Genesis and Evolution of Karst

Dusan Gavrilovic

    There are five morphostructural and geotectonic units in Serbia: the Dinaric Mountains, the Carpatho-Balkan Mountains, the Serbo-Macedonian Belt of horsts and rift-valleys, and the Pannonian and Pontian Basins. In the present-day relief, carbonate rocks are more or less widespread only in the two mountain systems mentioned. In the zone of the Serbo-Macedonian Belt and in the Pannonian Basin, they occur sporadically or are found under a thick covering of Tertiary sediments.
    The oldest carbonate rocks have to date been recorded in the Serbo-Macedonian Belt and at the foot of the Carpatho-Balkan Mountains. They occur for the most part in the form of intercalations (up to 20 m thick) and large lenses of marble and metamorphized limestones inserted in gneisses and chlorite schists. Apart from calcite and dolomite, they often contain quartz and other minerals. On the basis of paleontological investigations, they were preliminarily concluded to be of Proterozoic age. The largest complex of older Proterozic marbles (11 km²) was discovered near the town of Prokuplje in Southern Serbia. These marbles arose from shallow-water marine limestones and are now collectors of underground water, that is to say the karst process is expressed in them.
    Primarily represented by marbles and orthomarbles, carbonate rocks of the Cambrian, Ordovician, and Silurian have been discovered at the base of the Carpatho-Balkan Mountains, in the Serbo-Macedonian Belt, and at several localities in the Inner Dinarides. They occur in the form of smaller lenses and intercalations in schists and are fairly impure, without karst phenomena. Dolomites and limestones of Devonian age have been discovered northwest of the West Serbian city of Valjevo, but they still are not distinct from the Carboniferous. Little is known about this carbonate formation because it is mostly covered by younger sediments.
    Thus, carbonate rocks of Precambrian and older Paleozoic age represent a local phenomenon. They are of little thickness, impure, enclosed in impermeable rocks, and karstified in exceptionally rare cases.
The carbonate complex of the Dinarides rests on an older Paleozoic base. The Carboniferous series is primarily constructed of clayey schists, sandstones, and conglomerates, with larger and smaller insertions of limestones and dolomites, in which karst phenomena are encountered. Permian deposits are likewise in large measure composed of clastites, and west of Valjevo are represented by bituminous limestones.
    Triassic limestones, for the most part dolomitic, are very widely disseminated in the Inner and Central Dinarides. The Triassic carbonate series is discordantly deposited over the Paleozoic and is fairly heterogeneous. In the Central Dinarides, it is covered by extensive layers of Jurassic diabase-cherts, serpentinites, and amphibolites, whose thickness ranges from several tens to hundreds of meters. In most cases, Cretaceous limestones are deposited over serpentinites. In view of the position of serpentine covers within the Mesozoic carbonate series, it is certain that they played a very significant role in development of the karst process, both in upper and lower parts of the limestone. In the Inner and Central Dinarides, the Mesozoic carbonate series is covered by Senonian flysch.
    During the Paleogene, this entire area was dry land, which is to say that the possibility existed for development of the karst process. By action of the Laramide, Pyrenean, and Savic Orogenies, the Mesozoic sediment series was lifted and vaulted in the form of extensive monoclines, anticlines, and anticlinoria, in which horsts and graben occur as lower structural units. Due to strong lateral thrusts, nappes were formed in several places. The formation of deep graben where basins later arose and the covering of limestones and dolomites by impermeable rocks exerted great influence on the morphological and hydrological evolution of karst. The higher a moutain range was lifted, the deeper the karst process penetrated the carbonate mass, so that karst was successively developed already from the Upper Cretaceous or more than 70 million years ago.
    In the Carpatho-Balkan Mountains, karst is developed in Mesozoic carbonate rocks, primarily Cretaceous and Jurassic limestones and dolomites. The Mesozoic carbonate series lies over Paleozoic schists and sandstones. Lower parts of the series most often consist of Jurassic dolomites, which pass over first into bedded, then into massive limestones. The Mesozoic phase of sedimentation was completed at the end of the Cretaceous with the deposition of Senonian flysch. This was followed by the Paleogenic dry land phase, which was characterized by intensive tectonic movements, volcanism, and erosion.
    The karst is developed in smaller isolated areas of carbonate rocks, between which are found basins, river valleys, and terrains composed of impermeable rocks. Although thickness of the carbonate series is most often 100-300 m, there are places where it is more than 700 m thick. Besides being broken up in the horizontal direction, it is not uniform in the vertical direction either. Intercalations of marls and sandstones occur between pure limestones and dolomites. Veins of magmatic rocks very often penetrate limestones. On the other hand, the phenomenon of nappe formation is expressed along virtually the entire western rim of the Carpatho-Balkan Mountains. Nappes of Triassic and Jurassic limestones of the ”western belt” were formed over Permian-Triassic red sandstones, and then together with them over Miocene lacustrine sediments and Cretaceous limestones in the east. Thus, a broad zone of impermeable rocks is wedged between the western and eastern limestone belts. All of this has caused terrains of carbonate rocks in the Carpatho-Balkan Mountains to take on the characteristics of fluviokarst.
    Major marine-lacustrine transgressions from the Pannonian and Pontian Basins occurred at the beginning of the Miocene. Following tectonic depressions, they extended far to the south, flooding lower parts of the relief in the Carpatho-Balkan Mountains and the Inner and Outer Dinarides. In the course of the Miocene and Lower Pliocene, marine-lacustrine transgressions occurred again several times, leaving thick layers of sediments behind them. Thickness of the Neogene sedimentation series constitutes 100-400 m on the bottom of certain basins, while it achieves a value of 800-900 m in the Morava Depression. Considerable areas of karst were also covered by Neogene sediments. Lakes disappeared at the end of the Levantian in the Pannonian and Metohija Basins and already during the Pontian in other regions. Weakly expressed karst relief is also occasionally encountered in Miocene marls and clayey limestones, which are most widespread south of Belgrade.
    Even though conditions favorable for karst development were created already in the Proterozoic, there is no evidence to date indicating the existence of karst phenomena older than Carboniferous. If they were expressed at all, Precarboniferous karst phenomena became difficult to recognize due to subsequent mineralogical changes in carbonate rocks.
    According to currently available data, a fossil cave discovered north of the small town of Krupanj in Western Serbia could have the greatest age. It consists of a vertical shaft formed by corrosion in Middle Carboniferous limestones that was entirely filled with antimony ore. A pothole 5-10 m wide and 30 m deep appeared during exploitation of the ore deposit. This shaft is believed to have arisen through karstic erosion during the Upper Carboniferous and Lower Permian (Djurickovic A., 1982). However, in view of the fact that the orebody is of Oligo-Miocene age, the indicated shaft could also be of the same age. It follows that age of the shaft has not been established for sure.
    Numerous cave galleries almost entirely filled with terra rosa were discovered in the course of exploitation of marble on Vencac Mountain near the town of Arandjelovac. One such “closed cave” was surveyed for a length of 57 m (Jankovic M., 1997). Because the marble deposit is intensively exploited, these fossil cave galleries appear suddenly after blasting, but are also quickly destroyed. The marbles were initially dated as Paleozoic (Brkovic T. et al., 1980), which made it seem possible that the caves are of the same age. However, it is much more likely that the marbles appeared during the intrusion of granite of the nearby Bukulja Mountain. It was established by the strontium method that Bukulja granite is 13-20 million years old, which is to say that its crystallization occurred during the Miocene (Deleon G., 1969). The indicated cave galleries could thereby be of Miocene and Pliocene age.
    During exploratory drilling in the southern part of the Pannonian Basin, greater accumulations of underground water were discovered in cavernous Triassic limestones at two locations. They both have the nature of karst artesian aquifers. The first is located south of the mountain of Fruska Gora near the bottom of the Vrdnik Coal Basin. Below the Lower Miocene coal-bearing sediment series, at a depth of 330 m, a borehole located in the middle of the basin struck Lower Triassic limestones, from which water began to gush under pressure. Drilling through the limestones was continued for a distance of only 10 m, to a depth of approximately 340 m, that is to a point lying 120 m below sea level. The perforated limestones were highly cavernous, so that the bit in the course of drilling through them sometimes fell by as much as a whole meter at a time (Milojevic N. et al., 1968). In the Valjevo-Mionica Tertiary Basin, cavernous water-bearing Triassic limestones were likewise found below the impermeable Miocene sediment series. When a borehole near the village of Mionica had extended to a depth of 138 m, it reached Triassic limestones and water gushed from it under pressure (Mijatovic B., 1983). In our opinion, the caverns in Triassic limestones in both cases arose during the Paleogenic karst phase. Water in the fissure systems remained dormant under pressure until mine shafts and exploratory boreholes reached the Tertiary limestone floor.
    On the basis of exploratory boreholes, Paleogenic karst relief has been established along the western rim of Beljanica and Kucaj Mountains in the Carpatho-Balkan system. The paleorelief is developed in Cretaceous limestones and is covered by coal-bearing lacustrine sediments, whose thickness is in places greater than 400 m. Dolines, uvalas, and karstified valleys occur in the paleorelief. Two uvalas near the Resavica Mine are about 1 km wide and 100 m deep (Maksimovic B., 1956). Similar karst landforms of Paleogenic age were discovered during mining operations in the Jasenovac-Bliznak coal-bearing zone, on the western side of the Krepoljin-Krupaja Basin (Miljkovic Lj., 1986).
    Near the spring Krupacko Vrelo in the basin Belopalanacka Kotlina water-filled cavities were perforated in Cretaceous limestones at depths to 150 m. In the latest attempt, divers in the spring itself descended to a depth of 82 m or about 78 m below the Ni{ava River bed. The opinion has been expressed that the fissure system of the spring was formed already in the Paleogene, before the deposition of lacustrine Neogene sediments, which in the basin Belopalana~ka Kotlina reach a level of up to 350 m above the Nisava River bed. During the lacustrine phase, the underground stream that built the cave system probably exsurged as a sublacustrine spring, but at a height significantly greater than that of the present source (Petrovic D. et al., 1997). The assertion that “deep karst” is of Paleogenic age is difficult to prove because the fissure systems, regardless of the depth at which they are found, are fed by present-day surface waters. In the spring Zagubicko Vrelo, divers descended to a depth of 73 m, while Neogenic lacustrine sediments are encountered on the walls of the basin at levels up to 300 m above the spring. However, it was irrefutably established on the basis of the epigenetic gorge of the Velika Tisnica that all forms of relief higher than 600 m a.s.l. - including the Suvi Do and Tisnica valleys, which feed Zagubicko Vrelo - arose after the retreat of lakes from the basin Zagubicka Kotlina.
    Probably one of the most unusual karst landforms is the extensive cave system discovered in the Stari Trg Mine on the southern slopes of Kopaonik Mountain. In exploration for lead and zinc ores, galleries of the cave system have been surveyed from 730 to 135 m a.s.l. They were formed through the action of thermomineral waters in orthomarbles near their contact with schists. In addition to vertical channels, this system also contains horizontal galleries and halls measuring several tens of meters in diameter (Petrovic J., 1969). Prior to mining operations, all galleries lying below the bottom of the Ibar River Valley were constantly filled with water. Inasmuch as the potassium-argon method established a Middle Eocene age for Kopaonik granites (Deleon G. et al., 1961), it can be asserted with a fair degree of certainty that the indicated cave system arose in the younger Paleogene.
    Following the Savic orogenic phase in the Oligo-Miocene, when initial features of the present-day relief of Serbia were formed, karst proceeded to evolve more or less continuously throughout the entire Tertiary and Quaternary. Short breaks in the development of karst relief occurred only during lacustrine transgressions, which, in addition to the Pannonian Basin, encompassed all larger other ones as well. Although it is probable that fossil landforms have been preserved to the greatest extent among caves, it is difficult to prove their complete fossil nature in most cases. The situation is even more complicated in the case of surface landforms. Here it is sooner possible to speak of relict phenomena in the present-day relief than about fossil landforms. All recent karst of Serbia is a complex of Tertiary and Quaternary landforms. Traces of paleokarst of Neogene age have been discovered on the western side of Beljanica Mountain. Occurring under lacustrine sediments with bands of coal, relief consisting of dolines and conical limestone hills similar to “tropical karst” was exhumed here. The opinion has been expressed that this relief was formed in the Upper Miocene, under conditions of a humid and warm climate (Gavrilovic D., 1970). However, two potholes, 31 and 37 m deep, with no lacustrine sediments have been surveyed precisely in this region, which means that they are significantly younger. The potholes were probably formed during the Quaternary, that is after the exhumation of Miocene karst relief.
    Occurrences of paleokarst in several places have also been established in Sarmatian limestones south of Belgrade. The karst relief of the Pliocene and older Pleistocene is covered by loess deposits, originally more than 10 m thick. During the Holocene, the loess cover eroded to a considerable extent and became much thinner, so that the karst process was reactivated. Nevertheless, there are a fairly great number of dolines and vertical channels that have lost their former function and no longer have any expression on the topographic surface (Gavrilovic D., 1985).
    The question as the age of karst relief is very complicated because it was created parallel with the evolution of fluvial relief, which has still been inadequately studied. In determining the age of fluvial landforms, it is usual to proceed from the position and age of Neogene sediments in the relief, which are most often dated on the basis of paleontological finds. Just as geologists change their minds about the age of sediments and duration of the Miocene and Pliocene (Stevanovic P., 1988), so geomorphologists change theirs about the age of landforms in the relief. Already from the time of Cvijic (1922) and Jankovic (1909), it was believed that valleys in Serbia were for the most part downcut and that river terraces reflect brief interruptions in the deepening of valleys, that is to say only the superposition method was used in dating surfaces and river terraces. For all of these reasons, the age of karst relief can be determined only approximately, often with an error of several millions of years.
    In the valley of the Beli Timok, the reference point for determination of karst age is the epigenetically downcut Vratarnica Gorge. The opinion has been expressed that the surrounding limestone terrains were covered by Miocene lacustrine sediments up to 360 m a.s.l. or 200 m above the bed of the Beli Timok (Petrovic D., 1954). During removal of lacustrine sediments in the Pleistocene, an extensive valley pediment was formed whose parts now lie approximately 150 m above the river bed. In the course of further downcutting of the valley, karst landforms began to arise, initially on the pediment and then under it. Thus, the karst here is young, since karst landforms begin to arise only at the transition between the Pliocene and the Quaternary. Valley meanders in the gorge of the Svrljiski Timok River indicate that karst landforms here also began to arise only after the disappearance of a Miocene lake in the basin Knja`eva~ka Kotlina. It is interesting to note that on Tupi`nica Mountain as well, whose highest parts lie above 1000 m a.s.l., there are no landforms older than the Pliocene in the surface relief (Zeremski M., 1994).
    Due to little study and the broken-up nature of Neogene and Quaternary sediments in the basin Pirotska Kotlina, the age of karst on Vidlic and Tepos Mountains can be only generally determined. Although Miocene and Pliocene sediments were previously singled out in  the Odorovacko Polje (Milakovic B., 1967), there are no Miocene sediments at all on the “Pirot” sheet of the Basic Geological Map (1976). Around the rim of the basin Pirotska Kotlina, Pliocene gravels and sands occur up to 600 m a.s.l. or about 250 m of relative elevation. This approximately corresponds to the “middle lake level” in the Nisava Valley, which is of Pontian age (Jankovic P., 1909). Since the highest streamsinks in the Odorovacko Polje are found at of 730 m a.s.l., the bottom of the valley of the surface outlet must have been even higher, at approximately 800 m a.s.l. The most realistic hypothesis is that this old stream began on the northwestern margin of the polje and emptied into Lake Pirot east of the present-day Krupac Marsh. Downcutting of the outlet valley began on the level of a plateau at 900 m a.s.l., which is considered to belong to the “upper lake level” and to be probably of Upper Miocene age. Through later break-up of the relief under the influence of fluvial and karstic erosion, a “plateau of summits” consisting of conical limestone hills was left of this plateau. In view of the elevation at which downcutting of the Odorovac Lake outlet began, it is difficult to postulate that the karst landforms on Vidlic Mountain are older than those on Tepos Mountain. Thus, the oldest karst landforms on Vidlic and Tepos Mountains could be of Upper Miocene age, but they cannot be recognized and singled out in the present-day relief.
    On the mountains of Serbia, Pleistocene glaciation was of a local nature and did not significantly affect development of the karst process. On the proportionally small areas that were covered by ice, the karst process was slowed or curtailed, but it underwent intensification in the periglacial zone around the rim. Many dolines and uvalas were filled by glacial, glacio-fluvial, solifluction, and deluvial-proluvial sediments. In the majority of cases, meanwhile, the karst process was reactivated in the postglacial period, so that true fossil landforms are few.
    Of mountains higher than 2,000 m, the karst of the Mokra Gora Highlands and Zljeb Mountain-between the Ibar Valley and the Metohija Basin-has been studied to the greatest extent. Theoretically, conditions for development of the karst process existed here already in the Cretaceous, although only one cave filled with Miocene lacustrine sediments has been found up to now. Regardless of elevation above sea level, all surface karst landforms are of Pliocene and Quaternary age and were created after the retreat of lakes from the Metohija and Tutin Basins (Menkovic Lj., 1995).
    The direction of the karst process gradually changes under the influence of powerful differential tectonic movements that began at the end of the Cretaceous and grew constantly stronger. It increasingly takes on a vertical dimension and penetrates ever deeper into the carbonate mass. Surface corrosion and subsoil corrosion gradually lose their earlier significance, and deep-seated corrosion becomes dominant. The size of underground cavities steadily increases, and the focus of the corrosive action of water shifts over into them. Thus, the transition from a Mesozoic tropical peneplain through Miocene cone karst to Pliocene and Quaternary deep karst, although accompanied by climatic changes, was primarily dictated by tectonic denivellation of the relief.
    Since several horizons of paleokarst of varying age exist in structure of the Dinarides and Carpatho-Balkanides, the question can be posed as to the direction in which karstification and corrosive expansion of fissure systems advanced here. Was it only from the surface to the depths of the carbonate mass, or was it from above and below at the same time? Although conclusions as to the genesis and nature of karst are drawn for the most part on the basis of its surface and subsurface manifestations, the numerous instances of hydrothermal ore deposits and thermal waters must not be disregarded. Springs giving thermomineral waters of vadose origin are fairly frequent in the karst of Serbia, and many function for a considerably longer time than the majority of ordinary karst springs. Thus, for example, springs in the spa of Ni{ka Banja have already given water continuously for at least 27,000 years (Vucic A. et al., 1958). During this time, they have yielded more than 4.8 million cubic meters of limestone dissolved in water, of which a cube could be made with sides 364 m long.
    Study of the age and evolution of karst relief is unquestionably a very complex task involving many difficulties. At the same time, however, it opens up new possibilities for visualizing the essence of the karst phenomenon as a whole. It is probable that future investigations will significantly alter existing ideas about the genesis and age of landforms in the karst relief of Serbia.