Acta carsologica

Citation
Acta carsologica

Material Information

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

Subjects

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

Notes

Restriction:
Open Access - Permission by Publisher
Original Version:
Vol. 36, no. 1 (2007)

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
K26-00156 ( USFLDC DOI )
k26.156 ( USFLDC Handle )
5217 ( 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:

On Denudation Rates In Karst / Franci Gabrovšek ( .pdf )

Variation In Rates Of Karst Processes / Arthur N. Palmer ( .pdf )

Time Scales In The Evolution Of Solution Porosity In Porous Coastal Carbonate Aquifers By Mixing Corrosion In The Saltwater-Freshwater Transition Zone / Wolfgang Dreybrodt - Douchko Romanov ( .pdf )

The Age Of Karst Relief In West Slovenia / Andrej Mihevc ( .pdf )

Evolution And Age Relations Of Karst Landscapes / William B. White ( .pdf )

Cave And Karst Evolution In The Alps And Their Relation To Paleoclimate And Paleotopography / Philippe Audra - Alfredo Bini - Franci Gabrovšek - Philipp Häuselmann - Fabien Hobléa - Pierre-Yves Jeannin - Jurij Kunaver - Michel Monbaron - France Šušteršic - Paola Tognini - Hubert Trimmel - Andres Wildberger ( .pdf )

Aspects Of The Evolution Of An Important Geo-Ecosystem In The Lessinian Mountain (Venetian Prealps, Italy) / Leonardo Latella - Ugo Sauro ( .pdf )

What Does The Distribution Of Stygobiotic Copepoda (Crustacea) Tell Us About Their Age? / David C. Culver - Tanja Pipan ( .pdf )

How To Date Nothing With Cosmogenic Nuclides / Philipp Häuselmann ( .pdf )

Upper Cretaceous To Paleogene Forbulge Unconformity Associated With Foreland Basin Evolution (Kras, Matarsko Podolje And Istria; Sw Slovenia And Nw Croatia) / Bojan Otonicar ( .pdf )

A Review Of Coalesced, Collapsed-Paleocave Systems And Associated Suprastratal Deformation / Robert G. Loucks ( .pdf )

Clastic Sediments In Caves - Imperfect Recorders Of Processes In Karst / Ira D. Sasowsky ( .pdf )

Analysis Of Long-Term (1878-2004) Mean Annual Discharges Of The Karst Spring Fontaine De Vaucluse (France) / Ognjen Bonacci ( .pdf )

Timing Of Passage Development And Sedimentation At Cave Of The Winds, Manitou Springs, Colorado, Usa / Fred G. Luiszer ( .pdf )

How Long Does Evolution Of The Troglomorphic Form Take: Estimating Divergence Times In Astyanax Mexicanus / Megan L. Porter - Katharina Dittmar - Marcos Pérez-Losada ( .pdf )

Age Estimates For Some Subterranean Taxa And Lineages In The Dinaric Karst.- Acta Carsologica / Peter Trontelj - Špela Goricki - Slavko Polak - Rudi Verovnik - Valerija Zakšek - Boris Sket ( .pdf )

The Challenge Of Estimating The Age Of Subterranean Lineages: Examples From Brazil / Eleonora Trajano ( .pdf )


Full Text

PAGE 1

ON DENUDATION RATES IN KARST O HITROSTI DENUDACIJE NA K RASU Franci G ABROVEK 1 Izvleek UDK 551.331.24:551.44 Franci Gabrovek: O hitrosti denudacije na Krasu V prispevku predstavim enostaven matematini model s katerim raziskujem dinamiko znievanja krakega povrja. Predpostavim enakomerno napajanje s povrja in vertikalno pronicanje vode. Denudacijsko stopnjo izraunam iz asa, ki je potreben za odstranitev doloene debeline kamninskega sloja. Konkretno to naredim na primeru apnenca v katerem se voda pretaka v sistemu vertikalnih razpok. Hitrost denudacije naraa z debelino odstranjene plasti in dosee zgornjo mejo, ki je doloena z enabami, ki temeljijo na predpostavki, da se celo ten korozivni potencial vode manifestira v znievanju povrja. Kljune besede: kras, denudacijska stopnja, raztapljanje ap nenca, matematini model. 1 Karst Research Institute ZRC SAZU, Postojna, Slovenia, e-mail: gabrovsek@zrc-sazu.si Received/Prejeto: 01.02.2007 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 7 Abstract UDC 551.331.24:551.44 Franci Gabrovek: On denudation rates in Karst Paper presents a simple mathematical model, which enables study of denudation rates in karst. A vertical ow of water which is uniformly inltrated at the surface is assumed. Denu dation rate is calculated from the time needed to remove certain thickness of rock. is is done concretely on a limestone block dissected by a vertical array of fractures. It is shown that denu dation rate increases with the thickness of removed layer and approaches an upper limit which is dened by the maximum denudation equations, which are based on assumption that all dissolution potential is projected into a surface lowering. Keywords: karst, denudation rate, limestone dissolution, math ematical model. I NTRODUCTION Uniform lowering or surface denudation is a dominant karstication process (Dreybrodt, 1988; Ford & W illiams, 1989; W hite, 1988). e denudation rate is dened as the rate (LT -1 ) of lowering of a karst surface due to the dissolution of bedrock. A common approach used to estimate the denudation rate is based on the presumed equilibrium concentration (or hardness) and the amount of water which inltrates into the subsurface. It is sum marized in the famous Corbels equation (Corbel, 1959): 1 e inltrated water in mm/y is the dierence between precipitation P and evapotranspiration E. H is the equilib rium concentration (Hardness) in mg/L of dissolved rock, is the density of limestone in g/cm 3 f denotes the portion of soluble mineral in the rock, which will be 1 in this paper. e factor 1000 corrects for the mixture of units used in the equation. ere are more general equations of this kind like that of W hite (1984, this issue). For a Limestone terrain in a temperate climate all these equations give denuda tion rate of the order of several tens of meters per million years. Similar results are obtained from ow and con centration measurements in rivers which drain a known catchment area. From the measured data the total rock volume removed from the area in a given time period can be calculated. Dividing the removed volume by the surface of the area and the time interval gives the denu dation rate.

PAGE 2

Eq.1 implies that all dissolution capacity of water is used in the rock column, i.e. the solution at the exit of rock block is close to saturation. Among the many assumptions behind such estimations of the denuda tion rate I will address two of which at least one must be valid: 1. Most of the dissolution occurs close to the sur face, i.e. within epikarst. 2. In the long term, the dissolution at depth is inte grated into a surface denudation. It is the intention of this paper to theoretically vali date maximum denudation approach. S URFACE LO W ERING AND THE VOLUME OF DISSOLVED ROCK Dissolution of any rock is not instantaneous, but proceeds at some nite rates. In conditions of diuse inltration through the karst surface and prevailing vertical ow, the concentration of dissolved rock in the inltrating water will normally increase with the depth as schematically shown by color intensity in Fig.1. Fig. 2 presents point at some depth z below the sur face. e volume V of rock dissolved per unit surface area S in time t between the surface and the point is given by 2 where c(z) is the concentration of dissolved rock [M/L 3 ] at the depth z, q is the inltration rate at the surface [L 3 / (L 2 T)] and is the density of the rock [M/L 3 ]. Due to the surface lowering, the depth of the point is decreasing according to z(t) = z 0 Dt, where z 0 is the depth at t = 0 and D is the denudation rate (Fig. 2). e volume of dissolved rock per surface area in time T above the point is then given by: 3 Introducing a new variable z=z 0 Dt into the right hand integral in Eq. 3 gives: 4 Fig. 1. Section of a terrain with a uniform surface inltration of aggressive solution and prevailing vertical ow. Color intensity denotes that the concentration of dissolved rock increases with depth. Fig. 2: Idealized prole through the rock column at time t = 0 (le) and t > 0 (right). e depth of the point which is at z 0 decreases in time due to the surface lowering. F RANCI G ABROVEK TIME in KARST 2007 8

PAGE 3

e complete volume of rock initially above z 0 is z 0 S. To remove this volume a time T D is needed, where z 0 = DT D Using all this in Eq. 4, we obtain: 5 6 As given, D is an average denudation rate, cal culated from the time T D needed to remove a layer of thickness z 0 from the rock column with initial a uni form porosity distribution in vertical direction. If a rock layer has a nite thickness, z 0 can be taken as the layer thickness, T D the time needed to remove the com plete layer and D an avarege denudation rate. It is easy to see that if the solution quickly attains equilibrium Eq. 6 gives maximum denudation rates: 7 If this is not the case D will be below D c since inte gral with c eq is maximal. In this case we rewrite Eq. 6 as: 8 W ith increasing layer thickness an average concen tration within the layer increases and average denudation rates approach maximal. C ALCULATION OF THE CONCENTRATION PROFILE e results given so far are valid for any natural c(z). To obtain some quantitative results we revert to a special case where the calcite aggressive water is inltrating into a vertical fracture network. erefore we need to couple the rate equation for limestone and ow of laminar lm down a vertical fracture wall. L IMESTONE DISSOLUTION RATES Recently Kaufmann & Dreybrodt (2007) published the corrected rate equation with two linear regions and a non-linear region of dissolution kinetics: 9 e kinetic constants and rate orders are derived from theoretical and experimental results (Buhmann & Dreybrodt, 1985; Dreybrodt, 1988; Eisenlohr et al., 1999; Kaufmann & Dreybrodt, 2007). Values depend on the temperature, p CO2 and laminar layer thickness and are given in Kaufmann & Dreybrodt (2007). W e will use 1 = 3 -4 cm/s and 2 = 8 -6 cm/s, values which are valid at 10C for the open system dissolution (Kaufmann & Drey brodt, 2007). Nonlinear kinetics will not be discussed here. It is valid close to equilibrium and does not change the results substantially. c eq depends on the p CO2 temper ature, the presence of the foreign ions and the nature of the system where dissolution proceeds (open, closed, in termediate) (Appelo & Postma, 1993; Dreybrodt, 1988). e calcium equilibrium concentration normally takes values between 0.5 mmol/l 3 mmol/l, which in terms of dissolved calcite means 50 mg/l L AMINAR FLO W DO W N A SMOOTH VERTICAL W ALL Only rough assumptions can be made about the ow re gime of inltrated water. W e assume a laminar lm ow down the walls of vertical fractures. e velocity of such lm is given by (Bird et al., 2002): 10 where is the lm thickness, g gravitational acceleration and the kinematic viscosity. More suitable master vari able is a ow density along the fracture walls q f (cm 2 /s). Applying we get: 11 ON DENUDATION RATES IN KARST TIME in KARST 2007 9

PAGE 4

TIME in KARST 2007 10 T HE COUPLING OF FLO W AND DISSOLUTION To get the evolution of a concentration in a falling lm, the mass balance for dissolved calcite must be coupled to the rate laws given in Eq. 9. In a small volume of water lm with thickness width b, length dz and concentra tion c, the mass balance requires: 12 where A is the surface of the water rock contact (bdz). Integration of Eq. 12 gives the evolution of concentration in time as the lm ows down the fracture wall: 13 where Note that 1 is more than an order of magnitude smaller than 2 The time domain can easi ly be converted into the space domain using and Fig.3: Concentration prole in a lm owing down a smooth vertical fracture and dissolving limestone walls. V alues of are calculated from the fracture ow density obtained if the fracture spacing is 1 m and inltration intensity is 0.114 mm/h, 10 mm/h, 20 mm/h, 30 mm/h and 40 mm/h for curves 1-5 respectively. 14 where Fig. 3 presents the evolution of satu ration ratio c(z)/c eq for dierent 2 For most reasonable scenarios, the rst linear kinetics is active only close to the surface. erefore, it will be integrated directly into the surface lowering (i.e. c(z= 0) = 0.3 c eq ). S ATURATION LENGTH AND THE FRACTURE FLO W DENSIT Y Saturation length 2 controls the vertical evolution of concentration prole. It depends on the kinetic constant and the fracture ow density. To estimate the latter we assume that the rain falling to the surface with an inten sity q is evenly inltrated into a regular grid of fractures as shown on Fig. 4. e ow density in each fracture is proportional to the ratio between the surface of the inl tration area and the total breadth of the fractures drain ing the area. In a regular grid of fractures with fracture spacing d we obtain: 15 Fig. 4: Rain falling with intensity q [L T -1] is uniformly distributed into the fractures with ow density qf [L T -1] according to Eq.15. F RANCI G ABROVEK

PAGE 5

TIME in KARST 2007 11 where N depends on the geometry of fracture grid (e.g. for a series of parallel fractures and for a square grid). For thin lms ( < 0.005 cm) the rates are controlled by conversion of CO 2 into H + and HCO 3 and therefore increase linearly with lm thickness. Inltration intensity below 3 mm/h into a series of fractures with d = 100 cm produces lm thicknesses below 0.005 cm. Extremely low inltration intensities (< 1 mm/h) and fracture spacing in the range of few centimeters result in lm thicknesses in the order of 0.001 cm. is reduces the kinetic con stant approximately by a factor of 5. For thin lms, the lm thickness and decrease with q f 1/3 Consequently 2 is proportional to the q f 2/3 is has limited consequences for the dissolved volume and denudation rate discussed in the next section. In the early stages, the fracture ow is not expected to be in the form of a free surface lm, but full fracture ow instead. e saturation lengths in that case would be smaller than those derived here. e evolution of such fractures is given in Dreybrodt et al. (2005). R ESULTS AND DISCUSSION Inserting the concentration prole from Eq. 14 (second linear region only) into Eq. 7 gives: 16 Although it is a matter of a denition, the average denudation rate given in Eqs. 8 and 16 are not exactly what we are aer. W hat we look for is the actual lowering of karst surface, which is given by dz 0 / dT D W e will not go into mathematical details of derivation, but instead discuss its consequences on a plot of z 0 (TD). Note that the z 0 (T D ) has no explicit form, but its inverse function does: 17 W e will demonstrate the results on a characteristic data for a moderate climate with I=1000 mm/y and rela tively bare karst area with c eq = 1mmol/l or H = 100 mg/L. For = 2.5 g/cm 3 D C for this case is 40 m/Ma. W e assume that the rain inltrates into a parallel set of fractures with spacing d = 1 m. Fig. 5a shows z 0 (T D ) for four different saturation lengths arising from different infiltration intensities. Y early infiltration is 1000 mm/y for all curves. There fore, the time period of dissolution is inversely propor tional to the infiltration intensity. Dashed line shows the uniform lowering by D C W e wee that all lines become practically parallel to maximum denudation line for z 0 > 2 2 .The actual denudation rate becomes maximal when the removed thickness is larger than 2 2 This is about the depth where the concentration reaches 90% of saturation. The slope of the dotted lines presents the averaged denudation rates for curve with 2 = 70 m. Fig. 5b shows the averaged rate for the same scenar ios as Fig. 5a. Red dashed red curve clearly shows the fast approach of the actual rate to maximal for 2 = 52.5 m. Another interesting conclusion can be made from Fig. 5a. Dierent saturation lengths can also arise from dierent fracture spacing (see Eq. 15 for qf) If we imag ine a region with high fracture density within a region of low fracture density, the rst will initially be denuded faster, but latter on both actual rates will become the same. erefore the dierence made at the onset will stay projected in the surface. is is shown by the double ar row between lines 3 and 4. ON DENUDATION RATES IN KARST

PAGE 6

TIME in KARST 2007 12 Fig. 5: a) e time dependence of removed thickness for several inltration intensities. I=1000 mm/y, H = 100 mg/L, = 2.5 g/cm 3 d = 100 cm, N = 2. D ashed line show the maximum denudation rate which is 40 m/M a. D otted lines present the time averaged denudation rates (Eq.16). D ouble arrow demonstrates the dierence between the denuded thicknesses which is kept in time due to the initial rate dierences. b) D ependence of average denudation rates on the removed thickness for the same scenarios as in Fig. 5a. D ashed line presents the actual surface lowering for 2 = 52.5 m. F RANCI G ABROVEK

PAGE 7

TIME in KARST 2007 13 C ONCLUSION Denudation rate in a block with initially uniform porosity increases as the denudation proceeds and becomes maxi mum denudation (Eq.1), when the thickness of removed layer is about twice the typical saturation length. Initial dierences arising from dierent saturation lengths re main imprinted in the surface. If a soluble layer has a nite thickness, the average denudation rate increases with the thickness, i.e. denuda tion is more eective on thick rock layers. e presented results are based on many assump tions which might not be valid. Nevertheless, it gives some theoretical validation of maximum denudation formulae and suggest some mechanisms that can cause irregularities in karst surface. REFERENCES Appelo, C. A. J. & D. Postma, 1993: Geochemistry, ground water and pollution. A.A. Balkema, xvi, 536 pp, Rot terdam; Brookeld, VT. Bird, R. B., Stewart, W E. & E.N. Lightfoot, 2002: T rans port phenomena. John W iley & Sons, Inc., xii, 895 p. pp, New Y ork, Chichester. Buhmann, D. & W Dreybrodt, 1985: e kinetics of cal cite dissolution and precipitation in geologically relevant situations of karst areas.1. Open system.Chemical geology, 48, 189-211. Corbel, J., 1959: Vitesse de lerosion.Zeitschri fur Geo morphologie, 3, 1-2. Dreybrodt, W ., Gabrovek, F. & D. Romanov, 2005: Pro cesses of speleogenesis: A modeling approach. Vol. 4, Carsologica Zaloba ZRC, 375 pp, Ljubljana. Dreybrodt, W ., 1988: Processes in karst systems: physics, chemistry, and geology. Springer-Verlag, xii, 288 p. pp, Berlin; New Y ork. Eisenlohr, L., Meteva, K., Gabrovek, F. & W Dreybrodt, 1999: e inhibiting action of intrinsic impurities in natural calcium carbonate minerals to their dissolu tion kinetics in aqueous H 2 O-CO 2 solutions.Geo chimica Et Cosmochimica Acta, 63, 989-1001. Ford, D.C. & P. W illiams, 1989: Karst geomorphology and hydrology. Unwin Hyman, 601 pp, London. Kaufmann, G. & W Dreybrodt, 2007: Calcite dissolutio n kinetics in the system CaCO 3 -H 2 O-CaCO 3 at high undersaturation.Geochimica Et Cosmochimica Acta, In Press. W hite, W .B., 1984: Rate processes: chemical kinetics and karst landform development. In: La Fleur (Ed.): Groundwater as a geomorphic agent. Allen and Un win, 227-248. W hite, W B., 1988: Geomorphology and hydrology of karst terrains. Oxford University Press, ix, 464 pp, New Y ork. ON DENUDATION RATES IN KARST



PAGE 1

VARIATION IN RATES OF KARST PROCESSES SPREMENLJIVOST HITROSTI KRA KIH PROCESOV A rthur N. P ALMER 1 Izvleek UDK 551.44 Arthur N. Palmer: Spremenljivost hitrosti krakih procesov Razvoj krasa ni lineareni proces, pa pa poteka s spremenljivo hitrostjo, znailna so tudi obdobja stagnacije in obdobja, ko je razvoj obrnjen v smeri manj zrele faze. Velikost in narava sprememb sta odvisni tudi od asovnega merila v katerem jih opazujemo. Dananja merjenja v jamah kaejo, da je hitrost raztapljanja odvisna od letnega asa, pretoka in pogojev v prsti. Raztapljanje je obasno prekinjeno z obdobjem izloanja. Izmerjene hitrosti raztapljanja lahko ekstrapoliramo v asu in na osnovi tega sklepamo o rasti doloene korozijske oblike. Vendar bomo pri tem storili napako, saj merjenja ne vsebujejo dolgoasovnih sprememb. Te so lahko posledica razlinih de javnikov, kot so klimatske spremembe in spremembe, ki nas tanejo zaradi samega razvoja krasa (npr. uhajanje CO 2 zaradi odpiranja jamskih vhodov). V asovnem merilu 10 5 -10 6 let raz voj krasa prekinjajo ali pospeujejo spremembe erozijske baze in spremembe povrinskih vodotokov. Tak primer je povezava med razvojem krasa in doline reke Ohio v vzhodnem delu cen tralnih ZDA. V asovnem merilu 10 6 -10 8 let tektonski in strati grafski dogodki povzroajo dolgoasovne spremembe v razvoju krasu. Tak primer je kras v Skalnem gorovju v Severni Ameriki. Dvem fazam zakrasevanja v karbonu je sledil pokop in miner alna zapolnitev med permom in kredo. Temu je sledil obiren razvoj jam med paleocensko-eocenskim dvigom ter stagnacija in delna mineralna zapolnitev v poznoterciarni agradaciji. V tako velikem asovnem merilu je teko doloiti hitrost razvoja krasa, e sploh. Primerneje je, da razvojno zgodovino razdelimo v obdobja, ki ustrezajo pomembnejim regionalnim tektonskim in stratigrafskim dogajanjem. Kljune besede: razvoj krasa, hitrost raztapljanja, procesi naza dovanja, paleokras. 1 Department of Earth Sciences, State University of New Y ork, Oneonta, NY 13820-4015, U.S.A. e-mail: palmeran@oneonta.edu Received/Prejeto: 27.11.2006 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 15 Abstract UDC 551.44 Arthur N. Palmer: Variation in rates of Karst processes e development of karst is not a linear process but instead takes place at irregular rates that typically include episodes of stagnation and even retrograde processes in which the evolu tion toward maturity is reversed. e magnitude and nature of these irregularities diers with the length of time considered. Contemporary measurements in caves show uctuations in dissolution rate with changes in season, discharge, and soil conditions. Dissolution is sometimes interrupted by intervals of mineral deposition. Observed dissolution rates can be ex trapolated to obtain estimates of long-term growth of a solu tion feature. But this approach is awed, because as the time scale increases, the rates are disrupted by climate changes, and by variations that are inherent within the evolutionary history of the karst feature (e.g., increased CO 2 loss from caves as en trances develop). At time scales of 10 5 -10 6 years, karst evolution can be interrupted or accelerated by widespread uctuations in base level and surface river patterns. An example is the relation between karst and the development of the Ohio River valley in east-central U.S.A. At a scale of 10 6 -10 8 years, tectonic and stratigraphic events cause long-term changes in the mechanism and style of karst development. For example, much of the karst in the Rocky Mountains of North America has experienced two phases of pre-burial Carboniferous karst, mineral accretion during deep burial from Permian to Cretaceous, extensive cave development during Paleocene-Eocene upli, and stagnation and partial mineral deposition caused by late Tertiary aggrada tion. At such large time scales, it is dircult to determine rates of karst development precisely, if at all. Instead it is appropriate to divide the evolutionary history into discrete episodes that cor relate with regional tectonic and stratigraphic events. Key words: Karst evolution, dissolution rates, retrograde pro cesses, paleokarst.

PAGE 2

TIME in KARST 2007 16 INTRODUCTION In any discussion of the age of karst, one must consider the rates of the genetic processes and how they vary with time. ese are inuenced by the length of time over which they have operated. Karst development undergoes large variations in rate and is commonly interrupted by periods of stagnation or even retrograde processes in which mass is accumulated instead of removed. is pa per focuses on several eld examples that illustrate these processes and the dirculty of quantifying them. ese studies are still in progress and are used here only as points for discussion. SHORTTERM VARIATIONS IN DISSOLUTION RATE One approach to interpreting karst history is to measure current rates of bedrock dissolution, for example by ap plying the mass balance, or by measuring rates of bed rock retreat with micrometers or standardized bedrock tablets. In the two following studies, empirical kinetic equations are applied. On the basis of prior dissolution experiments, eld measurements of water chemistry are used to estimate dissolution or accretion rates at specic locations and times. Field example: eastern New York State Chemical measurements were made during 1985-1996 in streams of McFails Cave, New Y ork (Fig. 1; Palmer, 1996). Suitable data-loggers were not available for use in this ood-prone cave, so measurements were made ran domly at every opportunity. Although statistically shaky compared to continuous or short-interval sampling, this approach allowed full chemical analyses. e cave, in Silurian-Devonian limestones, consists of stream passages fed by dolines and ponors. Local soil P CO2 is 0.02-0.04 atm, but in this well-aerated cave the mean P CO2 of streams is only ~0.003 atm. Most measure ments were made in the main passage and were correlated with discharge, but this location was not accessible during high ow. To provide broader coverage, additional mea surements were made in similar passages with year-round accessibility. Chemical variations between sampling sites were negligible compared to variations with time. To al low extrapolation, the measurements were combined in a probability plot (Fig. 2), in which SI = log (IAP/K), IAP = (Ca 2+ )(CO 32), and K = calcite solubility product. Although the passages involved are active canyons, the water is conspicuously supersaturated except during the highest 20-30% of ow. At low ow the calcite SI oen exceeds +0.4 (~138% saturation). Calcite can precipitate Fig. 1: M ap of M cFails Cave, New Y ork, showing location of sampling sites. Fig. 2: Probability plot of calcite saturation index in M cFails Cave for the period 1985-1996, where SI = log(IAP/K). D ata points are triangles; X = example of probability interval used in T able 1. A RTHUR N. P ALMER

PAGE 3

TIME in KARST 2007 17 at approximately SI > +0.2, so why does it not precipitate in the cave at those times? During a particularly dry summer (1995), a con spicuous calcite layer did accumulate on the canyon oors. is coating averaged 0.3 mm thick with inclu sions of clay and quartz silt (Figs. 3 & 4). It was limited to surfaces that remained water-covered during lowest ow and formed a continuous layer in areas of steep gradi ent (supercritical ow) but only discontinuous patches in ponded water. In mid-January, 1996, heavy rain fell on rapidly melting snow and produced a ood with a return period of ~50 years. e main cave entrance was covered by 5 m of water, and smaller inputs contained roaring waterfalls. e calcite SI of the water entering the cave averaged 1.9 (cf. Fig. 2). is sample is not included in the statis tics, as it was not random, but obtained purposely at the ood peak, and it is not in the same class as the in-cave samples. However, it illustrates the high dissolutional ca pacity of extreme oodwater. e rate of limestone removal can be estimated by S = 31.56 k (1 C/C s ) n / cm/yr (Palmer, 1991), where S = rate of bedrock retreat, k = rate constant (mg-cm/L-sec), n = reaction order (di mensionless), C s = calcite saturation concentration, C = actual concentration of dissolved calcite, and = rock density (g/cm 3 ). C /C s is the saturation ratio, where 1.0 represents calcite saturation. From computer analysis, C/C s ~ (IAP/K) 0.35 For the cave conditions (mean P CO2 = 0.003 atm and T = 8C), laboratory measurements by Plummer et al. (1978) show that k ~0.01 and n ~2.2 at C/C s < 0.6, and k ~0.05 and n ~4 at C/C s > 0.6 in opensystem turbulent ow. Bedrock density is ~2.7 g/cm 3 in this low-porosity rock. From chemical measurements during the winter and spring of 1996, it was predicted that the entire calcite coating of 1995 should have been removed by the time the cave became accessible in May. In fact, all but a few sheltered remnants of the calcite had been removed by then. Although mechanical abrasion may have aided the removal in places, the agreement between prediction and result is mild support for the validity of this approach. Fig. 2 includes a best-t regression line through the chemical data. W here this line extends below saturation, the probability scale was divided into 5% increments. From the mean SI in each increment, a net dissolution rate of 1.3 x 10 -3 cm/yr was calculated for the period of study (Table 1). At that rate, the main cave stream would have deepened about 18 cm since the last glacial retreat in the region about 14,000 years ago. is is compatible with the presence of varved clays no more than a few cen timeters above the lowest bedrock oors. e clay was deposited when retreating glaciers blocked the local sur face river, ooding the valley and neighboring caves. Probability range Mean C/C s Mean S (cm/yr) Net annual entrenchment (cm) <0.05 ~0.52 ~0.017 ~8.5 x 10 -4 0.05 0.10 0.65 0.0064 3.2 x 10 -4 0.10 0.15 0.74 0.0019 9.5 x 10 -5 0.15 0.20 0.88 8.8 x 10 -5 4.4 x 10 -6 0.20 0.25 0.89 7.4 x 10 -5 3.7 x 10 -6 0.25 0.30 0.95 2.1 x 10 -6 1.1 x 10 -7 TOTAL: 1.3 x 10 -3 cm/yr 13 mm/1000 yrs T ab. 1: Net dissolution rate in M cFails Cave canyons, 19851996,where the best-t line in Fig. 2 falls below SI = 0. Entrenchment rates are calculated from the regression line, rather than from specic data points, and provide only a rough approximation. At the estimated entrenchment rate, the 10 m depth of the main McFails canyon would have required more than 700,000 years to form. is rate seems low for an active canyon with a gradient of 1.2 degrees, but it is Fig. 3: M ain stream of M cFails Cave during the summer of 1995, with calcite coating on oor of canyon. VARIATION IN RATES OF KARST PROCESSES

PAGE 4

TIME in KARST 2007 18 compatible with U/ speleothem dates. Related caves at the same elevation as McFails contain speleothems dated up to 277 ka (Dumont, 1995; Lauritzen & Mylroie, 2000; Mylroie & Mylroie, 2004). Some speleothems were lo cated near the cave oors, so the passages themselves are far older. But the entrenchment rate during this period must have varied because of climate changes and burial beneath glacial ice for several tens of thousands of years. (e only known glaciation in the area was W isconsin an.) e coarse bedload in parts of the cave also suggests mechanical abrasion during high ow. e entrenchment rate has probably decreased with time. W hen entrances were blocked by glacial sediment, or had not yet enlarged enough to form open holes, es cape of CO 2 to the surface must have been severely lim Fig. 4: in-section photomicrograph showing calcite crust on a limestone pebble from M cFails Cave (September, 1995). ited and the mean aggressiveness would have been higher than it is today. Also, calcareous glacial deposits cause low-ow inputs to be saturated with calcite before they even reach the cave. e main canyon of the cave has an entirely vadose origin because it extends exactly down the local dip of the strata, except where it is deected by joints (Fig. 1). erefore the canyon originated aer surface rivers had entrenched below its level (currently about 300 m above sea level). Although the age of the landscape is dircult to determine from the surface, data from the cave can provide helpful information. Mammoth Cave, Kentucky Meiman & Groves (1997), Anthony & Groves (1997), and Groves & Meiman (2005) conducted a similar study in the main river passage of Mammoth Cave, Kentucky. ey made a high-frequency record of water levels in monitor wells, combined with periodic measurements of water chemistry. To calculate dissolution rates, they used the kinetic equation described above. Because of thick sediment, cave enlargement rates could not be estimated precisely. However, the authors determined that during the highest 5% of ow, 38% of the mass was removed (vs. about 65% in McFails). e dierence is probably due, at least partly, to the lack of entrances near the sampling sites in Mammoth Cave through which CO 2 is lost, the higher carbonate content of soils in the New Y ork karst, and the dominance of sinking-stream inputs to McFails Cave during severe oods. VARIATION IN KARST PROCESSES AT TIME SCALES OF 10 5 10 6 Y EARS e low-relief karst plateaus of Kentucky and Indiana, U.S.A., are developed on early Carboniferous carbonates and include extensive doline elds bordered by sinking streams. ese include the Pennyroyal Plateau in Ken tucky and the Mitchell Plain in Indiana. ey are dissect ed to a maximum of 50-65 m by river valleys. Near rivers, inter-doline divides and residual at areas lie 175-190 m above sea level, and up to a few tens of meters higher elsewhere. Although resistant beds form local at areas, the overall surface is discordant to the strata. e surface is mantled in many places by residual, colluvial, and al luvial sediment up to 30 m thick, the surface of which is concordant with the erosion surface on nearby bed rock. In the Mitchell Plain the deposits are attributed to a widespread Tertiary rise in base level (Palmer & Palmer, 1975). On the Pennyroyal, Ray (1996) calls this relatively at surface the Green River Strath and attributes it to u vial processes. Caves are common in the karst plains and in adja cent sandstone-capped uplands. Mammoth Cave, Ken tucky, is the best-known upland example. Its highest pas sages correlate with nearby low-relief areas of the Pen nyroyal (Fig. 5), and passage patterns and gradients show that the Pennyroyal was the source of the cave water (Palmer, 1981). ese passages are mostly large canyons lled partly or completely with stream sediment (Fig. 6). Dating of these sediments by cosmogenic radionuclides gives ages up to 4 Ma (Granger et al. 2001), but in ar eas bordering the Green River (the outlet for Mammoth Cave water), most samples date to ~2.2 Ma (see also An thony & Granger, 2004, 2006). ese passages record a history of slow Tertiary entrenchment interspersed with aggradation, and with a widespread rise in base level of more than 20 m at ~2.2 Ma. e fragmentary sediment surfaces at the same elevation in the Pennyroyal must be correlative. e cause of the widespread aggradation at A RTHUR N. P ALMER

PAGE 5

TIME in KARST 2007 19 2.2 Ma is uncertain. It correlates roughly with the onset of widespread continental glaciation at higher latitudes, but it may relate more directly to a drying climate during the late Pliocene, which would have favored the accumu lation of sediments in lowlands. Pleistocene continental glaciers extended southward as far as northern Kentucky and caused much rearrange ment of surface drainage. Initial entrenchment below the uppermost passages in Mammoth Cave may have been triggered by the establishment of drainage from the Ap palachian Mountains westward to the Mississippi River, to form the so-called Teays River (Fig. 7; see Granger et al., 2001). Later, the previously tiny Ohio River be came one of the largest rivers on the continent when the Teays was diverted into it (Fig. 7). ese shis en hanced the rate of river entrenchment into the sediment-mantled plains of carbonate rock. Subsurface karst drainage developed and the surfaces became sinkhole plains. Pleistocene cave passages formed at various levels as much as 60-70 m below the Ter tiary passages. Again, caves provide clues to the interpretation of surface landscapes that cannot be discerned from surface observa tions alone. Could the karst plateaus have retained vestiges of their original at surface for 2 Ma without signicant lowering? Al though dolines extend deeply into them, nearly at rem nants of the sediment-covered and resistant bedrock sur faces remain at approximately the same elevations as the sediment in the upper-level passages of Mammoth Cave, which suggests that parts of the original surface have sur vived with little or no lowering. W hat is the current karst denudation rate? Much of the Mammoth Cave area is drained by the Turnhole Spring basin, which has an area of 220 km 2 (Q uinlan et al. 1983). In this basin, Hess (1974) measured a meanannual Ca content of ~60 mg/L and Mg of ~7.5 mg/L (see also Hess & W hite, 1993). ese measurements represent a mean dissolved load of ~0.044 cm 3 /L calcite and ~0.020 cm 3 /L dolomite (with the simplifying assumption that Fig.5: Location of M ammoth Cave and surrounding landscapes. M = M itchell Plain, P = Pennyroyal Plateau, U = sandstone-capped uplands. X = pre-Pleistocene head of Ohio River. 1, 2, 3 = sequence of drainage from Appalachian M ountains. 1 is probable but entirely hypothetical. 2 = late T ertiary T eays River, which is well known by its former valley, now lled with glacial sediment. 3 = course of the Ohio River since the early Pleistocene. Aer Palmer (1981); see also Granger et al., ( 2001) for explanation. Fig. 6: Simplied cross section through the Pennyroyal Plateau and M ammoth Cave, Kentucky (aer Palmer, 1981). Fig. 7: T ypical upper-level passage in M ammoth Cave with detrital sediment ll. is is a former tourist trail that is no longer open to the public. Sediment once lled the passage almost half-way but later subsided into an underlying passage. Note banks of remaining sediment on the le. VARIATION IN RATES OF KARST PROCESSES

PAGE 6

TIME in KARST 2007 20 dolomite = Mg and calcite = Ca Mg, in moles/L). e annual precipitation is 1.26 m/yr, and about 2/3 of it lost to evapotranspiration, so a 220 km 2 basin would have a mean runo of roughly 9 x 10 7 m 3 /yr. e loss of carbon ate rock is therefore about 6000 m 3 /yr. Roughly half of the basin consists of exposed carbonates, so the denuda tion rate on that half is about 5.5 cm/1000 years. is g ure corresponds to some of the lowest measured rates of carbonate denudation elsewhere (Ford & W illiams, 1989, p. 112). Transport of solids is neglected, as is subsur face dissolution. e Mammoth Cave System represents a maximum porosity of about than 4%, even in areas of maximum passage density (Palmer, 1995). W hen the denudation rate is extrapolated to 2 million years, it indicates an overall lowering of the solu ble Pennyroyal surface of roughly 100 m. is is impos sible, because it exceeds the total relief between the origi nal surface and the Green River. ere is no doubt that most of the surface has been lowered (Fig. 6), but there were evidently long periods of stagnation, especially at the beginning, when large parts of the surface were man tled with thick sediment. Most of the denudation is in the form of doline growth. Gams (1965) points out that corrosion accelerates in dolines as they grow, because of enhanced CO 2 production in their thickening soils. Ap parently the rate of karst denudation is higher today than during the early Pleistocene. KARST DEVELOPMENT AT TIME SCALES OF 10 7 10 8 Y EARS Karst that evolves throughout entire geologic periods or eras tends to do so in discontinuous steps in which lengthy episodes of stagnation exceed those of active karst processes. For example, certain karst areas of the Rocky Mountains and Black Hills (western U.S.A.) have undergone at least 7 dierent stages over the past 350 my but were actively forming only about 20% of that time. Jewel and W ind Caves in South Dakota are good exam ples (Fig. 8). W ith mapped lengths of 218 and 196 km, they are among the most complex caves in both pattern and diversity of geologic history. Each successive set of features was superposed on the previous ones, because each provided favorable sites for those that followed. Os borne et al. (2006) describe a similarly complex history in the Jenolan Caves of Australia. e major stages of karst development in the Black Hills are outlined below (Palmer & Palmer, 1989, 1995): 1. Early Carboniferous carbonates of the Madison Formation were deposited on a low-gradient continental shelf. Interbedded sulfates were included in the middle and upper Madison. 2. Brecciation and early voids formed by dissolution and reduction of sulfates, plus production of sulfuric acid (Fig. 9). Sulfate rocks were almost completely removed. 3. A mid-Carboniferous karst formed throughout much of western North America (Sando, 1988). Surface features included ssures and dolines up to 30 m deep. Caves concentrated at 20-50 m below the sur face along former sulfate zones and intersect ear lier breccias and caves (Fig. 10). Comparison with modern caves sug gests some freshwatersaltwater mixing dissolu tion. 4. e karst was buried by late Carbon iferous detrital sedi ment, and most caves were completely lled. e sedimentary burial continued through the Cretaceous to a depth of Fig. 8: Geologic setting of Wind Cave, South D akota. L = M adison Limestone (early Carboniferous) underlain by thin Cambrian sandstone, S = late Carboniferous sandstone, S H = mainly shale, K = Cretaceous sandstone, OS = Oligocene sediment (mainly siltstone, widely eroded). e upper surface of the M adison is irregular paleokarst. W T = water table in lowest passage of Wind Cave. e cave extends only a few meters below the water table. Arrows show dominant ow pattern of today. A RTHUR N. P ALMER

PAGE 7

TIME in KARST 2007 21 at least 2 km. Buried caves and vugs, as well as voids in the Carboniferous sediment, were lined by white scale nohedral calcite about 1-2 cm thick (Fig. 11). Pre-burial Fig. 9: Early solution voids and brecciation related to early Carboniferous sulfate-carbonate interactions in J ewel Cave, South D akota. ese are exposed by collapse of wall of a later cave. H eight of photo is about 2 m. voids can be recognized by this distinctive coating. Along faults, surfaces were coated by euhedral quartz up to a 5 mm thick. 6. e Black Hills and Rocky Mountains were uplied by the Laramide orogeny (latest Cretaceous through Eocene; Fig. 8). e climate was more humid than todays, and the present topography above the caves was formed by the end of the Eocene. Enhanced groundwater ow enlarged earlier caves to their present form (Fig. 12). eir layout shows evidence for mixing between shallow and deep water (Palmer and Palmer, 1989), although Bakalowicz et al. (1997) suggest a purely thermal origin. 7. e caves drained and were exposed to subaerial weathering, which produced thick carbonate deposits in many passages. 8. Most of the Eocene landscape was buried by Oligocene sediments during a drying of the climate. Al though much of this sediment has been removed by later Fig. 10: M id-Carboniferous paleokarst, B ighorn M ountains, Wyoming. Caves in cli were once lled with late Carboniferous sediment, but much of it has been removed by weathering and stream erosion. Fig. 11: T op: Scalenohedral calcite coating of M esozoic age on walls of Carboniferous vug, Wind Cave (crystal length ~1.5 cm). B ottom: Rhombohedral calcite coating of late T ertiary age on weathered walls of an early T ertiary passage, J ewel Cave (maximum thickness of calcite = 15 cm). VARIATION IN RATES OF KARST PROCESSES

PAGE 8

TIME in KARST 2007 22 erosion, the Eocene landscape on the resistant PaleozoicMesozoic rocks has survived almost intact, as have the underlying caves, thanks to the present semi-arid cli mate. 9. Partial blockage of springs by Oligocene sedi ments caused a second phase of calcite coating (mainly rhombohedral) averaging 15 cm thick in Jewel Cave (Fig. 11) but thinner in W ind Cave. e earlier scalenohedral coating is still visible in pockets and vugs that were iso lated from the cave development and exposed by later breakdown. In this sequence there is little information about de velopmental rates. Instead, the karst history is portrayed as a series of discrete episodes, which span a wide range of processes, groundwater conditions, tectonic relation ships, and levels of diagenetic maturity of the host strata. All eects have overlapped, and in some caves it is pos sible to stand in a single spot and distinguish every phase of their history. Fig. 12: T ypical cave passage of Eocene age in Wind Cave, showing remnants of earlier breccia (B) and paleo-ll (P).H eight of photo is about 2 m. CONCLUSIONS Karst processes operate at rates that vary considerably with time, and the magnitude of that variation is gener ally greater as the developmental time span increases. At every time scale, the developmental history of karst (at least in the examples described here) includes episodes of stagnation and of retrograde development when material is deposited instead of removed. Modern measurements of the rates of karst process es can be extrapolated into the past, but this extrapolation becomes more suspect as the time span increases. Over the entire growth history of major cave systems (usually 10 6 -10 7 years), many disruptions in rate are caused by changes in climate, base level, and river patterns. At time scales of 10 7 -10 8 years, interpretation of evolutionary rates becomes dircult, and the history of karst is usually subdivided into discrete episodes, in the same manner as tectonic and sedimentary events. As a karst feature develops toward maturity, it tends to undergo inherent changes in developmental rate. For example, a cave may decrease in enlargement rate as en trances open and enlarge, allowing greater rates of CO 2 loss. Rates of karst development may increase with time as dolines develop and enlarge, owing to greater expo sure of soluble rock and accumulation of high-CO 2 soils in depressions. It is impossible to interpret caves and karst without a solid understanding of their surrounding geology and physiography. But, despite uncertainties about their rates of development, karst features can provide more informa tion about the surrounding landscape than vice versa. A RTHUR N. P ALMER

PAGE 9

TIME in KARST 2007 23 REFERENCES Anthony, D.M. & D.E. Granger, 2004: A Late Tertiary origin for multilevel caves along the western escarp ment of the Cumberland Plateau, Tennessee and Kentucky, established by cosmogenic 26 Al and 10 Be. Journal of Cave and Karst Studies, 66, 2, 46-55. Anthony, D.M. & D.E. Granger, 2006: Five million years of Appalachian landscape evolution preserved in cave sediments. In R.S. Harmon and C.M. W icks (eds.): Perspectives on karst geomorphology, hydrol ogy, and geochemistry A tribute volume to D erek C. Ford and William B White: Geological Society of America, Special Paper 404, 39-50. Anthony, D.M. & C.G. Groves, 1997: Preliminary inves tigations of seasonal changes in the geochemical evolution of the Logdson River, Mammoth Cave, Kentucky. Proceedings of 6 th Science Conference, 15-23, Mammoth Cave, Kentucky. Bakalowicz, M.J., D.C. Ford, T.E. Miller, A.N. Palmer & M.V. Palmer, 1987: ermal genesis of dissolution caves in the Black Hills, South Dakota. Geological Society of America Bulletin, 99, 729-738. Dumont, K.A., 1995: Karst hydrology and geomorphology of the B arrack Z ourie Cave System, Schoharie County, New Y ork. M.S. thesis, Mississippi State University, p. 71, Mississippi State, Mississippi. Gams, I., 1965: Types of accelerated corrosion. In O. telcl (ed.): Problems of the speleological research. International Congress of Speleology, 133, Brno, Czech. Granger, D.E., D. Fabel & A.N. Palmer, 2001: PliocenePleistocene incision of the Green River, Kentucky, determined from radioactive decay of cosmogenic 26 Al and 10 Be in Mammoth Cave sediments. Geo logical Society of America Bulletin, 113, 7, 825-836. Granger, D.E., J.W Kirchner, & R.C. Finkel, 1997: Q ua ternary downcutting rate of the New River, Virgin ia, measured from dierential decay of cosmogenic 26 Al and 10 Be in cave-deposited alluvium. Geology, 25, 107. Groves, C. & J. Meiman, 2005: W eathering, geomorphic work, and karst landscape evolution in the Cave City groundwater basin, Mammoth Cave, Kentucky. Geomorphology, 67, 115-126. Hess, J.W ., 1974: H ydrochemical investigations of the cen tral Kentucky karst aquifer system. Ph.D. disserta tion, Pennsylvania State University, p. 218, Univer sity Park, Pennsylvania. Hess, J.W & W .B. W hite, 1993: Groundwater geochemis try of the carbonate aquifer, south-central Kentucky, U.S.A. Applied Geochemistry, 8, 189-204. Lauritzen, S.-E. & J.E. Mylroie, 2000: Results of a speleo them U/ dating reconnaissance from the Helder berg Plateau, New Y ork. Journal of Cave and Karst Studies, 62, 1, 20-26, Huntsville, Alabama. Meiman, J. & C. Groves, 1997: Magnitude/frequency analysis of cave passage development in the Cen tral Kentucky Karst. Proceedings of 6th Science Conference, 11-13, Mammoth Cave National Park, Kentucky. Mylroie, J.E. & J.R. Mylroie, 2004: Glaciated karst: How the Helderberg Plateau revised the geologic percep tion. Northeastern Geology and Environmental Sciences, 26, 1-2, 82-92, Troy, New Y ork. Osborne, R.A.L., H. Zwingmann, R.E. Pogson & D.M. Colchester, 2006: Carboniferous clay deposits from Jenolan Caves, New South W ales: Implications for timing of speleogenesis and regional geology. Aus tralian Journal of Earth Science, 53, 377-406. Palmer, A.N., 1981: A geological guide to Mammoth Cave National Park. Zephyrus Press, p. 210, Tean eck, New Jersey. Palmer, A.N., 1991: Origin and morphology of limestone caves. Geological Society of America Bulletin, 103, 1-21. Palmer, A.N.,1995: Geochemical models for the origin of macroscopic solution porosity in carbonate rocks. In Budd, D.A., P.M. Harris, & A. Saller (eds.): Un conformities in carbonate strata: eir recognition and the signicance of associated porosity. Ameri can Association of Petroleum Geologists, Memoir 63, 77. Palmer, A.N., 1996: Rates of limestone dissolution and calcite precipitation in cave streams of east-central New Y ork State. Abstracts of Northeastern Section meeting, Geological Society of America, 28, 3, 89. Palmer, A.N. & M.V. Palmer, 1989: Geologic history of the Black Hills caves, South Dakota. National Spe leological Society Bulletin, 51, 2, 72-99. Palmer, A.N. & M.V. Palmer, 1995: e Kaskaskia paleo karst of the Northern Rocky Mountains and Black Hills, northwestern U.S.A. Carbonates and Evapo rites, 10, 2, 148-160, Troy, New Y ork. Palmer, M.V. & A.N. Palmer, 1975: Landform develop ment in the Mitchell Plain of southern Indiana: Origin of a partially karsted plain. Zeitschri fr Geomorphologie, 19, 1-39. VARIATION IN RATES OF KARST PROCESSES

PAGE 10

TIME in KARST 2007 24 Plummer, L.N., T.M.L. W igley, T.M.L. & D.L. Parkhurst, 1978: e kinetics of calcite dissolution in CO 2 -wa ter systems at 5 to 60 C and 0.0 to 1.0 atm CO 2 American Journal of Science, 278, 179-216. Q uinlan, J.F., R.O. Ewers, J.A. Ray, R.L. Powell & N.C. Krothe, 1983: Ground-water hydrology and geo morphology of the Mammoth Cave Region, Ken tucky, and of the Mitchell Plain, Indiana. Indiana Geological Survey, Field trips in Midwestern geol ogy, 2, 1-85, Bloomington, Indiana. Ray, J.A., 1996: Fluvial features of the karst-plain erosion surface in the Mammoth Cave region. Proceed ings of 5th Science Conference, 137-156, Mammoth Cave, Kentucky. Sando, W .J., 1988: Madison Limestone (Mississippian) paleokarst: A geologic synthesis. In N.P. James and P.W Choquette (eds.): Paleokarst: Springer-Verlag, 256-277, New Y ork. A RTHUR N. P ALMER



PAGE 1

T IME SCALES IN THE EVOLUTION OF SOLUTION POROSIT Y IN POROUS COASTAL CARBONATE A QUIFERS B Y MI X ING CORRO SION IN THE SALT W ATER FRESH W ATER TRANSITION ZONE f ASOVNO MERILO RAZVOJA POROZNOSTI ZARADI KOROZIJE MEANICE V MEJNEM OBMOfJU SLADKOVODNIH LEf V MEDZRNSKO POROZNEM KARBONATNEM OBALNEM VODONOSNIKU W olfgang D RE Y BRODT 1 & Douchko R OMANOV 2 Izvleek UDK 556.3:552.54:539.217 Wolfgang Dreybrodt and Douchko Romanov: asovno merilo razvoja poroznosti zaradi korozije meanice v mejnem obmoju sladkovodnih le v medzrnsko poroznem karbonatnem obalnem vodonosniku Dosedanji modeli raztapljanja kalcijevega karbonata v obmoju meanja sladke in slane vode temeljijo na zdruitvi geokeminih ravnotenih in reakcijsko transportnih modelov. Dobljeni sistem nelinearnih enab zahteva veliko raunske moi. fe je hitrost raztapljanja dovolj visoka in lahko predpostavimo, da je raztopina ves as v ravnoteju glede na kalcit, reimo problem z poenostavljenim modelskim pristopom. Zaetno spreminjanje poroznosti v kamninski matriki doloa advekcijsko tranportna enabo, ki opisuje slanost v sladkovodni lei in pre hodnem obmoju pod njo. Pri reevanju porabimo dostopne programske kode. Tokove nastale zaradi razlik v gostoti modeli ramo s programom SEA W AT, topnost kalcita v meanici sladke in slane vode v odvisnosti od slanosti pa izraunamo s pro gramom PHREEQ C-2. Zaetno spreminjanje poroznosti lahko nato izraunamo z enostavnim analitinim izrazom gradienta prostorske razporeditve slanosti s(x,y), razporeditve gostot toka q(x,y) in drugega odvoda ravnotene koncentracije kalcija po slanosti. Tak modelski pristop uporabimo pri raunanju razvoja po roznosti v homogenih in heterogenih karbonatnih otokih in obalnih vodonosnikih. Podrobno so prikazani vzroki in geometrijski vzorci spreminjanja poroznosti. Rezultati kaejo, da je zaetna hitrost spremembe poroznosti reda velikosti 10 -6 na leto. To postavi asovno merilo razvoja jam v obmoje nekaj deset tiso do sto tiso let. Kljune besede: Raztapljanje kalcita, korozija meanice, o b mo je meanja sladke in slane vode, obalni vodonosnik, razvoj po roznosti. 1 Universitaet Bremen, FB1, Karst Processes Research Group, Bremen, Germany, e-mail: dreybrodt@ifp.uni-bremen.de 2 Freie Universitaet Berlin, Fachbereich Geowissenschaen, Berlin, Germany, e-mail: dromanov@zedat.fu-berlin.de Received/Prejeto: 21.12.2006 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 25 Abstract UDC 556.3:552.54:539.217 Wolfgang Dreybrodt and Douchko Romanov: Time scales in the evolution of solution porosity in porous coastal carbonate aquifers by mixing corrosion in the saltwater-freshwater tran sition zone. Dissolution of calcium carbonate in the saltwater-freshwater mixing zone of coastal carbonate aquifers up to now has been treated by coupling geochemical equilibrium codes to a reactivetransport model. e result is a complex nonlinear coupled set of dierential transport-advection equations, which need high computational eorts. However, if dissolution rates of calcite are surciently fast, such that one can assume the solution to be in equilibrium with respect to calcite a highly simplied modelling approach can be used. To calculate initial changes of porosity in the rock matrix one only needs to solve the advection-transport equation for salinity s in the freshwater lens and its transition zone below the island. Current codes on density driven ow such as SEA W AT can be used. To obtain the dissolution capacity of the mixed saltwater-freshwater solutions the calcium equilib rium concentration c eq (s) is obtained as a function of salinity by PHREE Q C-2. Initial porosity changes can then be calculated by a simple analytical expression of the gradient of the spatial dis tribution s(x, y) of salinity, the distribution of ow uxes q(x,y) and the second derivative of the calcium equilibrium concentra tion c eq (s) with respect to salinity s. is modelling approach is employed to porosity evolution in homogeneous and heterogeneous carbonate islands and coastal aquifers. e geometrical patterns of porosity changes and the reasons of their origin will be discussed in detail. e results reveal initial changes of porosity in the order of several 10 -6 per year. is places the time scale of cavern evolution to orders from several tens of thousands to a hundred thousand years. Keywords: Calcite dissolution, mixing corrosion, saltwaterfreshwater, mixing zone, coastal aquifer, evolution of porosity.

PAGE 2

TIME in KARST 2007 26 I NTRODUCTION Carbonate islands consisting of porous rocks show typical karst features characterized by large dissolution chambers close to the coast, which have been created by mixing corrosion in the fresh-saltwater transition zone (Mylroie and Carew, 2000). Figure1 represents the basic concept. Due to meteoric precipitation a freshwater lens builds up, oating on the denser saltwater (Vacher, 1988) e transition from freshwater to seawater is not sharp. De pending on many factors, such as tidal pumping, periodic ity of annual recharge, and the heterogeneity of the rocks properties in the aquifer it exhibits a transition zone. is zone can range from a few meters to half the depth of the lens. In this zone mixing between saltwater and freshwater activates mixing corrosion, which creates large chambers. ese are called ank-margin caves. Figure 2 shows such a cave with its typical solutional features on its ceiling. W hen seawater mixes with a solution of H 2 O-CO 2 CaCO 3 saturated with respect to CaCO 3 the mixture is no longer in equilibrium with respect to calcite. Depending on the chlorine concentration s, termed as chlorinity further on, of the mixture undersaturation or supersaturation may result. Figure 3 gives an example. It depicts the dierence of the calcium concentration c mix (s) of the mixture and that of its corresponding equilibrium concentration c eq (s) as a function of s. is is the amount of calcium, which can be dissolved or precipitated, when the mixed solution is in contact with carbonate rock. e H 2 O-CaCO 3 -CO 2 solution used to calculate this data is in equilibrium with a partial pressure of CO 2 of 0.01 atm at a temperature of 20C. e seawater also is at 20C. e data in Fig. 3 were obtained by use of the code PHREE Q C-2 (Parkhurst and Apello, 1999 ). From Figure 3 it is evident that mixtures with low content of seawater, chlorinity s 0.3 mol/ can dissolve calcite, whereas mixtures with higher chlorinity may precipitate calcite. Renewed aggressivity due to mixing therefore occurs only at the freshwater side of the mixing zone where chlorinity is low. If one assumes that dissolu tion of calcite proceeds surciently fast the solution there will be saturated with respect to calcite. Dissolution of minerals under such conditions is termed a gradient reaction (Phillips, 1991). Here we use this as a novel instrument to explain the evolution of po rosity in carbonate islands. Dissolution rates of limestone are surciently fast, such that aer mixing between salt water and freshwater we assume saturation with respect to calcite in the entire lens. Aer attaining equilibrium the local distribution of calcium concentration c eq (s(x,z)) becomes stationary and exhibits gradients. Necessarily advection and diusion must transport the dissolved limestone to the outow of the aquifer. Fig. 1: Conceptual representation of a carbonate island from M ylroie and Carew (2000). Fig. 2: Flank-margin cave. Fig. 3: as a function of chlorine concentr a tion. e curve extends from pure freshwater (right) to pure seawater (le). WOLFGANG D RE Y BRODT & D OUCHKO R OMANOV

PAGE 3

TIME in KARST 2007 27 e advection term: In Figure 4 we consider a volume element dxdydz at posi tion (x,y,z), into which ow, with components q x and q z enters perpendicular to dydz or dxdy. e ux q is dened by the volume of uid per time unit entering through a unit of surface area and is given in [cm 3 /(cm 2 s) = cms -1 ]. e component q x transports solution from the neighbouring elementary cell at position (x-dx, y,z) via the area dydz into the cell dxdydz. is solution has al ready attained equilibrium c eq (s(x-dx,y,z)) at position x-dx. W hen it enters into the volume element dxdydz it must dissolve or precipitate limestone to adjust its calcium concentration to equilibrium c eq (s(x,y,z)) at position x. On the other hand solution from the element dxdydz ows out into the neighbouring cell with ux q x (x,y,z). Mass conservation requires that the amount of limestone dis solved per time unit in the element dxdydz must be equal to the dierence of mass transported into the cell and that transported out of it. From this one nds (1) An analogue equation exists for Q z the amount of limestone dissolved by the ux component q z entering via the surfaces (dx,dy). (2) erefore (3) Because the ux q follows the Darcy law of incom pressible uids, =0. (4) whereby we have replaced is the calcium concentration resulting from the mixing of sea water and freshwater and is a linear function of Cl-con centration s. D ISSOLUTION IN THE MI X ING ZONE (5) is the increase of equilibrium concentration as given in Figure 3, s sea is chlorinity of seawater. e diusion term: Our mass balance so far, however, is incomplete because gradients of c eq cause transport by diusion. e rate Q D of mass transport by diu sion is given by (6) where D = qd/n + D m is the coercient of dispersion. n is the porosity of the rock and d its grain size. (Phillips, 1991). Dm is the constant of molecular dif fusion (10 -5 cm s -1 ). e total rate: e total dissolution rate Q tot is then given by Q D +Q adv (7) Due to the linearity of c mix with salinity s (eqn. 5) one nds grad(c mix ) proportional to grad(s). e distribution of salinity is governed by the ad vection-diusion equation (8) because the distribution s is stationary. From the linearity of s with c mix we have (9) e total dissolution rate Q tot is given by the master equation (10) Since c eq (s(x,z)) is a function of local distribution s(x,z) by dierentiating and using the chain rule, one nds using equation 8 (11) is master equation relates the amount of dissolved material per unit volume of the rock matrix [mol cm -3 s -1 ] Fig. 4: M ass balance for the advection term. T IME SCALES IN THE EVOLUTION OF SOLUTION POROSIT Y IN POROUS COASTAL CARBONATE A QUIFERS ...

PAGE 4

TIME in KARST 2007 28 to the gradient of salinity s, to the second derivative of and the ux q. can be obtained by dierentiating twice the data set of Figure 3. is data set was obtained by using the program PHRE EQ C2 and calculating about 50 closely spaced points to avoid numerical errors, when dierentiating twice. e result is shown in Figure 5. e function and the ow distribution can be obtained by the numerical hydrologic model SEA W AT, as will be shown in the next sections. To calculate the initial change of porosity it is surcient to obtain the ux and salinity distribution of an island without consid ering calcite dissolution, because the time to establish a stationary state of the lens is in the order of 100 years. It is a good approximation to assume that during this time the change of porosity is insignicant. Equation. 11 can be written in terms of the change of porosity as (1/s)(12) (12) M = 100 g/mol is the molecular weight at CaCO 3 =2.7 g/cm 3 is the density of compact CaCO 3 Q tot the mass of CaCO 3 dissolved per time from a unit volume of the rock matrix is given in mol s -1 cm -3 is the amount of volume dissolved per time from a unit volume of the rock matrix (cm 3 s -1 /cm 3 ). By use of equation 12 it is now possible to construct a conceptual frame for the evolution of porosity. Tests of this approach on simple benchmark models have shown its reliability and have found agreement to experimental data (Romanov and Dreybrodt, 2006). Fig. 5: Second derivative I NITIAL CHANGES OF POROSIT Y IN A HOMOGENEOUS ISLAND To obtain the initial distributions of ux q and chlorinity s in the lens of a carbonate island we have used SEA W AT by USGS (Guo and Langevin, 2002). e modeling do main is shown in Figure 6. e island is a strip of 1 km width. Porosity and the hydraulic conductivity K are uniform ( ).etransversaldisper. e transversal disper sivity is a T = d = 0.01 cm, the longitudinal dispers i vity is a l = 0.1 cm. Inltration is 3 10 -3 m/day =1.11m/year. is way the maximal depth of the lens is about 50 m below sea level. e lower border of the domain reaches down to 70 m. At that boundary an impermeable layer imposes no-ow conditions. e grid size in the domain is 1 m x 1 m in the part below sea level. In the part above sea level (2 m) the grid size is 0.2 m by 1 m. In its initial state when the island emerges out from the sea the entire aquifer is lled with seawater. W hen the island receives recharge from meteoric freshwater the lens builds up. A stable sta tionary lens is obtained aer about 30 years. Fig.7 shows the results of the model run. Figure 7a shows the freshwater lens (white), the tran sition zone and its distribution of Cl-concentration by a color code. From this distribution of chlorinity one can extract the scalar value and and Figure Figure 7b shows the chlorinity in units normalized to its maxi mum value along several horizontal sections as depicted in Figure 7a. e lowest section at -68 m is entirely in saltwater with maximum Cl-concentration. e section at -55 m extends through the almost horizontal base of Fig. 6: M odeling domain of a carbonate island. WOLFGANG D RE Y BRODT & D OUCHKO R OMANOV

PAGE 5

TIME in KARST 2007 29 the lens and shows a wide zone where the concentration raises to that of seawater. e upper sections cut through the mixing zone and there the rise in concentration from freshwater to seawater becomes steeper. e square of the gradient is shown by Figures 8a,b also normalized to its maximum value in Figure 8b. Figure 8a illustrates its local distribution, which exhibits large values only in the region of the transition zone. e horizontal distribution along horizontal sections is de picted in Figure 8b. e second derivative obtained from the Cl-concentration in Figure 7a is given in Figure 9a. Its distribution is limited to that part of the transition zone with 0 < s < 0.03 mole/ See Figure 5. is corresponds to a narrow fringe at the freshwater side of the transition zone with seawater content from zero up to about 4%. In any case creation of porosity is possible only in this restricted region. Figure 9b for completeness depicts some distribu tions of along horizontal sections. To calculate the initial rate of change in porosity (conf. eqn 12) the Darcy uxes q must be known. ey are also obtained from the model run and shown in Fig 10. e ux is low in the center of the island m/year), but increases by orders of magnitude when the uid moves coastward, where it becomes about 0.2 m/day at the outow. e dispersion coercient D = qd/ + D m (conf. eqn. 11) depends on the ux q, but also on the coer cient of molecular diusion D m =10 -5 cm 2 /s. For low ux q<10 -4 cms -1 and particle diameters d -2 cm dispersion is dominated by molecular diusion. In the following scenarios we have used d=10 -2 cm, a realistic value in po rous limestone. erefore in the range of ux, which can be read from Figure 10b the dispersion coercient in the center of the island is D=10 -5 cm 2 s -1 It increases by about 60% of this value at the coast. From the data given in Figures 7a, 8a, and 9a the initial porosity is obtained by use of eqn. 12. Figure 11 illustrates these results. Changes in porosity are restricted to a small fringe in the transition zone and are fairly even along it. ey are in the order of 10 -6 year -1 is is surcient to create substantial porosity within 100,000 years. At the outow ank margin caves can develop in 10,000 years. One has to keep in mind, however, that the approximation as a ho mogeneous island is a high idealization. Any disturbances, which increase the width of the transition zone, will reduce the gradients of chlorinity and therefore on more realistic settings the initial porosity changes accordingly. Fig. 7: H omogeneous island. a) Local distribution of chlorinity e white region designates the freshwater lens. b) chlorinity along horizontal sections as indicated in a). Fig. 8: H omogeneous island. a) Local distribution of the square of gradients b) square of gradients along horizontal sections as indicated in a). T IME SCALES IN THE EVOLUTION OF SOLUTION POROSIT Y IN POROUS COASTAL CARBONATE A QUIFERS ...

PAGE 6

TIME in KARST 2007 30 Fig. 9: H omogeneous island. Local distribution of the second derivative b) second derivative along horizontal sections as indica ted in a). As we have stated already, the second derivative is restricted to narrow regions in the freshwater side of the transition zone. It exhibits signicant values only at lo cations where the water contains between zero and 4% saltwater (see Figure 5). On the other hand the gradient in salinity is maximal at mixtures of about 50% seawater, because it arises from a diusive process. In the region of maximal gradients, however, the second derivative is small. Vice versa in the region of high values of the sec ond derivative, the gradients of salinity are low. is is il lustrated in Figure 12. is gure is an overlay of the hor izontal distributions (grads) 2 in Figure 8b (red curves), the second derivative in Figure 9b (green curves), and the initial porosity change in Figure 11b (black curves). All curves are normalized to their individual maximum values. erefore their values are not comparable in this gure. W hat can be compared, are the locations. Evi dently the curves for gradients and second derivative are well separated. e curves of porosity change are propor tional to the product of the square of the gradient and the second derivative. Porosity change displays high values in between their maxima but close to the region of high values of the second derivative. Figure 13 further illustrates this qualitatively. e red region depicts the locations of the modeling domain where (grads) 2 exhibits values is the maximal value. e green region shows these locations for the second derivative and nally the black region shows the locations of signicant changes of porosity. ese ndings agree with those of Sanford and Konikow (1989) who also found that changes in porosity are re stricted to regions where waters contain between 0.5% and 3% of seawater. It should be noted here that any mechanism, which changes the sigmoid shape of the salinity distribution to a linear prole would enhance evolution of porosity dra matically. In this case salinity gradients become constant in the entire mixing zone and their value is at least one order of magnitude higher at the maximal value of the second derivative. One could speculate that tidal pump ing and uctuations of the water table due to seasonal changes of inltration could cause such linear mixing zones. Present observations in boreholes give some evi dence for such transition zones. Fig. 10: H omogeneous island. a) Local distribution of ux b) ux along horizontal sections as indicated in a). WOLFGANG D RE Y BRODT & D OUCHKO R OMANOV

PAGE 7

TIME in KARST 2007 31 Fig. 11: H omogeneous island. a) Local distribution of initial change of porosity b) along horizontal sections as indicated in a). Fig. 12: H omogeneous island. (grads) 2 (red), (green), and) (black) along horizontal sections of the island. Numbers on the sets of curves give the depth of the section. Fig. 13: H omogeneous island. Regions of (grads) 2 (red), of (green), and change of porosity (black). I NITIAL CHANGES OF POROSIT Y IN A HETEROGENEOUS ISLAND A more realistic approach to nature can be taken by em ploying a geo-statistical distribution of hydraulic con ductivities. Figure 14 shows such a distribution gener ated with the soware of Chiang and Kinzelbach (1998). It covers conductivities of two orders of magnitude from about 380 m/day (red) down to 2 m/day (dark blue). Most of the aquifer is occupied by values between 10-200 m/day. Otherwise all previous boundary conditions are unchanged. e ow eld is illustrated in Figure 15. Flux is unevenly distributed, because the heterogeneous dis tribution of conductivities distorts the pathways of uid elements in comparison to the regular ones in a homoge neous island. Consequently the freshwater lens in Figure 16 shows a wide transition zone (compare to Figure 7a). Fig. 14: H eterogeneous island. Statistical distribution of hydraulic conductivity in the modeling domain. T IME SCALES IN THE EVOLUTION OF SOLUTION POROSIT Y IN POROUS COASTAL CARBONATE A QUIFERS ...

PAGE 8

TIME in KARST 2007 32 e square of the gradient is limited to the seawa ter side of the transition zone, as can be visualized from Figure 17. e region of 0-4% mixtures extends far into Fig. 15: H eterogeneous island. Local distribution of ux q. Fig. 16: H eterogeneous island. Local distribution of chlorinity e white region designates the freshwater lens. Fig. 17: H eterogeneous island. Local distribution of the freshwater lens. is can be also visualized from the second derivatives as shown in Figure 18. Figure 19 illustrates the initial change of porosity, which exhibits high values of 3 10 -6 year -1 (red) at only a few locations close to the freshwater side of the transition Fig. 18: H eterogeneous island. Local distribution of derivatives Fig. 19: H eterogeneous island. Local distribution of initial porosity change Fig. 20: H eterogeneous island. Regions of high values of (grads) 2 (red), (green), and change of porosity (black) in the modeling domain. WOLFGANG D RE Y BRODT & D OUCHKO R OMANOV

PAGE 9

TIME in KARST 2007 33 zone. At some favorable locations (red and yellow) caves may evolve there in several 10,000 to 100,000 years. is is further illustrated by Figure 20, which shows the regions of high values for (grads) 2 (red), d 2 c eq /ds 2 (green), and (black) in the modeling domain. Fig ure 21 depicts (grads) 2 (red), d 2 c eq /ds 2 (green), and (black) along selected horizontal sections. In both gures we nd that the regions of (grads) 2 (red), d 2 c eq /ds 2 (green) are well separated and porosity develops in between. Due to the heterogeneity, however, the patterns become complex. I NITIAL CHANGES OF POROSIT Y IN SALT W ATER TONGUES W hen impermeable strata underlay an island surciently close to its surface the freshwater lens cannot extend be low this layer and a saltwater tongue intrudes from the coastland inward until it reaches the impermeable layer. From thereon the freshwater lens is truncated by this layer. In this situation mixing of waters is restricted to the transition zone of the tongue and one expects high dissolution rates in this region. Fig. 22 shows the local distribution of chlorinity and the initial change of porosity using the statistical distribution in Figure 14 for the upper permeable part. e mixing zone exhibits a structure similar to that of the heterogeneous island at the corresponding loca tions. Porosity changes at the outow are low, but we nd values up to 10 -6 1/year land inward at various lo cations and also at the contact of the tongue with the impermeable rock. Fig. 21: a) H eterogeneous island. D istributions of (grads) 2 (red), (green), and porosity change (black) along selected horizontal sections. Number on the sets of curves give the depth of the section. Fig. 22: Coastal aquifer with heterogeneous conductivity down to 29 m as used in Fig. 14. e strata below 29 m are impermeable (grey). Local distribution of Cl-concentration s(x) and initial porosity change T IME SCALES IN THE EVOLUTION OF SOLUTION POROSIT Y IN POROUS COASTAL CARBONATE A QUIFERS ...

PAGE 10

TIME in KARST 2007 34 C ONCLUSION W e have taken substantial steps in modeling the initial porosity changes in fresh water saltwater mixing zones under various geological settings. W e have demonstrated that the knowledge of the distribution of salinity and the amount of the ow velocity is surcient to calculate the initial distribution of porosity changes. ese are the rst modeling results, which do not need scaling parameters for calibration as the models of Sanford and Konikow (1989), or more complex non-linearly coupled sets of dierential transportadvection equations (Saaltink et al. 2004). W e now understand that in a stationary island high porosity can develop only if high salinity gradients exist in the re gion of low saltwater content to up to 3%. For the highly idealized scenarios of islands with homogenous hydraulic conductivity we nd clear rules about the geometric dis tribution of porosity. ese, however, are destroyed for islands with a statistical distribution of hydraulic conduc tivity. Each realization of a statistical distribution then will give dierent results, and a general standard scenario can not be used as a tool. In heterogeneous settings porosity could occur at any place below the island and at favorable settings also ank-margin caves will arise. is explains why presence of these caves is not the rule, as it should be in homogeneous settings, where porosity develops only at the base and the outow of the lens. In view of the re stricted knowledge about the hydraulic properties and the initial porosity of the rock, which one necessarily has in carbonate platforms, detailed applied modeling at present, and most likely in the next decades will not be available. On the other hand our ndings give a rm basis for understanding the evolution of porosity on time scales of several ten to several hundred thousand years. Using an iterative procedure to implement changes of porosity and hydraulic conductivity in each time step will reveal the ba sic properties of processes involved in creating macro-po rosity such as caves and conduits. Feed back mechanisms, which enhance dissolution in the regions of increased po rosity and hydraulic conductivity could accelerate the evo lution of porosity. erefore time scales derived from the initial change in porosity represent upper limits only. Further work into this direction is needed. A CKNO W LEDGEMENT W e thank TOTAL S.A. for supporting this work. WOLFGANG D RE Y BRODT & D OUCHKO R OMANOV R EFERENCES Chiang, W H., Kinzelbach, W ., 1998: Processing Mod ow. A simulation system for modeling groundwa ter ow and pollution Guo, W ., Langevin, C. D., 2002: Users guide to SEA W AT: a computer program for simulation of three-dimen sional variable-density ground-water ow. U. S. Geological Survey Techniques of W ater-resources Investigations Book 6, USA. Mylroye, J. E., and Carew, J., L., 2000: Speleogenesis in coastal and oceanic settings. In: Klimchouk, A., Ford, D. C., Palmer, A. and Dreybrodt, W (Editors), Speleogenesis: Evolution of karst aquifers. National Speleological Society, Huntsville, 226-233. Parkhurst, D. L., Apello, C. A. J., 1999 (Version 2): Users Guide to PHREEQ C a Computer Program for Speciation, Reaction-path, 1D-transport, and In verse Geochemical Calculations, Technical Report 99-4259. U. S. Geological Survey, USA. Phillips O. M., 1991: Flow and reactions in the perme able rocks. Cambridge University Press. Cambridge, New Y ork, Port Chester, Melbourne, Sydney. 1991. Romanov D. and Dreybrodt W ., 2006: Evolution of po rosity in the saltwater-freshwater mixing zone of coastal carbonate aquifers: An alternative modeling approach. Journal of Hydrology 329, 661-673 Saaltink, M.W ., Batlle, F., Ayora, C., Carrera, J., Oliv ella, S., 2004: RETRASO, a code for modeling reac tive transport in saturated and unsaturated porous media. Geologica Acta 2 (3), 235 Sanford W E., Konikow L. F., 1989: Simulation of Cal cite Dissolution and Porosity Changes in Saltwater Mixing Zones in Coastal Aquifers. W ater Resources Research, v. 25, No. 4, p. 655-667. 1989. Vacher H. L., 1988: Dupuit-Ghyben-Herzberg analysis of strip-island lenses. Geological Society of America Bulletin, v. 100 p 580-591. 1988.



PAGE 1

T HE AGE OF K ARST RELIEF IN WEST S LOVENIA STAROST KRAKEGA RELIEFA V Z AHODNI S LOVENIJI Andrej M IHEVC 1 Izvleek UDK 551.435.8 (497.4 Kras) Andrej Mihevc: Starost krakega reliefa v zahodni Sloveniji Starost krasa lahko doloimo s trenutkom, ko so bile krake kamnine dvignjene iz morja. Drugi nain opredelitve starosti krasa je z datiranjem reliefnih oblik ali skupin reliefnih oblik. Planoto Kras sestavlja vrsta zelo razlinih reliefnih oblik, ki so nastale v razlinem asu, vendar so se zaradi posebnosti razvoja krasa ohranile in sobivajo v sedanjem reliefu. Na planoti, ki se poasi dviguje se hidroloke cone in krako povrje pomikajo navzdol. Vodotoki s strani so prenehali dotekati na kras in nek danje v viini talne vode nastalo uravnano povrje so razlenile tevilne vrtae. Na robu krasa so vrezane slepe doline, nekat ere od njih kaejo sledove tudi recentnih tektonskih premikov. Znievanje reliefa zaradi korozije je razgalilo jame, ki so se ob likovale globoko pod povrjem in ustvarilo brezstrope jame, ki so postale del dananje topograje povrja. Z morfoloko primerjavo brezstropih jam, slepih dolin in uravnav in datiran jem sedimentov ter upotevanjem starosti tektonskih faz lahko rekonstruiramo razvoj reliefa in ocenimo starost krake pokra jine. Kljune besede: kras, morfologija, starost, Kras, Slovenija. 1 Karst Research Institute, ZRC SAZU, Titov trg 2, Sl 6230 Postojna, Fax: +386 5 7001999, Andrej.Mihevc@guest.arnes.si Received/Prejeto: 01.02.2007 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 35 Abstract UDC 551.435.8 (497.4 Kras) Andrej Mihevc: e age of Karst relief in west Slovenia e age of a karst can be dened as the time when the karst rocks were uplied out of the sea. e other view of the age of karst is to dene the age of certain karst features or assem blages of karst features. On the Kras plateau there is a variety of forms that were formed at quite dierent times, but due to karst evolution, they coexist in todays relief. On the plateau, that is slowly rising, the hydrological zones in karst surface are moving downwards. Streams from the side ceased to ow on the karst and former leveled surface that was formed in condi tions of high ground water is dissected by numerous dolines. Blind valleys are incised at the side and some of them show the inuence of recent tectonics. e lowering of relief by corro sion exposes caves that have formed deep beneath the surface and creates unroofed caves that become a part of the surface topography. From the morphological comparison of the un roofed caves, blind valleys and levelled surfaces and by dating of the sediment and considering the age of tectonic phases we can reconstruct the evolution and estimate the age of the karst landscape. Key words: karst, morphology, age, Kras, Slovenia. I NTRODUCTION e question about time, like velocity of processes or age of karst surfaces and caves is a very important issue in karst studies. e age and evolution of karst is also im portant when we study karst as a specic ecosystem. It can tell us when karst and especially the caves start to form in a given area and how the landscape is changing. The first explanation of geomorphic evolution and the age of the karst in W Slovenia were made by geologists. To estimate the age they used geologic data the age of last marine sedimentation and the tectonic evolution of Dinaric mountains and the Alps (Grund 1914).

PAGE 2

TIME in KARST 2007 36 At rst karstologists were focused on understand ing karst processes and the evolution of karst features like dolines, poljes and corrosion plains. ey were much inuenced by the ideas of a geographic cycle promoted by Cvijit (1924). Karst evolution was divided into similar steps in the cycle but they also added a pre-karst phase of relief evolution with which they explained some mor phological elements in karst. e cyclic explanation of the karst evolution was lat er modied with climatic geomorphology (Rogli 1957, Radinja 1972). It emphasised the importance of climate on the morphological processes. is meant that some forms of relief, like conical hills and levelled surfaces were explained as a relicts from tropical climate. Because such a climate was present at the end of the Tertiary, these forms were determining the age of that relief features. Another important climatic signal in the morphol ogy of the Kras they estimate were the cold Pleistocene climates with periglacial processes in lower positions. Scree slopes, collapses in caves, uvial deposits in con tact karst areas and some ner sediment were explained as extremes of climate control and not normal karst phe nomena. ey were also used for determination of the age of features (Melik 1955, Gospodari 1985). Fig. 1: e location of the Kras plateau and the study areas. Geomorphologists have abandoned the cyclic mod el of relief and are now paying more attention to struc tural elements in karst morphology like recent tectonic (Habi, 1982), eld measurements and observations on karst denudation (Gams 1963), comparative studies of dierent karst features or types of karst, like contact karst (Gams 1962, Mihevc 1994), the study of dolines and col lapsed dolines (Mihevc 2001) and new geomorphologic features like unroofed caves (Mihevc 1996, 2001, Slabe 1997) as an important remnants of former landscapes and a source of sediments. Flowstones in the caves were dated (Hajna 1991, Mihevc 2001) and paleomagnetic methods were used in cave and karst sediments (Bosak & al. 1999, 2004). Very important data were provided by latest research on the plate tectonics. e tectonic evolution of the area is characterized since late Tertiary rst by northward motion of Adria micro plate which caused contraction deformations. e contraction was exhausted at about 6 Ma ago and was followed by rotation accompanied with uplis, folding and strike-slip basins formation. ese events take place in two distinct phases (Vrabec & al. 2006, Fodor & al. 1998). A NDREJ M IHEVC

PAGE 3

TIME in KARST 2007 37 G EOMORPHIC EVIDENCES ON THE AGE OF K RAS Kras is a low NW SE trending longitudinal plateau along Trieste Bay (Adriatic Sea) between ysch Brkini hills on SE, Vipava Valley in NE, and the Soa River low lands in NW e plateau is about 45 km long and 14 km wide. e surface of the plateau is slightly tilted from 500 m a.s.l on SE towards NE where it ends at about 100 m above the Soa river. e central part of Kras is built from highly per meable Cretaceous carbonate platform shallow marine limestone and less permeable dolomite. Eocene ysch that acts as an important impermeable barrier surrounds the carbonate massif. e age of the karst of Kras plateau can be dened as the time when the karst rocks were uplied out of the sea. For the most of Dinaric karst in Slovenia this occurred aer the Eocene, since aer that there is there is no evidence of younger marine sediments. As soon as the carbonate rocks were exposed, we can expect that the karst was formed, but there are no remnants of karst features from that time. Most likely denudation has destroyed them. e other view on the age of karst is to dene the age of those karst features for which we know how and when they were formed and which evolution was stopped long time ago. Such features are levelled surfaces, which evolve at the level of the karst water and blind valleys that were formed by alogenic rivers. W e can compare them with evolution of uvial relief and unroofed caves, which are caves exposed to surface by denudation. On the Kras plateau there is a variety of forms that were formed at quite dierent conditions and time but due to peculiarities of karst evolution they coexist in to days relief. is can make the determination of the one age of a karst landscape dircult or impossible, but it tells us about the genesis of the landscape trough dierent phases. Here we present the study of the part of the Kras, Divaki kras and Matarsko podolje and the edge of Pod gorski kras from which there are some evidences about the evolution and age of Kras. T HE UNROOFED CAVES OF D IVAKI KRAS e Divaki kras is tilted slightly towards N W at eleva tions between 450 and 400 m a.s.l, on the SE part of the Kras plateau. It is built up mostly by Cretaceous and Paleogene limestone. e karst features here are exceptional; there are the sinking of the Reka river into kocjanske jame cave via large collapse dolines with and hundreds of dolines. e largest caves of the area are the 12,500 m long and 275 m deep Kana jama and the 5800 m long and 250 m deep kocjanske jame. e caves were formed by the Reka river which can be reached at a depth of 195 m in kocjanske jame and 156 m a.s.l. in Kana jama. e main morphologic features of the area are col lapsed dolines and dolines which together cover about 12% of the area. e collapsed dolines are connected with active water caves. e solution dolines cover less than 4% of the area. e rest of the surface (88%) is level. ese points out the prevailing surface leveling process in the present conditions In this levelled surface there are several large un roofed caves (Mihevc 1996). As such caves appear on the surface due to denudation, and we may call their remains denuded caves. A cave ceiling will be the rst to be removed by denudation, which is why they are also called unroofed caves. ey were rst found and described in the Divaa Karst. e unroofed caves share on the surface is small, only about 0.16% of the entire surface. ree important unroofed caves have been found. e rst is a 350 m long unroofed cave near Povir village at 400 m above the sea level. ere is a remnant of a cave passage that was 6 m wide and over 5 m high. e en tire volume of the passage has been lled by allochtonous uvial sediments of clay, silicate sands and gravel with pebbles up to 25 cm in diameter. e second is an unroofed cave near Divaa on the slopes of doline Radvanj at the altitude of 390 415 m above sea level. It is exposed on the slope that dissects large cave passage, which is entirely lled with sedi ments. Similar sediments can be seen in the Divaka jama cave. is is a 600 m long cave, whose continua tion towards 250 m distant unroofed cave is completely lled. e cave was also lled, but the sediment was later washed from it by the seepage water. Here we can see that a part of the unroofed cave that still exists as an underground cave. e longest rooess cave is 1.800 m long remnant of caves whose passages were about 20 m large, and therein ew a great underground river. e cave was lled with uvial sediments and massive owstone. It is located T HE AGE OF K ARST RELIEF IN W S LOVENIA

PAGE 4

TIME in KARST 2007 38 partly above the kocjanske jame, where the actual river bed in the cave is 230 m below the unroofed cave. On the basis of the shape of walls and sediments we may reconstruct some evolution of the caves and later the surface. e caves are remnants of larger cave sys tems, which conducted wa ters from dierent sinking streams. Growth of speleo thems in them was frequent ly interrupted by phases of erosion or backll. e caves were aerwards lled up with uvial sediments. e large pebbles in the caves testify the great gradient of the sur face streams. Later all caves were lled with ner sedi ment, which could mean the lowering of the gradient in karst and aplanation. Later, the surface was tilted and up lied which caused lowering of the karst water level. e age of the unroofed caves can be established by comparative methods ac cording the denudation rate of the surface. If we pre sume, that it is about 50 m/ Ma (Gams 1962) and there was some 100 m 200 m of rock removed from above the caves that they are at least 2 4 Ma old, and probably older (Mihevc 1996, 2001). Similar time frames 1.6 1.8 Ma or/and 3.8 to 5 Ma were given also by paleomag netic datation of clastic sedi ments (Bosak & al. 1998) and by the timing with tectonic phase that started at 6 Ma (Vrabec & al. 2006). e age of the rooess cave can also be illustrated by the time, in which the water table in Kras lowered for 240 m, from about 400 m to 160 m a.s.l. Fig. 2: e map of the D ivaa karst. On the levelled surface the large collapse dolines are dominating features, solution dolines are frequent, but they represent only small proportion of the surface. e outlines of the main caves and the main unroofed caves are marked. On the map made of D EM with 12.5 m grid the road cuts or causeways are also seen. Legend: 1. Outline of the active river caves, 2. D ivaka jama cave, 3. Unroofed cave, 4. Unroofed caves mentioned in the text: A: Unroofed cave near Povir, B: Unroofed cave in doline Radvanj, continuation of D ivaka jama, C: Unroofed cave above kocjanske jame, 5. H eight of the surface, 6. H eight of the water level in caves, 7. Reka river and ponors, 8. e supposed direction of water ow, 9. Outline of the town D ivaa. A NDREJ M IHEVC

PAGE 5

TIME in KARST 2007 39 Fig. 3: Formation of the unroofed cave. e idealised drawing is representing actual cases of unroofed or partly denuded caves from the D ivaa karst, where probably more than hundred meters of the rock above unroofed caves were removed. e transformation of cave to the unroofed cave is here presented in three stages: a: Epiphreatic cave passage was formed deep below the surface, some owstone was deposited aer the cave became inactive; b: Surface approached the cave. At one side the slope cut the cave and made the entrance into the passage; from the horizontal surface former vadose shas transformed into vertical entrance. T rough both entrances piles of boulders and scree deposited. c: Great deal of the ceiling dissolved, some collapsed and formed relief oblong depression of the unroofed cave ending in front of the entrance to the cave. Fig. 4: Formation of the unroofed cave. e idealised drawings are representing the actual cases of unroofed or partly unroofed caves from the D ivaa karst which were completely lled with allogenic uvial sediment. e transformation is here presented in three stages: a: Cave passage was formed deep below the surface. ere was alternation of the sedimentation of owstone and allogenic sediments of the underground river. T owards the top of the prole sediments became ner. b: Surface approached the cave. At the side the slope cuts passage and exposed the cave sediments on the surface. c: Aer disintegration of the ceiling from the top oblong depression formed. In it there are alochtonous sediments and few blocks of limestone and some owstone. e unroofed cave ends with steep limestone wall or slope from where the karst surface continues. T HE AGE OF K ARST RELIEF IN W S LOVENIA

PAGE 6

TIME in KARST 2007 40 T HE BLIND VALLE Y S OF THE M ATARSKO PODOLJE CONTACT KARST Alogene rivers owing to karst enhance the karstication process and form particular relief features. Phenomena and forms that develop at the contact of uvial and karst relief are the result of the interaction of both morpho logical systems. e Matarsko Podolje is a 25 km long and 2-5 km wide tilted karst surface. In longitudinal section it gen tly raises from about 490 m on N W to 650 m on SE side. e karst surface continues towards SE but from the highest point there is an abrupt change and relief lowers over the distance of 2 km for 200 m to Brgudsko podolje karst surface. From the ysch Brkini hills that are NE of podolje there are 17 sinking streams that formed a row of large blind valleys in the edge of the Matarsko Podolje. e bottoms of these valleys are situated between 490 to 510 m. As the valleys are incised in the border of the karst, uplied towards SE, the blind valleys lying more to the south are deeper. e most NW lying, Brezovica and Od olina blind valley are cut for about 50 m only while the deepest is the last one, Brdanska dana on SE, deepened into limestone for 250 m. e blind valleys started to cut into the corrosion plain with small transverse and longitudinal gradient as in the other case the uvial valleys should develop in them. ey should be preserved on karst as dry valleys. e corrosion plains along the ponors were controlled by the piezometric level this is why they are all at same altitude. In the SE part where the upli was stronger, the blind valleys show the disturbances caused by fast tectonic upli and are preserved on the karst surface. Above the Raika Dana blind valley there is a fossil one, on the bottom of which are some old sedi ment from ysch. is is now higher than the ysch hills where the sediment came from. e other case is the most SE blind valley Brdan ska Dana. It developed in the SE edge of the Matarsko Po dolje. e tectonic structure along which the Matarsko Podolje ends caused also the asymmetric development of the blind valley. e W side of the blind valley was up lied and developed two fos sil higher levels in the side of the blind valley. e Brkini series of blind valleys oer enough data to follow the sequence of the morphological events and dominant factors which were decisive for the forma tion of the actual relief forms. e former shape along the ponors on the border of im permeable hills was karst corrosion plain. e water ow ing on it had a modest gradient in karst and was capable of the aplanation of the surface only. e lowering of the piezometric level in the karst enabled the development of the relief depressions along the ponors. e deepen Fig. 5: B lind valleys B rezovica (B r) and Odolina (O) on the NW part of the M atarsko podolje karst. B lind valleys cut for about 50 m into the edge of the levelled karst surface where dolines and larger collapse dolines prevail. ere are no traces of dry valleys or dry blind valleys. Legend: 1. Sinking streams, 2. B oundary ysch limestone. A NDREJ M IHEVC

PAGE 7

TIME in KARST 2007 41 Fig. 6: B lind valleys Raika dana (R) and B rdanska dana (B) with fossil blind valleys (f1, f2). ese valleys developed in SE part of M atarsko podolje during the tectonic upli. Upli deformed older corrosion plain and created height dierence between M atarsko and B rgudsko podolje. Further SE there is another blind valley () which developed at the edge of B rgudsko podolje that was not uplied. Raika peina cave that was once formed by sinking streams is at elevation about 600 m high above the recent ponors. Legend: 1. Sinking streams, 2. B oundary ysch limestone, 3. Cave Raika peina. ing and the contemporane ous widening of the valleys followed the lowering of the karst water to the altitudes about 500 m. e incision of blind valleys into the leveled sur face probably started and continued trough the last tec tonic phase that is 6 Ma be fore present. is is also ac cordance with the age of the cave sediments from Raika peina which were dated by paleomagnetic method and correlated with palaeonto logic data to 3.5 Ma (Pruner & al. 2003). T HE UNROOFED CAVES OF THE EDGE OF THE P ODGORSKI KRAS Podgora karst is small 5 km wide and long karst plateau, SW continuation of the Kras. Its surface is located at 500 to 450 m a.s.l. e plateau surface is leveled and dismem bered only by numerous dolines. ere is a sharp edge of the plateau and towards W in less than 1 km relief drops for 400 m. At the foot of the plateau there are recent karst springs of the rivers Riana and Osapska reka at altitudes of about 50 m a.s.l. In the frnotie quarry, that is located on the edge of the plateau, several caves were opened. Shas with sta lagmites and stalactites on the walls were lled by gravel as well as numerous bones of large Pleistocene mammals felt down to shas. ere are also large remnants of horizontal caves. e largest, 150 m long partly unroofed passage with the diameter of more than 10 m was opened in the western part of the quarry. e passage was entirely lled by mas sive owstones deposited over the uvial sediments, lay ers of gravel and conglomerate mixed up with sand and clay layers. Sedimentary ll was 17 m thick at least. In the cave calcareous tubes a serpulids were found both in sediments and still attached to the scalloped wall. ey match the morphology of extant serpulid tubes of M arifugia cavatica (Mihevc 2000; Mihevc et al., 2001a). Marifugia cavatica Absolon and Hrabe, 1930 is the only fresh-water species of the Serpulidae family and the only T HE AGE OF K ARST RELIEF IN W S LOVENIA

PAGE 8

TIME in KARST 2007 42 Fig. 8: e view of the unroofed cave in a quarry face. Lower part of the cave passage was lled with mostly laminated yellowish brown uvial sediments. Upper part is lled with owstone. e karst denudation already unroofed the cave, so that the owstone is exposed to the surface. T ubes of M arifugia cavatica are on the scalloped walls in the lower part of the cave prole, which were protected by ne uvial sediments. Fig. 7: D EM of the Podgorski kras. Levelled karst surface of Paleocene limestone and some intercalated ysch is in sharp contrast with uvial relief that developed on Eocene ysch. At the foot of the karst there are the major karst springs where M arifugia cavatica still lives today. e fossil tubes were found in the large cave exposed in the rnotie quarry. Legend: 1. Unroofed cave, 2: Flysch, 3: Limestone. known tube worm inhabiting continental caves. Stable isotope analysis (Mihevc et al., 2002) of fossil tubes from frnotie quarry site is comparable with stable isotope compositions of recent fresh-water species and greatly diers from those of marine serpulids. M arifugia cavatica is lter feeder with freeswimming larvae (Matjai & Sket 1966). It is widely dis tributed within the Dinaric Karst and lives in springs of rivers Riana and Osapska reka which are only few km and 370 m apart from the quarry. Two proles were anal ysed within the cave and dat ed back to 1.76 Ma (frnotie I) and 2.5.6 Ma (frnotie II site) (Bosak & al. 1999, Bo sak & al. 2004). Geomorphologic evo lution of the plateau shows similarities to those of Kras and Matarsko podolje. Epi phreatic caves of the sinking rivers were lled with sedi ments; the surface was lev elled and uplied to present altitude. In the quarry there are several unroofed caves or remains old caves. e evolution of vertical shas with dominance of later autochthonous ll resulted from younger vadose speleogenesis and Pleistocene sedimen tation. A NDREJ M IHEVC

PAGE 9

TIME in KARST 2007 43 C ONCLUSIONS ree dierent relief settings on the Kras, Matarsko podolje and Podgorski kras plateau show quite similar evolution. ere are old caves present everywhere, which are now exposed by denudation. ey were epiphreatic caves that were formed by sinking rivers, bringing allo genic sediments to caves. At the end of the morphogenet ic phase all these caves were completely lled with uvial sediments. is indicates the diminishing of the gradient in the whole area. Aer the caves were lled the three ar eas were levelled. Planation occurred in the similar con ditions, most likely close to the level of the karst water. Diminishing of the gradient which ended with pla nation could mean the same tectonic phase which ended at about 6 Ma ago. Aer that a new tectonic phase started. ree areas faced upli and tilting for several hundreds meters. e upli was stronger in the SE part of the area. Karst denudation was evenly lowering the surface, so the surface remained well preserved, dissected on central parts of karst with dolines, which represent few percent of total area only. e even denudation exposed former caves to the surface. Some of them are lled with sedi ments, from some sediments were washed away or were never lled. On the edges of Matarsko podolje there were several sinking streams shaping blind valleys. eir incision was controlled by the piezometric level of the water in karst or the Matarsko podolje and by the tectonic upli, they are getting deeper towards SE. Tilt of planation surface, dierent depth and asymmetric or fossil blind valleys are clear indicators of the recent tectonics. Ages of sediments in the unroofed caves and the morphological datations are in accordance with the ages of main tectonic phases. From these data we can con clude that the oldest elements of the relief are the un roofed caves. e blind valleys are of same age even if they dier by the dimensions. e main process on the surface is even denudation and formation of dolines that form only small proportion of the surface. e remains of tubes of M arifugia cavatica pr eserved in a quarry, high above the recent water caves in dicate that the karst environment suitable for cave animals has been present for at least 6 Ma and that there was no inter ruption from the time of the formation of the caves in the frnotie quarry and drop of water table and/or tectonic upli for at least 370 m. R EFERENCES Absolon, K. & S. Hrabe, 1930: ber einen neuen Sss wasser-Polychaten aus den Hhlengewssern der Herzegowina. Zool. Anz., 88, 9-10, 259-264. Aguilar, J. P., Crochet J.Y ., Krivic K., Marandat B., J. Mi chaux J., Mihevc A., ebela S. & B. Sige, 1998: Pleis tocene small mammals from karstic llings of Slo venia. Acta carsologica, 27/2, 141-150, Ljubljana. Bosak P., Pruner P., & N. Zupan Hajna 1998: Palaeomag netic research of cave sediments in SW Slovenia. Acta carsologica, 1998, let. 27, t. 2, str. 151-179. Bosak P., Mihevc A., Pruner P., Melka K., Venhodov D. & A. Langrov, 1999: Cave ll in the frnotie Q uarry, SW Slovenia: Palaeomagnetic, mineralogi cal and geochemical study. Acta carsologica, 28/2, 2, 15-39, Ljubljana. Bosk, P., Mihevc A. & P. Pruner 2004: Geomorphologi cal evolution of the Podgorski Karst, SW Slovenia: contribution of magnetostratigraphic research of the frnotie II site with Marifugia sp. Acta carso logica, 2004, letn. 33, t. 1, str. 175-204, Ljubljana. Cvijit, J., 1924: Geomorfologija I, 324, Beograd. Fodor L. Jelen B., Marton E., Skaberne D., far J. & M. Vrabec, 1998: Miocene Pliocene tectonic evolution of the Slovenian Periadriatic fault: Implications for Alpine-Carpatian extrusion models. Tectonics, vol. 17, 5, 690-709. Gams, I., 1962: Meritve korozijske intenzitete v Sloveniji in njihov pomen za geomorfologijo, Geografski vestnik 34/1962, 3-20, Ljubljana. Gospodari R., 1985: On the spelogenesis of Divaka jama and Trhlovca Cave. Acta carsologica, XIII: 5-32, Ljubljana. Gospodari R., 1988: Paleoclimatic record of cave sedi ments from Postojna Karts. Ann. Soc. geol. Belg., 111, 91-95. Grund A., 1914: Der geographishes zyclus um Karst. Zeitsch. d. Gesell f. Erdkunde, S. 621-640, Berlin Habi P.,1982: Kraki relief in tektonika. Acta carsologi ca, 4, 23-43, Ljubljana. T HE AGE OF K ARST RELIEF IN W S LOVENIA

PAGE 10

TIME in KARST 2007 44 Knez M. & T. Slabe, 2005: Unroofed caves are an impor tant feature of karst surfaces: examples from the classical karst. Z. Geomorphol., 46, t. 2, str. 181191 Kratochvil, J., 1939: Marifugia cavatica edini sladkovodni serpulid, ostanek starodavnega ivalstva na jugo slovenskem krasu. Proteus, 6, 92-96, Ljubljana. Matjai, J. & B. Sket, 1996: Developpement larvaire du Serpulien cavernicole Marifugia cavatica Absolon et Hrabe. Int. J. Speleol., 25B, 1, 9-16, LAquilla. Melik, A., 1955: Kraka polja Slovenije v pleistocenu. Dela Intituta za geograjo SAZU, 3, 1-163, Ljubljana. Mihevc, A., 1993: Contact karst of Brkini Hills. Acta carsologica, 23, 100-109, Ljubljana Mihevc, A., 1996: Brezstropa jama pri Povirju. Nae jame 38, 92-101, Ljubljana. Mihevc, A. & N. Zupan Hajna, 1996: Clastic sediments from dolines and caves found during the construc tion of the motorway bear Divaa, on the Classical Karst. Acta carsologica, 25, 169-191, Ljubljana. Mihevc, A. Slabe T. & S. ebela, 1998: Denuded cavesan inherited element in the karst morphology; the case from Kras. Acta carsologica, 27/1, 165-174, Ljubljana. Mihevc, A., 1999: e caves and the karst surface-case study from Kras, Slovenia. Etudes de gographie physique, suppl. XX VIII, Colloque europen-Karst 99, 141-144. Mihevc, A., 2000: Fosilne cevke iz brezstrope jame ver jetno najstareji ostanki jamskega cevkarja Marifu gia (Annelida: Polychaeta). Acta carsologica, 29/2, 261-270, Ljubljana. Mihevc, A., 2001: Speleogeneza Divakega krasa. Zbirka ZRC, 27: 1-180. Ljubljana. Mihevc, A. Sket B., Pruner P. & P. Bosk, 2001: Fossil re mains of a cave tube worm (Polychaeta: Serpulidae) in an ancient cave in Slovenia. Proc., 13 th Inter national Speleological Congress, 4 th Speleological Congress of Latin America and the Carribean, 26 th Brazilian Congress of Speleology, Brasilia, July 1522, 2001, 20-24, Brasilia. Mihevc, A., Bosak P. Pruner P. & B. Vokal 2002: Fossil remains of the cave animal Marifugia cavatica in the unroofed cave in the fernotie quarry, W Slovenia. Geologija, 45, 2, str. 471-474, Ljubljana. Pruner, P., Bosk P. Mihevc A. Kadlec J. Man O. & P. Schnabl, 2003: Preliminary report on palaeomag netic research on Raika peina Cave, SW Slovenia. 11th International Karstological School Classical Karst. Karst Terminology. Guide booklet of the ex cursions and abstracts of lectures or poster presen tations, Postojna, July 2003: 35-37. Postojna. Radinja, D., 1972: Zakrasevanje v Sloveniji v lui celot nega morfogenetskega razvoja. Geografski zbornik, 13, SAZU, Ljubljana. Rogli, J., 1957: Zaravni u vapnencima. Geografski glasnik 19, 103-134, Zagreb. Sket, B., 1970: ber Struktur und Herkun der unterir dischen Fauna Jugoslawiens. Biol. Vestn., 18, 6978, Ljubljana. Slabe, T., 1997: Karst features discovered during mo torway construction in Slovenia. Environ. geol. (Berl.), 1997, letn. 32, t. 3, str. 186-190. Vrabec, M., & L. Fodor, 2006: Late Cenozoic tectonics of Slovenia: structural styles at the Northeastern corner of the Adriatic microplate. e Adria micro plate: GPS geodesy, tectonics and hazards, NATO Science Series, IV, Earth and Environmental Sciences, vol. 61). Dordrecht: Springer, 151-168. Zupan Hajna, N., 1991: Flowstone datations in Slovenia. Acta carsologica, 1991, let. 20, str. 187-204. A NDREJ M IHEVC



PAGE 1

EVOLUTION AND AGE RELATIONS OF KARST LANDSCAPES RAZVOJ IN STAROSTNI ODNOSI KRAKIH POKRAJIN W illiam B. WHITE 1 Izvleek UDK 551.44 William B. White: Razvoj in starostni odnosi krakih pokra jin Vsaka kraka pokrajina je delo, ki napreduje. Razvoj pokrajine, ki ga je mogoe opazovati, je odvisen od medsebojno tekmujoih procesov povrinske denudacije, vrezovanja povrinskih tokov, razvoja jam in tektonskega dvigovanja. tevilni podatki o teh procesih, zbrani za dve ziografski enoti v gorovju Apalai na vzhodu ZDA kaejo, da se starost in asovna skala ujemata s prejnjimi geomorfnimi razlagami. Izsledke bolj ohlapno potrjuje nekaj podatkov, dobljenih za jamske sedimente. bal pa so spremembe razmerja hitrosti zaradi lokalnih posebnosti v velikosti cele magnitude in je torej regionalna interpretacija v najboljem primeru le grob pribliek. Kljune besede: kraka denudacija, razvoj pokrajine. 1 Materials Research Institute and Department of Geosciences,e Pennsylvania State University,University Park, PA 16802 USA; e-mail: wbw2@psu.edu Received/Prejeto: 20.12.2006 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 45 Abstract UDC 551.44 William B. White: Evolution and age relations of karst land scapes Any karst landscape is a work in progress. e observed evolu tion of the landscape is dictated by competing rate processes of surface denudation, stream downcutting, cave development, and tectonic upli. Q uantitative data on these processes, ap plied to two physiographic provinces of the Appalachian Mountains of eastern United States gives ages and time scales that are in agreement with previous geomorphic interpreta tion. e results are anchored, very loosely, by the few dates that have been established for cave sediments. Unfortunately, the measured rates vary over an order of magnitude as a result of local circumstances making regional interpretation a rough approximation at best. Key words: uviokarst, karst denudation, landscape evolution. I NTRODUCTION : WHAT D O WE M EAN B Y THE A GE OF A K ARST L ANDSCAPE ? By landscape, we usually mean some dened area of the earths surface as it exists at a single moment of time. Although most of the landforms remain constant on a human time scale, they are actually in the process of con tinuous evolution. In at least a microscopic way, todays landscape is not quite the same as yesterdays landscape. If the time scale is extended to thousands or millions of years, very large changes will have occurred to the land scape. Caves will have come and gone. A karst landscape, such as a doline plain, might supercially look the same but they wouldnt be the same dolines. e land surface is continuously lowered by dissolution. Old dolines disap pear and new dolines are formed. us when we speak of the age of a karst land scape we must carefully specify both spatial scales and time scales. At the largest scales we can talk about global chemical erosion over geologic time (Gibbs et al., 1999). W e can talk about the general lowering of a karst land scape, the phenomenon generally called karst denuda tion. W e can talk about the dierential dissolution that produces surface karst landforms. W e can talk about subsurface dissolution that produces caves. W e can talk about the relative rates of landscape evolution on karstic and non-karstic rocks. W e can talk about rates of tectonic upli that provide the gravitational gradients that drive all of the processes. e observed landscape in any geo

PAGE 2

TIME in KARST 2007 46 logic setting is the result of the interaction of all of these competing rate processes. As a result, age becomes a very slippery concept. e objective of the present paper is to determine what constraints on the time evolution of karst land scapes can be extracted from known rates of the land scape processes. e discussion will be limited to uvio karst. is means that consideration much be given to mass transport by surface streams on both carbonate and non-carbonate rocks as well as subsurface mass transport by dissolution. Illustrative examples are taken from the Appalachian Mountains of eastern United States. In the Appalachians are displayed two geologic settings: (1) e limestone valleys of the folded Appalachians where the karst surface is exposed across wide valley oors so that the disolutional dissection of the karst is primarily ver tical and distributed across the surface. (2) e Appala chian Plateaus where the carbonate rocks are protected by clastic caprock and where the dissolutional attack is primarily by valley incision around the perimeter. U NIFORM L ANDSCAPE L O W ERING : K ARST D ENUDATION Setting aside the necessity for also removing insoluble residue, the evolution of a carbonate rock landscape can be considered to be a purely chemical process. e rock mass is taken into solution and carried away by the con tinuous ux of water that moves through the system. Any measure of the rate of carbonate removal can be recalcu lated as an average lowering of the karst surface, a quan tity known as the karst denudation rate. Various methods have been devised for the direct measurement of denudation rate (summarized by W hite, 2000). e rate of surface lowering can be measured di rectly on exposed rock surfaces using embedded reference pins and a precision micrometer (High and Hanna, 1970). e micrometer works best on bare rock surfaces. Most limestone dissolution takes place under a soil mantle. A technique to measure dissolution rates in soil is to bury carefully weighed plaques of limestone for a known time, then re-excavate and weigh them again (Gams, 1981). On the scale of the entire drainage basin, it is pos sible to estimate denudation rate by a mass balance cal culation using the volume of water leaving the basin and the concentration of dissolved carbonates contained in the water. e denudation rate is then given by [1] In this equation, D n is the denudation rate in m 3 km -2 yr -1 (numerically equivalent to the more com mon unit of mm/ka), A is the basin area in km 2 N L is the fraction of the basin underlain by carbonate rocks, is the density of carbonate rock in gcm -3 t R is the period of record in years, Q(t) is the instantaneous discharge in m 3 s -1 (i.e. the hydrograph) and H(t) is the instantaneous (Ca + Mg) hardness in gcm -3 (i.e. the chemograph). e constant, K, contains unit conversions and has the value 10 -12 for the units given. Because the mass balance equa tion requires continuous records of both discharge and hardness which are not oen available, a variety of ap proximations have been proposed. If the reaction between inltrating water and car bonate rock at the base of the epikarst is assumed to reach equilibrium, the denudation rate can be calculated from rst principles (W hite, 1984). exposed rock surfaces using embedded reference pins and a precision micrometer (High and Hanna, 1970). The micrometer works best on bare rock surfaces. Most limestone dissolution takes place under a soil mantle. A technique to measure dissolution rates in soil is to bury carefully weighed plaques of limestone for a known tim e, then re-excavate and weigh them again (Gams, 1981). On the scale of the entire drainage basin, it is possible to estimate denudation rate by a mass balance calculation using the volume of water leaving the basin and the concentration of dissolved carbonates containe d in the water. The denudation rate is then given by 2 1 ) ( ) ( 1 t t R L n dt t H t Q t K A N D [1] In this equation, D n is the denudation rate in m 3 km -2 yr -1 (numerically equivalent to the more common unit of mm/ka), A is the basin area in km 2 N L is the fraction of the basin underlain by carbonate rocks, is the density of carbonate rock in gcm -3 t R is the period of record in years, Q(t) is the instantaneous discharge in m 3 s -1 (i.e. the hydrograph) and H(t) is the instantaneous (Ca + Mg) hardness in gcm -3 (i.e. the chemograph). The constant, K, contains unit conversions and has the value 10 -12 for the units given. Because the mass balance equation requires co ntinuous records of both discharge and hardness which are not often available, a variety of approximations have been proposed. If the reaction between infiltrating water and carbonate rock at the base of the epikarst is assumed to reach equilibrium, the denudation rate can be calculated from first principles (White, 1984). ) ( 4 3 1 3 1 2 2 1 3 2 3 2 2 E P P K K K K M D CO HCO Ca CO c cal n [2] In this equation, D n is the denudation rate in mm/ka. M cal is the molecular weight of calcite (or a weighted mix of the molecular weights of calcite and dolomite) and is the rock density in gcm -3 reactions and the evapotranspiration) is the annual runoff in mm/yr Many of the earlier measurements of karst denudation rates were reviewed and analyzed by Smith and Atkinson (1976). A selection of more recent data are displayed in Figure 1. The chosen examples include data from each of the three measurement methods described above and these give comparable results. The regional environments represented in Figure 1 include arid, alpine, northern, and temperate. Denudation rates vary by a factor of 5-10 within each group but the groups are almost completely overlapping. Local conditions at the sampling site, including soil c over, available water, and rock lithology, all contribute so that lo cal site variation masks regional scale [2] In this equation, D n is the denudation rate in mm/ ka. Mcal is the molecular weight of calcite (or a weighted mix of the molecular weights of calcite and dolomite) and is the rock density in gcm -3 e Ks are the usual equilibrium constants for carbonate reactions and the s are the activity coercients. P-E (precipitation minus evapotranspiration) is the annual runo in mm/yr Many of the earlier measurements of karst denu dation rates were reviewed and analyzed by Smith and Atkinson (1976). A selection of more recent data are displayed in Figure 1. e chosen examples include data from each of the three measurement methods described above and these give comparable results. e regional environments represented in Figure 1 include arid, al pine, northern, and temperate. Denudation rates vary by a factor of 5-10 within each group but the groups are almost completely overlapping. Local conditions at the sampling site, including soil cover, available water, and rock lithology, all contribute so that local site variation masks regional scale variations. ere is also the ques tion of how denudation rates have changed in response to climatic uctuations of the Pleistocene. For the re gional scale landscape evolution of interest in this paper, WILLIAM B. W HITE

PAGE 3

TIME in KARST 2007 47 about the best that can be said is that exposed karst surfaces in the Appala chian Mountains would be lowered by dissolution at a rate of 20-30 mm/ka. Fig. 1: M easured denudation rates. All data have been converted to units of mm/ ka. K R Mt. Kruterin, Austria (buried tablets) (Zhang et al., 1995). L G Logatec D oline, Slovenia (buried tablets) (Gams, 1981). H S H ochschwab M assif, Austrian Alps (buried tablets) (Plan, 2005). C-Y Cooleman Plain and Y arrangobilly Caves area, New South Wales, Australia (microerosion meter) (Smith et al., 1995). AK southeastern Alaska (microerosion meter) (Allred, 2004). S-S SaltelletSvartisen area, northern Norway (mass balance) (Lauritzen, 1984). Regional rivers draining through areas of uviokarst cut normal valleys in the clastic rocks that overlie, under lie, or border the karstic rocks and may appear as sur face streams in valleys cut into the karstic rocks. Meas urements of the downcutting rates of larger rivers are dircult because many of them, in their lower reaches, are at grade with a sediment load balanced against the discharge. Lowering of the bedrock channel can be very slow. A few data are given in Table 1. Lowering rates in the tectonically stable Appalachians fall in the same 2030 mm/ka range as is found for denudation of karst sur faces. Only one example, the Bighorn Basin in western United States is a factor of ten higher and may represent a higher rate of tectonic upli. Small tributary streams that ow from surround ing non-karstic lands onto the karst and then sink at the contact with the soluble rocks seem to have a much higher rate of channel lowering. Some direct micrometer measurements in the beds of sinking streams are given in Table 2. Sinking stream waters are generally highly un saturated so that sinking streams downcut rapidly into the carbonate rock at their sink points. Similar measure ments at spring outlets produce much smaller numbers. e highest values yet reported were for a muskeg-drain ing stream in Alaska (Allred, 2004) where there is an im plication that organic acids may also play a role. R ATES OF V ALLE Y D EEPENING T ab. 1. D owncutting Rate of Some M oderate-Size Rivers Name and Location Rate (mm/ka) Reference Bighorn River, Wyoming 350 Stock et al. (2006) East Fork, Obey River, Tennessee 30 Sasowsky et al. (1995) Anthony & Granger (2004) Green River at Mammoth Cave, Kentucky 30 Granger et al. (2001) Juniata River, Newport, Pennsylvania 27 Sevon (1989) New River at Pearisburg, Virginia 27 Granger et al. (1997) EVOLUTION AND AGE RELATIONS OF KARST LANDSCAPES

PAGE 4

TIME in KARST 2007 48 Caves here considered to be master trunk caves related to surface base-level streams have a three-stage develop ment. (1) e initiation phase is the evolution of an initial mechanical fracture to a critical-size protoconduit about one centimeter in aperture. (2) e enlargement phase takes the protoconduit up to the meters to tens of meters diameter of a typical cave passage. (3) e stagnation and decay phase is that period aer the cave passage has been drained and abandoned by lowering base levels. As the stagnation phase progresses, entrances are developed and process of collapse, speleothem growth, and sediment inlling choke o the once continuous conduit. Deepening of surface valleys breaks the cave into fragments. e initiation phase is almost purely chemical. Nearly saturated water percolates along alternative paths in the carbonate rock, slowly enlarging them. e initia tion phase ends when one pathway becomes surciently large to permit critically undersaturated water to pass completely through the aquifer. As a result, the nal lay out of the conduit system is largely determined during the initiation phase. e initiation phase is particularly amenable to geochemical modeling and some very el egant models have been constructed (Dreybrodt et al., 2005). e time scale for the initiation depends on assumed initial conditions but ap pears to be in the range of 10,000 to 20,000 years. e enlargement phase is large ly independent of outside factors. e rate of retreat of passage walls can be described by the PalmerDreybrodt equation (Palmer, 1991). The enlargement phase is largely independent of outside factors. The rate of retreat of passage walls can be described by the Palmer-Dreybrodt equation (Palmer, 1991). R n S C C k S 1 56 31 [3] S is the rate of wall retreat in cm/yr. Some calculations for passage enlargement are plotted in Fig. 2. The rate constant, k, was taken from Palmer (1991). The rock density, R was set equal to 2.65 g/cm 3 The reaction order, n = 1, in the fast dissolution regime. The only environmentally sensitive parameter is the saturation concentration of calcium carbonate which depends on the carbon dioxide partial pressure. Figure 2 shows the passage enlargement rates expected for a reasonable range of CO 2 pressures. Although the details are site-specific, even rough calculations suggest that 50,000 to 100,000 years are sufficient to allow a master cave to develop. The relationship between hydraulic gradient, h f /L, discharge, Q, and passage radius, R, is given by a form of the Darcy-Weisbach equation 5 2 2 4 R g Q f L h f [4] Some maximum gradients that can be supported by a given size conduit are plotted in Figure 3 for a selection of discharges. Because of the low hydraulic resistance of conduit systems, the elevation difference between the headwaters and the downstream reaches of surface streams can provide sufficient head to drive the cave-forming process. By this process of autopiracy, caves develop beneath surface valleys (or more often in the valley walls) and drain off the flow from the surface stream. Such caves generally have flatter gradients than the valleys that they underdrain. Unlike karst surfaces or surface valleys which are continuously evolving, caves remain as fixed elevation markers and are the only features of the karst landscape for which the age is locked in. Caves may ride upward with tectonic uplift, but otherwise remain fixed as the surface landscape falls around them. This is the stagnation and decay once-continuous conduit is fragmented as the surface lowers and valleys deepen. In terms of importance as biological habitat, the final stage is very important. Unfortunately, the details of the conduit decay of the conduit depends on local circumstances does not lend itself to numerical analysis. [3] S is the rate of wall retreat in cm/yr. Some calculations for passage Fig. 2: Enlargement phase for typical conduits assuming various carbon dioxide partial pressures based on the Palmer-D reybrodt equation. C AVE D EVELOPMENT IN F LUVIOKARST T ab. 2. D owncutting Rate in Small Karst Streams Name and Location Rate (mm/ka) Reference Cataract Cave, southeast Alaska 137 Allred (2004) County Clare, Ireland 500 400 High and Hanna (1970) Muskeg Inow Cave, southeast Alaska 1670 1080 Allred (2004)) Slate Cave, southeast Alaska 180 Allred (2004) Yarrangobilly, NSW, Australia 200 Smith et al. (1995) WILLIAM B. W HITE

PAGE 5

TIME in KARST 2007 49 Fig. 3: Supportable hydraulic head as a function of conduit radius for various discharges. e D arcy-Weisbach friction factor, f = 100. e gravitational acceleration, g = 9.8 msec -1 enlargement are plotted in Fig. 2. e rate constant, k, was taken from Palmer (1991). e rock density, R was set equal to 2.65 g/cm 3 e reaction order, n = 1, in the fast dissolution regime. e only environmentally sensi tive parameter is the saturation concentration of calcium carbonate which depends on the carbon dioxide partial pressure. Figure 2 shows the passage enlargement rates expected for a reasonable range of CO 2 pressures. Al though the details are site-specic, even rough calcula tions suggest that 50,000 to 100,000 years are surcient to allow a master cave to develop. e relationship between hydraulic gradient, h f /L, discharge, Q and passage radius, R, is given by a form of the Darcy-W eisbach equation The enlargement phase is largely independent of outside factors. The rate of retreat of passage walls can be described by the Palmer-Dreybrodt equation (Palmer, 1991). R n S C C k S 1 56 31 [3] S is the rate of wall retreat in cm/yr. Some calculations for passage enlargement are plotted in Fig. 2. The rate constant, k, was taken from Palmer (1991). The rock density, R was set equal to 2.65 g/cm 3 The reaction order, n = 1, in the fast dissolution regime. The only environmentally sensitive parameter is the saturation concentration of calcium carbonate which depends on the carbon dioxide partial pressure. Figure 2 shows the passage enlargement rates expected for a reasonable range of CO 2 pressures. Although the details are site-specific, even rough calculations suggest that 50,000 to 100,000 years are sufficient to allow a master cave to develop. The relationship between hydraulic gradient, h f /L, discharge, Q, and passage radius, R, is given by a form of the Darcy-Weisbach equation 5 2 2 4 R g Q f L h f [4] Some maximum gradients that can be supported by a given size conduit are plotted in Figure 3 for a selection of discharges. Because of the low hydraulic resistance of conduit systems, the elevation difference between the headwaters and the downstream reaches of surface streams can provide sufficient head to drive the cave-forming process. By this process of autopiracy, caves develop beneath surface valleys (or more often in the valley walls) and drain off the flow from the surface stream. Such caves generally have flatter gradients than the valleys that they underdrain. Unlike karst surfaces or surface valleys which are continuously evolving, caves remain as fixed elevation markers and are the only features of the karst landscape for which the age is locked in. Caves may ride upward with tectonic uplift, but otherwise remain fixed as the surface landscape falls around them. This is the stagnation and decay once-continuous conduit is fragmented as the surface lowers and valleys deepen. In terms of importance as biological habitat, the final stage is very important. Unfortunately, the details of the conduit decay of the conduit depends on local circumstances does not lend itself to numerical analysis. [4] Some maximum gradients that can be supported by a given size conduit are plotted in Figure 3 for a selection of discharges. Because of the low hydraulic resistance of conduit systems, the elevation dierence between the headwaters and the downstream reaches of surface streams can pro vide surcient head to drive the cave-forming process. By this process of autopiracy, caves develop beneath surface valleys (or more oen in the valley walls) and drain o the ow from the surface stream. Such caves generally have atter gradients than the valleys that they underdrain. Unlike karst surfaces or surface valleys which are continuously evolving, caves remain as xed elevation markers and are the only features of the karst landscape for which the age is locked in. Caves may ride upward with tectonic upli, but otherwise remain xed as the surface landscape falls around them. is is the stagna tion and decay phase in the caves history and is the phase in which entrances are developed and the once-continu ous conduit is fragmented as the surface lowers and val leys deepen. In terms of importance as biological habitat, the nal stage is very important. Unfortunately, the de tails of the conduit decay of the conduit depends on local circumstances does not lend itself to numerical analysis. e Cumberland Plateau is the southern-most exten sion of the great Appalachian plateaus that extend from New Y ork State into Alabama. e Cumberland Plateau in Tennessee and Alabama is an upland of low-dip Mis sissippian rocks. e plateau is capped with a highly re sistant quartzite which provides a reference elevation at about 550 to 600 meters. e denudation of the resistant quartzite is very slow, 3-5 mm/ka, according to Anthony and Granger (2004). e plateau is bounded by a pro nounced escarpment into which deep valleys (known locally as coves) have been incised. At the base of the western escarpment is a karst surface known as the High land Rim. e doline surface of the Highland Rim ex tends into many of the deeper coves. Mississippian lime stones underlie the valley walls of the coves and much of the Highland Rim (Fig. 4). e downcutting rate of one incised valley, that of the East Fork of the Obey River in north-central Ten nessee was rst calculated from magnetic reversals in the sediments of one of the caves in the valley wall (Sasowsky et al. 1995). is number was revised when cosmogenic isotope dating of the same cave showed that A GE R ELATIONSHIPS IN THE P LATEAU F LUVIOKARST S ETTING EVOLUTION AND AGE RELATIONS OF KARST LANDSCAPES

PAGE 6

TIME in KARST 2007 50 of 30 mm/ka (Table 1) is similar the downcutting rate of other moderate size rivers and also very similar to the expected denuda tion rate. e Highland Rim surface at the base of the western escarp ment has nearly eroded to the bot tom of the carbonate sequence. It all about 150 meters of limestone have been removed. If the High land Rim is raised according to the 30 mm/ka denudation rate, approximately 5 million years ago, the erosion surface was at the top of the limestone. e sediments in Big Bone Cave were dated at 5.7 Ma (Anthony and Granger, 2004) and it was claimed that this date represents a time when the Cumberland River was owing at the elevation of the Highland Rim. Fig. 4: Schematic cross-section through the western escarpment of the Cumberland Plateau. icknesses of individual beds are nominal values; bed thicknesses vary considerably over short distances (M ilici et al., 1979). the paleomagnetic measurements referred to an earlier reversal (Anthony and Ganger, 2004). e revised value e karst surfaces of the Great Valley and Valley and Ridge Provinces of the folded Appalachians are breached anticlines. Deep erosion along the anticlines has exposed the Ordovician and Cambrian limestones and dolomites which now form the valley oors. e more resistant quartzites on the anks of the anticlines remain as long nearlyparallel ridges bounding the valleys (Fig. 5). Contemporary surface streams have downcut 50 to 75 meters into the valley surface. ere must have been a time when the anticlines were rst breached to expose the carbonate rocks to denudation. Figure 6 shows the se quence of events (without time scale) and includes the recognized erosion surfaces identied in central Pennsyl vania. e Nittany Valley near State College, Pennsylvania is an interuve area. Here are found residual soils Fig. 5: Sketch showing topographic relations in central Pennsylvania. Ridges are supported by resistant quartzite; most of the valley oors are underlain by Cambrian and Ordovician carbonate rocks. Aer D eike (1961). A GE R ELATIONSHIPS IN A PPALACHIAN V ALLE Y F LUVIOKARST S ETTING WILLIAM B. W HITE

PAGE 7

TIME in KARST 2007 51 Fig. 6: e evolution of the Nittany V alley in central Pennsylvania showing traditional erosion surfaces. Aer Gardner (1980). rocks, a calculation based on insoluble residue content and bulk density suggests that more than 425 meters of carbonate rock were removed to accumulate this thick ness of soil ( W hite and W hite, 1991). On the (quite pos sibly unreasonable) assumption that the denudation rate has been 30 mm/ka, the removal of 425 meters of car bonates would require on the order of 14 million years, placing the beginning of what has been a uniform denu dation process in mid-Miocene time. e present relief between the valley oor and the ridge tops is about 250 meters. e carbonate surface at the beginning of the de nudation process would be 175 meters above the pres ent-day ridge tops. However, the estimated denudation would not include the entire carbonate section so it does not represent the breaching of the anticline which must have taken place earlier. e accordant ridge-lines of the folded Appalachians are oen taken to represent the Schooley Peneplain. If these quartzite-topped ridges erode as slowly as similar rocks on the Cumberland Plateau, the limestone would have lled the valley to the level of the ridge tops only 8 9 Ma ago. e age of the Schooley Peneplain would be much less than many ages that have been assigned to it, some setting the age as far back as the Jurassic. e valley oors which represent the Harrisburg Survey have been dissected by present day streams to produce an internal relief of about 60 meters. e caves of the Valley and Ridge Province are found within this in terval. Some are inlet caves with high gradients due to the rapid downcutting of sinking streams. Others are frag ments of base-level conduits. Given the observed rates of stream downcutting, the time span available for the development of these caves is 2 3 million years. with thicknesses averaging 50 meters. On the assump tion that these are let-down soils consisting of the in soluble residues from the dissolution of the carbonate Although doline plains give the impression of stable ero sion surfaces, denudation measurements suggest the rate of lowering is comparable to the rate of downcutting of surface valleys. e horizontal surface is maintained be cause of the internal drainage through the dolines. It is, therefore, problematic to attempt to assign and age to karst surfaces. Cave development is very rapid compared with the evolution of the surface landscape. Caves in tectonically stable areas serve as better markers of temporarily sta ble pauses in base level lowering than do either surface streams or the elevations of karst erosion surfaces. is conclusion has been suspected at least since the work of Davies (1960) but was given much stronger support by recent cosmogenic isotope dating (Granger et al., 1963; Anthony and Granger, 1964). It is also supported by the present geochemical calculations and mass balance argu ments. C ONCLUSIONS EVOLUTION AND AGE RELATIONS OF KARST LANDSCAPES

PAGE 8

TIME in KARST 2007 52 Allred, K., 2004: Some carbonate erosion rates of south east Alaska J ournal of Cave and Karst Studies 66, 89-97. Anthony, D.M. & D.E. Granger, 2004: A late Tertiary origin for multilevel caves along the western escarp ment of the Cumberland Plateau, Tennessee and Kentucky, established by cosmogenic 26 Al and 10 Be J ournal of Cave and Karst Studies 66, 46-55. Davies, W .E., 1960: Origin of caves in folded limestone National Speleological Society B ulletin 22, 5-18. Deike, R.G., 1961: Karst development in Brush Valley, PA Nittany Grotto Newsletter 9, 121-128. Dreybrodt, W ., F. Gabrovek & D. Romanov, 2005: Pro cesses of speleogenesis: A modeling approach Car sologica, No. 4, 375 p. Gams, I., 1981: Comparative research of limestone solu tion by means of standard tablets 8 th International Congress of Speleology Proceedings, Bowling Green, Kentucky, USA, p. 273-275. Gardner, T., 1980: Geomorphology of Nittany Valley Chapter 5 in Soils and Geology of Nittany V alley, E.J. Ciolkosz, R.R. Parizek, G. W Petersen, R.L. Cunningham, T.W Gardner, J.W Hatch, & R.D. Shipman, Eds., e Pennsylvania State University Agronomy Series No. 64, p. 52-75, University Park, PA. Gibbs, M.T., G.J.S. Bluth, P.J. Fawcett, & L.R. Kump, 1999: Global chemical erosion over the last 250 MY : Vari ations due to changes in paleogeography, paleocli mate, and paleogeology American J ournal of Sci ence 299, 811-51. Granger, D.E., D. Fabel & A.N. Palmer, 2001: Pliocene Pleistocene incision of the Green River, Kentucky, determined from radioactive decay of cosmogenic 26 Al and 10 Be in Mammoth Cave Sediments Geo logical Society of America B ulletin 113, 825-836. Granger, D.E., J.W Kirchner & R.C. Finkel, 1997: Q ua ternary downcutting rate of the New River, Virginia, measured from dierential dcay of cosmogenic 26 Al and 10 Be in cave-deposited alluvium Geology 25, 107-110. High, D. & F.K. Hanna, 1970:A method for the direct measurement of erosion on rock surfaces B ritish Geomorpological Research Group T echnical B ulletin No. 5, p. 24. Lauritzen, S.-E., 1984: Some estimates of denudation rates in karstic areas of the Saltellet Svartisen Re gion, North Norway Catena 11, 97-104. Milici, R.C., G. Briggs, L.M. Knox, P.D. Sitterly & A.T. Statler, 1979: e Mississippian and Pennsylvanian (Carboniferous) Systems in the United States Ten nessee U.S. Geological Survey Professional Paper 1110-G, 38 p. Palmer, A.N., 1991: Origin and morphology of limestone caves Geological Society of America B ulletin 103, 1-21. Plan, L., 2005: Factors controlling carbonate dissolution rates quantied in a eld test in the Austrian Alps Geomorphology 68, 201-212. Sasowsky, I.D., W .B. W hite & V.A. Schmidt, 1995: Deter mination of stream-incision rate in the Appalachian plateaus by using cave-sediment magnetostratigra phy Geology 23, 415-418. Sevon, W .D., 1989: Erosion in the Juniata River drainage basin, Pennsylvania Geomorphology 2, 303-318. Smith, D.I. & T.C. Atkinson, 1976: Process, landforms and climate in limestone regions Chapter 13 in Geomor phology and Climate, E. Derbyshire, Ed., John W iley, p. 367-409, London. Smith, D.I., M.A. Greenaway, C. Moses, & A.P. Spate, 1995: Limestone weathering in eastern Australia. Part I. Erosion rates Earth Surface Processes and Landforms 20, 451-463. Stock, G.M., C.A. Riihimaki, & R.S. Anderson, 2006: Age constraints on cave development and landscape evolution in the Bighorn Basin of W yoming, USA J ournal of Cave and Karst Studies 68, 76-84. W hite, W .B., 1984: Rate processes: chemical kinetics and karst landform development in Groundwater as a Geomorphic Agent, R.G. LaFleur, Ed., Allen & Un win, p. 227-248, London. W hite, W .B., 2000: D issolution of limestone from eld ob servations in Speleogenesis, A. Klimchouk, D.C. Ford, A.N. Palmer & W Dreybrodt, Eds., National Speleological Society, p. 149-155, Huntsville, AL. W hite, W .B. and E.L. W hite, 1991: Karst erosion surface in the Appalachian H ighlands in Appalachian Karst, E.H. Kastning and K.M. Kastning, Eds. National Speleological Society, p. 1-10, Huntsville, AL. Zhang, D., H. Fischer, B. Bauer, R. Pavuza & K. Mais, 1995: Field tests of limestone dissolution rates in karstic Mt. Kruterin, Austria Cave and Karst Sci ence 21, 101-104. R EFERENCES WILLIAM B. W HITE



PAGE 1

C AVE AND K ARST EVOLUTION IN THE A LPS AND THEIR RELATION TO PALEOCLIMATE AND PALEOTOPOGRAPH Y RAZVOJ JAM IN KRASA V ALPAH V LUfI PALEOKLIME IN PALEOTOPOGRAFIJE Philippe A UDRA 1 Alfredo B INI 2 Franci G ABROV EK 3 Philipp H USELMANN 4 Fabien H OBL A 5 Pierre-Y ves J EANNIN 6 Jurij K UNAVER 7 Michel M ONBARON 8 France U TER I f 9 Paola T OGNINI 10 Hubert T RIMMEL 11 & Andres WILDBERGER 12 1 quipe Gestion et valorisation de lenvironnement, UMR 6012 ESPACE CNRS, University of Nice Sophia-Antipolis, 98 boule vard douard Herriot, BP 209, 06204 Nice cedex, France (audra@unice.fr). 2 Dipartimento di Scienze della terra, Universit di Milano, via Mangiagalli 34, 20133 Milano, Italy (alfredo.bini@unimi.it) 3 Karst research Institute ZRC SAZU, Titov trg 2, 66230 Postojna, Slovenia (gabrovsek@zrc-sazu.si) 4 Institut suisse de splologie et de karstologie (ISSKA), CP 818, 2301 La Chaux-de-Fonds, Switzerland (praezis@speleo.ch) 5 ED Y TEM, Universit de Savoie, 73376 Le Bourget cdex (Fabien.Hoblea @ univ-savoie.fr) 6 Institut suisse de splologie et de karstologie (ISSKA), CP 818, 2301 La Chaux-de-Fonds, Switzerland (info@isska.ch) 7 Hubadova ulica 16, 61000 Ljubljana, Slovenia (jurij.kunaver@siol.net) 8 Dpartement de gosciences/gographie, ch. du Muse 4, Universit de Fribourg, 1700 Fribourg, Switzerland (michel.monbaron@unifr.ch) 9 Dept. of Geology NTF, University of Ljubljana, 1001 Ljubljana, Slovenia (france.sustersic@ntfgeo.uni-lj.si) 10 via Santuario inferiore, 33/D, 23890 Barzago (LC), Italy (paolatognini@iol.it) 11 Draschestrasse 77, 1230 W ien, Austria (Hubert.Trimmel@reex.at) 12 Dr. von Moos AG, Engineering Geology, 8037 Zrich, Switzerland (wildsch@bluewin.ch) Received/Prejeto: 01.12.2006 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 53 Abstract UDC 551.435.84(234.3) Philippe Audra, Alfredo Bini, Franci Gabrovek, Philipp Huselmann, Fabien Hobla, Pierre-Yves Jeannin, Jurij Ku naver, Michel Monbaron, France uteri, Paola Tognini, Hubert Trimmel & Andres Wildberger: Cave and Karst evolu tion in the Alps and their relation to paleoclimate and paleo topography Progress in the understanding of cave genesis processes, as well as the intensive research carried out in the Alps during the last decades, permit to summarize the latest knowledge about Alpine caves. e phreatic parts of cave systems develop close to the karst water table, which depends on the spring position, which in turn is generally related to the valley bottom. us, caves are directly linked with the geomorphic evolution of the surface and reect valley deepening. e sediments deposited in the caves help to reconstruct the morphologic succession and the paleoclimatic evolution. Moreover, they are the only means to date the caves and thus the landscape evolution. Caves appear as soon as there is an emersion of limestone from the sea and a water table gradient. Mesozoic and early tertiary paleokarsts within the alpine range prove of these ancient emersions. Hy drothermal karst seems to be more widespread than previously Izvleek UDK 551.435.84(234.3) Philippe Audra, Alfredo Bini, Franci Gabrovek, Philipp Huselmann, Fabien Hobla, Pierre-Yves Jeannin, Jurij Ku naver, Michel Monbaron, France uteri, Paola Tognini, Hubert Trimmel & Andres Wildberger: Razvoj krasa in jam v Alpah v lui paleoklime in paleotopograje V lanku predstavimo nova spoznanja o razvoju alpskih jam. Ta temeljijo na sintezi novih dognanj o procesih speleogeneze in rezultatih intenzivnih terenskih raziskav v Alpah v zadnjih desetletjih. Razvoj freatinih delov jamskih sistemov poteka v bliini freatine povrine, ki je vezana na poloaj izvirov, ti pa so vezani na dno alpskih dolin. Torej je razvoj jam neposredno vezan na geomorfoloki razvoj terena in poglabljanje dolin. Jamski sedimenti nosijo informacijo o zaporedju morfolokih in klimatskih dogodkov. e ve, doloanje starosti jam in pote ka razvoja povrja, je mono edino z datacijo jamskih sedimen tov. Razvoj jam se zane ob emerziji apnenca in vzpostavitvi hidravlinega gradienta. Mezocojski in zgodnje terciarni pa leokras v obmoju Alp so dokaz starih emerzij. Hidrotermalni kras je oitno bolj razirjen, kot so domnevali v preteklosti. Te jame so bile pozneje preoblikovane z meteorno vodo, ki je za brisala sledi zgodnjega hipogenega zakrasevanja. Ledeniki zavi

PAGE 2

TIME in KARST 2007 54 presumed. is is mostly due to the fact that usually, hydro thermal caves are later reused (and reshaped) by meteoric wa ters. Rock-ghost weathering is described as a new cave genesis agent. On the contrary, glaciers hinder cave genesis processes and ll caves. ey mainly inuence cave genesis indirectly by valley deepening and abrasion of the caprock. All present dat ings suggest that many alpine caves (excluding paleokarst) are of Pliocene or even Miocene age. Progress in dating methods (mainly the recent evolution with cosmogenic nuclides) should permit, in the near future, to date not only Pleistocene, but also Pliocene cave sediments absolutely. Key Words: Karst, Cave genesis, Alps, Glaciations, Messinian event, Paleoclimate, Paleotopography. rajo procese speleogeneze in zapolnjujejo jame. Na razvoj jam vplivajo posredno, preko poglabljanja dolin in bruenja povrja. Noveje datacije kaejo, da so tevilne jame v Alpah pliocenske ali celo miocenske starosti. Nove datacijske metode predvsem metoda kozmogenih nuklidov bodo omogoile absolutno datacijo sedimentov do pliocenske starosti. Kljune besede: kras, geneza jam, Alpe, poledenitve, mesinska stopnja. I NTRODUCTION Progress in cave exploration and cave genesis studies (Audra 1994, Jeannin 1996, Palmer 2000) permitted to recognize the potential of caves for the study of land scape evolution, valley deepening and thus erosion rates and climate changes (Huselmann et al. 2002; Bini et al. 1997). Most of the information that is sheltered within the caves morphology and sediments is no more avail able at the surface, mainly due to the intensive erosion, especially during the glaciations. is article gives information about cave genesis and its potential for the reconstruction of the evolution and timing of the landscape: Part I presents the latest results concerning cave genesis and their link with the landscape. Part II deals with new concepts about early cave genesis, including pre-existing karst systems (paleo karst), hydrothermal karst, and pseudokarst. Many caves are older than the glaciations and glaciers generally are rather hindering cave genesis processes. Part III conse quently presents evidences supporting a high age of many cave systems. In Part IV, ages obtained by dierent dating methods prove that karst genesis in the Alps started far before the Q uaternary, as far as the Cretaceous. S ETTING e Alpine belt extends from Nice (France) to Vienna (Austria) into seven countries (France, Switzerland, Italy, Liechtenstein, Austria, Germany and Slovenia). Karsts and caves are found in each country, the largest karst ar eas being located in periphery (Fig. 1). All massifs are dissected by deep ly entrenched valleys which divide continuous structures into dierent physiographic units. Annual pre cipitation range from 1500 to more than 3000 mm. e French W estern Prealps consist of folded and thrusted mas sifs of mainly Cretaceous rocks. e elevation ranges generally between Fig. 1: M ap of the alpine karsts (dark color) with location of the mentioned massifs (karst areas aer: B uzio & Faverjon 1996; M ihevc 1998; Stummer & Pavuza 2001; Wildberger & Preiswerk 1997. M ap: D Cardis). P HILIPPE A UDRA A LFREDO B INI F RANCI G ABROV EK P HILIPP H USELMANN F ABIEN H OBL A P IERRE YVES J EANNIN ...

PAGE 3

TIME in KARST 2007 55 1000 and 2000 m. e Vercors displays a landscape of ridges and valleys, whereas the Chartreuse presents a steep, inverted relief. e Central Swiss Alps harbors the highest alpine karst areas at Jungfrau (3470 m ASL). e Siebenhengste (2000 m ASL) and the Hlloch-Silberen (2450 m ASL) consist of nappes of Cretaceous and Eocene rocks. e Italian Southern Alps are located to the south of the Insubric Line. e carbonate rocks range in age from Carboniferous to Cretaceous-Eocene. ey are deformed and displaced by S-vergent thrusting and large scale fold ing. e elevation ranges from 200 m to 2400 m ASL. e Northern Calcareous Alps in Austria are com posed of a slightly folded succession of Trias limestones and dolomites with a thickness of more than 1000 m. Large plateaus extend from 1800 to 2200 m ASL. In the Slovenian Alps, the Julian and the Kamnik Alps correspond to the roots of the Austrian nappes. us the landscape is oen similar, with plateaus and narrow steep ridges dominated by high peaks reaching more than 2800 m ASL. G ENERAL CONCEPTS OF CAVE GENESIS e basics of cave genesis are beyond the scope of this paper. e reader can refer to the most comprehen sive and up-to-date work Speleogenesis: Evolution of Karst Aquifers (Klimchouk et al. 2000). G ENESIS OF CAVES AND MORPHOLOG Y OF PASSAGES RELATED TO W ATER TABLE POSITION W ater owing into limestone corrodes and erodes the rock. Driven by gravity and geological structure, it ows down more or less vertically, until it reaches either the karst water table or impermeable strata. en it continues owing more or less horizontally towards the spring, col lecting water from other lateral passages. W ater owing in the vadose (unsaturated) zone can only erode the oor of a gallery creating a meandering canyon. On the other hand, water owing within the phreatic (saturated) zone corrodes over its whole cross-section, giving a rounded cross-section (Fig. 2). e morphologies that are pre served once the watercourses have been abandoned give information about the prevailing position of the phreatic zone during the genesis of the galleries. R ECOGNITION OF CAVE GENESIS PHASES AND RELATION TO THE SPRING W ithin the saturated zone, two geometric types of con duits prevail (Ford 1977, 2000): 1) the water table caves, represented by horizontal conduits located at the top of the saturated zone; 2) the looping caves, represented by vertically lowering and rising conduits, whose amplitude may reach as much as 300 m, or even more. A phase of cave genesis corresponds to the net work of active conduits related to a given (paleo)spring. As springs move together with valley bottoms, we usually nd many dierent phases of cave genesis in a given karst region. As described on gure 2 the transition between phreatic conduits (elliptical shape) and vadose ones occurs at the top of the epiphreatic zone, i.e. more or less at the top level reached by water during highwater stages. Due to headlosses, highwater level is inclined to wards the outlet of the system, namely the karst spring (Jeannin 2001, Huselmann 2003). Most of the time conduits are located within a given range of altitudes (sometimes more than 300 m) below the (inclined) wa ter table limit. ese conduits go up and down (hence their name: loops) within this range and towards the spring. Sometimes main conduits of a given phase can be followed for kilometers and display a phreatic morphology all along. Sometimes the highest passages clearly show vadose entrenchment because they were located higher than the top of the epiphreatic zone, at least most of the time. Reality is a little more complicated that exposed here (see Huselmann et al. 2003 for instance), but the principle is the same. e main exceptions to this model, linking quite directly the phases of cave genesis to the (paleo)spring positions, i.e. valley bottom, occur when Fig. 2: An undulating phreatic tube is co-fed by a vadose meandering canyon, whose shape turns into a tube below the oodwater table. e arrow marks the transition from vadose to phreatic. C AVE AND K ARST EVOLUTION IN THE A LPS AND THEIR RELATION TO PALEOCLIMATE AND PALEOTOPOGRAPH Y

PAGE 4

TIME in KARST 2007 56 impervious barriers dam water somewhere inside the aquifer. S UCCESSION OF CAVE GENESIS PHASES CAVE LEVELS RECORDING BASE LEVEL CHANGES If the spring lowers gradually, the cave system behind also adapts gradually by entrenchment to the new situ ation: no distinct phases exist. If the spring lowers in a stepwise manner, followed by a time of relative stability, the owpath readjustment in the cave also occurs rap idly and a new cave genesis phase develops. Calculations show that, once a proto-conduit has been formed, caves may evolve very rapidly, in the order of 10 years, to reach penetrable size (Palmer 2000). erefore, aer a new entrenchment of a valley, pre-existing or newly cre ated soutirages (Huselmann et al. 2003) allow for the wa ter to reach the spring level quite quickly and a new water table, i.e. phase of cave genesis is created (Fig. 3). Former conduits, perched in the vadose zone aer the deepen ing of the karst system, are abandoned and remain dry (fossil passages). Provided that the cave genesis phases reect the deepening of the valleys through time, they give information for the reconstruction of paleorelief. Equivalent information at the surface is usually no longer present, mostly due to river or glacier erosion. In some cases, the base level may rise again aer a period of deepening (e.g. post-messinian inlling of the overdeepened canyons in the southern part of the Alps; Felber & Bini 1997). is caused a ooding of pre-exist ing karst systems and a reactivation of previously vadose or abandoned passages (Tognini 2001). T HE RELATION BET W EEN MORPHOLOG Y, CLIMATE AND SEDIMENTS Cave morphology depends on the position of the epi phreatic water table. e size of the passage, however, depends (among others, mostly geological factors) on time and ow rate. W orthington (1991) puts forward that there is an equilibrium size of a phreatic passage for a given ow rate. Aer this size is reached, the pas sage hardly grows anymore, and a growth above this size is mainly dependent on an increase in ow rate, either by capturing another catchment, or related to an increase in precipitation. For example, in the Siebenhengste system, the size of the main conduits doubles between two phases (700 m and 660 m). is very probably corresponds to the capture of the Schrattenuh catchment, which sig nicantly increased the size of the catchment area (Fig. 4). Conversely, a reduction in the catchment area due to valley entrenchment produces rearrangement of the cave system. Newly formed passages will be smaller than in the previous phase. Beside the size of the conduits, sediments also pro vide direct information about the ow velocities, i.e. discharge rates in the conduits. Grainsize distribution of cave sediments and conduit size make it possible to as sess paleodischarge rates quite precisely. Fig. 3: Schematic ow system. B lack = main (epiphreatic) gallery; light grey = soutirages (downward) and upow (upward); dark grey = perennial phreatic conduit. Fig. 4: N-S-projection of B renschacht and St. B eatus Cave with the recognized phases. e numbers are the elevations ( in m ASL) of the corresponding spring. Phase 558 is the present one. P HILIPPE A UDRA A LFREDO B INI F RANCI G ABROV EK P HILIPP H USELMANN F ABIEN H OBL A P IERRE YVES J EANNIN ...

PAGE 5

TIME in KARST 2007 57 N E W CONCEPTS ABOUT CAVE GENESIS T HE INFLUENCE OF EARL Y PHASES : PRE E X ISTING KARST S Y STEMS PALEOKARST H Y POGENIC KARST AND PSEUDOKARST Synand post-sedimentary paleokarsts Paleokarst are features that are not related to any present water circulation and completely obstructed. Since most of the caves (including fossil tubes) are related to present rivers and valleys, they are not considered as paleokarst. Some paleokarsts have been formed during or im mediately aer the sedimentation of carbonate platforms (Upper Triassic, for example Calcare di Esino/Grigna; Dachsteinkalk/Northern Limestone Alps). Dolines, pock ets and red paleosoils interfere within the cyclic sedimen tation of the so-called loferitic succession. Under a pre mature diagenesis, dissolution and concretion produced evinosponges (Bini & Pellegrini 1998) and dolomite-lled fractures that contain iron oxides from paleosoils. In the Julian Alps, paleokarstic conduits have been lled with carbonate mud and later lithied, so that presently a paleoconduit is just a portion of somehow dierently col ored solid rock. Other paleokarst had been set up aer the emersion of the limestone strata. ey are fossilized by Jurassic sediments (Swiss Prealps, Julian Alps), Up per Cretaceous sandstone (Siebenhengste), Eocene sands (Vercors), or Miocene conglomerates (Chartreuse). ose paleokarsts features may form highly porous discontinuities that may have guided the placement of later cave systems. Hydrothermal caves related to tectonic build-up Some caves have a hydrothermal origin, which can be recognized aer their typical corrosional cupolas originating from convection cells and their sediments like large calcite spar (Audra et al. 2002a; Audra & Hof mann 2004; Bini & Pellegrini 1998; Sustersic 2001; W ild berger & Preiswerk 1997). ose hydrothermal upows are usually located near huge thrust and strike-slip faults. Such karstications created well connected cave systems which later had generally been re-used by normal me teoric water ow aer upli above the base level. Since this change has mostly deleted the marks of their origin, they are only conserved when rapidly fossilized. Pseudokarst creating rock-ghosts (cave phan toms) Models of apparent karst features created by pro cesses other than pure dissolution are called pseudo karst. e phantomisation (rock-ghost weathering) was recently described as a major agent of karstication in impure limestones (Vergari & Q uinif 1997). In such limestones ow remained guided by fractures but par tially occurred in the matrix around the fracture. In a favorable context, warm and humid climate and longterm stability of the base level, this type of ow could dissolve the limestone cement, but impurities remained in place, in place, preserving the parent material tex Fig. 5: Pseudoendokarst cave system in the marly-silicated M oltrasio Limestone of Mt. B isbino, Lake of Como ( T ognini 1999, 2001). 1 Late Oligocene-Early M iocene: e tectonic structure was achieved during the neo-alpine phase. Upli raised the area above sea level, producing a gentle relief dissected by valleys. 2-3 M iddle-Late M iocene: According to very long base level stability under warm and humid climate, deep soils develop. With very low gradient and water movements, weathering progressively penetrates deeply into the water-lled zone. Upli gradually deepens the valleys. 4 M essinian: V alleys dramatically entrench, water table lowers, inducing an active ow. e weathered rock-ghosts are eroded away by piping, causing the formation of cave systems, which extend progressively in size and complexity. Steep hydraulic gradients prevent a further weathering at depth. e present remnants of rock-ghosts mark the maximal depth (700 m) reached by weathering that corresponds to the present 500-600 m altitude. With a continuous entrenchment, pseudoendokarsts become perched and only classical cave system develop below. 5 Early-M iddle Pliocene: Pseudoendokarstic caves systems stopped developing. 6 Late Pliocene-Quaternary: Sequences of erosion and deposition are developing (e.g. lacustrine caves sediments recording the presence of the Adda glacier close to the caves entrances; speleothem deposition is enhanced during interglacials). C AVE AND K ARST EVOLUTION IN THE A LPS AND THEIR RELATION TO PALEOCLIMATE AND PALEOTOPOGRAPH Y

PAGE 6

TIME in KARST 2007 58 tures and structures. Rock porosity increased up to 35%, causing a dramatic increase in hydraulic conduc tivity. is weathered material is called rock-ghosts, or phantoms. e downstream part of such systems, close to the surface, can be eroded by piping because of the absence of cement. is may produce caves (Tognini 2001). Some peculiar features may point out their dif ferent origin (weathered walls, regularly spaced 3D net work, brisk change in passage morphology, dead-end at gallery terminations with conservation of the ghost of the weathered host-rock). Aer the piping event, the rock-ghosts remained perched on an unweathered rock, in which only classical karst processes adapted to the new base level began to be active. C OMPLE X RELATIONSHIPS TO GLACIERS Some older theories supported a direct relationship be tween glaciations and genesis of cave systems through glacial meltwater. However, recent datings (U/, paleo magnetism) and eldwork has clearly proven that many caves are older than the glaciations. e role of the gla ciers seems to be mostly limited to valley deepening, base level rising during glacial periods and related sedi mentation in the conduits (Audra 1994, 2004; Bini 1994; Huselmann 2002). e genesis of new caves only takes place in certain contexts, where the glacial inuence of ten is only indirect. Glacial processes mainly ll caves In the Alps, glaciers were temperate with owing water. As valley bottoms were lled by ice, base levels raised all along the valleys. Furthermore, tills obstructed the pre existing springs. erefore, a large glacier body may have raised karstwater level by several hundreds of meters, for instance 500 to 600 m in the Bergerhhle/Tennengebirge (g. 6). Such a rising karstwater level reactivated many older conduits, increasing drastically ow cross-sections and leaving only restricted ow velocities in each con duit. Fine-grained carbonate-rich sediments found in very many caves are good indicators of these stages. Since this carbonate our could obviously not be dissolved by the natural aggressivity of the water, it implies that a chemical erosion of cave walls was very probably neg ligible. is is conrmed by old speleothems, preceding such phases, that are hardly dissolved (Bini et al. 1998). Mechanical abrasion in the ooded zone is also improb able because of the small ow velocity. erefore, it must be postulated that the genesis of deep-seated cave con duits is not favored by glaciations (Audra 1994, 2001a; Bini et al. 1998; Maire 1990). In contrast, interglacials induce the presence of vegetation and soil at the surface. Both elements greatly enhance the CO 2 content of the water (Bgli 1978), and reduce the amount of debris washed into the cave. So, water has a much higher initial acidity and can therefore enlarge caves (Audra 2004). During the same time, wa ter from the ne ssures and matrix, which entered the system below the soil and epikarst, where pCO2 is high, is oversaturated with respect to calcite when it reached a (ventilated) cave passage. erefore many speleothems formed. In some low valleys with at bottoms, lakes lled the previously overdeepened valley and kept the water table high. erefore, in spite of the sometimes consid erable valley deepening by glaciers, the karst water table could never reach the total depth of the valley, blocking thus the genesis of deeper cave levels (Kanin). Neverthe less, in the South Alpine domain, the uvial valley deep ening may have allowed deep (and today submerged) karstication. Fig. 6: e Cosa Nostra-B ergerhhle system/T ennengebirge, Salzburg Alps (Audra et al. 2002b). T o the le (3), relationship between cave passage altitude and old karst levels. Karst development began during the Oligocene beneath the Augensteine (1). D uring the M iocene, horizontal systems developed with alpine water inputs (2), showing dierent levels (3) related to successive phases of stability: Ruinenhhlen (4) and Riesenhhlen (e.g. Eisriesenwelt 5). Following Pliocene upli, alpine systems developed (e.g. Cosa NostraB ergerhhle 6). H orizontal tubes at the entrance correspond to a M iocene level (7). A sha series (6) connect to horizontal tubes from B ergerhhle-B ierloch (8), corresponding to a Pliocene base level (9). e present water table at 700 m (10) pours into B runnecker Cave, which connects to the Salzach base level (11). P HILIPPE A UDRA A LFREDO B INI F RANCI G ABROV EK P HILIPP H USELMANN F ABIEN H OBL A P IERRE YVES J EANNIN ...

PAGE 7

TIME in KARST 2007 59 Fr anc e S witzerland Ital y A ustria Slov enia T ab. 1: Synthesis of information about the quoted caves systems C AVE AND K ARST EVOLUTION IN THE A LPS AND THEIR RELATION TO PALEOCLIMATE AND PALEOTOPOGRAPH Y

PAGE 8

TIME in KARST 2007 60 As a conclusion, a warm climate induces passage growth and speleothem deposition, whereas a cold cli mate generally tends to obstruct the lower passages by sediments. Glacial sediments covering older speleothems: cave systems may predate glaciations Some cave sediments correspond to very old glaciations, according to paleomagnetic measurements that show inverse polarity: Ofenloch/Churrsten (Mller 1995), grotte Vallier/Vercors (Audra & Rochette 1993), Crnel sko brezno/Kanin (Audra 2000). ese sediments oen overlie successions of alterites or massive owstone de posits, which in turn prove the existence of a warm and humid climate, thus showing that the cave systems pre date those glaciations. Some of the old speleothems are more or less intensely corroded by owing water postdat ing their deposition. Cave development and glacial activity Glacial abrasion at the surface and erosion in the va dose zone. At the surface, the glacial activity is without doubt responsible for the abrasion of a variable amount of bedrock (50-250 m), which has surfaced old conduits that previously were deeply buried. is is manifested by wide open shas, cut galleries and arches. During gla cial melt, meltwater disappeared into distinct sectors. As soon as fractures were connected to preexisting con duits, they enlarged quickly and thus formed the inva sion vadose shas (Ford 1977), which can reach several hundred meters of depth: Granier, Silberen, Kanin (Ku naver 1983, 1996). e eectiveness of such meltwater is mainly due to its velocity in the vertical cascades as well as their abrasive mineral load originating from bedrock and till material. Some new cave systems appeared in the intra-Al pine karst area due to glacial erosion. in limestone belts or marbles intercalated with metamorphic series were freed from their impervious cover by glacial erosion. Some caves are still in direct relationship with the peri glacial ow, and act as swallowholes. eir morphology reects the cascading waterow and has a juvenile form: Perte du Grand Marchet/Vanoise, Sur Crap/Graubn den (W ildberger et al. 2001). At the Grotte ophile/ Grandes Rousses, U/ datings evidenced that the cave was active at least along the two glacial-interglacial cycles that are marked by the sequence of passage-forming/ll ing with gravel/sinter deposition (Audra & Q uinif 1997). Since cave development mainly occurred during inter glacial, the eect of the glacier is only indirect, by eroding the impervious covers (Audra 2004). e lower phases of huge cave systems are indi rectly generated by glacial valley-deepening. W hile the uppermost cave systems are oen older than the glacia tions (infra), the lower passages are oen of Q uaternary age, since they are related to valleys evidently deepened by glaciers. In this respect, glaciers are indirectly respon sible for the creation of new cave passages (Siebenheng ste, Chartreuse, Vercors). is strongly contrasts with the South Alpine domain, where valleys were deepened dur ing the Messinian event. Here, glaciations contributed merely to the inlling of the preexisting valleys. us, most of the South Alpine cave systems are thought to be older than the glaciations. M ORPHOLOGIC AND TOPOGRAPHIC EVIDENCES FOR A HIGH AGE OF CAVE S Y STEMS Some existing caves and karst features clearly correspond to a strongly dierent topography than today. ey are therefore supposed to be older. In the following para graphs, the position and morphology of caves are com pared to todays landscape. en cave sediment charac teristics are presented and discussed. In a third part, links between caves and well-recognized paleotopographies are explained. All those indications are clear evidences for a high age of cave systems. C AVE S Y STEMS VS PRESENT TOPOGRAPH Y Perched phreatic tubes Conduits with an elliptical morphology are sometimes perched considerably above the present base level (Tab. 1, 3 rd column). ey developed close to a paleo base level, long before todays valley deepening. At the Siebenheng ste, the highest phases even show a ow direction oppo site to the present one. Caves intersected by current topography Old perched caves are oen segmented by a subsequent lowering of the surface. Two situations are usually found in the eld: Old phreatic caves at the surface of karst plateaus, which have been eroded by glacial abrasion (Grigna, Do lomites, Triglav, Kanin, Tennengebirge) Old phreatic caves along valley anks, obviously cut by the lowering of the topography (Adda, Adige, Sal zach, Isre): Pian del Tivano, Mt. Bisbino, Mt. Tremez P HILIPPE A UDRA A LFREDO B INI F RANCI G ABROV EK P HILIPP H USELMANN F ABIEN H OBL A P IERRE YVES J EANNIN ...

PAGE 9

TIME in KARST 2007 61 zo, Campo dei Fiori (Southern Alps), Paganella (Dolo mites). Dimensions too large with respect to the present catch ment and climate e dimensions of some conduits are far too large com pared to the present catchment area, thus proving that the older catchment areas had been much larger, but are now truncated by erosion (Eisriesenwelt/Tennengebirge (g. 6); Antre de Vnus/Vercors; Snezna jama na Raduhi/ Kamnik Alps, caves at Pokljuka and Jelovica plateaus at Julian Alps, Siebenhengste, Pian del Tivano, Campo dei Fiori/Southern Alps). Spring location vs. present base level If the position of a spring is not due to a geologic perch ing above an impervious layer, it has to be close to base level (see part I). However, in some cases springs did not lowered down to todays base level. In other cases springs are obviously located far below the base level. is can be explained by the following hypotheses: Some springs are perched, because the valley inci sion is very recent and rapid (Pis del Pesio/Marguareis). Others are presently submerged below the base level and hidden by alluvial ll or till (Emergence du Tour/Ara vis; Campo dei Fiori). ey were set into their place before the base level raised and they continue to function due to the high transmissivity of the sediment ll. A specialty is given when old vertical vadose caves are suddenly stopped by the present water table, proving that the horizontal drains are at much greater depth and completely drowned. Typical vadose morphologies (spe leothems, karren) are known in some drowned conduits (Grotta Masera, Grotta di Fiumelatte/Lake of Como; Fontaine de Vaucluse/Provence). Here, the spring loca tion is adapted to the present base level, but the caves are proof that the base level may, in some cases, also rise. is is especially true for areas aected by the Messinian crisis (Bini 1994; Audra & al. 2004). C AVE SEDIMENTS SHO W ING EVIDENCE OF A REMOTE ORIGIN DIFFERENT CLIMATE AND OLD AGE (tab. 1) Old uvial material e presence of some caves sediments is inexplicable with the present waterpaths. Big rounded pebbles found in caves perched high up on top of clis mean that a valley bottom had to exist at this level. Aerwards, the valleys deepened so much that they are far below such perched massifs (Salzach/Salzburg Alps; Granier/Char treuse). Oen, gravels found in these caves have a petrog raphy and mineralogy that is not found in the present rocks. ey are issued either from caprock that has dis appeared a long time ago (Fontana Marella, Campo dei Fiori) or from distant catchments, as proven by uvial pebbles (Augensteine/Northern Limestone Alps in Aus tria), quartz sandstones (Slovenian Alps), uvioglacial sediments (Lake of Como). Dating of uvial pebbles by cosmogenic nuclides from the Grotta Masera (Como), yielded a probable age comprised between 2.6 to 7.2 Ma, showing a pliocene age, or maybe older (Huselmann unpub.; Bini & Zuccoli 2004). In the Granier system, this method yields ages comprised between 1.8 to 5.3 Ma (Hobla & Huselmann 2007). Record of climatic changes in subterranean sediments Oen, the analysis of the sediments evidences climate changes, with a change from biostatic conditions, marked by the rarity of allogenic sediments, towards rhexistatic conditions, with lots of allogenic sediments. ese sediments come from the erosion of soils in a con text of climate degradation and general cooling. ey usually are interpreted to reect the climatic change in the Pliocene, before the onset of the glaciations. Such sediments are present in most of the old cave phases, which therefore should be older than the end of the Pliocene: Grotte Vallier/Vercors; Tennengebirge (Audra 1994, 1995), Campo dei Fiori (Bini et al. 1997), Monte Bisbino (Tognini 1999, 2001). In the Dachstein-Mam muthhle, which dates back to the Tertiary and shows a phreatic tube perched 1000 m above the Traun val ley, owstones grown during the interglacials intern ger with a series of debris-ow conglomerates of glacial origin (Trimmel 1992). In the Grotta di Conturines/Do lomites (2775 m ASL), the mean annual temperatures deduced from the 18 O of speleothems were between 15 and 25, which implies that speleothemes deposited in a warmer climate within the Tertiary, probably also at a lower altitude than it is found today (Frisia et al. 1994). Furthermore, in many caves, either conduits or owstones have been deformed by late Alpine tectonic movements: Grotta Marelli, Grotta Frassino/Campo dei Fiori (Uggeri 1992; Bini et al. 1992, 1993). Dating results prove the antiquity of cave systems e calculated age of old speleothems are regularly above the U/ limits (700 ka, even 1.5 Ma according to the 234 U/ 238 U equilibrium (Bini e t al. 1997); Tab. 1). e pale omagnetic measurements oen show inverse magnetism, sometimes with multiple inversion sequences, proving of a very old age of the cave sediments (Audra 1996, 2000; Audra & Rochette 1993; Audra et al. 2002b). e use of the new cosmonucleide method to date old quartz sedi ments also conrms this trend and yield ages reaching C AVE AND K ARST EVOLUTION IN THE A LPS AND THEIR RELATION TO PALEOCLIMATE AND PALEOTOPOGRAPH Y

PAGE 10

TIME in KARST 2007 62 back to about 5 Ma (see the details for Siebenhengste ex ample in this volume). R ELATIONS TO AN OLD TOPOGRAPH Y e geomorphologic approach, which uses external markers of old base levels (paleovalleys, paleoshelves with associated sediments) that are well dated, oers pre cious possibilities for the dating of karst systems. Sadly, correlations are almost impossible up-to-date due to the scarcity of such information. In the northern ank of the Alps, the glaciations oen caused the remnants of an old topography to disappear. e southern Alps, less glaci ated and better studied in this context, oer more pos sibilities, also thanks to the presence of guiding events like the Messinian incision and the following Pliocene marine highstand. Old erosion surfaces e identication of old erosion surfaces is a precious tool in geomorphology. Large surfaces oen top the re lief and cut across very old caves that are dircult to link to an old drainage system because of their fragmented character. e cave systems developing below those high surfaces are more recent, such as the stacked surfaces in the Vercors, of Eocene, infra-Miocene and Pliocene age (Delannoy 1997). Shelves along slopes, created by lateral corrosion of the rim of ancient depressions, have the same signicance as perched valley bottoms. In Vercors, Pliocene caves could be associated on them, such as the Antre de Vnus and the Grotte Vallier (Delannoy 1997). In the area of Varese (Lombardy), the Oligo-Miocene surface that cuts across limestone, porphyritic rocks and granites, is dissected by the late Miocene valleys that had been deepened during the Messinian (Bini et al. 1978, 1994; Cita & Corselli 1990; Finckh 1978; Finckh et al. 1984). Morphological and sedimentological evidences of prepliocene paleovalleys A uvial drainage pattern of Oligo-Miocene age, incised in the relief, predated the Alpine tectonic events of the late Miocene. e drainage originated in the internal massifs, cut through the calcareous border chains, and ended in alluvial fans in the molasse basins. In the border chains, perched paleovalleys are found more than 1500 m above the present ones (Salzburg Alps), as well as u vial deposits coming from siliceous rocks (Augensteine/ Northern Calcareous Alps; siliceous sands/Julian Alps (Habic 1992)), sometimes buried in caves near the valley slopes (Grotta di Monte Fenera/Piemont, Grotta Fontana Marella/Campo dei Fiori). In the northern ank of the Alps, these valleys have been destroyed by the deepening of the hydrographic net work, aided by the action of the glaciers. In the South, the old valleys have been deepened by the Messinian incision and lled by Pliocene sediments (Lake of Como/Adda, Varese, Tessin, Adige, Durance). As a consequence, the horizontal karstic drains that were linked to the old val leys had been truncated by slope recession, and are pres ently perched (Grotta Battisti/Paganella; Grotte Vallier/ Vercors; Pian del Tivano, Monte Bisbino (Tognini 2001); Campo dei Fiori (Uggeri 1992)). e almost generally observed input of allogenic waters coming from imper meable rocks upstream, combined with a tropical humid climate with considerable oods, explains the giant di mensions of those caves. A GE OF A LPINE KARSTIFICATION : FROM PALEOKARSTS TO RECENT MOUNTAIN D Y NAMICS P ALEOKARST A MILESTONE FOR OLD KARSTS e study of paleokarsts is a separate domain. No cave system has survived in its integrality from the periods predating the Miocene. In the Northern Limestone Alps of Austria, the possibility that caves of the highest level (Ruinenhhlen) may be relicts of an oligocene karsti cation has been discussed (Frisch et al. 2002). However, Paleogene paleokarsts are frequent, as evidenced by nat ural or articial removal of their lling: In Siebenhengste, upper Cretaceous paleotubes and fractures are found in Lower Cretaceous limestone, lled with Upper Cretaceous Sandstone (Huselmann et al. 1999). In many places, (Switzerland, Vercors, Chartreuse) vast pockets covering a karst relief and lling up some conduits can be observed. In Southern Alps, upper Eocene and lower Oli gocene sediments have been found into large cavities inlled by basaltic intrusions (Covoli di Velo, Ponte di Veia/Monte Lessini) eir age could be determined by K/Ar datings (Rossi & Zorzin 1993). In several regions (Vercors and Chartreuse, Monte Lessini), karstication is more or less continuous from the Eocene onwards. However, the tectonic and paleo P HILIPPE A UDRA A LFREDO B INI F RANCI G ABROV EK P HILIPP H USELMANN F ABIEN H OBL A P IERRE YVES J EANNIN ...

PAGE 11

TIME in KARST 2007 63 geographic changes have only le dispersed paleokarsts. Since the Miocene on, several massifs emerged from the molasse basins, thus allowing a karstication that con tinues today. E STIMATION OF THE FIRST E X POSURE ACCORDING TO MOLASSE PETROGRAPH Y e main phase of karstication begins when suitable rocks are exposed at land surface. Since the oldest rem nants of karst are oen eroded, it is possible to calibrate the beginning of the karstication by the foreland sedi ments (mainly the Molasse), which contain limestone pebbles eroded away at the surface. However, absence of evidence is not evidence of absence: sedimentary gaps are frequent, and a karst in biostatic conditions does not spread detritic elements towards the foreland. As a gen eral rule, the Miocene molasse registered the beginning of the last big karstication phase, earlier in Italy, later in Switzerland: Upper Oligocene-Lower Miocene (30 to 20 Ma) in the Southern Molasse, based on dated uvial sediments located in paleovalleys (Gelati et al. 1988). Lower Miocene (20 Ma) in the molasse south of Grenoble, corresponding to the erosion of the emerged anticlines of the Vercors and Chartreuse (Delannoy 1997). Lower Miocene (20 Ma) in the Austrian Nord-Al pine molasse, corresponding to the erosion of the Augen steine cover, which is of Upper and Middle Oligocene age (Lemke 1984; Frisch et al. 2000). Upper Freshwater Molasse in the Eastern Swiss basin (Hrnli fan, Middle Miocene 17-11 Ma) which contains pebbles of the rst erosion of Helvetic nappes (Siebenhengste, Silberen, Speck 1953; Brgisser 1980). D ATING THE Y OUNGEST PHASES AND E X TRAPOLATION e most generally applied dating method for cave sedi ments is U/. It makes it possible to date speleothems. In best cases, it allows for going back to as far as 700 ka dating only the sediment contained within the cave and not the cave itself. e use of paleomagnetic dating makes it possible, in some scarce cases, to push back the datable range to 2.5 Ma. e use of cosmogenic isotopes (Grang er et al. 2001) is the only recent method that opens new possibilities, having a dating range between 300 ka and 5 Ma. Another solution consists in dating lower cave phas es that are supposed to be younger, and in progressively going up the phases towards the oldest cave systems, un til reaching the limits of the used methods. From the cal culated rate of valley deepening, one can then extrapolate the age of the uppermost phases. Of course, such an ap proach can only give a general idea about the age. e lowermost phases of the Siebenhengste cave sys tem, St. Beatus Cave and Brenschacht, have been dated by U/. e following ages have been obtained: Phase 558 (youngest) began at 39 ka (max. 114 ka) and is still active today; Phase 660 was active between 135 and 114 ka; Phase 700 was active between 180 and 135 ka; and Phase 760 started before 350 ka and ended at 235 ka (Fig. 4). ese age values indicate a general valley incision rate of 0.5 to 0.8 mm/a, with a tendency to slow down as the age gets higher. Extrapolation indicated an age of about 2.6 Ma for the oldest cave systems, at 1850 m ASL. Ab solute cosmogenic dating yielded an age of 4.4 Ma for the oldest sediment, contained in the second-highest cave phase at 1800 m, showing a slower entrenchment in the older phases (Huselmann & Granger 2005; see also this volume). Dating of the cave systems at Hlloch/Sil beren gave maximal rates of valley incision in the range of about 1.5 to 3.5 mm/a. R ELATIVE UPLIFT RATES AND EROSION VOLUMES IN FORELAND SEDIMENTS Upli rates are generally calculated for long periods of time, taking the average of variable rhythms and inte grating vast parts of the area, without taking into account block tectonics which can dier considerably from one massif to the other. In the same range, the estimated vol ume of the foreland basins only gives a global approach. Such results only may give a general frame for a valida tion. Modeling the ssion-track measurements of the Swiss Central Alps (Reuss valley) give an average upli of 0.55 mm/a (Kohl, oral comm. 2000) comparable to cal culations of recent upli (0.5 mm/a; Labhart 1992) and consistent with the rates inferred from dating in caves. Upli is maximal in the central parts of the mountain chains, therefore the rocks are more deeply eroded in this area. As a consequence, the oldest caves had to have dis appeared from the central zones, compared to the border chains where they are better preserved due to the slower erosion. C AVE AND K ARST EVOLUTION IN THE A LPS AND THEIR RELATION TO PALEOCLIMATE AND PALEOTOPOGRAPH Y

PAGE 12

TIME in KARST 2007 64 A CKNO W LEDGEMENTS PH, PYJ and MM acknowledge the Swiss National Sci ence Foundation for support of the Habkern W orkshop (Grant No. 21-62451.00) and for research support (Grant No. 2100-053990.98/1). C ONCLUSION e examples mentioned above are distributed through out the Alpine belt. erefore, the conclusions drawn here are valid for Alpine Caves at least, but they may be applied to other cave systems also. e main following conclusions can be drawn from the above synthesis: In contrast to some earlier views, caves are not directly linked to glaciations. On the contrary, there is evidence that during glaciations caves are mainly lled with sediments, while they are enlarged during the inter glacials. e main inuence of glaciers upon cave genesis is the deepening of the base level valley, thus inducing a new cave genesis phase to be formed. U/ datings, coupled with paleomagnetism, in ferred a Lower Pleistocene to Pliocene age for several cave sediments. Fossil or radiometric datings of solidied cave lls (sandstone, volcanic rocks) gave ages reaching back to the Upper Cretaceous. It follows that caves are not inherent to the Q uaternary period, but are created whenever karstiable rocks are exposed to weathering. Due to later inll, however, most explorable caves range from Miocene to present age. W e have shown that caves are related to their spring, which is controlled by a base level that usually consists of a valley bottom. So, the study of caves gives very valuable information about valley deepening pro cesses and therefore about landscape evolution. Caves constitute real archives, where sediments are preserved despite the openness of the system. e study of cave sediments gives information about paleo climates. Moreover, the combination of cave morphology and datable sediments allow to reconstruct the timing of both paleoclimatic changes as well as landscape evolu tion between the Tertiary and today. Dierential erosion rates and valley deepenings can be retraced. Information of this density and completeness has disappeared at the surface due to the erosion of the last glacial cycles and the present vegetation. Correlations between well-dated cave systems can signicantly contribute to the geodynamic understand ing of the Alpine belt as a whole. e location of most cave systems at the Alpine border chains is very lucky: since they are dependent on base level (in the foreland), recharge and topography (towards the central Alps). ey inevitably registered changes in both domains. Caves are therefore not only a tool of local importance, but may have a wide regional/interregional signicance. e dating method by cosmogenic nuclides was recently applied in some French, Italian and Swiss alpine cave systems which partially contain pre-glacial uvial deposits. e dated sediments yielded ages ranging be tween 0.18 and 5 Ma, which are consistent with other approaches. Advances in modern dating techniques (cosmogenic isotopes, U/Pb in speleothems) therefore open a huge eld of investigations that will very signi cantly contribute to the reconstruction of paleoclimates and topography evolution along the last 5, possibly 15 to 20 Ma. e messinian event inuenced cave genesis over the whole southern and western sides of the Alps by overdeepening valleys. However, the subsequent base level rising ooded those deep systems creating huge deep phreatic aquifers and vauclusian springs (Audra et al. 2004). R EFERENCES e reader will nd a complete bibliography, compiled by the authors of this paper, in: Huselmann, P. & Monbaron, M., Editors (2001): Cave Genesis in the Alpine Belt. Rapports de recherche, Institut de Gographie Universit de Fribourg, 156 p. Audra, P. 1994: Karsts alpins, gense de grands rseaux souterrains. Exemples: le T ennengebirge (Autriche), lIle de Crmieu, la Chartreuse et le V ercors (France). PhD esis, University of Grenoble. Karstologia Mmoires 5, 280 p. P HILIPPE A UDRA A LFREDO B INI F RANCI G ABROV EK P HILIPP H USELMANN F ABIEN H OBL A P IERRE YVES J EANNIN ...

PAGE 13

TIME in KARST 2007 65 Audra, P. 1995: Signication des remplissages des karsts de montagne.Karstologia, 25, 13-20. Audra, P. 1996: Lapport de ltude des remplissages la connaissance de la karstogense: le cas du chourum du Goutourier (massif du Dvoluy, Hautes-Alpes).Revue danalyse spatiale quantitative et applique, 38, 109-120. Nice. Audra, P. 2000: Le karst haut alpin du Kanin (Alpes ju liennes, Slovnie-Italie). tat des connaissances et donnes rcentes sur le fonctionnement actuel et lvolution plio-quaternaire des structures karst iques.Karstologia, 35, 27-38. Audra, P. 2001a: Feichtner cave (Kitzsteinhorn, Salzburg, Austria), a deep cave system developing into calcar eous schists in a glacial environment.Acta Carso logica, 30, 2, 165-174. Ljubljana. Audra, P. 2001b: French Alps karst: study methods and recent advances. In: Cave Genesis in the Alpine B elt (Eds. Ph. Huselmann, M. Monbaron).Rapports de recherche, Institut de Gographie, Universit de Fribourg, 7-28. Audra, P. 2004: Kitzsteinhorn high-alpine karst (Salz burg, Austria): Evidence of non-glacial speleogen esis.Die Hhle, 55, 1-4, 12-18. Audra, P. & Hofmann B. A. 2004: Les cavits hypognes associes aux dpts de sulfures mtalliques (MVT).Le Grotte dItalia. Table-ronde internatio nale, 2-5 septembre 2004, Valsassina, Italie. Com munit Montana della Valsassina & Association franaise de karstologie, 5, 35-56. Jeso. Audra, P. & Rochette, P. 1993: Premires traces de gla ciations du Plistocne infrieur dans le massif des Alpes. Datation par palomagntisme de remplis sages la grotte Vallier (Vercors, Isre, France).Compte-rendu lAcadmie des Sciences, S. 2, 11, 1403-1409, Paris. Audra, P. & Q uinif, Y 1997: Une cavit de haute-mon tagne originale: la grotte ophile (Alpe dHuez, France). Rle des paloclimats plistocnes dans la splogense.Splochronos, 8, 23-32, Mons. Audra, Ph., Bigot, J.-Y & Mocochain, L. 2002a: Hypo genic caves in Provence (France). Specic features and sediments.Acta Carsologica, 31, 3, 33-49, Lju bljana. Audra, Ph., Q uinif,Y & Rochette, P. 2002b: e genesis of the Tennengebirge karst and caves (Salzburg, Austria).Journal of Cave and Karst Science, 64, 3, 153-164. Audra, Ph., Mocochain, L., Camus, H., Gilli, ., Clauzon, G. & Bigot, J.-Y 2004. e eect of the Messinian Deep Stage on karst development around the French Mediterranean.Geodinamica Acta, 17, 6, 27-38. Bini, A. 1994: Rapports entre la karstication primdi terranenne et la crise de salinit du Messinien: lexemple du karst lombard (Italie).Karstologia, 23, 33-53. Bini, A., Cita, M. B. & Gaetani, M. 1978: Southern al pine lakes: hypothesis of an erosional origin related to the Messinian entrenchment.Marine Geology, 27, 271-288. Bini, A., Q uinif, Y ., Sules, O. & Uggeri, A. 1992: Les mou vements tectoniques rcents dans les grottes du Monte Campo dei Fiori (Lombardie, Italie).Kar stologia, 19, 23-30. Bini, A., Rigamonti, I. & Uggeri, A. 1993: Evidenze di tettonica recente nell area Monte Campo dei Fiori Lago di Varese.Il Q uaternario, 6, 1, 3-14. Bini, A., Breviglieri, P., Felber, M., Ferliga, C., Ghezzi, E., Tabacco, I. & Uggeri A. 1994: Il problema dellorigine delle valli. I depositi Plio-Q uaternari e levoluzione del territorio varesino.Guida alle escursioni Ri unione autunnale Gruppo Nazionale Geograa Fisica e Geomorfologia CNR, Varese, 100-149. Bini, A., Uggeri, A. & Q uinif, Y 1997: Datazioni U/ ef fettuate in grotte delle Alpi (1986-1997). Considera zioni sullevoluzione del carsismo e del paleoclima.Geologia Insubrica, 2, 1, 31-58. Bini, A. & Pellegrini, A. 1998: Il carsimo del M oncodeno. Geologia Insubrica, 3, 2, 296 p. Bini, A., Tognini, P. & Zuccoli, L. 1998: Rapport entre karst et glaciers durant les glaciations dans les val les pralpines du sud des Alpes.Karstologia 32, 7-26. Bini, A. & Zuccoli, L. 2004: Le problme de la grotte Masera (2213 LO), Lecco, Italie.14e Rencontre doctobre, Florac, 12-17. Splo-club de Paris Bgli, A. 1978: Karsthydrographie und physische Spelolo gie. Springer Verlag, Berlin, 292 p. Bosak, P., Hercman, H., Mihevc, A. & Pruner, P. 2002: High-resolution magnetostratigraphy of speleo thems from Snena jama.Acta Carsologica, 31, 3, 15-32. Ljubljana Brgisser, H.M. 1980: Z ur mittel-mioznen Sedimenta tion im nordalpinen M olassebecken: D as Appenzel lergranit-Leitniveau des H rnli-Schuttfchers (Obe re Ssswassermolasse, Nordostschweiz). Mitt. Geol. Inst. ETH + Univ. Zrich, N.F. 232, 196 p. Buzio, A. & Faverjon, M. 1996: Grottes et splologie en Italie.Spelunca, 61, 31-50. Ciszewski, A. & Recielski, Kr. 2001: Caves of the Kitzs teinhorn.Polish Caving 1997-2001, Caving Com mission of Polish Mountaineering Association, Kra kow, 22-24. C AVE AND K ARST EVOLUTION IN THE A LPS AND THEIR RELATION TO PALEOCLIMATE AND PALEOTOPOGRAPH Y

PAGE 14

TIME in KARST 2007 66 Cita, M.B. & Corselli, C. 1990: Messinian paleogeogra phy and erosional surfaces in Italy: an overview.Palaeogeography, Palaeoclimatology, Palaeoecol ogy, 77, 67-82. Delannoy, J.-J. 1997: Recherches gomorphologiques sur les massifs karstiques du V ercors et de la T ransversale de Ronda (Andalousie). Les apports morphogniques du karst. se d tat, Institut de gographie alpine, Grenoble, 678 p. Fantoni, E. & Fantoni, R. 1991: Geologia del monte Fen era: ipotesi sulla genesi del sistema carsico.De Valle Sicida 2, 1, 11-22. Felber, M. & Bini, A. 1997: Seismic survey in alpine and prealpine valleys of Ticino (Switzerland): evidences of a Late Tertiary uvial origin.Geologia Insubrica, 2 (Southern Alps Q uaternary Geology, IGCP 378 Meeting, Lugano, ottobre 1995), 46-47. Finckh, P. G. 1978: Are Southern Alpine lakes former Messinian canyons? Geophysical evidence for preglacial erosion in the Southern Alpine lakes.Ma rine Geol., 27, 289-302. Finckh, P. G., Kelts, K. & Lambert, A. 1984: Seismic stra tigraphy and bedrock forms in perialpine lakes.Bull. Geol. Soc. Am., 95, 1118-1128. Ford, D. C. 1977. Genetic classication of solutional cave systems.Proceedings of the 7 th International Con gress of Speleology, Shereld, International Union of Speleology & British Cave Research Association, Bridgwater, 189-192. Ford, D. C. 2000: Speleogenesis under unconned set tings.Speleogenesis: Evolution of Karst Aquifers (Eds. Klimchouk et al. ), National Speleological So ciety, Huntsville, 319-324. Frisch, W ., Szkely, B., Kuhlemann, J. & Dunkl, I. 2000: Geomorphological evolution of the Eastern Alps in response to Miocene tectonics.Zeitschr. fr Geo morphologie N. F., 44, 103-138. Frisch, H., Kuhlemann, J., Dunkl, I., Szkely, B., Vennemann, T. & Rettenbacher, A. 2002: DachsteinAltche, Augenstein-Formation und Hhlenent wicklung die Geschichte der lezten 35 Millionen Jahre in den zentralen Nrdlichen Kalkalpen.Die Hhle, 53, 1, 1-36. Frisia, S., Bini, A. & Q uinif, Y 1994: Morphologic, crys tallographic and isotopic study of an ancient ow stone (Grotta di Cunturines, Dolomites) Implica tions for paleoenvironmental reconstructions.Spe leochronos, 5, 3-18. Mons Gelati, R., Napolitano, A., Valdisturlo, A. 1988: La Gonfo lite lombarda: stratigraa e signicato nellevoluzione del margine sudalpino.Riv. Ital. Pal. Strat., 94, 285332. Granger, D.E., Fabel, D. & Palmer, A.N. 2001: PliocenePleistocene incision of the Green River, Kentucky, determined from radioactive decay of cosmogenic 26 Al and 10 Be in Mammoth Cave sediments.GSA Bulletin, 113, 7, 825-836. Habic, P. 1992: Les phnomnes palokarstiques du karst alpin et dinarique en Slovnie.Karst et volutions climatiques. Hommage Jean Nicod (Eds. J.-N. Salomon, R. Maire), Presses universitaires de Bor deaux, 411-428. Huselmann, P. 2002: Cave Genesis and its relationship with surface processes: Investigations in the Sieben hengste region (BE, Switzerland) .PhD thesis, Uni versity of Fribourg, Switzerland. Hhlenforschung in der Region Siebenhengste-Hohgant 6, 168 p. Huselmann, Ph. & Granger, D. E. 2005: Dating of caves by cosmogenic nucleides: Method, possibilities, and the Siebenhengste example (Switzerland).Acta Carsologica, 34, 1, 43-50. Ljubljana Huselmann, Ph., Jeannin, P.-Y & Bitterli, T. 1999: Rela tions between karst and tectonics: the case-study of the cave system north of Lake un (Bern, Switzer land).Geodinamica Acta, 12, 377-388. Huselmann, Ph., Jeannin, P.-Y ., Monbaron, M. & Laurit zen, S. E. 2002: Reconstruction of Alpine Cenozoic paleorelief through the analysis of Caves at Sieben hengste (BE, Switzerland).Geodinamica Acta, 15, 261-276. Huselmann, Ph., Jeannin, P.-Y & Monbaron, M. 2003: Role of epiphreatic ow and soutirages in conduit morphogenesis: the Brenschacht example (BE, Switzerland).Zeitschri fr Geomorphologie N.F., 47, 2, 171-190. Hobla, F. 1999: Contribution la connaissance et la gestion environnementale des gosystmes karstiques montagnards: tudes savoyardes. PhD esis, Uni versity of Lyon 2, 995 p. Hobla, F. & Huselmann, Ph. 2007: Dating of the Gra nier system by cosmogenic nucleides, Chartreuse, France.is volume Jeannin, P.-Y 1996: Structure et comportement hydrau lique des aquifres karstiques. PhD thesis, Univer sity of Neuchtel. 248 p. Klimchouk, A., Ford, D.C., Palmer, A.N. & Dreybrodt, W (Editors) 2000: Speleogenesis: Evolution of Karst Aquifers.National Speleological Society, Hunts ville, 528 p. Kunaver, J. 1983: Geomorfoloki razvoj Kaninskega pogorja s posebnim ozirom na glaciokrake pojave (Geomorphological development of Kanin Mts with special regard to glaciokarstic phenomena).Geo grafski zbornik, 22, 197-346. P HILIPPE A UDRA A LFREDO B INI F RANCI G ABROV EK P HILIPP H USELMANN F ABIEN H OBL A P IERRE YVES J EANNIN ...

PAGE 15

TIME in KARST 2007 67 Kunaver, J. 1996: On the location factor of the caves in upper Soca valley, with special regard to Kanin Mts.Alpine Caves: Alpine Karst Systems and their Environmental Context. Proceedings of the Inter national Congress, Asiago, Italy, 275-282. Labhart, T.P. 1992: Geologie der Schweiz. Ott-Verlag, un, 211 p. Lemke, K. 1984: Geologische Vorgnge in den Alpen ab Obereozn im Spiegel vor allem der deutschen Mo lasse.Geologische Rundschau, 73, 1, 371-398. Maire, R. 1990: La haute montagne calcaire. se dEtat, University of Nice, Karstologia Mmoires 3, 731 p. Mihevc, A. 1998: Krako povrje.Geografski atlas Slo venije (Ed. D. Perko), Ljubljana, 90-91. Mller, B.U. 1995: Die Hhlensedimente des Ofenlochs.Stalactite 45, 1, 25-35. Palmer, A.N. 2000: Digital modeling of individual solu tional conduits. Speleogenesis: Evolution of Karst Aquifers (Eds. Klimchouk et al. ), NSS, Huntsville, 194-200. Rossi, G. & Zorzin, R. 1993: Nuovi dati sui fenomeni pa leocarsici dei Covoli di Velo (Monti Lessini VR).Le Grotte dItalia, X VI (Atti X VI Congr. Naz. Spe leol., Udine 1990), 169-174. Speck, J. 1953: Gerllstudien in der Subalpinen M olasse am Z ugersee und V ersuch einer palogeographischen Auswertung. Kalt-Zehender, Zug, 175 p. Stummer, G. & Pavuza, R. 2001: Karstverbreitungskarten sterreichs.Speleo-Austria (Eds. Geyer et al. ), Verein fr Hhlenkunde in Obersteir (VHO), Bad Mitterndorf, p. 45. uteri, Fr. 2001: Some characteristics of Alpine karst in Slovenia.Cave Genesis in the Alpine Belt (Eds. Ph. Huselmann, M. Monbaron), Rapports de re cherche, Institut de Gographie Universit de Fri bourg, 125-140. Tognini, P. 1999: Individuazione di un nuovo processo speleogenetico: il carsismo del M. B isbino (Lago di Como). PhD esis, Universit degli Studi di Mi lano. 433 p. Tognini, P. 2001: Lombard southalpine karst: main fea tures and evolution related to tectonic, palaeogeo graphic and palaeoclimatic regional history two examples of a global approach.Cave Genesis in the Alpine Belt (Eds. Ph. Huselmann, M. Monbaron), Rapports de recherche, Institut de Gographie Uni versit de Fribourg, 81-114. Trimmel, H. 1961: Hhlenausfllung, Hhlenent-wick lung und die Frage der Hhlenbildungszyklen.Memoria V della Rassegna Speleologica Italiana (Symposium Internazionale di Speleologia, Varenna 1960), 49-65. Trimmel, H. 1992: Dveloppement des grottes des Alpes orientales au Plistocne et lHolocne.Karst et volutions climatiques. Hommage Jean Nicod (Eds. J.-N. Salomon, R. Maire), Presses universita ires de Bordeaux, 285-292. Uggeri, S. 1992: Analisi geologico ambientale di un mas siccio carbonatico prealpino (M. Campo dei Fiori, Varese): geologia, geologia del Q uaternario, idro geologia.PhD esis, Universit di Milano, 153 p. Vergari, A. & Q uinif, Y 1997: Les palokarst du Hainaut (Belgique).Geodinamica Acta, 10, 175-187. W ildberger, A. & Preiswerk, Ch. 1997: Karst und H hlen der Schweiz/Karst et grottes de Suisse/Carso e grotte della Svizzera/Karst and Caves of Switzerland. Spe leo Projects, Basel, 208 p. W ildberger, A., W eidmann, Y ., Pulfer, T., Adank, M. & Strebel, R. 2001: Ruosna el Plaun Cumin (Laax), eine Ponorhhle im Bndner Oberland.Actes du 11e Congrs national de splologie, 231-234. W orthington, S.R.H. 1991: Karst hydrogeology of the Ca nadian Rocky M ountains.PhD thesis, McMaster University, 227 p. Zanalda, E. 1994: Dinaromys bogdanovi (Mammalia: Rodentia) from the middle Pleistocene of western Lombardy (Italy).Riv. It. Paleont. Strat., 100, 143148 C AVE AND K ARST EVOLUTION IN THE A LPS AND THEIR RELATION TO PALEOCLIMATE AND PALEOTOPOGRAPH Y



PAGE 1

A SPECTS OF THE EVOLUTION OF AN IMPORTANT GEO ECOS Y STEM IN THE L ESSINIAN M OUNTAIN V ENETIAN P REALPS I TAL Y P OGLEDI NA RAZVOJ POMEMBNEGA GEO EKOSISTEMA V GORAH L ESSINI B ENEKE P REDALPE I TALIJA Leonardo L ATELLA 1 & Ugo S AURO 2 Izvleek UDK 551.442:574.4 (234.323.4) Leonardo Latella & Ugo Sauro: Pogledi na razvoj pomembne ga geo-ekosistema v gorah Lessini (Beneke Predalpe, Italija) Jama Grotta dellArena (476 V/VR) v gorah Lessini, 1512 m n.m., je zelo pomemben podzemeljski kraki sistem. feprav je dolga le 74 m, vsebuje geoloke, geomorfoloke in okoljske elemente, znailne za krako podzemlje Visokih Lessini. Grot ta dellArena ima nekaj geolokih in favnistinih znailnosti skupnih z drugimi pomembnimi in znanimi krakimi sistemi. Jama je med tistimi z najvejim tevilom troglobiontskih vrst v vseh Benekih Predalpah, od katerih nekatere verjetno izvi rajo izpred kvartarja. Z geolokega vidika predstavlja jama kon taktni kras, kjer so vzdol stratigrafskega in tektonskega stika razlini apnenci. Grotta dellArena je na stratigrafskem stiku med apnenci Calcari del Gruppo di San Vigilio in Rosso Ammonitico in je zelo blizu prelomne ploskve, vzdol katere se vertikalno stikata omenjeni formaciji s formacijo Bianco ne, to je vrsta drobnoplastovitega in gosto prepokanega, slabo odpornega apnenca. Zanimiva je prisotnost precejnjega tevila terciarnih oziroma sploneje predkvartarnih vrst. To je verjetno v zvezi z jamsko geologijo. V prispevku so podrobneje obravna vane razline vrste podzemskih krakih sistemov v sami jami Grottta dellArena kot tudi v gorah Lessini in tudi njihovi odno si z razporeditvijo jamskega ivalstva. Kljune besede: razvoj krasa, geomorfologija, biospeleologija, invazija favne, Beneke Predalpe, Italija. 1 Museo Civico di Storia Naturale di Verona. Lungadige Porta Vittoria, 9, 37129 Verona, Italy. E-mail: leonardo.latella@comune.verona.it 2 Universit degli Studi di Padova, Dipartimento di Geograa. Via del Santo 26, 35123 Padova, Italy. E-mail: ugo.sauro@unipd.it Received/Prejeto: 21.12.2006 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 69 Abstract UDC 551.442:574.4 (234.323.4) Leonardo Latella & Ugo Sauro: Aspects of the evolution of an important geo-ecosystem in the Lessinian Mountain (Venetian Prealps, Italy) e Grotta dellArena (476 V/VR), located in the Lessinian Mountain, at the elevation of 1512 m a.s.l., is a very important underground karst system. Although it is only 74 m long, sev eral of the geological, geomorphological and environmental features of the High Lessinian underground karst are present in this cave. e Grotta dellArena shares some common geological and faunistic characters with other important and well known karst systems. is cave has also one of the highest number of troglobitic species in all Venetian Prealps and some of them possibly originated in the pre-Q uaternary. From the geological point of view the cave is the expression of a contact karst, where dierent limestone types come in contact both stratigraphically and along tectonic structures. e Grotta dellArena is located at the stratigraphic contact between the Calcari del Gruppo di San Vigilio and the Rosso Ammonitico and it is very close to a fault plane putting in vertical contact the two above forma tions with the Biancone, a kind of limestone closely stratied and densely fractured, very sensible to frost weathering. It is interesting to note the presence of a good number of species of Tertiary, or more generally pre-Q uaternary, originate in the Grotta dellArena. is presence is possibly related to the geol ogy of caves. In this paper the dierent kinds of underground karst systems in the Grottta dellArena and Lessinian Mountain, are analyzed and the relation with the cave fauna distribution are taken in consideration. Key words: karst evolution, geomorphology, biospeleology, faunistic invasions, Venetian Prealps, Italy.

PAGE 2

TIME in KARST 2007 70 I NTRODUCTION e Grotta dell Arena is registered with the number 476 in the Cadastre of the Caves of Veneto Region (the cave has been surveyed by A. Pasa in 1942, and GAS USV in 1972); the karst area is ML03 (Monti Lessini 03). e cave is 74 m long with a dierence in elevation of 22 m. It is located in the Lessinian Mountain district of Bosco Chiesanuova, in Malga Bagorno area. G.C: 11 06 02 E 45 39 56 N, elevation 1512 m a.s.l. e Grotta dellArena is a signicant kind of under ground karst system in Lessinian Mountain in fact: it is a type of speleogenetic style in the morphodynamic context of the High Lessinians, several of the geological, geomorphological and environmental features of the High Lessinian under ground karst are present in this cave and played a signi cant role in karst evolution, some of the best known karst systems in the Les sinian Mountain (Mietto & Sauro, 2000; Rossi & Sauro, 1977), such as the Abisso della Preta, the Covolo di Cam posilvano, the Abisso del Giacinto, the Abisso dei Lesi, the Ponte di Veja, share some common characters with the Grotta dellArena, from the biospeleological point of view, this cave has one of the highest number of troglobitic species in all Venetian Prealps, several troglobitic species are endemic for the Grotta dellArena or the Lessinian Mountains and some of them possibly originated in the pre-Q uaternary. e Grotta dellArena is a large chamber, roughly el liptical in plane section, with a main diameter of about 50 m. e roof coincides mostly with bedding planes. e southern part of the oor is characterized by a large, asymmetrical, funnel-shaped depression, a kind of sub terranean doline developed in the collapse debris. e chamber is situated a few meters below the top ographical surface; it is connected to the surface through Gastropoda Opiliones Diplopoda Orthoptera Zospeum sp. Ischyropsalis strandi Lessinosoma paolettii Troglophilus sp. Anellida Copepoda Collembola Coleoptera Marionina n.sp. Speocyclops cfr. infernus Onychiurus hauseri Orotrechus vicentinus juccii Araneae Lessinocamptus caoduroi Pseudosinella concii Orotrechus pominii Troglohyphantes sp. Moraria n. sp. Sincarida Italaphaenops dimaioi Pseudoscorpiones Elaphoidella n. sp. Bathynella sp. Lessinodytes pivai Chthonius lessiniensis Isopoda Amphipoda Laemostenus schreibersi Neobisium torrei Androniscus degener Niphargus galvagnii similis Halberrria zorzii Balkanoroncus boldorii T ab. 1: List of the cave-dwelling species in the Grotta dellArena. Fig. 1: e collapse depression called Arena. Fig. 2: e large chamber in the Arena cave. In the foreground the debris blocks, in the background the inner doline. L EONARDO L ATELLA & U GO S AURO

PAGE 3

TIME in KARST 2007 71 some narrow passages which start from an open collapse depression located on a slope, which resembles a Ro man theatre (i.e. an Arena, hence the name of the cave) (Fig.1, Fig. 2). e depression is the result of the collapse of part of the subterranean room. To understand the signicance of this cave it is nec essary to: delineate the geological, geomorphological, and, in general, environmental characteristics of this cave, reconstruct the framework of the spatial and tem poral evolution of the High Lessini karst, taking into account the climatic and environmen tal changes of the external environment surrounding the cave that occurred during the Pleistocene. analyse the phyilogeographical and taxonomical arnities of the troglobitic elements of its fauna. T HE ENVIRONMENTAL CONTE X T e Grotta dellArena had been previously dened not as a distinct structure, but as a window on a subterranean space, that allows us to see only some features of a karst system (Castiglioni & Sauro, 2002). In fact, the subter ranean environment is a much more complex system, mostly hidden to the human perception. From the geological point of view the cave is expres sion of a contact karst, where dierent limestone types come in contact both stratigraphically and along tectonic structures (Capello et al. 1954; Pasa, 1954; Sauro, 1973, 1974, 2001). In particular, the limestone formations pres ent here are: Calcari del Gruppo di San Vigilio of lowermiddle Jurassic, about 60 m in depth, pure both oolitic and bio-sparitic/ruditic, or reef limestones, relatively densely fractured, Rosso Ammonitico, a condensed rock unit of middleupper Jurassic age, about 30 m in depth, made up by nodular micritic limestone very resistant to ero sion, crossed by widely spaced fractures, Biancone, a chalk type unit, from the lower and middle Cretaceous, 100-200 meters in depth, made up by whitish marly limestone closely stratied and densely fractured, very sensible to frost weathering. e Scaglia Rossa formation of the upper Creta ceous, and the Eocene limestone, which lie above the Bi ancone in the western and southern part of the plateau are not present in the studied area because they have been completely eroded. Below the Calcari del Gruppo di San Vigilio there is the formation Calcari Grigi di Noriglio, of lower Jurassic, which is about 300 m in depth and out crops in the slopes of the main valleys, a kind of uvio karstic canyons. e Grotta dellArena is located at the stratigraphic contact between the Calcari del Gruppo di San Vigilio and the Rosso Ammonitico and it is very close to a fault plane putting in vertical contact the two above forma tions with the Biancone (Fig. 3). e cover rocks of the cave are made mostly by the massive beds of lower Rosso Ammonitico, whereas the inner cave is mostly developed inside the Calcari del Gruppo di San Vigilio. At the topo graphical surface, the line of the normal fault runs along Fig. 3: Sketches of the Arena cave system: I Plan of the system; the grey corresponds to the B iancone rock unit. II V ertical model of the karst system. Legend: 1) B iancone formation, 2) Rosso Ammonitico, 3) Calcari del Gruppo di San V igilio and Calcari Grigi, 4) debris pipe in the cave, 5) bedding plane karst zone at the contact Rosso AmmoniticoCalcari del Gruppo di San V igilio, 6) fault plane karst zone, a) at the B iancone side, b) at the Rosso Ammonitico side, 7) lateral ow inside and from the B iancone aquifer, 8) vertical karst ow. A SPECTS OF THE EVOLUTION OF AN IMPORTANT GEO ECOS Y STEM IN THE L ESSINIAN M OUNTAIN

PAGE 4

TIME in KARST 2007 72 a small valley, a few meters to the east of the cave; the displacement of the fault is about 100 m. From the geomorphological point of view, Biancone is dissected by a network of dry valleys, whereas Rosso Ammonitico generates a rocky landscape with large at karren separated by corridors, or rock cities of large blocks. From the hydrological viewpoint, the water circu lates diusely inside the dense network of discontinuities of the Biancone unit; the preferential ows is sub-parallel to the topographical surface and occurs mostly below the dry valley bottoms, but is also inuenced by the structural setting; vertical losses occur along the fault and fracture zones. In contrast, water circulation is more concentrated and mostly vertical in the Rosso Ammonitico. S PATIAL AND TEMPORAL EVOLUTION OF THE KARST S Y STEM It is easy to understand that the Grotta dellArena results of dierent spatial and temporal processes which oc curred as a consequence of several predisposing factors. In fact, the cave is at the same time, an example of litho logical contact karst, of intra-stratal karst, of fault zone karst and of a subterranean hydrological transition from a dispersed and sub-horizontal water ow to a more con centrated and sub-vertical one. e Grotta dellArena system is fed by a lateral water ow coming from the Biancone aquifer and crossing the fault zone, facilitated by the westward dipping of the stra ta. e speleogenesis of the cave has taken place in the lithological, tectonic and hydrological transition zone. Each cave we visit represents a moment of a long history, it is like the picture of a movie. Surely the present aspect of this cave and of its collapsed part are the result of relatively recent processes, occurred mostly during the middle and upper Pleistocene. But the karst system of which the cave is expression has surely begun to develop much earlier. Some caves, located in middle of the Lessinian pla teau and in the Berici hills, are the result of the re-activa tion of old paleokarstic nets developed during the Paleo gene (Rossi & Zorzin, 1989, 1991; Dal Molin et al. 2000); other caves with llings from the early middle Pleisto cene developed mostly during the lower Pleistocene. e Grotta dellArena chamber seems to be related with the second group. e fault to the east side of the cave is a paleotec tonic feature of Jurassic age, reactivated during the Cre taceous and later by the Alpine orogenesis during the Paleogene and the Neogene. e area where the cave is located probably emerged from the sea during the Oli gocene, as the southern part of the Lessinian plateau. e erosion of the Eocene rock unit occurred during late Pa leogene and early Neogene. e Scaglia Rossa formation was probably eroded during middle to late Neogene. At the beginning of the Q uaternary these two formations disappeared completely in the area (remnants of Scaglia are still present in the western High Lessinian). A model showing the sequence of landscapes devel oped in the dierent rocks by the erosion can be created, based on present-day landscapes of other parts of the Lessini Mountains, where the eroded geological forma tions are still present. us in the southwestern Lessinian Mountain (High Valpolicella) there is an active hydro graphic network with gorges entrenched in both Eocene Limestones and in the Scaglia Rossa. Here, the early morphogenesis, aer the emersion and the upli, has been mostly of the uvial type, marked by the development of a network of valleys strongly con trolled by the tectonic structure. So, a valley developed along the fault line. Following the incision of the Scaglia Rossa, the karst process begun to aect the fault zone. But, it is especially aer the erosion of the Scaglia Rossa that the aquifer hosted in the Biancone started to feed a new underground karst system located near to the fault zone of which the Grotta dellArena is the present day ex pression. From this simple model it is possible to infer that the evolution of the underground karst system started since Neogene, probably since middleupperMiocene. e transition from the uvial environment to the karst environment has been accompanied by the development of a uviokarstic milieu in the Biancone. In this milieu, which is still present, there is not surface runo except during exceptional events, but there is a diuse circula tion inside the rock, for some aspects similar to that oc curring below the river beds, inside the alluvial deposits (Fig. 4). L EONARDO L ATELLA & U GO S AURO

PAGE 5

TIME in KARST 2007 73 T HE CLIMATE AND ENVIRONMENTAL CHANGES DURING THE P LEISTOCENE e Lessinian Mountain plateau was aected by the cli matic and environmental changes of the Pleistocene. In the cave area there is no evidence of past development of local glaciers (the nearest local glacier was more than 1 km to the northwest). However, traces of strong perigla cial processes, such as remnants of small rock glaciers, nivation niches, etc. are present (Sauro, 2002). During the last W rm sporadic permafrost was present in the area. e material resulting from the collapse of the Are na depression has been aected by cryoclastic processes, as shown by a large soliuction lobe located to the north side of the same hollow. e climate and enviromental change occurred in the Pleistocene, aected the colonization of the subterra nean environment by some actual troglobitic species and shaped the distribution of the species that colonized this environments before the Pleistocene. Fig. 4: Sketch of the morphological evolution of the alti Lessini according with the erosional stages reached by the relief (progressive erosion of the rock units). T HE CAVE FAUNA AT PRESENT e cave fauna of the Grotta dellArena is characterized by the presence of high number of troglobitic and en demic species (Caoduro & Ruo, 1998). Colonization of the cave by the troglobitic elements occurred in dier ent times. Ancient elements of this fauna colonized the subterranean environments before the Pleistocene, and other species invaded the cave in dierent periods along the Q uaternary. A SPECTS OF THE EVOLUTION OF AN IMPORTANT GEO ECOS Y STEM IN THE L ESSINIAN M OUNTAIN

PAGE 6

TIME in KARST 2007 74 Today, this cave has a high number of cave-adapted animals. Of the 43 taxa known for the cave, 24 could be considered eucavernicolous species (sensu Ruo, 1955: eutroglophiles+troglobites). e specialization index (eutroglobites/ eucavernicolous), has a value of 0.91, this means that 91% of the cave species in the Grotta dellArena are troglobionts. F INAL R EMARKS e subterranean karst of the alti Lessini is much more spatially developed than what is perceived by a speleolo gist. It consists not only in large pits and chambers but in a network of smaller cavities and ssures. In the two horizontal dimensions it is a kind of net, even if aniso tropic, better developed along the fault zones and some bedding planes. In the vertical dimension the anisotropy is even greater, and the thickness overpasses one thou sand of meters. In the time dimension, this karst network has evolved progressively, even with dierent speeds inu enced by the changes of the morphostructural setting and of the external environment. e karst morphogen esis occurred as result of the co-occurrence of various favourable conditions. e hydro-geological condition of the alluvial de posits of the water courses of the early erosional stage, during middle Neogene, are no present here nowadays, but there are situation for some aspects similar both be low the valley bottoms of the Biancone and in the diuse net of karst ssures developed inside this rock unit. is diuse aquifer is in contact with the more typical karst aquifer of the limestone of the Jurassic rock units. Likewise, some of the larger karst pockets developed in the Eocene limestone, may have had some connections with the karst cavities in the Scaglia, and, along the main fault structures or volcanic structures, also with the karst voids in the Jurassic rock units. Here, sudden and sharp changes of conditions of the underground environments have not occurred dur ing the late Neogene and the Pleistocene. Even the abrupt climatic changes of the Pleistocene have had a limited inuence on the underground environments, according with the large thickness reached by it before the end of Neogene. It is interesting to note the presence of a good num ber of species of Terziary, or generally pre-Q uaternary, origin in the Grotta dellArena. e most important relict species are: B alkanoroncus boldorii (Beier, 1931), Lessino camptus caoduroi Stoch, 1997, Italaphaenops dimaioi Ghi dini, 1964 and Lessinodytes pivai Vigna Taglianti e Sciaky, 1988 (Casale & Vigna Taglianti, 1975; Vigna Taglianti & Sciaky, 1988; Gardini, 1991; Galassi pers. com.). e presence and distribution of these species inside the caves of Lessinia (particularly the terrestrial species) has been usually related to certain environmental char acteristic like temperature, humidity, air circulation etc. However, on the basis of the actual knowledge (Latella & Verdari, 2006), it appears that all these species are present in caves with a large range of temperatures, altitude and morphology. All these caves are developed inside, or in contact with, the Biancone or Scaglia (Cretaceous lime stone) formations. It is likely that the geomorphology of the cave plays an important role not only in shaping the historical distribution, but also the actual presence, of cave animals in Lessinian area. A CKNO W LEDGMENTS W e are grateful to Sandro Ruo for the helpful discus sions and for the reading of the manuscript. W e also thanks Augusto Vigna Taglianti for the informations re garding Trechinae, Diana Galassi for the informations on Copepoda and Beatrice Sambugar for the Anellida. anks to Cristina Bruno for the linguistic review. L EONARDO L ATELLA & U GO S AURO

PAGE 7

TIME in KARST 2007 75 R EFERENCES Caoduro, G. & Ruo, S., 1998: La Grotta dellArena, un biotopo di eccezionale interesse negli alti Lessini. La Lessinia ieri oggi domani: quaderno culturale 1998, 39-44. Casale, A. & Vigna TAglianti, A., 1976: Note su Itala phaenops dimaioi Ghidini (Coleoptera, Carabide). Bollettino del Museo Civico di Storia Naturale di Verona, 2 (1975): 293-314. Capello, C.F., Nangeroni, G., Pasa, A., Lippi Boncampi, C., Antonelli, C. & Malesani, E., 1954: Les phno mnes karstiques et lhydrologie souterraine dans certaines rgions de lltalie. Assoc. Intern. Hydrol., vol. 37, n. 2, pp. 408-437, gg. 5, Paris. Castiglioni, B. & Sauro, U., 2002: Paesaggi e geosistemi carsici: proposte metodologiche per una didattica dellambiente. In: Varotto M. & Zunica M. (a cura di) Scritti in ricordo di Giovanna Brunetta. Dipar timento di Geograa G. Morandini, Universit di Padova, 51-67. Dal Molin, L., Mietto, P. & Sauro, U., 2000: Considera zioni sul paleocarsismo terziario dei Monti Berici: la Grotta della Guerra a Lumignano (Longare Vicen za). Natura Vicentina 4, 33-48 (ISSN 1591-3791). Gardini, G., 1991: Pseudoscorpioni cavernicoli del Vene to (Arachnida). (Pseudoscorpioni dItalia XIX). Bollettino del Museo Civico di Storia Naturale di Verona, 15 (1988): 167-214. Latella, L. & Verdari, N., 2006: Biodiversity and bioge ography of Italian Alps and Prealps cave fauna. Ab stracts 18th International Symposium of Biospeleol ogy, Cluj-Napoca, Romania, 10-5 July 2006: 9-10. Mietto, P. & Sauro, U., (eds), 2000: Le Grotte del Veneto: paesaggi carsici e grotte del Veneto. Second edition, Regione del Veneto-La Graca Editrice, 480 pp. Pasa, A., 1954: Carsismo ed idrograa carsica del Gruppo del Monte Baldo e dei Lessini Veronesi. C.N.R., Cen tro Studi per la Geograa Fisica, Bologna, Ricerche sulla morfologia e idrograa carsica, n. 5, 150 pp. Rossi, G. & Sauro, U., 1977: LAbisso di Lesi: analisi mor fologica e ipotesi genetiche. Le Grotte dItalia, (4) 6 1976): 73-100. Rossi, G. & Zorzin, R., 1989: Fenomeni paleocarsici nei Lessinian Mountain Centrali Veronesi. La Lessinia ieri oggi domani: quaderno culturale 1989, 47-54. Rossi, G. & Zorzin, R., 1991: Nuovi dati sui fenomeni pa leocarsici dei Covoli di Velo (M.ti Lessini Verona). Atti X VI Congr. Naz. di Speleologia, Udine, 169174. Ruo, S., 1955: Le attuali conoscenze sulla fauna caver nicola della regione pugliese. Memorie di Biogeo graa adriatica, 3: 1-143. Sauro U., 2002: Q uando in Lessinia cera il grande gelo. Q uaderno Culturale La Lessinia ieri oggi domani 2002, 85-94. Sauro, U., 1973: Il Paesaggio degli alti Lessini. Studio geo morfologico. Museo Civ. di St. Nat. di Verona, Mem. f. s., 6, 161 pp. Sauro, U., 1974: Aspetti dellevoluzione carsica legata a particolari condizioni litologiche e tettoniche negli Alti Lessini. Boll. Soc. Geol. It., 93, 945-969. Sauro, U., 2001: Aspects of contact karst in the Venetian Fore-Alps. Acta Carsologica, Ljubljana, 30(2), 89102, 2001. Vigna Taglianti & Sciaky R., 1988: Il genere Lessinodytes Vigna Taglianti, 1982 (Coleoptera, Carabidae, Tre chinae). Fragmenta Entomologica, 20 (2): 159-180. A SPECTS OF THE EVOLUTION OF AN IMPORTANT GEO ECOS Y STEM IN THE L ESSINIAN M OUNTAIN



PAGE 1

W HAT DOES THE DISTRIBUTION OF ST Y GOBIOTIC COPEPODA CRUSTACEA TELL US ABOUT THEIR AGE? KAJ NAM POVE RAZIRJENOST STIGOBIONTSKIH CEPONObNIH RAKOV CRUSTACEA: COPEPODA O NJIHOVI STAROSTI? David C. CULVER 1 & Tanja PIPAN 2 Izvleek UDK 595.3-15 591.5:595.3 David C. Culver & Tanja Pipan: Kaj nam pove razirjenost stigobiontskih cepononih rakov (Crustacea: Copepoda) o nji hovi starosti? Ob predpostavki, da je nastajanje novih vrst posledica vikari ance, brez naknadne disperzije, se za ocenjevanje starosti ivalskih skupin pogosto uporablja geografska razirjenost sti gobiontov. Ob domnevi, da je disperzija merilo za doloanje sta rosti vrst, smo prouevali primernost podatkov o razirjenosti stigobiontskih kopepodov v Sloveniji. Na osnovi analize obsega naselitve in pogostosti naseljevanja znotraj obmoja smo v prispevku priloili seznam nekaterih vrst cepononih rakov, ki naj bi bile evolucijsko stareje. Telesna velikost ne doloa ob sega naselitve in pogostosti pojavljanja. Kljune besede: speleobiologija, Copepoda, stigobionti, dis perzijska biogeograja. 1 Department of Biology, American University, 4400 Massachusetts Ave., NW W ashington D.C., U.S.A.; e-mail: dculver@american.edu 2 Karst Research Institute ZRC-SAZU, Titov trg 2, SI-6230 Postojna, Slovenia; e-mail: pipan@zrc-sazu.si Received/Prejeto: 27.11.2006 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 87 Abstract UDC 595.3-15 591.5:595.3 David C. Culver & Tanja Pipan: What Does the Distribution of Stygobiotic Copepoda (Crustacea) Tell Us About eir Age? Geographic distribution of stygobionts is oen used to estimate age of a group by assuming vicariant speciation with little or no subsequent dispersal. W e investigated the utility of using distri butional data for Slovenian stygobiotic copepods by assuming that dispersal is a way to measure age of a species. W e list some species of Copepoda that, on the basis of their range and fre quency of occupancy within their range, should be older. Body size is not predictor either of range or frequency of occupancy. Key words: Speleobiology, Copepoda, stygobionts, dispersal biogeography. I NTRODUCTION e distribution of stygobionts has oen been used to infer the age of a fauna. e general procedure has been to assume that little or no migration has occurred, and that the extant distribution represents the site of original colonization and isolation in subterranean habitats. e vicariance biogeographic view, now dominant in modern biogeography (e.g., Crisci, Katinas, and Posadas 2003) largely supplanted the old idea of centers of origin with species dispersing out from this central place (Matthew 1915). Given the reduced opportunities for dispersal of subterranean animals, it is not surprising that there have been a number of studies that show a correspondence between ancient shorelines and current distributions, especially Tethyan and Paratethyan distributions (Culver and Pipan in press). In some cases, it has been possible to match distributions to historical events and to obtain support from molecular clock data (see Verovnik, Sket, and Trontelj 2004). However, not all subterranean dis

PAGE 2

TIME in KARST 2007 88 tributions can be explained solely by vicariance. A par ticularly interesting example is the cirolanid isopod An trolana lira Bowman. In general, subterranean cirolanids are found near to marine shores (Botosaneanu, Bruce, and Notenboom 1986), suggesting a marine ancestor with vicariant isolation. But, A. lira is found in caves in the Appalachian Valley of Virginia and more than 200 km from not only the present ocean shore, but from any ocean dating back at least to the Paleozoic (Holsinger, Hubbard, and Bowman 1994). In this contribution, we take a dispersalist rather a vicariance view of subterranean biogeography. W e con sider a model of colonization and isolation as follows. A species colonizes and is isolated in a subterranean site. As adaptation occurs, the species occupies more sites in the vicinity of the colonization. at is, the frequency of occupancy of subterranean sites increases. In the next stage, the species expands its range, with a high occupan cy of suitable sites in its range. Finally, as other species also evolve, the original species may be out-competed or it may become specialized in response to competition. In this scenario, it will then occupy a lower frequency of sites within its range. us, we can rank the ages of spe cies in increasing age as follows: 1. Species with small ranges and occupying few (sometimes only one) sites 2. Species with small ranges but occupying a high frequency of sites within its range 3. Species with large ranges and occupying a high frequency of sites within its range 4. Species with large ranges and occupying a low frequency of sites within its range. W e examine this hypothesis using distributional data of subterranean copepods from Slovenia (see Pipan 2005), and make assess the utility of this approach. M ETHODS AND M ATERIALS From information in Pipan (2005) and Culver, Pipan, and Schneider (in press) we generated list of stygobiotic copepods from seven Slovenian caves, with informa tion on ranges, frequency of occupancy of well-sampled caves, and average body size. Ranges were categorized into three groups: 1. Slovenian endemics 2. Balkan endemics 3. European endemics 4. Cosmopolitan species To measure frequency of occupancy, we used data from Pipan (2005), which was intensive enough that it is likely that most species were found (Pipan and Cul ver in press). Body sizes were taken from original species descriptions and direct measurement by TP. Data were available for 37 species. Analysis was done by grouping ranges into two cat egories (Smallcategories 1 and 2 and Largecategories 3 and 4), frequency of occupancy into two categories (Low to 3 caves and High to 7 caves), and size into two categories (Smallless than the overall mean of 0.61 and largegreater than or equal to the overall mean of 0.61). e resulting 2X2 contingency tables were analyzed for independence using Fishers Exact Test in JMP TM (Sall, Creighton, and Lehman 2005). R ESULTS Available data for the 37 species of stybobiotic copepods are shown in Table 1. In Table 2, all species are categorized into four groups based on range and occupancy. ere was no signicant dierence between observed and ex pected although there was a small excess of species with large ranges that were also found in a high frequency of caves. ose species hypothesized to be the oldest (large ranges, low occupancy) were: Acanthocyclops kieferi Acanthocyclops venustus stammeri D iacyclops clandestinus D icyclops languidoides Elaphoidella elaphoides M orariopsis scotenophila e group hypothesized to be the next oldest are those with large ranges and high occupancy: Elaphoidella jeanneli B ryocamptus balcanicus Acanthocyclops venustus Parastenocaris nolli alpina D AVID C. CULVER & T ANJA PIPAN

PAGE 3

TIME in KARST 2007 89 T ab. 1: Stygobiotic copepod species found in seven well-sampled caves in Slovenia. See Pipan (2005) and Culver et al. (in press). Species Name/Taxonomic Authority Mean Body Size No. Caves Occupied Range Acanthocyclops kieferi (Chappuis, 1925) 0.73 2 3 Acanthocyclops venustus (Norman & Scott, 1906) 1.07 1 3 Acanthocyclops venustus stammeri (Kiefer, 1930) 1.07 5 3 Bryocamptus balcanicus (Kiefer 1933) 0.40 4 3 Bryocamptus borus Karanovic & Bobic, 1998 1 2 Bryocamptus pyrenaicus (Chappuis, 1923) 0.80 7 3 Bryocamptus sp. 2 1 cf. Stygepactophanes sp. 0.35 3 1 Diacyclops charon (Kiefer, 1931) 1.00 7 2 Diacyclops clandestinus (Kiefer, 1926) 0.40 3 4 Diacyclops hypogeus (Kiefer, 1930) 0.50 2 1 Diacyclops languidoides (Lilljeborg, 1901) 0.80 3 4 Diacyclops slovenicus (Petkovski, 1954) 0.68 3 1 Echinocamptus georgevitchi (Chappuis, 1924) 0.70 1 2 Elaphoidella cvetkae Petkovski, 1983 0.75 4 2 Elaphoidella elaphoides (Chappuis, 1924) 0.60 1 3 Elaphoidella franci Petkovski, 1983 0.64 1 1 Elaphoidella jeannelli Chappuis, 1928 0.60 4 3 Elaphoidellakarstica Dussart & Defaye (1990) 1 1 Elaphoidella sp. A 2 1 Elaphoidella sp. B 2 1 Elaphoidella stammeri Chappuis, 1936 0.62 4 1 Maraenobiotus cf. brucei 0.60 1 1 Metacyclops postojnae Brancelj, 1990 >0.61 1 2 Moraria sp. A 2 1 Moraria sp. B 1 1 Moraria stankovitchi Chappuis, 1924 0.55 2 2 Morariopsis dumonti Brancelj, 2000 0.39 2 1 Morariopsis scotenophila (Kiefer 1930) 0.49 3 4 Nitocrella sp. 0.50 2 1 Parastenocaris cf. andreji 0.40 2 1 Parastenocaris nolli alpina (Kiefer, 1938) 0.42 5 3 Parastenocaris sp. A 0.40 2 1 Parastenocaris sp. B 0.40 4 1 Parastenocaris sp. C 0.40 2 1 Speocyclops infernus (Kiefer 1930) 0.47 6 2 Troglodiaptomus sketi Petkovski, 1978 0.88 3 2 W e investigated whether there was a body size bias for occupancy or range size. Smaller copepods might be able to disperse more easily but they may also be more subject to the vagaries of water movement in epikarst (Pi pan and Culver 2006). In any case, there was no relation ship between frequency of occupancy and body size and no relationship between range and body size (Table 3). W HAT DOES THE DISTRIBUTION OF ST Y GOBIOTIC COPEPODA CRUSTACEA TELL US ABOUT THEIR AGE?

PAGE 4

TIME in KARST 2007 90 T ab. 2: Number of species of stygobiotic copepods in categories of large and small range and high and low frequency of site occupancy. Numbers in parentheses are the expected numbers. Observed and expected numbers do not signicantly dier (p=0.21, Fishers Exact T est). High Occupancy Low Occupancy Large Range 4 (2.4) 5 (6.6) Small Range 6 (7.6) 22 (20.4) T ab. 3: Number of species of stygobiotic copepods in categories of high and low frequency of site occupancy (A), large and small range (B) and body size. Numbers in parentheses are the expected numbers. Neither association was statistically signicant (p=0.71 for A, p=0.71 for B Fishers Exact T est). A. High Occupancy Low Occupancy Large Body Size 5 (4.3) 8 (8.7) Small Body Size 5 (6.7) 12 (11.3) B. Large Range Small Range Large Body Size 5 (4.5) 8 (8.5) Small Body Size 5 (5.5) 11 (10.5) Finally, we investigated the taxonomic position of the putative older species, i.e., those with larger ranges. Of the ten species listed above, ve are cyclopoids and ve are harpacticoids. ere is an excess of large ranged cyclopoids but the dierence was only signicant at p~0.10 (Table 4). Acanthocyclops is especially notewor thy. All three stygobiotic species (A. kieferi, A. venustus, and A. venustus stammeri) had large ranges. In contrast none of the three species of Moraria (M. stankovitchi, sp. A, and sp. B) have large ranges. e lone calanoid species ( T roglodiaptomus sketi) also has a small range. T ab. 4: Relationship between range and taxonomic group (Cyclopoida vs. H arpacticoida). Expected numbers are given in parentheses. e relationship was marginally signicant (p~0.10, Fishers Exact test). Large Range Small Range Cyclopoida 5 (2.8) 5 (7.2) Harpacticoida 5 (7.2) 21 (18.8) D ISCUSSION W e have created a list of copepod species that, according to the hypothesis outlined in the introduction, should be older than other stygobiotic copepod species discussed in this study. Unfortunately, we know of no detailed phy logeny that would allow for such a comparison but we think that it would make for a very interesting study to do so. W hat is known about copepod phylogeny is that the Cyclopoida seem to be a more recent group than the Harpacticoida, accoding to the phylogeny of Huys and Boxshall (1991). e fact that cyclopoids are over-repre sented among species with large ranges (Table 4) contra dicts the hypothesis put forward. Of course, just because cyclopoids as a group are younger does not mean that the species are all younger than harpacticoids. Alternatively, it may be that harpacticoids are in general being outcom peted by cyclopoids, and this has resulted, not only in reduction in occupancy frequency, but also in range con traction. W e think that examination of the kinds of distribu tion patterns (range size and occupancy) discussed here will yield interesting results. is analysis would enrich phylogeography studies as well as provide additional hy potheses about the origin and evolution of subterranean groups. A CKNO W LEDGEMENTS e authors were supported by funds from the Center for Subterranean Biodiversity of the Karst W aters Insti tute and the Ministry of Higher Education, Science, and Technology of the Republic of Slovenia. D AVID C. CULVER & T ANJA PIPAN

PAGE 5

TIME in KARST 2007 91 R EFERENCES Botosaneanu, L., N. Bruce, & J. Notenboom., 1986: Isopoda: Cirolanidae, pp. 412-421, in L. Botosaneanu [ed.] Stygofauna M undi. E.J. Brill, Leiden, e Nether lands. Crisci, J.V., L. Katinas, & P. Posadas., 2003: H istorical B io geography. An Introduction. p. 250, Harvard Univ. Press, Cambridge. Culver, D.C., T. Pipan., & K. Schneider., in press: Vicari ance, dispersal, and scale in the aquatic subterra nean fauna of karst regions. Freshwater B iology Culver, D.C., & T. Pipan., in press: Subterranean ecosys tems. In S.A. Levin [ed.] Encyclopedia of B iodiver sity, second edition. Elsevier, Amsterdam. Holsinger, J. R., D. A. Hubbard, Jr & T. E. Bowman., 1994: Biogeographic and ecological implicationd of newly discovered populations of the stygobiont iso pod crustacean Antrolana lira Bowman (Cirolani dae). J ournal of Natural H istory 28, 1047-1058. Huys, R., and G. Boxshall., 1991: Copepod evolution. p. 468, e Ray Society, London, Matthew, W .D., 1915: Climate and evolution. Annals of the New Y ork Academy of Sciences 24, 171-318. Pipan, T., 2005: Epikarst a Promising H abitat. 100 p. Karst Researach Institute at ZRC-SAZU, ZRC Pub lishing, Postojna. Pipan, T., & D.C. Culver., 2006: Copepod distribution as an indicator of epikarst system connectivity. H ydro geology J ournal Pipan, T., & D.C. Culver., in press: Regional species rich ness in an obligate subterranean dwelling fauna epikarst. J ournal of B iogeography. Sall, J., L. Creighton, & A. Lehman., 2005: JMP Start Sta tistics. Brook/Coleomson Learning, Belmont, California. Verovnik, R., B. Sket, & P. Trontelj., 2004: Phylogeog raphy of subterranean and surface populations of water lice Asellus aquaticus (Crustacea: Isopoda). M olecular Ecology 13, 1519-1532. W HAT DOES THE DISTRIBUTION OF ST Y GOBIOTIC COPEPODA CRUSTACEA TELL US ABOUT THEIR AGE?



PAGE 1

H O W TO DATE NOTHING W ITH COSMOGENIC NUCLIDES K AKO DATIRATI PRAZNINE S KOZMOGENIMI NUKLIDI Philipp H USELMANN 1 Izvleek UDK 539.16:552.5(494) Philipp Huselmann: Kako datirati praznine s kozmogenimi nuklidi Jame predstavljajo praznino v kamninski masi in jim kot takim ne moremo doloiti starost. Zato z datiranjem jamskih sedimen tov sklepamo tudi o starosti jame, pri emer seveda ne moremo trditi, da je dobljena starost tudi prava starost jame. V lanku predstavimo metodo pri kateri z zdruitvijo sedimentarnih in morfolokih izsledkov sklepamo o relativni kronologiji dogod kov. Datiranje v oviru relativne kronologije lahko uporabimo za doloevanje starosti razlinih oblik, procesov in sedimentov. Dobljene rezultate pa lahko uporabimo kot pomembne mejnike v kronologiji, npr. pri intepretaciji klimatskih sprememb. Veliko jam je starejiih od zgornje meje starosti (350 do 700 ka), ki jo lahko doloimo z uran-torijevo metodo, ki je zelo razirjena. V zadnjem asu se zato uveljavlja metoda datacije s kozmogenimi nuklidi, ki omogoa datiranje dogodkov do starosti 5 Ma. Ker je teoretino ozadje te metode predstavljeno drugje (npr. Granger v tej tevilki), se tu omejimo le na uporabo metode v jamskem sistemu Siebenhengste (vica). Kljune besede: relativna kronologija, kozmogeni nuklidi, metodika datiranja jam, Siebenhengste. 1 Swiss Institute of Speleology and Karst studies SISKA, c.p. 818, 2301 La Chaux-de-Fonds, Switzerland, Fax 0041 32 913 3555, e-mail: praezis@speleo.ch Received/Prejeto: 11.12.2006 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 93 Abstract UDC 539.16:552.5(494) Philipp Huselmann: How to date nothing with cosmogenic nuclides A cave is a natural void in the rock. erefore, a cave in itself cannot be dated, and one has to resort to datable sediments to get ideas about the age of the void itself. e problem then is that it is never very certain that the obtained age really is coin cident with the true age of the cave. Here, we present the use of a method which couples sedimentary and morphologic infor mation to get a relative chronology of events. Datings within this relative chronology can be used for assessing ages of forms, processes, and sediments, and the obtained dates also x some milestones within the chronology, which then can be used to retrace, among other things, paleoclimatic variations. For many cave systems, the dating limits of the most widely used U/ method on speleothems are too low (350 to max. 700 ka) to get ages that inform us about the age of the cave. e recent use of cosmogenic nuclides on quartz-containing sediment permits to push the datable range back to 5 Ma. W hile the theoretical background is explained elsewhere (Granger, this volume), we concentrate on the Siebenhengste example (Switzerland). Key words: relative chronology, cosmogenic nuclides, cave dat ing methodology, Siebenhengste. I NTRODUCTION For many cave scientists, it might not be evident that a cave does not exist only the surrounding rock gives ex istence to the void called cave. erefore, a cave cannot be dated by conventional methods (Sasowsky 1998), but one has to use datable sediments. In karstic caves, the age of the surrounding rock gives a maximal age of the cave, while the sediments found within the cave give variable ages from today (in the case of still active speleothems) up to the last stages of speleogenesis (in the case of spe cic sand deposits dated by cosmogenic nuclides) and therefore to the age of nothing itself.

PAGE 2

TIME in KARST 2007 94 is paper contains two parts. In the rst part, the concept of relative chronology is explained. e link be tween morphology and sediment succession leads to a relative chronology of erosional and depositional events. Any dating of sediment with the purpose of studying the age of nothing basically requires such a relative chronol ogy, which places the obtained data into a timeframe. In the second part, the dating of sandy cave sedi ments with cosmogenic nuclides is briey presented. Es pecially when dealing with sands, a relative chronology is very important to date only meaningful sediments. e theoretical background is only very briey presented, and the reader is referred to Granger (this volume) for more thorough information. e Siebenhengste example, the use of the relative chronology, and the obtained results are presented in more detail. T HE CONCEPT OF RELATIVE CHRONOLOG Y I NTRODUCTION Geologists and other scientists are usually aware of the laws of stratigraphy, which say that a younger sediment overlies an older one. ese laws are the base of a relative chronology. is chronology is normally used to assess the correctness of an obtained age the numerical value has to be concordant with stratigraphy, or the dated age may not be correct. Most of the time, this principle is used with stalagmites, where the obtained ages must be older at the base and younger at the top (e.g. Sptl et al. 2002). Morphological indications, on the other hand, also give chronological information. A keyhole passage in forms us that a phreatic phase was followed by a vadose one. Successions of speleogenetic phases are found in many cave systems. W hile some of them indicate base level rises (Audra et al., 2004), most of them indicate a downcutting of the regional base level (upli, valley deepening, e.g. Ford & W illiams 1989; Rossi, Cortel & Arcenegui 1997). is in itself is also a chronological in formation: the oldest cave passages are on top, the young est ones near the present baselevel. e dirculty now is to connect the sediments of several, basically independent, sedimentary proles and to link them with the morphological succession of the cave passages. us, the sedimentary proles are not in dependent from each other, and a relative chronology of erosional and depositional events over the whole cave can be made. EX AMPLE Figure 1 shows a real situation encountered in St. Beatus Cave (Switzerland): To the right side is a typical keyhole passage which proves that a phreatic initiation of the ellipse on top was followed by a canyon incision. In the middle part of the gure, the meander gradually disappears and is replaced by a more or less elliptic passage that continues towards the le side of the gure. W e see therefore a transition of a vadose feature into a phreatic one, and thus an old wa ter level. In the prole to the right, we observe owstone deposition that was truncated by the river incising the meander. erefore, the owstone predates the canyon, but postdates the initial genesis of the elliptic passage to the right. e meander changes into an elliptic passage, thus the two forms are contemporaneous. Consequently, the older owstone disappears in the area of this transi tion. W ithin all the passages, silts were deposited. ey are younger than the meander incision, and younger than the passage to the le, and prove of an inundation of the whole cave. Stalagmites grow on the silts and are partially still active. is example can be written as a ta ble (Tab. 1). ----------------------------------------Phreatic genesis of top ellipse ----------------------------------------Water level lowering Deposition of owstone Erosion of owstone Erosion of meander ----------------------------------------Water level lowering Silt deposition Stalagmite growth ----------------------------------------T ab. 1: Chronology of erosional and depositional events (Fig. 1) is table is a rst relative chronology that links the sediments and the morphology of the cave.For practical reasons, the table presenting the chronology of events in a large cave system is not rewritten with each sedimen tary succession found. Instead, the single sedimentary sequence is coupled with morphology, and is written as a column in the table. e next sedimentary sequence, again coupled with morphology, is written as another column. us, the above example would then look like Table 2. P HILIPP H USELMANN

PAGE 3

TIME in KARST 2007 95 Sequence at left Sequence at right Phreatic genesis of top ellipse Water level lowering Phreatic genesis Deposition of owstone Phreatic genesis Erosion of owstone Phreatic genesis Erosion of meander Water level lowering Silt deposition Silt deposition Stalagmite growth Stalagmite growth T ab. 2: Chronological table with columnar writing of Fig. 1 EX PANSION If we continue upand downstream of that prole, we nd several other morphological indications and sedi mentary successions, each of them having a link with our initial prole until we encounter the next paleo-water level and thus the next morphological change. ere, the links have to be established again. e table thus slowly grows and gets more complete. Of course, the example presented above is an ideal case. Oen, the passages lack some information, thus making it dircult to establish an unambiguous chron ological table. Table 3 give an example: here, the upper passage lacks incision of a canyon. erefore, it is not clear whether the sediments found in the upper passage were all deposited while the lower passage was still in its initial genesis, or whether the sediments can be partly correlated. In this case, a relative correlation of the sedi ments by observation only is not possible: some absolute dates have to be obtained. Of course, these ages have to be in stratigraphic order of both the sediment succession and the morphologic indications. e above example had been dated by U/ on speleothems. e resulting table is presented in Table 4. Here, the speleothems with roughly the same age have been grouped together. en, laminated silt deposits that are thought to be a product of glacial damming (Bini, Tognini & Zuccoli 1998; Audra et al. this volume), are parallelized, inferring that the whole cave was ooded in such conditions. Of course, some un certainties still persist. Lower passage Upper passage Phreatic genesis Phreatic genesis W ater level lowering W ater level lowering Speleothem Speleothem Speleothem Speleothem Speleothem Speleothem Speleothem Pebble deposition Sand deposition Silt deposition Silt deposition Silt deposition Silt deposition Silt deposition Silt deposition Speleothem Erosion Erosion Sand deposition Silt deposition Pebble deposition Sand deposition Silt deposition Erosion Erosion Erosion ? ? ? T able 3: A more complicated example from St. Beatus Cave T ab. 3: A more complicated example from St. B eatus Cave Lower passage Upper passage Phreatic genesis Phreatic genesis W ater level lowering W ater level lowering Speleothem (>350 ka) Speleothem (337 ka) Speleothem (1 14 ka) Speleothem Speleothem (235 ka) Speleothem (180 ka) Speleothem (91 ka) Pebble deposition Sand deposition Silt deposition Silt deposition Silt deposition Silt deposition Silt deposition Silt deposition Speleothem (99 ka) Erosion Erosion Sand deposition Silt deposition Pebble deposition Sand deposition Silt deposition Erosion Erosion Erosion T able 4: The more complicated example, dated and expanded T ab. 4: e more complicated example, dated and expanded H O W TO DATE NOTHING W ITH COSMOGENIC NUCLIDES

PAGE 4

TIME in KARST 2007 96 WH Y A RELATIVE CHRONOLOG Y? e huge advantage of such a table of relative chronology is that it oers more control on the correct stratigraphic order than single sections, in ideal cases also the cave genesis can be dated, and last but not least, when having Fig. 1: Schematic section through a part of St. B eatus Cave (Switzerland), showing the relationship between sediments and morphology. parallelized all the sedimentary sequences, it is possible to make a synthetic and dated sediment prole of the whole cave, which can then be used to get information on climatic variations and the presence or absence of gla ciers damming the caves exit (Huselmann 2002). D ATING W ITH COSMOGENIC NUCLIDES I NTRODUCTION Cosmogenic nuclides are generated by the interaction of cosmic rays (mainly protons, neutrons, and muons) with atoms in the Earths atmosphere and lithosphere. e production rate of cosmogenic isotopes depends on the intensity of the cosmic rays, which is subject to change. e atmosphere then absorbs most of the primary rays and thus causes production rates to depend on elevation. Finally, the geometry of the sample location (and eventu al snow or soil cover) also has its eects. e radioactive nuclides most widely used for dating purposes are 10 Be and 26 Al produced in Q uartz. T HE PRINCIPLE AND POSSIBILITIES OF BURIAL DATING Burial dating of cave sediments is a relatively new tech nique that indicates the time sediment has been under ground (Granger, Fabel & Palmer 2001). It relays on the radioactive decay of the nuclides that were previ ously accumulated when the sediment was exposed at the surface. W hereas the intensity of the cosmic rays may vary with time, the ratio of produced 10 Be to 26 Al remains always approximately 1:7. e 10 Be/ 26 Al ratio can thus be calculated from the production rates and radioactive decay. If a sample that contains 10 Be and 26 Al is washed underground to surcient depth to be shielded from further radiation, the nuclide concen trations diminish. Since 26 Al has a half-life of 720 ka, opposed to the one of 10 Be of 1.34 Ma, the ratio of 1:7 is gradually lowered. Measurement of that ratio there fore gives a direct indication of the time the sample re mained underground. Of course, several prerequisites have to be fullled in order to get a burial age: First of all, the sediment must contain quartz that was irradiated surciently prior to burial. e grain size should be minimally ne sand (otherwise the cleaning process also eliminates the quartz), but may reach pebble size without problem. en, burial should ideally be 20-30 m below the surface to be surciently shielded from radiation. In order to make a measurement meaningful, the stratigraphic relationship of the sampled sand with the passage and other sediments should be clearly estab lished the relative chronology is needed. Burial dating has a range from about 100 years up to 5 Ma. Aer that time, the amount of remaining isotopes is usually too small to be measured accurately (Granger & Muzikar 2001). It is one of only a few radio metric methods that date lower Q uarternary and Plio cene deposits. It is of great interest for cave dating, rst because many old caves were created in the Pliocene or even earlier, and second because caves are very eective at shielding the sediment from further cosmic ray bom bardment. As with other cave-dating methods, burial dating may also be used to date the age of the passage, P HILIPP H USELMANN

PAGE 5

TIME in KARST 2007 97 thus indicating valley deepening rates and evolution of the surface outside the cave. T HE S IEBENHENGSTE E X AMPLE W e used burial dating to date the old passages of the Siebenhengste cave system in Switzerland. e Sieben hengste region is situated in the north-western part of the Alps, adjacent to the molasse basin. From Lake un, the mountain range extends to the Schrattenuh, 20 km away. e cave region is one of the longest and deep est worldwide, with the Rseau Siebenhengste-Hohgant having 154 km length and -1340 m depth. e caves comprise 14 dierent speleogenetic phases, which can be related to paleo-valley bottoms (Jeannin, Bitterli & Huselmann 2000). e highest and oldest ve phases (at presumed spring elevations of >1900, 1800, 1720, 1585, and 1505 m a.s.l.) had their springs in the Eriz valley (Fig. 2). e next phase, at 1440 m, shows a change in ow direction of 180. e spring was then located in the area of Lake un. e inuence of todays Aare valley (the site of Lake un today) therefore became predominant. All subsequent springs (at 1145, 1050, 890, 805, 760, 700, 660, and 558 m a.s.l.) drained to wards the Aare valley. In the area between Lake un and Hohgant, a total of 23 sites were selected for sampling (see Fig. 2: stars indicate sites). Selection was made on the basis of a relative chronology, and care has been given to ensure that ei ther the oldest possible sediment, or a series in stratigraphic order, was sampled. Due to the limited amount of time in which sam pling could be done, the relative chronology is incomplete (Tab. 5), although the main events were re traced. 21 samples were analysed (Huselmann & Granger 2005). e results show a great diversity of ages, ranging from 118 ka up to Fig. 2: Projection (370 degrees) of the Siebenhengste caves with the speleogenetic phases. Stars indicate sampling places for cosmogenic dates. From H uselmann & Granger (2005), modied. 0 0.5 1 1.5 km Caves of the region Siebenhengste Hohgant Perspective view from 370 g as of January 2002 Only the most important passages shown Brenschacht St. Beatus Caves Fitzlischacht Faustloch A2 Siebenhengste F1 K2 by HRH & toporobot Hagltsch Zone profonde Eriz Aare valley m a.s.l. Btterich Hohgant 558 1 150 1440 1505 1720 1950 890 760 Spring for 1950-1505 Spring for 1440 558 1050 A201 SHP RBL FSTL A2 FITZ BG1 HGL Fig. 3: Plot of ages (vertical) versus altitude (horizontal). H O W TO DATE NOTHING W ITH COSMOGENIC NUCLIDES

PAGE 6

TIME in KARST 2007 98 4.4 Ma (Tab. 6). e surface sample (MW A) has a burial age of 106 176 ka. us, the value is indistinguishable from zero, and we may assume that the sample was never buried. e sample from St. Beatus Cave (BG1) has an age of 182 122 ka. Its true value, bracketed by U/ 2000 1500 1000 500 0 1 2 3 4 5 Burial age (10 6 years) Elevation (m a.s.l.) 0.12 km/My 1.2 km/My Fig. 4: Rate of valley lowering in the Siebenhengste. Only maximum and minimum ages are displayed; however the valley deepening rates as well as the knickpoint at ~800 ka are easily visible. = morphologic event (a denotes phreatic genesis, a vadose enlargement), italic = dated event A201 ShP low SHP up Hagltsch A2TR A2CHU A2NS RBL L18 Faustloch Beatus Age Interpretation ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------SHP 7 Sediment 4.39 SHP2 2.35 SHP 3 Silt ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------Flowst. Flowst. Silt lake Silt Flowst. Sand A201 SHP5 1.9-1.84 Sand Sand Silt Silt ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------Flowst. Flowst. Silt Flowst. lake SHP1 SHP6 HGLP 1.54-1.60 Flowst. Flowst. Flowst. HGLS L18 1.04-1.09 (.93?) Silt ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------Flowst. SHP4 HGLT A2TR A2CHU A2NS RBL2 0.78-0.80 (.93?) RBL1 0.63 ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------Flowst. flooding FSTL 0.47 ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------BG23 0.23 BG1 0.18 BG20 0.16 ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------T ab. 5: Relative chronology of events around the Siebenhengste P HILIPP H USELMANN

PAGE 7

TIME in KARST 2007 99 ages, should be between 160 and 235 ka, which is again the case. ese values indicate that the method yields young ages where expected. A dirculty for dating with cosmogenic nuclides is mobility of the sediment. For instance, recent sand can be transported into a fossilized cave by a ood and then be deposited. Our results show that this process happens: for any speleogenetic phase, there is a range of ages ob served (Fig. 3). However, Fig. 3 also indicates that the re-mobilization and re-deposition of old sediments is rarely observable: if this would be the case, we would expect a random distribution of ages throughout the phases. However, the maximum age decreases with the next lower phase. W e can thus construct a gradual valley lowering with time which is represented in Fig. 4. W e see a knickpoint in the line connecting the ages: this knick point occurs at around 800 ka and 1500 m. is point reects a dramatic increase in valley deepening rate and coincides with the change in ow direction from Eriz to the Aare valley. T ab. 6: Results of dating. C ONCLUSIONS A relative chronology of events, albeit incomplete, cou pled with burial age dating by cosmogenic nuclides, per mitted to obtain a continuous history of valley incision in the Alps. Such data cannot be obtained in the same precision with other methods or at the surface. e re sults presented here are the rst cosmogenic dates for an Alpine cave system in a glacially inuenced area. e re sults indicate an onset of karstication in the Siebenheng ste before 4.4 Ma, that is in the Pliocene or even earlier. Together with U/ dates obtained earlier (Huselmann 2002), the history of the Siebenhengste cave system and its surrounding environment can be traced back over a huge time span. e construction of a complete relative chronology is very time-consuming, but can be extremely reward ing given the information one can extract from the cave. If speleogenetic phases, which are related to the overall geomorphic evolution of an area, can be expanded by such relative chronologies as well as absolute dates, the rate, duration, and extent of valley deepenings can be as sessed, and a paleoclimatic history can be drawn as well. H O W TO DATE NOTHING W ITH COSMOGENIC NUCLIDES

PAGE 8

TIME in KARST 2007 100 REFERENCES Audra, Ph., L. Mocochain, H. Camus, E. Gilli, G. Clauzon & J.-Y Bigot, 2004: e eent of the Messinian Deep Stage on karst development around the Mediterra nean Sea. Examples from Southern France. Geo dinamica Acta, 17, 6, 27-38. Bini, A., P. Tognini, & L. Zuccoli, 1998: Rapport entre karst et glaciers durant les glaciations dans les val les pralpines du Sud des Alpes. Karstologia, 32, 2, 7-26. Ford, D.C. & P. W illiams, 1989: Karst geomorphology and hydrology. Chapman & Hall, London, 601 p. Granger, D.E, D. Fabel & A.N. Palmer, 2001: PliocenePleistocene incision of the Green River, Kentucky, determined from radioactive decay of cosmogenic 26Al and 10Be in Mammoth Cave sediments. GSA Bulletin, 113, 7, 825-836. Granger, D.E. & P.F. Muzikar, 2001: Dating sediment burial with in-situ produced cosmogenic nuclides: theory, techniques, and limitations. Earth and Planetary Science Letters, 188, 269-281. Huselmann, Ph., 2002: Cave genesis and its relationship to surface processes: Investigations in the Siebenheng ste region (BE, Switzerland). PhD thesis, Universit de Fribourg, 168 p. Huselmann, Ph. & D.E. Granger, 2005: Dating of caves by cosmogenic nuclides: Method, possibilities, and the Siebenhengste example (Switzerland). Acta Carsologica, 34, 1, 43-50. Jeannin, P.-Y ., T. Bitterli, T. & Ph. Huselmann, 2000: Genesis of a large cave system: the case study of the North of Lake un system (Canton Bern, Switzer land). In: A. Klimchouk, D. C. Ford, A. N. Palmer, & W Dreybrodt (Eds.), Speleogenesis: Evolution of Karst Aquifers, pp. 338-347. Rossi, C., A. Cortel & R. Arcenegui, 1997: Multiple pa leo-water tables in Agujas Cave System (Sierra de Penalabra, Cantabrian Mountains, N Spain): Crite ria for recognition and model for vertical evolution. Proceedings 12th Int. Congress of Speleology, La Chaux-de-Fonds, Switzerland, 1, 183-187. Sasowsky, I.D., 1998: Determining the age of what is not there. Science, 279, 1874. Sptl, C., M. Unterwurzacher, A. Mangini & F.J. Long stae, 2002: Carbonate speleothems in the dry, in neralpine Vinschgau valley, northernmost Italy: W itnesses of changes in climate and hydrology since the last glacial maximum. Journal of Sedimentary Research, 72, 6, 793-808 P HILIPP H USELMANN



PAGE 1

U PPER C RETACEOUS TO P ALEOGENE FORBULGE UNCONFORMIT Y ASSOCIATED W ITH FORELAND BASIN EVOLUTION K RAS M ATARSKO P ODOLJE AND I STRIA ; S W S LOVENIA AND N W C ROATIA Z AKRASELA PERIFERNA IZBOKLINA POVEZANA Z RAZVOJEM ZGORNJEKREDNO PALEOGENSKEGA PREDGORSKEGA BAZENA ; K RAS M ATARSKO PODOLJE IN I STRA JZ S LOVENIJA IN SZ H RVAKA Bojan O TONIfAR 1 1 Karst Research Institute ZRC SAZU, Titov trg 2, Si-6230 Postojna, Slovenia, e-mail: otonicar@zrc-sazu.si Received/Prejeto: 01.02.2006 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 101 Abstract UDC 551.44.551.7(497.4-14) 552.541.551.7(497.4-14) Bojan Otoniar: Upper Cretaceous to Paleogene forbulge un conformity associated with foreland basin evolution (Kras, Matarsko Podolje and Istria; SW Slovenia and NW Croatia) A regional unconformity separates the Cretaceous passive mar gin shallow-marine carbonate sequence of Adriatic Carbonate Platform from the Upper Cretaceous and/or Paleogene shal low-marine sequences of synorogenic carbonate platform in southwestern Slovenia and Istria (a part of southwestern Slove nia and northwestern Croatia). e unconformity is expressed by irregular paleokarstic surface, locally marked by bauxite de posits. Distinctive subsurface paleokarstic features occur below the surface (e.g. lled phreatic caves, spongework horizons). e age of the limestones that immediately underlie the un conformity and the extent of the chronostratigraphic gap in southwestern Slovenia and Istria systematically increase from northeast towards southwest, while the age of the overlying limestones decreases in this direction. Similarly, the deposits of synorogenic carbonate platform, pelagic marls and ysch (i.e. underlled trinity), deposits typical of underlled peripheral foreland basin, are also diachronous over the area and had been advancing from northeast towards southwest from Campan ian to Eocene. Systematic trends of isochrones of the carbonate rocks that immediately underand overlie the paleokarstic sur face, and consequently, of the extent of the chronostratigraphic gap can be explained mainly by the evolution and topography of peripheral foreland bulge (the forebulge). e advancing exural foreland prole was the result of vertical loading of the foreland lithospheric plate (Adria microplate) by the evolving orogenic wedge. Because of synand post-orogenic tectonic processes, and time discrepancy between adjacent foreland ba sin deposits and tectonic (orogenic) phases it is dircult to dene the exact tectonic phase responsible for the evolution of the foreland complex. According to position and migration of the subaerially exposed forebulge, distribution of the foreland Izvleek UDK 551.44.551.7(497.4-14) 552.541.551.7(497.4-14) Bojan Otoniar: Zakrasela periferna izboklina povezana z razvojem zgornjekredno-paleogenskega predgorskega bazena; Kras, Matarsko podolje in Istra (JZ Slovenija in SZ Hrvaka) V jugozahodni Sloveniji in Istri so kredna karbonatna za poredja Jadranske karbonatne platforme pasivnega obrobja Jadranske mikroploe loena z regionalno diskordanco od zgornjekrednih in paleogenskih karbonatnih zaporedij sino rogene karbonatne platforme. Razgibano paleokrako povrje, ki diskordanco oznauje, je lokalno prekrito z boksitom. Pod povrjem se pojavljajo razline podpovrinske paleokrake oblike, med drugim veje zapolnjene freatine jame in dis kretni horizonti drobnih prepletajoih se kanalov. Starost apnencev neposredno pod paleokrakim povrjem in obseg stratigrafske vrzeli v jugozahodni Sloveniji in Istri sistematino naraata od severovzhoda proti jugozahodu, nasprotno pa starost apnencev, ki paleokrako povrje pokrivajo v tej smeri upada. Preko obravnavanega obmoja so med campanijem in eocenom od severovzhoda proti jugozahodu napredovala tudi sedimentna zaporedja sinorogenih karbonatnih platform (karbonatne kamnine Krake grupe) ter pelaginih laporjev in ia, ki predstavljajo sedimente podhranjenega predgorskega bazena. Sistematine trende izohron karbonatnih kamnin, ki leijo neposredno pod in nad paleokrakim povrjem in posledino razpona stratigrafske vrzeli lahko v veliki meri razloimo z evolucijo in topograjo periferne predgorske iz bokline. Napredujoi eksurni predgorski prol je nastal zaradi vertikalne obremenitve predgorske litosferske ploe (Jadran ske mikroploe) z nastajajoim orogenim klinom. Zaradi soasnih in postorogenih tektonskih procesov ter asovnega neskladja med sedimenti sosednjih predgorskih bazenov in med razlinimi tektonskimi (orogenimi) fazami tega dela zahodne Tetide v kredi in paleogenu, je opredelitev tektonske faze, ki je neposredno odgovorna za evolucijo obravnavanega predgorja oteena. Glede na poloaj in migracijo periferne iz

PAGE 2

TIME in KARST 2007 102 related macrofacies and orientation of tectonic structures, espe cially of Dinaric nappes, and Dinaric mountain chain I suggest that the foreland basin complex in western Slovenia and Istria was formed during mesoalpine (Dinaric) tectonic phase due to oblique collision between Austroalpine terrane/Tisia micro plate and Adria microplate when probably also a segmentation of the foreland plate (Adria microplate) occurred. Key words: forebulge unconformity, paleokarst, chronostrati graphic gap, ysch, Adriatic Carbonate Platform, synorogenic carbonate platform, foreland basin, Adria microplate, Dinaric orogene, Cretaceous, Paleogene, SW Slovenia, Istria. bokline, razporeditev makrofaciesov podhranjenega predgor skega bazena ter usmerjenost tektonskih struktur, predvsem Dinarskih pokrovov, in Dinarskega gorstva v celoti domnevam, da je nastal predgorski sistem v zahodni Sloveniji in Istri med mezoalpidsko (Dinarsko) tektonsko fazo, kot posledica bone kolizije med Avstroalpidskim terranom in/ali Tisa mikroploo ter Jadransko mikroploo, pri emer je verjetno prilo tudi do segmentacije Jadranske mikroploe. Kljune besede: diskordanca, paleokras, kronostratigrafska vr zel, i, Jadranska karbonatna platforma, periferna predgorska izboklina, sinorogena karbonatna platforma, predgorski bazen, Jadranska mikroploa, Dinarski orogen, kreda, paleogen, ju gozahodna Slovenija, Istra. I NTRODUCTION Plate tectonics theory had a crucial impact on our un derstanding of sedimentary basins, and consequently, of carbonate sedimentary systems. Plate tectonics de termines not only the gross architecture (dimension and shape) and lithological/structural characteristics of carbonate platforms (Bosellini, 1989), but also their evolution and the longevity. ose characteristics are largely dened by specic geotectonic setting in which certain carbonate platform begin to grow. Carbon ate platform(s), which colonize certain area through longer or shorter period of geologic history, constantly change its/their position in relation to the equator and plate boundaries and pass through dierent phases of the W ilson cycle. e sedimentary and diagenetic char acter of the carbonate platform(s) constantly change(s) during this journey and at a stretch, the platform evo lution may be stoped. In this case, the area formerly inhabited by the carbonate platform may fall under conditions which are not favourable for considerable carbonate production. In one scenario it may imme diately aer the deposition or later in the geologic his tory be uplied, subaerially exposed and karstied. Similarly as the plate tectonics governs the sedimen tary evolution of the carbonate platforms, it may also determins their diagenetic evolution, including karsti cation. e gross architecture, lithological/structural caharacteristics, and the evolution and the longevity of the uplied area with subaerially exposed carbonate platform are mainly dependent on its geotectonic posi tion regard to plate boundaries, former geodynamics and consequently topography of the area, especially of the carbonate platform. Although important for the ap pearance of the karstic landscape, the eects of other variables, such as climate and ground water level, may be just superimposed on the geotectonically predis posed framework. Each karstic landscape carries its specic geotecton ic signature which can be read from and explained with specic evolution of karstic features and a karst system as a whole. In addition, studies of sedimentary succes sions of rocks that underand overlie the (paleo-) karstic surface and that of the adjacent sedimentary basins as well as the general geologic conditions of the region may signicantly improve our knowledge on geodynamics of the uplied area. e paper documents an example of paleokarst that occurred during the upli of the Adriatic Carbon ate Platform (sensu Vlahovit et al., 2005) in the distant foreland region of the evolving collision related orogenic belt between the Adria microplate (sensu Stampi et al. 1998) and the Austroalpine terrane and/or Tisia micro plate (sensu Neugebauer et al., 2001) in the Late Creta ceous and the Early Paleogene. e study is based on 36 geological proles from the karstic regions of southwestern Slovenia, both Slo venian and Croatian part of Istria peninsula and the area between Trieste bay and Italian-Slovenian border in northeastern Italy (Figs. 1, 2). To get a whole picture of conditions that dominated the region during the emersion period, I expend the area of interest to syno rogenic carbonate platform that onlap the paleokarstic surface and to siliciclastic ysch regions of afore men tioned areas and the adjacent regions of western Slove nia and northeastern Italy (along the border between Italy and Slovenia). e aim of this work is to show the causes of the upli and subaerial exposure of the northwestern part of the Cretaceous Adriatic Carbonate Platform in Late Cretaceous and Early Paleogene. e data presented here were provided mainly from the studies of paleogeograph ic and topographic extent of the emersion, stratigraphy of the carbonate successions that immediately underand B OJAN O TONIfAR

PAGE 3

TIME in KARST 2007 103 Fig. 1: Geographical position and simplied geological map of the western Slovenia and Istria showing major structural elements (modied from Placer, 1999). overlain the paleokarstic surface, stratigraphy and sedi mentology of the onlapping synorogenic carbonate plat form and adjacent deeper marine basin as well as from regional geotectonic and general geologic situation. e time of the upli is correlated with events on the adjacent plate boundaries of the western Tethian domain (tradi tional orogenic phases) and global eustatic curve. G EOLOG Y OF THE AREA e geology of southwestern Slovenia and Istria has been studied from the late 19 th century on. Since that time also a regional unconformity which separates shallow-marine carbonate successions of dierent Cretaceous formations from shallow-marine limestones of the Upper Creta ceous/Lower Paleogene Liburnia Formation or Eocene U PPER C RETACEOUS TO P ALEOGENE FORBULGE UNCONFORMIT Y ASSOCIATED W ITH FORELAND BASIN EVOLUTION

PAGE 4

TIME in KARST 2007 104 Fig. 2: Simplied lithostratigraphic columns of Cretaceous to Eocene successions in southwestern Slovenia and Istria (NW Croatia and SW Slovenia), with one column from NW Italy. Authors of original geological columns are listed below: 1) ribar (1995); Rinar (1997), 2) D robne (1977, 1979), 3) D robne et al. (1988, 1996); ribar (1995); J urkovek et al. (1996), 4) J urkovek et al. (1996), 5) D robne (1981); J urkovek et al. (1996), 6) J urkovek et al. (1997), 7) H amrla (1959); D robne (1977, 1979); Pavlovec et al. (1991), 8) H amrla (1959, 1960); J urkovek et al. (1996), 9) B razzatti et al. (1996), 10) H amrla (1960); D robne et al. (1991); J urkovek et al. (1996), 11) H amrla (1959); B user & Lukacs (1979); D elvalle & B user (1990); J urkovek et al. (1997); this study, 12) D robne (1977); D elvalle & B user (1990), 13) D elvalle & B user (1990); ribar (1995); B user & Radoiin (1987), 14) ikin et al. (1972); D robne (1977), 15) D robne (1977); this study, 16) D robne (1977, 1981); H ottinger & D robne (1980); D robne & Pavlovec (1979); D robne et al. (1991); T urnek & D robne (1998); this study, 17) D robne (1977), 18) iki et al. (1972); D robne (1977), 19) B iondin et al. (1995), 20) ikin et al. (1972); D robne (1977), 21) ikin et al. (1968); D robne (1977), 22) Pleniar et al. (1969); D robne (1977); Gabrin et al. (1995), 23) Pleniar et al. (1969); D robne (1977), 24) H amrla (1959); Pleniar et al. (1973); D robne (1977); V elin & V lahovin (1994); M atiec et al. (1996), 25) ikin et al. (1968); D robne (1977); H ottinger & D robne (1980); D robne et al. (1991), 26) M atiec et al. (1996), 27) T arlao et al. (1995), 28) B user & Lukacs (1972); D robne (1977); H ottinger & D robne (1980); M atiec et al. (1996), 29) Polak & ikin (1973); D robne (1977), 30) D robne et al. (1991), 31) 34) M atiec et al. (1996), 35) ikin et al. (1968); M aga (1973); ikin et al. (1973); ikin & Polak (1973); H ttinger & D robne (1980); Otoniar et al. (2003), 36) Polak (1970); D robne (1977); M atiec et al. (1996). B OJAN O TONIfAR

PAGE 5

TIME in KARST 2007 105 Alveolina-Nummulites Limestone has been known. e Liburnia Formation, Alveolina-Nummulites Limestone and intermediate Trstelj Beds represent the Kras Group (Koir, 2003) (Fig. 3), which corresponds to the lower unit of the underlled peripheral foreland basin stratigraphy (i.e. the lower unit of the underlled trinity of Sinclair, 1997). us the unconformity represents a megasequence boundary and typically separates the underlying passive margin carbonate succession from the overlying deposits of the synorogenic carbonate platform at periphery of the foreland basin (Koir & Otoniar, 2001). e synorogen ic carbonate platform was nally buried by prograding hemipelagic marls (i.e. the middle unit of the underlled trinity of Sinclair, 1997) and deep-water clastics (ysch) (i.e. the upper unit of the underlled trinity of Sinclair, 1997) (Fig. 3). Because the name of the carbonatre plat form that overlie the unconformity has not been dened yet, I will use in this paper only the general geodynamic term i.e. the synorogenic carbonate platform. e unconformity is expressed by an irregular paleokarstic surface, locally marked by bauxite depos its. Although the unconformity has been repeatedly mentioned, no systematic study of paleokrast has been performed. Relatively numerous papers on biostratig raphy, especially on the carbonate successions of the Kras Group, have been published (see list of references attached to Fig. 2), yet not more than few attempts on explanation of the sedimentology of the paleokarstic deposits and onlapping beds have been done (Otoniar, 1997; Debeljak et al. 1999; Durn et al. 2003). Only oc casionally, the geotectonic conditions under which the paleokarst (upli) evolved have been briey mentioned (Koir & Otoniar, 2001; Otoniar & Koir, 2001; Durn et al., 2003). Tectonically, the discussed area corresponds to three macrotectonic units, the Southern Alps, the Ex Fig. 3: Generalized stratigraphic column of Upper CretaceousEocene succession in the Kras (Karst) and M atarsko podolje regions, SW Slovenia, showing major lithostratigraphic units (modied from Koir, 2004). Fig. 4: Illustrative geological map showing distribution of ysch deposits and major structural elements in western Slovenia. e map is based mainly on data from basic geological maps of Y ugoslavia, 1:100.000, sheets B eljak & Ponteba ( J urkovek, 1986), Udine-T olmin & V idem (Udine) (B user, 1986), Kranj (Grad & Ferjani, 1974), Gorica (B user et al., 1968), Postojna (B user et al., 1967), T rst (Pleniar et al., 1969) and Ilirska B istrica ( ikin et al., 1972). Copyright: Geoloki zavod Slovenije (Geological survey of Slovenia), 2002 All rights reserved. U PPER C RETACEOUS TO P ALEOGENE FORBULGE UNCONFORMIT Y ASSOCIATED W ITH FORELAND BASIN EVOLUTION

PAGE 6

TIME in KARST 2007 106 ternal Dinarides and the Dinaric foreland (Placer, 1999) (Fig. 1). W hile the ysch-related sediments can be fol lowed across the all three units (Fig. 4), the unconfor mity and the overlying carbonate successions of the Kras Group correspond to the most external thrust unit of the Dinaric fold and thrust belt the northwestern External Dinarides in southwestern Slovenia, Italian part of the Kras plateau and northeastern Istria, and to more stable foreland domain of the Dinaric mountain belt in other parts of Istria (Figs. 1, 2). e nappe structure of northwestern part of the Ex ternal Dinarides comprises ve successively lower and younger thrust units from northeast to southwest: Trno vo Nappe, Hruica Nappe, Snenik rust Sheet, Komen rust Sheet and Kras rust Edge (Placer, 1981, 1999, 2002) (Fig. 1). e External Dinarides and the Dinaric foreland correspond to the northwestern part of the Cretaceous Adriatic Carbonate Platform and the Upper CretaceousEocene synorogenic carbonate platform which occupied northeastern part of the Adria microplate s.s. (Fig. 5). In the Cretaceous the area of present day Southern Alps was a part of deeper marine realm which comprised the Slovenian Basin formed in the Middle Triassic (Cousin, 1981; Buser, 1989) and the area of former Julian Carbon ate Platform which was drowned in the Lower and Mid dle Jurassic (Cousin, 1981; Buser, 1989). e geologic and paleogeographic situation started to change severely in the Late Cretaceous (see below). It is important to note, that the described region is re cently conned from the north side by the Periadriatic fault zone, from the west by the deposits of the Southern Alpine Molasse Basin and from the south and southwest by the Adriatic Sea and its sediments (Fig. 1). To understand the mechanisms that governed the upli and emersion, regional geotectonic conditions of the wider area of the Late Cretaceous-Early Paleogene W estern Tethys were taken into consideration. During the Mesozoic, the area between Eurasia and Gondwana or the western part of the extensive Tethys bay of the Pangea was occupied by more or less uniform Adria microplate surrounded by smaller tectonic units or terranes (Fig. 5). W ith regard to major geotectonic events, the extent and shape of Adria microplate was Fig. 5: A) Paleogeographical map showing major geotectonic units at Santonian-Campanian boundary in western T ethys and central Atlantic (modied from Neugebauer et al., 2001). B) Geotectonic and paleogeographic units of Adria microplate and adjacent areas. B OJAN O TONIfAR

PAGE 7

TIME in KARST 2007 107 changing constantly through the geologic history. e re sults of these events (e.g. tecto-sedimentary successions or cycles) could be correlated between geographically and geologically distant parts of the Adria microplate. Aer substantial Permian to Middle Triassic and Triassic/Jurassic extensional tectonics, the Adria domain became encircled by oceanic bays and dissected by nu merous deepwater basins and drowned carbonate plat forms (Fig. 5). It is considered that since Early Jurassic the Adriatic Carbonate Platform had been isolated by deeper marine realms (Vlahovit et al., 2005). At the Middle/Late Jurassic boundary compression al tectonic regime prevailed over the peri-Adriatic region. It was caused by the beginning of closure (subduction) of the adjacent oceanic bays of the W estern Tethys. During the Late Jurassic and Cretaceous gentle broad-scale posi tive and negative lithospheric deections periodically oc curred on the Adriatic Carbonate Platform. e deec tions were expressed by coexistence of karstic areas and somewhat deeper marine intra-platform basins (Tiljar et al., 1995; 1998; Vlahovit et al., 2005). Distinctive de ections correspond to period of ophiolite emplace ment [e.g. the Late Jurassic/Early Cretaceous obduction of ophiolite suite of the Dinaric Tethys on the E margin of the Adria microplate (Pamit et al., 1998; 2000)] and distant collisions [e.g. the mid-Cretaceous Eoalpine oro genesis in the Pelso/Austroalpine/Tisia domain (Faupl & W agreich, 2000; Neugebauer et al., 2001)]. Topographic disunity over the platform gave rise to irregular facies distribution and thickness of carbonate successions of dierent parts of the platform. Signicant interruptions of carbonate successions are also related to global eustatic oscillations and/or oce anic anoxic events, but they are mainly superimposed on tectonically induced changes of relative sea-level. us before the beginning of the upli of northern part of the Adriatic Carbonate Platform in the Late Cre taceous and the synchronous onset of ysch sedimenta tion in the area north and north-eastern of the platform, the whole region was already topographically distinctly heterogeneous. Flysch started to deposit in deeper ma rine basin with partly inherited bathymetry from former deeper marine domain of Slovenian Basin and drowned Julian Carbonate Platform (Fig. 5). Deeper marine realms with more or less uninterrupted sedimentation had still encircled the carbonate platform from its western and southwestern side (Vlahovit et al., 2005) (darker grey area on Fig. 5). Later tectonic activity which shortened the area and displaced dierent parts of the region, prevent more ac curate interpretation of geotectonic conditions at those time. Namely, except the substantial shortening of the region due to dierent thrusting phases of Alpine oro geney, the area north from the Periadriatic Fault Zone was displaced for at least 100 km eastward during the Miocene (Ratschbacher et al., 1991; Frisch et al., 1998; Vrabec & Fodor, 2006), in some estimates up to 500 km (Haas et al., 1995). It should be noted that western Istria (i.e. Dinaric Foreland on Figure 1) experienced signi cant counterclockwise rotation most likely between the end of Miocene and the earliest Pliocene (Mrton et al., 1995; Mrton, 2006). P ALEOKARST In the investigated area both surface and subsurface pale okarstic features occur. In places the paleokarstic surface is denoted by surface karst forms like karrens, dolines and depressions of decimetric amplitude (Fig. 6a). Pedogenic features and enlarged root-related channels characterize the upper part of the vadose zone, the epikarst. Vadose channels, shas and pits penetrate up to a few tens of meters bellow the paleokarstic surface, where they may merge with originally horizontally oriented phreatic cav ities. e latter comprise characteristics of caves forming in fresh/brackish water lenses. At least some of them may be dened as ank margin caves (Fig. 6b, 6c). In extensive outcrops, the remains of such caves can be followed as much as few hundreds of meters along strike. In one case a breccia body which was dened as paleokrastic cave re lated deposit (Otoniar et al., 2003), is so extensive that was used even as mappable unit for Basic geologic map of Y ugoslavia 1:100.000 (see Maga, 1965). e cavities are usually irregular and elongated in shape, and could be up to few tens of meters long and up few meters high (Fig. 6b). Depending on locality, the phreatic cavities were found in dierent positions regarding to the paleokrastic surface, the lowest one some 75 meters below it. e cavi ties had been subsequently partly reshaped and entirely lled with sediments and owstones in the upper part of the phreatic, epiphreatic and vadose zones (Figs. 6b, 6c). Similarly, the vadose channels and voids are also lled by sediments and owstones, but they usually dier from these of phreatic cavities in higher content of noncar bonate material, lower C values of carbonate material and more distinctive pedogenic modication. e denu dation had frequently exposed lled paleokarstic subsur face cavities on the paleokarstic surface, where they may be identied only by the remains of their ll (Otoniar U PPER C RETACEOUS TO P ALEOGENE FORBULGE UNCONFORMIT Y ASSOCIATED W ITH FORELAND BASIN EVOLUTION

PAGE 8

TIME in KARST 2007 108 et al., 2003) (Fig. 6c). e internal sediments and ow stones oen occur as grains in deposits that cover the paleokarstic surface or ll subsurface paleokarstic cavi ties of dierent generations. Paleokarstic surface with its depressions as well as subsurface channels and voids are oen covered and lled by bauxite deposits which were locally exploited (Fig. 6d) (Gabrit et al., 1995). Certain limestone lithofacies of immediate cover of the unconformity are commonly locally conned, sug Fig. 6: A) Paleokarstic surface is locally denoted by small scale depressions (motorway road-cut at Kozina village, SW Slovenia). Note colour contrast between Upper Cretaceous shallow marine limestone of the Lipica Formation and dark grey palustrine limestone of the Liburnia Formation. H ammer for scale is about 30 cm high. B) H orizontally oriented cave of irregular shape largely lled with reddish-stained calcareous mudstone/siltstone (Podgrad, M atarsko Podolje, SW Slovenia). e maximal height of the cave is approximately 4 meters. e cave deposits are articially marked by reddish transparent colour on the photograph. C) B reccia body represents a part of lled rooess paleokarstic phreatic cave at Koromano in Istria, NW Croatia. (1,8 m tall geologist for scale in the upper right corner) D) Excavated paleokarstic cavity (vadose sha?) originally lled with bauxite (M injera, Istria, NW Croatia). 6 A 6 C 6 B 6 D B OJAN O TONIfAR

PAGE 9

TIME in KARST 2007 109 gesting highly irregular topography of the karstic surface before the beginning of transgression. In places it is clear that the incipient transgression involved gradual increase of groundwater table and, eventually, ponds or blue holes were formed in karstic depressions (Durn et al., 2003). In the Kozina site (southwestern Slovenia) during the blue hole stage of the transgression, a paleokarstic pit was lled by coarse grained breccia with vertebrate remains, mainly dinosaurian and crocodilian bone frag ments and teeth (Debeljak et al., 1999, 2002). General ly, the cover sequence (i.e. the Liburnian Formation of Maastrichtian and early Paleogene age) is characterized by restricted, marginal marine and palustrine lithofacies, which frequently show pedogenic modications. E VOLUTION OF THE PERIPHERAL BULGE THE FOREBULGE Besides the research on paleokarst related phenomena, the study of sedimentary successions of the host rock in which the paleokarstic features occur and those that overlie the paleokarstic surface is of crucial importance to understand the upli of substantial part of the Adriatic Carbonate Platform above the sea-level in the Late Cre taceous and Paleogene. To explain the mechanisms that govern the upli, regional and global geotectonic and eu static conditions were taken into consideration, too. STRATIGRAPH Y e age of the limestones that immediately underlie the unconformity and the extent of the chronostratigraphic gap in southwestern Slovenia and Istria systematically increase from northeast towards southwest (Figs. 2, 7a, 7b), while the age of the overlying limestones decrease in this direction (Figs. 2, 7c). In western part of Istria the orientation of the isochrones is slightly dierent and Fig. 7: A) Isochrones of carbonate rocks that immediately underlie the unconformity. B) Isochrones of the extent of the chronostratigraphic gap. C) Isochrones of carbonate rocks that immediately overlie the unconformity. Isochrones in all gures are in M a. M ajor structural elements of the area (see Fig. 1) and positions of the geological proles used in the research (see Fig. 2) are also shown in the gures. U PPER C RETACEOUS TO P ALEOGENE FORBULGE UNCONFORMIT Y ASSOCIATED W ITH FORELAND BASIN EVOLUTION

PAGE 10

TIME in KARST 2007 110 shows a dome-like topography of the forebulge. e iso chrones represent a statistic result acquired by kriging in Surfer programme version 8.00 ( Golden Soware, Inc.). e data were provided from 36 geological proles from the karstic regions of southwestern Slovenia, both Slovenian and Croatian part of Istria peninsula and the area between Trieste bay and Italian-Slovenian border in northeastern Italy (Figure 2 and red dots on Figures 7a, 7b, 7c). e youngest rocks below the unconformity be long to mid-Campanian and occur in the central and northeastern part of the Kras (Karst) plateau (the Komen thrust sheet) (Fig. 1) (Jurkovek et al., 1996) and close to Postojna (the Snenik thrust sheet) (Fig. 1) (ribar, 1995; Rinar, 1997) in southeastern Slovenia, while the oldest one, Valanginian and Hauterivian in age, crop out in the western part of Istria (Matiec et al., 1996) (Figs. 2, 7a). e beds that cover the unconformity correspond to dierent ages, litofacies, members and formations. As mentioned afore, the age trend of the immediate cover is opposite to that of the footwall. In this case the oldest rocks occur in southwestern Slovenia and belong to the youngest stage of the Late Cretaceous the Maastrich tian. Towards southwest, progressively younger deposits onlap the paleokarstic surface (Figs. 2, 7c). However, the youngest strata that onlap the unconformity dont t ex actly with the oldest one immediately below it. W ith re gard to described situation, the chronostratigraphic gap increases considerably from few Ma on the Kras plateau (southwestern Slovenia) to more than 80 Ma in western Istria (Figs. 2, 7b). e lithofacies of the lower part of the cover sequence (e Liburnian formation) frequently show features typi cal of subaerial exposure surfaces, including calcrete, pseudomicrokarst, brecciated horizons and karstic sur faces. Locally, the lowermost subaerial exposure surface of the Liburnija Formation, which shows karstic topog raphy of decimetric amplitude, and the main paleokarstic surface form a composite unconformity. Sporadically, thin coal beds and seams occur in the lower part of the sequence. Although the stratigraphy of the Kras Group, Transitional Beds and Flysch (Fig. 3) shows overall deepening of the basin, prominent subaerial exposure surfaces also occur in carbonate successions of Trstelj Beds and Alveolina-Nummulites Limestone (Koir & Otoniar, 1997; Koir, 2003). Much thicker successions of paralic sediments with more frequent unconformities and marsh related sediments occur in southwestern Slo venia and northeastern Istria in comparison with other parts of Istria, yet local variation can be signicant (Figs. 2, 8). In western Istria, where the chronostratigraphic gap is the most extensive, the foraminiferal limestones fre quently lie directly on the paleokarstic surface (Matiec at al., 1996). e thickness of the Kras Group generally decreases from northeast toward southwest, although also in this case signicant deviations may occur (Figs. 2, 8). e point where the unconformity pinch-out to wards the foreland basin occurs somewhere between the northeastern part of the Kras plateau on the Komen rust Sheet and some 10 km (approximately 25 km in original position see Placer, 1999) distant Mt. Nanos on the Hruica nappe (Fig. 1). From this point on towards the foreland basin, the upli of the forebulge didnt take place because the area was so close to the orogene that experienced only a subsidence. Here, the sedimentary succession of the Adriatic Carbonate Platform gradually passes into progressively deeper-marine carbonate suc cession of synorogenic carbonate platform. Namely, on the Mt. Nanos at Campanian-Maastrichtian boundary, the deepening of the shallow marine carbonate platform without any evidence of preceding emersion is docu mented (ribar, 1995). Further towards the northeast, in the Julian Alps (the eastern part of the Southern Calcareous Alps) and in the most northern part of recent Dinaric mountain belt in western Slovenia and northeastern Italy (the Trnovo Nappe), the turbiditic siliciclastic sediments (ysch) started to deposit in Campanian and Maastrichtian over the rocks of dierent lithology, age and origin (Pavi, 1994). Flysch oen overlies deeper marine pelagic marls of scaglia type and alodapic carbonates, which were re ceiving the material from Adriatic Carbonate Platform. It is important to note that in this part of western Slovenia deep-marine basin existed before ysch or above men tioned deeper marine pelagic marls started to deposit. However, the oldest pelagic marls (pre-ysch deposits) which overlie the Upper Cretaceous shallow marine car bonates of the northeastern margin of the Adriatic Car bonate Platform also belong to Maastrichtian. Similar as I stated for chronostratigraphic gap, the pelagic marls and ysch deposits are also diachronous over the area. From northeast toward southwest, successively younger strata onlap the pre-foreland basin deposits (Fig. 4). e successions of pelagic marls and especially si liciclastic ysch were periodically interrupted by depo sition of calcarenitic and calcruditic beds/megabeds, locally even of olistostrome character. ose beds were supplied by turbiditic currents from the fault-related es carpments of distorted and seismically active marginal areas of former Adriatic Carbonate Platform (Skaberne, 1987; Tunis & Venturini, 1987) and later also from outer parts of synorogenic carbonate platforms (distally steep ened ramps?) (Fig. 9). e synorogenic carbonate and siliciclastic deposits of other parts of External Dinarides (e.g. Dalmatia) are B OJAN O TONIfAR

PAGE 11

TIME in KARST 2007 111 younger than these described here. ey started to de posit not before Eocene (Marjanac & osovit, 2000) and probably represent deposits of dierent foreland system or at least of dierent segment of the one described here. D ISCUSSION Systematic trends expressed by isochrones showing the age of the carbonate rocks that immediately underand overlie the paleokarstic surface (Figs. 7a, 7c), and conse quently, the extent of the chronostratigraphic gap (Fig. 7b), can be explained mainly by the evolution and topography of peripheral foreland bulge (the forebulge) (Fig. 9). W hen the foreland continental lithospheric plate is vertically loaded by the fold and thrust belt, it responds with exure. In front of the evolving orogen an asym metric foreland basin is formed; the deepest part of the basin (the foredeep) is located adjacent to the orogenic wedge (Fig. 9). Because of the isostatic rebound on ver tical loading of the lithosphere, the opposite side of the basin (opposite to the orogenic wedge) is instantaneously upwarped and the bulge with subtle relief is formed, the peripheral bulge or the forebulge. e bulge is especially well expressed in early, ysch stage of the foreland basin evolution (Crampton & Allen, 1995). W hile the wave length of the deection is approximately the same for both, foreland basin and peripheral bulge, the ampli tude of the basin subsidence is typically much greater as the upli of the bulge (Crampton & Allen, 1995; Miall, 1995). If the conditions are suitable, synorogenic car bonate platforms with distinctive ramp topography may colonise the gentle slope of the forebulge toward the fore deep (Dorobek, 1995). Signicantly, as the whole complex of the orogenic wedge advances forelandward, the exural prole pro duced by the orogenic wedge advances with it. Topog raphy of the forebulge is controlled by numerous factors, among which the rigidity of the foreland lithospheric plate and the rate of emplacement of the load are the most important (Allen & Allen, 1992; Dorobek, 1995; Miall, 1995). An expected maximal height of the forebulge above the sea level (if the foreland plate is at or close to sea-level prior to exural loading) would be in the range of up to a few tens to few hundreds meters (Crampton & Fig. 8: Lithostratigraphic columns for three adjacent sites in M atarsko Podolje and Mt. Slavnik (SW Slovenia). Note signicant variations in thickness of lithostratigraphic units and in time span of stratigraphic gap. U PPER C RETACEOUS TO P ALEOGENE FORBULGE UNCONFORMIT Y ASSOCIATED W ITH FORELAND BASIN EVOLUTION

PAGE 12

TIME in KARST 2007 112 Allen, 1995; Miall, 1995). According to topography of the forebulge, the rate of erosion (see W hite, 2000) and the style of migration of the orogenic wedge, the area of max imal denudation should occur in the central part of the region, which is over-passed by the bulge (Crampton & Allen, 1995). In addition, non-exural deformations (e.g. reactivation of pre-existing heterogeneities, enhanced deections because of horizontal in-plane stresses) and inherited topography may signicantly inuence the evolution and topography of the forebulge (Allen & Al len, 1992; Dorobek, 1995; Miall, 1995; Crampton & Al len, 1995). On Mt. Nanos (Hruica Nappe; Fig. 1) shallow wa ter rudist limestone of the Adriatic Carbonate Platform gradually passes over limestone with orbitoidiform larger foraminifera into pelagic marls without any emersion at the base of the deepening sequence the erosional gap reduces to conformity. e age span of this transition falls within a period of the shortest documented chronostrati graphic gap between the northeastern part of former Adriatic Carbonate Platform and the overlying synogen ic carbonate platform (Fig. 2), which extends from midCampanian to Late Maastrichtian. Maastrichtian in age are also the oldest pelagic marls which in places directly overlie the Upper Cretaceous shallow water carbonates of former northeastern margin of the Adriatic Carbon ate Platform. Although the oldest turbiditic siliciclastic ysch was deposited in a basin with inherited deeper ma rine bathymetry (former Slovenian Basin) its Campan ian and Maastrichtian age could be correlated with other incipient foreland related deposits and phenomena. W ith regard to these criteria and trends of unconformity re lated isochrones elsewhere (Figs. 2, 7a, 7b, 7c), I suggest that northern part of the Adriatic Carbonate Platform had thrived more or less prosperously till the end of Campanian, when an initial upli of the forebulge occurred. e carbonate sediments that had origi nally been deposited till that time, and are now missing in carbonate successions immediately below the unconformity, had been erased dur ing the paleokarstic period by the karstic denudation processes. According to topography of the forebulge and advancing nature of the foreland geodynamic complex as a whole, the most extensive denuda tion is expected in the central area over which the forebulge migrates. e western part of Istria, where the chronostratigraphic gap is the larg est and the beds immediately below the unconformity are the oldest (Fifs. 2, 7a, 7b), most probably corresponds to this zone. However, in an ideal conceptual/mathematical model of the forebulge unconformity, the amount of ero sion should remain more or less constant over vast area in the central part of the region over-passed by the bulge, and decreases on its distal slope towards back-bulge ba sin (Crampton & Allen, 1995). Instead, in western Istria the isochrones of the beds underlying the unconformity show distinctive condensation compared to situation in northeastern Istria and southwestern Slovenia (Fig. 7a). I suggest that this is not the result of rapid increase of the amount of footwall eroded but rather of denudation of primarily much thinner Cretaceous carbonate succes sions in western Istria. Namely, in this part of Istria the carbonate successions are relatively thin (Matiec et al., 1996), partly because of repeating emersions throughout the Cretaceous (Velit et al., 1989) and partly because of reduced accommodation space of Cretaceous shallow marine environments. Evidence of considerable Late Ju rassic and Cretaceous land areas in the vicinity of west ern Istria (probably oshore form its recent west coast), came also from dinosaur record (footprints and bones) (Della Vecchia et al., 2000; Mauko & Florjani, 2003; Mezga et al., 2003) and distribution of sedimentary fa cies of the adjacent peritidal to deeper marine environ ments of intraplatform basins (Tiljar et al., 1995; 1998). W hy was the area of western Istria beeing preferentially uplied during the Cretaceous is still questionable, but the reasons for deections should be searched at adjacent plate boundaries where their reorganisation and dier ent collision-related events and processes (see Faupl & W agreich, 2000; Neugebauer et al., 2001) produced hori Fig. 9: Schematic block diagram of foreland basin complex showing the position of the orogenic wedge, foredeep and forebulge with distribution of macrofacies belts before plate convergence ended (modied from B radley & Kidd, 1991). B OJAN O TONIfAR

PAGE 13

TIME in KARST 2007 113 zontal in-plane stresses that may be transmitted many hundreds of kilometers inboard of actual collision (Zei gler et al., 1995). It is also possible that the central zone of the foreb ulge and the slope towards back-bulge basin in their nal position occurred oshore of recent Istrian west coast. However, we should be aware that the Late Cretaceous Adriatic Carbonate Platform was surrounded from the western side by deeper marine interplatform basins (Vlahovit et al., 2005) what might considerably aected the appearance of the forebulge and the back-bulge area. Although the abnormal thickness of denuded stratigraphy in western Istria is mainly the result of previous sedimentary history, some uncertainties may also arise from dierential upli/subsidence of certain parts of the forebulge. Evidence for dierential subsid ence along reactivated ancient tectonic structures is for example well documented in carbonate successions of the Kras Group, where the thickness of chronoand lithostratigraphic units may vary considerably over short distances (Figs. 2, 8). In conclusion I suggest that the denudation exposed the oldest carbonate rocks in the western Istria partly be cause of specic evolution (migration) and topography of the forebulge and partly because of primarily thinner carbonate successions in this part of Istria compared to more northeastern parts of the investigated area. e rate of transgression over the paleokarstic sur face is expressed by the isochrones of the strata that onlap the unconformity (Fig. 7c). W hile the large scale diachro nism of the onlapping strata shown in Figure 7c is the result of specic large-scale topography and migration of the forebulge as a whole, local smaller scale spatial dif ferences in the onlap pattern (not observable in Figure 7c) are due to shorter oscillations of relative-sea level and deposition over topographically irregular paleokarstic surface (e.g. dolines, shas a blue hole phase of the transgression). e pattern of the isochrones shown in Figure 7c suggests that the transgression during its earlier stages (southwestern Slovenia and northeastern Istria) was slower compared to its later stages (western Istria). Although subsequent tectonic deformations, such as tectonic shortening, faulting and rotation, substantially aected the area, the rate of the onlap in southwestern Slovenia and northeastern Istria is estimated to about 2-3 km/Ma while in southwestern Istria to about 4-5 km/Ma. W e should be aware that some apparent anomalies, espe cially at terminations of the isochrones may be the result not only of later tectonic deformations of the area but also of limited number of data points which are not uniformly distributed, spatially conned area of the investigation along the strike of the forebulge and defectiveness of sta tistic method (kriging) used. Slightly dierent orienta tion of the isochrones in western part of Istria compared to those in southwestern Slovenia and northeastern Istria (Figs. 7a, 7b, 7c) may also be the result of dierent syde positional or synorogenic orientation of the prevailing stresses (see Marinit & Matiec, 1991; Matiec et al., 1996) during the Cretaceous and Paleogene and later counterclockwise rotation of the area (see Mrton et al., 1995 amd Mrton, 2006). In spite of all that, the reasons for dierent stratigraphic pinch-out rate are many sided and may arise from dierential rheologic and structural characteristics of the foreland plate itself, events at colli sion zone and adjacent plate boundaries, sublithospheric processes and external reasons like eustatic sea-level os cillations and climate changes. In our case it is dircult to determine the exact reason for the increasing rate of the onlap in Lower Eocene, not only because dierent pro cesses may lead to the same result, but also because they can act simultaneously. Long term sea-level fall (i.e. second-order cycle of Haq et al., 1987) may for example slow-down the onlap rate and vice-versa long term sea-level rise may increase the onlap rate. If we observe the eustatic curve for the Cretaceous and Paleogene (Haq et al., 1988) we can no tice that the rate of the onlap is in relatively good agree ment with mid-Campanian to Late Paleocene second-or der fall and Early Eocene rise of the sea-level. However, the foreland basin should progressively widen and pinchout migration rate would increase also if, for example, the orogenic wedge loaded a progressively stronger elastic lithosphere (Allen & Allen, 1992). Although not all local variations of relative sea-level oscillations and so the onlap rate could be identied from isochrones in the Figure 7c, they could be observed in the eld. Namely, the subaerial exposure surfaces that periodically interrupt the carbonate sedimentation of the Liburnia Formation reect relative sea-level falls. Short term falls (i.e. third-order cycles of Haq et al., 1987), which were documented in Late Maastrichtian, Late Pa leocene and Early Eocene (Haq et al., 1988), could cause these unconformities. On the other hand, a few other processes may inu ence the rate of the onlap. e forebulge should increase in height and migrate toward the orogenic wedge over time if the foreland lithosphere behaves viscoelastical ly even when the load is unchanging (Tankard, 1986). However, estimations for time constants of the viscous relaxation of stresses are longer than actual amount of time available for the forebulge migration (Allen & Al len, 1992; Dorobek, 1995). Variation in onlap rate may reect also changes in sediment supply, or within the orogenic wedge, such as the formation of a new thrust complex (Crampton & Allen, 1995) or transition from passive to active thrusting phase. An increase in com U PPER C RETACEOUS TO P ALEOGENE FORBULGE UNCONFORMIT Y ASSOCIATED W ITH FORELAND BASIN EVOLUTION

PAGE 14

TIME in KARST 2007 114 pressive in-plane stress produced during convergence also might enhance upli of the forebulge and causing shoreline regression along its ank (Allen & Allen, 1992; Dorobek, 1995). Evidence of short term sea-level oscillations could also be recognized from the specic evolution of the pa leokarst, especially phreatic caves. If the majority of len ticular caves with irregular walls and discrete horizons of spongework or swisse-cheese like vugs on young carbon ate islands originated at/in fresh/brackish water lenses (see Mylroie & Carew, 1995), then in our case the major part of the cavities had been emplaced in the vadose zone prior to submergence and burial. Namely, the caves are frequently completely lled with deposits originated in vadose zone, like owstone and bauxite, or they had been opened to the paleokarstic surface by complete denuda tion of the roof (i.e. rooess caves of Mihevc, 2001). If the water-level is stagnant and the forebulge migrates, than in the conceptual sense only those phreatic cavi ties developed below that forebulge ank that facing to wards back-bulge basin should be uplied in the vadose zone before subsidence. On the contrary, phreatic caves developed below the ank facing the foreland basin and the advancing orogenic wedge should suer nothing but subsidence and subsequent burial. eoretically it is possible that because of the advancing character of the forebulge, caves formed in dierent sides of the forebulge may occur in the same karstic prole. Phreatic cavities developed below the ank facing towards the back-bulge region should be uplied and modied in the vadose zone. Subsequently, aer the crest of the forebulge mi grates over the back-bulge ank, the back-bulge phre atic caves should re-immerge into phreatic zone, but this time below the ank facing towards the foreland basin. It is important to note that frequently observed multiphase modications of originally phreatic caves could also be the result of the same causes of relative sea-level oscil lations that govern the onlap character of the beds that overlie the unconformity (e.g. relaxation of the viscoelas tic bulge, formation of a new thrust complex, increase of horizontal in-plane stress, eustatic sea-level fall). e carbonate platform was subsequently re-estab lished and nally buried by prograding deeper-marine clastics (pelagic marls and ysch) of the migrating fore land basin (Fig. 9). As it has been already discussed, shal low-water carbonate successions that cover the uncon formity may yield a considerable amount of information about relative sea-level oscillations and geodynamics of the forebulge. Paralic/shallow-marine successions with frequent unconformities and palustrine deposits of the Liburnia Formation (Fig. 3) are usually much thicker in south western Slovenia and northeastern Istria than in central and western Istria (Fig. 2). ere the paleokarstic surface is frequently directly overlain by foraminiferal limestones (Matiec at al., 1996). e general trend of thickness and the rate of transition from shallow to deep marine envi ronments (drowning) (Fig. 2) are in good agreement with the rate of the onlap (Fig. 7c) and should be the result of the same processes that caused the dierentiations in the onlap pattern. I suggest that the anomalies in thickness and facies distribution that could be in places quite dis tinctive may arise from reactivation of inherited geologi cal structures due to the approaching orogenic wedge. It has been discussed already, that the orogenic phas es could be recognised from structural and stratigraphic data even in areas that are located at some distance from the source of tectonic activity at plate boundaries (e.g. collision and orogenesis). Because of later tectonic defor mations it is sometimes dircult to dene the exact tec tonic phase which aects the area and the actual source of tectonic activity. In our case, the structural and stratigraphic data in dicate the evolution of migrating synorogenic foreland basin complex, which should be the result of collision processes and the evolution of the advancing orogenic wedge (see e.g. Allen & Allen, 1992; Crampton & Allen, 1995; Miall, 1995). At rst sight it seems normal to link the foreland complex to tectonic phase that generated structures by mainly NE-SW compression (mesoalpine phase of some authors; see Doglioni & Bosellini, 1987) and gave rise to Dinaric mountain belt during its nal stages. However, the Dinaric orogenic belt of which nal upli occurred during the Oligocene-Miocene (Vlahovit et al., 2005) is supposed to be the result of collision be tween Tisia and Adria microplates with onset of colli sion during the Eocene (Pamit et al., 1998; Pamit, 2002), what is also the age of the oldest synorogenic deposits of the coastal part of the External Dinarides (Marjanac & osovit, 2000). On the contrary, although the nappe structures of western Slovenia and Late Cretaceous Pa leogene compressional deformations of northeastern Italy indicate NE-SW or ENE-W SW compression, and so Dinaric orientation of prevailing regional stress, the oldest foreland basin deposits in these regions are much older than those of other parts of the External Dinarides and belong to the latest stages of Late Cretaceous (Pavi, 1994; Doglioni, 1987; Doglioni & Bosellini, 1987). As it is shown on Figure 4 the age distribution of ysch de posits indicates the advancing nature of foreland basin from northeast towards southwest what is in accordance with Dinaric orientation of the prevailing regional stress. W hile south of Zagreb-Zemplen fault line, the remnants of oceanic lithosphere (i.e. ophiolite melange) as well as subduction and collision related rocks of In ternal Dinarides (i.e. the Sava-Vardar zone by Pamit et B OJAN O TONIfAR

PAGE 15

TIME in KARST 2007 115 al, 1998), which could be linked to closing processes of the Vardar Ocean and collision between Tisia and Adria (Pamit, 2000) are widespread, north of Zagreb-Zemplen line no such rock has been found so far. It seems possible that in central Slovenia, in prolongation of the Sava-Var dar zone, such rocks have been buried by Tertiary sedi ments and Southern Alpine nappes. In addition, on the NNE side the nappe structure of W estern Slovenia was cut from its root zone by Periadriatic fault. e root zone should be displaced for at least 100 km eastward during the Miocene (Ratschbacher et al., 1991; Frisch et al., 1998; Vrabec & Fodor, 2006). Although, the structural and sedimentary features of eoalpine tectonic phase which culminated in midCretaceous orogeny in the Austroalpine domain (Faupl & W agreich, 2000) and also aected the central and west ern part of the Italian Southern Alps (Doglioni, 1987; Doglioni & Bosellini, 1987) mostly pre-date the foreland related features and sediments described here, it should be noted that in Istria Tertiary tectonic cycle (from Eo cene on) display distinctively dierent orientation of the prevailing stress than Mesozoic one (Marinit & Matiec, 1991; Matiec et al., 1996). In conclusion, the foreland basin complex in west ern Slovenia and Istria was probably formed during me soalpine (Dinaric) tectonic phase, although some inu ences of eoalpine tectonic phase could be important in earlier stages of its evolution. e time discrepancy and also the exact orientation of prevailing regional stress are probably the result of oblique collision between Adria and Tisia microplates (and/or Austroalpine terrane?) and/or segmentation of the foreland plate (see Ricci-Luc chi, 1986; Allen & Allen, 1992). Oligocene to recent tectonic events especially in Di narides and Apennines, and conter-clockwise rotation of Adria importantly modied the area formerly occupied by the forebulge, but this is already beyond the scope of this paper. C ONCLUSIONS In spite of all structural and depositional heterogenei ties and subsequent tectonic deformation of the area the paleokarstic unconformity marked by distinctive surface and subsurface paleokarstic features exhibits characteris tics typical of a forebulge unconformity: 1) From northeast towards southwest the uncon formity cuts progressively older units which are onlapped by progressively younger shallow water carbonates; the chronostratigraphic gap progressively increases. 2) Deepening upward sequences of synorogenic ramp-like carbonate systems overlie the unconformity. In marginal parts of the former Adriatic Carbonate Plat form towards the foreland basin, a deepening upward sequence is documented also without intermediate un conformity here the sequence is conformable because the orogenic wedge was so close that the area experi enced only subsidence and forbulge upli had no taken place. 3) e foreland basin with siliciclastic turbiditic ysch deposits was developing synchronously with the forebulge and synorogenic carbonate platforms. It was also advancing synchronously in the same direction as they were forebulge and synorogenic carbonate plat forms. e stratigraphy overlying the unconformity (i.e. underlled trinity) representing subsidence in under lled peripheral foreland basin. 4) Evidence of contemporary seismic activity arises from periodic carbonate resediments (megabeds, olistostromes) nd in siliciclastic ysch successions. ey were supplied by turbiditic currents from the fault related escarpments of the forebulge slope (reactivated ancient faults). Besides exural upwarping because of the isostatic rebound on vertical loading of the foreland lithosphere, other smaller scale exural and non-exural deformations signicantly inuenced the evolution and appearance of the forebulge (incuding its diagenesis and karstication), lithofacies distribution and thickness of the carbonate successions above the unconformity. At least some inuence of eustatic sea-level oscillations can not be excluded. 5) e subaerially exposed area and the facies belts of progressive forelandward advancing shallowmarine, pelagic, and turbiditic depositional environ ments ahead of the orogenic front are roughly parallel to the Dinaric mountain chain. However, the Dinaric fore land-related system supposedly began to evolve during the Eocene when Tisia and Adria microplates began to collide what is much later comparing to Late Cretaceous onset of foreland basin evolution and forebulge upli in western Slovenia and Istria. In Istria the orientation of the prevailing regional stress during Cretaceous tectonic cycle diers signicantly from Eocene one. I suggest that the foreland basin complex in western Slovenia and Is tria was probably formed during mesoalpine (Dinaric) tectonic phase, due to oblique collision of Adria and Ti sia microplates (and/or Austroalpine terrane?) and seg mentation of the foreland plate. U PPER C RETACEOUS TO P ALEOGENE FORBULGE UNCONFORMIT Y ASSOCIATED W ITH FORELAND BASIN EVOLUTION

PAGE 16

TIME in KARST 2007 116 A CKNO W LEDGMENTS Most of this work was performed as part of doctoral dis sertation at University of Ljubljana. I am grateful to David Culver for English correction of an earlier version of the manuscript and Adrijan Koir for constructive reviews. R EFERENCES Allen, P.A. & Allen, J.R., 1992: Basin analysis principles & applications.Blackwell science, p. 451, London. Biondit, B., Braun, K., Vlahovit, I., Mlinar, b., Andrit, M., Balen, I., Pollak, D., Fuek, L., Otrit, N., Prtol jan, B., aban, B., Blagus, Z., Dukarit, F., Buljan, R. & Biondit, R., 1995: Inenjerskogeoloki model eljeznikog tunela itarija.In: I. Vlahovi, I. Velit & M. parica (Eds.), 1. hrvatski geoloki kongres: Zbornik radova, Opatija, 18.-21. oktober 1995. In stitut za geoloka istraivanja i Hrvatsko geoloko drutvo, 81-87, Zagreb. Bosellini, A., 1989: Dynamics of Tethyan carbonate plat forms.In: P.D. Crevello, W ilson, J.L., Sarg, J.F. & Read, J.F. (Eds.), Controls on carbonate platforms and basin development. Society of Economic Pale ontologists and Mineralogists, 3-13, Tulsa. Brazzatti, T., Caau, M., Cozzi, A., Cucchi, F., Drobne, K. & Pugliese, N., 1996: Padriciano section (Karst of Trieste, Italy).In: K. Drobne, Gorian & B. Kotnik. (Eds.), e role of impact processes in the geological and biological evolution of planet Earth: international workshop, Postojna, Slovenia, Septem ber 27 th October 2 nd 1996. Znanstvenoraziskovalni center SAZU, 189-198, Ljubljana. Bradley, D. C. & Kidd, W S. F., 1991: Flexeal extension of the upper continental crust in collisional fore dewps.Geological Society of America Bulletin, 103, 1416 1438. Buser, S. & Lukacs, E., 1972: Geologija boksitov v okolici Karojbe v Istri.Zbornik 7. kongresa geologa Jugo slavije, Zagreb, 1972., 13-30, Zagreb. Buser, S. & Lukacs, E., 1979: Bauxite in Slowenien.Ann. Inst. Geol. Hungarici, 54, 3, 209-220. Buser, S. & Radoiit, R., 1987: Dazikladacejske alge u srednjopaleocenskim miliolidskim krenjacima na Krasu u Sloveniji = Dasycladacean algae in Middle Paleocene miliolid limestones in Kras in Slovenia.Geologija, 28/29 (1985/1986), 69-91. Buser, S., 1986: Osnovna geoloka karta SFRJ. List Tol min in Videm (Udine) [Kartografsko gradivo]. 1:100.000. Zvezni geoloki zavod, Beograd. Buser, S., 1989: Development of the Dinaric and the Ju lian Carbonate Platforms and of the intermediate Slovenian Basin (NW Y ugoslavia).In: G. Battista Carulli, F. Cucchi in C. Pirini Radrizzani (Eds.), Evolution of the karstic carbonate platform: Relation with other periadriatic carbonate platforms, Trieste, 1987. Mem. Soc. Geol. It., 40, 313-320, Trieste. Buser, S., Grad, K. & Pleniar, M. 1967: Osnovna geoloka karta SFRJ. List Postojna [Kartografsko gradivo]. 1:100.000. Zvezni geoloki zavod, Beograd. Buser, S., Grad, K. & Pleniar, M. 1968: Osnovna geoloka karta SFRJ. List Gorica [Kartografsko gradivo]. 1:100.000. Zvezni geoloki zavod, Beograd. Cousin, M., 1981: Les rapports AlpesDinarides.Les conns de lItalie et de la Y ougoslavie. Soc. Gol. Nord, 5, 1. Crampton, S. L. & Allen, P. A., 1995: Recognition of fore bulge unconformities associated with early stage foreland basin development: example from the North Alpine Foreland Basin.AAPG Bulletin, 79, 10, 1495 1514. Debeljak, I., Koir, A. & Otoniar, B., 1999: A preliminary note on dinosaurs and non-dinosaurian reptiles from the Upper Cretaceous carbonate platform suc cesion at Kozina (SW Slovenia) = Preliminarno ob vestilo o dinozavrih in drugih reptilih iz zaporedja zgornjekrednih platformnih karbonatov pri Kozini (JZ Slovenija).Razpr. Slov. akad. znan. umet., Razr. naravosl. vede, 40, 3-25. Debeljak, I., Koir, A., Buetaut, E. & Otoniar, B., 2002: e Late Cretaceous dinosaurs and crocodiles of Kozina (SW Slovenia): a preliminary study.Mem. Soc. Geol. Ital., 57, 193-201. Della Vecchia, F., Tarlao, A., Tentor, M., Tunis, G. & Ven turini, S., 2000: First record of Hauterivian dino saur footprints in southern Istria (Croatia).In: I. Vlahovit & R. Biondit (Eds.), 2. hrvatski geoloki kongres: Zbornik radova, Cavtat-Dubrovnik, 17-20. maj 2000. Institut za geoloka istraivanja, 143-149, Zagreb. B OJAN O TONIfAR

PAGE 17

TIME in KARST 2007 117 Delvalle, D. & Buser, S., 1990: Macrofacies analysis of limestones from the Upper Cretaceous to the Lower Eocene of SW Slovenia (Y ugoslavia).Geologija, 31 (1988/1989), 351 394. Doglioni, C & Bosellini A., 1987: Eoalpine and Meso alpine tectonics in the Southern Alps. Geologische Rundschau, 76, 3, 735-754. Doglioni, C., 1987: Tectonics of the Dolomites (Southern Alps, Northern Italy).Journal of structural geology, 9, 2,181-193. Dorobek, S.L., 1995: Synorogenic carbonate platforms and reefs in foreland basins: controls on stratigraph ic evolution and platform/reef morphology.In: S.L. Dorobek & G.M. Ross (Eds.), Stratigraphic evolu tion of foreland basins: Special publication No. 52. SEPM, 127-147. Tulsa Drobne, K. & Pavlovec, R., 1979: Excursion K: Gole paleocene, ilerdian, cuisian.In: K. Drobne (Ed.), Geological development in Slovenia and Croatia: guidebook, 16 th european micropalaentological col loquium, Zagreb Bled, 8.. september 1979. Slovensko geoloko drutvo in Hrvako geoloko drutvo, 217-223, Ljubljana. Drobne, K., 1977: Alveolines paleogenes de la Slovenie et de lIstrie.Mem. suisses de Paleont., 99, p. 174. Drobne, K., 1979: Paleocene and Eocene beds in Slove nia and Istria.In: K. Drobne (Ed.), Geological de velopment in Slovenia and Croatia: guidebook, 16 th european micropalaentological colloquium, Zagreb Bled, 8.. september 1979. Slovensko geoloko drutvo in Hrvako geoloko drutvo, 49-64, Ljubljana Drobne, K., 1981: Znailne foraminifere in njih zdrube v podlagi danijskih plasti.Simpozij o problemih danija v Jugoslaviji: Zbornik referatov, 2, 85-97, Ljubljana. Drobne, K., Ogorelec, B., Dolenec, T., Marton, E. & Palinka, L., 1996: Biota and abiota at the K/T boundary in the Dolenja vas sections, Slovenia.In: K. Drobne, Gorian & B. Kotnik. (Eds.), e role of impact processes in the geological and biological evolution of planet Earth: international workshop, Postojna, Slovenia, September 27 th October 2 nd 1996. Znanstvenoraziskovalni center SAZU, 163181, Ljubljana. Drobne, K., Ogorelec, B., Pleniar, M., Zucchi-Stolfa, M. L. & Turnek, D., 1988: Maastrichtian, Danian and anetian Beds in Dolenja vas (NW Dinarides, Y ugoslavia): mikrofacies, foraminifers, rudists and corals.Razpr. Slov. akad. znan. umet., Razr. nara vosl. vede, 29, 147-224. Drobne, K., Turnek, D., Cimerman, F., Knez, M. & Og orelec, B., 1991: Introduction to the Paleogene SW Slovenia and Istria.IGCP project 286 Early Paleo gene bentos: eld-trip guidebook, p. 85, Ljubljana. Durn, G., Ottner, F., Tiljar, J., Mindszenty, A. & Barudija, U., 2003: Regional subaerial unconfor mities in shallow-marine carbonate sequences of Is tria: sedimentology, mineralogy, geochemistry and micromorphology of associated bauxites, palaeosols and pedo-sedimentary complexes.In: I. Vlahovit & J. Tiljar (Eds.), 22 nd IAS Meeting of Sedimentol ogy: eld trip guidebook, Opatija, 2003. Institute of geology, 209-254, Zagreb. Faupl, P. & W agreich, M., 2000: Late Jurassic to Eocene palaeogeography and geodynamic evolution of the Eastern Alps.Mitt. sterr. Geol. Ges. (1999), 92, 79-94. Frisch, W ., Kuhlemann, J., Dunkl, I. & Brugel, A., 1998: Palinspastic reconstruction and topographic evolu tion of the eastern Alps during late Tertiary tectonic extrusion.Tectonophysics, 297, 1-4, 1-15. Gabrit, A., Galovit, I., Saka, K. & Hvala, M., 1995: Min eral deposits of Istria Some deposits of bauxite, building stone and quartz sand.In: I. Vlahovit & I. Velit (Eds.), First Croatian geological congress: Excursion guide-book, Opatija, 18-21.10.1995. In stitut za geoloka istraivanja i Hrvatsko geoloko drutvo, 111 137, Zagreb. Grad, K. & Ferjani, L. 1974. Osnovna geoloka karta SFRJ. List Kranj [Kartografsko gradivo]. 1:100.000. Zvezni geoloki zavod, Beograd. Haas, J., Kovacs, S., Krystyn, L. & Lein, R., 1995: Signi cance of Late Permian Triassic facies zones in ter rane reconstructions in the alpine north Pannonian domain.Tectonophysics, 242, 1-2, 19-40. Hamrla, M., 1959: O pogojih nastanka premogi na kra su = On the conditions of origin of the coal beds in the karst region.Geologija, 5, 180-264. Hamrla, M., 1960: K razvoju in stratigraji produktivnih liburnijskih plasti Primorskega krasa.Rudarsko metalurki zbornik, 3, 203-216. Haq, B.U., Hardenbol, J. & Vail, P.R., 1987: Chronology of uctuating sea levels since the Triassic.Science, 235, 4793, 1156-1167. Haq, B.U., Hardenbol, J. & Vail, P.R., 1988: Mesozoic and Cenozoic chronostratigraphy and cycles of sea level change.In: C.K. W ilgus, B.S. Hastings, C.A. Ross, H. Posamentier, J. Van W agoner & C.G.C. Kendall (Eds.), Sea level changes: an integrated approach: Special publication No. 42. Society for Sedimentary Geology, 71-108, Tulsa. U PPER C RETACEOUS TO P ALEOGENE FORBULGE UNCONFORMIT Y ASSOCIATED W ITH FORELAND BASIN EVOLUTION

PAGE 18

TIME in KARST 2007 118 Hottinger, L. & Drobne, K., 1980: Early Tertiary conical imperforate foraminifera.Razprave SAZU, 22, 3, 189 276. Jurkovek, B., 1986: Osnovna geoloka karta SFRJ. List Beljak in Ponteba [Kartografsko gradivo]. 1:100.000. Zvezni geoloki zavod, Beograd. Jurkovek, B., Kolar-Jurkovek, T. & Ogorelec, B., 1997: Rezultati geoloke spremljave gradbenih del na odseku avtoceste med Divao in Kozino.Annales, 11, 161-186. Jurkovek, B., Toman, M., Ogorelec, B., ribar, L., Drobne, K., Poljak, M. & ribar, Lj., 1996: Formacijska geoloka karta 1:50.000 junega dela Trako-Ko menske planote.Intitut za geologijo, geotehniko in geoziko, p. 143, Ljubljana. Koir, A. & Otoniar, B., 1997: Diskonformnosti v zgorn jekredno paleogenskem zaporedju Jadransko Di narske karbonatne platforme v jugozahodni Sloveni ji.In: A. Horvat, M. Fister and N. Zupani (Eds.), 13. Posvetovanje slovenskih geologov: Povzetki referatov, Ljubljana, 4. in 5. april, 1997. Geoloki zbornik, 11, Naravoslovnotehnika fakulteta, Odd elek za geologijo, 15-16, Ljubljana. Koir, A. & Otoniar, B., 2001: e evolution of Upper Cretaceous and Paleogene synorogenic carbon ate platforms in NW Dinaric Foreland Basin.In: I. Dragievit & I. Velit (Eds.), 1 st scientic meet ing on carbonate platform or carbonate platforms of Dinarids: abstracts, Zagreb, Oct. 1 st -2 nd 2001. Rudarsko-geoloko-nani fakultet, Prirodoslovnomatemetiki fakultet, Institut za geoloka istraivanja i Hrvatsko geoloko drutvo, 62-63, Zagreb. Koir, A., 2003: Litostratigrafska revizija zgornje krede in paleogena v jugozahodni Sloveniji.In: A. Horvat (Ed.), 16. Posvetovanje slovenskih geologov, Raz prave. Poroila: Treatises. Reports (Geoloki zborn ik, 17). Univerza v Ljubljani, Naravoslovnotehnika fakulteta, Oddelek za geologijo, 92-98, Ljubljana. Koir, A., 2004: Microcodium revisited: Root calcica tion products of terrestrial plants on carbonate-rich substrates.Journal of sedimentary research, 74, 6, 845-857. Maga, N., 1965: Osnovna geoloka karta SFRJ. List Cres [Kartografsko gradivo]. 1:100.000. Zvezni geoloki zavod, Beograd. Maga, N., 1973: Osnovna geoloka karta SFRJ. 1:100.000. Tolma za list Cres.Zvezni geoloki zavod, p. 42, Beograd. Marinit, S. & Matiec, D., 1991: Tektonika i kinematika deformacija na primjeru Istre.Geoloki vjesnik, 44, 257 268. Marjanac, T. & osovit, V., 2000: Tertiary depositional history of eastern Adriatic realm.In: B. Tomljenovit, D. Balen & B. Sait (Eds.), Pancardi 2000: eldtrip guidebook, Dubrovnik, Croatia, 1-3.oktober 2000. Vijesti hrvatskoga geolokog drutva: special issue, 37, 2, 93-103, Zagreb. Martini, E., 1971: Standard Tertiary and Q uaternary cal careous nannoplankton zonation. In: A. Farinacci (Ed.), 2 nd International conference of planktonic microfossils: proceedings, Rome. Tecnoscience, 2, 739. Mrton, E., 2006: Paleomagnetic evidence for Tertiary counterclockwise rotation of Adria with respect to Africa.In: N. Pinter, G. Gyula, J. W eber, S. Stein & D. Medak (Eds.), e Adria Microplate: GPS Geod esy, Tectonics and Hazards, NATO Science Series: IV. Earth and Environmental Sciences. Springer Netherlands, 71-80. Mrton, E., Drobne, K., Cimerman, F., osovit, V. & Koir, A., 1995: Paleomagnetism of latest Maas trichtian through Oligocene rocks in Istria (Croa tia), the Karst region, and S of the Sava fault (Slo venia).In: I. Vlahovit, I. Velit & M. parica (Eds.), 1. hrvatski geoloki kongres: Proceedings, Opatija, 18-21.10.1995. Institut za geoloka istraivanja i Hrvatsko geoloko drutvo, 355-360. Matiec, D., Vlahovit, I., Velit, I.& Tiljar, J., 1996: Eocene limestones overlying lower Cretaceous deposits of W estern Istria (Croatia): Did some parts of present Istria form land during the Cretaceous?.Geologia Croatica, 49, 1, 117 127. Mauko, A. & Florjani, B., 2003: Dinosaur footprints in the Upper Turonian Coniacian limestone in the Krnica Bay (NE Istria, Croatia) = Stopinje dinoza vrov v zgornjeturonijsko coniacijskem apnencu v zalivu Krnica (NE Istra, Hrvaka).Geologija, 46, 1, 93-100. Mezga, A., Bajraktarevit, Z., Cvetko Teovit, B. & Guit, I., 2003: Dinosaur tracks as an evidence for the ter restriality in the Late Jurassic sediments of Istria, Croatia.In: I. Vlahovit (Ed.), 22 nd IAS Meeting of Sedimentology: abstracts book, Opatija, 2003. Insti tute of geology, 126, Zagreb. Miall, A.D., 1995: Collision-related froeland basins.In: C.J. Busby & R.V. Ingersoll (Eds.), Tectonics of sedi mentary basins. Blackwell science, 393-424, Cam brige. Mihevc, A., 2001: Speleogeneza Divakega krasa.Zaloba ZRC, ZRC SAZU. Zbirka ZRC, 27, p. 180, Ljubljana. B OJAN O TONIfAR

PAGE 19

TIME in KARST 2007 119 Mylroie, J.E. & Carew, J.L., 1995: Karst development on carbonate islands.In: D. A. Budd, A. H. Saller & P. M. Harris (Eds.), Unconformities and porosity in carbonate strata. American Association of Petro leum Geologists, AAPG Memoir, 63, 55 76, Tulsa. Neugebauer, J., Greiner, B. & Appel, E., 2001: Kinematics of the Alpine-W est Carpathian orogen and palaeo geographic implications.Journal of the Geological Society, 158, 97-110. Otoniar, B. & Koir, A., 2001: Jurassic-Eocene geody namic evolution of northern Dinarids: Controls on carbonate depositional systems.In: I. Dragievit & I. Velit (Eds.), 1 st scientic meeting on carbonate plat form or carbonate platforms of Dinarids: abstracts, Zagreb, Oct. 1 st -2 nd 2001. Rudarsko-geoloko-nani fakultet, Prirodoslovno-matemetiki fakultet, In stitut za geoloka istraivanja i Hrvatsko geoloko drutvo, 59-61, Zagreb. Otoniar, B., 1997: Macroscopic paleokarstic features in Upper Cretaceous limestones of the AdriaticDinaric Carbonate Platform (SW Slovenia).In: P.-Y Jeannin (Ed.), 12 th International Congress of Speleology. Swiss Speleological Society (SSS/SGH), Speleo Projects, La Chaux-de-Fonds, Switzerland, 417-420. Otoniar, B., Koir, A., & osovit, V., 2003: CretaceousTertiary paleokarst from Koromano (Istria, NW Croatia): an example of paleokarstic denuded cave.In: I. Vlahovit (Ed.), 22 nd IAS Meeting of Sedimen tology: abstracts book, Opatija, 2003. Institute of geology, 146, Zagreb. Pamit, J., 2000: e Sava-Vardar zone (SVZ).In: B. Tomljenovit, D. Balen & B. Sait (Eds.), Pancardi 2000: Fieldtrip guidebook, Dubrovnik, Croatia, 1-3.oktober 2000. Vijesti hrvatskoga geolokog drutva special issue, 37, 2, 19-22, Zagreb. Pamit, J., 2002: e Sava-Vardar Zone of the Dinarides and Hellenides versus the Vardar Ocean.Eclogae Geologicae Helvetiae, 95, 1, 99-113. Pamit, J., Guiit, I. & Jelaska, V., 1998: Geodynamic evo lution of the Central Dinarides.Tectonophysics, 297, 1-4, 251-268. Pamit, J., Guit, I. & Jelaska, V., 2000: Basic geological features of the Dinarids and South Tisia.In: B. Tomljenovit, D. Balen & B. Sait (Eds.), Pancardi 2000: Fieldtrip guidebook, Dubrovnik, Croatia, 1-3.oktober 2000. Vijesti hrvatskoga geolokog drutva special issue, 37, 2, 9-18, Zagreb. Pavlovec, R., Knez, M., Drobne, K. & Pavi, J., 1991: Pro les: Koana, Sv. Trojica and Leskovec; the disinte gration of the carbonate platform.In: K. Drobne, D. Turnek, F. Cimerman, M. Knez & B. Ogorelec (Eds.), Introduction to the Paleogene SW Slovenia and Istria: eld-trip guidebook, 69-72, Ljubljana. Pavi, J., 1994: Biostratigraphy of Cretaceous, Paleocene and Eocene clastics of Slovenia = Biostratigraja krednih, paleocenskih in eocenskih klastitov Slo venije.Razpr. Slov. akad. znan. umet., Razr. nara vosl. vede, 35, 66-84. Placer, L., 1981: Geoloka zgradba jugozahodne Slovenije = Geologic structure of southwestern Slovenia.Ge ologija, 24, 27 60. Placer, L., 1999: Contribution to the macrotectonic sub division of the border region between Southern Alps and External Dinarides = Prispevek k makrotekton ski rajonizaciji mejnega ozemlja med Junimi Alpa mi in Zunanjimi Dinaridi.Geologija, 41, 223-255. Placer, L., 2002: Predhodna objava rezultatov strukturne ga proliranja Krakega roba in Istre (AC Kozina Srmin, Seovlje) = Preliminary results of structural proling of the Kras edge and Istria (AC Kozina Srmin, Seovlje).Geologija, 45, 1, 277-278. Pleniar, M., Polak, A. & ikit, D., 1969: Osnovna geoloka karta SFRJ. List Trst [Kartografsko gradi vo]. 1:100.000. Zvezni geoloki zavod, Beograd. Pleniar, M., Polak, A. & ikit, D., 1973: Osnovna geoloka karta SFRJ. 1:100.000. Tolma za list Trst.Zvezni geoloki zavod, p. 68. Beograd. Polak, A. & ikit, D., 1973: Osnovna geoloka karta SFRJ. 1:100.000. Tolma za list Rovinj.Zvezni geoloki zavod, p. 51, Beograd. Polak, A., 1970: Osnovna geoloka karta SFRJ. 1:100.000. Tolma za list Pula.Zvezni geoloki zavod. p. 44, Beograd. Ratschbacher, L., Frisch, W ., Linzer, H.G. & Merle, O., 1991: Lateral extrusion in the Eastern Alps 2: struc tural analysis.Tectonics, 10, 2, 257-271. Ricci Lucchi, F., 1986: e Oligocene to Recent basins of the northern Apennines.In: P.A. Allen & P. Home wood (Eds.), Foreland basins. Spec. Publs. Int. Ass. Sediment. IAS. Blackwell, p. 105-139, Oxford. Rinar, I., 1997: Geologija okolice Postojne.magistrsko delo, p. 78, Ljubljana. Serra-Kiel, J. et al., 1998: Larger foraminiferal biostratig raphy of the Tethyan Paleocene and Eocene.Bul letin De La Socit Gologique De France, 169, 2, 281-299. Sinclair, H.D., 1997: Tectonostratigraphic model for un derlled peripheral foreland basins: An Alpine per spective.Geological Society of America Bulletin, 109, 3, 324-346. Skaberne, D., 1987: e Paleocene megaturbidites An hovo.In: E. Colizza (Ed.), E. Friuli, Karst of Gorizia and of W Slovenia: guidebook, Trieste, 5 th and 6 th June 1987. Universita degli studi, Istituto di geologia e paleontologia, 37-45, Trieste. U PPER C RETACEOUS TO P ALEOGENE FORBULGE UNCONFORMIT Y ASSOCIATED W ITH FORELAND BASIN EVOLUTION

PAGE 20

TIME in KARST 2007 120 Stampi, G. M., Mosar, J., Marquer, D., Marchant, R., Baudin T. & Borel, G., 1998: Subduction and obduc tion processes in the Swiss Alps.Tectonophysics, 296, 1-2, 159-204. ikit, D. & Polak, A., 1973: Osnovna geoloka karta SFRJ. 1:100.000. Tolma za list Labin.Zvezni geoloki za vod. p. 55, Beograd. ikit, D., Pleniar, M. & parica, M., 1972: Osnovna geoloka karta SFRJ. List Ilirska Bistrica [Karto grafsko gradivo]. 1:100 000. Zvezni geoloki zavod, Beograd. ikit, D., Polak, A. & Maga, N., 1968: Osnovna geoloka karta SFRJ. List Labin [Kartografsko gradivo]. 1:100.000. Zvezni geoloki zavod, Beograd. ribar, L., 1995: Evolucija Gornjekredne Jadransko-Di narske karbonatne platforme u jugozapadnoj Slo veniji.magistrski rad, p. 89, Zagreb. Tankard, A.J., 1986: On the depositional response to thrusting and lithospheric exure: examples from the Appalachian and Rocky Mountain basins.In: P.A. Allen and P. Homewood (Eds.), Foreland ba sins. IAS, Blackwell Scientic Publications, p. 369392, Oxford, Boston. Tarlao, A., Tunis, G. & Venturini, S., 1995: Lutetian transgression in central Istria: the Rogoviti-Meari section case.In: I. Vlahovi, I. Velit & M. parica (Eds.), 1. hrvatski geoloki kongres: zbornik radova, Opatija, 18.-21. oktober 1995, Institut za geoloka istraivanja i Hrvatsko geoloko drutvo, 613-618, Zagreb. Tiljar, J., Vlahovit, I., Matiec, D. & Velit, I., 1995: Plat formni faciesi od gornjega titona do gornjega alba u zapadnoj Istri i prijelaz u tempestite, klinoformne i rudistne biolititne faciese donjega cenomana u junoj Istri (ekskurzija B).In: I. Vlahovit & I. Velit. (Eds.), 1. hrvatski geoloki kongres: vodi ekskurzija, Opati ja, 18.-21.10.1995. Institut za geoloka istraivanja i Hrvatsko geoloko drutvo, 67-110, Zagreb. Tiljar, J., Vlahovit, I., Velit, I., Matiec, D. & Robson J., 1998: Carbonate facies evolution from the Late Albian to Middle Cenomanian in Southern Istria (Croatia): Inuence of synsedimentary tectonics and extensive organic carbonate production.Fa cies, 38, 137. Tunis, G. & Venturini, S., 1987: Volzana limestone and Drenchia unit p.so Solarie (Colovrat); Megabeds and turbidites sequence of Eastern Friuli (the italce menti quarry of Vernasso).International syposium on the Evolution of the karstic carbonate platform: relation with other periadriatic carbonatne plat forms: excursion guide-book, Trieste, 1.-6. 6. 1987. Universita degli studi di Trieste, Institut odi geolo gia e paleontologia, 3-12, Trieste. Turnek, D. & Drobne, K., 1998: Paleocene corals from the northern Adriatic Platform.In: L. Hottinger & K. Drobne (Eds.), Paleogene shallow benthos of the Tethys. 2. Slovenska akademija znanosti in umet nosti, Dela. Znanstvenoraziskovalni center SAZU, Paleontoloki intitut Ivana Rakovca, 34, 2, 5, 129154, Ljubljana. Velit, I. & Vlahovit, I., 1994: Foraminiferal assemblag es in the Cenomanian of the Buzet-Savudrija area (Northwestern Istria, Croatia).Geologia Croatica, 47, 1, 25 43. Velit, I., Tiljar, J. & Soka, B., 1989: e variability of thicknesses of the Barremian, Aptian and Albian carbonates as a consequence of changing deposi tional environments and emersion in W estern Istria (Croatia, Y ugoslavia).Mem. Soc. Geol. Ital. (1987), 40, 209 218. Vlahovit, I., Tiljar, J., Velit, I. & Matiec, D., 2005: Evo lution of the Adriatic Carbonate Platform: Palaeo geography, main events and depositional dynam ics.Palaeogeography Palaeoclimatology Palaeo ecology, 220, 3-4, 333-360. Vrabec, M. & Fodor, L., 2006: Late Cenozoic tectonics of Slovenia: structural styles at the Northeastern corner of the Adriatic microplate. In: N. Pinter, G. Grenerczy, J. W eber, S. Stein & D. Medak (Eds.), e Adria microplate: GPS geodesy, tectonics and haz ards. NATO Science Series, IV, Earth and Environ mental Sciences. Springer, 151-168, Dordrecht. W hite, W .B., 2000: Dissolution of limestone from eld observations.In: A.B. Klimchouk, D.C. Ford, A.N. Palmer & W Dreybrodt (Eds). Speleogenesis Evo lution of karst aquifers. National speleological soci ety, 149-155, Huntsville. Ziegler, P.A., Cloetingh, S. & Van W ees, J.D., 1995: Dy namics of intra-plate compressional deformation: e Alpine foreland and other examples.Tectono physics, 252, 1-4, 7-59. B OJAN O TONIfAR



PAGE 1

A R EVIE W OF C OALESCED C OLLAPSED P ALEOCAVE SY STEMS AND A SSOCIATED S UPRASTRATAL D EFORMATION M EDSEBOJNO ZDRUbENI PORUENI PALEOKRAKI JAMSKI SISTEMI IN DEFORMACIJE NAD NJIMI LEbEfIH PLASTI PREGLED Robert G. L OUCKS 1 Izvleek UDK 551.44 Robert G. Loucks: Medsebojno zdrueni porueni paleokraki jamski sistemi in deformacije nad njimi leeih plasti pregled Medsebojno zdrueni porueni paleokraki jamski sistemi in deformacije nad njimi leeih plasti predstavljajo izrazite dia genetsko/strukturne oblike karbonatnih zaporedij v bliini ses tavljenih geolokih nezveznosti. Osnovno zgradbo posamezne ga sistema lahko razdelimo na dva dela. Spodnji zakraseli del, kjer je gostota jam velika, je ohranjen v obliki masivnih bre, ki pogosto kaejo v tlorisu vzorec sestavljen iz ravnih odsekov. Za deformirane plasti, ki prekrivajo poruene jamske sisteme, so znailne velike skledaste do katlaste uleknine, ki jih sekajo pre lomi in razpoke. Regionalno gradijo zdrueni paleokraki jamski sistemi tega tipa vzorec velikega merila (zajemajo obmoja velika stotine do tisoe kvadratnih kilometrov), sestavljen iz ravnih odsekov in vkljuuje obmoja zgoenih zdruenih breastih teles, loenih z relativno neprizadeto prikamnino. Tak vzorec lah ko kae na razvoj paleokrakega jamskega sistema vzdol razpok linskih con. Porueni paleokraki jamski sistemi predstavljajo velike kompleksne pojave, ki odraajo organiziranost velikega merila. Za opredelitev zgradbe in razprostranjenosti popolnega paleokrakega jamskega sistema teh dimenzij potrebujemo po datke seizminih raziskav ali izdanke dimenzij gorovja. Kljune besede: pelokraki jamski sistemi, paleokras, defor macije, jamski sistemi. 1 Bureau of Economic Geology John A. and Katherine G. Jackson School of Geosciences, e University of Texas at Austin, Uni versity Station Box X, Austin, Texas 78713-8924 U.S.A., Fax: 512-471-0140 email: bob.loucks@beg.utexas.edu Received/Prejeto: 27.11.2006 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 121 Abstract UDC 551.44 Robert G. Loucks: A Review of Coalesced, Collapsed-Paleo cave Systems and Associated Suprastratal Deformation Coalesced, collapsed-paleocave systems and associated supra stratal deformation appear to be prominent diagenetic/struc tural features in carbonate sections at/near composite uncon formities. e basic architecture of the system can be divided into two sections. e lower karsted section, where high-den sity cave formation took place, is preserved as massive breccias commonly displaying a rectilinear pattern in map view. e overlying suprastratal deformation section is characterized by large, circular to linear sag structures containing faults and fractures. Regional distribution of coalesced, collapsed-cave systems commonly appears as large-scale (hundreds to thou sands of square kilometers in area), rectilinear patterns with areas of concentrated, coalesced breccias separated by relatively undisturbed host rock. is pattern may reect development of the paleocave system along fracture swarms. Collapsed-paleocave systems are large, complex features that show broad-scale organization. e complete paleocave system may need seismic data or large, mountain-scale outcrops to de ne their architecture and distribution. Key Words: Paleocaves, Paleokarst, karst, suprastratal defor mation, cave systems.

PAGE 2

TIME in KARST 2007 122 I NTRODUCTION At several composite unconformities in the stratigraphic record, carbonate sections display extensive karsting that leads to multiple development of cave systems (Esteban, 1991). ese cave systems underwent extensive collapse and mechanical compaction with burial. Deformation of the overlying strata is associated with burial collapse of the cave system. e eects of this suprastratal deforma tion can be noted 700+ m up section above the karsted interval. is review will describe the evolution of cave sys tems during burial and what the characteristics of the cave systems are at dierent stages of burial. Also the characteristics of suprastratal deformation will be de scribed. Paleocave systems have been investigated by sev eral authors including Lucia (1968, 1995, 1996), Loucks and Anderson (1980, 1985), Kerans (1988, 1989, 1990), W ilson et al. (1991) W right et al. (1991), Candelaria and Reed (1992), Loucks and Handford (1992), Lucia et al. (1992), Kerans et al. (1994), Hammes et al. (1996), Maz zullo and Chilingarian (1996), McMechan et al. (1998), Loucks (1999, 2001, 2003), Loucks et al. (2000, 2004), Loucks and Mescher (2001), McMechan et al. (2002), and Combs et al. (2003). e review will mainly synthe size material from these studies. C LASSIFICATIONS OF C AVE P RODUCTS AND F ACIES Loucks (1999) and Loucks and Mescher (2001) pro duced classications of cave products and cave facies. Loucks (1999) used a ternary diagram (Fig. 1) to show the relationships between crackle breccias, mosaic bre ccias, chaotic breccias, and cave sediments. Crackle breccias are highly fractured rock, with thin fractures separating the clasts and only minor displacement ex isting between the clasts. Mosaic breccias show more displacement than crackle breccias, but the clasts can still be tted back together. Chaotic breccias are com posed of mixtures of clasts that have been transported vertically by collapse or laterally by uvial or densityow mechanisms. Clasts show no inherent association with their neighbors. Chaotic breccias grade from ma trix-free, clast-supported breccias; through matrix-rich, clast-supported breccias; to matrix-rich, matrix-sup ported breccias. Cave-sediment ll can consist of any material, texture, or fabric. Loucks and Mescher (2001) proposed a classication of six common paleocave facies (Fig. 2): (1) Undisturbed strata, which are interpreted as un disturbed host rock. In this facies bedding continuity is excellent for tens of hundreds of meters. (2) Dis turbed strata that are disturbed host rock around the collapsed passage. Bedding continuity is high, but it is folded and oset by small faults. It is commonly overprinted by crackle and mosaic brecciation. (3) Highly disturbed strata, which is collapsed host rock adjacent to or immediate ly above passages. (4) Coarse-clast chaotic breccia that is interpreted as collapsed-breccia cavern ll pro duced by ceiling and wall collapse. It is characterized by a mass of very poorly sorted, granuleto bouldersized chaotic-breccia clasts approxi mately 0.3 to 3 m long that form a ribbon-to tabular-shaped body as much as 15 m across and hundreds of meters long. It is commonly clast Fig. 1: Cave-sediment lls and breccias can be separated into three end members: crackle breccia, chaotic breccia, and cave-sediment ll. M odied from Loucks (1999) and reprinted by permission of the AAPG whose permission is required for further use. R OBERT G. L OUCKS

PAGE 3

TIME in KARST 2007 123 supported, but can contain matrix material. (5) Fine-clast chaotic brec cia interpreted as laterally (hydro dynamically) sorted, transportedbreccia cavern ll. Characterized by a mass of clast-supported, mod erately sorted, granuleto cobblesized clasts with varying amounts of matrix. Clasts can be imbricated or graded. Resulting bodies are ribbonto tabular-shaped and are as much as 15 m across and hundreds of meters long. (6) Cave-sediment cavern ll that can be carbonate and/or silici clastic debris of any texture or fabric and commonly displaying sedimen tary structures. Fig. 2: Six basic cave facies are recognized in a paleocave system and are classied by rock fabrics and structures. M odied from Loucks and M escher (2001) and reprinted by permission of the AAPG whose permission is required for further use. E VOLUTION OF C AVE P ASSAGES Knowledge of the processes by which a modern cave passage forms at the surface and evolves into a col lapsed paleocave passage in the subsurface is necessary to understand the features of paleocave systems. Loucks (1999) described this evolutionary process (Fig. 3), and the review pre sented here is mainly from that in vestigation. A cave passage is a product of near-surface karst processes that in clude dissolutional excavation of the passage, partial to total breakdown of the passage, and sedimentation in the passage (Fig. 4). During lat er-burial cave collapse, mechanical compaction takes place. Fig. 3: Schematic diagram showing evolution of a single cave passage from its formation in the phreatic zone of a nearsurface karst environment to burial in the deeper subsurface. M odied from Loucks (1999) and reprinted by permission of the AAPG whose permission is required for further use. A R EVIE W OF C OALESCED C OLLAPSED P ALEOCAVE SY STEMS AND A SSOCIATED S UPRASTRATAL D EFORMATION

PAGE 4

TIME in KARST 2007 124 Initial passages form in phreatic and/or vadose zones (Fig. 3). Passages are excavated where surface recharge is concentrated by preexisting pore systems, such as bed ding planes or fractures (Palmer, 1991), that form a con tinuous link between groundwater input, such as sink holes, and groundwater output, such as springs (Ford, 1988). Cave passages are under stress from the weight of Fig. 4: B lock diagram of a near-surface modern karst system. e diagram depicts four levels of cave development (upper-right corner of block model), with some older passages (shallowest) having sediment ll and chaotic breakdown breccias. M odied from Loucks (1999) and reprinted by permission of the AAPG whose permission is required for further use. overlying strata. A tension dome, or zone of maximum shear stress, is induced by the presence of the passage or cavity (W hite, 1988). Stress is relieved by collapse of the rock mass within the stress zone. is collapse produces chaotic breakdown breccia on the oor of the cave pas sage (Figs. 3 and 4). e associated stress release around the cavity produces crackle and mosaic breccias in the adjacent host rock. As cave-bearing strata are bur ied, extensive mechanical compac tion begins, resulting in collapse of the remaining void (Fig. 3). Multiple stages of collapse occur over a broad depth range. Meter-scale bit drops in wells (indication of cavernous pores) are not uncommon down to depths of 2,000 m and are observed to occur to depths of 3,000 m (Loucks, 1999). e collapsed passages become pods of chaotic breccia (Fig. 3). e ar eal cross-sectional extent of brec ciation and fracturing aer burial and collapse is greater than that of the original passage because the ad jacent fractured and brecciated host rock has become part of the brec ciated pod. Sag features, faults, and fractures (Fig. 3) occur over the col lapsed passages. E VOLUTION OF C OALESCED C OLLAPSED P ALEOCAVE SY STEMS A coalesced, collapsed-paleocave system can be divided into two parts: (1) a lower section of strata that contains collapsed paleocaves and (2) an upper section of strata that is deformed to varying degrees by the collapse and compaction of the section of paleocave-bearing strata (Fig. 5). e deformed upper section of strata is termed suprastratal deformation (Loucks, 2003) and is discussed in a later section. Cave systems are composed of numerous passages. If the areal density of passages is low, the collapsed cave system will feature isolated, collapsed passages (nonco alescing paleocave system; Fig. 6). If the cave system has a high density of passages, as is common at composite third-order unconformities (Esteban, 1991; Lucia, 1995; Fig. 6: Schematic diagram showing burial and collapse of lowdensity cave system (noncoalescing, collapsed-cave system) and reprinted by permission of the AAPG whose permission is required for further use. R OBERT G. L OUCKS

PAGE 5

TIME in KARST 2007 125 Fig. 5: Schematic diagram showing the stages of development of a coalesced, collapsed-paleocave system. M odied from Loucks et al., (2004) and reprinted by permission of the AAPG whose permission is required for further use. Loucks, 1999), then upon burial and collapse the system can form largescale, coalesced, brecciated and fractured breccia bodies upon burial and collapse that are the amalgama tion of many passages and interven ing disturbed host rock (coalescing paleocave system; Fig. 5). e bodies are hundreds to several thousands of meters across, thousands of meters long, and tens of meters to more than 100 m thick. Internal spatial complexity is high, resulting from the collapse and coalescence of nu merous passages and cave-wall and cave-ceiling strata. S UPRASTRATAL D EFORMATION Collapse and compaction of cave systems provide poten tial for development of large-scale fracture/fault systems that can extend from the collapsed interval upward to more than 700 m (Kerans, 1990; Hardage et al., 1996a; Loucks, 1999, 2003; McDonnell et al. in press). ese fracture/fault systems are not related to regional tectonic stresses. Large-scale suprastratal deformation occurs above the collapsed-cave system. As the cave system collapses during burial, overlying strata will sag or subside over the collapsed area. is phenomenon is well documented in mining literature (Kratzsch, 1983; W ittaker and Red dish, 1989). Kratzsch (1983, p. 147) presented a diagram (Fig. 7) that shows the stress eld above a collapsed mine passage and associated subsidence. e overlying stress eld widens from the edges of the excavation, and the overlying strata are under compression directly over the excavation. Near the edges of the excavation, between a vertical line extending from the edge of the cavity and the limit line, strata are under extension (tension). W ithin this zone of stress the overlying strata have the poten tial to sag, creating faults and fractures for some distance upward, depending on the mechanical properties of the strata and the thickness of the beds within the strata. Fig. 8 is a scatterplot showing a number of examples of the magnitude of subsidence over coal mines. e graph in dicates that subsidence is recorded at horizons more than 800 m above the cavity. ese data indicate the magni tude of the eect that the collapse of a cavity can have on overlying strata. Applying the above concept of stress elds over cav ities to the collapse of a cave passage during burial sug Fig. 7: D iagram of a collapsed mine showing collapsed breccia zone and suprastratal deformation. e center of the subsidence trough is under compression, whereas the wings are under extension. M odied from Kratzch (1983). A R EVIE W OF C OALESCED C OLLAPSED P ALEOCAVE SY STEMS AND A SSOCIATED S UPRASTRATAL D EFORMATION

PAGE 6

TIME in KARST 2007 126 gests that similar stress elds will develop. As the cave passage collapses, it has the potential to aect a consider Fig. 8: Scatterplot showing thickness of overburden that can be aected by mine collapse. Graph shows a trend of greater subsidence with less overburden. able number of overlying strata. W ithin a cave system, numerous passages will collapse with burial. Each pas sage will develop a stress eld above it, and these stress elds will interact to create a larger, combined stress eld. is concept was presented by W ittaker and Reddish (1989; p. 47), who detailed instances in which multiple mining excavations are collapsing. e stress eld above a collapsing cave system will be complex because the dif ferent cave passages do not collapse and compact uni formly over time. As local areas collapse, dierent stress elds will develop, producing fractures and faults related to that individual stress eld. Resulting suprastratal de formation will show variable fracture and fault patterns within an overall subsidence sag. A unique circular fault pattern above collapsed cave systems is recognized by cy lindrical faults (Hardage et al. 1996a; Loucks, 1999; Mc Donnell et al. in press). M EGASCALE A RCHITECTURE P ATTERNS OF C OALESCED C OLLAPSED P ALEOCAVE SY STEMS Coalesced, collapsed-paleocave systems are megascale diagenetic/structural features that can aect more than 700 m of section and be regional in scale. As discussed earlier, the karsted section reects the coalescing of col lapsed breccias that formed by collapse of passages and associated disturbed host rock. e vertical extent of the breccias commonly aects the upper 100 m of section (Loucks and Handford, 1992; Loucks 1999) and as much as 300 m of the total section (Lucia, 1996). e intensity of brecciation can vary throughout the aected interval. Kerans (1990), Loucks (1999), Loucks et al. 2004), and many others have published descriptions of collapsed, brecciated paleocave zones. Fig. 9 shows examples of cave facies from the Lower Ordovician Ellenburger Group in central Texas (Loucks et al. 2004). e regional pattern of the collapsed paleocave system is commonly rectilinear (Loucks, 1999). is rectilinear pattern is probably an artifact of the original cave system developing along an early-formed fracture system. In a detailed study of a paleocave system in the Fig. 10: Slice map through a collapsed-paleocave system in the Lower Ordovician Ellenburger Group in central T exas. M odied from Loucks (2004) and reprinted by permission of the AAPG whose permission is required for further use. R OBERT G. L OUCKS

PAGE 7

TIME in KARST 2007 127 Fig. 9: Representative cores from paleocave facies. (a) Crackle-fractured disturbed host rock. (b) Collapsed chaotic breccia with large slabs and cave-sediment ll. (c) T ransported chaotic breccias in carbonate cave-ll matrix. Sample on right is under UV light. Samples from Lower Ordovician Ellenburger Group in central T exas. M odied from Loucks (2004) and reprinted by permission of the AAPG whose permission is required for further use. Lower Ordovician in central Texas, Loucks et al. (2004) presented maps (Fig. 10) and cross sections of the threedimensional, ne-scale architecture of a coalesced, col lapsed-paleocave system. e coalesced, collapsed-pas sage breccias range in size to as much as 350 m and are separated by disturbed and undisturbed host rock rang ing in size up to 200 m. Lucia (1995) also presented a map of brecciated collapsed passages (Fig. 11) from out crops in the Franklin Mountains of far west Texas, which displays a crude rectilinear pattern. is rectilinear pattern can be seen on seismic data as well. Loucks (1999) presented seismic-based maps from Benedum eld in W est Texas that display a rectilin ear pattern of sags and circular faults induced by collapse of the Ellenburger paleocave system below (Fig. 12). A similar rectilinear pattern is evidenced on seismic data in Boonsville eld (Fig. 13) in the northern Fort W orth Basin in Texas (Hardage et al. 1996a; McDonnell et al. in press). In both the Benedum and Boonesville datasets, suprastratal deformation aects up to 700 m of section above the karsted interval (Figs. 12 and 13). A R EVIE W OF C OALESCED C OLLAPSED P ALEOCAVE SY STEMS AND A SSOCIATED S UPRASTRATAL D EFORMATION

PAGE 8

TIME in KARST 2007 128 Fig. 11: (a) Photograph of the Great M cKelligon Sag in the Franklin M ountains of far West T exas. Photograph and general interpretation are from Lucia (1995) but have been modied by current author. is outcrop is an outstanding example of a collapsedpaleocave system with associated overlying suprastratal deformation. (b) M ap produced by Lucia (1995) of several paleocave systems within the Franklin M ountains. Paleocave trend lines are by current author. C ONCLUSIONS Coalesced, collapsed-paleocave systems are megascale diagenetic/structural features that can aect more than 700 m of section and be regional in scale. e architec ture of the complete system can be divided into the lower collapsed zone, where the dense system of caves formed and collapsed with later burial, producing a complex zone of brecciation. e upper, suprastratal deformation section formed during the collapse of the karsted section. e overlying strata were generally lithied, but the sag also aected concurrent sedimentation patterns (Hard age et al. 1996b). e deformation in the deformed su prastratal zone consists of normal, reverse, and cylindri cal faults and fractures (Loucks, 1999; McDonnell et al. in press). It is important to emphasize that large-scale structural features can develop above karsted zones and not be related to regional tectonic stresses. R OBERT G. L OUCKS

PAGE 9

TIME in KARST 2007 129 Fig. 12: 3-D seismic example over an Ellenburger paleocave system from B enedum eld in West T exas. (a) Second-order derivative map in the Fusselman interval displaying sag zones produced by Ellenburger paleocave collapse. (b) Seismic line showing missing sections (collapse in Ellenburger section), cylindrical faults, and sag structures. Suprastratal deformation is >1,000 thick in this section. M odied from Loucks (1999) and reprinted by permission of the AAPG whose permission is required for further use. Coalesced, collapsed-paleocave systems and associ ated suprastratal deformation are complex systems, and large-scale outcrops or datasets are necessary to dene them. However, with the model presented in this paper, individual data points can lead to recognition that the system is a coalesced, collapsed-paleocave feature. A R EVIE W OF C OALESCED C OLLAPSED P ALEOCAVE SY STEMS AND A SSOCIATED S UPRASTRATAL D EFORMATION

PAGE 10

TIME in KARST 2007 130 Fig. 13: Suprastratal deformation sag features in post-Lower Ordovician Ellenburger strata in Fort Worth B asin in north T exas. (a) Curvature map at M ississippian Forestburg Limestone horizon displaying sag features and faults produced by collapse in the Ellenburger interval. From M cD onnell et al. (in press). (b) 3D seismic line at 1:1 scale showing sag features produced by paleocave collapse in the Ellenburger section. Line-of-section location is shown by dashed line in Fig. 13a. A CKNO W LEDGEMENTS I would like to express my appreciation to Angela Mc Donnell for reviewing this manuscript. Lana Deiterich edited the text. Published with the permission of the Di rector, Bureau of Economic Geology, John A. and Kath erine G. Jackson School of Geosciences, e University of Texas at Austin. R OBERT G. L OUCKS

PAGE 11

TIME in KARST 2007 131 R EFERENCES Candelaria, M. P. & C. L. Reed, eds., 1992: Paleokarst, karst related diagenesis and reservoir development: examples from Ordovician-Devonian age strata of W est Texas and the Mid-Continent.Permian Basin Section SEPM Publication No. 92-33, p. 202. Combs, D. M., R. G. Loucks, & S. C. Ruppel, 2003: Lower Ordovician Ellenburger Group collapsed paleocave facies and associated pore network in the Barnhart eld, Texas.in T. J. Hunt & P. H. Luolm, e Permian Basin: back to basics: W est Texas Geologi cal Society Fall Symposium: W est Texas Geological Society Publication No. 03-112, 397-418. Ford, D. C., 1988: Characteristics of dissolutional cave systems in carbonate rocks.in N. P. James & P. W Choquette, eds., Paleokarst: Springer-Verlag, Berlin, 25-57. Esteban, M., 1991: Palaeokarst: practical applications.in V. P. W right, M. Esteban, & P. L. Smart, eds., Pal aeokarst and palaeokarstic reservoirs: University of Reading, Postgraduate Research for Sedimentology, PRIS Contribution No. 152, 89-119. Hammes, Ursula, F. J. Lucia, & Charles Kerans, 1996: Reservoir heterogeneity in karst-related reservoirs: Lower Ordovician Ellenburger Group, W est Texas.in E. L. Stoudt, ed., Precambrian-Devonian geology of the Franklin Mountains, W est Texasanalogs for exploration and production in Ordovician and Silu rian karsted reservoirs in the Permian Basin: W est Texas Geological Society, Publication No. 96-100, 99-117. Hardage, B. A., D. L. Carr, D. E. Lancaster, J. L. Simmons Jr., R. Y Elphick, V. M. Pendleton, & R. A. Johns, 1996a: 3-D seismic evidence of the eects of carbon ate karst collapse on overlying clastic stratigraphy and reservoir compartmentalization.Geophysics, 61, 1336-1350. Hardage, B. A., D. L. Carr, D. E. Lancaster, J. L. Simmons, Jr., D. S. Hamilton, R. Y Elphick, K. L. Oliver, & R. A. Johns, 1996b: 3-D seismic imaging and seismic attribute analysis of genetic sequences deposited in low-accommodation conditions.Geophysics, 61, 1351-1362. Kerans, Charles, 1988: Karst-controlled reservoir hetero geneity in Ellenburger Group carbonates of W est Texas.reply: American Association of Petroleum Geologists Bulletin, 72, p. 1160-1183. Kerans, Charles, 1989: Karst-controlled reservoir hetero geneity and an example from the Ellenburger Group (Lower Ordovician) of W est Texas.e University of Texas at Austin, Bureau of Economic Geology Re port of Investigations No. 186, p. 40. Kerans, Charles, 1990: Depositional systems and karst geology of the Ellenburger Group (Lower Ordovi cian), subsurface W est Texas.e University of Texas at Austin, Bureau of Economic Geology Re port of Investigations No. 193, p. 63. Kerans, Charles, F. J. Lucia, & R. K. Senger, 1994: Inte grated characterization of carbonate ramp reser voirs using Permian San Andres outcrop analogs.American Association of Petroleum Geologists Bul letin, 78, 181-216. Kratzsch, H., 1983: Mining Subsidence Engineering.Springer-Verlag, Berlin, Translated by R. F. S. Flem ing, p. 543. Loucks, R. G., 1999: Paleocave carbonate reservoirs: ori gins, burial-depth modications, spatial complexity, and reservoir implications.American Association of Petroleum Geologists Bulletin, 83, 1795-1834. Loucks, R. G., 2001: Modern analogs for paleocave-sedi ment lls and their importance in identifying paleo cave reservoirs.Gulf Coast Association of Geologi cal Societies Transactions, 46, 195-206. Loucks, R. G., 2003: Understanding the development of breccias and fractures in Ordovician carbonate res ervoirs.in T. J. Hunt & P. H. Luolm, e Permian Basin: back to basics: W est Texas Geological Soci ety Fall Symposium: W est Texas Geological Society Publication No. 03-112, 231-252. Loucks, R. G. & J. H. Anderson, 1980: Depositional fa cies and porosity development in Lower Ordovician Ellenburger dolomite, Puckett Field, Pecos County, Texas.in R. B. Halley & R. G. Loucks, eds., Car bonate reservoir rocks: SEPM Core W orkshop No. 1, 1-31. Loucks, R. G. & J. H. Anderson: 1985, Depositional fa cies, diagenetic terrains, and porosity development in Lower Ordovician Ellenburger Dolomite, Puckett Field, W est Texas.in P. O. Roehl & P. W Choquette, eds., Carbonate petroleum reservoirs: Springer-Ver lag, 19-38. Loucks, R. G. & R. H. Handford, 1992: Origin and rec ognition of fractures, breccias, and sediment lls in paleocave-reservoir networks.in M. P. Candelaria & C. L. Reed, eds., Paleokarst, karst related diagen esis and reservoir development: examples from Or dovician-Devonian age strata of W est Texas and the Mid-Continent: Permian Basin Section SEPM Pub lication No. 92-33, 31-44. A R EVIE W OF C OALESCED C OLLAPSED P ALEOCAVE SY STEMS AND A SSOCIATED S UPRASTRATAL D EFORMATION

PAGE 12

TIME in KARST 2007 132 Loucks, R. G. & P. Mescher, 2001: Paleocave facies clas sication and associated pore types.American As sociation of Petroleum Geologists, Southwest Sec tion, Annual Meeting, Dallas, Texas, March 11-13, CD-ROM, p.18. Loucks, R. G., P. Mescher, & G. A. McMechan, 2000: Ar chitecture of a coalesced, collapsed-paleocave sys tem in the Lower Ordovician Ellenburger Group, Dean W ord Q uarry, Marble Falls, Texas.Final re port prepared for the Gas Research Institute, GRI00/0122, CD-ROM. Loucks, R. G., P. Mescher, & G. A. McMechan, 2004: ree-dimensional architecture of a coalesced, col lapsed-paleocave system in the Lower Ordovician Ellenburger Group, Central Texas.American As sociation of Petroleum Geologists Bulletin, 88, 545564. Lucia, F. J., 1968: Sedimentation and paleogeography of the El Paso Group.in W J. Stewart, ed., Delaware basin exploration: W est Texas Geological Society Guidebook No. 68-55, 61-75. Lucia, F. J., 1995: Lower Paleozoic cavern development, collapse, and dolomitization, Franklin Mountains, El Paso, Texas.in D. A. Budd, A. H. Saller, and P. M. Harris, eds., Unconformities and porosity in car bonate strata: American Association of Petroleum Geologists Memoir 63, 279-300. Lucia, F. J., 1996: Structural and fracture implications of Franklin Mountains collapse brecciation.in E. L. Stoudt, ed., Precambrian-Devonian geology of the Franklin Mountains, W est Texas-Analogs for explo ration and production in Ordovician and Silurian karsted reservoirs in the Permian basin: W est Texas Geological Society 1996 Annual Field Trip Guide book, W TGS Publication No. 96-100, 117-123. Lucia, F. J., Charles Kerans, & G. W Vander Stoep, 1992: Characterization of a karsted, high-energy, rampmargin carbonate reservoir: Taylor-Link W est San Andres Unit, Pecos County, Texas.e University of Texas at Austin, Bureau of Economic Geology Re port of Investigations No. 208, p. 46. Mazzullo, S. J. & G. V. Chilingarian, 1996: Hydrocarbon reservoirs in karsted carbonate rocks.in G. V. Chil ingarian, S. J. Mazzullo, & H. H. Rieke, eds., Car bonate reservoir characterization: a geologic-engi neering analysis, Part II: Elsevier, 797-685. McDonnell, A., R.G. Loucks, & T. Dooley, (in press): Q uantifying the origin and geometry of circular sag structures in northern Fort W orth Basin, Texas: paleocave collapse, pull-apart fault systems or hy drothermal alteration?.American Association of Petroleum Geologists Bulletin. McMechan, G. A., R. G. Loucks, P. A. Mescher, & X ia oxian Zeng, 2002: Characterization of a coalesced, collapsed paleocave reservoir analog using GPR and well-core data.Geophysics, 67, 1148-1158. McMechan, G. A., R. G. Loucks, X. Zeng, & P. A., Me scher, 1998: Ground penetrating radar imaging of a collapsed paleocave system in the Ellenburger dolo mite, Central Texas.Journal of Applied Geophys ics, 39, 1-10. Orchard, R. J., 1975: Prediction of the magnitude of sur face movements.in Proceedings, European Con gress on Ground Movement, 39-46. Palmer, A. N., 1991: Origin and morphology of limestone caves.Geological Society of America Bulletin, 103, 1-21. W hite, W B., 1988: Geomorphology and hydrology of karst terrains.Oxford University Press, New Y ork, p. 464. W ilson, J. L., R. L. Medlock, R. D. Fritz, K. L. Canter, & R. G. Geesaman, 1992: A review of Cambro-Orodovi cian breccias in North America.in M. P. Candelaria & C. L. Reed, eds., Paleokarst, karst related diagen esis and reservoir development: examples from Or dovician-Devonian age strata of W est Texas and the Mid-Continent: Permian Basin Section SEPM Pub lication No. 92-33, 19-29. W ittaker, B. N. & D. J. Reddish, 1989: Subsidence; Occur rence, Prediction and Control: Elsevier, Develop ment in Geotechnical Engineering, No. 56, p.528. W right, V. P., M. Esteban, & P. L. Smart, eds., 1991: Pal aeokarst and palaeokarstic reservoirs: Postgraduate Research for Sedimentology, University, PRIS Con tribution No. 152, p.158. R OBERT G. L OUCKS



PAGE 1

C LASTIC SEDIMENTS IN CAVES IMPERFECT RECORDERS OF PROCESSES IN KARST K LASTIfNI SEDIMENTI V JAMAH NEPOPOLNI ZAPIS KRAKIH PROCESOV Ira D. S ASO W SK Y 1 Izvleek UDK 552.517:551.7 Ira D. Sasowsky Klastini sedimenti v jamah nepopolni za pis krakih procesov be od nekdaj so klastini sedimenti pomembno orodje pri raz biranju geoloke zgodovine. V zgodnjih tudijah so uporabili naela stratigraje tudi pri raziskovanju jamskih sedimentov Glavnino jamskih sedimentov sestavljajo podori in aluvij. Ra ziskave aluvija so se uspeno izkazale pri dataciji jam, doloanju razvoja povrja, paleookolja in paleontologije. Zaradi monega tlanega toka in hitrih sprememb stanj, so jamski sedimenti drugani od povrinskih. To, poleg ostalih dejavnikov, pred stavlja teave pri interpretaciji zapisov, ki jih hranijo jame. Po drugi strani pa je obstojnost jamskih sedimentov dalja zaradi zavetja, ki jim ga nudi jama. Po vsem svetu poznamo tevilne uspene datacije jamskih sedimentov z magnetizmom ali izo topi. Jame znotraj erozijskega cikla vzdrijo do10 milijonov let, zato naj jamski sedimenti ne bi bili znatno stareji. K ljune besede: klastini sedimenti, paleoklima, sedimen tologija, stratigraja, datiranje. 1 Orce for Terrestrial Records of Environmental Change, Department of Geology and Environmental Science, University of Ak ron, Akron, OH 44325-4101, USA. Received/Prejeto: 24.01.2007 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 143 Abstract UDC 552.517:551.7 Ira D. Sasowsky: Clastic sediments in caves imperfect re corders of processes in karst Clastic sediments have played an important role in deciphering geologic history and processes since the inception of the disci pline. Early studies of caves applied stratigraphic principles to karst deposits. e majority of cave deposits are breakdown and alluvium. e alluvial materials have been successfully investi gated to determine ages of caves, landscape evolution, paleoen vironmental conditions, and paleobiota. Rapid stage changes and the possibility of pipe-full ow make cave deposits dierent than surface deposits. is and other factors present dirculties with interpreting the cave record, but extended preservation is aorded by the roong of deposits. Dating by magnetism or isotopes has been successful in many locations. Caves can be expected to persist for 10 Ma in a single erosive cycle; most cave sediments should be no older than this. Key words: clastic sediments, paleoclimate, sedimentology, stratigraphy, dating. I NTRODUCTION Geology is undeniably a science of history, and since the earliest practice of the discipline, that history has been revealed in clastic sedimentary deposits. W illiam Smith, for example, created maps of the sedimentary rocks in England in the late 1700s, and established a relative chro nology of their deposition using stratigraphic position and fossils. It has been natural, therefore, that karst scien tists examine clastic deposits in caves, in order to explore geologic time. In doing so, they are in large part applying the same principles and techniques developed by classi cal stratigraphers. An early example of this was a study by Kukla and Loek (1958) examining the processes of cave sediment deposition and preservation. In the present day, work such as that by Granger et al. (2001) and Polyak et. al. (1998) builds upon those classical techniques and ap plies laboratory methods to develop absolute chronolo

PAGE 2

TIME in KARST 2007 144 gies. ese chronologies in turn have allowed insight to such processes as river incision, water-table lowering, and landscape/climate linkages. is paper is a brief evaluation of clastic sediments as they apply to deciphering historical processes and events in karst terrane. Advantages and problems of working with these unique deposits are presented. For purposes of this paper, the age of a given cave sediment refers to the time of deposition of the material in the cave. M ATERIALS AND PROCESSES e processes that result in clastic sedimentation in caves are quite varied. Reviews and details including classi cation of deposits are presented in several texts (W hite, 1988; Ford and W illiams, 1989; Sasowsky and Mylroie, 2004). A perspective is given here. A useful broad-level classication is genetic, and based upon whether the clastic material originated within the cave (autogenic) or was carried in from the surface (allogenic). e former class is mainly bedrock breakdown (incasion), but encompasses ne grained sediments sourced from insoluble residue during phre atic enlargement, collapse of secondary mineralization (speleothems), and so forth. Allogenic sediments include alluvium, windblown material, animal feces, fossil mat ter, till, etc. In practice, the most commonly occurring materi als by far are bedrock breakdown and alluvium. Conse quently, autogenic cave sediments are mainly limestone. Allogenic sediments are usually resistant siliciclastics, because carbonates do not typically persist in the uvial environment. ere is no satisfying overall term for the clastic de posits found in caves. e word soil has been applied to the ne grained deposits, but this is a misnomer by most denitions, and is not recommended. Cave ll and cave earth have also been used. Regolith seems applica ble in spirit, but, because this material does not strictly .form(s) the surface of the land .... (Jackson, 1997) some may object to such usage. B REAKDO W N e collapse of cave bedrock walls and ceilings results in material that is angular, and ranges in size from sand to boulders. It is possible many times to visually t larger blocks to their point of origin on the adjacent cave walls and ceilings. e process of breakdown is not a common occurrence on human timescales. Only a few cases of present-day natural failure have been documented. For example, in Mammoth Cave, Kentucky only one large col lapse was noted in 189 years of mining and tourism (May et al. 2005). However, on geologic timescales, the proc ess is pervasive and evident in most caves. Failures occur along existing planes of weakness (joints, faults, bedding planes). Causes of collapse can include removal of under lying support (particularly loss of buoyancy caused by the transition from phreatic to vadose conditions), removal of overlying arch support, cryoclastism (wedging by ice), and secondary mineral wedging (W hite and W hite, 2003). Triggering by earthquakes has also been observed, for example in Sistem Zeleke Jame-Karlovica (personal communication, F. Drole). Davies (1951) published an early analysis of expected collapse parameters in the cave environment. is was expanded on by W hite (1988, p. 232) to evaluate stability of ceilings relative to limestone bed thickness. Greater spans can be maintained by thick er beds. Jameson (1991) provides a comprehensive over view and classication of breakdown. Breakdown is frequently most prolic at 1) the intersections of cave passages, presumably due to the greater span lengths present at such points, and 2) where the cave is close to the surface, due to lack of thinning of the span and resulting decreased competency. In evalua tions of causes for passage terminations (W hite, 1960) it was noted that many cave passages ended in breakdown blockage (referred to by explorers as terminal break down). Although pervasive, breakdown has not found sig nicant utility for deciphering earth history in karst ter ranes. A LLUVIUM Alluvium enters caves by sinking stream, and occasion ally by colluvial mechanisms. e transport processes are for the most part similar to those in surface chan nels. e full range of sediment sizes are seen, structures such as cross-bedding and pebble imbrications develop, and cut-and-ll stratigraphy is possible. However, there are two important dierences exhibited for stream ow in caves when compared to most surface channels. First, channel width is severely constrained by bedrock walls. is promotes rapid stage increase during ooding, akin to that of slot canyons in surface streams (Fig. 1). Second, I RA D. S ASO W SK Y

PAGE 3

TIME in KARST 2007 145 because the channel is roofed over, it is possible to have conned (pipe-full) rather than open channel ow. Tak en in conjunction, the results of these two conditions are the likelihood of high ow velocities, and the possibil ity of upwards phreatic ow. A striking example of rapid Fig. 1: Subterranean stream channels are typically narrow, and have no oodplain (a). is leads to rapid stage changes. Similar conditions in the surface environment are only seen in slot canyons such as the V irgin River, Utah, USA (b). stage change is seen in Hlloch, Switzerland, where rises of 250 m in a single ood have been recorded (W ildberger and Preiswerk, 1997; Jeannin, 2001). Cases of phreatic lis are seen in many cave systems. In Castleguard Cave (Rocky Moun tains, Canada) a seasonally active li of 9 m is observed (Schroeder and Ford, 1983). In that situation wellrounded cobbles are accumulated at the base, where they reside until communition reduces them sur ciently to allow transport up the li tube. e composition of the allu vium reects the source of the ma terial, as well as some other factors. It is interesting to note that a high proportion of clay sized material found in cave alluvium is actually ne-grained silica, not a clay min eral (W hite, 1988). e residuum found on the surface of many karst terranes frequently contains high amounts of clay and chert. e clay results from insoluble residues of the weathered limestone. e chert behaves in a very persistent way, being found throughout cave passages. I NFORMATION REVEALED In the investigation of clastic sedimentary deposits, either cave related or not, answers are sought to such questions as: How old? W hat was the paleoenvironment? W hat was the ow direction? W hat organisms were present? ese in turn allow an understanding of geologic history, environments of deposition, past climates, and potential for sedimentary deposits to act as mineral and fuel res ervoirs. In the case of cave studies, it is primarily the rst question which has been addressed. Caves can only be numerically dated by the deposits that they hold, and this age is usually reported as a minimum value. Alluvial materials are considered superior to speleothems in this undertaking, because they are emplaced much earlier in the existence of the cave. Once a date has been obtained, subsequent inferences such as rates of river incision, de nudation, and so forth, can then be made based upon the relation of the cave to the landscape. Dating has been accomplished by radiocarbon, magnetism, and cosmo genic isotopes. Paleoenvironmental information is revealed through studies of sedimentary structures and sequences, as well as via analyses of clay mineralogy and environmental magnetism. Paleohydrology can be deduced using tra ditional stratigraphic indicators such as cross-bedding, pebble imbrication, etc. Fossil deposits of organisms are actually rather rare within caves most cave depsits are barren of these materials. Signicant deposits are known, though, and many excavations made in caves (particu larly in the entrance facies) serve as irreplaceable records of terrestrial fauna. C LASTIC SEDIMENTS IN CAVES IMPERFECT RECORDERS OF PROCESSES IN KARST

PAGE 4

TIME in KARST 2007 146 L IMITS ON TIMESCALE Caves are erosional landforms, which have a limited pe riod of existence. Excluding those caves which have been subjected to burial, this places a practical limit on their duration as potential recorders of nearby processes. In any case, the cave sediments can be no older than the cave they are emplaced in (Sasowsky, 1998). erefore, the ultimate limit on preservation of sediments within a cave is the persistence (lifetime) of the cave in the envi ronment. In most limestone terranes epigenetic processes occur, with dissolution taking place both at the surface (forming pavements, dolines, etc.) and in the subsur face (forming caves). As base level lowers, denudation of the upland surfaces is also occurring and uppermost caves are eventually breached and destroyed. In certain settings examples of various states of decay can be seen in the landscape, and the sedimentary lls of breached (unroofed) caves may even be observed (e.g. uteri, 2004). In settings such as the Appalachian Valley and Ridge, hundreds of meters of carbonate have been de nuded from anticlinal valleys (W hite, 1988), and one may imagine extensive systems of caves which have been obliterated with no remaining trace. Bounds on the expected lifetime of an epigene cave may be evaluated by considering the two main control ling factors: initial depth of formation and rate of land surface lowering (denudation, Fig. 2). Although caves may form at any depth, a practical limit of 300 m is rea sonable, and the majority of caves are much shallower (Milanovic, 1981). Note that this depth is not correla tive to the frequently reported mapped depth of caves, which refers to the maximum vertical extent of survey. In the context of the present evaluation, depth is the posi tion below surface (thickness of overlying rock) at a given point in the cave. Denudation rates can be quite variable, and tend to correlate with rainfall (W hite, 1988, p. 218). Envelopes of expected cave persistence can be construct ed (Fig. 2) using these 2 parameters. Based upon this cal culation, epigene caves would usually exist in the erosive environment for up to 10 Ma. In practice, dating has not yet resulted in identica tion of caves this old within the present erosional cycle. Paleomagnetic dating has been used back to 4.4 Ma (Cave of the W inds, Colorado, USA; Luiszer, 1994). Cosmogen ic isotope dating has documented cave sediments as old as 5.7 (.1) Ma (Bone Cave, Tennessee, USA; Anthony and Granger, 2004). e absence of older values may be a consequence of limitations of dating methods, or reect the relative dearth of older caves in the environment, or both. e challenges of paleomagnetic dating include ab sence of ne-grained sediments, lack of uninterrupted sedimentation, and uncertainties of correlation with the global magnetic polarity scale. Cos mogenic dating is constrained by the absence of quartzose sediments, un certainties in parent isotope values, and the cost/eort of analyses. If consideration is extended beyond the present erosional cycle, lled and buried caves (paleokarst) are found in the rock record. Such materials have been recognized in many places, and the lls described in some detail (e.g. Loucks, 1999). Interest has been strong in the con text of exploration for minerals or petroleum. ese deposits also rep resent a potential trove of informa tion on far past hydrologic and en vironmental conditions because of their capacity to preserve. Fig. 2: eoretical persistence of caves in an erosional environment. e length of time that a given cave will exist depends upon the initial depth of formation (position on y-axis) and the denudation rate (slope of line). Gray regions envelope a range of reasonable denudation pathways for two examples. In case A, a cave formed at 200 m depth, the expected lifetime is 2.5 to 10 M a. For a cave formed at 100 m depth (case B), the lifetime is reduced to 1.25 to 5 M a. Solid sloping lines are the average denudation rate, 69 m/M a, for 33 major drainage basins (calculated from data in Summereld and H ulton, 1994). I RA D. S ASO W SK Y

PAGE 5

TIME in KARST 2007 147 R ESOLUTION CONTINUIT Y, AND VERACIT Y Stratigraphers have traditionally examined marine or paralic sediments because of their resolution, continuity, and veracity. Compared to terrestrial deposits, marine/ paralic strata are much more laterally and vertically ex tensive, they are of economic interest, and they potential ly function as continuous recorders for long periods of time. Terrestrial deposits are of interest though, particu larly because they contain information about the on-con tinent setting. W ithin the terrestrial environment lacus trine deposits and uvial terraces have seen the greatest attention as recorders of Cenozoic paleo-conditions. Lakes probably represent the highest quality records in the terrestrial environment their environment many times is one of high preservation potential. Lacustrine deposits can be sampled by coring; duplication of cores can serve as a quality control; accumulation rates can be rapid; sediment properties are well tied to local environ mental conditions; and spatial variability is usually well understood. Terraces tend to preserve a partial record of the uvial environment, depending upon regional upli or down-cutting of the stream. In comparison, most caves contain spatially irregu lar deposits that can be aected by factors such as plug ging of swallets, extreme ow events, and back-ooding. Hydrologic complexity is common (Bosk et al. 2003), even more so than surface uvial environments. Analysis of the paleohydrology of the depositional setting through cave passage morphometry is usually necessary, and may be quite time consuming if detailed maps are not avail able. Stratigraphic sections may be discontinuous, and require compilation. Caves are dircult sampling loca tions, due to logistics, remoteness, lack of light, and con straints on sampling equipment transport. Nevertheless, the cave environment is one that pro vides some advantages in recording the history of a re gion. e greatest advantage is that of potential preserva tion. Because caves are roofed over deposits are likely to be protected (at least on intermediate time scales), from Fig. 3: Comparison of sedimentary records from Lake B aikal, Russia (3 columns on le), and Cave of the Winds, USA (3 columns on right). B aikal data used with permission from King and Peck, 2001. Cave of the Winds data used with permission from Luiszer, 1994. C LASTIC SEDIMENTS IN CAVES IMPERFECT RECORDERS OF PROCESSES IN KARST

PAGE 6

TIME in KARST 2007 148 surcial erosion. is is particularly germane for the u vial deposits. W eathering and erosion of surface uvial terraces is commonplace. In the cave, such materials may sit undisturbed for years. For example, in Xanadu Cave, Tennessee, USA, a pristine, non-indurated uvial deposit that is greater than 780 ka was sampled (Sasowsky, et al. Fig. 4: Episodic inlling and removal of sediments is commonly observed in caves. In this section of Windy M outh Cave (West V irginia, USA) a diamict was almost completely removed aer being covered with owstone. e conduit is presently dry. 1995). Although rare, in exceptional settings the quality of the cave re cord may approach that of lakes (Fig. 3). Conditions amenable to this are stable recharge conguration, dif fuse recharge, minimal variation of discharge, and deep circulation. In Figure 3 two exceptional records are compared. e Lake Baikal record was constructed from cores taken on watercra. In that setting, about 40 m of sediment accumulate in 1 Ma. In contrast, at Cave of the W inds the accumulation rate is slower by more than an order or magnitude. In many settings caves appear to undergo episodic lling and exca vation (Fig. 4). In certain cases this may be locally controlled by cata strophic storms (e.g. Doehring and Vierbuchen 1971). However, the presence of broadly similar depos its/incisions within many caves in a region supports the idea that cave clastic materials reect regional paleoclimatic conditions. ese deposits hold much information that will be revealed with continued advances in conceptual frameworks and improved labo ratory methods. R EFERENCES Anthony, D.M. & D.E. Granger, 2004: A Late Tertiary origin for multilevel caves along the western escarp ment of the Cumberland Plateau, Tennessee and Kentucky, established by cosmogenic 26 Al and 10 Be. Journal of Cave and Karst Studies, v. 66, no. 2, p. 46-55. Bosk P., P. Pruner, & J. Kadlec, 2003: Magnetostratigra phy of cave sediments: Application and limits. Stu dia Geophysica et Geodaetica, v. 47, p. 301-330. Davies, W E., 1951: Mechanics of cavern breakdown. National Speleological Society Bulletin, v. 13, p. 3643. Doehring, D.O. & R.C. Vierbuchen, 1971: Cave Develop ment during a catastrophic storm in the Great Val ley of Virginia. Science, v. 174, no. 4016, p. 13271329. Ford, D.C. & P.W W illiams, 1989: Karst geomorphology and hydrology. Unwin Hyman, London, 601 p. Granger, D.E., D. Fabel, D. & A.N. Palmer, 2001: Plio cenePleistocene incision of the Green River, Ken tucky, determined from radioactive decay of cosmo genic 26 Al and 10 Be in Mammoth Cave sediments. Geological Society of America Bulletin, v. 113; no. 7, p. 825-836. Jackson, J.A. (ed.), 1997: Glossary of geology. 4 th ed., American Geological Institute, Falls Church, Vir ginia, 769 p. Jameson, R.A., 1991: Concept and classication of cave breakdown: An analysis of patterns of collapse in Friars Hole Cave System, W est Virginia: In, Kast ning, E.H. and Kastning, K.M. (eds.), Appalachian Karst. National Speleological Society, Huntsville, Alabama, USA, p. 35-44. Jeannin, P.-Y ., 2001: Modeling ow in phreatic and epi phreatic karst conduits in the Hlloch cave (Muo tatal, Switzerland). W ater Resources Research, v. 37, no. 2 p. 191-200. I RA D. S ASO W SK Y

PAGE 7

TIME in KARST 2007 149 King, J. & J. Peck, 2001: Use of paleomagnetism in studies of lake sediments: In: Last, W .M. & J.P. Smol, (eds.), Tracking environmental change using lake sedi ments Volume 1: Basin analysis, coring and chron ological techniques. Kluwer Academic Publishers, Dordrecht, p. 371-389. Kukla, J. & V. Loek, 1958: K promlematice vyzkumu jeskynnich vyplni (To the problems of investigation of the cave deposits). feskoslovensy Kras, v. 11, p. 19-83. Loucks, R.G., 1999: Paleocave carbonate reservoirs: Ori gins, burial-depth modications, spatial complexity, and reservoir implications. AAPG Bulletin, v. 83; no. 11; p. 1795-1834. Luiszer, F. G., 1994: Speleogenesis of Cave of the W inds, Manitou Springs, Colorado: In Sasowsky, I. D., and Palmer, M. V. (eds.) Breakthroughs in karst geomi crobiology and redox geochemistry (Special Publi cation 1). Charles Town, W est Virginia, Karst W a ters Institute, p. 91-109. May, M.T., K.W Kuehn, C.G. Groves, C.G., & J. Meiman, 2005: Karst geomorphology and environmental concerns of the Mammoth Cave region, Kentucky. American Institute of Professional Geologists 2005 Annual Meeting Guidebook, Lexington, Kentucky, 44 p. Milanovic, P. T., 1981: Karst Hydrogeology (translated from the Y ugoslavian by J. J. Buhac). W ater Re sources Publications, Littleton, Colorado, 434 p. Polyak, V.J., W .C. McIntosh, N. Gven, N., & P. Proven cio, 1998: Age and origin of Carlsbad Cavern and related caves from 40 Ar/ 39 Ar of alunite. Science, v. 279, no. 5358, p. 1919 1922 Sasowsky, I.D., 1998: Determining the age of what is not there. Science, v. 279, no. 5358, p. 1874 Sasowsky, I. D. & J.W Mylroie (eds.), 2004: Studies of cave sediments: Physical and chemical recorders of climate change. Kluwer Academic/Plenum Pub lishers, New Y ork, 329 p. Sasowsky, I. D., W .B. W hite, & V.A. Schmidt, 1995: Deter mination of stream incision rate in the Appalachian Plateaus by using cave-sediment magnetostratigra phy. Geology, v. 23, no. 5, p. 415-418. Schroeder, J. & D.C. Ford, 1983: Clastic sediments in Cas tleguard Cave, Columbia iceelds, Canada. Arctic and Alpine Research, v. 15, no. 4, p. 451-461. Summereld, M. A., & N.J. Hulton, 1994: Natural con trols of uvial denudation rates in major world drainage basins. Journal of Geophysical Research, v. 99(B7), p. 13,871,884. uteri, F., 2004: Cave sediments and denuded caverns in the Laki Ravnik, classical Karst of Slovenia: In: Sasowsky, I.D. and Mylroie, J.W (eds.), Studies of cave sediments: Physical and chemical recorders of climate change. Kluwer Academic/Plenum Pub lishers, New Y ork, p. 123-134. W hite, W B., 1960: Termination of passages in Appala chian Caves as evidence for a shallow phreatic ori gin. Bulletin of the National Speleological Society, v. 22, no. 1, p. 43-53. W hite, W .B., 1988: Geomorphology and hydrology of karst terranes. Oxford University Press, 464 p. W hite, W .B. & E.L. W hite, 2003: Gypsum wedging and cavern breakdown: Studies in the Mammoth Cave System Kentucky. Journal of Cave and Karst Stud ies, v. 65, no. 1, p. 43-52. W ildberger, A. & C. Preiswerk, 1997: Karst and caves of Switzerland. SpeleoProjects, Basil, Switzerland, 208 p. C LASTIC SEDIMENTS IN CAVES IMPERFECT RECORDERS OF PROCESSES IN KARST



PAGE 1

A NAL Y SIS OF LONG TERM 18782004 MEAN ANNUAL DISCHARGES OF THE KARST SPRING F ONTAINE DE V AUCLUSE F RANCE A NALIZA DOLGOfASOVNEGA 18782004 POVPREfNEGA LETNEGA PRETOKA KRAKEGA IZVIRA F ONTAINE DE V AUCLUSE F RANCIJA Ognjen BONACCI 1 Izvleek UDK 556.36(44) Ognjen Bonacci: Analiza dolgoasovnega (1878-2004) po v prenega letnega pretoka krakega izvira Fontaine de Vaucluse (Francija) V prispevku predstavim statistino analizo asovne vrste povprenega letnega pretoka in letnih padavin v zaledju slavne ga izvira Fontaine de Vaucluse v Franciji. Fontaine de Vaucluse je tipini kraki izvir pri katerem voda priteka iz velike globine. Nahaja se v jugovzhodni Franciji. Povpreni pretok izvira je 23,4 m 3 /s. Povprena koliina letnih padavin v zaledju, ki meri 1130 km 2 je 1096 mm. Z uporabo metode umerjenih delnih vsot (RAPS) smo doloili pet statistino pomembnih razlinih podobdobij: 1) 1878-1910: 2) 1911-1941; 3) 1942-1959: 4) 19601964; 5) 1965-2004. Vzrokov za razline pretoke preko celotne ga obdobja (1878-2004) je lahko ve; npr. klimatske spremem be in antropogeni vplivi. V tem trenutku moramo poudariti, da objektivne znanstvene razlage za razline hidroloke znailnosti v petih podobdobjih e ne poznamo. Kljune besede: hidrologija krasa, povpreni letni pretok, koliina letnih padavin, krakih izvir, Fontaine de Vaucluse, Francija. 1 Faculty of Civil Engineering and Architecture, University of Split, 21000 Split, Matice hrvatske 15, Croatia, E-mail: obonacci@ gradst.hr Received/Prejeto: 27.11.2006 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 151 Abstract UDC 556.36(44) Ognjen Bonacci: Analysis of long-term (1878-2004) mean annual discharges of the karst spring Fontaine de Vaucluse (France) Statistical analyses have been carried out on a long-term (18782004) series of mean annual discharges of the famous karst spring Fontaine de Vaucluse (France) and the mean annual rainfall in its catchment. e Fontaine de Vaucluse is a typical ascending karst spring situated in the south-eastern region of France. e spring has an average discharge of 23.3 m 3 /s. e average annual rainfall is 1096 mm. Its catchment area covers 1130 km 2 Using the rescaled adjusted partial sums (RAPS) method the existence of next ve statistically signicant dier ent sub-series was established: 1) 1878-1910: 2) 1911-1941; 3) 1942-1959: 4) 1960-1964; 5) 1965-2004. e dierent spring discharge characteristics during this long period (1878-2004) can be caused by natural climatic variations, by anthropogenic inuences, and possibly by climate changes. At this moment it should be stressed that objective and scientically based rea sons for dierent hydrological behaviour in ve time sub-peri ods could not be found. Keywords: karst hydrology, mean annual discharges, annual catchment rainfall, karst spring, Fontaine de Vaucluse, France. I NTRODUCTION e Fontaine de Vaucluse represents one of the most famous and most important karst springs on the Earth. It is located in the south-eastern karst region of France (Figure 1), about 30 km eastward of the town of Avignon. It represents the only ow exit from the 1500 m thick karst aquifer of Lower Cretaceous limestone (Blavoux et al., 1992b). e karst system of the Fontaine de Vauc luse is characterised by an approximately 800 m unsatu rated zone. Emblanch et al., (1998) and Emblanch et al., (2003) stressed important role of this zone for the trans formation of rainfall into runo. e Fontaine de Vauc luse karst spring catchment area is estimated to be 1130

PAGE 2

TIME in KARST 2007 152 km 2 (Cognard-Plancq et al., 2006a; 2006b). e average catchment altitude is 870 m a. s. l. e average annual air temperature of the catchment is 9,6 C. e Fontaine de Vaucluse is typical ascending karst spring (Michelot & Mudry 1985; Blavoux et al., 1991/1992; 1992a). Its limestone channel ranges in diam eter from 8 to 30 m (Mudry & Puig, 1991). e lowest Fig. 2: T ime data series of annual rainfall P at the Fontaine de V aucluse catchment with trend line for the period 1878-2004. Fig. 1: Location map of karst spring Fontaine de V aucluse. depth reached by diver was -308 m below the gauging station datum of 84.45 m a. s. l. is depth is still not at the bottom of the ascending karst channel. e maximum water level measured at the gaug ing station was 24.10 m above the datum, the minimum was a few centimetres below the datum. e rate of the maximum discharge of the spring has never been pre cisely measured, but it is estimated that it cannot exceed 100 m 3 /s (Blavoux et al., 1991/1992; 1992a). CognardPlancq et al. (2006b) state that maximum spring dis charge varies between 100 and 120 m 3 /s. is surmise identies a karst spring with limited discharge capacity (Bonacci 2001). e historical minimum discharge is 3.7 m 3 /s (Blavoux et al., 1991/1992). Every karst aquifer has complex hydrodynamic be haviour. e Fontaine de Vaucluse karst system respons es to rainfall quite rapid in comparison with the large recharge area. e peak of hydrograph occurred 24 to 72 hours aer the rainfall events. e spring water level and discharge recessions are slow, which can be explained by the existence of a large storage capacity of the aquifer (Cognard-Plancq 2006b). e primary objective of the investigation was to dene sub-periods with dierent hydrological behav iour of the Fontaine de Vaucluse karst spring during 127 years period (1878-2004), analysing time series of mean annual spring discharges. It should be the rst step in explanation of this extremely important and interesting phenomenon. ANAL Y SIS OF CATCHMENT ANNUAL RAINFALL TIME SERIES e climate in the catchment is Mediterranean. Rainfall distribution over the year as well as over the large spring catchment is irregular. Intensive and signicant rainfall events occurred during autumn and spring, while sum mer and winter are generally dry. Interannual uctua tions of rainfall on the catchment are very high. In order to dene an historical homogeneous catch ment rainfall database Cognard-Plancq et al., (2006b) used six rainfall gauging stations. e mean elevation of these stations is 445 m a. s. l., while the mean elevation of the spring catchment is 870 m a. s. l. Transformation of the measured monthly rainfall to the altitude of 870 m a. s. l. was made. e average annual catchment rainfall in the 1878-2004 period is 1096 mm, while the minimum and maximum observed values were 641 mm (1953) and 1740 mm (1977) respectively. Data series with linear trend line of the annual rain fall on the Fontaine de Vaucluse catchment for the pe riod 1878-2004 are presented in Figure 2. e increasing trend of the catchment rainfall of 1.045 mm per year is not statistically signicant but should not be neglected in further analyses. O GNJEN BONACCI

PAGE 3

TIME in KARST 2007 153 ANALISIS OF MEAN ANNUAL DISCHARGES TIME SERIES Data series with linear trend line of the mean annual spring discharges Q for the period 1878-2004 are pre sented in Figure 3. e decreasing trend of the mean an nual discharges of 0.0468 m 3 /s per year is not statistically signicant. e average annual catchment discharge in the 1878-2004 period was 23.3 m 3 /s, while the minimum and maximum observed values were 7.61 m 3 /s (1990) and 53.4 m 3 /s (1915) respectively. It should be stressed that annual catchment rain fall during the same period has an increasing trend. In Figure 4 linear regression between the mean annual the Fontaine de Vaucluse discharges Q and the Fontaine de Vaucluse catchment annual rainfall P is shown. e lin ear correlation coercient is only 0.713, which is rela tively low. A special problem is that the regression line cut abscissa line at 222 mm of annual rainfall P, which is relatively low value. Explanation of so unusual rainfallruno relationship can be found in fact that accuracy of discharges and rainfalls are not very high, and maybe the value of catchment area of 1130 km 2 is not precisely dened. It should be stressed that determination of exact catchment area in karst is one of the greatest and very oen unsolved problems. is may be the case with the catchment of the Fontaine de Vaucluse spring. e weak relationship between runo and rainfall means that some others factors (probably: air temperature, groundwater level, interannual rainfall distribution, changes of catch ment area during the time, preceding soil wetness, an thropological inuences, climate change etc) have inu ence on it. A time series analysis can detect and quantify trends and uctuations in records. In this paper the Rescaled Adjusted Partial Sums (RAPS) method (Garbrecht & Fer nandez 1994) was used for this purpose. A visualisation approach based on the RAPS overcomes small systematic changes in records and variability of the data values them selves. e RAPS visualisation highlights trends, shis, data clustering, irregular uctuations, and periodicities in the record (Garbrecht & Fernandez 1994). It should be stressed that the RAPS method is not without shortcom ings. e values of RAPS are dened by equation: where is sample mean; is standard deviation; n is number of values in the time series; (k=1, 2,n) is counter limit of the current summation. e plot of the RAPS versus time is the visualisation of the trends and uctuations of Y t Time data series of Rescaled Adjusted Partial Sums (RAPS) for mean annual spring discharges in the period 1878-2004 are given in Figure 5. erefore, the total data Fig. 3: T ime data series of mean annual discharges Q at the Fontaine de V aucluse karst spring with trend line for the period 1878-2004. Fig. 4: Linear regression between the mean annual the Fontaine de V aucluse discharges Q and annual the Fontaine de V aucluse catchment rainfall P. Fig. 5: T ime data series of the Rescaled Adjusted Partial Sums (RAPS) for mean annual discharges Q for the period 1878-2004 with designated next ve sub-periods: 1) 1878-1910; 2) 19111941; 3) 1942-1959; 4) 1960-1964; 5) 1965-2004. A NAL Y SIS OF LONG TERM 18782004 MEAN ANNUAL DISCHARGES OF THE KARST SPRING F ONTAINE DE V AUCLUSE

PAGE 4

TIME in KARST 2007 154 the 5 % and even more 1 %. At the same time the neigh bouring sub-series averages of the catchment rainfall are not statistically signicant. Figure 7 shows ve linear regressions between mean annual discharges Q and annual catchment rainfall P de ned for ve dierent sub-periods. It can be seen that linear correlation coercients for all sub-series, except for third (1942-1959) and fourth (1960-1964) ones are higher than the linear correlation coercient for whole time series. series was divided into next ve subsets: 1) 1878-1910; 2) 1911-1941; 3) 1942-1959; 4) 1960-1964; 5) 1965-2004. Cognard-Plancq et al., (2006a; 2006b) dened the same ve stationary sub-periods using dierent methodology. Five time data sub-series of the Fontaine de Vau cluse karst spring mean annual discharges Q with trend lines for ve dened sub-periods are shown in Figure 6. In order to investigate statistically signicant dierences between the averages of ve time sub-series for Q and P the t-test was used. e neighbouring averages of dis charges for all ve sub-series are statistically signicant at Fig. 7: Linear regressions between mean annual discharges Q and annual catchment rainfall P dened for ve dierent sub-periods. Fig. 6: Five time data sub-series of the Fontaine de V aucluse karst spring mean annual discharges Q with trend line for ve dened sub-periods. CONCLUSION e rescaled adjusted partial sums (RAPS) method es tablished existence of next ve statistical, and hydrologi cal signicant dierent time sub-series: 1) 1878-1910: 2) 1911-1941; 3) 1942-1959: 4) 1960-1964; 5) 1965-2004. Variations in the Fontaine de Vaucluse karst spring hy drological regime during relatively short period of 127 years are very strong and cannot be neglected. Anthro pogenic impacts are probably the main cause of such be haviour of the mean annual spring discharges time series analysed, but the natural pattern of drought and wet years is also possible. Land-use changes and overexploitation of surface water and groundwater at the spring catch ment on hydrological regime of the Fontaine de Vaucluse spring certainly exists. eir exact quantication during analysed period is extremely questionable due to miss ing of many parameters. Strict division of natural and anthropogenic inuences on the hydrological regime is hardly possible. The significant changes of spring discharge characteristics during 127 years long period (18782004) can be caused by natural climatic variations, by anthropogenic influences, and possibly by climate changes. It is extremely hard, but at the same time extremely practically and theoretically important, to find correct and scientifically based explanation of this phenomenon. Cognard-Plancq et al., (2006a) consider that rain fall-runo data have shown the large impact of clima tologic variations on the hydrogeological system. ey conclude that the underground storage zone is an impor tant inuence on karst spring outow, which depends on rainfall amount over 2 or 3 previous years. Investigations made in this paper do not conrm this statement. Correct answers on many questions dealing with changes in hydrological-hydrogeological regime of the Fontaine de Vaucluse karst spring cannot be done using only annual data. Some processes can be explained mea suring and analysing climatologic, hydrologic, hydrogeo logical and geochemical interactions in shorter as well as larger time increments. e problem is that most of pa rameters required for these analyses were not monitored in the past. O GNJEN BONACCI

PAGE 5

TIME in KARST 2007 155 More accurate and precise delineation and deni tion of the Fontaine de Vaucluse spring catchment should be done. It is possible that its catchment area changes as a function of groundwater level. is means that ground water level measurements in deep piezometers should be organized across the catchment. e second task which should be considered in further analyses is detailed anal ysis of inuence of rainfall distribution during the year on the spring runo. is can have very strong inuence on the relationship between rainfall and runo, especial ly in karst areas. It can be stated that main dilemmas about variations of mean annual discharges of the Fontaine de Vaucluse karst spring during 127 years long period have not been solved. ey should be explained using number of dif ferent procedures and climatic as well as other indica tors, and performing further detailed measurements and analyses. e paper presents the need for interdisciplin ary analyses incorporating several approaches and tech niques. For the sustainable development and the protec tion of such valuable water resource it is very important to establish prerequisites for the denition of a causes and consequences of its hydrological changes. ACKNO W LEDGEMENT e author thanks to Anne-Laure Cognard-Plancq and Christophe Emblanch from Laboratoire dHydrogologie, Facult des Sciences, Universit dAvignon et Pays de Vaucluse 84000 Avignon, 33 Rue Louis Pasteur, France, which kindly provide me with data analysed in this pa per. REFERENCE Blavoux, B., Mudry, J. & Puig, J.-M., 1991/1992: Bilan, fonctionnement et protection du systme karstique de la Fontaine de Vaucluse (sud-est de la France). Geodinamica Acta, 5 (3), 153-172., Paris. Blavoux, B., Mudry, J. & Puig, J.-M., 1992a: e karst system of the Fontaine de Vaucluse (Southeastern France). Environ. Geol. W ater. Sci., 19 (3), 215-225. Blavoux, B., Mudry, J., & Puig, J.-M., 1992b: Role du con texte geologique et climatique dans la genese et le fonctionnement du karst de Vaucluse (Sud-Est de la France). In: H Paine, W Back (eds.) Hydrogeology of Selected Karst Regions. IAH International Con tributions to Hydrogeology, Vol. 13, 115-131. Bonacci, O., 2001: Analysis of the maximum discharge of karst springs. Hydrogeol. J., 9, 328-338. Cognard-Plancq, A.-L., Gvaudan, C. & Emblanch, C., 2006a: Apports conjoints de suivis climatologique et hydrochimique sur le rle de ltre des aquifres karstiques dans ltude de la problmatique de changement climatique; Application au systme de la Fontaine de Vaucluse. Proceedings of the 8th Conference on Limestone Hydrogeology. Neucha tel, Sep. 21-23, 2006, 67-70. Cognard-Plancq, A.-L., Gvaudan, C., &, Emblanch, C., 2006b: Historical monthly rainfall-runo database on Fontaine de Vaucluse karst system: review and lessons. IIIme Symposium International Sur le Karst Groundwater in the Mediterranean Coun tries, Malaga, Spain. In: J J Duran, B Andreo, F Y Carrasco (eds.) Karst, Cambio Climatico y Aguas Subterraneas. Publicaciones des Instituto Geological y Minero de Espana. Serie: Hidrogeologia y Aguas Subterrraneas, N: 465-475. Emblanch, C., Puig, J. M., Zuppi, G. M., Mudry, J., & Bla voux, B., 1998: Comportement particulier lors des montes de crues dans les aquifres karstiques, mise en vidence dune double fracturation et/ou de cir culation profonde: Example de la Fontaine de Vau cluse. Ecologae Geol. Helv., 92: 251-257. Emblanch, C., Zuppi, G. M., Mudry, J., Blavoux, B. & Batitot, C., 2003: Carbon 13 of TDIC to quantify the role of the unsaturated zone: e example of the Vaucluse karst systems (Southeastern France). J. of Hydrol., 279 (1-4): 262-274. A NAL Y SIS OF LONG TERM 18782004 MEAN ANNUAL DISCHARGES OF THE KARST SPRING F ONTAINE DE V AUCLUSE

PAGE 6

TIME in KARST 2007 156 Garbrecht, J. & Fernandez, G. P., 1994. Visualization of trends and uctuations in climatic records. W ater Resources Bulletin, 30 (2): 297-306. Michelot, C. & Mudry, J., 1985: Remarques sur les exu toires de laquifre karstique de la Fontaine de Vau cluse. Karstologia, 6(2): 11-14. Mudry, J., Puig, J.-M., 1991: Le karst de la Fontaine de Vaucluse (Vaucluse, Alpes de Haute-Provence, Drme). Karstologia, 18 (2): 29-38. O GNJEN BONACCI



PAGE 1

TIMING OF PASSAGE DEVELOPMENT AND SEDIMENTATION AT CAVE OF THE W INDS, MANITOU SPRINGS, COLORADO, USA fASOVNO USKLAJEVANJE RAZVOJA JAMSKIH PROSTOROV IN SEDIMENTACIJA V JAMI CAVE OF THE W INDS, MANITOU SPRINGS, COLORADO, ZDA Fred G. L UISZER 1 Izvleek UDK 551.3:551.44:550.38 550.38:551.44 Fred G. Luiszer: asovno usklajevanje razvoja jamskih pro storov in sedimentacija v jami Cave of the Winds, Manitou Springs, Colorado, ZDA flanek se osredotoa na zaetek razvoja jamskih prostorov in asovno sosledje sedimentacije v jami Cave of the winds, Manitou Springs, Kolorado. V bliini jame se nahajajo alu vialne terase, ki so bile datirane z radiometrino metodo. Z geomorfoloko metodo so bile povezane z jamo Cave of the W inds. V teh aluvialnih terasah so bili najdeni fosilni ostanki kopenskih polev, na katerih so bile opravljene datacije z ami nokislinami, ki so pokazale starost ~ 2 Ma let. Starost aluvialnih teras in njihova geomorfoloka povezava z jamo Cave of the W inds, sta sluila kot izhodie za natannejo asovno umes titev 10 metrov debele sekvence jamskih sedimentov, ki so bili magnetostratigrafsko opredeljeni. Raziskava je pokazala, da se je raztapljanje v jami prielo pred ~4.5 Ma leti, medtem ko se je odlaganje klastinih sedimentov prenehalo pred ~1.5 Ma let. Kljune besede: Cave of the W inds, Manitou Springs, ZDA, magnetostratigraja, aminostratigraja, kopenski poli. 1 University of Colorado, Boulder, Department of Geological Sciences, Campus Box 399, Boulder, CO 80302, USA. Received/Prejeto: 13.12.2006 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 157 Abstract UDC 551.3:551.44:550.38 550.38:551.44 Fred G. Luiszer: Timing of Passage Development and Sedi mentation at Cave of the Winds, Manitou Springs, Colorado, USA. In this study the age of the onset of passage development and the timing of sedimentation in the cave passages at the Cave of the W inds, Manitou Springs, Colorado are determined. e amino acid rations of land snails located in nearby radiometri cally dated alluvial terraces and an alluvial terrace geomorphi cally associated with Cave of the W inds were used to construct an aminostratigraphic record. is indicated that the terrace was ~ 2 Ma. e age of the terrace and its geomorphic rela tion to the Cave of the W inds was use to calibrate the magne tostratigraphy of a 10 meter thick cave sediment sequence. e results indicated that cave dissolution started ~4.5 Ma and cave clastic sedimentation stopped ~1.5 Ma. Key words: Cave of the W inds, Manitou Springs, magneto stratigraphy, aminostratigraphy, land snails. INTRODUCTION Cave of the W inds, which is 1.5 km north of Manitou Springs (Figure 1), is a solutional cave developed in the Ordovician Manitou Formation and Mississippian W il liams Canyon Formation. Commercialized soon aer its discovery in the1880s it has been visited by millions of visitors in the last 125 years. As part of an extensive study (Luiszer, 1997) of the speleogenesis of the cave the timing of passage development and sedimentation needed to be determined. e task of dating the age of caves has al ways been an enigma because dating something that has been removed is not possible. Sediments deposited in the cave passages, however, can be dated, which then can be used to estimate the timing of the onset of cave dis solution and when the local streams abandoned the cave.

PAGE 2

TIME in KARST 2007 158 FIELD AND LABORATOR Y PROCEDURES Amino Acid Dating Snails were collected from outcrops of the Nussbaum Al luvium, and from younger radiometrically dated alluvia (Fig. 2) for the purpose of dating the Nussbaum Alluvi um by means of amino-acid racemizatio. Approximately 50 kg of sandy silt was collected at each site. To minimize sample contamination, washed plastic buckets and fresh plastic bags were used. e samples were loaded into containers with a clean metal shovel and with minimal hand contact. In the lab, the samples were disaggregated by putting them in buckets lled with tap water and letting them soak overnight. e samples were then washed with tap water through 0.5-mm mesh scree. Following air drying, the mollusks were hand picked from the remaining matrix by means of a small paint brush dipped in tap water. e mollusks were then identied. Only shells that were free of sediment and discoloration were selected for fur ther processing. ese shells were washed at least ve times in distilled water while being sonically agitated. e amino-acid ratios were determined on a high-per formance liquid chromatograph (HPLC) at the Institute of Arctic and Alpine Research (University of Colorado, Boulder). Paleomagnetism A coring device was used to sample the cave sediments at six cored holes in the Grand Concert Hall (Fig. 3). e core samples were obtained by means of a coring device in which a hand-powered hydraulic cylinder drives a stainless-steel, knife-edged barrel down into the sedi ments. Up to 40 cm of sediment could be cored each trip into the hole without sediment distortion. Samples were also collected from hand-dug pits at Mummys Alcove and Sniders Hall (Fig. 3). Additionally, samples were col lected from a vertical outcrop in Heavenly Hall (Fig. 3). e pits and outcrops were sampled for paleomagnetic study by carving at vertical surfaces and pushing plastic sampling cubes into the sediment at stratigraphic inter vals ranging from 3.0 to 10.0 cm. e samples were ori ented by means of a Brunton compass. e core barrel and all pieces of drill rod that at tached to the barrel were engraved with a vertical line so that the orientation of the core barrel could be measured with a Brunton compass within A hand-operated hydraulic device was used to extract the sediment core from the barrels. As the core was extruded, a xed thin wire sliced it in half, lengthwise. Plastic sampling cubes were then pushed into the so sediment along the center line of the at surface of the core half at regular intervals A specially constructed coring device was utilized to core several locations in the cave. e natural remnant mag netization (NRM) of samples taken from the cores were use to construct a magnetostratigraphic record. is record by itself could not be used to date the age of the sediments because sedimentation in the cave stopped sometime in the past and part of the record was missing. An alluvial terrace, which overlies the Cave of the W inds, is geomorphically related to the cave. e age of the alluvial terrace, which had not been previously dated, can be used to determine the age of the youngest stream deposited sediments in the cave. An abundant number and variety of land snails were found when this alluvium was closely searched. Biostratigraphy could not be used to determine the age of the terrace because all of the snail species found were extant, however, the amino acid ra tions of the snails collected from this terrace and nearby radiometrically dated terraces were used to construct an aminostratigraphy that was used to date the alluvium. Once the age of the terrace was determined the age of the youngest magnetic chron of the magnetostratigraphic re cord could be assigned thus enabling the dating of cave dissolution and sedimentation. Fig. 1: Location of study area. Colorado El Paso Count y Cave of the Wi nd s Colorado Spring s Manitou Sprin gs 25 24 25 Denve r F RED G. L UISZER

PAGE 3

TIME in KARST 2007 159 (generally ~5.0 cm). e samples at Sniders Hall, Mum mys Alcove, and Hole 1 were taken with 3.2 cm 3 sampling cubes; all other samples were taken with 13.5 cm 3 cubes. In the lab, the NRM (Natural Remanent Magnetiza tion) of all samples was initially measured. Subsequently, the samples were subjected to alternating-eld (A. F.) demagnetization and remeasured. All samples were rst demagnetized at 10, and then at 15 millitesla (mT). Some samples at the bottom of Hole 5 that displayed aberrant inclinations and declinations were additionally demag netized at elds up to 30 mT. All remanence measure ments were made on a Schonstedt SSM 1A spinner mag netometer with a sensitivity of 1X10 -4 A/m. Repeat mea surements indicate an angular reproducibility of ~2 at an intensity of 1X10-6 A/m 2 Age Of Cave Passages Because Cave of the W inds is an erosional feature, its exact age cannot be determined. However, geologic and geomorphic features related to the cave can be used to bracket the age of incipient and major cave development. Solution breccia in the Manitou Formation indicate that there may have been some Middle Ordovician to Devo nian cave development (Forster, 1977). Sediment-lled paleo-caves and paleo-sinkholes at Cave of the W inds in Qlo Qes Qp Qb Kpl Qv Qs Tn Qrf KPr Pf Ypp MCr Xg nb PINEY CREEK ALLUVIUM (UPPER HOLOCENE ) EOLIAN SAND (HOLOCENE) BROADW AY ALLUVIUM (LAT E PLEISTOCENE ) LOUVIERS ALLUVIUM (LAT E PLEISTOCENE ) SLOCUM ALLUVIUM (LAT E PLEISTOCENE ) VERDOS ALLUVIUM (LAT E PLEISTOCENE ) ROCKY FLAT S ALLUVIUM (EARLY PLEISTOCENE ) NUSSBAUM ALLUVIUM (LAT E PLIOCENE) LARAMIE, FOX HILLS, and PIERRE FORMATIONS (CRETACEOUS) CRE TACEOUS, JURASSIC, TRIASSIC, and PERMIAN ROCK S FOUNTAIN FORMATIO N (PENNSYLVANIAN ) MISSISSIPPIAN, ORDOVICIAN, AND CAMBRIAN ROCKS PIKES PEAK GRANITE (PRECAMBIAN) BIOTITE GNEISS (PRECAMBIAN) Collection sites for amino acid racemization stud y 4 3 3 2 2 1 1 0 0 4 MILES KILOMETERS Qlo Qlo Qlo Qlo Qlo Qlo Qes Kpl Kpl Qp Qrf Qrf Qv Ypp Xg nb Xg nb Ypp Qv Qv Qv Qs Qv Qs Kpl Qv Qv Qp Qv Kpl Qs Qv Kpl Qp Kpl Qv Qv Qb Qes Qb Qp Qb Qb Qp Qes Kpl Kpl Tn MCr Tn Yp p Qrf Qp Ypp Pf Fil lmore Str e et I25 Garden of the Gods Road I25 Colora do Ave. Neva da Ave. U D Manitou Springs Ypp KPr KPr KPr ST AR LIGHT A CRES FILLMOR E CENTENNI AL CHESNUT COLOR AD O CITY CA VE OF THE WIND S N Figure 2. GEOLOGY MA P OF COLOR AD O SPRINGS A ND MA NITOU SPRINGS AR EA wi th locations of snail collection sites. Geology adapted fr om Trimble and Machette, (1979). U D U D U D U D U D BLA CK CA NYON MCr MCr MCr Pf Pf Pf Pf Pf Pf N3 8 N3 8 MODERN FLOOD PL AI N dicate Devonian to Late Mississippian karst development (Hose & Esch, 1992). Subsequent Cenozoic dissolution along some of these paleokarst features has resulted in the formation of cave passage (Fish, 1988). Between the Pennsylvanian and Late Cretaceous, about 3000 m of sediments, which contain abundant shale beds, were de posited over the initial cave. Very little, if any, cave de velopment could take place during this period of deep burial under the thick blanket of the nearly impervious rock. e Laramide Orogeny, beginning in the Late Cretaceous (~75 Ma, Mutschler et al., 1987), was asso ciated with the upli of the Rocky Mountains. e up li, which included the Rampart Range and Pikes Peak, caused the activation of the Ute Pass and Rampart Range Faults (Morgan, 1950; Bianchi, 1967). In the Manitou Springs area, movement on the Ute Pass Fault resulted in the folding, jointing and minor faulting of the rocks (Hamil, 1965; Blanton, 1973). e subsequent ow of corrosive water along the fractures related to the fold ing and faulting would produce most of the passages in Cave of the W inds and nearby caves. Upli during the early Laramide Orogeny increased the topographic relief in the Manitou Springs area, resulting in the initiation of erosion of the overlying sediments and also increased TIMING OF PASSAGE DEVELOPMENT AND SEDIMENTATION AT CAVE OF THE W INDS, MANITOU SPRINGS, COLORADO, USA

PAGE 4

TIME in KARST 2007 160 the local hydraulic head. e erosion of some of the im pervious shale along with the increased hydraulic head may have initiated some minor water ow through the joints and faults, causing incipient dissolution. However, in the rst 25 m. y. of the Laramide Orogeny, erosional stripping almost equaled upli (Tweto, 1975) resulting in a subdued topography with a maximum elevation of about 1000 m (Epis and Chapin, 1975). It was unlikely, therefore, that a large hydraulic head existed--a necessary hydraulic head that would have had to be present to force through the rock the large volumes of water needed for development of a large cave system. A Late Miocene-Early Pliocene alluvial deposit on the Rampart Range, 18 km northwest of Manitou Springs, indicates renewed MiocenePliocene upli, which in some places was up to 3000 m (Epis and Chapin, 1975). At the same time, movement along the Ute Pass Fault caused re direction of Fountain Creek from its former position near the abovementioned alluvial deposit to its present position (Scott, 1975). Val ley entrenchment along the Ute Pass Fault by Fountain Creek, in conjunction with upli, created the hydraulic head needed to drive the mineral springs, the mixing zone, and limestone dissolution (Luiszer, 1997). It is likely, therefore, that the age for the onset of major dissolu tion at Cave of the W inds is prob ably Late Miocene-Early Pliocene (7 Ma to 4 Ma). Age Of Cave Fill Sedimentation in the cave appears to have been contemporaneous with passage development. ere are a few problems in proving this chro nology. One is the lack of datable materials in the sediments, such as fossils or volcanic ashes. Preliminary study of the sediments indicated that magnetic reversal stratigraphy (mag netostratigraphy) might be useful in dating the sediments. e use of this method, however, presents another problem: it requires that the polarity sequence be constrained by at least one independent date. e Nussbaum Alluvium, which crops out east of the cave and is ~20 m higher in elevation, is apparently related to coarse sediments present at the top of sediment sequences in Cave of the W inds. If an age can be assigned to the Nussbaum Alluvium and the relation ship of the Nussbaum Alluvium to the coarse sediments in the cave deciphered, then an independent date can be as signed to at least one polarity reversal in the cave. e age of the Nussbaum Alluvium will be dealt with rst, because the age of the Nussbaum Alluvium is poorly constrained. Various authors have assigned that range from Late Plio cene to early Pleistocene (Scott, 1963; Soister, 1967; Scott, 1975). For the purpose of correlating the Nussbaum Allu Figure 3. Map Of Cav e Of The Wi nds, Manitou Springs, Colorado show ing locations of samplings sites Modi fi ed from Paul Burger, 199 6 Tunnel Entranc e Natural Entrance Cliffhanger Entranc e Manitou Gran d Cavern Entrance (Sealed ) Thieves Canyon Snider Ha ll Grand Concert Ha ll Silent Splendo r Heavenly Hall Mummys Alcove Snider Pi t Hole 6 Hole 5 Hole 1 Hole 3 Hole 4 Hole 2 0 12. 2 24. 4 Meters Fee t 0 40 80 N T A' A Passage below main cave (in red) Passage above main cave (in red) F RED G. L UISZER

PAGE 5

TIME in KARST 2007 161 vium with a paleomagnetic reversal, a more accurate date of the Nussbaum Alluvium was needed. is problem was solved by aminostratigraphy. Aminostratigraphy Most amino acids exist as two forms: Land D-isomers (Miller & Brigham-Grette, 1989). In a living organism, the amino acids are L-isomers. Aer the death of an or ganism, the amino acids racemize, which is the natural conversion of the L-isomers into D-isomers. Eventu ally the amino acids in the dead organism equilibrate to a 50/50 mixture of Land D-isomers. e amino acids used in the present study are D-alloisoleucine and Lisoleucine (A/I). ese amino acids are somewhat more complex, because L-isoleucine actually changes to a dif ferent molecule, D-alloisoleucine. is reaction, simi lar to racemization, is called epimerization (Miller and Brigham-Grette, 1989). e rate at which this reaction takes place is a func tion of temperature. For example, if the burial-tempera ture history for a group of mollusks of dierent ages has been the same, the ratio of the two amino acids alloiso leucine and isoleucine (A/I) in the mollusk shells can be used for relative dating and in some cases, absolute dating (Miller and Brigham-Grette, 1989). Because tem perature controls the rate of racemization, the tempera ture history of buried fossils must be considered before using A/I to derive ages. Solar insolation, re, altitude, and climate can eect the burial temperature of fossils. Diurnal or seasonal solar heating of fossils buried at shallow depths may accelerate racemization and increase the apparent age of the samples (Goodfriend, 1987; Miller and Brigham-Grette, 1989). erefore, samples should be obtained from depths that exceed 2 m (Miller and Brigham-Grette, 1989). During the intense heat associated with a re, racemization can also be greatly accelerated. For example, charcoal, which has a 14 C age of ~1500 years, found with snails at Manitou Cave suggests that the snails were exposed to a forest re before being transported into the cave. If so, the A/I of the snails may be anomalously high for their age. e altitude of the collection site can also aect ra cemization rates. For example, snails in this study were collected at altitudes between 1890 and 2195 m above sea level. Because of the normal adiabatic eect, the highest site averages about 1.7 C less than the lowest site. An other temperature variable is long-term climate change. For example, the Nussbaum Alluvium has probably been exposed to episodes of higher or lower temperatures for much longer periods of time than the younger alluvia. Because post-depositional thermal histories are impos sible to ascertain, the burial temperature for all alluvia in this study are assumed to be the same. Mollusks Results In all, over 10,000 mollusks, which included one spe cies of slug, one species of clam, and 24 species of snail, were identied and counted. e tabulated number for each species is the number of shells that could be iden tied. For example, the Louviers site had ~3,000 snails that could not be identied because they were too small (juvenile) or broke. Because of the small weight of the individual snails (0.3 to 5.0 mg) in relation to the 30 mg necessary for testing, only abundant species that oc curred at multiple sites could be used for the amino-acid study. e species chosen for the Nussbaum (Black Can yon) were V allonia cyclophorella and Pupilla muscorum and from the Verdos, V allonia cyclophorella and Gastro copta armifera (Table 1). All of the alloisoleucine and isoleucine (A/I) ratios of the snails along with laboratory identication numbers are tabulated in Table 2A. Table 2B contains the average and standard deviation of the A/I of selected snails from each site. Discussion Of A-I Ratios e epimerization rates of the four species used in this study are very similar. is is indicated in Table 2A by the comparable A/I values of dierent snail species at the same sample location. Moreover, shell size did not appear to greatly aect the A/I. For example, the average Gastro copta armifera shell weighs 5 mg; the V allonia cyclophore lla 1 mg; yet, the A/I for these shells from Manitou Cave are similar (Table 2A). e snails from the Verdos Alluvium, which include the Starlight, Fillmore, and Colorado City sites (Loca tions on Fig. 2), were used to test the A/I reproducibility of samples from one site and to compare the A/I from the dierent sites. Extra eort was put into assuring that the amino acid ratios of snails from Verdos Alluvium were as accurate as possible, because any errors in their A/I determination would greatly amplify the inaccuracies of the extrapolated age of the Nussbaum Alluvium (Fig. 4). erefore, the Starlight site was sampled three times and the Fillmore site, twice. Although each of the two sub-sites at Colorado City were sampled twice, the scarcity of V al lonia cyclophorella and Gastrocopta armifera necessitated combining all snail shells of similar species from the en tire site and from both sampling trips. One Starlight-site sample (Table 2A, Lab # AAL-5768) was excluded from the nal curve tting because it had an anomalously high ratio as compared to the others from that site. e snails that made up this sample (AAL-5768) may have been from an older reworked alluvium or there may have been a problem with their preparation or analysis. e data from the Colorado City site were also ex cluded from the nal curve tting, primarily because the A/I of the two species were very dierent from each other TIMING OF PASSAGE DEVELOPMENT AND SEDIMENTATION AT CAVE OF THE W INDS, MANITOU SPRINGS, COLORADO, USA

PAGE 6

TIME in KARST 2007 162 and both A/I were much lower than those from the Star light and Fillmore sites. eir low A/I would indicate that the Colorado City site may actually be either the Slocum or Louviers Alluvium. e anomalously low ratios, of course, could also be the result of contamination from modern shells or organic material. Determination of anomalously high or low ratios would be impossible without multiple sampling. Tak ing one sample per site would have been useless for this study. Two samples per site was acceptable when the A/I ratios were about the same for both species. W ith 12 sam ples from the Verdos Alluvium, it was quite appropriate to discard the highest and lowest ratios. Age Of e Alluvia e higher A/I of the snails from Black Canyon site, which is mapped as Nussbaum Alluvium, indicates that it is older than the other alluvia (Table 2A). Furthermore, by ascertaining the ages of the younger alluvia, plotting those against their relative A/I ratio, and tting a curve to the resultant plot, a equation can be derived that can be used to calculate the approximate age of the Nussbaum Alluvium. Snails from the modern ood plain (Fig. 2) were used to ascertain the A/I ratio of modern snails. About 50% of the snails at this site were alive when collected. e live snails were not analyzed because the esh, which may have dierent amino-acid ratios than the shells, might have contaminated the shell A/I ratios. Empty shells were used for analysis and assumed to be about one year old. ere is a pos sibility that the modern shells were reworked from older sediments such as the Piney Creek Alluvium. How ever, the abundant live snails mixed with the dead snails of the same spe cies suggests that all snail specimens were contemporaneous. e Piney Creek Alluvium site (Fig. 2) has been mapped as Piney Creek and Post-Piney Creek (Trim ble and Machette, 1979). Charcoal collected from the Piney Creek site (location on Fig. 2) was 14 C dated at 1542 130 years old (Table 3) indicating that the site should be mapped as Post-Piney Creek Alluvi um. e snails collected at Manitou Cave, which have relatively high A/I ratios, were initially thought to be about the same age as dated deposits at Narrows Cave. Narrows Cave is located ~0.4 km north of Manitou Cave contains ood deposits intercalated with owstone that has been dated and found to have a maximum uranium-thorium age of 32 2 Ka (Table 3). ey were thought to be the same age because the snails at Manitou Cave and the deposits at were both deposited by paleo-oods and both had similar heights above W il liams Canyon Creek. However, charcoal associated with the snails in Manitou Cave was 14 C dated with an age of 1552 75 years (Table 3). Apparently, either young char coal mixed with old snails during the paleo-ood or the snails were aected by a forest re that induced anoma lously high A/I ratios. is conicting evidence made it necessary to exclude the Manitou Cave data from the curve tting. e Louviers Alluvium site was mapped by Trim ble and Machette (1979). Elsewhere in Colorado the Louviers has been dated at 115 Ka by Machette (1975). Szabo (1980) gave a minimum age of 102 Ka and in ferred that the maximum age was ~150 Ka. e Fill more, Colorado City, and Starlight sites are all mapped as Verdos Alluvium (Trimble and Machette, 1979), which, in the Denver area, contains the 640-Ka Lava Creek B ash near its base (Sawyer et. al., 1995; Izett et. al., 1989; Machette, 1975). Because the Lava Creek B ash gives the maximum age for the Verdos Alluvium, AGE (ka ) LO W A/I CURVE FIT Age = 2002.4y 2 + 782.0y 4.2 AVERAGE A/I CURVE FIT Age = 1268.6y 2 + 863.8y 10. 0 Modern Centennia l Chesnut Starlight and Filmor e Black Cany on 2300 1900 1700 Figure 4. The A/I of sn ails from the Starlight and Filmore (Verdos Alluvium), Chesnut (Louviers Alluvium), Centennial (Piney Creek), and Modern Flood Plain sites are used fo r curve f itting. The diamond s represent the curve f it of the av erage A/I. T he circles and squares represent the curve f its of +/one standard de via tion of the A/I. The A/I of the snails from the Black Cany on site w ith the appropriate equations are used to ex trapolate the age of the Nussbaum Allu vi um (1.9+0.4/-0.2 Ma ). HIGH A/I CURVE FIT Age = 1064.2y 2 + 782.3y 11. 2 X X X Ln((1+A/I)/(1-0.77*A/I))-0.032 1. 0 0. 9 0. 8 0. 7 0. 6 0. 5 0. 4 0. 3 0. 2 0. 1 0. 0 0 500 1000 1500 2000 2500 F RED G. L UISZER

PAGE 7

TIME in KARST 2007 163 TIMING OF PASSAGE DEVELOPMENT AND SEDIMENTATION AT CAVE OF THE W INDS, MANITOU SPRINGS, COLORADO, USA T ab. 1: Species identied, their location, and amount of shells counted. Modern Flood Plain Centennial Manitou Cave Chesnut Filmore (Verdos) Starlight (Verdos) Colorado City Black Canyon (Nussbaum) East (Verdos) West (Verdos) Carychium exiguum (Say) 8 3 105 35 18 5 Cionella lubrica (Muller) 20 3 1 0 2 1 Columella alticola (ingersoll) 4 0 1 0 Derocerus spp. 1 0 4 1 Discus whitneyi 4 1 4 0 Euconulus fulvus (Muller) 16 2 20 2 2 1 Fossaria parva (Lea) 19 2 3 1 8 3 10 3 Gastrocopta armifera (Say) 1 0 1 0 47 7 5 1 38 7 17 7 5 2 Gastrocopta cristata Pilsbry 1 0 1 0 15 6 19 6 10 3 Gastrocopta holzingeri (Sterki) 6 2 6 1 96 39 37 12 3 1 Gastrocopta pellucida (Pfeier) 1 0 193 27 3 1 6 2 3 1 136 28 Gastrocopta procera (Gould) 6 2 5 1 35 6 5 2 14 5 3 1 Gyraulus parvus (Say) 2 1 53 13 Hawaiia minuscula (Binney) 23 6 43 5 82 12 134 14 84 14 12 5 20 7 12 3 26 5 Oreohelix spp. 28 4 Oxyloma spp. 20 2 6 2 5 2 Physa spp. 1 0 10 3 Pisidium casertanum (Poli) 13 4 200 50 Pupilla muscorum (Linne) 70 18 174 22 20 3 4 0 2 0 9 4 12 4 2 1 52 11 Pupoides albilabris (C.B. Adams) 7 1 4 2 8 3 Pupoides hordaceous (Gabb) 133 28 Pupoides inornata Vanatta 35 9 7 1 1 0 130 22 3 1 6 2 9 2 7 1 Stagnicola spp. 10 3 Succinea spp. 4 1 45 8 3 1 Vallonia cyclophorella (Sterki) 250 64 579 72 197 28 381 39 240 41 60 24 20 7 32 8 123 26 Vertigo gouldi and ovata 1 0 4 1 390 39 Zonitoides arboreus (Say) 1 0 96 14 1 0 7 3 20 7 15 4 6 1 TOTAL 394 100 809 100 713 100 989 100 585 100 248 100 304 100 397 100 483 100 # % # % # % # % # % # % # % # % # %

PAGE 8

TIME in KARST 2007 164 T ab. 2A: Alloisoleucine and isoleucine (A/I) ratios of snails. Sample Location Species Lab Number Results Average Standard Deviation Modern P AAL-5990 0.020 0.022 0.021 0.001 Modern V AAL-5989 0.021 0.020 0.021 0.001 Centennial P AAL-5970 0.023 0.021 0.022 0.001 Centennial V AAL-5969 0.024 0.032 0.028 0.004 Manitou Cave G AAL-5993 0.042 0.056 0.051 0.050 0.006 Manitou Cave P AAL-5992 0.039 0.044 0.041 0.041 0.002 Manitou Cave V AAL-5991 0.043 0.069 0.056 0.013 Chesnut GO AAL-5972 0.106 0.124 0.115 0.009 Chesnut V AAL-5971 0.106 0.103 0.105 0.002 Colorado City G AAL-5986 0.154 0.210 0.163 0.176 0.025 Colorado City V AAL-5985 0.076 0.083 0.080 0.004 Fillmore 1 G AAL-5976 0.279 0.274 0.277 0.003 Fillmore 1 V AAL-5975 0.298 0.275 0.287 0.012 Fillmore 2 G AAL-5988 0.239 0.233 0.236 0.003 Fillmore 2 V AAL-5987 0.283 0.270 0.277 0.007 Starlight P AAL-5768 0.423 0.414 0.420 0.419 0.004 Starlight V AAL-5767 0.276 0.317 0.296 0.296 0.017 Starlight 1 G AAL-5974 0.302 0.322 0.312 0.010 Starlight 1 V AAL-5973 0.307 0.231 0.224 0.254 0.038 Starlight 2 G AAL-5978 0.292 0.298 0.295 0.003 Starlight 2 V AAL-5977 0.246 0.244 0.245 0.001 Black Canyon P AAL-5766 0.502 0.531 0.543 0.529 0.526 0.015 Black Canyon V AAL-5765 0.545 0.545 0.546 0.544 0.515 0.576 0.545 0.018 V = Vallonia cyclophorella P = Pupilla muscorum G = Gastrocopta armifera GO = Vertigo gouldii and Vertigo ovata T ab. 2B: Average values and standard deviation of A/I ratios of selected snails from each site. Average standard deviation Average Average + standard deviation Modern Flood Plain 0.020 0.021 0.022 Centennial 0.021 0.025 0.029 Chesnut 0.103 0.110 0.117 Filmore and Starlight 0.249 0.275 0.301 Black Canyon 0.523 0.536 0.549 T ab. 3: Uranium-thorium and 14 C dates. 14 C Age (years B.P.) Lab. Number* Centennial site 1495 130 GX-15992 Manitou Cave 1505 75 GX-15993 Krueger Enterprises Inc. Uranium-thorium Age** (years B.P.) Narrows Cave 32,000 2,000 ** Dan Muhs, U.S.G.S., 1990, per. comm. F RED G. L UISZER the inferred maximum age of ~150 Ka was assigned to the Louviers Alluvium. e interpolated age of the Nussbaum Alluvium, therefore, represents its maximum age. Parabolic Curve Fitting Ages and A/I data (Table 2B ) from four of the younger alluvia, together with A/I data from the Nussbaum, were used to extrapolate the age of the Nussbaum. Various authors have applied linear and parabolic curve tting to amino acid data for both interpolation and extrapolation of age (Miller & Brigham-Grette, 1989). Mitterer & Kriausakul (1989) have employed the parabolic func tion (y=x 2 ) with good results. Ap plying the generalized parabolic equation (y=A+Bx+Cx 2 ) to my data resulted in a better curve t than the specialized parabolic function (y=x 2 ). Use of the specialized para bolic function assumes that the A/I ratio starts at 0.0 and that at an ini tial age near zero, the racemization rate is innitely large. e data from my study area suggest that both of these assumptions are invalid (Table 2B and Fig. 4). Ignoring the A+Bx terms ap pears to have little eect on curve tting of relatively young snails (<100 Ka). e generalized para bolic function, however, was used in this study because the age of the Nussbaum Alluvium is extrapolated 3 to 4 times beyond the oldest cali bration point. Parabolic-curve ts for the average ratio with error bars of one standard deviation indicate an extrapolated age for the Nussbaum Alluvium of 1.9 +0.4/-0.2 Ma (Fig. 4).

PAGE 9

TIME in KARST 2007 165 Extrapolating a date that is 3 to 4 times more than the maximum calibration date is a practice generally frowned up. I believe that by carefully collecting and handling samples, obtaining precise analysis of the ami no acids, acquiring the best age determinations of the younger deposits, and curve tting with the generalized parabolic function, I have ameliorated problems usually associated with such extrapolation. e 1.9-Ma date for the Nussbaum Alluvium is appropriate only for the unit mapped in the Manitou Springs area; it may not be cor relative with the type section in Pueblo, Colorado. e date, 1.9 +0.4/-0.2 Ma, which is the most accurate date available for the Nussbaum Alluvium, was used to cali brate the magnetostratigraphy of the sediments in Cave of the W inds. Magnetostratigraphy Rocks and unconsolidated sediments can be magnetized by the magnetic eld of the earth (Tarling 1983), acquir ing natural remanent magnetization (NRM). A type of NRM in sediments is detrital remanent magnetization (DRM), which is formed when the magnetic grains of a sediment, such as magnetite or hematite, are aligned with the earths magnetic eld during or soon aer deposition (Verosub, 1977). e DRM of a sediment has the same orientation as and its intensity is proportional to, the earths magnetic eld (Verosub, 1977). e magnetic eld of the earth has reversed many times in the past (Tarling, 1983). Polarity time scales have been constructed by compiling the reversals and the radiometrically derived dates of the rock in which the reversals are preserved, (Mankinen & Dalrymple, 1979; Harland et. al., 1982; Hailwood, 1989; Cande and Kent, 1992). ere are several ways to use this time scale to date sediments. By assuming that the top of a sediment section starts at the present and sedimentation has been uninterrupted, such as in deep ocean basins, it is a simple matter of counting the reversals and correlat ing them with the polarity time scale. Because of ero sion or a hiatus in deposition, however, the top of many sediment sections will have an older age that must be ascertained by some other technique before reversals in the section can be correlated with the polarity time scale. Another way of dating sediments is by pattern matching. If the sedimentation rate of an undated sec tion is constant or known and there are many reversals (5-10), the polarity record can be matched to the pattern of the polarity time scale to provide dating. is is pos sible because the timing of reversals is apparently ran dom (Tarling, 1983). erefore, the timing of a sequence of reversals is seldom repeated. Both of these techniques mentioned here were used to rene the age of the sedi ments at Cave of the W inds. Paleomagnetic Results All the paleomagnetic data from Hole 6 are presented to give an example of all the raw data from all sampling sites and how the samples responded to demagnetization (Ta ble 4). Inspection of the complete data set revealed that all samples responded very similarly to demagnetization. e complete data set of sites included in this study as well as other miscellaneous sites not used in this study are available from the author on computer storage disks. Sample depth and magnetic declination aer 15-mT AF demagnetization from each site were used to correlate the magnetic polarity within and between the Grand Concert Hall and nearby Heavenly Hall (Fig. 5). An ex ception to use of the 15-mT-AF demagnetization is Hole 5, where samples from 6.5 to 10.0 m were subjected to 20-, 25-, and 30-mT-AF demagnetization. e higher elds were applied in an attempt to remove secondary overprints. Even with the increasing demagnetization, TIMING OF PASSAGE DEVELOPMENT AND SEDIMENTATION AT CAVE OF THE W INDS, MANITOU SPRINGS, COLORADO, USA -100 -5 0 0 50 10 0 15 0 20 0 25 0 30 0 10 0 20 0 30 0 40 0 50 0 60 0 70 0 80 0 90 0 1000 0 -100 -5 0 0 50 10 0 15 0 20 0 25 0 30 0 10 0 20 0 30 0 40 0 50 0 60 0 70 0 80 0 90 0 1000 0 -100 -5 0 0 50 10 0 15 0 20 0 25 0 30 0 10 0 20 0 30 0 40 0 50 0 60 0 70 0 80 0 90 0 1000 0 -100 -5 0 0 50 10 0 15 0 20 0 25 0 30 0 100 200 300 400 500 600 700 800 900 100 0 0 -100 -5 0 0 50 10 0 15 0 20 0 25 0 30 0 10 0 20 0 30 0 40 0 50 0 60 0 70 0 80 0 90 0 1000 0 -100 -5 0 0 50 10 0 15 0 20 0 25 0 30 0 100 200 300 400 500 600 700 800 900 100 0 0 -100 -5 0 0 50 10 0 15 0 20 0 25 0 30 0 100 200 300 400 500 600 700 800 900 100 0 0 -100 -5 0 0 50 10 0 15 0 20 0 25 0 30 0 10 0 20 0 30 0 40 0 50 0 60 0 70 0 80 0 90 0 1000 0 -100 -5 0 0 50 10 0 15 0 20 0 25 0 30 0 100 200 300 400 500 600 700 800 900 100 0 0 Sniders Hall C entimeter s Hole 1 Hole 5 MA TUYA MA GA USS Kaena Olduv ai Polarity Chrons Polarity Subchrons Mummys Alcov e Hole 2 Hole 3 Hole 4 Hole 6 Heavenly Hall Figure 5. Cross-section and correlation of the magnetic declination of sampled pits and cored holes from the Grand Concert Hall and Hea ve nl y Hall. (See Figure 3 f or locations of pits and holes.) A A'

PAGE 10

TIME in KARST 2007 166 T ab. 4: Complete Paleomagnetic results of H ole 6, Grand Concert H all Sample Number Depth cm Natural 10 mT 15 mT Dec. Inc. Int. Dec. Inc. Int. Dec. Inc. Int. 11 80 -10 62 1.2E-4 -21 55 5.4E-5 -14 58 4.7E-5 12 93 -30 45 1.5E-4 -7 54 6.8E-5 -16 52 5.0E-5 13 105 -11 52 2.0E-4 -4 52 1.2E-4 -9 60 1.1E-4 14 118 -23 27 1.5E-4 -23 26 1.1E-4 -24 24 9.7E-5 21 121 -14 44 1.8E-4 -21 40 1.2E-4 -20 39 9.6E-5 22 131 -11 46 2.7E-4 -7 39 1.7E-4 -6 40 1.5E-4 23 141 -0 41 1.4E-4 -13 34 7.2E-5 -7 37 6.6E-5 24 151 4 42 3.2E-4 3 38 2.1E-4 2 38 1.8E-4 31 154 -11 40 3.8E-4 -15 32 2.1E-4 -14 34 1.9E-4 32 163 -27 40 2.2E-4 -21 38 1.4E-4 -24 36 1.3E-4 33 172 -24 41 2.6E-4 -29 36 1.8E-4 -30 35 1.6E-4 34 182 -17 40 2.5E-4 -19 38 1.7E-4 -20 37 1.6E-4 41 184 -18 38 4.4E-4 -18 40 3.1E-4 -18 39 2.9E-4 42 194 -15 41 1.8E-4 -24 42 1.1E-4 -20 36 9.6E-5 43 204 -17 58 1.4E-4 -45 62 4.7E-5 -50 64 3.4E-5 44 213 165 -32 1.1E-4 162 -28 9.9E-5 162 -29 9.1E-5 51 216 115 11 7.0E-5 149 -9 6.1E-5 156 -11 5.7E-5 52 226 155 -29 9.5E-5 165 -36 1.0E-4 168 -35 9.2E-5 53 236 -14 52 7.4E-5 144 75 1.4E-5 149 59 1.1E-5 54 246 30 50 1.3E-4 54 30 5.5E-5 62 30 5.0E-5 61 249 81 47 6.7E-5 114 14 4.6E-5 119 3 3.7E-5 62 257 -2 51 1.2E-4 34 41 2.3E-5 36 16 1.0E-5 63 264 176 58 2.7E-5 164 -7 2.3E-5 170 -16 2.3E-5 64 271 177 -18 5.0E-5 184 3 6.4E-5 183 3 6.2E-5 71 273 -12 88 6.7E-4 169 -9 4.7E-5 172 -8 4.5E-5 72 281 203 14 3.9E-5 142 12 7.4E-5 144 8 7.0E-5 81 283 127 -11 1.2E-4 131 -19 9.6E-5 131 -18 8.6E-5 82 294 150 -17 6.4E-5 156 -21 6.2E-5 158 -24 5.2E-5 83 305 93 16 5.4E-5 130 -16 5.2E-5 128 -17 4.5E-5 84 316 59 30 2.2E-5 140 -34 2.2E-5 142 -38 2.0E-5 91 319 28 62 7.2E-5 82 47 2.2E-5 100 29 1.7E-5 92 329 85 67 4.5E-5 148 5 3.8E-5 154 -2 3.5E-5 93 339 -2 53 5.5E-5 41 61 1.4E-5 41 56 1.3E-5 94 349 17 75 6.3E-5 140 34 2.5E-5 162 26 2.2E-5 101 352 101 -26 5.4E-5 159 15 4.5E-5 153 3 4.1E-5 102 362 81 76 5.2E-5 131 33 2.0E-5 150 17 1.5E-5 103 372 67 79 7.7E-5 139 -1 2.6E-5 137 -6 3.2E-5 104 382 178 11 4.7E-5 179 -16 4.7E-5 186 -21 4.7E-5 111 385 170 24 5.5E-5 165 -3 4.9E-5 165 -2 4.5E-5 112 396 159 -12 3.7E-5 156 -30 3.5E-5 157 -29 3.3E-5 113 407 184 -10 3.9E-5 174 -32 4.8E-5 172 -32 4.5E-5 F RED G. L UISZER

PAGE 11

TIME in KARST 2007 167 Sample Number Depth cm Natural 10 mT 15 mT Dec. Inc. Int. Dec. Inc. Int. Dec. Inc. Int. 121 420 -37 -30 2.4E-5 -70 -57 1.2E-5 -49 -54 1.0E-5 122 431 228 52 2.5E-5 214 25 1.4E-5 208 18 1.1E-5 123 439 -44 55 3.1E-5 249 16 1.1E-5 266 16 1.1E-5 124 451 126 3 2.1E-5 145 -32 2.2E-5 142 -36 1.9E-5 131 453 211 7 1.6E-5 136 -38 9.8E-6 164 -55 9.8E-6 132 463 172 37 1.5E-5 144 -43 2.4E-5 147 -40 1.9E-5 133 472 69 -21 9.3E-6 149 -46 1.6E-5 152 -50 1.6E-5 134 481 263 19 5.4E-6 167 -32 1.3E-5 177 -28 1.2E-5 141 484 115 8 1.9E-5 183 -35 1.4E-5 174 -40 1.1E-5 142 493 189 -6 7.2E-5 188 -19 5.8E-5 192 -22 5.1E-5 143 503 176 62 2.1E-5 158 23 9.6E-6 163 20 8.2E-6 144 512 68 35 1.3E-5 186 -36 6.3E-6 175 -36 6.6E-6 151 514 -19 40 3.4E-5 -39 15 8.3E-6 -15 15 5.9E-6 152 525 41 70 4.7E-5 139 66 1.8E-5 146 60 1.6E-5 153 535 -6 46 8.1E-5 -4 33 3.3E-5 -7 36 2.6E-5 154 545 189 33 1.2E-4 185 21 1.1E-4 182 21 1.1E-4 161 547 173 57 2.8E-5 186 2 2.4E-5 187 -0 2.4E-5 162 558 208 7 2.6E-5 200 -28 3.8E-5 201 -28 3.2E-5 163 568 10 68 1.7E-5 206 -25 9.1E-6 211 -33 1.1E-5 171 570 146 73 4.4E-5 175 27 2.9E-5 173 31 2.7E-5 172 580 255 54 10.0E-6 190 -46 1.5E-5 176 -43 1.6E-5 173 591 131 63 4.3E-5 161 14 2.1E-5 160 10 2.2E-5 174 601 39 67 2.2E-5 148 14 6.4E-6 150 2 7.4E-6 181 603 92 52 7.9E-6 163 -18 6.4E-6 135 -31 4.9E-6 182 613 2 69 2.6E-5 152 87 1.2E-5 113 78 8.2E-6 183 624 -44 19 2.1E-5 -69 -45 1.7E-5 -67 -53 1.4E-5 184 634 -16 5 1.6E-5 267 -55 1.1E-5 -86 -59 8.6E-6 191 636 106 60 2.0E-6 112 -37 3.0E-6 160 1 1.4E-6 192 646 -38 7 2.3E-5 -54 -26 1.8E-5 -55 -40 1.2E-5 193 657 2 47 2.6E-5 -4 19 5.7E-6 -26 -27 3.0E-6 194 667 14 64 4.3E-5 42 63 1.6E-5 45 61 1.1E-5 201 669 8 45 4.5E-5 25 43 2.5E-5 24 45 1.6E-5 202 677 22 84 5.0E-5 183 84 2.6E-5 192 80 1.8E-5 203 685 48 52 3.3E-5 74 50 1.7E-5 78 44 1.2E-5 204 692 25 48 1.9E-5 56 25 6.3E-6 78 -17 1.6E-6 however, the declination of the deeper samples at Hole 5 have greater variability than those of shallower samples (Fig. 5). Additionally, the polarity results from Hole 5 are shown in Fig. 6, which also shows the correlation with the known paleomagnetic record and stratigraphy of the cave sediments. Criteria For Reversal Assignment Sequences of samples that had an average declination of ~0.0 and an average inclination of ~35.0 were assigned to normal polarity. e ideal inclination for DRM in the Manitou Springs area should be ~60. e low values recorded at Cave of the W inds are considered to be the TIMING OF PASSAGE DEVELOPMENT AND SEDIMENTATION AT CAVE OF THE W INDS, MANITOU SPRINGS, COLORADO, USA

PAGE 12

TIME in KARST 2007 168 result of inclination error resulting from sediment com paction (Verosub, 1977). Sequences of samples that had an average declination of ~180 were assigned a reversed polarity. In most cases, inclinations of these samples were variable, ranging mostly between -35.0 and +10.0. Because of this variability, the sample declinations were used to determine reversals (Fig. 5). ese anomalous inclinations appear to be related to post-depositional ac quisition of remanent magnetization. Eects Of Post-Depositional Remanent Magnetization On Sediments Most post-depositional remanent magnetization (PDRM) is the result of realignment of the magnetic par ticles during compaction and especially dewatering, both of which can take place thousands to millions of years aer deposition (Verosub, 1977). A possible explana tion of the variability of the inclination of the reversed samples is that these samples were compacted and dewatered during a normal polarity interval, overprint ing a normal component. e de watering and compaction may have occurred rather quickly following rapid draining of the water in the cave passages related to downcut ting by Fountain Creek. Mud cracks present in the top two meters of the sediments at the Grand Concert Hall combined with their mostly normal polarity (Fig. 6) indicate that this is a plausible explanation. Further mi cro-sampling and precision analysis would be necessary to ascertain the mechanism responsible for the dif ference in the inclinations. Chemical remanent magne tization (CRM) may also be a con tributing factor to the inclination anomalies. Alteration and oxidation of iron-bearing minerals in the sedi ments may contribute to the CRM. is could only be a factor in the top two meters of coarse sediments (Fig. 6), which contain unaltered miner als; because the general oxidizing conditions and neutral to slightly alkaline pH of percolating cave wa ters through these sediments would preclude mobilization or precipita tion of iron oxides. e underlying soil-derived clays, which have al ready undergone prolonged oxida tion before being deposited in the cave, are chemically stable and would not be vulnerable to CRM. Paleomagnetic Correlation Because there are no independent dates on the cave sedi ments, correlation of the magnetic polarity record of the Cave of the W inds sediments with the accepted polarity time scale is dircult. It requires matching the sequence of known polarity events with the Cave of the W inds record. e age of the Nussbaum Alluvium, which is ap parently related to the uppermost coarse cave sediments, however, can be used to help constrain the paleomag netic correlation. e relationship of the Nussbaum Al luvium to the detrital sediments in Cave of the W inds will be discussed in detail. As discussed previously, the clay is deposited in the cave below the phreatic-vadose interface where sediment-laden streams enter water-lled passages. e 1.5 2.0 2.5 3.0 3.5 4. 0 4.5 Pleistocene Ti me (Ma) Epoch Pliocen e MA TU YAMA GAUSS GILBERT Polarity Chrons Polarity Subchrons Mammoth Kaena Reunion Cochiti Nunivak 0 2 3 4 5 6 7 8 9 10 Meter s Declination Angular limestone clasts up to 50 cm contains silt, clay, smal l (bat) and large bones near base. Appears to be artificial fill. Floor Flow stone Brow n, laminated, micaceous silt contains clay intraclasts. Brow n, laminated, micaceous silt interbedded wi th mottled red clay Mostly silt near top an d clay near bottom. Beds dip 20 degress east. Green Clay Red Sandstone Limestone Chert Clay Intraclast Wh ite clay Reddish bro wn clay contains red, bro wn purple, green and blue mottles. Reddish bro wn clay contains yello w and purple mottles. Contains limestone and purple sandstone clasts. Bedrock Figure 6. Paleomagnetic correlation and stratigraphy of Grand Concert Hall Hole 5. Paleomagnetic time scale adapted from Harland and others (1982) 1 Thin bed (5 cm) of Mn Fe-oxides and solution residue. Normal Polarity Reversed Polarity Olduv ai F RED G. L UISZER

PAGE 13

TIME in KARST 2007 169 Nussbaum Alluvium was being deposited at the same time that clay was being deposited in the Grand Con cert Hall the (Fig. 7A). As Fountain Creek downcut and moved to the south, the water table dropped (Fig. 7B). e drop in the water table coincided with drop in the water depth in rooms like the Grand Concert Hall. As the water depth dropped, the velocity of the water passing through the room increased. e increased stream en ergy changed the sedimenta tion regime from clay depo sition to silt, sand, and gravel deposition (Fig. 7B). Fluvial sedimentation at Cave of the W inds stopped as Fountain Creek moved further to the south and downcut further (Fig. 7C). e relationship between the Nussbaum Al luvium and the sediments in the cave indicate that the siltclay interface in the Grand Concert Hall took place aer the Nussbaum was deposit ed. More specically, the siltclay interface should be the same age as the Nussbaum Alluvium minus the time it took for Fountain Creek to downcut and drop the water table to the level of the Grand Concert Hall (Fig. 7B). e sediment oor of the Grand Concert Hall, where the paleomagnetic data was obtained, is about 20 m below the Nussbaum Alluvium. e age of the Nussbaum Alluvium (~1.9 Ma) and its height above modern streams (200 m) provides an estimate of the average down-cutting rate of 10.5 cm/1000 years. Accord ingly, accumulation of coarse sediments in the cave 20 m below the Nussbaum Allu vium probably would have begun ~1.7 Ma. e estimated 1.7 Ma age of the clay-coarse sedi ment interface correlates well with the onset of the Olduvai Subchron at 1.9 Ma ( ~2.2 m depth, Fig. 6). is is the most probable correlation. Alter natively, one could match the normal-polarity sequence (1.0 to 2.2 m depth, Fig. 6) with the Jaramillo Subchron (Harland et al., 1982) or the Gauss Chron (Fig. 6). ese correlations, however, would result in an age of ~1.0 Ma or ~ 2.6 Ma, respectively, for the clay-coarse-sediment in terface, which is estimated to be 1.7 Ma, thereby making these alternate correlations unlikely. TIMING OF PASSAGE DEVELOPMENT AND SEDIMENTATION AT CAVE OF THE W INDS, MANITOU SPRINGS, COLORADO, USA = Nussbaum Allu vi um Fountain Cree k Wa ter tab le W illiams Canyon Creek Grav el Sand Silt Clay Mn-oxide Fe-oxide Solution debris Grav el Sand Silt Clay Mn-oxide Grav el Sand Silt Clay Mixing Zone A. ~2 Ma Wa ter ascen ding from Ute Pass Faul t = Grand Concert Hall, Cave Of The Wi nd s Sout h Nort h Fountain Cree k Wa ter tab le B. ~1.8 Ma Sout h Nort h C. Present Wa ter tab le Manitou Cave Fountain Creek W illiams Canyon Creek Figure 7. Schematic cross sections showing the sequence of changes in the wa ter table, topographic setting, and depositional phases ov er the last ~2 Ma. Sout h Nort h Tex t in green are sediments that are being deposited in the ca ve = outline of cav er n

PAGE 14

TIME in KARST 2007 170 e complete paleomagnetic correlation shown on Fig. 6 follows from correlation of the normal-polarity in terval between 1.0 and 2.2 m in depth with the Olduvai Subchron. According to the correlation suggested here, the oldest cave sediment was deposited about 4.3 Ma, a date that agrees quite well with the previously discussed probable age of the major onset of cave formation (7 Ma to 4 Ma). Cave of the W inds is a phreatic cave dissolved from the calcite-rich Manitou, W illiams Canyon, and Leadville Formations. Dissolution occurred along joints associ ated with Laramide faulting and folding. Paleokarst fea tures, such as sediment-lled ssures and caves, indicate that some of the passages at Cave of the W inds are re lated to cave-forming episodes that started soon aer the deposition of the Ordovician Manitou Formation and continued to the beginning of the Cretaceous Laramide Orogeny. Most speleogenesis, however, occurred in the last ~5.0 Ma. e Nussbaum Alluvium was assigned an age of ~1.9 Ma by means of aminostratigraphy. e age of the Nussbaum Alluvium and its relation to coarse grained sediments at Cave of the W inds were used to x an age of ~1.7 Ma for the onset of coarse grained sedimentation in the cave. is enabled the identication of the Olduvai Polarity Subchron in the coarse grained sediments. Cor relation of the magnetostratigraphy of cave sediments with the accepted polarity time scale indicates that the dissolution of cave passage started ~4.2 Ma and stopped ~1.5 Ma. CONCLUSIONS REFERENCES Blanton, T. L., 1973: e Cavern Gulch Faults and the Fountain Creek Flexure, Manitou Spur, Colorado [M.S. thesis]: Syracuse University, New Y ork, 90 p. Bianchi, L., 1967: Geology of the Manitou-Cascade Area, El Paso County, Colorado with a study of the perme ability of Its crystalline rocks [M.S. esis]: Golden, Colorado School of Mines. Cande, S. C., and D. Kent., 1992: A new geomagnetic po larity time scale for the Late Cretaceous and Ceno zoic: Journal of Geophysical Research, 97, 10, 1317. Epis, R. C., and C.E. Chapi, 1975: Geomorphic and tec tonic implications of the Post-Laramide, Late Eo cene Erosion surface in the Southern Rocky Moun tains, in Curtis, B.F., ed., Cenozoic History of the Southern Rocky Mountains: Geological Society of America Memoir 144, 45-74. Fish, L., 1988: e real story of how Cave of the W inds Formed: Rocky Mountain Caving, 5, 2, 16-19. Forster, J. R., 1977: Middle Ordovician subaerial expo sure and deep weathering of the Lower Ordovician Manitou Formation along the Ute Pass Fault zone: Geological Society of America Abstracts with Pro grams, 9, 722. Goodfriend, G. A., 1987: Evaluation of amino-acid race mization/epimerization dating using radiocarbondated fossil land snails: Geology 15, 698-700. Hailwood, E. A., 1989: e role of magnetostratigraphy in the development of geological time scales; Pale oceanography, 4, 1, 1-18. Hamil, M. M., 1965: Breccias of the Manitou Springs area, Colorado [M.S. thesis]: Louisiana State Uni versity, 43 p. Harland, W B., et al., 1982: A geologic time scale: Cam bridge, Great Britain, Cambridge University Press, 66 p. Hose, L. D., & Esch, C. J., 1992: Paleo-cavity lls formed by upward injection of clastic sediments to lithostat ic load: exposures in Cave of the W inds, Colorado [abs.]: National Speleological Society Convention Program, Salem, Indiana, p.50 Izett, G. A., Obradovich, J. D., & H.H. Mehnert., 1989: e Bishop Ash Bed (Middle Pleistocene) and some older (Pliocene and Pleistocene) chemically and mineralogically similar ash beds in California, Ne vada, and Utah: U. S. Geological Survey Bulletin, 1675, 37 p. Luiszer, F. G., 1997: Genesis of Cave of the W inds, Mani tou Springs, Colorado, [Ph. D. thesis]: Boulder, Uni versity of Colorado, 112 p. Machette, M. M., 1975: e Q uaternary geology of the Lafayette Q uadrangle, Colorado, [M. S. thesis]: Boulder, University of Colorado, 83 p. F RED G. L UISZER

PAGE 15

TIME in KARST 2007 171 TIMING OF PASSAGE DEVELOPMENT AND SEDIMENTATION AT CAVE OF THE W INDS, MANITOU SPRINGS, COLORADO, USA Mankinen, E. A., & Dalrymple, G. B., 1979: Revised geo magnetic polarity time scale for the interval 0-5 m. y. B. P.; Journal of Geophysical Research, 84, B2, 615-626. Miller, G. H., & Brigham-Grette, J., 1989: Amino acid geochronology: Resolution and precision in carbon ate fossils in IN Q UA Q uat. Dating Methods, Rutter and Brigham-Grette Eds. Pergamon Press. Mitterer, R. M., & Kriausakul, 1989: Calculation of amino acid racemization ages based on apparent parabolic kinetics: Q uaternary Science Reviews, 8, 353-357. Morgan, G. B., 1950: Geology of W illiams Canyon area, north of Manitou Springs, El Paso County, Colo rado (Masters thesis): Golden, Colorado School of Mines, 80 p. Mutschler, F. E., Larson, E. E., & R.M. Bruce: 1987: Laramide and younger magmatism in ColoradoNew petrologic and tectonic variations on old themes: Colorado School of Mines Q uarterly 82, 4, 1-47. Sawyer, D. A. et al., 1995: New chemical criteria for Q uaternary Y ellowstone tephra layers in central and western North America: Geological Society of America Abstracts with Programs, 27, 6, 109. Scott, G. R., 1963, Nussbaum Alluvium of Pleistocene(?) age at Pueblo, Colorado. U. S. Geological Survey Professional Paper, 475-C, C49-C52 Scott, G. R., 1975, Cenozoic surfaces and deposits in Curtis, B. F., ed., Cenozoic History of the Southern Rocky Mountains: Geological Society of America Memoir 144, 227-248. Soister, E., 1967, Relation of Nussbaum Alluvium (Pleis tocene) to the Ogallala Formation (Pliocene) and to the Platte-Arkansas divide, Southern Denver Basin, Colorado. U. S. Geological Survey Professional Pa per 575-D, p.D39-D46. Szabo, B. J., 1980, Results and assessment of uraniumseries dating of vertebrate fossils from Q uaternary alluviums in Colorado: Arctic and Alpine Research, 12, 95-100. Tarling, D. H., 1983, Palaeomagnetism; principles and applications in geology, geophysics and archaeol ogy: Chapman and Hall Ltd., London, 379 p. Trimble, D. E., & Machette, M. M., 1979, Geologic map of the Colorado Springs-Castle Rock Area, Front Range Urban Corridor, Colorado; U. S. Geological Survey, 1:100,000, Map I-857-F Tweto, O., 1975, Laramide (Late Cretaceous-Early Ter tiary) Orogeny in the Southern Rocky Mountains in Curtis, B.F., ed., Cenozoic History of the Southern Rocky Mountains: Geological Society of America Memoir 144, 1-44. Verosub, K. L., 1977, Depositional and post-depositional processes in the magnetization of sediments: Re views of Geophysics and Space Physics, 15, 129143.

PAGE 17

H O W LONG DOES EVOLUTION OF THE TROGLOMORPHIC FORM TAKE ? E STIMATING DIVERGENCE TIMES IN A ST Y ANA X ME X ICANUS KAKO DOLGO TRAJA EVOLUCIJA TROGLOMORFNIH OBLIK? OCENJEVANJE DIVERGENfNIH fASOV PRI AST Y ANA X ME X ICANUS Megan L. PORTER 1 Katharina DITTMAR 2 & Marcos P REZ-LOSADA 3 Izvleek UDK 551.44:597 591.542 Megan L. Porter, Katharina Dittmar & Marcos Prez-Losada: Kako dolgo traja evolucija troglomorfnih oblik? Ocenjevanje divergennih asov pri Astyanax mexicanus Znailnosti, ki vkljuujejo tudi kolonizacijske poti in obstoj tako epigejinih kot hipogejinih populacij vrste Astyanax mexica nus, ji omogoajo, da predstavlja zanimiv sistem za prouevanje evolucije in asa, potrebnega za razvoj podzemeljskih troglo morfnih oblik. Za A. mexicanus smo na podlagi e objavljenih sekvenc ocenili divergenni as ob uporabi: 1) dveh razlinih populacijskih mitohondrialnih podatkovnih baz (citokrom b in NADH dehidrogenaze 2), obe z natanno in sproeno metodo molekularne ure, in 2) razirjenega logenetskega pris topa v kombinaciji s fosilno kalibracijo ter tirimi jedrnimi geni (rekombinacijski aktivacijski gen, forkhead kontrolni gen in -tropomiozin) in dvema mitohondrialnima genoma (16S rDNA in citokrom b). Ob uporabi navedenih podatkovnih baz smo ocenili divergenni as za tri dogodke v zgodovini razvoja troglomorfnih populacij A. mexicanus. Prvi, razhajanje med podzemeljskimi haplotipi se je zgodilo v Pleistocenu, verjetno v odvisnosti od nihanja vode, ki je omogoilo kolonizacijo in posledino izolacijo v novih podzemeljskih habitatih. Drugi, verjetno je v povezavi s pleistocenskimi dogodki pri eni liniji podzemeljskih populacij A. mexicanus prilo do introgresivne hibridizacije s takratnimi povrinskimi populacijami (0.26-2.0 Ma). Z uporabo divergennega asa povrinskih populacij tistih linij, ki ne kaejo introgresije ocenjujemo, da je troglomorfna oblika A. mexicanus mlaja od 2,2 (ocene fosilne kalibracije) do 5,2 milijona let (cytb ocena) (Pliocen). Kljune besede: Astyanax mexicanus, divergenni as, troglo morzem, podzemlje, speleobiologija, evolucija. 1 Dept. of Biological Sciences, University of Maryland Baltimore County, Baltimore, MD, USA; e-mail: porter@umbc.edu 2 Dept. of Molecular Biology, University of W yoming, Laramie, WY USA 3 GENOMA LLC, 50E W oodland Hills, Provo, UT 84653-2052, USA Received/Prejeto: 06.12.2006 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 173 Abstract UDC 551.44:597 591.542 Megan L. Porter, Katharina Dittmar & Marcos Prez-Losada: How long does evolution of the troglomorphic form take? Esti mating divergence times in Astyanax mexicanus Features including colonization routes (stream capture) and the existence of both epigean and cave-adapted hypogean popula tions make Astyanax mexicanus an attractive system for investi gating the subterranean evolutionary time necessary for acqui sition of the troglomorphic form. Using published sequences, we have estimated divergence times for A. mexicanus using: 1) two dierent population-level mitochondrial datasets (cyto chrome b and NADH dehydrogenase 2) with both strict and relaxed molecular clock methods, and 2) broad phylogenetic approaches combining fossil calibrations and with four nuclear (recombination activating gene, seven in absentia, forkhead, and -tropomyosin) and two mitochondrial (16S rDNA and cytochrome b) genes. Using these datasets, we have estimated divergence times for three events in the evolutionary history of troglomorphic A. mexicanus populations. First, divergence among cave haplotypes occurred in the Pleistocene, possibly correlating with uctuating water levels allowing the coloni zation and subsequent isolation of new subterranean habitats. Second, in one lineage, A. mexicanus cave populations expe rienced introgressive hybridization events with recent surface populations (0.26-2.0 Ma), possibly also correlated with Pleis tocene events. Finally, using divergence times from surface populations in the lineage without evidence of introgression as an estimate, the acquisition of the troglomorphic form in A. mexicanus is younger than 2.2 (fossil calibration estimates) 5.2 (cytb estimate) Ma (Pliocene). Key words: Astyanax mexicanus, divergence time, troglomor phy, subterranean, evolution.



PAGE 1

H O W LONG DOES EVOLUTION OF THE TROGLOMORPHIC FORM TAKE ? E STIMATING DIVERGENCE TIMES IN A ST Y ANA X ME X ICANUS KAKO DOLGO TRAJA EVOLUCIJA TROGLOMORFNIH OBLIK? OCENJEVANJE DIVERGENfNIH fASOV PRI AST Y ANA X ME X ICANUS Megan L. PORTER 1 Katharina DITTMAR 2 & Marcos P REZ-LOSADA 3 Izvleek UDK 551.44:597 591.542 Megan L. Porter, Katharina Dittmar & Marcos Prez-Losada: Kako dolgo traja evolucija troglomorfnih oblik? Ocenjevanje divergennih asov pri Astyanax mexicanus Znailnosti, ki vkljuujejo tudi kolonizacijske poti in obstoj tako epigejinih kot hipogejinih populacij vrste Astyanax mexica nus, ji omogoajo, da predstavlja zanimiv sistem za prouevanje evolucije in asa, potrebnega za razvoj podzemeljskih troglo morfnih oblik. Za A. mexicanus smo na podlagi e objavljenih sekvenc ocenili divergenni as ob uporabi: 1) dveh razlinih populacijskih mitohondrialnih podatkovnih baz (citokrom b in NADH dehidrogenaze 2), obe z natanno in sproeno metodo molekularne ure, in 2) razirjenega logenetskega pris topa v kombinaciji s fosilno kalibracijo ter tirimi jedrnimi geni (rekombinacijski aktivacijski gen, forkhead kontrolni gen in -tropomiozin) in dvema mitohondrialnima genoma (16S rDNA in citokrom b). Ob uporabi navedenih podatkovnih baz smo ocenili divergenni as za tri dogodke v zgodovini razvoja troglomorfnih populacij A. mexicanus. Prvi, razhajanje med podzemeljskimi haplotipi se je zgodilo v Pleistocenu, verjetno v odvisnosti od nihanja vode, ki je omogoilo kolonizacijo in posledino izolacijo v novih podzemeljskih habitatih. Drugi, verjetno je v povezavi s pleistocenskimi dogodki pri eni liniji podzemeljskih populacij A. mexicanus prilo do introgresivne hibridizacije s takratnimi povrinskimi populacijami (0.26-2.0 Ma). Z uporabo divergennega asa povrinskih populacij tistih linij, ki ne kaejo introgresije ocenjujemo, da je troglomorfna oblika A. mexicanus mlaja od 2,2 (ocene fosilne kalibracije) do 5,2 milijona let (cytb ocena) (Pliocen). Kljune besede: Astyanax mexicanus, divergenni as, troglo morzem, podzemlje, speleobiologija, evolucija. 1 Dept. of Biological Sciences, University of Maryland Baltimore County, Baltimore, MD, USA; e-mail: porter@umbc.edu 2 Dept. of Molecular Biology, University of W yoming, Laramie, WY USA 3 GENOMA LLC, 50E W oodland Hills, Provo, UT 84653-2052, USA Received/Prejeto: 06.12.2006 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 173 Abstract UDC 551.44:597 591.542 Megan L. Porter, Katharina Dittmar & Marcos Prez-Losada: How long does evolution of the troglomorphic form take? Esti mating divergence times in Astyanax mexicanus Features including colonization routes (stream capture) and the existence of both epigean and cave-adapted hypogean popula tions make Astyanax mexicanus an attractive system for investi gating the subterranean evolutionary time necessary for acqui sition of the troglomorphic form. Using published sequences, we have estimated divergence times for A. mexicanus using: 1) two dierent population-level mitochondrial datasets (cyto chrome b and NADH dehydrogenase 2) with both strict and relaxed molecular clock methods, and 2) broad phylogenetic approaches combining fossil calibrations and with four nuclear (recombination activating gene, seven in absentia, forkhead, and -tropomyosin) and two mitochondrial (16S rDNA and cytochrome b) genes. Using these datasets, we have estimated divergence times for three events in the evolutionary history of troglomorphic A. mexicanus populations. First, divergence among cave haplotypes occurred in the Pleistocene, possibly correlating with uctuating water levels allowing the coloni zation and subsequent isolation of new subterranean habitats. Second, in one lineage, A. mexicanus cave populations expe rienced introgressive hybridization events with recent surface populations (0.26-2.0 Ma), possibly also correlated with Pleis tocene events. Finally, using divergence times from surface populations in the lineage without evidence of introgression as an estimate, the acquisition of the troglomorphic form in A. mexicanus is younger than 2.2 (fossil calibration estimates) 5.2 (cytb estimate) Ma (Pliocene). Key words: Astyanax mexicanus, divergence time, troglomor phy, subterranean, evolution.

PAGE 2

TIME in KARST 2007 174 I NTRODUCTION Understanding the evolution of the cave form has fasci nated biologists interested in subterranean faunas since Darwin. Termed troglomorphy, the suite of progressive and regressive characters associated with cavernicolous animals can be observed in the worldwide convergence of form found in the cave environment, exhibited in similar structural, functional, and behavioral changes across di verse taxonomic groups. Much of the debate over troglo morphy has centered on the evolutionary mechanisms responsible for character regression, generally argued to be either neutral mutation or natural selection. Several studies, (Gammarus minus Culver et al. 1995; Astyanax mexicanus Jeery, 2005) have shown eye degeneration is the result of selection, and, in the case of A. mexica nus, is caused by the pleiotropic eects of natural selec tion for constructive traits. Another, less studied, aspect of understanding troglomorphy is the evolutionary time required to gain the cave form. Because it is generally dif cult to pinpoint the time of subterranean colonization and isolation from surface ancestors, few troglomorphic species oer the opportunity for quantitative estimates of the evolutionary time spent in the subterranean realm. erefore, the time of cave adaptation is thought of in relative terms, where the degree of eye and pigment re duction indicates the period of cavernicolous evolution and therefore the relative phylogenetic age of each spe cies (Aden, 2005). In evolutionary studies of cave adaptation, Asty anax mexicanus has become a model system (Jeery, 2001). e advantageous features of A. mexicanus as a model system include the existence of both surface and troglomorphic cavesh populations, with several cave sh populations having evolved constructive and regres sive changes independently (Jeery, 2001). Furthermore, since the discovery of the species in 1936 (Hubbs & Innes, 1936), there has been an extensive amount of research devoted to characterizing developmental, phylogenetic, taxonomic, and biogeographic aspects of the species (Jef fery, 2001; Mitchell et al. 1977; W iley & Mitchell, 1971;). In terms of being a model system for understanding the evolution of the troglomorphic form, A. mexicanus has at least one additional favorable attribute. e primary mode of A. mexicanus subterranean colonization is via stream capture, with most of the captured surface drain ages no longer supporting epigean populations (Mitchell et al. 1977). ese captures provide discrete coloniza tion events correlated with divergence time from surface populations and therefore with the time of subterranean evolution. Molecular studies that have looked at A. mexicanus phylogeography indicate that at least two independent invasions of surface Astyanax have occurred (Dowling et al. 2002a; Strecker et al. 2003, 2004). ese two distinct A. mexicanus genetic lineages consist of cave sh from La Cueva Chica, La Cueva de El Pachn, El Stano de Y erbaniz, El Stano de Molino, El Stano de Pichijumo, and La Cueva del Ro Subterrneo (lineage A) and from La Cueva de los Sabinos, El Stano de la Tinaja, La Cueva de la Curva, and El Stano de Las Piedras (Lineage B) with dierent evolutionary histories Lineage A clus ters with closely related epigean populations while lin eage B has no closely related epigean counterparts. e close association of Lineage A to epigean populations (as estimated by mitochondrial markers) is thought to be the result of either recent subterranean colonization or reect recent introgressive hybridization with surface populations, while lineage B is considered to be a more ancient colonization event from surface populations that are extinct in the region (Dowling et al. 2002a; Strecker et al. 2004). Although the evolutionary histories of dif ferent hypogean A. mexicanus populations are complex, the two lineages oer the unique opportunity to estimate the divergence time required for the evolution of the tro glomorphic form based on discrete times of colonization and the previous molecular studies of their phylogeogra phy. At least one other study has estimated lineage ages in A. mexicanus populations; however, this study was based on a single gene molecular clock estimate and did not specically estimate the divergence times of the cave populations (Strecker et al. 2003). Here we use three dif ferent sets of publicly available sequence data and known fossil calibrations and apply multiple phylogenetic ap proaches to estimate the age of cave colonization and stream capture events, and to provide an estimate of the time necessary to acquire the troglomorphic form in A. mexicanus. M EGAN L. PORTER, K ATHARINA DITTMAR & M ARCOS P REZLOSADA

PAGE 3

TIME in KARST 2007 175 Sequence Data Data were acquired from Genbank (http://www.ncbi. nlm.nih.gov/) from previously published studies of A. mexicanus and characiform shes (Tab. 1). ese studies provided three dierent datasets, consisting of: 1) popu lation-level haplotype datasets for the mitochondrial cy tochrome b (cytb; Strecker et al. 2004) and NADH dehy drogenase 2 (ND2; Dowling et al. 2002a) genes, and 2) a species-level dataset of four nuclear (recombination ac tivating gene RAG2; seven in absentia sina; forkhead h; and r -tropomyosin trop) and two mitochondrial genes (16S rDNA and cyt b ) from representatives within the Otophysi (Calcagnotto et al. 2005). Divergence times from all three data sets were estimated and compared. Species-level Phylogenetic Analyses e species-level dataset included selected Otophysi, Characiformes, and Characidae sequences (see Tab. 1), and was analyzed using Anotophysi species as outgroups. Representative A. mexicanus cyt b haplotype sequences from the Strecker et al., (2004) study were included in the dataset of characiform species to estimate diver gence times based on fossil calibrations for comparison with population-based estimates utilizing substitution rates. Alignments of protein-coding regions were trivial and were accomplished using amino acid translations. Sequences of the trop gene spanned an intron, which was removed due to signicant length variation (70-836 bp) leading to ambiguous alignments. e alignment of the 16s rDNA gene was generated using the E-INS-i accuracy-oriented strategy of MAFFT v.5 (Katoh et al., 2005). All of the individually aligned genes were then concatenated to form a single dataset consisting 3770bp in length. e concatenated dataset was analyzed with PAUP* 4.0b10 (Swoord, 2000) using maximum parsi mony and implementing the parsimony ratchet method (Nixon, 1999) using a batch le generated by PAUPRat with the default parameters for 5000 replicates (Sikes & Lewis, 2001). Divergence time estimation Population analysis. Dates of divergence were inferred for A. mexicanus lineage A and B cave sh populations using the cyt b and ND2 datasets with BEASTv1.4 (Drummond & Rambaut, 2003). Because the cytb and ND2 haplotype datasets were generated from dierent studies, they can not be combined. erefore, each dataset was used to independently estimate the divergence times of the A. mexicanus cave-adapted haplotype sequences. Each da taset was analyzed using both strict and relaxed clock models (Drummond et al., 2006) tested under constant and skyline models of population growth. As part of BEAST divergence time estimation, either a calibration point (fossil or geologic) or a gene-specic substitution rate is required. Because there are no geologic dates cor responding to A. mexicanus populations invading sub terranean systems, substitution rates were used. For each gene, the range of substitution rates calculated for other freshwater sh were used. For cyt b mean substitution rates ranged from 0.005 to 0.017 substitutions/site/mil lion year (my) (Bermingham et al. 1997; Burridge et al. 2006; Dowling et al., 2002b; Perdices & Doadrio, 2001; Sivasundar et al., 2001; Zardoya & Doadrio, 1999) and for ND2 mean substitution rates ranged from 0.011 to 0.026 substitutions/site/my (Near et al., 2003; Mateos, 2005). ese independent rates were used to calibrate the rate of evolution of our datasets by either xing the rate to the lowest and highest value estimated for each gene or using strong prior distributions on the substitution rates. Two independent MCMC analyses 2x10 7 steps long were performed sampling every 2,000 th generation, with a burn-in of 2x10 6 generations. All the Bayesian MCMC output generated by BEAST was analyzed in Tracer v1.3 (Drummond & Rambaut, 2003). Likelihood-based A H RS method. W e used the likeli hood heuristic rate-smoothing algorithm of ( Y ang, 2004) as implemented in PAML3.14 ( Y ang, 2001). Sequence data were analyzed using the F84+ model. Branches at each locus were classied into four rate groups accord ing to their estimated rates. e oldest known fossil rep resentatives of major lineages within the Ostariophysi are well established in recent literature (see Briggs, 2005 and references therein), and have been used in recent studies estimating molecular-based divergence times of Otocephalan clades (Peng et al., 2006). ese fossil representatives were used as calibration points for the AHRS divergence time analysis (Fig. 1, Tab. 2,). Fos sil calibrations were accommodated as xed ages and mapped to the basal node of the clade of interest. Given that most fossils are dated to an age range, the minimum and maximum ages of each fossil were used for diver gence time estimations under separate analyses. Fossil dates were determined using the 1999 GSA Geologic Time Scale. METHODS H O W LONG DOES EVOLUTION OF THE TROGLOMORPHIC FORM TAKE ? E STIMATING DIVERGENCE TIMES ...

PAGE 4

TIME in KARST 2007 176 T ab. 1: T axonomy, gene data, and Genbank accession numbers for sequences used in Characiformes phylogeny reconstruction. Abbreviations of mitochondrial gene sequences: 16S = 16S rD NA, cytb = cytochrome b; abbreviations for nuclear gene sequences: th = forkhead, RAG2 = recombination activating gene, sina = seven in absentia, trop = r-tropomyosin. 16S cytb fkh RAG2 sina trop Anotophysi (outgroup) Chanidae Chanos chanos NC004693 NC004693 --------Gonorynchidae Gonorynchus greyi NC004702 NC004702 --------Kneriidae Cromeria nilotica NC007881 NC007881 --------Parakneria cameronensis NC007891 NC007891 --------Otophysi (ingroup) CHARACIFORMES Anostomidae Leporinus sp. AY788044 AY791416 AY817370 AY804095 AY790102 AY817252 Chilodontidae Chilodus punctatus AY787997 --AY817325 --AY790056 AY817215 Prochilodontidae Prochilodus nigricans AY788075 AY791437 AY817400 AY804120 AY790133 AY817278 Hemiodontidae Hemiodus gracilis AY788027 AY791405 AY817353 AY804084 AY790086 AY817240 Parodontidae Parodon sp. AY788065 AY791427 AY817390 AY804110 AY790123 AY817269 Serrasalmidae Colossoma macropomum AY788000 AY791386 AY817328 AY804061 AY790059 AY817218 Cynodontidae Hydrolycus pectoralis AY788033 --AY817359 AY804088 AY790091 AY817244 Characidae Acestrorhynchus sp. AY787956 AY791353 AY817288 AY804026 AY790014 AY817181 Aphyocheirodon sp. AY787966 AY791363 AY817298 AY804031 AY790025 --Astyanacinus sp.1 AY787969 AY791365 AY817301 AY804033 AY790028 AY817190 Astyanacinus sp.2 AY787987 --AY817317 AY804051 AY790046 AY817209 Astyanax bimaculatus AY787955 --AY817287 AY804025 AY790013 AY817180 Astyanax mexicanus (Brazil) --AY177206 --------Astyanax mexicanus (haplotype AB) -AY639041 ----Astyanax mexicanus (haplotype AL) -AY639051 ----Astyanax mexicanus (haplotype EA) -AY639075 ----Astyanax mexicanus (haplotype FA) -AY639084 ----Astyanax mexicanus (haplotype GA) -AY639089 ----Astyanax mexicanus (haplotype GB) -AY639090 ----Astyanax scabripinis AY787967 --AY817299 --AY790026 AY817188 Brycon hilarii AY787976 AY791370 AY817307 AY804040 AY790035 AY817198 Bryconamericus diaphanus AY787984 AY791375 AY817314 AY804048 AY790043 AY817206 Bryconops sp. AY787985 AY791376 AY817315 AY804049 AY790044 AY817207 Chalceus erythrurus AY787990 AY791379 AY817320 AY804053 AY790049 AY817211 Chalceus macrolepidotus AY787999 AY791385 AY817327 AY804060 AY790058 AY817217 Cheirodon sp. AY787995 AY791382 AY817324 AY804057 AY790054 --Cheirodontops sp. AY787996 AY791383 --AY804058 AY790055 --Creagrutus sp. AY788001 ----AY804062 AY790060 AY817219 Exodon paradoxus AY788013 AY791397 AY817340 AY804072 AY790072 AY817227 Gephyrocharax sp. AY788014 AY791398 AY817341 AY804073 AY790073 AY817228 Hemibrycon beni AY788020 AY791402 AY817346 AY804079 AY790079 AY817234 Hemigrammus bleheri AY788017 --AY817343 AY804076 AY790076 AY817231 M EGAN L. PORTER, K ATHARINA DITTMAR & M ARCOS P REZLOSADA

PAGE 5

TIME in KARST 2007 177 16S cytb fkh RAG2 sina trop Hemigrammus erythrozonus AY788023 --AY817349 AY804081 AY790082 AY817236 Hemigrammus rodwayi AY788034 --AY817360 AY804089 AY790092 AY817245 Hyphessobrycon eques AY788022 --AY817348 AY804080 AY790081 AY817235 Inpaichthys kerri AY788039 --AY817365 AY804093 AY790097 AY817248 Knodus sp. AY788041 AY791414 AY817367 AY804094 AY790099 AY817249 Moenkhausia sanctaphilomenae AY788054 ----AY804104 AY790112 AY817261 Mimagoniates lateralis AY788051 AY791420 AY817377 AY804101 AY790109 AY817259 Prodontocharax sp. AY788064 AY791426 AY817389 AY804109 AY790122 --Roeboides sp. AY787994 AY791381 AY817323 AY804056 AY790053 AY817214 Salminus maxillosus AY788080 AY791438 AY817405 AY804124 AY790137 AY817282 Triportheus angulatus AY788082 --AY817407 AY804125 AY790139 AY817283 Ctenolucidae Ctenolucius hujeta AY787998 AY791384 AY817326 AY804059 AY790057 AY817216 Lebiasinidae Nannostomus beckfordi AY788059 --AY817384 --AY790117 AY817265 Crenuchidae Characidium fasciatum AY787992 AY791380 AY817322 AY804055 AY790051 AY817213 Erythrinidae Hoplias sp. AY788031 AY791409 AY817357 AY804087 AY790090 AY817242 Alestidae Arnoldichthys spilopterus AY787968 AY791364 AY817300 AY804032 AY790027 AY817189 Brycinus nurse AY787970 AY791366 AY817302 AY804034 AY790029 AY817191 Phenacogrammus aurantiacus AY788066 AY791428 AY817391 AY804111 AY790124 AY817270 Hepsetidae Hepsetus odoe AY788030 AY791408 AY817356 AY804086 AY790089 AY817241 Citharinidae Citharinus citharus AY787989 AY791378 AY817319 --AY790048 --Distichodontidae Distichodus sexfasciatus AY788012 AY791396 AY817339 AY804071 AY790071 AY817226 Neolebias trilineatus AY788063 AY791425 AY817388 AY804108 AY790121 AY817268 CYPRINIFORMES Cobitidae Misgurnus sp. AY788053 --AY817379 AY804103 AY790111 --Cyprinidae Danio rerio AY788011 --AY817338 AY804070 AY790070 AY817225 L abeo sorex AY788043 AY791415 AY817369 --AY790101 AY817251 Gyrinocheilidae Gyrinocheilus sp. AY788015 AY791399 --AY804074 AY790074 AY817229 SILURIFORMES Callichthyidae Corydoras rabauti NC004698 NC004698 --------Loricariidae Ancistrus sp. AY787958 AY791354 AY817290 --AY790016 AY817183 Bagridae Chrysichthys sp. AY787957 AY791355 ----AY790017 AY817193 Heptapteridae Pimelodella sp. AY787953 AY791351 AY817285 --AY790011 AY817178 Ictaluridae Ictalurus punctatus AY788040 AY791413 AY817366 --AY790098 --H O W LONG DOES EVOLUTION OF THE TROGLOMORPHIC FORM TAKE ? E STIMATING DIVERGENCE TIMES ...

PAGE 6

TIME in KARST 2007 178 Fig. 1: Characiform divergence time chronogram estimated using a representative topology chosen from the set of 867 most parsimonious trees. White branches indicate branches where less than 75% of the most parsimonious trees were topologically congruent. e grey box indicates the clade of Astyanax mexicanus sequences. Fossil calibration nodes are numbered and correspond to T ab. 2. e major geologic periods are mapped onto the phylogeny. M EGAN L. PORTER, K ATHARINA DITTMAR & M ARCOS P REZLOSADA

PAGE 7

TIME in KARST 2007 179 Population-level divergence time estimations. Estimates of the mean divergence times were not signicantly dif ferent between strict and relaxed clock and population growth models and calibration methods of the substitu tion rate, but condence intervals under the xed substi tution rate approach were narrower, as expected. Hence only the time estimates under the strict clock model, con stant population size and minimum and maximum mean substitution rates for both genes are provided. Compar ing the cytb and ND2 estimates of divergence times for the A. mexicanus A and B lineages show several features. First, the estimated ranges of divergence for cave hap lotypes within each lineage were similar between genes (cyt b and ND2) and lineages (A and B), placing the di vergence among hypogean populations between 0.1410.885 Ma for lineage A, and 0.084-0.575 Ma for lineage B (Tab. 3). W hen comparing the estimates among genes within a lineage, however, the divergence times of hypo gean and epigean haplotypes are dierent, with cytb esti mates providing generally older estimates. Species-level divergence time estimation. Using the maximum parsimony ratchet, the selected Characidae, Characiform, and Otophysi sequences generated 867 trees of score 11758. e 50% majority rule consensus of these trees was similar to the published research that generated the data (Calcagnotto et al. 2005). Because a fully resolved tree with branch lengths is required for AHRS divergence time estimation and because very few branches in the consensus tree collapsed (e.g. were in conict), a random tree from the set of 867 was used (Fig. 1). e A. mexicanus sequences included in the analysis clustered with other Characidae species, although were not monophyletic with other Astyanax species (A. bi maculatus and A. scabripinnis). e divergence time estimates for the representative A. mexicanus cave sh populations generated using this phylogeny with Oto physi fossil calibrations agreed well with the estimates of hypogean haplotype divergence from cyt b and ND2 us ing substitution rates (Tab. 3). However, the estimates of cave versus surface population divergence times based on fossil calibrations were in better agreement with ND2 than with cytb estimates. is is particularly interesting, as the only gene included in this dataset for A. mexica nus was cyt b RESULTS T ab. 2: T axonomy and ages of fossils used as calibrations for divergence time estimation. Node # refers to Fig. 1. Taxonomy Reference Geologic age (MYA) Node # Otophysi Characiformes Gayet, 1982 Late Cretaceous (65-99) 1 Cypriniformes Catostomidae Cavender, 1986 Paleocene (54.8-65) 4 Siluriformes Gayet & Meunier, 2003 late Campanian-early Maastrichtian (68.2-77.4) 3 Corydoras Cockerell, 1925 Late Palaeocene (61-65) 2 T ab. 3: Comparison of divergence time estimates using substitution rates and molecular clock methods for cytochrome b (cytb) and NADH dehydrogenase 2 (ND2) mitochondrial genes, and for molecular methods incorporating fossil dates as calibrations. Substitution Rates Fossil Calibration Cytb ND2 Min Max (Ma) Min Max (Ma) Min Max (Ma) Lineage A cave 0.261 0.885 0.141 0.331 0.2-0.3 cave vs. surface 0.588 2.00 0.256 0.599 0.4-0.5 Lineage B cave 0.169 0.575 0.084 0.196 0.1-0.1 cave vs. surface 1.524 5.181 0.877 2.055 1.7-2.2 Lineage A vs. Lineage B 1.741 5.922 1.053 2.472 1.7-2.2 H O W LONG DOES EVOLUTION OF THE TROGLOMORPHIC FORM TAKE ? E STIMATING DIVERGENCE TIMES ...

PAGE 8

TIME in KARST 2007 180 D ISCUSSION Previous molecular studies of A. mexicanus phylogeog raphy indicate that at least two independent invasions of surface Astyanax have occurred (Dowling et al., 2002a; Strecker et al., 2003, 2004). Our estimates of divergence time from two dierent methods and three dierent datasets are in general agreement about the divergence times among the cave haplotypes in each lineage (Tab. 3). ese estimates place cave haplotype divergence times in the Pleistocene, when it is suggested that climatic cool ing of surface waters led to the extinction of Astyanax in North America (Strecker et al., 2004). In particular, our data show an interesting pattern for lineage B haplotypes, which are proposed to be the older of the two lineages. e recent divergence times estimated for lineage B hap lotypes (0.084-0.575 Ma) supports the hypothesis that aer the initial colonization event, subterranean routes of colonization were associated with uctuating ground water levels in the Pleistocene (Strecker et al., 2004). e fact that estimated times of within lineage divergence were similar also suggests that the divergence of subter ranean haplotypes in both lineages were inuenced by the same processes. In order to determine the evolutionary age of the subterranean lineage, and therefore estimate the time re quired for evolution of the troglomorphic form, the di vergence of the hypogean haplotypes from epigean popu lations is needed. However, the estimates from our three datasets did not agree, with cyt b molecular clock meth ods estimating older divergence times than either ND2 or fossil calibrated estimates. Some of the discrepancy is due to the fact that dierent sets of surface popula tions were sampled in each study (Dowling et al., 2002a; Strecker et al., 2004). For example, the most closely relat ed surface population in the cyt b study were from Belize (Strecker et al., 2004) while there were no closely related surface populations to lineage B haplotypes in the ND2 study (Dowling et al., 2002a). However, this makes the older cytb estimates even more notable because lineage B haplotypes have no evidence of introgressive hybridiza tion with surface populations. If we consider just lineage B hypogean divergence from surface ancestors as an es timate of subterranean evolution, the estimated time for acquisition of the troglomorphic form is 0.877-2.055 Ma (Q uaternary Tertiary boundary) based on ND2 and fossil calibrations, while it is 1.524-5.181 Ma (Pliocene) based on cyt b Although the estimates of divergence times among the three dierent datasets did not agree, comparison of estimates between the lineages show that lineage A diverged from surface ancestors more recently than lineage B (Tab. 3). is more recent divergence from epigean populations is congruent with previous hypoth eses, that either lineage A populations represent a more recent subterranean invasion, or that they are an older invasion masked by more recent mitochondrial intro gressive hybridization with surface forms (Dowling et al., 2002a). In the few studies that have looked at other markers (allozymes, microsatellites, and RAPDs), it has been suggested that at least Chica and Pachn popula tions are the result of surface introgression (Avise & Se lander, 1972; Espinasa & Borowsky, 2001; Strecker et al., 2003). Furthermore, based on the degree of variability in troglomorphic features of each lineage A population, it has been suggested that dierent populations represent dierent degrees and patterns of surface introgression. In order to more accurately determine both the patterns of introgression in the lineage A populations, as well as the underlying relationships of the cave populations to each other in order to estimate subterranean evolution ary times, studies investigating more types of markers are needed. Previous research of A. mexicanus populations throughout Mexico (including cavesh lineages A and B) estimated haplotype divergences to range from 1.8 4.5 Ma (Strecker e t al., 2004). Our estimates suggest that di vergence times among cave haplotypes and between lin eage A cave and epigean haplotypes are much younger than this; however, hypogean divergences from surface ancestors in lineage B are concordant with these older dates. e evolutionary history of cave adaptation in A. mexicanus is complex. Based on mitochondrial molecu lar clock estimates, our estimates of divergence times are congruent with previous hypotheses by showing lineage B to be a phylogenetically older subterranean lineage, with more recent divergence among subterranean sys tems. However, this study also provides quantitative dates for these events. Lineage A populations are estimated to be younger; however, these dates only represent mito chondrial lineages. Several of the populations in lineage A have been shown to be introgressed with surface forms (Chica, Pachn, and Subterraneo). To our knowledge, the hypothesis of surface introgression has not been investi gated in the remaining lineage A populations (Molino, Pichijumo, and Y erbaniz). Understanding the patterns of introgression in all of the lineage A populations, and estimating the actual subterranean evolutionary time, re quires investigating additional nuclear markers. M EGAN L. PORTER, K ATHARINA DITTMAR & M ARCOS P REZLOSADA

PAGE 9

TIME in KARST 2007 181 C ONCLUSIONS Features including colonization routes (stream capture) and the existence of both epigean and cave-adapted hy pogean populations make A. mexicanus an attractive system for investigating the subterranean evolution ary time necessary for acquisition of the troglomorphic form. If it is possible to estimate the divergence time of closely related cave versus surface populations, we can estimate the age of subterranean occupancy. is same divergence time also has relevancy to geologic proc esses in the karst system by providing a rough estimate of the age of subterranean stream capture in particular regions. Based on published sequence data, we have esti mated divergence times for three events in the evolution ary history of troglomorphic A. mexicanus populations. First, divergence times among cave haplotypes in both lineages occurred in the Pleistocene, possibly correlating with uctuating water levels allowing the colonization, and subsequent isolation of, new subterranean habitats. Second, in lineage A, A. mexicanus cave populations ex perienced introgressive hybridization events with surface populations recently. Finally, using divergence times of lineage B from surface populations as an estimate, the acquisition of the troglomorphic form in A. mexicanus is younger than 2.2 (fossil calibration) 5.2 (cytb) Ma (Pliocene). Given that there are at least 30 caves known to contain populations of A. mexicanus (Espinasa et al., 2001; Mitchell et al., 1977), the number of independent invasions and instances of introgressive hybridization may be even higher than currently understood. In order to fully understand the number of independent invasions, the history of introgression with surface populations, and the divergence times of cave and surface populations, a broader survey of cave sh populations and of both nu clear and mitochondrial markers is needed. L ITERATURE C ITED Aden, E., 2005: Adaptation to darkness. In Culver, D.C., & W hite, W .B., (eds.), Encyclopedia of Caves, Else vier Academic Press, pp.1-3. Avise, J.C., & R.K. Selander., 1972: Genetics of cavedwelling shes of the genus Astyanax Evolution, 26, 1-19. Bermingham, E., McCaerty, S.S., & A.P. Martin., 1999: Fish biogeography and molecular clocks: perspec tives from the Panamanian isthmus. In Kocher, T.D., & Stepien, C.A., (eds.), Molecular systematics of shes. Academic Press, San Diego, CA. pp.113-128. Briggs, J.C., 2005: e biogeography of otophysan shes (Ostariophysi: Otophysi): a new appraisal. -Journal of Biogeography, 32, 287-294. Burridge, C.P., Craw, D., & J.M. W aters., 2006: River cap ture, range expansion, and cladogenesis: e genetic signature of freshwater vicariance. Evolution, 60, 1038-1049. Calcagnotto, D., Schaefer, S.A., & R. DeSalle., 2005: Re lationships among characiform shes inferred from analysis of nuclear and mitochondrial gene se quences. -Molecular and Phylogenetics and Evolu tion, 36, 135-153. Cavender, T.M., 1986: Review of the fossil history of North American freshwater shes. In Hocutt, C.H., & W iley, E.O. (eds.), e zoogeography of North American freshwater shes. John W iley, New Y ork, pp.699-724. Cockerell, T.D., 1925: A fossil sh of the family Callich thyidae. Science, 62, 317-322. Culver, D.C., Kane, T.C., & D.W Fong., 1995: Adaptation and natural selection in caves. Harvard University Press, Cambridge, 223p. Dowling, T.E., Martasian, D.P., & W .R. Jeery., 2002a: Evidence for multiple genetic forms with similar eyeless phenotypes in the blind cavesh, Astyanax mexicanus. -Molecular Biology and Evolution, 19, 446-455. Dowling, T.E., Tibbets, C.A., Minckley, W .L., & G.R. Smith, 2002b: Evolutionary relationships of the plagopterins (Teleostei: Cyprinidae) from cyto chrome b sequences. Copeia, 2002, 665-678. Drummond, A.J., & Rambaut, A., 2003: BEAST version 1.4 [computer program]. Available: http://evolve. zoo.ox.ac.uk/beast. Accessed 25 November 2006. Drummond, A.J., Ho, S.Y .W ., Phillips, M.J., & A. Ram baut., 2006: Relaxed phylogenetics and dating with condence. -PLoS Biology 4, e88. Espinasa, L., & R.B. Borowsky., 2001: Origins and rela tionship of cave populations of the blind Mexican tetra, Astyanax fasciatus, in the Sierra de El Abra. Environmental Biology of Fishes, 62, 233-237. Espinasa, L., Rivas-Manzano, P., & H. Espinosa-Prez., 2001: A new blind cave sh population of the genus Astyanax : geography, morphology and behavior. Environmental Biology of Fishes, 62, 339-344. H O W LONG DOES EVOLUTION OF THE TROGLOMORPHIC FORM TAKE ? E STIMATING DIVERGENCE TIMES ...

PAGE 10

TIME in KARST 2007 182 Gayet, M., 1982: Consideration sur la phylogenie et la paleobiographie des Ostariophysaries. -Geobios Memoir, 6, 39-52. Gayet, M., & F.J. Meunier., 2003: Palaeontology and pal aeobiogeography of catshes. In Arratia, G., Kapoor, B.G., Chardon, M., & Diogo, R., (eds.), Catshes, Vol. 2, pp.491-522. Science Publishes, Eneld, NH. Hubbs, C.L., & W .T. Innes., 1936: e rst known blind sh of the family Characidae: A new genus from Mexico. Occasional Papers of the Museum of Zo ology University of Michigan, 342, 1-7. Jeery, W .R., 2001: Cavesh as a model system in evo lutionary developmental biology. -Developmental Biology, 231, 1-12. Jeery, W .R., 2005: Adaptive evolution of eye degenera tion in the Mexican blind cavesh. -Journal of He redity, 96, 185-196. Katoh, K., Kuma, K., Toh, H., & T.Miyata., 2005: MAFFT version 5: improvement in accuracy of multiple sequence alignment. -Nucleic Acids Research, 33, 511-518. Mateos, M., 2005: Comparative phylogeography of live bearing shes in the genera Poeciliopsis and Poecilia (Poeciliidae: Cyprinodontiformes) in central Mexi co. -Journal of Biogeography, 32, 775-780. Mitchell, R.W ., Russell, W .H., & W .R. Elliott., 1977: Mexi can eyeless characin shes, Genus Astyanax : envi ronment, distribution, and evolution. Texas Tech University Special Publications of the Museum, 12, 89pp. Near, T.J., Kassler, T.W ., Koppelman, J.B., Dillman, C.B., & D.P. Philipp., 2003: Speciation in North American black basses, M icropterus (Actinopterygii: Centrar chidae). Evolution, 57, 1610-1621. Nixon, K.C., 1999: e Parsimony Ratchet, a new meth od for rapid parsimony analysis. Cladistics, 15, 407-414. Peng, Z., He, S., W ang, J., W ang, W ., & R. Diogo., 2006: Mitochondrial molecular clocks and the origin of the major Otocephalan clades (Pisces: Teleostei): A new insight. Gene, 370, 113-124. Perdices, A., & I. Doadrio., 2001: e molecular system atics and biogeography of the European cobitids based on mitochondrial DNA sequences. -Molecu lar Phylogenetics and Evolution, 19, 468-478. Sikes, D.S., & P.O. Lewis., 2001: Beta soware, version 1. PAUPRat: PAUP* implementation of the parsimo ny ratchet. [computer program]. Available: http:// www.ucalgary.ca/~dsikes/soware2.htm. Accessed 25 November 2006. Sivasundar, A., Bermingham, E., & G. Ort., 2001: Pop ulation structure and biogeography of migratory freshwater shes (Prochilodus: Characiformes) in major South American rivers. -Molecular Ecology, 10, 407-417. Strecker, U., Bernatchez, L., & H. W ilkens., 2003: Genetic divergence between cave and surface populations of Astyanax in Mexico (Characidae, Teleostei). -Mo lecular Ecology, 12, 699-710. Strecker, U., Fandez, V.H., & H. W ilkens., 2004: Phylo geography of surface and cave Astyanax (Teleostei) from Central and North America based on cyto chrome b sequence data. -Molecular Phylogenetics and Evolution, 33, 469-481. Swoord, D.L., 2000: PAUP*. Phylogenetic analysis using parsimony (*and other methods). Version 4. Sinau er Associates, Sunderland, MA. W iley, S. & R.W Mitchell., 1971: A bibliography of the Mexican eyeless characin shes of the genus Asty anax -Association for Mexican Studies Bulletin, 4, 231-239. Y ang, Z., 2001: PAML: Phylogenetic Analysis by Maxi mum Likelihood. University College London, Lon don. Y ang, Z., 2004: A heuristic rate smoothing procedure for maximum likelihood estimation of species diver gence times. -Acta Zoologica Sinica, 50, 645-656. Zardoya, R., & I. Doadrio., 1999: Molecular evidence on the evolutionary and biogeographical patterns of European cyprinids. -Journal of Molecular Evolu tion, 49, 227-237. M EGAN L. PORTER, K ATHARINA DITTMAR & M ARCOS P REZLOSADA



PAGE 1

A GE ESTIMATES FOR SOME SUBTERRANEAN TA X A AND LINEAGES IN THE D INARIC K ARST O CENE STAROSTI ZA NEKATERE PODZEMELJSKE TAKSONE IN bIVALSKE LINIJE NA D INARSKEM KRASU Peter T RONTELJ 1 pela G ORIfKI 1 Slavko P OLAK 2 Rudi V EROVNIK 1 Valerija Z AKEK 1 & Boris S KET 1 Izvleek UDK 575.8:551.442(234.422.1) 591.542(234.422.1) Peter Trontelj, pela Goriki, Slavko Polak, Rudi Verovnik, Valerija Zakek & Boris Sket: Ocene starosti za nekatere podzemeljske taksone in ivalske linije na Dinarskem krasu Z uporabo primerjalnega logeografskega pristopa in neod visnih molekularnih ur smo predlagali asovni potek evolucije troglobiontov Dinarskega krasa, ki velja za sorazmerno veliko tevilo taksonov. Zdi se, da kljuni dogodki pripadajo dvema obdobjema. (1) Glavne razdelitve znotraj holodinarskih tak sonov so iz obdobje srednjega miocena. Predstavljajo zgornji potencialni asovni limit za naselitev jam. (2) Regionalna dife renciacija, vkljuno s speciacijo, ki je lahko vsaj deloma pove zana s podzemeljsko fazo, naj bi se zgodila med zgodnjim in srednjim pleistocenom. Ocenjujemo, da se je zaela invazija veine prouevanih ivalskih linij v podzemlje Dinarskega kra sa v obdobju med dvema in petimi milijoni let. Kljune besede: podzemlje, molekularna ura, molekularna logenija, logeograja, Dinarski kras. 1 Oddelek za biologijo, Biotehnika fakulteta, Univerza v Ljubljani, Vena pot 111, 1000 Ljubljana, Slovenia, fax: +386 1 2573390, e-mail: peter.trontelj@bf.uni-lj.si 2 Notranjski muzej Postojna, Ljubljanska 10, 6230 Postojna, Slovenia. Received/Prejeto: 30.01.2007 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 183 Abstract UDC 575.8:551.442(234.422.1) 591.542(234.422.1) Peter Trontelj, pela Goriki, Slavko Polak, Rudi Verovnik, Valerija Zakek & Boris Sket: Age estimates for some subter ranean taxa and lineages in the Dinaric Karst Using a comparative phylogeographic approach and dierent independent molecular clocks we propose a timescale for the evolution of troglobionts in the Dinaric Karst that is relatively consistent over a wide taxonomic range. Keystone events seem to belong to two age classes. (1) Major splits within holodinaric taxa are from the mid-Miocene. ey present the potential up per limit for the age of cave invasions. (2) Regional dierentia tion, including speciation, which can at least in part be associ ated with a subterranean phase, took place from early Pliocene to mid-Pleistocene. W e suggest two to ve million years as the time when most of the analyzed lineages started invading the Dinaric Karst underground. Key words: subterranean, molecular clock, molecular phylog eny, phylogeography, Dinaric Karst. I NTRODUCTION e use of new molecular and systematic techniques us ing allozymes and DNA sequences has enabled us to see a new picture of the evolution and diversity of subter ranean fauna (e.g. Avise & Selander 1972; Sbordoni et al., 2000; Caccone & Sbordoni 2001; Leys et al., 2003; Verovnik et al., 2004; Goriki & Trontelj 2006; Lefbure et al., 2006; Zakek et al., 2007). Molecular clock ap proaches should, at least in theory, enable us to date, to verify or to falsify previous hypotheses about the age of subterranean species. To be exact, it is usually not the age of a lineage or a taxon itself that is of special inter est or under dispute, but the time since it has attained its subterranean nature, making it even more challeng ing. Hypotheses and models explaining cave invasions

PAGE 2

TIME in KARST 2007 184 and speciation in caves are well-elaborated (e.g. Rouch & Danielopol 1987; Holsinger 2000; Trajano 2005) and should thus oer good grounds for the timing of such events and for testing their correlation with geographical, geological and hydrographical counterparts. For exam ple, Leys et al., (2003) have shown that all evolutionary transitions to subterranean life in Australian dytiscids took place during the Late Miocene and Early Pliocene as a result of aridication. However, reliable data on the age of these events is surprisingly scarce. W hen such data are available, the accuracy is oen below that of molecular clock rates. In fact, the use of molecular dating methods itself has introduced considerable uncertainty about how old subterranean species might be. W hile the youngest estimation, based on classical biological reasoning, is no more than 10,000 years (Sket 1997), the upper limit for the divergence of two subterranean sister species has been pushed to an incredible 110,000,000 years (Buhay & Crandall 2005). Boutin and Coineau (2000) have argued that dating of cladogenetic events by a molecular clock is particular ly useful in the case when the dates are corroborated by other methods. Since the obvious problem of the Dinaric Karst area is that reliable dating for clearly dened vicari ant events or the age of available subterranean habitat is lacking, it has been impossible to corroborate molecu lar clock divergence by independent data. In this case a comparative phylogeographic approach might provide the means for an independent validation of age esti mates. Comparative phylogeography seeks, as does his torical biogeography, concordant geographical patterns of codistributed lineages (e.g. Arbogast & Kenagy 2001). e evolution of codistributed phylogeographic groups of dierent taxa is likely to have been driven by the same historical factors, like vicariant events or climatic shis. In this contribution we (1) identify common phylo geographic patterns among those troglobiotic taxa from the Dinaric Karst for which such data are available, and (2) estimate the timeframe of the corresponding cladoge netic events using a global molecular clock approach. MATERIAL AND METHODS e presented data were taken from several phylogeo graphic studies of subterranean animals in the Dinaric Karst, including the ubiquitous aquatic isopod Asellus aquaticus Linne (Verovnik et al., 2004, 2005), the cave salamander P roteus anguinus Laurenti (Goriki 2006, Goriki & Trontelj 2006), and the cave shrimp T roglo caris s. lato (Zakek et al., 2007). Further, we included unpublished sequences from studies that are in progress, including leptodirine cave beetles and aquatic sphaero matid isopods from the genus M onolistra. e age esti mations for the last two groups should be regarded as preliminary because in-depth analyses of phylogenetic relationships and corroboration by further loci are still under way. W e were only interested in a small number of well-supported splits and therefore used straightfor ward minimum evolution searches with bootstrapping as implemented in MEGA (Kumar et al., 2004). Divergence time estimates are based on available clock-rate data for groups that are as closely related as possible (Caccone & Sbordoni [2001] for leptodirines, Ketmaier et al., [2003] for Asellus aquaticus, and Sturmbauer et al., [1996] and Schubart et al., [1998] for M onolistra). To assure compat ibility between molecular divergences we used the same models as were used in the original works describing the rates (Tamura-Nei distances with a gamma distributed rate variation among sites). W here more than one hap lotype per population or lineage was analyzed we used net between group distances to correct for ancestral in traspecic diversity. RESULTS e split between major geographically dened lineages e geographical distribution of troglobiotic (in cluding stygobiotic) sister taxa can be used to infer inde pendent cave invasions. For example, if the present-day ranges of two troglobionts are separated by large areas of non-karstic terrain without hypogean habitat, we can postulate an epigean last common ancestor. Examples of that kind can be found in the shrimp genus T roglo caris, with the H ercegovinensis lineage inhabiting Trans caucasian and SE parts of the Dinaric Karst where it is sympatric with the SE populations of the Anophthalmus lineage (Zakek et al., 2007). eir split estimated at 6 P ETER T RONTELJ PELA G ORIfKI S LAVKO P OLAK R UDI V EROVNIK V ALERIJA Z AKEK & B ORIS S KET

PAGE 3

TIME in KARST 2007 185 Myr ago is the oldest, although unlikely, possible time of cave invasion. e youngest split that could be reliably inferred from the phylogenetic tree and probably still oc curred in surface waters, was the one between the Bos nian lineage and other Anophthalmus lineages. Because the karst area in Bosanska Krajina, to which the Bosnian clade is restricted, is so remote and isolated from the rest of the Dinaric populations, it is reasonable to assume that an underground connection between them could never have existed. e estimated time of this split, 3.7.3Myr ago, is hence the oldest possible age at which T roglocaris anophthalmus might have invaded the Dinaric Karst un derground (Tab. 1). For the cave salamander Proteus anguinus, exhibit ing a distribution pattern similar to that of T roglocaris, the corresponding age of the Bosanska Krajina lineage was estimated at 4.4.4 Myr (Goriki 2006). However, older lineages exist that, theorethically, might have in vaded caves even as early as 8.8 Myr ago (see also Fig. 1). Another troglobiotic group restricted to the Dinaric Karst area and having a non-troglomorphic sister group is the Dinaric clade of Asellus aquaticus (see Verovnik et al., 2005). e time of this split, and hence the maxi mum possible age of cave invasion is 3.8.8 Myr. T IMING OF MORPHOLOGICAL CHANGES W here possible, we tried to combine the biology (e.g. degree of troglomorphism, lack of gene ow) of taxa with corresponding data on paleogeography and paleo hydrography to infer speculative scenarios on how and when lineages might have switched to subterranean life and evolved troglomorphic traits. For example, we have some indication about how long at most it takes a sala mander population to become troglomorphic. Since the subspecies P. a. parkelj Sket et Arntzen has retained its ancestral, non-troglomorphic characteristics, it is reason able to conclude that its sister lineage must have evolved troglomorphoses independently from other, less related troglomorphic lineages (Sket & Arntzen 1994; Goriki & Trontelj 2006; see Fig. 1). e split between the nontroglomorphic lineage and its last troglomorphic sister lineage was estimated at 0.5.6 Myr based on mitochon drial rDNA sequences, 1.1.4 Myr based on the mtDNA control region (Goriki 2006), and at 1.1.5 Myr by an allozyme clock (Sket & Arntzen 1994). Asellus aquaticus has evolved several separate sub terranean and troglomorphic populations. One of them, from the subterranean Reka River below the Kras/Carso Plateau, is genetically completely isolated from epigean populations at the Reka resurgence while there are no epi gean populations in the Reka before the sink (Verovnik e t al., 2003, 2004, 2005; Fig. 2). Further, it has no mtDNA T ab. 1. Estimated time (in million years) of some keystone events in the evolution of troglobionts in the D inaric Karst. Taxon Age of holodinaric group Age of merodinaric group Mid-Dinaric split Northwest split Troglocaris (Dinaric and Caucasian lineages) 1 7.9.1 n.a. n.a. n.a. Troglocaris anophthalmus agg. 1 n.a. 3.7.3 1.3.3 1.5.1 Troglocaris hercegovinensis agg. 1 n.a. 3.8.8 n.a. n.a. Proteus anguinus 2 8.8.0 n.a. 8.8.0 4.2.2 Asellus aquaticus (Dinaric clade) 3 n.a. 3.8.8 n.a. 0.8.2 Microlistra 4 n.a. 1.1.3 n.a. n.a. Pseudomonolistra hercegoviniensis 4 n.a. 0.3.0 n.a. n.a. Monolistra caeca 4 n.a. 1.8.7 n.a. n.a. Leptodirus hochenwartii hochenwartii et L. h. reticulatus 5 n.a. 1.9.0 n.a. n.a. 1 Using COI clock for shrimps (see Knowlton & Weigt 1998; Z akek et al., 2007) 2 Using 12S and 16S rD NA clock for Newts (see Cacconesee et al., 1997; Goriki 2006) 3 Using COI clock for subterranean Asellota (see Ketmaier et al., 2003; V erovnik et al 2005) 4 Using 16S r-RNA clock for ddler crabs (Sturmbauer et al 1996) and land crabs (Schubart et all 1998) 5 Using COI clock for subterranean leptodirine beetles (Caccone & Sbordoni 2001) A GE ESTIMATES FOR SOME SUBTERRANEAN TA X A AND LINEAGES IN THE D INARIC K ARST

PAGE 4

TIME in KARST 2007 186 and nuclear rDNA haplotypes in common with hypoge an populations from the Ljubljanica River drainage with which the Reka drainage has been connected many times during the Pleistocene and occasionally even nowadays (Habi 1989). It is thus reasonable to assume that the an cestor of the subterranean Reka River population invad ed hypogean waters and became cave-adapted before any secondary contact could occur. e estimated age of the Reka River lineage is 3.1.1 Myr (Verovnik et al., 2004), making it a pre-Pleistocene troglobiotic relict (Verovnik et al., 2004). M onolistra, a troglobiotic group of freshawater sphaeromatid isopods, shows a high taxonomic and mor phological diversity restricted to the Dinaric Karst and parts of the Southern Calcareous Alps. According to our preliminary results of a molecular phylogenetic analysis based on nuclear and mitochondrial DNA sequences, there are at least three well-supported monophyla. ese are the subgenus M. (M icrolistra), M. (M onolistra) caeca Gerstaecker, and the polytypic M. (Pseudomonolistra) hercegoviniensis Absolon. Several lines of evidence sug gest that the common ancestors of each of these groups invaded cave waters polytopically (Sket 1986, 1994). W hile we remain ignorant about when and how oen an cestral M onolistra lineages invaded subterranean waters, we can expect that the radiation of at least some of the three groups took place in the underground. eir ages (maximally 0.4.7 Myr) give us an idea for how long some M onolistra lineages have been dwelling in the Di naric Karst underground. Leptodirus hochenwartii Schmidt, a highly troglo morphic leptodirine cave beetle, is the only terrestrial Dinaric troglobiont with available molecular dating. Us ing a leptodirine COI clock calibration by Caccone and Sbordoni (2001) we estimated the age of the Leptodirus lineage by dating the split with Astagobius angustatus Schmidt, its slightly less troglomorphic sister lineage. e estimated time of this split (8.7.8 Myr ago) is the oldest possible age at which the extremely specialized morphology of Leptodirus could have started evolving. Moreover, taking into account recent unpublished phy logenetic ndings based on nuclear and mitochondrial gene sequences, the traditional subspecies of Leptodirus in fact represent distinct lineages with divergences well in the range of between species comparisons. ese lineages all share the same constructive apomorphic troglomor phic characters, and it seems probable these troglomor phies have already existed at least at the time of their last common ancestor. e time of divergence between basal Leptodirus lineages hence represents the youngest pos sible age at which Leptodirus has evolved its full array Fig. 1: A simplied view of the phylogenetic relationships obetween troglomorphic and non-troglomorphic Proteus anguinus populations (from Goriki and T rontelj 2006). Postulating a non-troglomorphic ancestor and unidirectional evolution toward troglomorphism, we can take the split between the black subspecies (non-troglomorphic) and its unpigmented sister lineage to estimate the maximal time ( T1) needed for a salamander lineage to evolve the entire array of cave-related traits known in this taxon. If one accepts the notion of multiple independent cave invasions for Proteus, than T2 is the potentially oldest time since it has become subterranean. Fig. 2: e case of troglomorphic and non-troglomorphic lineages of Asellus aquaticus in the D inaric Karst, highly simplied (from V erovnik et al 2004, 2005). e Reka and the Ljubljanica (Ljub) basin lineages have independently invaded subterranean waters and thus constitute separate taxa, although traditionally assigned to the same subspecies, A. a. cavernicolus. e subterranean Reka River population presents the oldest stygobiotic lineage of Asellus aquaticus. B ecause it is genetically completely distinct, it must have escaped interbreeding during various times of hydrological contact with surface populations. We therefore believe that it became a specialized stygobiont soon aer the split at time T1. e Ljub lineage from the subterranean Ljubljanica River, although morphologically distinct, is still sharing mtD NA haplotypes with surface populations and thus represents a younger invasion. Eur and D in denote various epigean European and D inaric lineages, respectively. P ETER T RONTELJ PELA G ORIfKI S LAVKO P OLAK R UDI V EROVNIK V ALERIJA Z AKEK & B ORIS S KET

PAGE 5

TIME in KARST 2007 187 DISCUSSION Before we reach any conclusions we would like to note that dating of keystone events in the evolution of sub terranean life, as well as anywhere else in evolution (e.g. Graur & Martin 2004), remains a highly speculative en terprise. Of central concern should be the fact that we are relying on a more or less global clock within certain taxonomic boundaries. ese clocks usually rely on sin gle calibration points (e.g. the separation of the SardiniaCorsica microplate from the Iberian Peninsula; Ketmaier et al., 2003) and have mostly not been tested against in dependent geological events. Further, all our timings assume linear accumulation of substitutions over time, i.e. the existence of a valid mo lecular clock. Although we can be quite sure that this as sumption is violated to a certain extent, we can mitigate the problem by excluding those taxa from the analysis that violate the linearity assumption most. More sophisticated and realistically modeled approaches use a relaxed clock allowing for dierent local rates on dierent branches of the tree (e.g. Sanderson 2002). However, with single cali bration points only, such approaches yield quite hopeless and certainly unrealistic intervals. For example, the age of the deepest split in the Niphargus virei (subterranean amphipod from France) complex was estimated at 14 Myr using a global Stenasellus clock, whereas the relaxed clock estimate was 22 Myr (Lefbure et al., 2006). ird, it should be noted that even with the aid of molecular phylogenetic tools the timing is still suscep tible to incorrect estimations of relationships and incom plete taxonomic coverage. For example, the timing of the origin of the highly troglomorphic morphologies in Lep todirus depends on the most basal split in the taxon. By not having included all known subspecies, we are facing the risk that some other subspecies might have branched o earlier than the studied ones. One potentially useful way to improve our informal condence in the timing of evolutionary events in subter ranean animals is to look for phylogeographic correspon dence of timings derived from independent taxa with in dependent molecular clocks. At the present stage of most of our analyses such comparisons can only be preliminary. W e can nevertheless notice that specic groups of events belong to dierent age classes, most markedly the gap be tween the age of holodinaric troglobionts and those with narrower distributions within the Dinaric Karst (Tab. 1). e recent lineages of Proteus and T roglocaris probably both originate from the Miocene Dinaride Lake System (Krstit et al., 2003), and the age of both taxa reects their dierentiation long before they invaded the hypogean en vironment (Sket 1997; Goriki 2006; Zakek et al., 2007). Regional dierentiation, including speciation, which can at least in part be associated with a subterranean phase, appears to be much younger, ranging from Pliocene to mid-Pleistocene. Based on these estimates plus the esti mated age of the Reka River lineage of Asellus aquaticus (see above) we, tentatively, suggest two to ve million years as the time when most of the analyzed lineages started in vading the Dinaric Karst underground. e mid-Dinaric split of Proteus and T roglocaris anopthalmus does not seem to originate from the same of troglomorphic characters. Based on a yet incomplete taxonomic sample (L. h. hochenwartii Schmidt and L. h. reticulatus J. Mller) we tentatively dated it at 1.9.0 Mya. T IMING OF PALEOH Y DROGRAPHIC CHANGES For some stygobiotic taxa with a broader Dinaric range, we identied two concordant geographic patterns pos sibly pointing to common underlying historical events, like changes in hydrographic connections. ese vicari ant patterns include (1) a split between a northwestern and southeastern Dinaric clade (mid-Dinaric split), and (2) a younger subdivision of the northwestern clade (or of a part thereof) into a western and eastern Slovenian lineage (Tab. 1). It has been stated that some stygobiotic species in habit areas that are hydrographically fragmented. e most parsimonious explanation of such distributions is that their ranges were hydrographically interconnected in the past. is may as well include the surface paleohy drography that was heavily fragmented by karstication. e (polytopic) immigration underground could thus have proceeded simultaneously with the separation of ancestral populations. W e can illustrate this scenario by the case of some M onolistra lineages, namely of the sub genus M icrolistra and of the species M. (M.) caeca. Some ten M icrolistra spp. are perfectly allopatric in distribu tion, mainly bound to actual watersheds. Another group, M. caeca, inhabits at least three watersheds, in which four named subspecies have evolved. According to a 16S rDNA molecular clock (Sturmbauer et al., 1996, Schu bart et al., 1998), the system began to fragment about three million years ago. A GE ESTIMATES FOR SOME SUBTERRANEAN TA X A AND LINEAGES IN THE D INARIC K ARST

PAGE 6

TIME in KARST 2007 188 vicariant event as the latter was estimated to be younger by an order of magnitude. Another commonality of the phylogeographic pattern, the division between a west ern and an eastern clade in the Slovenian Dinaric Karst, might have a common hydrogeological cause in two sty gobiotic crustaceans (A. aquaticus and T anophthalmus) somewhere in the middle of the Pleistocene. In Proteus, however, the same split appears to be substantially older. In the Dinaric Karst we were, so far, unable to nd reliable time estimates for paleogeographic events to cali brate local molecular clocks in dierent lineages. Con versely, the timing of phylogenetic events can serve, inas much as we rely on global molecular clocks, to estimate the date of geographical, hydrographical, and geological changes (Sket 2002). e comparative phylogeographic approach and the use of dierent independent molecular clocks have enabled us for the rst time to propose a tim escale for the evolution of troglobionts that is relatively consistent over a wide taxonomic range. is timescale is a preliminary one, though. W e expect it to change with the inclusion of further taxa, the study of more genes and the use of more accurate molecular dating approaches. ACKNO W LEDGMENTS W e are indebted to many friends and colleagues who helped us with the acquisition of biological samples and accompanied us during eld work. W e thank Gregor Brako and Joica Murko-Buli for their indispensable assistance in the lab. e presented work is the result of several research projects funded by the Slovenian Re search Agency, and of the contract n EVK2-CT-200100121PASCALIS of the Fih Research and Technologi cal Development Framework Program supported by the European Community REFERENCES Arbogast, B.S. & Kenagy, G.J., 2001: Comparative phy logeography as an integrative approach to histori cal biogeography. Journal of Biogeography, 28, 819. Avise, J.C. & Selander, R.K., 1972: Evolutionary genetics of cave dwelling shes of the genus Astyanax. Evo lution, 26, 1. Boutin, C. & Coineau, N., 2000: Evolutionary rates and phylogenetic age of some stygobiontic species. In: Subterranean ecosystems Ecosystems of the W orld 30 (Eds. H. W ilkens, D.C. Culver & W .F. Hum phreys). Elsevier, Amsterdam etc., pp. 433. Buhay, J.E. & Crandall, K.A., 2005: Subterranean phylo geography of freshwater crayshes shows extensive gene ow and surprisingly large population sizes. Molecular Ecology, 14, 4259. Caccone, A., Milinkovitch, M.C., Sbordoni, V. & Powell, J.R., 1997: Mitochondrial DNA rates and biogeogra phy in European newts (genus Euproctus) . System atic Biology, 46, 126. Caccone, A. & Sbordoni, V., 2001: Molecular biogeogra phy of cave life: a study using mitochondrial DNA from bathysciine beetles. Evolution, 55, 122. Goriki, & Trontelj, P., 2006: Structure and evolution of the mitochondrial control region and anking sequences in the European cave salamander Proteus anguinus. Gene, 387, 31. Goriki, ., 2006: Phylogeographic and morphological analysis of European cave salamander (Proteus an guinus) populations. Doctoral Dissertation. Uni versity of Ljubljana, Biotechnical Faculty, Dept. of Biology, Ljubljana. Graur, D. & Martin, W ., 2004: Reading the entrails of chickens: molecular timescales of evolution and the illusion of precision. Trends in Genetics, 20, 80. Habi, P., 1989: Kraka bifurkacija Pivke na jadransko rnomorskem razvodju (Pivka carst bifurcation on Adriatic Black Sea watershed). Acta Carsologica, 18, 233. Holsinger, J.R., 2000: Ecological derivation, colonization, and speciation. In: Subterranean ecosystems Eco systems of the W orld 30 (Eds. H. W ilkens, D.C. Cul ver & W .F. Humphreys). Elsevier, Amsterdam etc., pp. 399. P ETER T RONTELJ PELA G ORIfKI S LAVKO P OLAK R UDI V EROVNIK V ALERIJA Z AKEK & B ORIS S KET

PAGE 7

TIME in KARST 2007 189 Ketmaier, V., Argano, R. & Caccone., A., 2003: Phylo geography and molecular rates of subterranean aquatic Stenasellid Isopods with a peri-yrrenian distribution. Molecular Ecology, 12, 547. Knowlton, N. & W eigt, L.A., 1998: New dates and new rates for divergence across the Isthmus of Panama. Proceedings of the Royal Society of London B, 265, 2257. Krstit, N., Savit, L., Jovanovit G. & Bodor, E., 2003: Lower Miocene lakes of the Balkan Land. Acta Geo logica Hungarica, 46/3, 291. Kumar, S., Tamura, K & Nei M., 2004: MEGA3: Inte grated soware for Molecular Evolutionary Genet ics Analysis and sequence alignment. Briengs in Bioinformatics, 5, 150. Lefbure, T., Douady, C.J., Gouy, M., Trontelj, P., Brio lay, J. & Gibert, J., 2006: Phylogeography of a sub terranean amphipod reveals cryptic diversity and dynamic evolution in extreme environments. Mo lecular Ecology, 15, 1797. Leys, R., W atts, C.H.S., Cooper, S.J.B. & Humphreys, W .F., 2003: Evolution of subterranean diving beetles (Co leoptera: Dytiscidae: Hydroporini, Bidessini) in the arid zone of Australia. Evolution, 57, 2819. Rouch, R. & Danielopol, D. L., 1987: Lorigine de la faune aquatique souterraine, entre le paradigme du refuge et le modele de la colonisation active. Stygologia, 3, 345. Sanderson, M.J., 2002: Estimating absolute rates of mo lecular evolution and divergence times: a penalized likelihood approach. Molecular Biology and Evo lution, 19, 101. Sbordoni, V., Allegrucci, G. & Cesaroni, D., 2000: Popu lation genetic structure, speciation and evolutionary rates in cavedwelling organisms. In: Subterranean ecosystems Ecosystems of the W orld 30 (Eds. H. W ilkens, D.C. Culver & W .F. Humphreys). Else vier, Amsterdam etc., pp. 453. Schubart, C.D., Diesel, R. & Hedges, S.B., 1998: Rapid evolution to terrestrial life in Jamaican crabs. Na ture, 393, 363. Sket, B., 1986: Evaluation of some taxonomically, zoo geographically, or ecologically interesting nds in the hypogean waters of Y ugoslavia (in the last de cades). Comunicaciones, 9. Congreso Internacional de Espeleologia, 1, 126. Sket, B., 1997: Distribution of Proteus (Amphibia: Urode la: Proteidae) and its possible explanation. Journal of Biogeography, 24, 263. Sket, B., 2002: e evolution of the karst versus the dis tribution and diversity of the hypogean fauna. In: Evolution of karst: from Prekarst to cessation (Ed. F. Gabrovek) Ljubljana-Postojna, pp. 225. Sket, B. & Arntzen, J.W ., 1994: A black, non-troglomor phic amphibian from the karst of Slovenia: Proteus anguinus parkelj n. ssp. (Urodela: Proteidae). Con tributions to Zoology, 64, 33. Sturmbauer, C., Levinton, J.S. & Christy, J., 1996: Mo lecular phylogeny analysis of ddler crabs: test of the hypothesis of increasing behavioral complexity in evolution. Proceedings of the National Acad emy of Sciences of the United States of America, 93, 10855. Trajano, E., 2005: Evolution of lineages. In: Encyclopedia of Caves (Eds. D.C. Culver & W .B. W hite). Else vier, Amsterdam etc., pp. 230. Verovnik, R, Sket, B. & Trontelj, P., 2004: Phylogeography of subterranean and surface populations of water lice Asellus aquaticus (Crustacea: Isopoda). Mo lecular Ecology, 13, 1519. Verovnik, R., Sket, B. & Trontelj, P., 2005: e coloniza tion of Europe by the freshwater crustacean Asellus aquaticus (Crustacea: Isopoda) proceeded from an cient refugia and was directed by habitat connectiv ity. Molecular Ecology, 14, 4355. Verovnik, R., Sket, B., Prevornik, S. & Trontelj, P., 2003: Random amplied polymorphic DNA diversity among surface and subterranean populations of Asellus aquaticus (Crustacea: Isopoda). Genetica, 119, 155. Zakek, V., Sket, B. & Trontelj, P., 2007: Phylogeny of the cave shrimp T roglocaris : evidence of a young con nection between Balkans and Caucasus. Molecular Phylogenetics and Evolution, 42, 223. A GE ESTIMATES FOR SOME SUBTERRANEAN TA X A AND LINEAGES IN THE D INARIC K ARST



PAGE 1

THE CHALLENGE OF ESTIMATING THE AGE OF SUBTERRANEAN LINEAGES: E X AMPLES FROM BRAZIL IZZIV OCENJEVANJA STAROSTI PODZEMELJSKIH bIVALSKIH LINIJ: PRIMERI IZ BRAZILIJE Eleonora T RAJANO 1 Izvleek UDK 551.44:597(81) 591.542(81) Eleonora Trajano: Izziv ocenjevanja starosti podzemeljskih ivalskih linij: primeri iz Brazilije V prispevku je opisana uporabnost in uinkovitost razlinih pristopov za ocenjevanje starosti ivalskih linij s pomojo morfologije, molekularne logenije, biogeograje in geologije. Predstavljeni so primeri podzemeljske favne iz Brazilije, pred vsem rib kot najbolj raziskane skupine. Molekularno-bioloke raziskave, ki vkljuujejo tudi troglobionte, opravljamo na zgolj treh taksonih. Molekularne ure zaenkrat e ne moreme upora biti, vendar zgolj posredne dokaze. Na voljo imamo le nekaj logenetskih podatkov, npr. za morske zmaje iz druin Hep tapteridae in Trichomycteridae. Teoretino so bazalni troglo bitski kladi stareji od apikalnih, eprav verjeten obstoj, sicer izumrlega epigejinega taksona, ki pripada takim kladom, ovira primerjavo. Omejitve uporabe troglomorzma za ocenjevanje logenetske starosti smo analizirali s poudarkom na komplek snosti mehanizmov, ki so osnova morfolokemu razloevanju. Razpololjiva paleoklimatska rekonstrukcija, ki temelji na dat iranju kapnikov iz jam severovzhodne in jugovzhodne Brazilije, je omejena na zadnjih 200.000 let in je kot taka uporabna le za relativno recentne linije. Topografska izolacija, ki verjetno velja za nekaj skupin rib iz osrednje Brazilije, spada v asovno ob dobje 10 5 let. Stareji datirani dogodki (obdobje 10 6 let ali ve), ki naj bi predstavljali vikariantske dogodke in ki so pomembni za vodne linije podzemeljskih sorodnikov, so povezane z raz vojem dananjih glavnih junoamerikih poreij. Trenutno je, zaradi malotevilnih podatkov, najbolja metoda za ocenjevan je starosti brazilskih troglobitskih linij kombinacija pristopov, ki vkljuujejo morfologijo, sistematiko in biogeograjo. Kljune besede: evolucija troglobiontov, speleobiologija, stopnja troglomorzma, Brazilija, podzemeljske ribe, razloevalno razmerje. 1 Departamento de Zoologia, Instituto de Biocincias da Universidade de So Paulo, So Paulo, BRASIL; e-mail: etrajano@usp.br Received/Prejeto: 06.12.2006 COBISS: 1.01 TIME in KARST, POSTOJNA 2007, 191 Abstract UDC 551.44:597(81) 591.542(81) Eleonora Trajano: e challenge of estimating the age of sub terranean lineages: examples from Brazil e applicability and eectiveness of dierent kinds of evidence used to estimate the age of lineages morphological, molecu lar, phylogenetic, biogeographical, geological are discussed. Examples from the Brazilian subterranean fauna are presented, using mainly shes, one of the best studied groups, as a model. Only three taxa including troglobites are object of molecular studies, all in progress. erefore, molecular clocks cannot be applied yet, and indirect evidence is used. Few phylogenies are available, e.g. for the catsh families Heptapteridae and Tricho mycteridae. eoretically, basal troglobitic clades are older than apical ones, but the possible existence of extinct epigean taxa belonging to such clades hampers the comparison. As well, the limitations of the use degrees of troglomorphism to estimate phylogenetic ages are analyzed with focus on the complexity of the mechanisms underlying morphological dierentiation. Paleoclimatic reconstructions based on dating of speleothems from caves in northeastern and southeastern Brazil are avail able, but limited up to the last 200,000 years, thus useful for relatively recent lineages. Topographic isolation, probable for some sh groups from Central Brazil, is also within the time range of 10 5 years. Older dated events (in the order of 10 6 years or more) that may represent vicariant events aecting aquatic lineages with subterranean derivatives are related to the estab lishment of the modern South American main river basins. In view of the paucity of data useful for estimating the age of Bra zilian troglobitic lineages, combined evidence, including mor phology, systematics and biogeography, seems to be the best approach at the moment. Key words: evolution of troglobites, degree of troglomorphism, Brazil, subterranean shes, dierentiation rates.

PAGE 2

TIME in KARST 2007 192 D EGREE OF TROGLOMORPHISM AND PH Y LOGENETIC AGE : e use of the degree of troglomorphism to infer relative phylogenetic ages is based on the assumption that the rates of morphological dierentiation are fairly constant among subterranean taxa, at least those regarding eyes and pigmentation, which tend to be lost along the isola tion in subterranean habitats. To accept this notion, it is necessary to assume that the mechanisms of reduction are the same for each of these characters and that their reduction progress in parallel. However, there is strong evidence in contrary. e occurrence of dierent mosaics of character states in closely related taxa suggests dierent mecha I NTRODUCTION e problem of estimating ages for subterranean or any other lineages starts with the very denition of age, whether the time since the isolation from the immediate sister-group (age of the cladogenetic event) or the begin ning of dierentiation, either the genetic or the morpho logical one (see Boutin & Coineau, 2000, for a discussion about the concept of phylogenetic ages). Dierent kinds of evidence have been used to establish ages of lineages, but their applicability depends on the aspect of age con sidered. Molecular studies may provide ages of genetic dierentiation, independently of morphological change. Dating of potential geological isolation events, such as periods of climatic stress and large scale geological changes, may be used to infer the time in isolation. Infer ences about relative times of isolation or dierentiation also come out from comparative morphological studies within a phylogenetic and biogeographic framework. Ideally, all evidence should be combined to produce co herent hypotheses about the evolution of subterranean lineages in the temporal scale. In Brazil, robust molecular studies encompassing exclusively subterranean (troglobitic) taxa started very recently and focus on a few sh groups with very special ized troglomorphic derivatives. Basically three groups are under study with focus on populations or species: the phreatobitic characiform Stygicthys typhlops, from a karst area in eastern Brazil (studied by F. P. L. Marques & C. R. Moreira); the Amazonian catsh genus Phreatobius with phreatobic species collected in wells situated in al luvial plains (studied by J. Muriel Cunha); and the hep tapterid subterranean catsh from Chapada Diamantina, northeastern Brazil, belonging to the genus Rhamdiopsis (F.A.Bockmann, pers. comm.), previously cited as a new genus (studied by R. Borowsky & M. E. Bichuette). Few phylogenetic studies with biogeographic analyses of larg er groups including Brazilian troglobites are available. Studies aiming to establish the ages of paleoclimatic uctuations based on speleothem dating are also recent in Brazil, but are progressing quickly. Important climatic changes have been recorded in dierent karst areas, from the presently semiarid northeast to wet areas in the sub tropical southeast. However, these studies are restricted to the late Q uaternary, imposing limits to its application to the problem of establishing ages for subterranean lin eages because many of these lineages probably have a more ancient origin. Older geological events, such as the Miocene Plio-Pleistocene important changes that pro duced the modern Amazon River system, are useful to estimate the age of some Brazilian lineages. Classically, the degree of troglomorphism, basically the reduction of eyes and pigmentation, has been used as a measure of the phylogenetic age for troglobitic animals (Poulson, 1963; W ilkens, 1973, 1982; Langecker, 2000). In spite of the many restrictions to its generalized appli cation (see below), the degree of morphological special ization may, in certain cases, provide relative ages of iso lation in the subterranean environment, being a supple ment to molecular and geological evidence. In the phylogenetic context, a lineage is a branch which departs from one node to another (hypothetical ancestor), from a node to a terminal, or an ancestral branch plus all the derived terminals, including extinct taxa (which remain unnoticed unless a fossil is known). e present discussion deals lineages including termi nals. It must be noted that the ever present possibility of extinction of epigean terminals in a lineage leading to a troglobitic taxon is a source of bias that may produce overestimations of its time of isolation in the subterra nean environment. Among Brazilian subterranean taxa, shes are by far the best studied group with focus on the currently dis cussed aspects. us, I took basically examples from these animals. For the sake of simplicity, I use herein the term subterranean as synonym of troglobitic (exclusively subterranean) species, to the exclusion of the equally subterranean, although not exclusively, troglophilic and trogloxenic populations. E LEONORA T RAJANO

PAGE 3

TIME in KARST 2007 193 nisms acting at dierent rates in each population. An il lustrative example is provided by the armored catshes, Ancistrus cryptophthalmus, from Central Brazil: in the large population found in Anglica Cave, pigmentation is more reduced but eyes are less reduced than in the much smaller population from Passa Trs Caves (Reis et al., 2006). A study based on geometric morphomet rics showed that the four known populations also dier in general body shape, with a mosaic in the deformation axes, indicating divergence probably due (at least partial ly) to topographic isolation (Reis et al., op. cit.). Other mosaics are also observed among related heptapterids among the Rhamdiini, Pimelodella kronei presents eyes more reduced than Rhamdia enfurnada, the opposite be ing observed for melanic pigmentation. Such mosaics may encompass a larger number of characters, including behavioral and physiological ones. is is the case with the troglobitic amblyopsids, tradi tionally ranked in order of increasing degree of reduction of eyes and pigmentation as: T yphlichthys subterraneus < Amblyopsis spelaea < A. rosae (Poulson, 1963). Never theless, A. spelaea presents more specialized life history traits and feeding behavior, while A. rosae is more de rived as regards to agonistic behavior and metabolic rates (both subject to reduction); the otherwise less derived T yphlichthys subterraneus is intermediate in relation to agonistic behavior and metabolic rates (Poulson, 1963; Bechler, 1983). Distinct selective pressures are likely to explain such mosaics. For this reason, attempts to rank species like these according to their degree of adapta tion or specialization to the cave life are unconvincing. In fact, the reduction of melanic pigmentation in subterranean shes results from dierent, independent mechanisms, which may superpose. Morphological mechanisms aect the size and number of melanocytes, whereas physiological ones aect the ability to synthe size melanin. Apparently, this ability may be lost due to dierent mutations aecting at least distinct two steps in the synthesis of eumelanin, one upstream and the other downstream the synthesis of DOPA: the rst cor responds to completely depigmented sh which respond to the administration of L-DOPA by synthesizing mela nin, referred as DOPA(+) by Trajano & Pinna (1996) and tyrosinase-positive by Jeery (2006); the second correspond to depigmented sh which to not respond to L-DOPA (DOPA(-) albinos; Trajano & Pinna, op. cit.). Among Brazilian completely depigmented subterranean shes, Stygichthys typhlops, the new Rhamdiopsis from Chapada Diamantina and the armored catsh, Ancistrus formoso are DOPA(+), the heptapterid T aunayia sp. (actually a Rhamdiopsis F.A. Bockmann, pers. comm.) is DOPA(-) (M.A. Visconti and V. Felice, pers. comm.), and one third of the population of the trichomycterid T richomycterus itacarambiensis is DOPA(-), whereas the remaining two thirds have functional melanophores re duced in density. e morphological mechanism is based on an ad ditive polygenic system (W ilkens, 1988), resulting in a continuous variation in the rst evolutionary steps and progressing towards complete depigmentation through out the population at slower rates than that caused by the loss of the ability to synthesize melanin, which is based on monogenic systems (W ilkens, 1988). For instance, it has been demonstrated that albinism in dierent popu lations of Mexican Astyanax is caused by independent mutations in the same gene, Oca2 (Protas et al., 2005). erefore, very pale but still pigmented sh species, with scattered micromelanophores (such as the T richomycter us undescribed species respectively from Bodoquena and from Serra do Ramalho karst areas, and the Ituglanis spp. from So Domingos karst areas) may be younger than any of those DOPA(+) albinos. us, the use of troglo morphic pigmentation as a measure of relative age should be restricted to related taxa retaining melanin (i.e., to the exclusion of DOPA albinos), where the degree of pale ness is due to mutations in the additive polygenic system underlying the morphological, gradual mechanism. Regression of eyes is also due to complex genetic systems. In the blind Mexican tetra characins, genus As tyanax it has been shown that regression is caused by the inactivation of several genes that take part in the developmental control, and that growth factors acting at a lower level of this control appear to be involved in the degeneration of the eyes (Langecker, 2000). Clearly, studies on a much large sample of troglobitic species are needed before any inference about dierentiation rates can be made. Two other factors inuence the rates of divergence: population sizes and life cycle strategies. Small popula tions tend to dierentiate faster due to phenomena as ge netic dri. Population sizes are highly inuenced by eco logical factors such as nutrient availability and the extent of habitats suitable for colonization. It is noteworthy that energy is higher in streams (higher carrying capacity), but phreatic habitats occupy larger areas and volumes. Because there is no taxonomic correlation with these fac tors, related species may dier in population sizes (for instance, populations respectively with 20,000 and 1,000 individuals were estimated for A. cryptophthalmus in An glica and in Passa Trs caves Trajano, 2001a), thus in divergence rates. As well, nutrient availability may also be perceived dierently even by taxonomically related species, depending on the erciency of energy use. Such erciency may be improved along the adaptation to the subterranean life, allowing for increase in population sizes, then in lowered dierentiation rates. THE CHALLENGE OF ESTIMATING THE AGE OF SUBTERRANEAN LINEAGES: E X AMPLES FROM BRAZIL

PAGE 4

TIME in KARST 2007 194 K-selected life strategies imply lower dierentia tion rates due to delayed ages for rst maturation and low reproductive rates (few individuals reproducing at given times), which work on opposite directions: delayed rst maturation implies slow divergence rates (longer re productive generations), whereas low reproductive rates result in lowered eective populations, which would ac celerate divergence rates.. In conclusion, there is a complex balance between dierent genetic, ecological and biological factors, which may act in dierent directions to produce the actual di vergence rates. Such rates may dier among related taxa, and even among dierent characters. erefore, the de gree of troglomorphism as a measure of age of subterra nean lineages should be used with extreme caution. G EOLOGICAL PALEOCLIMATIC AND BIOGEOGRAPHICAL EVIDENCE : Dating of paleoclimatic events based on growth phases of speleothems and similar deposits may be applied to sub terranean lineages within the framework of the paleocli matic model (Barr, 1968; W ilkens et al., 2000). However, its cyclical nature imposes serious limitations because, without biological data (molecular, morphological, phy logenetic), it is not possible to establish in which phase the isolation rst took place. As a matter of fact, isolation with dierentiation may occur along several subsequent unfavorable phases intercalated with coalition phases, thus what really counts to produce genetic and/or mor phological divergence is the sum of isolation periods (Trajano, 1995), and not simply the time since the rst isolation event. For instance, in northeastern Brazil there were nine dry phases (no speleothem growth) in the last 210,000 years, intercalated with short wet phases lasting from several hundreds to a few thousand years each. Overall, these periods of speleothem growth occupied only 8% of the studied period, i.e., around 20,000 years in con trast with 190,000 years with dry conditions, like the one prevalent nowadays in the region (W ang et al. 2004). Hence, at least in the late Pleistocene, there was a much extended period of isolation for the hypogean fauna in northeastern Brazil for lineages already established in subterranean habitats, from 190,000 to 210,000 years, de pending on the occurrence or not of introgression with epigean relatives during the wet periods. As a matter of fact, several of the most highly specialized Brazilian tro globites have been found in this region (e.g., Rhamdiopsis catshes, Spelaeogammarus amphipods, Pongicarcinia xi phidophorum isopods, Coarazuphium beetles), as well as the only Brazilian troglobitic scorpions, cockroaches and Ctenid spiders. On the other hand, climatic changes were not as dominant in the subtropical southeast Brazil and dry phases were shorter, at least for the last 116,200 years (Cruz-Jr. et al., 2005). erefore, total time of isolation in subterranean habitats during the late Pleistocene was shorter in SE than in NE Brazil. Hypothetically, a pop ulation that became rst isolated at a given time in the northeast would be much more dierentiated, both ge netically and morphologically, than another population rst isolated at the same time in the southeast. If one con siders age as the time of the rst isolation, these two lineages have the same age; if age is the total time in isolation, then the rst one is older. It is clear that, in a cyclical model, the degree of genetic dierentiation do not provide a good evidence of age without a precise de termination of the duration of each phase. Geological and geographical events over larger temporal scales may provide more robust evidence. e genus Phreatobius is distributed around the Amazon basin, in tributary basins from both margins of the Am azon River. e rst described species, P. cisternarum, lives underground in the alluvial fan around the Ama zon delta, being collected in shallow hand-dug wells. Much latter, in the 1990s, other species were found deeply buried in submerged litter banks in shallow igaraps (small tributaries) along the le margin of the Negro and Amazon rivers. More recently, a second phreatobic species was discovered in wells in the State of Rondnia, Rio Madeira basin, in the right margin of the Amazon drainage (J. Muriel-Cunha & J. Zuanon, pers. comm..; description in progress by J. Muriel-Cunha & M. de Pinna). is wide, peripheral distribution of the Phreatobius genus around the Amazon basin may be explained by an origin between the late Miocene and the late Pliocene (~2.5 Ma), when a gigantic lake, or a series of interconnected mega-lakes occasionally united to cover most or all of lowland Amazonia to a shallow depth (Campbell et al., 2006). In fact, Phreatobius cat shes are adapted to shallow, hypoxic conditions, with dark pink to red skin indicating cutaneous breathing; since all known species exhibit this conspicuous trait, this is probably an ancestral condition for the genus. I suggest that the fragmentation of the lacustrine habitat during the late Pliocene, leading to the establishment of the modern Amazon River drainage system, may have been an isolation event for the ancestors of the extant E LEONORA T RAJANO

PAGE 5

TIME in KARST 2007 195 species. Nevertheless, an older origin for cannot be ruled out. On the other hand, P. cisternarum has been found not only north and south of the Amazon River mouth but also in the large Maraj Island in between, with no unequivo cal morphological dierentiation so far detected between these localities (Muriel-Cunha & Pinna, 2005). ese pop ulations were isolated during the formation of the Amazon delta, ~2.5 Ma ago, suggesting a high evolutionary stability, at least at the morphological level, possibly due to the envi ronmental stability of the subterranean habitat. e disjunct distribution also points to a very ancient origin for the Calabozoidea isopods. So far, this taxon is composed exclusively by three extant phreatobic species, one from the Orinoco basin, in Venezuela (Calabozoa pellucida), e two from Brazil, respectively from the So Francisco (Pongycarcinia xiphidiourus) and the Paraguay (undescribed species) river basins. e only connection between these regions is through the Amazon basin, and I speculate that the ancestors may have dispersed during or prior to the formation of the huge Lago Amazonas. Actually, the So Francisco lineage would be older, at least 5 Ma, which is the estimated age of separation of this basin based on studies of the biogeographical patterns in Brazilian freshwater shes (Hubert & Renno, 2006). Mes sana et al., (2002) argue for a close relationship between the Calabozoidea and the Oniscoidea isopods, thus both lineages have the same phylogenetic age, which goes back to the Jurassic-Cretaceous (gondwanic origin L. A. Souza, pers. comm.). A phylogenetic study, that could add more light to this interesting problem, is waiting for the collection of additional specimens, what is proving to be very dircult in spite of the eorts of biologists and cave divers. Apparently these animals are very rare and/ or live mainly in inaccessible, deep phreatic habitats. Geomorphological events as alluvial erosion pro ducing waterfalls that split populations (topographical isolation), once dated, also provide data useful to esti mate the age of lineages such as the dierent populations of the armored catsh, A. cryptophthalmus. P H Y LOGENETIC AND MOLECULAR EVIDENCE : In order to be minimally reliable and useful, molecular clocks must be based on well corroborated phylogenies with at least one node correlated to geographic or geo logical isolation events of known age. In cyclical models, such correlation is hampered when cycles are relatively short and repetitive, as is the case with the paleoclimatic uctuations in the late Pleistocene in Brazil, adding a great deal of uncertainty to the molecular clock. Ma rine transgressions, which have been used to establish dates for vicariant events in epigean Brazilian taxa such as freshwater shes, are of no use for subterranean lin eages because almost all karst areas in Brazil are above the maximum sea levels. In any case, the conclusion of the molecular studies on Phreatobius spp., S. typhlops and Rhamdiopsis sp. from Chapada Diamantina will certainly open new interesting avenues in this eld. As already mentioned, few phylogenetic studies of groups including Brazilian troglobites are available, most at the genus level and incomplete in terms of taxa encom passed. Among shes, the heptapterid catshes were ob ject of a phylogenetic study, but the cave species were not included (Bockmann, 1998). Phylogenetic and molecular studies on heptapterids are in progress, but the position of the Phreatobius genus and of the troglobitic Rhamdi opsis species within this genus are still unclear. Recently analyzed morphological data indicate that, within the ge nus Rhamdiopsis, T aunayia sp. is basal whereas the spe cies from Chapada Diamantina have a more apical posi tion in the phylogeny (F. A. Bockmann, pers. comm.). ese two species independently adapted to the same kind of habitat, the upper phreatic zone connected to the surface through caves (Trajano, 2001b), having devel oped advanced characters states related to the hypogean life, including miniaturization. T aunayia sp., however, is even more specialized, presenting a hypertrophied lat eral line system in the head, with behavioral evidence of enhanced mechano-sensory sensitivity. is, associated with its putative basal position in the Rhamdiopsis phy logeny, points to an older age for the lineage to which the troglobitic T aunyaia sp. belongs, much anterior to the late Pleistocene. e phylogeny of the catsh family Trichomycteri dae was also studied (W osiack, 2002), but only one among 10+ troglobitic species presently known, T richomycterus itacarambiensis, was included. It is an apical taxa in the phylogeny, indicating a relatively recent origin. A recent derivation of T itacarambiensis from an epigean ancestor from the Upper So Francisco River basin is consistent with the morphological variation observed in eyes and pigmentation and also with the notion of a quick xation of genes for albinism, since one third of the population is made of albinos. However, in the absence of a correlation between some node and dated geographic or geological isolation events, it is not possible to estimate an absolute age, even approximate, for this cave lineage. THE CHALLENGE OF ESTIMATING THE AGE OF SUBTERRANEAN LINEAGES: E X AMPLES FROM BRAZIL

PAGE 6

TIME in KARST 2007 196 C OMBINED EVIDENCE : For extremes in the inter-taxa variation, the troglomor phism degree may provide good evidence of relative ages. For instance, it is reasonable to suppose that shes with slightly reduced eyes and pigmentation such as the heptapterids Rhamdiopsis sp. from Cordisburgo (east ern Brazil) and Pimelodella spelaea, from So Domingos (Central Brazil), are younger than the highly troglomor phic Rhamdiopsis sp. from Chapada Diamantina and T aunaya sp., from Campo Formoso. e two former species probably have been isolated topographically be cause they inhabit streams several meters above the base level, and an isolation period in the order of 10 5 years (es timated time for the erosional processes lead to the cur rent landscape A. Auler, pers. comm..) may be estimat ed. e two latter species inhabit presently semiarid karst areas in northeastern Brazil subject to extended periods of isolation at least during the last 210,000 years, but they probably became isolated well before. us, an estimate in the order of 10 5 -10 6 years seems reasonable. A molecular study focusing on the hypervari able Region I of MtDNA did not nd any evidence of divergence between the cave populations of Ancistrus cryptophthalmus (Moller & Parzefall, 2001). However, geometric morphometric analyses showed a clear, statis tically signicant dierence between these populations, but with some superposition with the epigean closest relatives (Reis et al., 2006). Taken together, these data in dicate a recent isolation of the cave populations from the epigean ones and also from each other, in the order of 10 4 -10 5 years. Preliminary molecular studies on Ituglanis spe cies from So Domingos karst area are consistent with the observed morphological dierences (Bichuette et al., 2001) justifying the recognition of four species, each one in a separate microbasin that runs parallel westwards (Bichuette & Trajano, 2004). ese catshes are sym patric with the morphologically less specialized A. cryp tophthalmus, P. spelaea and Eigenmannia vicentespelaea (Gymnotiformes), making So Domingos karst area a world hotspot of biodiversity for subterranean shes. All the Ituglanis catshes have eyes more reduced and are paler than the other species, presenting scattered mela nophores, i.e., they are not DOPA albinos. ree among these Ituglanis species occupy a very specialized habitat, with adaptations to the phreatic environment that include miniaturization. Moreover, I. epikarsticus, and prob ably also I. bambui and I. ramiroi (Trajano & Bichuette, unpubl. data), live and disperse through the epikarst, whereas the other species are typical stream-dwellers, like their epigean relatives. In spite of intensive collect ing eorts, no epigean Ituglanis catsh was found in So Domingos (the same is true for Pimelodella; Bichuette & Trajano, 2003). Taken together, these evidences indicate a longer time in isolation for the Ituglanis catshes. In con clusion, the rich troglobitic ichthyofauna from So Do mingos seems to be the result of anachronous isolation events, including both the extinction of epigean relatives due to unknown factors (for Ituglanis and Pimelodella ) and topographic isolation (for Ituglanis spp. and also A. cryptophthalmus) Anachronous isolation, possibly in association with dierent divergence rates, may also explain the disparity in troglomorphic degree observed for the subterranean fauna from the Upper Ribeira Valley karst area, SE Brazil. is fauna includes very specialized species, such as the pseudoscorpion Spelaeobochica muchmorei and the deca pod Aegla microphthalma, to moderately troglomorphic species, such as the opilionid Pachylospeleus strinatii, the carabid beetle Schizogenius ocellatus and the catsh Pimelodella kronei. W ithin the framework of the paleo climatic model, in view of the short isolation periods (= dry phases) during the late Pleistocene (see above) it is probable that all these species became rst isolated in caves before this period. A CKNO W LEDGEMENTS I am grateful to Augusto Auler, Fernando P. L. Marques and Janice Muriel Cunha for the discussion of ideas and criticisms, and to Richard Borowsky for the critical read ing and revision of the English style of an early version of the manuscript. Many data were gathered during stud ies sponsored by the Fundao de Amparo Pesquisa do Estado de So Paulo FAPESP (grant n. 03/00794-5, among others). e author studies are also supported by the CNPq (fellowship and grant n. 302174/2004-4). E LEONORA T RAJANO

PAGE 7

TIME in KARST 2007 197 R EFERENCES : Bechler, D. L., 1983: e evolution of agonistic behavior in amblyopsid shes. Behavioral Ecology and Socio biology, 12, 35-42, Heidelberg. Bichuette, M. E., Garz, A. & D. Mller., 2001: Preiminary study on cave-dwelling catshes, Ituglanis sp., from Gois, Brazil, p. 22. In: E. Trajano. & R. Pinto-daRocha (eds.). X V International Symposium of Bio speleology, Socit Internationale de Biospologie, So Paulo. Bichuette, M. E. & E. Trajano., 2003: Epigean and subter ranean ichthyofauna from So Domingos karst area, Upper Tocantins river basin, Central Brazil. Journal of Fish Biology, 63, 1100-1121, London. Bichuette, M. E. & E. Trajano., 2004: ree new subterra nean species of Ituglanis from Central Brazil (Siluri formes: Trichomycteridae). Ichthyological Explora tions of Freshwaters, 15, 3, 243-256, Mnchen. Bockmann, F. A., 1998: Anlise logentica da famlia H eptapteridae ( T eleostei, Ostariophysi, Siluriformes) e redenio de seus gneros. PhD esis, Universi dade de So Paulo, So Paulo, 589 p. Boutin, C. & N. Coineau., 2000: Evolutionary rates and phylogenetic age in some stygobiontic species, p. 433-451. In: W ilkens, H., Culver, D.C. & W .F. Hum phreys (eds.). Ecosystems of the W orld 30. Subter ranean Ecosystems (eds.). Elsevier, Amsterdan. Campbell Jr., K. E., Frailey, C. D. & L. Romero-Pittman., 2006. e Pan-Amazonian Ucayali Peneplain, late Neogene sedimentation in Amazonia, and the birth of the modern River system. Palaeogeography, Pa laeoclimatology, Palaeoecology, 239, 166-219, Am sterdan. Cruz Jr., F. W ., Burns, S. J., Karmann, I., Sharp, W D., Vuille, M., Cardoso, A. O., Ferrari, J. A., Dias, P. L. S. & O. Viana Jr., 2005: Insolation-friven changes in atmospheric circulation over the past 116,000 years in subtropical Brazil. Nature, 434, 63-66, London. Hubert, N. & J.-F. Renno., 2006: Historical biogeogra phy of South American freshwater shes. Journal of Biogeography, p. 1-23 [www.blackwellpublishing. com/jbi] Jeery, W R., 2006: Convergence of pigment regression in cave animals: developmental, biochemical, and genetic progress toward understanding evolution of the colorless phenotype., p. 38. In: Moldovan, O. T. (ed.). X VIII th International Symposium of Biospel eology 100 years of Biospeleology, Cluj-Napoca, SIBIOS Socit Internationale de Biospologie. Langecker, T.G., 2000: e eects of continuous darkness on cave ecology and and cavernicolous evolution, p. 135-157. In: W ilkens, H., Culver, D.C. & W .F. Hum phreys (eds.). Ecosystems of the W orld 30. Subter ranean Ecosystems (Eds.). Elsevier, Amsterdan. Messana, G., Baratti, M. & D. Benvenuti., 2002: Pongy carcinia xiphidiourus n. gen. n. sp., a new Brazilian Calabozoidae (Crustacea Isopoda). Tropical Zool ogy, 15, 243-252, Firenze. Muriel-Cunha, J. & M. de Pinna., 2005: New data on cistern catsh, Phreatobius cisternarum, from sub terranean waters at the mouth of the Amazon River (Siluriformes, Incertae Sedis). Papis avulses de Zoologia, 45, 26, 327-339, So Paulo. Moller, D. & J. Parzefall., 2001: Single or multiple origin of the subterranean catsh Ancistrus cryptophthal mus. W hat we can learn from molecular data, p. 38. In: Trajano, E. & R. Pinto-da-Rocha (eds.). X V International Symposium of Biospeleology, Socit Internationale de Biospologie, So Paulo. Poulson, T. L., 1963: Cave adaptation in Amblyopsid shes. American Midland Naturalist, 70, 2, 257-290, Notre Dame. Protas, M. E., Hersey, C., Kochanek, D., Zhou, Y ., W ilkens, H., Jeery, W R., Zon, L. I. Borowsky, R. & C. J. Tabin., 2005: Genetic analysis of cavesh reveals molecular convergence in the evolution of albinism. Nature Genetics, Advance Online Publication, Let ters, p. 1-5 [published online 11 December 2005] Reis, R. E., Trajano, E. & E. Hingst-Zaher., 2006: Shape variation in surface and cave populations of the armoured catsh Ancistrus (Siluriformes: Lori cariidae) from the So Domingos karst area, Upper Tocantins River, Brazil. Journal of Fish Biology, 68, 414-429, London. Trajano, E., 1995: Evolution of tropical troglobites: Ap plicability of the model of Q uaternary climatic uc tuations. Mmoires de Biospologie, 22, 203-209, Moulis. Trajano, E., 2001: Habitat and population data of tro globitic armoured cave catshes, Ancistrus cryp tophthalmus Reis 1987, from Central Brazil (Silu riformes: Loricariidae). Environmental Biology of Fishes, 62, 1-3, 195-200, Dordrecht. Trajano, E., 2001: Ecology of subterranean shes: an overview. Environmental Biology of Fishes, 62, 1-3, 133-160, Dordrecht. THE CHALLENGE OF ESTIMATING THE AGE OF SUBTERRANEAN LINEAGES: E X AMPLES FROM BRAZIL

PAGE 8

TIME in KARST 2007 198 Trajano, E. & M.C.C. Pinna., 1996: A new cave species of T richomycterus from eastern Brazil (Siluriformes, Trichomycteridae). Revue franaise dAquariologie, 23, 3-4, 85-90, Nancy. W ang, X., Auler, A. S., Edwards, R. L., Cheng, H., Cris talli, P. S., Smart, P. L., Richards, D. A. & C.-C. Shen., 2004: W et periods in northeastern Brazil over the past 210 kyr linked to distant climate anomalies. Nature, 432, 740-743, London. W ilkens, H., 1973: Anciennet phylognetique et degrees de reduction chez les animaux cavernicoles. Annal es de Splologie, 28, 2, 327-330, Paris. W ilkens, H., 1982: Regressive evolution and phylogenetic age: the history of colonization of freshwaters of Y u catan by sh and Crustacea. Texas Memorial Mu seum Bulletin, 28, 237-243. W ilkens, H., 1988: Evolution and genetics of epigean and cave Astyanax fasciatus (Characidae, Pisces). Evolu tionary Biology, 23, 271-367, New Y ork. E LEONORA T RAJANO


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.