ACTA CARSOLOGICA	XXVII/2	2	25-40	LJUBLJANA 1998
COBISS: 1.08
SPELEOMORPHOLOGY OF DRY PASSAGE IN PROVALA CAVE (CROATIA)
MORFOLOŠKE ZNAČILNOSTI "SUHEGA ROVA" V JAMI PROVALA (HRVAŠKA)
NEVEN BOČIC1 & NENAD BUZJAK2
1	Speleological society Karlovac, Kurelčeva 3, 47000 KARLOVAC, CROATIA
2	Speleological society Dinaridi, Marulicev trg 19/II, 10000 ZAGREB, CROATIA
Prejeto / received: 3. 11. 1998
Izvleček	UDK: 551.44 (497.5)
Neven Bočič & Nenad Buzjak: Morfološke značilnosti "Suhega rova" v jami Provala (Hrvaška)
Jama Provala, 1687 m dolga in 55 m globoka, razvita v dveh nivojih, se odpira v fluviokraškem svetu Zumberačke gore (SZ Hrvaška). Skozi spodnji nivo teče stalni vodni tok. Zgornji nivo je, glede na vodne značilnosti, mogoče razdeliti na tri dele: 1. južni s stalnim vodnim tokom, 2. del z občasnim vodnim tokom in 3. suhi del. V zadnjih dveh delih, ki se imenujeta Suhi rov, so bile opravljene morfološke raziskave. Ključne besede: speleomorfologija, jamska terasa, slepi nivo, polslepi nivo, Provala, Suhi rov, Zumberačka gora, Hrvaška.
Abstract	UDC 551.44 (497.5)
Neven Bočič & Nenad Buzjak: Speleomorphology of dry passage in Provala cave (Croatia)
Provala cave is located in the fluviokarst area of Zumberačka gora mountain (NW Croatia). This two-level cave is 1687 m long and 55 m deep. In the lower level there is a permanent water stream. The upper level, according to its hydrological characteristics, can be divided into three parts: 1. the southern part with permanent water stream, 2. the part with periodical water stream, and 3. dry part. The last two parts called "Dry passage" were speleomorphologically observed.
Key words: speleomorphology, cave terrace, blind cave level, half-blind cave level, Provala.
INTRODUCTION
Zumberačka gora mountain massif is a wide hilly area between the Sava and Kupa rivers and the border of Slovenia in NW Croatia (Fig. 1). From the highest NW part with its highest peak, Sv. Gera or Trdinov vrh (1178 m), its intersected relief is steeply inclined towards the Kupa and Sava river valleys, deeply incised by numerous creeks and small river valleys.
The whole area is characterized by fluviokarst relief, which mostly developed in Upper Triassic dolomite, Cretaceous carbonate beds (limestone and breccia) and Miocene Litothamnium limestone. There are numerous karst relief forms (dolines, uvalas, poljes, blind and dry valleys, caves, shafts) and hydrogeological features (maze-like subsurface circulation with ponors and karst springs). But, due to small thickness of karstifiable rocks the depth of karstification is also small. Because of
Fig. 1: Position map of Zumberačka gora mountain and Provala cave.
Fig. 2: The parts of Dry passage (the basic ground-plan and profile, except their contents and Fig. 5 and 7, were surveyed by SOPDJ-Samobor).
wide extension of younger impermeable cover which covers mentioned beds, branching surface drainage also occurs.
@umbera~ka gora has temperate warm humid climate with an annual air temperature between 8°-10°C and annual precipitation between 1000-1400 mm ([egota 1974). In higher parts it is cov-ered by mountain beech forest and in lower parts by oak and hornbeam forest (Rogli} 1974).
Except forests, there is a little of other natural resources. Due to that fact and also because of historical and other reasons (problems with water, lack of fertile soil, traffic isolation etc.) most of the area is of low population density.
Provala cave is located near the village of Bu~ari in the area of Donji O{trc, the central part of @umbera~ka gora mountain. Its entrance is located in the wide depression, at the end of a dry blind valley. It was found by cavers from SOPD Japeti}-Samobor in 1995, who carried out most of exploration. The cavers from SOPD Dubovac-Karlovac and SOPD Velebit-Zagreb participated at exploration also. Since the entrance was completely filled by rocks, the first step of exploration was digging. More than 20 longer or shorter survey excursions resulted with 1687 m of surveyed passages of total depth of 55 m. About 200 m of a very narrow passages are still waiting to be surveyed.
In this paper we present the results of speleomorphological research made in the upper level part of the cave, called Dry passage. We decided to process only this part because it is most representative in speleomorphological sense - it contains plenty of data valid for making conclusions about cave speleogenesis. So, we tried to collect as more data as possible to explain its origin, development and morphology. Because we had to observe this passage as a part of the whole cave, relevant data about the rest of the cave were also considered. In continuation of our work we will try to give the answer to question about origin and development of complete cave as a part of karstifi-cation process of a wider area.
GEOLOGICAL SITUATION
According to Milan Herak's geotectonic classification of Dinarides (1991), @umbera~ka gora mountain is an area at the contact between structural complexes of the Inner Dinarids (Supradinari-cum) and Dinaric carbonate platform (Dinaricum). It is characterized by a overthrust structure, where dominantly Upper Triassic dolomite (as a part of Supradinaricum) overlie Dinaricum which consists of Jurassic and Lower Cretaceous limestones partially covered by marginal flysch depos-its. Later tectonic movements divided overlaying beds into smaller tectonic units.
Bu~ari village area (and Provala cave) is located at above mentioned Supradinaricum-Dinari-cum contact. The base consists of Malmian (Oxfordian-Kimmeridgian) poorly or non-stratified oolitic limestone. By transgression it was covered by Upper Crecateous beds - Senonian flysch and in some parts breccia (Herak & Bukovac 1988; Pleni~ar & Premru 1977). These beds are in tectonic contact with Upper Triassic stromatolitic dolomites. These are thick to medium thick beds, very broken near the mentioned contact. As analyzed in the case of nearby tectonic window Duralije (Herak & Bukovac 1988), this contact is of overthrust type, as mentioned above.
In tectonic structure the most distinctive are Dinaric trending faults and joints (Pleni~ar & Premru 1977).
PROVALA CAVE BASIC DATA
At this moment Provala cave is the longest speleological feature in Zumberačka gora mountain. According to the morphological and hydrogeological classification of speleological features (Garasic 1991), it is a branching and multilevel cave consisting of one main passage and many narrow later-ally passages. The cave has two levels - the upper one which consists of active, periodically active and dry part, and the lower one with permanent water stream. The levels have 10 inter-connections.
Since the water stream enters and flows out through siphons, the cave was classified as a through-flowing cave. There are 7 mostly sieve-like siphons. The origin of water is uncertain. Probably the cave drains water from surrounding Donji Ostrc karst area, surface characterized by dry dolines, blind and dry valleys The water stream in the cave generally follows direction of the main passage (towards north) and probably rises as Jasevnica creek. This assumption is based on the fact that distance between the last cave siphon and Jasevnica spring is only 75 m and vertical difference is 3,5 m (T. Rubinic - oral communication).
SPELEOMORPHOLOGY OF DRY PASSAGE
According to morphological characteristics and hydrological conditions, the upper level of Provala cave can be divided into 3 parts (Fig. 2):
1.	the southern part with permanent water stream,
2.	the part with periodical water stream and
3.	the dry part.
The last two parts, called Dry passage, were observed. Its basic morphometric data are presented in Tab. 1.
Tab. 1: Dry passage - morphometric data
PART	LENGTH (m)
Per. active part	63,5
Dry part	42,5
DRY PASSAGE	106
WIDTH (m)	HEIGHT(m)
0,5 - 5	0,5 - 5,5
2 - 7	3,5 - 6
0,5 - 7	0,5 - 6
The periodically active part of Dry passage is a continuation of the cave permanent active upper level part. It is a simple, non-branching passage of generally N-S and E-W direction (about 50:50 ratio). It starts in a very low and narrow passage part (periodically siphon) and ends in a ponor-zone that consist of 2 ponors, which connect it with the lower active level. At this point the dry part starts. In morphological sense, it is more complicated than the last one. Due to the sharp changes of directions it has Z-like ground-plan. It is characterized by several remarkable side-wall levels (or terraces) which represent traces of older water levels or passage bottoms, which we called cave terraces. Due to the different mechanism of origin they should be distinguished from the flood-discharge terraces (Jakucs 1977).
The inclination of the passage bottom in whole part is subhorizontal. In the periodically active
part its mean value is -3° (in direction of water flow). In the dry part inclination is of similar value, but in some parts the bottom is, due to the debris and boulder piles, steeply inclined.
Geological composition and structure as factors of speleogenesis
Dry passage was formed in thick Malmian oolitic limestone. As mentioned above, it is very poorly stratified. Due to the fact that the passage walls are often covered by clay and flowstone or are heavily broken, probable, however rare bedding planes were not seen. According to size and dimensions of passages, which are not typical of this part of Croatian karst area (Božič 1971; Buzjak et all. 1996; Marjanac 1971), it is obvious that this limestone is a very good basis for speleogenesis and karstification in general. Except by its composition, it is controlled by secondary po-rosity due to tectonic movements observed in the whole cave.
The fractures determined the directions and cross-sections of the passage, appearance of some rocky relief forms and speleothems. Two faults (one of them was interpreted as reverse) and 8 major joints (diaclases) of different directions were noted. The ratio of their occurrence and directions is presented in Fig. 3.
%40
E-W NNW-SSE NNE-SSW NW-SE	N-S
direction
Fig. 3: The share of fractures directions in Dry passage.
Due to small number of specimens, the ratio between these fracture directions and passage directions data can not be validly statistically processed. But, if Fig. 3 is compared with speleomor-phological map (Fig. 4) and with the above mentioned data of Dry passage generally directions, it is obvious that the main role in the passage development had fault and joints of E-W directions. Al-though the number of N-S oriented fractures lag behind the number of NNW-SSE, NNE-SSW and NW-SE oriented fractures, the length of passage parts of N-S direction is bigger than the length of parts of other 3 directions. The reason for that lies in fact that in the 1/3 of the N-S oriented parts of Dry passage no joints of this or similar direction were noticed. They could be widened and rounded during the speleogenetic phases (Fig. 5, cross-sections 5 and 6) or filled and covered with sediments. The beginning part of passage is of typically phreatic cross-section and possibly determined by bedding plane (Fig. 5, cross-section 7).
In some parts of the passage, especially from point 2 (cross-section number on Fig. 4) towards the entrance shaft and ahead, along these fractures a larger amount of dripping water comes out, resulting in abundance of speleothems. In other parts of the passage (especially between points 4 and 7), fractures are stopped by fine sediment cover or less calcite and fine grained subaerial flow-stone coating (Hill & Forti 1997) somewhere at the ceiling.
The origin of the entrance shaft is determined by the fault of E-W direction.
Cave sediments
In whole passage man can find various types of sediments - both autochthonous and allochtho-nous (Kranjc 1989). The most widespread is clay. It mostly covers walls from point 4 to 7 (period-ically active part), up to 50 cm in thickness and up to 3,5 m high above the bottom (Fig. 5, cross-section 5). Its preservation at this position is due to the absence of dripping water and small discharge of periodical water stream. The clay also covers the bottom around points 2 and 3. At the point 2 it is intersected by desiccation cracks formed by drying of sediment after dragging and evaporation of water. In this part the clay is poorly preserved on the walls which are almost com-pletely washed out. Considering the presence and size of scallops, it is due to faster water stream, probably fast decrease of water level in past and larger amount of dripping water afterwards. In some parts the clay is preserved under flowstone (Fig. 4, big meander before point 2).
The pebble is mostly present along the periodically active part. It builts only a few centimeters thick bed on the bottom, but in places it can even be found 4 meters above it (in wall pockets or wider joints where it is preserved mixed with clay). Its grains are mostly up to 5 cm long, very flat and rounded. According to this characteristics, its beds are somewhere of imbricate structure. Among predominantly carbonate pebble, man can find non-carbonate, probably chert pebbles. This was transported from the southern part of upper level where chert is inserted into heavy folded carbonate beds. The bottom there is covered by mixed chert and limestone rock debris.
The rock debris and boulders are mostly present along fault planes and joint intersections. The biggest pile of rock debris is present along fault plane in the dry part (Fig. 4). Considering its ap-pearance and composition, it was deposited after retreat of water stream from this part. In distinction from it, smaller rock debris in periodically active part were washed out and only boulders covered by scallops remained (Fig. 4; point 3).
The speleothems deposit only in the meander at the older cave terrace (Fig. 4, before point 2),
Photo 2: Water leve1 cave terraces in the Dry part of Dry passage. The upper cover stands on the Ist generation cave terrace and the lower cover stands over Potior 3.
below heavy broken rock. In other parts of the passage there are no speleothems at all. Although the rocks are somewhere well jointed, there could be found only thin fine-grained subaerial flowstone coating. Probably the speleothems are absent because clay filled joints and fissures.
Cave rocky relief
Since the cave rocky relief is an important indicator of speleogenesis (Slabe 1995), it was also observed. Its distribution is shown on Fig. 6. It is mostly preserved and noticeable at the ceiling and at the washed out walls. At the ceiling, ceiling pockets predominate. They are concentrated at heavy broken rock surface, making in places a network of parallel features (Fig. 4; point 4). The pockets are mostly shallow and always elongated along the joints that control their origin.
The scallops are present between points 2 and 4, but at upper (older) water level horizons only.
Fig. 5: Selected cross-sections of Dry passage.
Their distribution (and distribution of there present fine-grained sediments) points to changes of velocity and amount (or discharge) of water which followed the changes of hydrological regimes; this will be discussed later.
Among other rocky relief features, water level horizons (wall notches) predominate (Photo 1). These are linear elongated semicircular or arch-like notches cut into walls or even ceiling by former water stream or even siphon lakes, which level they indicate (Ford & Williams 1989; Slabe 1995, 53-55). They could be find in whole Dry passage cut in the rock or in the clay deposited on the walls. Due to later erosion, corrosion or flowstone and speleothem deposition, their shapes were changed in places.
SPELEOGENESIS OF DRY PASSAGE
By analyses of the morphological characteristics of Dry passage (Fig. 4 and Fig. 5), several phases of its development were noted. They are due to lowering of the local base level (negative shift of Jasevnica spring). This process resulted in passage entrenchment leaving older bottoms dry and at higher level at passage sides as cave terraces (Phot. 2). It also resulted in steeplike passage profile due to the retrograde moving of water stream activity after openings of new ponors up-stream, and forming of the lower, the present active level. Often changes of water level were also followed by formation of many water level horizons described above.
The developement of Dry passage can be reconstructed as follows (Fig. 7).
The highest cave terrace found in the dry part is the oldest one (0 generation; Fig. 4, Fig. 7 and Phot. 1). In the periodically active part it is almost completely destroyed and preserved only in fragments. This cave terrace is a remain of a phreatic phase, during which the water flowed through today's highest dry level towards north. By opening of one of the ponors (0) in the continuation of the passage, it started to form today's lower level.
The opening of new ponor upstream (1), a cutting of new bottom started and the former one (0) remained dry and at higher level. That was the beginning of epiphreatic and vadose phase. Today it is visible as the 1st generation cave terrace. It is very good preserved and characterized by mean-ders, scallops and wall niches. The sediments are preserved only when covered by flowstone.
The next ponor opened by widening of fault plane (2). The previous bottom remained dry and new water stream incision started (the 2nd generation cave terrace). According to well preserved rocky relief (scallops especially) and sediments, there was probably rather fast change of hydrolog-ical conditions. From that presumption man can conclude that its activation was probably enforced by neotectonic movements. The inclination of ceiling behind Ponor 2 and its preserved rocky relief indicate that the rest of the passage upstream was submerged (siphon). The water level horizons and clay deposited at the walls point to changes of water level depending to inflow.
The opening of Ponor 3 (located also at the fault plane) resulted in further lowering of water level in the siphon lake which existed there. This change was followed by decreasing of water stream velocity and increasing of fine sediment deposition on the 2nd generation cave terrace. Such process is indicated by the inclination of bottom thick clay cover from Ponor 2 towards Ponor 3 (upstream direction), and by its surface intersected by desiccation cracks.
The last and today's periodically active Ponor 3a is at the almost same level as the last one (3). From the Ponor 3 it is divided by boulders of a fault origin. They may be an obstruction for water
stream, accelerating widening of joints and fault plane upstream till the activation of new ponor (3a). This ponor is active after heavy rain only when water pours out from the siphon upstream (in the active part of the upper level). In this way produced water stream is of very small discharge, indicated by its narrow and shallow bed (3rd generation water stream level) cut into well preserved older sediments and rocky passage bottom (without any rocky relief features characteristic for flowing water).
The water stream incision may be seen in fluvial sediments also (clay and pebble). The narrower parts of the passage are washed out (due to the faster water flow), but in the wider one (where water flow was slower) they are well preserved on the walls (where the clay dips steeply towards bottom) and in the bottom. In the periodically active part, the bottom fluvial sediment cover is different - the clay is washed out due to incision of water stream and replaced by pebble.
As already stated, the openings of ponors along the passage stream bed resulted in retrograde moving of water stream activity upstream. It resulted in the steeplike passage profile since in the earlier active passage parts behind new ponor incision stopped and they remained as higher passage level parts (Fig. 7). But, in the still active passage parts water stream incision (and passage en-trenchment) continued. Since this process is similar to the development of blind valleys on the karst surface (Sweeting 1973), the levels with ponors opened at their ends formed in this way are called blind passage levels. Since there is also the passage part with water flow only during high water periods (the part with periodical water stream together with the active part), we called it half-blind passage level. In Dry passage three such levels were formed: the oldest one between Ponors 1 and 2, the younger one between Ponors 2 and 3a and the youngest or half-blind passage level which ends at Ponor 3a. The first two levels are, due to the water stream incision only partially preseved, but they can be almost completely reconstructed by preserved cave terraces (Fig. 7).
Their ponors can be sieve-like and inaccessible for cavers or accessible connection to the lower cave levels.
Ëntrance
toward lower level
older water level (terrace)
water level horizons
cave karren
n
■CL « n

Bperlodlcally water stream
rz^Ti cross-section I- number
0
PROFILE
20m
Fig. 6: Distribution of cave rocky relief in Dry passage.
CONCLUSION
The research of speleomorphological characteristics of the upper level in Provala cave (Dry passage) shows that phases of its development were controlled by lowering of the local base level. This process influenced upon the changes in cave hydrological conditions (lowering of cave water stream level). This was connected with openings of new ponors along stream bed upstrems, result-ing in deepening of passage cross-sections (by incision of the water stream), forming the cave ter-races which represent old passage bottoms and steeplike passage profile.
According to the results of all above mentioned element analyses, the following phases in spe-leogenesis were established:
1.	phreatic phase - determined according to the highest preserved cave terrace in the dry part and its rocky relief,
2.	epiphreatic phase and vadose phase - noted in whole passage owing to well preserved water level horizons and 0, 1st and 2nd cave terraces incised by the water stream at different levels. It was caused by lowering of water level after activation of new ponors upstream, so former active part bottoms remained as above mentioned higher dry cave terraces. This process resulted also in steeplike passage profile due to forming of the blind and half-blind passage levels.
In the part with periodically active stream its incision is still in progress but it is, due to the small water discharge, very slow and not constant.
ACKNOWLEDGMENTS
For suggestions in interpretation of geological conditions in research area we wish to thank to Dr. S. Božicevic, Dr. J. Bukovac and Dr. I. Galovic (Institute of geology Zagreb) and for help in the field-work to T. Rubinic and I. Rasic (SOPD Japetic Samobor).
Fig. 7: Scheme of speleogenesis of Dry passage.
REFERENCES
Božic, V. 1971: Speleoloski objekti kanjona Slapnice. Nase planine, 9-10, 213-216, Zagreb
Buzjak, N., D. Perica, Z. Greguric 1996: Speleoloski objekti Samoborskog gorja. Zbornik radova 1. hrv. geogr. kongresa (Proceedings), 143-150, Zagreb
Ford, D. & Williams, P. 1989: Karst geomorphology and hydrology. 307-308, Chapman & Hall, London.
Herak, M. 1991: Dinaridi - mobilisticki osvrt na genezu i strukturu. Acta Geologica, 2, vol.21, 6769, Zagreb.
Herak, M. & Bukovac, J. 1988: Tektonsko okno Duralije u Zumberku. Geol. vjesnik, 41, 231-236, Zagreb.
Hill, C. & Forti, P. 1997: Cave minerals of the World. 70-71, NSS, Alabama.
Jakucs, L. 1977: Morphogenetics of karst regions. 217-220, Adam Hilger, Bristol.
Kranjc, A. 1989: Recent fluvial cave sediments, their origin and role in speleogenesis. Dela SAZU, 27, 21-22, Ljubljana.
Marjanac, S. 1972: Speleoloski objekti u plitkom krsu Zumberackog i Samoborskog gorja. Nase jame, 13 (1971), 79-83, Ljubljana.
Plenicar, M. & Premru, U. 1977: OGK 1:100000 i Tumac za list Novo Mesto. Sav. geol. zavod, Beograd.
Roglic, J. 1974: Biljni pokrov Sredisnje Hrvatske. In: Cvitanovic, A. (ed.) Geografija SR Hrvatske - knjiga 1: Sredisnja Hrvatska. 77-81, Skolska knjiga, Zagreb.
Slabe, T., 1995: Cave rocky relief. Zbirka ZRC, 10, 101-113, Ljubljana.
Sweeting, M. 1973: Karst landforms. 110-113, Columbia University Press, New York.
Segota, T. 1974: Klima Sredisnje Hrvatske. In: Cvitanovic, A. (ed.) Geografija SR Hrvatske -knjiga 1: Sredisnja Hrvatska. 61-67, Skolska knjiga, Zagreb.
MORFOLOGIJA SUHEGA ROVA V JAMI PROVALA (HRVAŠKA)
Povzetek
Jama Provala, 1687 m dolga in 55 m globoka, razvita v dveh nivojih, se odpira v fluviokra{kem svetu @umbera~ke gore (SZ Hrva{ka). Skozi spodnji nivo te~e stalni vodni tok. Zgornji nivo je, glede na vodne zna~ilnosti, mogo~e razdeliti na tri dele: 1. južni s stalnim vodnim tokom, 2. del z ob~asnim vodnim tokom in 3. suhi del. V zadnjih dveh delih, ki se imenujeta Suhi rov, so bile opravljene morfolo{ke raziskave.
Na podlagi teh raziskav je ugotovljeno, da so faze v razvoju Suhega rova povezane z nižanjem krajevne erozijske baze. To je vplivalo na vodne razmere v jami (zniževanje nivoja jamskega toka). S tem v zvezi je odpiranje novih ponorov ob strugi navzgor, zaradi ~esar se je nižalo dno rova (zaradi globinskega vrezovanja vodnega toka), nastajale so terase, ki predstavljajo nekdanje dno rova, in stopni~asti pre~ni prerez rova. Ugotovljeni sta bili slede~i speleogenetski fazi:
1.	Freati~na - predstavljena z najvišjo ohranjeno jamsko teraso v suhem delu in z njegovim skalnim reliefom.
2.	Epifreati~na in vadozna - je opazna vzdolž vsega rova zaradi dobro ohranjenih vodnih nivojev in zaradi ni~elne, prve in druge terase, ki jih je vrezal vodni tok v razli~nih nivojih. Temu je bil vzrok v nižanju nivoja vode, ki so ga povzro~ili novi ponori ob toku navzgor, tako da so nekdanji aktivni deli dna rova ostali kot že omenjene višje suhe jamske terase. Zaradi teh procesov je nastal tudi stopni~ast prerez rova, zaradi nastajanja slepih in polslepih rovov.
V ob~asno aktivnih delih rova se vrezovanje {e nadaljuje, vendar je zaradi majhnega pretoka zelo po~asno in z ob~asnimi prekinitvami.