COBISS: 1.01 GLACIAL DESTRUCTION OF CAVE SYSTEMS IN HIGH MOUNTAINS, wITH A SPECIAL REFERENCE TO THE ALADAGLAR MASSIF, CENTRAL TAURUS, TURKEY LEDENIŠKO UNIčENJE VISOKOGORSKIH JAMSKIH SISTEMOV: PRIMER MASIVA ALADAGLAR, CENTRALNI TAURUS, TURčIJA Aleksander KLIMCHOUK1, Serdar BAYARI2, Lütf NAZIK3 & Koray TÖRK3 Abstract UDC 551.331:551.44(560) Aleksander Klimchouk, Serdar Bayari, Lütf Nazik & Koray Törk: Glacial destruction of cave systems in high mountains, with a special reference to the Aladaglar massif, Central Tau-rus, Turkey Erasure of karst features and dissection of karst are among the main destructive efects of glacial action upon karst (Ford, 1983). Tey lead to destruction of functional relationship be-tween the relief and a karst system, and to glacial dissection of pre-glacial cave systems. Stripping of the epikarstic zone and upper parts of cave systems on sub-horizontal surfaces results in prevalence of decapitated shafs in high mountains afected by glaciations. Vertical dissection of a karst massif by glacial ero-sion creates cave openings in sub-vertical surfaces (clifs), a well known feature. Observations of vertical shafs exposed by clifs are less common. Such shafs, unwalled by surface geomorphic processes, are in a certain way an analogous to the “unroofed” caves, exposed by denudational lowering of sub-horizontal surfaces. Te Aladaglar Massif (Central Taurus, Turkey) is an outstanding example of high mountain karst. Te high-altitude part of the massif has been severely glaciated during quater-nary Glacial erosion was the dominant factor in the overall surface morphology development, resulting in the formation of numerous glacial valleys, cirques, ridges and pyramidal (horn) peaks. Te overall relief between the highest peaks and the lowest karst springs in Aladaglar is 3350 m. Te local vertical magnitude of relief between bottoms of glacial valleys and sur-rounding ridges is up to 1700 m. Recent studies suggest that the most recent major glaciation occurred in the Aladaglar massif during the Holocene Cooling and terminated between 9,300 and 8,300 years BP. Tis paper describes unwalled shafs at sub-vertical surfaces, a feature which is common in Aladaglar but is not so common, or overlooked, in other high mountain areas. Exposure of such shafs is mainly due to intense gravitational processes induced by the combined efect of the removal of the ice support to clifs and the glacial rebound. keywords: glaciations, karst, denuded caves, unwalled shafs, Aladaglar, Central Taurus, Turkey 1 Ukrainian Institute of Speleology and Karstology, Ministry of Education and Science, NASU, Ukraine; klim@speleogenesis.info 2 Hydrogeological Engineering Section of Hacettepe University, Turkey; e-mail: serdar@geo.hacettepe.edu.tr 3 General Directorate of Mineral Research and Exploration, Turkey Received / Prejeto: 27.07.2006 ACTA CARSOLOGICA 35/2, 111-121, LJUBLJANA 2006 Izvleček UDK UDK 551.331:551.44(560) Aleksander Klimchouk, Serdar Bayari, Lütf Nazik & Koray Törk: Ledeniško uničenje visokogorskih jamskih sistemov: primer masiva Aladaglar, Centralni Taurus, Turčija Erozija kraških površinskih oblik in zarezovanje v kras so najbolj uničujoče posledice ledeniškega delovanja na krasu (Ford, 1983). Rezultat je prekinjena funkcijska povezava med reljefom in kraškim sistemom ter razrez predglacialnih jamskih sistemov. Na površinah z majhnim naklonom, ledeniško delovanje prizadene predvsem epikras in vrhnje dele jamskih sistemov. Na takih površinah najdemo veliko brezen, ki jim je ledeniška erozija odstranila vrhnje dele (t.i. obglavljena brezna). Zaradi vertikalnega vrezovanja v kras, se v stenah masivov odpirajo jamski vhodi, redkeje pa naletimo na vzdolžno prerezana, izpostavljena brezna. Taka, »brezstenska« brezna, so na nek način analogija brezstropih jam, ki so nastale kot posledica denudacije sub-horizontalnih površin. Masiv Aladaglar (Centralni Taurus, Turčija) je izreden primer visokogorskega krasa. Višji deli masiva to bili tekom kvar-tarja močno poledeneli, zato tam prevladuje tipična ledeniška morfologija v vseh pojavnih oblikah. Višinska razlika med najvišjimi vrhovi Aladaglarja in najnižjimi kraškimi izviri je 3350 metrov, lokalni vertikalni razpon med ledeniškimi dolinami in okoliškimi grebeni doseže 1700 m. Novejše raziskave kažejo, da je bila zadnja velika poledenitev v Aladaglarju med holocensko ohladitvijo, ki je trajala med 9300 do 8300 leti pred današnjostjo. V članku obravnavamo »brezstenska brezna«, ki so v stenah Aladaglarja pogosta. Takih brezen je v drugih visokogorskih masivih malo, ali pa so bila prezrta. Razkritje brezen je delo gravitacijskih procesov ob umiku ledenikov. ključne besede: poledenitev, kras, denudirane jame, brezstenska brezna, Aladaglar, Centralni Taurus, Turčija. ALEKSANDER KLIMCHOUK, SERDAR BAYARI, LüTFI NAZIK & KORAY TÖRK INTRODUCTION During the last decade considerable attention has been given by many researchers to so-called unroofed (de-nuded) caves. It was generally appreciated long ago that lowering of the karst surface due to ongoing denudation ultimately results in uncovering and destruction of caves. However, it was the work of Mihevc and his colleagues (Mihevc, 1996; Mihevc et.al., 1998) that drew specifc attention to the topic. Subsequent publications by many scholars shed light on several aspects, to which the specifc study of unroofed caves gave useful information. Unroofed caves were recognized as adistinctive sub-type of surface karst, a cave partially transformed by surface processes. More understanding arises about their roles in the formation of karst landscape and about the overall denudation progress in karst. Based on observations in tropical karst, Klimchouk (2005) revived the view that unroofng of caves can be a large-scale geomorphic proc-ess. Te cosmogenic nuclide exposure dating (Gosse & Phillips, 2001) of rock surfaces exposed due to cave unroofng can give invaluable information on aging of unroofng events and relevant landforms - a sound possibil-ity which is still to be tested and realized. Most of studies of unroofed caves came from the areas of moderate to low relief karst topography, so that consequently they were focused on sub-horizontal pas-sages that were unroofed by the sub-horizontal denudation surface. Tis is refected by the very term “unroofed caves”, which implies opening of sub-horizontal cave ele-ments that had a roof. works that investigate this topic in high mountain environment, are rare (see Mais, 1999 for an examples from Alps). In high mountains there is considerable vertical relief, which introduces more complex-ity into the conceptual representation and genetic con-sideration of the phenomena: as a consequence the term “unroofng” seems to be insufcient to describe various relations of caves with the surface. Our recent karst and cave studies in the Aladaglar Massif, Central Taurus, Turkey, yielded a variety of in-structive observations on diferent types of caves exposed to the surface by various geomorphological processes. In particular, this study revealed extraordinary examples of the exposure of vertical caves (shafs) by sub-vertical surfaces. In this article we present these observations, which inspired discussion of some general terminological and geomorphological aspects and gave insights to some problems of local geomorphological evolution. GENERAL REMARKS ON TERMINOLOGY Te initial term “roofess caves” has been gradually re-placed by the more correct “unroofed caves”. A general defnition is that unroofed caves are old caves that have been exposed due to the lowering of karst relief. Tis tac-itly implies a sub-horizontal orientation of the lowering surface, hence - sub-horizontal caves are the most read-ily available for observations when truncated (unroofed) by such surface. Interestingly enough, vertical shafs cut by lowering of the sub-horizontal denudation surface are not considered as something of special interest in the context of “unroofng” as they retain the capacity of entrances to the underground space. Shafs retain their status as underground forms, and they are not going to be erased geologically as fast as sub-horizontal caves do when unroofed. Apparently, the term “unroofed caves” does not apply to vertical shafs. Four types of shafs can be distinguished accord-ing to the mode of their exposure to the sub-horizontal surface. Ponor shafs are those developed in a direct genetic (hydrologic) relationship with the surface and still retaining this relationship, such as shafs swallowing streams formed on adjacent non-karstic rocks or catch- 112 ACTA CARSOLOGICA 35/2 - 2006 ments where dispersed infltration is prevented by patch-es of a low-permeability cover. Epikarst shafs are those developed at the bottom of the epikarstic zone as epi-karst-draining paths, and opened to the surface due to its gradual lowering and collapse (Klimchouk, 2004). Col-lapse shafs are those formed by collapsing of large underground rooms. Decapitated shafs are those exposed due to erasure of the upper part of a massif by some high energy agency, commonly by glacial stripping. Te latter category is pertinent to the subject of this article. In the high mountain karst steep to vertical surfaces such as cirque headwalls are common. Tey can be hundreds of meters in height, and sometimes more than 1500 m. Interception of sub-horizontal passages by sub-vertical surfaces creates open cave entrances, a common phenomenon. Tis article focuses on vertical shafs opened by such sub-vertical surfaces, a less known phenomenon. Teir nature within the topography can difer because they can be created by fuvial incision or glacial erosion and/or gravity (rock detachment, fall and slide). In general terms we are considering the intersection of 3D daylight surfaces with 3D systems of underground GLACIAL DESTRUCTION OF CAVE SYSTEMS IN HIGH MOUNTAINS, wITH A SPECIAL REFERENCE TO THE ALADAGLAR ... caves. Te daylight surface is polygenetic, and its inter-section with a polygenetic 3D cave system will be guided by a number of geomorphic agents, as well as by the to-pology of both systems. Te Table 1 clarifes the termi-nology for intersection features based on simple geomet-ric considerations. struction of respective underground forms themselves. Unroofed caves and unwalled shafs are the two major types of disintegrating cavities, which can be collectively referred to as denuded (or exposed) caves. Te term “cave ruins” is also used to describe vari-ous kinds and states of cave disintegration (Mais, 1999). tab. 1: features resulting from intersection of the daylight surface and a 3d cave system Components of a 3D cave system Geomorphic agencies dominating in creation of diferently oriented surfaces that open caves in high mountains Sub-vertical faces Sub-horizontal surfaces fuvial incision, glacial erosion, gravitational destruction (rock detachment and slide) denudation due to dissolution, weathering mass wasting and areal erosion, glacial erosion by icecaps and at the bottoms of glaciers Sub-horizontal cave elements (passages) cave openings (entrances) unroofed caves Sub-vertical cave elements (shafts) unwalled shafts shaft openings (entrances) Inclined cave components and inclined surfaces can produce a variety of features at their intersection. How-ever, they can be assigned to one of the basic categories distinguished in the table and do not need specifc terms. Cave and shaf openings (entrances) do not imply de- It could be used to describe felds of closely spaced disintegrating underground forms, when individual unroofed passages are barely recognizable. THE ALADAGLAR KARST Aladaglar is an outstanding karst massif located in the Central Taurus Range in the Adana-Kayseri-Nigde prov-inces of Turkey It is situated between the regional Ecemis Fault to the west and the deeply incised Zamanti River valley to the east (Figs 1 and 2). Te southeastern part of Turkey is an active plate boundary where the Arabian and the Eurasian plates are colliding along the Bitlis-Za-gros suture. Tis determined the intensity of neotectonic processes and uplif since Late Oligocene, with the high-est rates occurring in the Plio-Pleistocene. Te Aladaglar Massif is composed chiefy by Triassic, Jurassic and Cre-taceous limestones and rises up to 3750 m in elevation. Te overall relief between the highest peaks and the low-est karst springs is 3350 m. Te altitudinal distribution of the principal components of the surface geomorphology and the major known cave system is illustrated in Fig. 3. Te Zamanti River valley provides the general base level of erosion, towards which the karstic underground drainage is directed. Te overall morphology is well illustrated by digital elevation models produced from the “Aladaglar Karst” GIS database developed during this study (Fig. 2). Morphologically, the northern (Black Aladag), central and southern (white Aladag) sectors can be distinguished, with the local relief increasing from north to south. Te high-altitude parts of the Aladaglar massif have been severely glaciated during quaternary Aladaglar belongs to the Pyrennean type of glacial landscape, i.e. glaciers there were confned to high areas but not occu-pied lower valleys where the outputs of a karst system oc-curred (Ford & williams, 1989). Glacial erosion was the dominant factor in the overall surface morphology devel-opment, resulting in the formation of numerous glacial trough valleys, cirques, aretes (narrow jagged ridges) and horn (or pyramidal) peaks. Our recent studies suggested that the magnitude and extension of quaternary glacia-tions in Aladaglar was greater than previously thought (Bayari et. al, 2003; Zreda et al., 2005). Although the glacial landforms indicate there have been numerous epi-sodes of glacial advance and retreat, evidences of the old-er glaciations are largely erased by the efects of the last, remarkably extensive, glaciation. Cosmogenic 36Cl dating of morainic boulders in the Hacer Valley suggests that it terminated between 9,300 and 8,300 years BP there. Glacial geomorphic processes acting on the karsti-fed limestone substratum gave rise to the distinct pecu- ACTA CARSOLOGICA 35/2 - 2006 113 ALEKSANDER KLIMCHOUK, SERDAR BAYARI, LüTFI NAZIK & KORAY TÖRK fig. 1: location of the Aladaglar massif (lef) and its overview digital elevation model. fig. 2: Physiography of the Aladaglar massif. Te dEm is based on the 1:25,000 topographic map, overlain by a landsat satellite image and data layers from the GiS “Aladaglar Karst and Caves”. distribution of explored caves is shown by small red dots and major springs are indicated by blue dots. 114 ACTA CARSOLOGICA 35/2 - 2006 GLACIAL DESTRUCTION OF CAVE SYSTEMS IN HIGH MOUNTAINS, wITH A SPECIAL REFERENCE TO THE ALADAGLAR ... fig. 3: Altitudinal distribution o f princip al denudation and erosion levels, springs and caves in the Aladaglar massif, Central taurus, turkey. A composite profle Sw-NE, sub-parallel to the Barazama valley. liar features known as glaciokarstic morphology. However, in local areas of moderate relief (glacial source areas and valley bottoms) recent glacial scouring of prominent mesoforms on the one hand, and flling of negative mesoforms by weath-ering (frost shatter) debris on the other hand, makes appearance of karstifed surfaces generally smoother that it can be typically seen in lower-al-titude Alpine karst massifs (Fig. 4). Glacial valleys created during the quater-nary glaciations were entrenched into an already intensely karstifed massif. Smaller glacial valleys extend from source areas at 3100-3300 m down to altitudes of about 1900-2300 m, while some large valleys (such as Hacer) incised as low as 1100 m asl. Te local altitudinal ranges between ridges and fig. 4: Characteristic high altitude glaciokarstic landscape in Aladaglar. Glacial scouring in the recent past has made appearance of karstifed surfaces generally smoother than can be typically seen in lower-altitude Alpine karst massifs, where the last glaciation terminated earlier. A = yedigoller Plateau, a paleo-source area for the hacer valley glacier at the altitudes of 3100-3400 m; B = harmancik Plateau, a paleo-source area for the Kemikli valley glacier at the altitudes of 3100-3200 m. Photos by E. medvedeva (A) and A. Kopchinsky (B). ACTA CARSOLOGICA 35/2 - 2006 115 ALEKSANDER KLIMCHOUK, SERDAR BAYARI, LüTFI NAZIK & KORAY TÖRK fig. 5: Section across the Karagol, hacer and Kemikli valleys, showing local elevation diferences and the typical occurrence of caves and shafs intersected by variably oriented surfaces. valley bottoms can be as great as 1700 m in large valleys, although in smaller val-leys they are only of 200-700 m (Fig. 5 and 6). Te steep valleys, with their many sub-vertical faces, are subject to intense gravi-tational processes. Te freshly glaciated rocky surfaces have many diferent orientations and thus expose numerous pre-glacial cavities (in the sense that they are older than at least the last major glaciation), creating all types of intersections outlined in Table 1: shaf openings and unroofed caves on sub-horizontal surfaces and cave openings and unwalled shafs on sub-vertical surfaces. fig. 6: Kemikli glacial valley. Note an unwalled shaf in the clif on the lef and a train of boulders on the scree apron and glacially stripped bedrock at the foot of the clif. Photo by A.Klimchouk. 116 ACTA CARSOLOGICA 35/2 – 2006 GLACIAL DESTRUCTION OF CAVE SYSTEMS IN HIGH MOUNTAINS, wITH A SPECIAL REFERENCE TO THE ALADAGLAR ... DECAPITATED SHAFTS During 2001-2004 over 150 caves were explored in the Aladaglar Massif. Tey were mainly vertical, with an aggregate total depth of 6640 m. Of them 32 caves are deeper than 50 m, and 12 caves are in excess of 100 m. Fify-seven caves are located above 3000m, the highest explored example being at 3410 m. Te great majority of shafs explored in the high karst zone are decapitated shafs that had been exposed due to erasure of the upper part of the massif by glacial stripping. Erasure of karst features by mechanical abra-sion of bedrock at the base of ice is a known efect of glaciations on karst (Ford, 1983). Abundant evidence in Aladaglar suggests that this efect can be greater than was previously thought. Our observations suggest that the bedrock thicknesses up to several tens of meters, includ-ing the entire epikarstic zone and large dolines, can be stripped away by the glacier action. Shafs entrances in valley bottoms and other low ar-eas are blocked and obscured by debris, the result both of plugging by debris during glaciations and intense post-glacial physical weathering. Most shaf entrances that remain open are found at the crests of ridges or topo-graphic eminences within valleys, such as roche moun-tonnées, - those places which were sites of intense glacial scour but have limited or no contemporary catchments to supply frost debris (Fig. 6). Te smoothed tops of some ridges at altitudes of 3100-3300 m, along with the pres-ence of polished surfaces and decapitated shafs, suggest that an icecap of some considerable thickness may have covered these ridges during the recent glaciation. when decapitating shaf openings, glacial erosion stripped the upper portion of the rock together with the epikarstic zone and some upper sections of pre-glacial cave systems. Discovery of a decapitated shaf entrance on the surface and likenhood of fnding an explorable cave beneath it depend on which particular component of a cave system was intersected and how the opening is situated in the relief (Fig. 7). If a large internal pit of fig. 7: Shafs decapitated by glacial erosion. lef panel - Aladaglar, right panel - Crowsnest Pass, Rocky mountains, Canada. Photos by A. Klimchouk. ACTA CARSOLOGICA 35/2 - 2006 117 ALEKSANDER KLIMCHOUK, SERDAR BAYARI, LüTFI NAZIK & KORAY TÖRK substantial diameter got exposed, it had little chance of remaining unplugged by frost shatter debris from weath-ering in the shaf mouth catchment. Hence, most shafs of this type are simple single pits blocked at the bottom. Some large shafs located in the tops of ridges are blocked with ice, which contains numerous bands of frost debris within it (e.g. there is about 100m in the Ice Cave). Com-plex and deep caves are commonly those which got ex-posed by stripping at the level of a narrow meander pas-sage that continues to a next vertical pit. UNwALLED SHAFTS In the steep to vertical slopes of the Aladaglar glacial valleys, many unwalled or partially unwalled shafs are clearly displayed. we illustrate this with examples from the particularly deep Hacer glacial valley (Figs 8 and 9) and an example from the smaller and shallower Kemikli valley (Fig. 10). only a few to 20 m thick. It is apparent that this cave will become unwalled in the near geological future. Fig. 10 shows an unwalled shaf in the southern clif of the Kemikli valley. Two other shafs, partly unwalled in the upper parts, are seen in the right background. Te presence of a pile of boulders beneath this cluster of un- fig. 8: Exposed caves in the southern side (northern face) of the hacer glacial valley. Te vertical extent of the clifs in the photographs is approx. 1000 m. Numbers indicate samples of: 1 = unwalled shafs, 2 = partly unwalled shafs, 3 = unroofed meanders with shaf openings. many similar features, recognisable on photos, are not indicated in order to avoid clutter. Fig. 8 shows the southern side of the central sector of the Hacer valley, where it has the maximum cross-sectional vertical extent of about 1700 m. Individual unwalled shafs more than 100 m in the vertical extent can be seen. Some shafs are partly unwalled, while others are open only at the upper or lower ends within inclined and hanging faces. On inclined faces some unroofed mean-dering passages, interspersed with shafs, can be traced for hundreds of meters. Fig. 9 shows the clif face in the upper Hacer valley, where there is a cave that is almost parallel to the exter-nal face. Te wall that separates the shaf from the clif is walled shafs (some of them with fragments of shaf solu-tion morphology) indicates that the unwalling occurred as a rockfall, and that this event post-dated the last glaci-ation. See also Fig. 6, where a fall of boulder-sized blocks is well seen on the lef, scattered on the scree apron at the foot of the clif and on the glacially scoured rock surface below it. If the shaf unwalling had occurred before or during the last glaciation, boulders would have been be removed by the glacier. 118 ACTA CARSOLOGICA 35/2 – 2006 GLACIAL DESTRUCTION OF CAVE SYSTEMS IN HIGH MOUNTAINS, wITH A SPECIAL REFERENCE TO THE ALADAGLAR ... fig. 9: Right: Exposed caves in the rock clif in the upper part of hacer glacial valley (northern facing). Numbers indicate: 1 = unwalled shafs, 2 = partly unwalled shafs. Black vertical arrows point to shaf openings, white vertical arrows indicate lower (downward-open) shaf outlets, a horizontal arrow indicates a cave opening. lef: Te profle of the Professors Cave, with the shaf entrance located some 10 m far from the drop and the lower outlet opened to the clif. Te wall between the shaf and the vertical rock face is 3-20 m wide. fig. 10: Unwalled shafs in the Kemikli valley. Photo by A. Klimchouk. ACTA CARSOLOGICA 35/2 – 2006 119 ALEKSANDER KLIMCHOUK, SERDAR BAYARI, LüTFI NAZIK & KORAY TÖRK DISCUSSION AND CONCLUSION Te creation of a considerable vertical component during the development of relief in highly energetic Mountain set-tings leads to the vertical dissection of previously formed cave systems. It is an important part of the cave disintegra-tion process. Cave (passage) openings in sub-vertical sur-faces (the holes-in-the-wall) are common and well-known features. Shafs unwalled in sub-vertical surfaces are the less acknowledged phenomena, but they are common and easily recognisable in the Aladaglar massif Although fuvial entrenchment can create very steep surfaces on valley sides, it usually cannot directly erode away any intercepted shafs to make them unwalled. Unless the stream is very large, it will be captured and channeled down the shaf, rather than dissecting it. Instead, by creat-ing a steep relief, fuvial erosion may induce gravitational rock falls and slides, which then can expose the shafs, un-walling them on sub-vertical surfaces. More commonly, however, fuvial downcutting exposes cave openings by intersecting passages that are sub-horizontal. Direct glacial erosion, applied to those parts of val-ley slopes that were in contact with ice, can be the shaf-unwalling agency. However, in Aladaglar unwalled shafs are also found on the higher sections of slopes that ap-parently were not in a direct contact with valley glaciers that produce the most of the lateral erosion (i.e. they are above the glacial trimlines). Gravitational processes, chiefy rock falls and slides, thus are the dominant processes in shaf unwalling. Glacial erosion results in destabilization of sub-vertical slopes due to both additional downcutting and lateral undercutting in valleys. Both the glacial erosion and the glacial load cause considerable rearrangement of the strain feld. when the ice recedes its support of the clif face is removed. Glacial rebound and stress release afer the ice removal further contribute to the clif destabilization. As a result, a clif may experience one major topple, fall or slide as a consequence of a particular glaciation. Tose shafs that turned to be near sub-vertical external faces, were readily unwalled soon afer the last glaciation. Subsequently there may be further falls, etc. but they are usually much lesser in scale. Although the mechanism of the shaf unwalling is quite obvious, it remains unclear why unwalled shafs are more abundant in Aladaglar, as compared to most of Alpine karst massifs and many other formerly glaci-ated mountain karsts. One of the reasons could be the diference in the rates of the surface processes in diferent climatic settings and at diferent altitudes (2800-3700 m in Aladaglar versus 1800-2700 m in most of the typical Apline karsts). Unwalled shafs are not long-living surface features. Most likely, their morphology gets reworked fast by denudation agencies, giving rise to various kinds of grooves and small gorges in steep to sub-vertical slopes. Terefore, the diference in the amount of time since the ice receded in various regions could be another reason. Te relatively large number of unwalled shafs in Aladaglar can probably be explained by the combination of both these reasons: very recent end of the last glaciation (Early Holocene) and slower rates of remodelling of the face morphologies since that due to the high altitudes. Other massifs of the comparable height and climate con-ditions may simply have a lower degree of pre-glacial karstifcation. And, eventually, the diferences in the up-lif rates between regions may also play a role. Abundance of decapitated and unwalled shafs in Aladaglar clearly suggests that extensive and deep cave systems were already well developed there before major glaciations commenced. Te most pronounced and immense efect of mountain glaciations on karst is the destruction of functional relationship between the relief and the karst system and glacial dissection of the karst system (Ford, 1983). Tis is afected through erasure of the epikarstic zone and upper parts of cave systems on sub-horizontal surfaces (decapi-tation of shafs) and vertical dissection of cave systems by overdeepened valleys (unwalling of shafs). ACKNOwLEDGEMENTS Tis study was supported by the General Directorate of greatly benefted from comments and editing by Derek Mineral Research and Exploration of Turkey (MTA) and Ford. Our thanks are also due to cavers from the Ukr.S.A the Ukrainian Speleological Association. Te feld sea- and Hacettepe University Speleological Club for their sons in 2004 and 2005 were supported by the expedition dedicated eforts on exploring and documenting the Ala- grants of the National Geographic Society. Tis paper has daglar caves. 120 ACTA CARSOLOGICA 35/2 – 2006 GLACIAL DESTRUCTION OF CAVE SYSTEMS IN HIGH MOUNTAINS, wITH A SPECIAL REFERENCE TO THE ALADAGLAR ... REFERENCES Bayari, S, Zreda, M., Ciner, A., Nazik, L., Törk, K., Ozy-urt, N., Klimchouk, A. & Sarikaya, A, 2003: Te Extent of Pleistocene ice cap, glacial deposits and glaciokarst in the Aladaglar Massif: Central Taurids Range, Southern Turkey. XVI INqUA Congress, Paper #55360, XVI INqUA Congress, Reno Nevada USA, 23-30 July 2003, Abstracts, 144. Gosse, J.C. & Fillips, F.M., 2001: Terrestrial in situ cos-mogenic nuclides: theory and application.- quater-nary Science Reviews, 20, 1475-1560. Ford, D., 1983: Efects of glaciations upon karst aquifers in Canada.- Journal of Hydrology, 61, 149-158. Ford D.C. & williams P.w., 1989: Karst Geomorphology and Hydrology London: Unwin Hyman, 600 pp. Mais, K., 1999: Roofess Caves, a polygenetic status of cave development with special references to cave regions in the Eastern Calcareous Alps in Salzburg and Central Alps, Austria.- Acta Carsologica, 28, 2, 145-158, Ljubljana. Klimchouk, A.B., 2004: Towards defning, delimiting and classifying epikarst: Its origin, processes and variants of geomorphic evolution. In: Jones, w.K, Culver, D.C. and Herman, J. (Eds.). Epikarst. Proc. of the symposium held October 1 through 4, 2003 Sheperdstown, west Virginia, USA. Karst water Institute special publication 9, 23-35. (republished in: Speleogenesis and Evolution of Karst Aquifers (www.speleogenesis.info), 2004, 2, 1. Klimchouk, A.B., 2005: Cave unroofng as a large-scale geomorphic process. Cave & Karst Science 32, 2-3, 93-98. Mihevc, A., 1996: Te cave Brezstropa jama near Povir (in Slovene, English abstract).- Nase jame, 38, 65 -75, Ljubljana. Mihevc, A., Slabe, T. & Šebela, S., 1998: Denuded caves -an inherited element in the karst morphology: the case from Kras.- Acta Carsologica, 27, 1, 165-174, Ljubljana. Zreda, M., Ciner, A., Bayari, S. & Sarikaya, A., 2005: Remarkably extensive early Holocene glaciation in Turkey. EGU 2005, 24-29 April 2005 Vienna, Geo-physical Research Abstracts, v. 7 (06068). ACTA CARSOLOGICA 35/2 – 2006 121