guano, was present only in t h e drip No. 3 wit h t h e hig h er values of conductivity and concentrations of nitrates and c h lo rides (Table 1). e concentration, C = 1 D and t h e S h annonW eaver diversity index, H are presented in t h e Fig. 4. D 3: Oligochaeta were the most abundant in the drips in Huda luknja, sampling from November 2005 till October 2006. F ig. 4: Shannon-Weaver biodiversity index (H) and concentra tion (C: 1-D) for each sampling site in Huda luknja from Novem ber 2005 till October 2006 (eggs not included). and H were in negative signicant correlation wit h con ductivity ( D vs. conductivity: r = -0.77, p = 0.004; H vs. conductivity: r = -0.60, p = 0.037), and D vs. total h ard ness ( r = -0.59, p = 0.043) and Ca 2+ ( r = -0.61, p = 0.037). DISCUSSION Comparison of an epikarst fauna from caves in Dinaric karst (SW Slovenia) and Huda luknja from isolated karst in NE Slovenia (Table 3) s h ows some important dier ences but also similarities. On t h e one h and, t h e karstic area wit h Huda luknja h arbours relatively poor biodi versity, partly because t h e fauna in percolation water is not dominated by copepods as is t h e case in limestone caves in classical karst area (Pipan 2005). On t h e ot h er h and, t h e nding of a new, probably endemic amp hibi ous lumbricid species in t h e percolation water is a furt h er indication of an isolated development of t h e fauna of t h e Paka isolated karst. e species found in t h e epikarst of Huda luknja are in s h arp contrast to t h e results of Pipan (2005) in t h e epikarst of caves in t h e Dinaric karst region of Slovenia (Table 3). Copepod species diversity was neg ligible in Hudna luknja compared to t h e six karst caves. e smallest number of copepod species Pipan (2005) found was ve and t h e median was eig h t, compared to two for Hudna luknja. is is unlikely t h e result of inade quate sampling. More drips were sampled in Huda luknja and Pipan and Culver (2007) s h owed t h at ve or six drips s h ould h ave been adequate to nd 90 percent of t h e cope pod species. Abundance dierences of copepods tell t h e same story. Density per drip per year was many times less in Huda luknja t h an in any of t h e karst caves (Table 3). Only Postojnska jama approac h ed Huda luknja in abun dance and even t h ere abundance was nearly t hree times greater. e abundance of ot h er aquatic species s h ows a similar pattern alt h oug h in t his case Postojnska jama and Huda luknja form a group distinct from t h e ot h er caves. Wh at is t h e explanation for t his dierence? Pipan et al. (2006a) s h owed t h at ceiling t hickness was an important determinant of species distribution, and t h e relative t hick ceiling in Hudna luknja (see Fig. 1) would seem to bear t his out, as does t h e negative correlation of overall abun dance of animals wit h distance. e similar pattern was s h own in t h e study of geograp hic distance on an epikarst copepod community composition in Organ Cave in W est Virginia, U.S.A. (Pipan et al. 2006b). e distribution of terrestrial species provides an important additional clue about structure of epikarst and a direction for future researc h In striking contrast to t h e aquatic pattern, Huda luknja h ad more terrestrial individuals in t h e samples t h an t h e caves in Dinaric karst. As Pipan (2005) did not identify terrestrial ani mals to t h e same level of detail as in t h is study, it is not FIRST STUDIES OF THE FAUNA OF PERCOLATION W ATER IN HUDA LUKNJA, A CAVE IN ISOLATED KARST IN SLOVENIA

PAGE 8

ACTA CARSOLOGICA 37/1 2008 148 possible to compare species ric h ness. Median number of terrestrial invertebrates per drip was 1.6 for t h e karst caves and 6.5 for Huda luknja (Table 3). is tantalizing data suggests t h at wit h increasing dept h t h ere may also be increasing amounts of terrestrial h abitat. If it is t h e epikarst and not t h e column of percolating water t h at is t h e primary h abitat as most evidence indicates (Culver & Pipan 2005, but see Sket et al. 2004a), t h en t h e cracks and ssures of t h e zone of percolation h arbors a terres trial fauna (as indicated by presented data), and acts as a lter as indicated by t h e importance of ceiling t h ickness (Pipan et al. 2006a, b). e distribution and t h e development of subterra nean fauna depend on geomorp h ologic and h ydrologic conditions since t h e Pliocene (Bole 1977). e idea t h at caves of various isolated karst patc h es could oer an op portunity for studying evolutionary and biogeograp hic processes in t h e h ypogean fauna, especially molluscs (Bole 1972), h as not been conrmed. It now seems t h at t his idea can be resurrected by referring to t h e epikarst fauna. Distribution of fauna in t h e epikarst w hic h is largely ruled by local conditions is in close relation wit h t h e h ydrogeological c h aracteristics of epikarst, con trolled by h ydraulic connectivity (Pipan & Culver 2007). For some subterranean organisms it was found out t h at t h eir reproduction is induced by decreasing tempera ture and increasing ux of organic material (Culver et al. 1995). W e can assume t h at t h e enc h ytraeids wit h t h eir newly h atc h ed juveniles in November are representatives of suc h a species well adapted to t h e h ypogean environ ments. e new lumbricid species, found only in t h e drip No. 3 w h ose water c h aracteristics diered considerably from ot h er drips especially by hig h er nitrate concentra tion, deserve a special treatment elsew h ere. e positive correlation between number of individuals and K + and NO 3probably concerns t h eir special nutritional value, but any furt h er conclusion would be speculative. Pi pan et al. (2006a) discuss a similar example of exclusive presence among copepods in percolation water as some species were found only in one drip in a cave. is was best explained as representing extreme environmental conditions in t h e cave. Niphargus scopicauda probably derived from unknown water courses above t h e cave, w hile t h e two h arpacticoid species are from t h e epikarst. B ryocamptus balcanicus s h ows troglomorp hic c h aracters w hile P arastenocaris nolli alpina wit h its slim body pre fers interstitial h abitats and crevices of t h e unsaturated zone w h ere it most easily escapes from being preyed by various predators. It is oen found in small depressions on cave walls but not in larger pools (Petkovski 1959, Pipan 2005). Groups of animals found in percolation water are common in many ot h er subterranean h abi tats (Moldovan et al. 2007), but some stygobiotic species from drips are specialized for particular micro h abitats and oen represent endemic taxa (Pipan 2005). So far, per h aps except for t h e enc h ytraeid species, all t h e ot h ers in t h e Medvedji rov seem to belong to representatives of taxa s h aring various h abitats. Negative correlations of D and H wit h conductivity, and wit h total h ardness and Ca 2+ probably indicate t h at t h e individuals in drips were mostly eluted into t h e passage by water relatively rapidly percolating t hroug h t h e limestones, alt h oug h t his result could also be coincidental. e signicant dierences in t h e p h ysical and c h emical properties between t h e drips indicate h etero geneous nature of t h e h abitats supplying water to t h e drips. According to t h eir terrestrial and aquatic taxa as semblages t h e drips are fed wit h t h e percolation water from t h e surface, from ssures and larger undiscovered passages above t h e Medvedji rov. In general, t h e epigean, epikarst and h ypogean faunas were present in t h e drips, alt h oug h fauna in t h e drips is relatively very poor as compared wit h t h e one in t h e Dinaric and Alpine karst (Pipan 2005; unpublis h ed data), especially wit h respect to t h e p h reatic taxa. e biodiversity of a particular cave is governed by p h ysical and ecological h eterogeneity in t h e epikarst w hic h inuence t h e biotic diversity in t h e lower unsatu rated zone (Pipan 2005, Moldovan et al. 2007). In prin ciple, suc h taxa also present an important source of par ticulate organic carbon (POC) entering a cave t hroug h t h e epikarst (Simon at al. 2007). In our case, all t hree: t h e epigean, t h e epikarstic and t h e h ypogean terrestrial as well as aquatic taxa contributed to t h e biotic diversity of t h e drips, but little to t h e input of organic matter. Sub terranean biodiversity is at t h e local scale controlled by productivity (Gibert & De h arveng 2002). e very lim ited amount of POC could be one of t h e reasons for t h e species scarcity in t h e area. e limited POC, t h e dried individuals of terrestrial h ypogean taxa as well as rela tively low concentrations on N and P compounds in most drips suggest t h at sudden pouring wit h t h e rain water are t h e crucial factor for driing fauna into t h e cave passage. e decrease of biodiversity from t h e rst toward t h e last sampling site was found in some previous studies of an epikarst fauna from Slovenian caves (Pipan 2005, Pipan et al. 2006a), and as is t h e case in Medvedji rov indicates h ap h azardous inux of individuals from various h abitats above t h e passage. T ANJA PIPAN, V ESNA NAVODNIK, F RANC JAN EKOVI T ONE NOVAK

PAGE 9

ACTA CARSOLOGICA 37/1 2008 149 CONCLUSIONS On t h e basis of t h e fauna assemblages composed of ter restrial and aquatic taxa, it is concluded t h at samples from Huda luknja are fed by t h e percolation water from t h e surface, from ssures and larger undiscovered pas sages above t h e Medvedji rov. Bot h t h e aquatic as well as t h e terrestrial taxa were randomly distributed in t h e drips irrespective of t h e p h ysical and c h emical proper ties of water. In general, t h e epikarst aquatic fauna of t h e Paka isolated karst is relatively poor. At t h e moment, it can not be concluded w h et h er t his is a general p h enom enon in t h e isolated karst concerning its all-over biotic poverty or it is a specic c h aracteristic of eac h isolated karst patc h es, like t h e Paka karst. More adequate knowl edge on t h e topic must be acquired in t h e isolated karst to make t h e decision. ACKNO W LEDGEMENTS W e are indebted to t h e sta of t h e c h emical laboratory of t h e Steam Power Station otanj w h ere t h e c h emical anal yses were carried out, especially to Mrs. Greta Srnovrnik. W e are grateful to t h e cavers of t h e Cavers Club Speleos Siga in Velenje for t h e support in t h e investigations and to Viktor Ocvirk for t h e h elp wit h t h e eld work. Boris Sket and Cene Fier determined t h e Niphargus speci mens. W e t h ank Janez Ravljen, Rajko Brai and Milan Podpean for t h e permission of publis hing t h e plans of Huda luknja. W e sincerely t h ank Janez Mulec and David Culver for critical remarks and insig h tful comments on t h e manuscript. David Culver is t h anked for his con structive review of t h e manuscript t h at included h elpful suggestions and Table 3 t h at improved t h e value of t h e nal paper. e study was partly supported by t h e Slov ene Ministry of Hig h Education, Science and Tec hnology wit hin t h e Biodiversity researc h programme (P1-0078). REFERENCES Bakalowicz, M., 2004: e epikarst, t h e skin of karst. In: Jones, W K., D. C. Culver, & J. S. Herman (eds.) P ro ceedings of the Epikarst Symposium, October 1-4, 2003, Sheperdstown, West V irginia, USA, Karst W a ters Institute Special Publication 9, pp. 16. Bole, J., 1972: Podzemeljski poli na osamljenem krasu Slovenije.Nae jame, 13, 55. Bole, J., 1977: Podzemeljski poli v osamljenem krasu Posavskega hribovja.Nae jame, 18, 31. Bole, J., 1985: Recentni podzemeljski poli in razvoj nekateri h poreij na dinarskem krasu.Razprave (Dissertationes), classis IV, SAZU, 24, 315. Brodar, M.,1993: Paleolitske in mezolitske najdbe iz jame pe h ovke pri Zgornjem Doliu.Ar h eoloki vest nik, 44, 7. Brodar, S., 1938: Das Palolitikum in Jugoslawien.Quartr, 1, 140+VII. Camac h o, A. I., A. G. Valdecasas, J. Rodrguez, S. Cuezva, J. Lario, & Snc h ez-Moral, S., 2006: Habitat con straints in epikarstic waters of an Iberian Peninsula system cave.Annales de Limnologie-International Journal of Limnology, 42, 2, 127. Culver, D. C., T. C. Kane, & Fong, D. W ., 1995: Adapta tions and natural selections in caves: t h e evolution of G ammarus minus. Harvard University Press, p. 233, Cambridge. Culver, D. C., M. C. C hristman, W R. Elliott, H. H. Hobbs III, & Reddell J. R., 2003. e Nort h American obli gate cave fauna: regional patterns.Biodiversity and Conservation, 12, 441-468. Culver, D. C. & Pipan, T., 2007: Wh at does t h e distribu tion of stygobiotic Copepoda (Crustacea) tell us about t h eir age?Acta carsologica, 36, 87. Culver, D. C. & Pipan, T., in press: Caves as Islands.In G. Rice [ed.] Encyclopedia of Islands, University of California Press, Berkeley. Fong, D. W ., D. C. Culver, H. H. Hobbs III, & Pipan, T., 2007: e invertebrate cave fauna of W est Virginia, Second edition.W est Virginia Speleological Sur vey, Bulletin No. 16, p. 163, Barrackville. Gams, I., 2004: Kras v Sloveniji v prostoru in asu.Intitut za raziskovanje krasa ZRC SAZU, p. 515, Ljubljana. FIRST STUDIES OF THE FAUNA OF PERCOLATION W ATER IN HUDA LUKNJA, A CAVE IN ISOLATED KARST IN SLOVENIA

PAGE 10

ACTA CARSOLOGICA 37/1 2008 150 Gibert, J., & Culver, D. C., 2002: Subterranean ecosys tems: a truncated functional biodiversity.Biosci ence, 52, 6, 473. Gospodari, R., 1980: Geoloka zgradba in kraki pojavi. In: Bati, F., M. Brodar, R. Gospodari, V. Kutor & T. Novak: Soteska Huda luknja. Obzorja.Kulturni in naravni spomeniki Slovenije, 102, 6. Habe, F., 1972: Nekatere speleoloke znailnosti osam ljenega krasa Slovenije.Nae jame, 13, 45. Habi, P., 1969: Hidrografska rajonizacija krasa v Slo veniji.Kr Jugoslavije, 6, 79. Kocbek, F., 1895: Huda luknja in njene podzemeljske jame.Planinski vestnik, 1(12), 177. Legendre, P. & Legendre, J., 2003: Numerical Ecology. Elsevier, Amsterdam. Moldovan, O. T., T. Pipan, S. Iepure, A. Mi h evc, & Mulec, J., 2007: Biodiversity and ecology of fauna in per colating water in selected Slovenian and Romanian caves.Acta carsologica, 36, 493-501. Mri, N., T. Novak, F. Potonik & Amon, T., 1996: Eco logical evaluation of oniscoids and diplopods from cavities in Slovenia.Mmoires de Biospologie, 13, 203. Novak, T., 2005: Terrestrial fauna from cavities in Nort h ern and Central Slovenia, and a review of systemati cally ecologically investigated cavities.Acta carso logica, 34, 169. Novak, T. & Kutor, V., 1982: Zur Fauna der W nde drei er H h len Nordostsloweniens (Jugoslawien).Die H h le, 33(3), 82. Petkovski, T. K., 1959: Fauna Copepoda peine Dona Duka kod Raa Skopje.Fragmenta balcanica, 2(14), 107. Pipan, T., 2003: Ekologija cepononi h rakov (Crusta cea: Copepoda) v prenikajoi void izbrani h kraki h jam.Doktorska disertacija (in Slovene wit h Englis h abstract and summary), Univerza v Ljubljani, Odd elek za biologijo, 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. Pipan, T., & Culver, D. C., 2005: Estimating biodiversity in t h e epikarstic zone of a W est Virginia cave.Jour nal of Cave and Karst Studies, 67(2), 103. Pipan, T., A. Blejec, & Brancelj, A., 2006a: Multivariate analysis of copepod assemblages in epikarstic wa ters of some Slovenian caves.Hydrobiologia, 559, 213. Pipan, T., M. C. C hristman, & Culver, D. C., 2006b: Dy namics of epikarst communities: microgeograp hic pattern and environmental determinants of epikarst copepods in Organ Cave, W est Virginia.American Midland Naturalist, 156, 75. Pipan, T., & Culver, D. C., 2007: Copepod distribution as an indicator of epikarst system connectivity.Hy drogeology Journal, 15, 817. Polsc h er, N., 1917: Die Huda luknja und i hre Grotten.Mitteilungen der Geograp hisc h en Gesellsc h a in W ien, 60, 117. Presetnik, P. & Hudoklin, A., 2005: Spodnja Klevevka jama pomembno zatoie netopirjev in novo najdie dolgokrilega netopirja (M iniopterus sch reibersii) na Dolenjskem (JV Slovenija).Natura Sloveniae, 7(1), 31. Pretner, E., 1949: Aphaobius (Aphaobiella subgen. nov.) budnar-lipoglaveki spec. nov, A. (A.) tisnicensis spec. nov. in opis samca P retneria saulii G. Mller (Coleoptera, Silp hidae).Razprave (Dissertationes), classis IV, SAZU, 143. Ravljen, J. (ur.), 1986: Soteska Huda luknja. Jamarski bilten. Glasilo jamarskega kluba Speleos Titovo Velenje, 4. Ravljen, J., 1989: Huda luknja neko in danes. Zbor tajerski h jamarjev Titovo Velenje, 3. Simon, K. S., T. Pipan, & Culver D. C., 2007: A concep tual model of t h e ow and distribution of organic carbon in caves.Journal of Cave and Karst Studies, 69, 279. Sket B., 1993. Cave fauna and speleobiology in Slovenia.Nae jame 35, 35. Sket B., 1994. Distribution of some subterranean Crus tacea in t h e territory of t h e former Y ugoslavia.Hy drobiologia 287, 65. Sket B., 1999a. e nature of biodiversity in h ypogean waters and h ow it is endangered.Biodiversity and Conservation 8, 1319. Sket B., 1999b. Hig h biodiversity in h ypogean waters and its endangerment t h e situation in Slovenia, t h e Di naric karst, and Europe.Crustaceana 72, 767. Sket, B., 2002: e evolution of t h e karst versus t h e dis tribution and diversity of t h e h ypogean fauna. In: Gabrovek, F. (ed.) Evolution of karst: from prek arst to cessation, Intitut za raziskovanje krasa ZRC SAZU, Zaloba ZRC, Postojna, pp. 225. Sket, B., P. Trontelj, & agar C., 2004a: 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, p. 104-113. In: Jones, W K., D. C. Culver, & Herman, J. S. (eds.) Epikarst. P ro ceedings of the symposium held October 1 through 4, 2003, Sheperdstown, West V irginia, USA, Karst W a ters Institute Special Publ. 9, C h arles Town, W .Va., pp. 104-113. T ANJA PIPAN, V ESNA NAVODNIK, F RANC JAN EKOVI T ONE NOVAK

PAGE 11

ACTA CARSOLOGICA 37/1 2008 151 Sket, B., K. Paragamian, & Trontelj P., 2004b: A census of t h e obligate subterranean fauna of t h e Balkan Pen insula. In: Grit h s, H. I., B. Krytufek, & Reed, J. M. (eds.) B alkan B iodiversity. P attern and P rocess in the European Hotspot, Kluwer Academic Publis h ers, Dordrec h t, pp. 309. SPSS 11.0 for W indows. Standard Met h ods for t h e Examination of W ater and W astewater, 17 t h edition, 1989.APHA-A WW AWPCF, p. 1088 p. Trontelj, P., Goriki, S. Polak, R. Verovnik, V. Zakek, & Sket, B.,: 2007: Age estimates for some subterranean taxa and lineages in t h e Dinaric karst.Acta carso logica, 36, 183. Zagmajster, M. & Kov, L., 2006: Distribution of pal pigrades (Arac hnida, Palpigradi) in Slovenia wit h a new record of Eukoenenia austriaca (Hansen, 1926).Natura Sloveniae, 8(1), 23. FIRST STUDIES OF THE FAUNA OF PERCOLATION W ATER IN HUDA LUKNJA, A CAVE IN ISOLATED KARST IN SLOVENIA



PAGE 1

MICROORGANISMS IN H Y POGEON: EXAMPLES FROM SLOVENIAN KARST CAVES MIKROORGANIZMI PODZEMLJA: PRIMERI IZ SLOVENSKIH KRAKIH JAM Janez MULEC 1 Izvleek UDK 579.2:551.44 Mikroorganizmi podzemlja: primeri iz slovenskih krakih jam Mikroorganizmi v jama h naseljujejo razline h abitate, kjer raz vijejo tevilne interakcije. Kot dokaz mikrobne aktivnosti la h ko naletimo na tevilne bioloko pogojene krake oblike. Mikroor ganizmi so vklueni tako v litogene kot litolitine procese. Poleg h eterotrofni h organizmov la h ko v jama h priakujemo tudi av totrofne. Nekatere cianobakterije in mikroalge la h ko preivijo v jama h tudi pri tisti h intenziteta h svetlobe, ki so nije od nji h ove fotosintetske kompenzacijske toke. V prispevku so predstav ljena na h ajalia in mikroorganizmi (bakterije, cianobakterije, mikroalge, glive in praivali), ki so bili do sedaj identicirani v slovenski h jama h. Posebej bakterije, kot najbolj raznovrstna skupina mikroorganizmov, ponujajo ogromen biote hnoloki in bioremediacijski potencial. Mikrobna biomasa predstavlja v jama h upotevanja vreden vir hrane za vije razvite jamske organizme. Jame v Sloveniji predstavljajo e veliko monosti za odkritje novi h vrst, kot je primer glive M ucor troglophilus, ki je bila odkrita na jamski kobilici Troglophilus neglectus Kljune besede: mikroorganizmi, jamska mikrobiologija, kras, Slovenija. 1 Karst Researc h Institute SRC SASA, Titov trg 2, SI-6230 Postojna, Slovenia; e-mail: janez.mulec@guest.arnes.si Received/Prejeto: 07.12.2007 COBISS: 1.02 ACTA CARSOLOGICA 37/1, 153-160, POSTOJNA 2008 Abstract UDC 579.2:551.44 Microorganisms in hypogeon: Examples from Slovenian karst caves In caves microorganisms in h abit distinct h abitats w h ere t h ey develop various interactions. As an evidence of microbial activ ity several features can be identied. Microorganisms are in volved bot h in lit h ogenic and lit h olitic processes. Besides h et erotrop h s in caves autotrop hic organisms can be also expected. Some cyanobacteria and microalgae in caves can survive even at p h oton ux densities lower t h an t h eir p h otosynt h etic compen sation point. In t h e paper up-to-date identied groups of mi croorganisms (bacteria, cyanobacteria, microalgae, fungi and protozoa) wit h t h eir localities in Slovenian caves are presented. Especially bacteria from caves, as t h e most diverse group, of fer immense biotec hnological and bioremediation potential. In caves microbial biomass can be considered a considerable food source for cave-dwelling hig h er organisms. Caves in Slovenia oer great c h ances to discover new species, as was fungus M u cor troglophilus discovered in association wit h t h e cave cricket Troglophilus neglectus. Key words: microorganisms, cave microbiology, karst, Slovenia. CAVE MICROBIOLOG Y Cave microbiology h as recently been establis h ed as a new interdisciplinary eld of microbiology, geology and c h emistry dealing wit h microscopic life t h at resides in caves and inuences natural cave processes. In t h e last decades t h e recognition of microorganisms in geologi cal processes in caves altered our perception of cave eco systems (Barton & Jurado 2007). In Slovenia as well as elsew h ere in t h e world cave microbiology is still in its infancy. Slovenian Classical Karst is of great interest for researc h ers of dierent professional backgrounds as it is well karstied and relatively well studied. Estimates obtained from rnotie Quarry (Slovenia) s h owed t h at void spaces cavernosity in Slovenian karst represents 3.9% (Knez et al. 2004), w hic h means t h at t his is a poten tial h abitat w hic h can be colonized by t h e adapted cave (micro)organisms. A potential to nd specialized hig h er organisms in suc h environments is very hig h (Pipan & Culver 2007), it can also be expected t h at a h uge variety of dierent groups of cave-adapted microorganisms will be found.

PAGE 2

ACTA CARSOLOGICA 37/1 2008 154 Barton (2006) dened several features wit hin caves w hic h can be identied as evidence of microbial activity: dots on surfaces, unusual coloration of speleot h ems, precipitates, corrosion residues, structural c h anges and biolms. In Slovenian caves microbial communities and t h eir inter actions h ave been studied in various microenvironments. One of t h e pioneering studies on microbial mat covering cave walls was conducted by Meguar and Sket (1977) in t h e cave Planinska jama. Anot h er interesting interac tion of microbes wit h t h e rock surface is formation of so called cave silver. e term is frequently used by cavers due to its weak silver uorescence w h en illuminated by a lig h t source. ese water droplets are formed near t h e cave entrance w h ere cold and warm air are mixed and t h eir presence is probably connected wit h condensation (Mulec et al. 2002). In caves many times cave gold is also observed by visitors (Fig. 1). e golden aspect of colonies usually appears w h en illuminated and water droplets frequently magnify t h e yellowis h pigment of t h e microbial mat beneat h t h e water lm. ese mats are ob served in places covered wit h sediments or ot h er energy sources broug h t by oods or percolation water; in s h ort at places w h ere organic matter enters t h e caves. Besides h eterotrop h s one h as to keep in mind also t h e presence of p h otoautotrop hic organisms in caves. In t h e illuminated parts of t h e cave, cave entrances and places deep in s h ow caves illuminated by articial lig h t (Fig. 1), greenis h mat is composed predominantly of cyanobacteria and green microalgae (Mulec et al. in press). Many microbes w hic h enter caves are not adapted to cave conditions and t h eir colonies on various substrata are of s h ort duration, but t h eir biomass can notably in uence usually nutrient-decient cave environment. is can be t h e case also in caves wit h hig h alloc h tonous or ganic input due to h uman activity w h ere decomposition processes start to take place. Microbial presence is very well observed in guano, animal excrements, and dead animals. TRACES OF MICROBIAL PRESENCE F ig. 1: Examples of microbial colonization in caves. (A) M icrobial mat known as cave gold, (B) stalactites of calcite moonmilk, (C) stromatolitic stalagmites, (D) lampenora. BIODETERIORATION AND BIOSPELEOGENESIS Microorganisms in caves are involved in lit h ogenic proc esses, e.g. speleot h em deposition and cavern enlarge ment (Engel et al. 2004, Caaveras et al 2006, Mulec et al. 2007) and lit h olitic processes. e latter, w hic h are completely undesirable from t h e h uman perspective, were extensively studied in caves wit h pre historic paint ings and t h eir role in degradation discussed (Asencio & Aboal 2001, Caaveras et al. 2001, Sc h abereiter-Gurtner et al. 2002). Many times microbial presence can be observed indirectly. Sometimes inside caves large surfaces of weat h ered limestone can be noticed, w hic h is a result of incomplete dissolution by carbonic acid in t h e cave en vironment under specic conditions, and remain on t h e passage walls (Zupan Hajna 2003). Corrosion residues can be result of microbial metabolic activity and are fur t h er involved in dissolving of t h e h ost rock. In samples of weat h ered limestone from t h e caves Peina v Bortu and in Martinska jama Mulec et al. (2002) retrieved 1.1 6 colony-forming units per gram (CFU/g) and 1.3 3 CFU/g of wet mass, respectively. An interesting example of carbonate precipitation in t h e form of calcite ras is a pond periodically lled wit h water in Peina v Bortu cave w h ere Mulec et al. (2002) obtained 2.5 4 CFU/ml. In t h e same study Mulec et al. (2002) sampled moonmilk in t h e cave Snena jama na Radu hi in t h e Alpine region of Slovenia w h ere t h ey cal culated 6.4 2 CFU/g (Fig. 1). Moonmilk is dened as t h e result of disintegration of bedrock and speleot h ems or as mixed deposition of calcite crystals and water (Hill & Forti 1997). In t h e study of calcite moonmilk from Al tamira cave, Caaveras et al. (2006) concluded t h at mi J ANEZ MULEC

PAGE 3

ACTA CARSOLOGICA 37/1 2008 155 TRANSPORT OF MICROBES As elsew h ere in t h e environment viable microbial prop agules and ot h er tiny droplets oat in t h e air in caves due to t h eir low specic gravities and are disseminated by air currents. Classical met h od for measurement of airborne microorganisms relies on culturing w h en an agar medium is exposed to t h e environment. Bioaero sol microbes on agar plates are collected primarily due to gravity (Buttner et al. 1997). Using t his procedure in a simple experiment performed in December of 2001 in t h e cave Peina v Bortu (total lengt h of t h e cave is 240 meters) close to t h e entrance 68 CFU was observed on an agar plate and at two sampling points close to t h e end of t h e cave 106 and 113 CFU, respectively. Petri plates wit h nutrient agar were exposed for 20 minutes to t h e cave at mosp h ere, transferred to t h e laboratory and cultivated at RT for 7 days. Due to t h e winter air inow in caves t h e hig h est CFU being at t h e end of t h e cave is not surpris ing. Microbes are passively transported by airows w hic h depend on t h e seasonal air circulation and t h us represent an important mode of spreading of inoculum in dierent parts of caves. Anot h er carrier of microbes in caves rep resents dripping and seeping water. W it h h eterotrop hic plate count w hic h gives us estimates on live h eterotro p hic bacteria in water Geri et al. (2004) found from a single trickle in kocjanske jame t h at eac h day 2.2 4 bacteria as CFU are released in t h e cave from areas above t h e cave ceiling. Besides air currents and water ow, in troduction of microbes by animals and h umans is an im portant mode of transport deep in t h e karst underground (Dobat 1970). crobes actively c h ange t h e p h ysicoc h emical precipitation of moonmilk, resulting in formation of variety of bre crystal morp h ologies, sizes and bre networks. Anot h er example is stromatolitic stalagmites (Fig. 1) and sta lactites, t h e growt h of w hic h is en h anced by carbonate deposition promoted by cyanoprokaryotes towards sun lig h t (Taboroi 2006). In Slovenia t h ese structures were studied in kocjanske jame (Mulec et al. 2007). It seems t h at suc h biogenic carbonate formations can be far more widespread in subaerial h abitats if microalgal propagules meet adequate ecological factors for growt h. ENERG Y INPUT Caves and ot h er subterranean h abitats h ave long been recognized as a nutrient-decient environment. En ergy sources and nutrients can enter caves as atmos p h eric gases, as soil-derived aromatic and polyaromatic compounds, percolating via surface water, and reduced metal ions suc h as Mn 2+ and Fe 2+ wit hin t h e rock itself. Some cave condensates contain various small aromatic compounds t h at microbes can use as carbon and energy sources (Barton & Jurado 2007). Some underground eco systems are totally self sucient. Movile cave in Romania is t h e most known example wit h ric h and diverse subter ranean fauna wit h a c h emolit h oautotrop h y-based eco system and bacteria as primary producers (Sarbu et al. 1996, Kinkle & Kane 2000). Among ot h ers caves in t h e world anot h er suc h example was recently discovered Ay allon cave in Israel (Por 2007). Caves are usually devoid of sunlig h t: t h at is w h y primary production based on p h otosynt h esis cannot de velop. is is not t h e case in cave entrances illuminated by sunlig h t and in s h ow caves around lamps w h ere a complex p h ototrop hic community named lampenora develops (Dobat 1972, 1998, Martini et al. 1981). In terestingly, in caves some cyanobacteria and microalgae in t his community can survive even at p h oton ux densi ties lower t h an t h eir p h otosynt h etic compensation point (Mulec et al., in press). A completely dierent situation on energy sources are caves wit h hig h energy input eit h er via a polluted un derground stream and by polluted epikarst water from t h e unsaturated zone (Pipan 2005). Bacteria w hic h enter caves by percolation water can be considered a consider able food source for metazoans (Geri et al. 2004). Sud den oods bring hig h amounts of alloc h tonous organic matter and organisms and consequently h eterotrop hic decomposition processes can begin. Sometimes even w h en t h e source of pollution is removed t h e previous condition is not restored. e great majority of microbial species do not in h abit suc h energetically favourable conditions. ey h ave to survive under extremes of near-starvation or oligo trop hic conditions, dened as h aving less t h an 2 mg of total organic carbon (TOC) per litre. Nutrient resources MICROORGANISMS IN H Y POGEON: EXAMPLES FROM SLOVENIAN KARST CAVES

PAGE 4

ACTA CARSOLOGICA 37/1 2008 156 in caves rarely reac h as hig h as 0.5 mg of TOC per litre (Barton & Jurado 2007). Wh at is interesting is t h e fact t h at in suc h oligotrop hic condition up to 10 6 cells of dif ferent bacterial species per gram of rock material is easy to nd (Barton & Jurado 2007). is is in accordance wit h ndings from some Slovenian caves w h ere relatively hig h bacterial biomass was retrieved by cultivation met h od (Mulec et al. 2002, Geri et al. 2004). To overcome limitation, sels h competition for resources is replaced by cooperative and mutualistic microbial associations (Barton & Jurado 2007). CAVE MICROORGANISMS In caves dierent h abitats are populated by dierent microorganisms: bacteria, fungi, algae and protozoa. Microbes are usually identied in water bodies, rocky surfaces, on sediments and in guano. Anot h er important group of microorganisms are t h ose w hic h developed dif ferent type of interaction wit h troglobiotic animals, e.g. epibionts and parasites on cave animals (Golemansky & Bonnet 1994). B ACTERIA As discussed above various microbial mats can be traced in caves. Meguar and Sket (1977) identied species composition of organic covers from t h e walls in t h e cave Planinska jama. Beneat h an etc h ed rocky surface was ob served. e majority of bacteria belonged to Gram posi tive cocci, rods and pleomorp hic s h aped cells. Several bacterial isolates were retrieved in laboratory conditions: B acillus subtilis, B acillus cereus, B acterium brevi, P roac tinomyces polychromogenes Based on t h e morp h ology and bioc h emical tests Mulec et al. (2002) identied dif ferent bacterial groups in silver as hing droplets known as cave silver, in a pond wit h calcite ras and in weat h ered limestone. In t h e all studied microenvironments u orescent pseudomonads seemed to be prevalent bacteria w hic h can be due to t h eir versatile metabolic pat h ways. In caves various types of mats are largely made up of ac tinomycetes (Grot h et al. 1999). In permanent drips wit h hig h biodiversity of meiofauna in kocjanske jame Geri et al. (2004) establis h ed t h at bacterial communities con tained low proportions of Gram-positive bacteria wit h hig h incidence of Enterobacteriaceae and Vibrionaceae and practically no culturable actinomycetes. CY ANOBACTERIA AND MICROALGAE In caves cyanobacteria and algae can be found in wa ter bodies (Kue hn et al. 1992, Sanc h ez et al. 2002) and in aerop h ytic subaerial h abitats (Golubi 1967, Dobat 1970). Aerop h ytic cyanobacteria and algae are easily observed in t h e cave entrance illuminated by direct or indirect sunlig h t and, in s h ow caves equipped wit h arti cial illumination, as a part of a lampenora community around lamps. Dobat (1972) named spots wit h growing lampenora ecosystems in formation. In Slovenian sub aerial environments 197 cyanobacterial and algal taxa h ave been identied up to date by dierent aut h ors: rna jama (Martini et al. 1981), Huda luknja (Krivograd Klemeni 2007), Kostanjevika jama (Mulec 2005), Krka jama (Krivograd-Klemeni & Vr h ovek 2005), a lead and zinc mine in Meica (Mulec 2005), a mercury mine in Id rija (Mulec 2005), Pekel pri Zalogu cave (Martini et al. 1981, Mulec 2005), Pilanca cave (Martini 1978), Pivka jama (Martini et al. 1981, Mulec 2005), Posto jnska jama (Dobat 1973, Martini et al. 1981, Mulec 2005), Raike ponikve cave (Mulec 2005), kocjanske jame (Golubi 1967, Martini et al. 1981, Mulec 2005), and upanova (Taborska) jama (Martini et al. 1981, Mulec 2005). In t h e last compre h ensive study on cave p h ototrop h s Mulec (2005) found out t h at Aphanocapsa muscicola, Chlorella sp., Lyngbya sp., Synechocystis sp. and Trentepohlia aurea are very common taxa in cave h abitats. In t h e same study on poorly illuminated walls in kocjanske jame a typi cal cave cyanobacteria Geitleria calcarea was reported for t h e rst time in Slovenia (Mulec 2005). From t h e cave entrance in kocjanske jame in t h e p h ototrop hic com munity of stromatolitic stalagmites, 35 taxa were identi ed wit h a low portion of coccoid cyanobacteria and taxa suc h as Calothrix sp., Homoeothrix sp. and Schizothrix sp. In carbonate depositions on t h e illuminated side of stalactites in t h e same cave entrance 14, mainly coccoid, cyanobacteria were identied (Mulec et al. 2007). F UNGI An important group of organisms found in caves are fun gi colonizing dierent substrates, including excrement of hig h er animals t h at occasionally enter or live inside caves. Fungi usually form complex microbial consortium wit h ot h er microorganisms. For example, among diverse bacterial biota in microbial mat covering rocky surface Meguar and Sket (1977) identied P enicillium sp. and Aspergillus sp. and in weat h ered limestone Mulec et al. (2002) isolated fungus belonging to Cladosporium her barum group. J ANEZ MULEC

PAGE 5

ACTA CARSOLOGICA 37/1 2008 157 Using t h eir adaptability and versatile metabolic pat h ways, fungi develop various types of interactions wit h ot h er organisms. During t h e investigation of t h e mycoora associated wit h cave-in h abiting fauna from Kozlov rob several h yp h omycetes and zygomycetes were identied on living larvae, adults and dead bodies of cave crickets Troglophilus neglectus (Zalar et al. 1997, GundeCimerman et al. 1998). In association wit h t h e cave cricket a new species of M ucor troglophilus was described w hic h was regularly found on living T. neglectus adults and on larvae (Zalar et al. 1997). From dead bodies of T. neglectus 30 fungal species were isolated representing 21 dierent genera, 18 species and 12 genera from t h e dead larvae and 15 genera and 19 species from t h e body of adult animals. Fungi isolated wit h t h e hig h est frequen cy from t h e larval stage belonged to t h e genus M ucor e dominant fungus isolated from t h e adult stage was an entomopat h ogen B eauveria bassiana Besides ot h er facultative pat h ogens suc h as Aspergillus, Cladosporium, P enicillium, M ortierella and Geotrichum several ubiqui tous soil saprop h ytic fungi were isolated as well (GundeCimerman et al. 1998). Troglop hilic mot h s Scoliopteryx libatrix and Triphosa dubitata are frequently infected wit h entomopat h ogenic fungi (Kubtov & Dvotk 2005). Life and dead imagos of troglop hilic mot h s S. libatrix and T. dubitata from 14 Slovenian caves (Apolonova jama, Beka jama, Dimnice, Huda luknja, Jama pri Pru h u, Jama koljka, Koblar ska jama, Koloevka, Makovca, Petrova jama, Pilanca, pe h ovka, Turkova jama, Zelke jame) were screened for t h e presence of entomopat h ogenic fungi. B eauveria bassiana compared wit h t h e incidence of P aecilomyves farinosus and Lecanicillum fusisporum, appeared wit h t h e hig h est frequency (Tkavc 2007). e neotenic cave amp hibian P roteus anguinus is t h e only obligate cave-dwelling vertebrate in Europe and t h e most c h aracteristic animal of underground waters in Dinaric karst. More t h an 250 localities of P anguinus are known (Sket 1997). At some specimens kept under laboratory conditions at t h e Department of Biology in t h e Biotec hnical faculty, University of Ljubljana, sudden development of disease symptoms and deat h of animals aer few days was recorded. e causative agent were fungi from t h e genus Saprolegnia w hic h h yp h ae penetrat ed deep into skin tissue and in t his way most probably destroyed t h e osmotic equilibrium in P anguinus. Fungi from t h e genus Saprolegnia are ubiquitous in fres h water bodies (Kogej 1999). P ROTOZOA Amoebae and ot h er protozoans are an integral part of all ecosystems. eir dynamics and community structure is an important indicator of biotic and abiotic c h anges in t h e environment. In Slovenian caves up to now no compre h ensive study h as been performed on t his topic. Mulec and W aloc hnik (2007) isolated from a cave pool wit h calcite ras in Peina v Bortu cave t h e potentially pat h ogenic Acanthamoeba castellanii genotype T4 and Hartmannella vermiformis Bot h amoebae can serve as vectors for intracellular pat h ogenic bacteria. From weat h ered limestone from t h e same cave va h lkampids were identied (Mulec & W aloc hnik 2007). BIOTECHNOLOGICAL ASPECT OF CAVE MICROORGANISMS Biotec hnological and bioremediation potential of cave in h abiting microorganisms is still not exploited enoug h. Many microbes h ave t h e potential to h arbour dierent important substances w hic h can be eective under low (cave) temperature and t h us interesting for industry suc h as antibiotics and tumour suppression substances. For example in Carlsbad Cavern novel spe cies of microorganisms t h at can degrade complex h az ardous aromatic compounds, suc h as benzot hiazola and benezenesulfonic, were isolated. Microbes can use t h ese compounds involved in t h e manufacture of plastics as an energy source for t h eir growt h (Barton 2006). Microor ganisms isolated from cave-dwelling fauna are also a sig nicant source of biotec hnological important substances. For example some fungal isolates from t h e cave cricket Troglophilus neglectus obtained in previous studies (Zalar et al. 1997, Gunde-Cimerman et al. 1998) s h owed hig h c hitinolytic, lipolytic and proteolytic activities (Glavan 1997). MICROORGANISMS IN H Y POGEON: EXAMPLES FROM SLOVENIAN KARST CAVES

PAGE 6

ACTA CARSOLOGICA 37/1 2008 158 REFERENCES Asencio, A.D. & M. Aboal, 2001: Biodeterioration of wall paintings in caves of Murcia (SE Spain) by epi lit hic and c h asmoendolit his microalgae.Arc hiv fr Hydrobiologie: Supplementband 140, Algological studies 103, 131-142, Stuttgart. Barton, H.A., 2006: Introduction to cave microbiology: a review for t h e non-specialists.Journal of cave and karst studies, 68, 2, 43-64, Huntsville. Barton, H.A. & V. Jurad o, 2007: Wh ats up down t h ere? Microbial diversity in caves.Microbe (W as hington, D.C.), 2, 3, 132-138, W as hington. Buttner, M.P., K. W illeke & S.A. Grins h pun, 1997: Sam pling and analysis of airborne microorganisms.In: Hurst, C.J., G.R. Knudsen, M.J. McInerney, L.D. Stetzenbac h & M.V. W alter (eds.), Manual of envi ronmental microbiology. American society for mi crobiology, 629-640, W as hington. Caaveras, J.C., S. Cuezva, S. Sanc h ez-Moral, J. Lario, L. Laiz, J.M. Gonzales &, C. Saiz-Jimenez, 2006: On t h e origin of ber calcite crystals in moonmilk depos its.Naturwissensc h aen, 93, 27-32, Heidelberg. Caaveras, J.C., S. Sanc h ez-Moral, V. Soler, & C. Saiz-Ji menez, 2001: Microorganisms and microbially in duced fabrics in cave walls.Geomicrobiology jour nal, 18, 3, 223-240, New Y ork. Dobat, K., 1970: Considrations sur la vgtation cryp togamique des grottes du Jura Souabe (sud-ouest de lAllemagne).Annales de splologie, 25, 4, 872907, Paris. Dobat, K., 1972: Ein kosystem in Auau: Die Lam penora Sc h au h h len.Umsc h au in W issensc h a und Tec hnik, 72, 15, 493-494, Frankfurt. Dobat, K., 1973: Ein Beitrag zur eingangs-, Lampenund Pilzora der Postojnska jama (Adelsberger Grotte bei Postojna, Jugoslawien).Razprave, Slovenska akademija znanosti in umetnosti, 16, 2, 123-143, Ljubljana. Dobat, K., 1998: Flore de la lumire articille (lampen ora-maladie verte).In: Jubert hie, C., & V. Decu (eds.), Encyclopaedia biospeleologica, Tome 2, So cit de Biospologie, 1325-1335, Moulis-Bucarest. Engel, A.S., L.A. Stern & and P.C.Bennett, 2004: Micro bial contributions to cave formation: New insig h ts into sulfuric acid speleogenesis.Geology, 32, 369372, Boulder. Geri, B., T. Pipan & J. Mulec, 2004: Diversity of cul turable bacteria and meiofauna in t h e epikarst of kocjanske jame Caves (Slovenia).Acta carsolog ica, 33, 1, 301-309, Ljubljana. Slovenian karst caves oer a great potential in many as pects for cave microbiology. It is wort h to continue locat ing and identifying microorganisms w hic h in h abit dif ferent nic h es in caves, e.g. various bio-mediated crusts on sediments, calcite ras, cave pearls, Studies s h ould go beyond t h e taxonomy in a direction of understand ing t h e role of microbes in geoc h emical processes. e evidence for extant microbial processes in caves became more compelling wit h t h e discovery of ot h er mineral de posits w h at were dicult to explain t hroug h purely geo logic or inorganic processes, suc h as pool ngers. ese structures were proved to be microbial in origin. Today we know t h at many carbonate minerals wit hin caves are associated wit h t h e h eterotrop hic bacteria, including metastable carbonates, suc h as valerite (Barton & Jurado 2007). To study cave microbiology it is of crucial impor tance to preserve caves in t h eir intact state; t h at is w h y in troduction of alien microbiota in caves must be reduced to minimum. Cave microbiology oers immense poten tial to study evolution relations hips and use of alternative sources of energy developed by microbes for scavenging of scarce nutrients in oligotrop hic environments. It can also give answers of limits of life and h elp us to identify t h e geoc h emical signatures of life. FUTURE PERSPECTIVES ACKNO W LEDGEMENT Some results included in t his paper were accomplis h ed in t h e framework of t h e project no. Z6-7072 Role and signicance of microorganisms in karst processes sup ported by t h e Slovenian Researc h Agency. J ANEZ MULEC

PAGE 7

ACTA CARSOLOGICA 37/1 2008 159 Glavan, G., 1997: Production of enzymes by M ucor fungi, isolated from cave cricket Troglophilus neglectus. Graduation t h esis, University of Ljubljana, Biotec h nical Faculty, p. 64, Ljubljana. Golemansky, V. & L. Bonnet, 1994: Protozoa.In: Juber t hie, C. & V. Decu (eds.), Encyclopaedia Biospele ologica, Vol. 1, Socit de Biospologie, 23-33 Mou lis, Bucarest. Golubi, S., 1967: Algenvegetation der Felsen: Eine kolo gisc h e Algenstudie im dinarisc h en Karstgebiet.In: Elster, H.J. & W O h le (eds.), Die Binnengewsser, Band XXIII, E. Sc h weizerbartsc h e Verlagsbuc hh an dlung, p. 183, Stuttgart. Grot h, I., R. Vettermann, B. Sc h uetze, P. Sc h umann & C. Saiz-Jimenez, 1999: Actinomycetes in karstic caves of nort h ern Spain (Altamira and Tito Bustillo).Journal of microbiological met h ods, 36, 115-122, Amsterdam. Gunde-Cimerman, N., P. Zalar & S. Jeram, 1998: Myco ora of cave cricket Troglophillus neglectus cadav ers.Mycopat h ologia (1975), 141, 111-114, Den Haag. Hill, C. & P. Forti, 1997: Cave Minerals of t h e W orld.2nd edition. National speleological society, p. 463, Huntsville. Kinkle, B.K. & T.C. Kane, 2000: C h emolit h oautotrop hic micro-organisms and t h eir potential role in subsur face environments.In: W ilkens, H., D.C. Culver & W Hump hreys (eds.), Ecosystems of t h e W orld 30, Subterranean ecosystems, Elsevier, 309-318, Am sterdam. Knez, M., T. Slabe & S. ebela, 2004: Karstication of t h e aquifer discovered during t h e construction of t h e expressway between Klanec and rni Kal, Classical Karst.Acta carsologica, 33, 1, 205-217, Ljubljana. Kogej, T., 1999: Infection of P roteus anguinus wit h t h e fungi of t h e genus Saprolegnia .Graduation t h esis, University of Ljubljana, Biotec hnical Faculty, p. 97, Ljubljana. Krivograd Klemeni, A, 2007: Algae in selected aquat ic and terrestrial h abitats oristic and ecological view.Dissertation t h esis, University of Ljubljana, Biotec hnical Faculty, p. 210, Ljubljana. Krivograd-Klemeni, A. & D. Vr h ovek, 2005: Algal ora of Krka jama cave, Slovenia.Sbornk nrod n h o muzea v Praze, ada B, Ptrodn vdy, Historia naturalis, 61, 1-2, 77-80, Pra h a. Kubtov, A. & L. Dvotk, 2005: Entomopat h ogenic fungi associated wit h insect hibernating in under ground s h elters.Czec h mycology, 57, 3-4, 221-237, Pra h a. Kue hn, K.A., R.M. Oneil & R.D. Koe hn, 1992: Viable p h otosynt h etic microalgal isolates from ap h otic en vironments of t h e Edwards aquifer (central Texas).Stygologia, 7, 3, 129-142, Leiden. Martini, A., 1978: Primarna produkcija v jamski h sistemi h, 1. faza.Raziskovalna skupnost Slovenije, p. 53, Ljubljana. Martini, A., D. Vr h ovek & F. Bati, 1981: Flora v ja ma h z umetno osvetlitvijo.Bioloki Vestnik, 29, 2, 27-56, Ljubljana. Meguar, F. & B. Sket, 1977: On t h e nature of some or ganic covers on t h e cave-walls.Proceedings of t h e 6 t h international congress of Speleology, Academia, 159-161, Olomouc. Mulec, J., 2005: Algae in t h e karst caves of Slovenia.Dis sertation t h esis, University of Ljubljana, Biotec hni cal Faculty, p. 149, Ljubljana. Mulec, J., G. Kosi & D. Vr h ovek, 2007: Algae promote growt h of stalagmites and stalactites in karst caves (kocjanske jame, Slovenia).Carbonates and evap orates, 22, 1, 6-9, Troy. Mulec, J., G. Kosi & D. Vr h ovek, in press: C h aracteriza tion of cave aerop h ytic algal communities and ef fects of irradiance levels on production of pigments.Journal of cave and karst studies, Huntsville. Mulec, J., P. Zalar, N. Zupan Hajna & M. Rupnik, 2002: Screening for culturable microorganisms from cave environments (Slovenia).Acta carsologica, 31, 2, 177-187, Ljubljana. Mulec, J. & J. W aloc hnik, 2007: Amoebae in carbonate precipitating microenvironments in karst caves.Geop h ysical Researc h Abstracts, v. 9, European Geosciences Union, W ien. Pipan, T., 2005: Epikarst a promising h abitat: copepod fauna, its diversity and ecology: a case study from Slovenia (Europe).ZRC Publis hing, p. 101, Lju bljana. Pipan, T. & C. Culver, 2007: Regional species ric hness in an obligate subterranean dwelling fauna epikarst copepods.Journal of biogeograp h y, 34, 854-861, Oxford. Por, F.D., 2007: Op h el: a groundwater biome based on c h emoautotrop hic resources. e global signi cance of t h e Ayyalon cave nds, Israel-. Hydrobio logia, 592, 1-10, Den Haag. Sanc h ez, M., J. Alcocer, E. Escobar & A. Lugo, 2002: P h y toplankton of cenotes and anc hialine caves along a distance gradient from t h e nort h eastern coast of Quintana Roo, Y ucatan Peninsula.Hydrobiologia, 467, 1-3, 79-89, Den Haag. Sarbu, S.M., T.C. Kane & B.K. Kinkle, 1996: A c h emo autotrop hically based cave ecosystem.Science, 272, 5270, 1953-1955, New Y ork. MICROORGANISMS IN H Y POGEON: EXAMPLES FROM SLOVENIAN KARST CAVES

PAGE 8

ACTA CARSOLOGICA 37/1 2008 160 Sc h abereiter-Gurtner, C., C. Saiz-Jimenez, G. Pinar, W Lubitz & S. Rolleke, 2002: Cave paleolit hic paintings h arbour complex and partly unknown microbial communities.FEMS microbiology letters, 211, 711, New Y ork. Sket, B., 1997: Distribution of P roteus (Amp hibia: Urode la: Proteidae) and its possible explanation.Journal of biogeograp h y, 24, 263-280, Oxford. Taboroi, D., 2006: Biologically inuenced carbonate speleot h ems.In: Harmon, R.S & C. W icks (eds.), Special Paper 404: Perspectives on Karst Geomor p h ology, Hydrology, and Geoc h emistry A Tribute Volume to Derek C. Ford and W illiam B. White, Geological Society of America, 307-317, Boulder. Tkavc, R., 2007: Identication and genotypization of entomopat h ogenic fungi isolated from troglop hile mot h s Scoliopteryx libatrix L. and Triphosa dubitata L..Graduation t h esis, University of Ljubljana, Bio tec hnical Faculty, p. 80, Ljubljana. Zalar, P., G.L. Hennebert, N. Gunde-Cimerman & A. Cimerman, 1997: M ucor troglophilus, a new spe cies from cave crickets.Mycotaxon, 65, 507-516, It h aca. Zupan Hajna, N., 2003: Incomplete solution: weat h ering of cave walls and t h e production, transport and de position of carbonate nes.ZRC Publis hing, p. 167, Ljubljana. J ANEZ MULEC


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.