Progress in 1996 (I)

1995-07-10KDL 4901




Marjorie M.Sweeting, Oxford 
School of Geography, Oxford University, England

The karstlands of S. E. Europe and of Southern China form the two largest and most varied areas of karst relief in the world. This contribution attempts to compare the development of the ideas of karst geomorphology in these two regions. Because of differing geological and geographical conditions, the emphases and explanations of karst phenomena have developed along quite different lines. Furthermore, Chinese attitudes to land use have led to a greater exploitation of karst areas than in Europe.

Because of the situation of Southern China within the S.E.Asian monsoon area, the Chinese karstlands are much more dissected by the great rivers and are more fluvio-karstic than the karstlands of Southern Europe. Karst studies in China have been much more integrated into general Chinese karst geomorphology than in Europe. Debates about karst water-levels and base-levels which formed a conspicuous part in the development of karst thinking in Europe, have been absent in China. Both regions have been subjected to neotectonism, but the proximity of the Chinese karst areas to Tibet and the Himalayas is probably of greater significance than the nearness of the classical karst and Dalmatia to the Alps. The Quaternary history of the two regions is quite different and is reflected in the legacy of the Quaternary period upon the landscapes. Chinese geomorphology has been less affected by debates about the importance of climatic geomorphology than in Europe. The opening up of the karstlands of China, together with the vast amount of work now being done by Chinese scientists, can lead to a useful reappraisal of the ideas and development of karst geomorphology in Europe and also in America.



Kohler, Heinz-Charles1, Auler, Augusto2, Catta nio, Mariab3
1 Museu de Historia Natural, Federal University of Minas Ger ais, B razil 
2 Geography and Geology Department, Western kentucky Un iversity, USA 
3 Federal University of Mato Grosso do sul, Brazil

The area is located in the eastern border of Brazilian Pantanal, 200 km away from Campo Grande, capital of Mato Grosso do Sul State. The karst occurs in folded and faulted calcitic and dolomitic limestone belonging to both Cuiaba and Corumba Formations of Upper Paleozoic age. The karst landforms develop in Serra d a Bodoquena(Bodoquena Hills) anticlinal and at Salobro and Formoso Rivers sincli nal, as a 100 km long stripe running norh-south.

Karst morphology is characterized by four distinct physiographical domains:A) Fluviokarst of Perdido River basin, developed in the high level karst plains, at an average altitude of 600m; B) Isolated hills 100 m high in average containing both solution and collapse sinkholes representing a residual form of domain A; C) Karst plain (at a 320 m topographical level) containing the rise of most of the rivers. Two types of springs can be distinguished, a diffuse discharge spring and a conduit flow spring; D) Fluviokarst of Formoso and Prata Rivers showing travertine deposition.

The underground karst is well developed throughout the area. The caves contain usually large descending passages that intercept the water table. Cave diving performed by the second author has shown that some underground lakes exceed 50 m in depth. The lake at Lago Azul cave show a very peculiar troglobitic fauna not found in other lakes. A slow movement of water was detected in this cave in contrast with the static water at other sites such as Anhumas Pit. The water level at Lago Azul cave lies 15 m below the regional base level measured at nearby rivers, suggesting a complex hydrogeological pattern. Cave cones up to 10 m high were located in caves with no air space. This speleothem is formed through accumulation of cave rafts. This data points toward a lower paleo water level in the caves. Stalactites found under 20 m deep water represent further evidence of a slow subsidence of the area.



Ford, Derek C.
Department of Geography, McMaster University, Ontario, Canada

Limestone, dolomite, gypsum and anhydrite outcrop over 1,200,000 square kilometres in Canada and there are more than 500,000 square kilometres of salt accessible to interstratal dissolution. Most of the country was repeatedly glaciated by alpine glaciers or continental ice sheets during the Quaternary; half of it is permafrozen today.

The interrelationships between glaciers and components of the karst system are complex and varied. Nine differing effects of glacier action upon karst are recognised here: 1. erasure by glacier scour, impacting chiefly upon karren; 2.dissection of cave systems, principally in alpine areas; 3. infilling of dolines and poljes with glacial detritus; 4. injection of detritus deep into the aquifer, rendering it partly or wholly inert; 5.shielding of soluble bedrocks from postglacial dissolution by detritus rich in soluble fragments; 6.sealing of limestone and dolomite pavements by melt-out tills; 7.acceleration of karst development by superimposing glacier aquifers upon karstic aquifers; 8.steepening groundwater hydraulic gradients by glacial entrenchment of valleys; 9.inducing interstratal dissolution by deep injection of groundwaters during glacial recession and isostatic rebound.

Relationships with modern permafrost are also complex. Glacieres (ice-filled caves) exist S. of permafrost boundaries. There is little impediment to karstic groundwater circulation in the discontinuous permafrost zone. Where permafrost is widespread or continuous, active karst progressively shrinks to the zones of highest hydraulic gradient or most soluble rocks. At the climatic extreme karst is of subglacial melt origin and is relict in postglacial environments.



Rai, R.K.
Department of Geography, North-Eastern Hill University, Shillong, India.

The Cherrapunji Plateau is the part of Meghalaya, geologically and physi ographically the Meghalaya Plateau is a part of Peninsular Shield located to the south of the north-eastern segment of the Himalayas.

The development of karst geomorphological landscapes in the vast karst regions of Meghalaya is closely related to lithological characteristics of rocks and is controlled by many natural conditions in particular lithology, structure of rocks, climate, height of the water table and nature of drainage as well as biogenic process. The southern part of the plateau is mainly composed of Cretaceous Tertiary sedimentary rocks and a narrow belt of Sylhet Traps. Numerous faults and folds are traced in the rocks. The uplift of the plateau in different periods has given rise to a complex multi cyclic landscape. In the region the structural events in later part of Tertiary period and climatic changes have interacted to influence and control karst development. The development of Mawsmai cave reflects the control of lithology and structure. The limestone deposits found in Cherrapunji plateau consists of limestone, sandstone, shale and thin bends of coal. The caves with stalactites and stalagmites, sinkholes, underground channels are the common features in the area.


Ngo Ngoc Cat, Eng. Le Uxuan Hong 
NCSR of Vietnam

Investigations on subsurfacial karstic development of karstic plains see m to be extremely important not only in theoretic side but also in practical side, especially for groundwater prospecting and investigations, construction of irrigation system (hydro-electric dam, water reservoir), basements, bridges and other industrial and military establishments.

Based on the multiannual investigation results on hydrogeology and engineering-geology conducted by Geological Inspecting Teams No.9T, 58, 63, 47 and 37 at different area of Northern Vietnam and the field investigation of the Authors at Thainguyen, Honggai, Phuly, Ninhbinh and Thanhhoa, the following conclusion can be proposed:

1. Subsurfacial karstic development is available in the two main d irections of NW-SE and NE-SW .
2. Karstification is being diminshed with the depth of 60-70 m. The most popular depth of subsurfacial karstic groundwater      bearing caverns is 10-40 metres.
3. The linear coefficient of Karstification is mainly in the range of 10 -15%.
4. It is the groundwater that has decisive role in subsurfacial karstifica tion velocity.
5. The above mentioned conclusions more or less correspond with the subsur face
    geophysical investigation results from electric-sounding and well-logging approaches.


Department of Geology, Shiraz University, Iran







A.B.Klimchouk, V.M.Nasedkin, K.I.Cunningham

Secondary cave formations of the aerosol origin. Peculiarities of morphology, structure and localization of some types of gypsum secondary formations (crystals, rims, hollow "stalactites") are considered from caves of the Western Ukraine, Kugitang Ridge(Turkmenistan) and Guadalupe Mountains (USA). It is shown that all these formations are controlled by air flow and caused by deposition from aerosols. Mechanisms of generation of cave (autochtonous) aerosols are suggested. Enhanced radioactivity and high ionization of cave air are major prerequisites for cave aerosol formation. As a result of radioactive decay condensation aerosols, aeroions of mineral combinations and solid aerosols are formed. Mechanisms of transportation and deposition of aerosol material in forms of above formations are considered. It is assumed that many cave subaerial formations are of an aerosol origin.



Guizho Science Fund sponsored project
An Yuguo1 He Fusheng 2 Rong Kunf ang 3 Li Jingyang 3
1 Guizhou Academy of Science
2Guizhou Normal University
3Guizhou Institute of Technology


Through detail investigation of stromatolites of speleothems in Zhijin Cave, Guizhou, the author has systematically collected samples of live algae and speleothem. The species of live algae are identified by biomicroscope, petrographic microscope and SEM observation: preliminary ascertainment of species related to lithogenesis and bilateral comparison with algae fossil in stromatolite, we have determined that algae fossil in stromatolite is similar to live blue-green algae which belongs to the Quaternary fossil of blue-green algae. Finally, structural characteristics of stromatolite has been assorted.

Keywords: Zhijin Cave, Stromatolite of Speleothem, Organic Build Up


1. The stratum that the cave develops is Huangchu Ba member, Yuelang group, lower Triassic period (T1Y2), which is composed of thin to thick bedded bright pale limestone. Its total thickness is 210 m. The overlying stratum, Juoji tan member(T1Y3), is purple shale which is 54 m thick. The underlying stratum, Shabaowang member(T1Y1), is yellow green shale 34.5m thick.

2. The cave situation: 1316 m a.s.l. at cave gate; 1391 m a.s.l. at Shiwan Dashan, the highest bottom in the cave; 1132 m a.s.l. at Shuixiang Zeguo, the lowest part in the cave. The relative elevation is 249 m. According to U-series dat ing, the oldest formation of the speleothem is 261thousand year s B.P(Zhao Sushen,1989).


The gate of Zhijin Cave is facing southwest 240°E . Longer time sunshine, air, humidity and temperature is suitable for algae living. Therefore, algae are prospering in the Yinbing Hall, especially for blue-green algae species. On the surface or wall of collapsed rock and the cave, "algae crust", constructed by kinds of live algae colony is growing. It has been identified that include: Gl oeocapsa Kutz., Nastoc Vanch., Chroococcus Nag., Phormidium Kutz, etc, all of which in Cyanobyta. Example 1: on a collapsed boulder's surface a solidified white speleothem stripe has been formed along the surface due to weeping effect; a "algae crust" cap, 0.5-2.0 mm thick, is growing on its surface (Fig.I-1). It is observed to be live blue-green algae colony that we have mentioned (Fig.I-2). Example 2. On another nodular stalactite (diameter 5-25 mm), a layer of purple, blue and green algae crust cap is growing on the surface, and CaCO3 powder like speleothem saturated with water and cream white color can be seen after removing the crust on the nodule. Downwards, on the base, there are solidified and striped nodular speleothem(Fig.I-4). Example 3: On a northwest fault wall, similar nodule are growing parallely, covered by a green algae crust which is also formed by the above mentioned live algae colony after observing it from SEM(Fig.I-5).

Fig.1 The Sketch Map of Z hijin Cave
W .cave entrance 1.yinbing hall 2.karst window 3.jian menguan 4.caizhulou 5.jinta palace 6.sun & moon lake 7.shouxing palace 8.wa nshou hill 9.wangshanhu 10.mangu corridor 11.nantianmen 12 .xuexang palace 13.linxiao palace 14.guanghan palace 15.yinyu palace 16.shiw an dashan 17.shuxiang zeguo 18.dining hall 19.beihai long 
20.dahai long 21.jinshu palace


Many scholars have done detail researches about distribution of blue-green algae colony. Many articles have pointed out that, in bright illuminated shallows, at dark or faint light circumstances, or even in condition of total darkness, blue-green algae can also survive. Therefore, at dim or light scattering places or tourist lamp spotting places, algae is growing at most places. After sampling it is identified to be: Lyngbya Ag., Phormidium Kutz., Scytonema Ag., Tolypothrix Kutz. of Cyanobyta. An extreme flimsy algae crust formed by algae colonies is observed under biomicroscope. Blue-green algae filament excretes calcic material, catchs and adheres tiny particles. This distinct calcification has proved that algae can seek life activity tenaciously in speleothem environment of Zhijin Cave and are primary lithogenesis.


Refering to some related articles (Reference Materials of southwest Geological Technology, 9th volume stromatolite Collection, 1975.Southwest Geological Technology Research Institute .)  and combining them with the st ructural characteristics of stromatolite of speleothem in Zhijin Cave, we have following assortment:

1. Tree stromatolite, depict the speleothem "Yingyu su" in Yingyu Palace(Fig.I-6), "Ta song"(Fig.I-7) and "Xue song"(Fig.I-8) in Jinta Palace as examples. We have found that they belong to "Huge Complex Merged Stromatolite", 7-20 m high, base diameter 0.6-3 m, conic helicoid(dextral), ordered growing schistose str ucture around the trunk. The center of schistose structure is circle or ellipse. Beside the center, laminae are steep arc-ridge and clear, 1-3 mm wide dark stripe, 1.5-2 mm wide bright stripe. They are sheaf and grain calcite. By identification and comparison of algae fossil sections of "Yingyu su"(Fig.II-1), "Ta song"(Fig.II-2) and "Xue song"(Fig.II-3) through petrographic microscope they are found similar to cave live blue-green algae: Lyngbya Ag. and Phormidium Kutz.

2. Slate stromatolite, depict the slaty stalacite at Jianmenguan as example (Fig.II-4). They also belong to "Huge Complex Merged Stromatolite", columnar stromatolite combination in microcosm(Fig.II-5), tabular stromatolite aggregate in macrocosm. The center is ellipse and one or two sides of it are steep arc-ridge lamination which is clear, 0.1-0.5mm wide dark stripe, 0.1-1 mm wide bright stripe, belongs to hoar sheaf calcite.

3.Scale stromatolite, depict the scaly stromatolite at Yinbin Hall and Jinta Palace as examples (Fig.II-6). The scales always are regularly destributed scalelets. The algae lamination is distinct and ripply, 0.1 mm wide dark stripe, 1 mm wide bright stripe, belongs to sheaf calcite crystal.

4.Arch stromatolite, depict the speleothem "Bamankui" at Lingxiao Palace as example(Fig.II-7). They are "Huge Complex Merged Stromatolite". Its cross section is concentric arch structure. And its lamination is distinct and beded or ripply, 0.1-1 mm wide dark stripe, 0.1-1 mm wide bright stripe, belongs to sheaf calcite crystal.

5.Column stromatolite, depict with the white deposition (Fig.II-8) at Wansanhu and stoneflower at Mangu Corridor. Regular or irregular, furcated or unfurcated(Fig.III-1) cross section is concentric columnar lamination (Fig.III-2,3). The lamination is compact, distinct and ripple arched, 0.16 mm wide dark stripe, 0.17 mm wide bright stripe, are grain calcites most of them. The identification and comparison of algae fossil section through SEM have discovered blue-green algae fossil which are similar to live blue-green algae: Lyngbya Ag., Phormidium Kutz(Fig.III-4).

6. Nodule stromatolite, depict with the nodular stalactite (Fig.I-4) at Yinbin Hall and Jianmenguan as examples (Fig.III-5). The center is stellar half sphere nodular structure that expands around and protruds upwards. It has distinct lamination, 0.1-1.62mm wide dark stripe, 1.39 mm bright stripe and it is c olumnar calcite crystal. The identification of algae fossil by SEM has shown blu e-green algae fossils that are similar to live algae: Gloeocapsa Kutz., Nostoc Vauch., Chroococcus Nag., Phormidium Kutz. (Fig.III-6)

7. Globe stromatolite, depict the "cavepearl" at Chaizulou as example (Fig.III-7). They are globular or elliptical, ordinarily 15-50 mm in diameter, mostly have one core that are mainly constituted by crumbs and gravels. The lamina is concentric, compact and distinct, 0.1-1 mm dark stripe, the same wide bright stripe. The identification and comparison of algae fossil by petrographic microscope have found that they are similar to live blue-green algae: Phormidium Kutz.


1. The intrinsic reason of forming grotesque speleothem is lithogenesis that takes major part in the blue-green algae's life activity seeking in cave speleothem.

2. The structures of live blue-green algae colony: algae crust and stromatolite of speleothems are basically the same, from observing by biomicroscope, petrographic microscope and SEM. By comparison algae fossil in stromatolite is found similar to the Quaternary blue-green algae fossil.

3. The structure, shape and characteristics of stromatolite are directly controled by related algae activities.


(1) John I.Wray, 1977, Calcareous Algae. Elsevier Scientific Publishing Company. Amsterdam-Oxford-New York.
(2) Fossil Algae. Recent Results and Developments. Edited by Erik Flugel. Springer-Verlag, Berlin-Heidelberg-New York,       1977.
(3) Zhao Susheng, 1989,U-series Ages of Karst Speleothem in East China. Karst of China. No.1
(4) Rong Kunfang et al. 1991, Organic Build-up of Dripstone deposition in Zhijin Cave. Journal of Guizhou Institute of     
     Technology, No.3


I-1. white speleothem formed by primary lithogenesis, being cov ered by an"algae crust" at Yingbin Hall. photo 
I-2. SEM image of I-1. ´EÀ 2000
I-3. SEM image of I-1. ´EÀ 350
I-4. nodular stalactite being covered by an "algae crust" at Yingbin Hall. photo
I-5. SEM image of I-4. ´EÀ 300
I-6. tree stromatolite "Yinyu su" at Yinyu Palace. photo
I-7. tree stromatolite "Ta song" at Jinta Palace. photo
I-8. tree stromatolite "Xue song" at Jinta Palace. photo
II-1. blue-green algae fossil of "Yinyu su" thin section, orthogon al petrography. ´EÀ 65 
II-2. blue-green algae fossil of "Ta song", thin section, orthogon al petrography. ´EÀ 65 
II-3. blue-green algae fossil of "Xue song", thin section, orthogonal petrography. ´EÀ 65 
II-4. thin section of tabular stromatolite sample at Jianmenguan. photo 
II-5. columnar lamination in the tabular stromatolite, orthogonal petrography. 
II-6. scaly stromatolite on the collapsed boulder. photo of II-7
II-7. "Bawankui" the arch stromatolite at Linxiao Palace. photo
II-8. white speleothem at Wansanhu. photo
III-1. bud furcation of columnar stromatolite in white speleothem at Wansanhu orthogonal petrography. ´EÀ 65 
III-2. Cross section of columnar stromatolite in white speleothem at Wansanhu, thin section. photo. ´EÀ 1 
III-3. vertical section of columnar stromatolite in white speleothem at Wansanhu, thin section. photo. ´EÀ 1 
III-4. blue-green algae fossil in the columnar stromatolite at Wansan hu. SEM image. ´EÀ 200 
III-5. vertical section of nodular stromatolite at Jianmenguan. thin section photo. ´EÀ 1 
III-6. blue-green algae fossil in the nodular stromatolite at Yingbin Hall. SEM image. ´EÀ 300 
III-7. "cavepearl", the globular stromatolite at Chaizulou. photo
III-8. blue-green algae fossil in the globular stromatolite. orthogona l petrography. ´EÀ 65



Geurts, Marie-Anne
Department of Geography, University of Ottawa, Canada

This research examines a growing model of the Holocene travertine deposits of Coal River Springs Ecological Reserve, in southeast Yukon Territory (Canada). This is an answer to a question addressed in the international literature and related to travertine dams and terraces.

The study is based on a systematic analysis of forms observed and measured along the site. It appears that there is a continuum from one type of form to another. The irregularities of the surface trap moss and algae that produce the first filter for clastic and ionic content of the saturated water. The deposition and precipitation of calcite produce the first rims and rimstone pools. The development of rimstone pools induces the formation of corbelling and cerebroid forms in relation with the water movement and degassing. The growth of the rimstone produce dams that become rather thin for their height; these dams are named phytoherms. These phytoherms are made by coalescent inversed inversed corbelling. This structure is due to the association of two mosses: Cratoneuron sp. and Bryum sp.. Discontinuous growth has been observed and tentatively interpreted as hydrological variation and ecological readjustment. They produce cementation of the dams be fore a new stage of building. It appears that downstream dams do not start growing before those upstream get filed with water and start to overflow.



M.j.Buck H.P.Schwarcz, 
Department of Geology, McMaster University, Canada

Caves formed by the oxidation of H2S are being described in an ever in creasing number of localities worldwide. Common to all such caves is the oxidation of H2S to sulphuric acid. The sulphuric acid reacts with carbonate, frequently p recipitating gypsum in a variety of distinctive deposits. These deposits are key to understanding the mechanisms of formation in both active as well as in relict H2S caves. Their isotope geochemistry demonstrates the involvement of H2S.

Comparison of two active H2S caves clearly shows strong similarities in their hydrogeology, geomorphology and gypsum deposits. Both are vadose and contain thermal water springs which rapidly degas H2S into the cave atmospheres. At Kan e Caves, Wyoming, U.S.A., Egemeier (1978) described the process of "replacement-solution" whereby carbonate on the walls and ceiling is continuously replaced by gypsum to form a replacement crust. This gypsum crust falls in small fragments and dissolves in the stream on the cave floor. This process enlarges the cave upward and outward. This can also be seen in Villa Luz Cave, Tabasco, Mexico. In both cases, the cave streams have low gradients and are spread over flat floors demonstrating that direct carbonate dissolution by these streams does not significantly contribute to cave enlargement.

Relict H2S caves are best known in the Guadalupe Mountains, New Mexico, U.S.A.. These include the celebrated Carlsbad Cavern and Lechuguilla Cave, with a combined length of over 130 km. None of these caves have active H2S springs, but their genesis can be deduced from the nature and isotope geochemistry of extensive gypsum deposits. These deposits are subdivided into three genetic types.



Wolfgang Dreybrodt
Institute of Experimental Physics, University of Bremen, Germany

Karstification is initiated, where a percolating pathway of linked fract ures of about 10-2cm aperture widths connects a site of water input to an out let, located below the input. Under natural conditions the hydraulic gradients driving water through the fractures are low (i < 0.01) and karstification resulting in tubes of a few mm diameter, takes at least several ten thousands of years.

Close to hydraulic structures, such as dams, vadose cave located below the flow of water reservoirs, or caves where the outlets are blocked for water storage, unnaturally steep hydraulic gradients arise. This can lead to enhanced karstification causing significant leakage in times of only several 10 years, and also might endanger the structural integrity of the bedrock foundation.

To model the development of karst channels in such situations a computer simulation is performed, using the dissolution rates of natural limestone, as derived from laboratory experiments on a variety of rock samples. The results show that initially a slow increase of water flow through the fractures is established until after times of seve ral ten years an extremely rapid increase is observed. After this event of breakthrough the widths of the channels widen rapidly with rates of up to about 10-1cm/year.

Using this model we discuss the dependence of breakthrough times on the initial parameters. These are the aperture widths of the fractures, the length of the percolating pathway, the hydraulic head which drives the water flow and also the chemical parameters of the solution, mainly the initial PCO2 in equil ibrium with the water which enters into the fractures, and also the initial concentration of calcium carbonate. We further investigate the influence of grouting to prevent formation of leakage channels.

In summary we show that leakage of hydraulic structures might likely result from solutional widening of preexisting initial fractures, and not only, as usually assumed, from failure of sediment plugs, in preexisting karst channels.



Zhang Jie 1, Li Shengfeng 1, Chen Shufan 2
1 Department of Geo & Ocean Sciences, Nanjing University, Ch ina
2Nanjing College of Education, China

The limestone Sculptures of Liang Dynasty Tomb are located in the east suburb of Nanjing, and have been exposed to subaerial weathering since they were built in AD 522-526. Until they were remounted onto limestone platforms from a rice field several years ago, some had been buried to half of their height by unknown earth accumulation (may be eolian versus fluvial deposition). Microorganisms observed on surface of the sculptures are endo- and epilithic blue-green algae and green algae as well as crustose lichen, which play important roles in surface weathering of the sculpture. Bio-erosional (solutional) phenomena on surfaces include: A) minor honeycomb pit (1.5-2 mm deep) along suture lines with algae covering; B) pin-prick and micro-grooves formed by lichen; C) honeycomb pit patches with algae covering which can only be observed with lens; D) boreholes formed by algal cell or/and fungi filament. With microscopic observation, some interesting mechanisms of bioerosion can be identified. Beside direct erosional effects algae covered with mucus can cause weathering related to the tension developed by drying mucus. Bio-effects combined with other erosional agents causes obvious damage to the sculpture. Bioerosion along suture lines may cause disintegration of the sculptures.





V.S.Kovalevsky, A.V.Efremenko
Water Problems Institute, Russian Academy of Sciences, Moscow, R ussia

As known, karst water regime and resources are defined by the peculiari ty of the geological structure of a territory, climatic conditions, the type and stages of karst development, the degree of relief dissection, landscape conditions and also by the economic activity. The revealing of variations of manifestations of these factors in the karst water regime and resources permits us not only to understand the peculiarities of karst formation in various natural conditions, but also to predict and even quantitatively estimate probable changes in karst water regime and resources, including those under influence of probable transformations of the climate. Such estimations are especially important in regions where karst waters are the main source of public water supply.

We shall examine these peculiarities in three characteristic regions of the Russian platform - the southern coast of the Crimea, Izhorskoe plateau in the B altic area and Ufimskoe Plateau of the West Ural region. The detailed description of the karst of these regions has been given in numerous monographs (Kruber, 1915; Dublyanskii and Kiknadze, 1984; Maksimovich, 1969; Gvozdetsky, 1954; Sokolov, 1962; Gorbunova, 1977, and others). We shall note only the most characteristic peculiarities of these territories, determining the specific of karst formation and, first of all, of karst water regime and resources formation.

The karst of the Mountainous Crimea can be referred to as the karst of the Mediterranean type /2, 9/. This is mountainous, open (from land surface) karst, developing in a region with a strongly dissected relief, up to 1000-1500 m above sea level, in monoclinically occurring (with beds dipping from the south to the north) organogenic limestones of the Upper Cretaceous age with a thickness of 70-80 m. Climatically the area is related to a semiarid subtropical zone where precipitation in the mountainous part reaches 800-1400 mm per year and in the piedmont part it is about 400 mm per year. The precipitation occurs mainly during the cold period of a year and its amount during summer is minimal. The evaporation is about 500 mm per year, the evaporativity is 900 mm per year, annual average surface flow is 300 mm per year, although in upper parts of some karst rivers it reaches 1200 mm per year. The recharge of karst ground water is mainly effected by inflow of precipitation (rain and, in a lesser degree, melted water) during the winter-spring period what determines the characteristic regime of spring discharges and rivers, formed by them /4, 5/.

The karst of Izhorskoe plateau is referred to as the karst of the platform type, where Ordovician limestones are deposited practically horizontally (with a slight inclination to the south), with a thickness of up to 50 m, with a relatively slightly dissected relief, up to 100-150 m above sea level and with an erosional downcutting of up to 50-70 m. On the surface, karstified rocks are overlapped by Quaternary glacial formations of various thickness (from 0 to 30 m), which are represented mainly by loamy moraine with sands, pebbles, and boulders lenses/2/. Climatically, the area is related to a humid coastal zone with a total average long-term amount of precipitation of approximatily 700 mm per year; evaporation is approximatily 450 mm per year, evaporativity is nearly 600 mm per year and surface runoff is nearly 250 mm per year. The precipitation occurs mainly during the warm summer period. However, due to the unstable winter freezing of the zone of aeration the most favourable conditions for ground water recharge are in autumn-winter and spring periods here. The predominant form of karst water recharge is infiltration due to the presence of covering deposits.

The karst of Ufimskoe plateau is also referred to as the karst of the plat form type. Karstified limestones, salts and gypsums of Early Permian age are deposited practically horizontally. Karstified massifs are confined more often to vaults and local rises, brachy-anticlines, are often overlapped by Upper Jurassic and Quaternary terrigenic deposits. The relief is moderately dissected, with absolute elevations of up to 200-250 m and an erosional downcutting of up to 100 m /3, 6, 8/. Climatically the area is related to a moderately continental climate with cold winters, stable freezing of the zone of aeration. The precipitation occurs mainly during the cold period of the year, accumulates in the form of snow and infiltrates during the spring period and during the period of autumn rains. The average long-term amount of precipitation is 700 mm per year, evaporation is 450 mm per year, evaporativity is 600 mm per year. Surface runoff is about 250 mm per year, changing from 100 to 600 mm per year. In spite of a great number of sinkholes (up to 100-500 per sq km) on the surface, ponors and even karren, the predominent type of recharge of karst waters is infiltration that determines the comparative regulation of karst water regime. The karst water regime is various, depending on peculiarities of hydrogeological and climatic conditions. By the time and conditions of recharge it is essentially zonal. In the northern part of the Russian Platform with a stable winter snow cover and the freezing of soils and the zone of aeration, the main recharge of karst aquifers occurs during the spring period. The rise of karst water levels begins with passing of temperatures through zero Centigrade sometimes 2-3 weeks earlier; due to active air circulation in karst hollowes the melting of snow cover begins from below even during minus air temperatures. A more intensive level rise begins during passing of air temperatures through zero. The time of the beginning of this rise of karst water levels ranges from the north to the south from July to April.

In the southern part of the Russian Platform (the Crimea, the Caucasus, the Dniester Area), in the zone with an unstable winter snow cover and sporadic freezing of the zone of aeration, karst water recharge, level rises and increase in spring yields begin with starting of autumn rains and in warm winters can proceed during the whole cold period, i.e., as long as spring. During colder winters, this recharge can be broken off because of formation of snow cover and continue again only in spring in February-March with the beginning of snow melting. In the first case, the continuous winter rise of levels and yields ends with a sprin g peak; in the second case, after the first, smaller by its size, autumn peak, the second spring peaks takes place that is higher, observed more often in March and February, more rarely in April. The rate of karst water recharge also depends on the climatic conditions, i.e., in some degree it is zonal, because the precipitation on the plain decreases from the north to the south, and evaporation increases. However, the rate of recharge depends considerably on hydrogeological conditions, primarily on the degree of permeability of the zone of aeration or the openness of the karst, and on the relief of an area. Izhorskoe plateau, where k arst is covered, and the relief is relatively slighty dissected, the modules of ground water flow change from 1-1.5 to 4-5, rarely 6 l/sec/km2, averaging about 2 l/sec/km2 /7/. At Ufimskoe Plateau with more opened karst and greater dissection of the relief (up to 0.2-0.7 km/km2) the modules of ground water flow change from 1-2 to 5-7, averaging about 4 l/sec/km2. In the Mountaino us Crimea with open karst and intensively dissected relief the modules of ground water flow reach 8-60 l/sec/km2, predominant values are 10-15 l/sec/km2.

The portion of precipitation involved in ground water recharg e also depends on the degree of karst openness and on the correlation between precipitation and evaporation in the region. On Izhorskoe Plateau the coefficient of ground water flow, determining karst river discharges, especially in upper reach, changes from 4 to 27%, averages 10%. On Ufimskoe Plateau this coefficient reaches 20%, in the Crimea it averages about 40%, on some watersheds it is 80% and more. In the basin of the Biuk-Uzenbash River (village of Shchastlivoe), where the area of the watershed is 6.55 km2, the surface runoff, formed by a group of springs, reaches 1250 mm per year, while the maximum long-term precipitation is 1400 mm per year. The module of annual runoff in the watershed is 39.7 l/sec/km2, tha t of the dry season in 5-10% of annual average precipitation.

As known, the degree of regime fluctuation of the karst water dep ends on the degree of drainage of the territory (the rate of water exchange), the size of watershed areas, and the degree of moisture availability of a year. The zones of intensive water exchanges which are situated, as a rule, above the present basis of drainage, are the most variable (with a ratio of maximum to minimum y ields of more than 20). The karst spring yields of this zone respond to all more or less considerable falls of liquid precipitation in a watershed. The graphs of the regime of spring yields, constructed even from month average values, have a saw-like character (Fig.1a). The springs of the Crimea are the most variable among springs of comparable regions. The coefficient of spring fluctuation here in the zone of intensive water exchange reaches 20-600. Such great changeability of yields is determined, on the one hand, by a great amount of precipitation in the mountainous part of the Crimea, greater openness of the karst, that provides a high coefficient of ground water flow. On the other hand, such a dynamics is defined by the greater discharge of the karst massifs, and as a result the most of infiltrated moisture is quickly (during one season) discharged through flow of springs which reaches minimal values during dry seasons and often dries up at all.

The springs of the second zone are characterized by a smaller variation (5-20); this zone can be called active, slightly regulated. This zone is situated both above and slightly below the local basis of discharge. The extent of runoff regulation depends on the size of watershed areas, smaller seepage parameters of karstified deposits, and the presence of covering deposits. The springs of this zone do not deplete, the character of their intra-annual regime is smoother, with clearly expressed seasonal extremums (Fig.1b). The degree of moistening of not only the current but also of the previous year affects the character of this regime.

The springs of the third zone - the zone of the slow water exchange, with the greatest regulation both of recharge and discharge of ground water, are characterized by lesser variation(<5). Such a regulation occurs due to som e greater depths of karst water circulation, greater thickness of covering formations, remoteness of recharge areas from discharge areas. The springs of this zone, both gravity ones and ascending ones, have the smoother character of intra-annual changeability with extremums, shifted to 1-2 months in comparison with the springs of the first and the second types (Fig.1c). Besides the humidity of the current year, that of one or two and even more previous years affects the formation of their regime. The belonging of springs to this or that type can be determined not only according to the character of the graphs of regime and coefficients of variation of their yields, but to some other indice s - by peculiarities of changeability of coefficients of depletion and the degree (ratio) of their seasonal depletion, also by the degree of intra-series interrelation (autocorrelation) of annual average values of long-term series.

In the springs of the first type (the upper zone) a change in coefficients of depletion in time, according to seasonal decrease of spring yields, i s traced more clearly. This change- ability is well expressed on semilogarithmic graphs of the interrelation between spring yields and time (Fig.2), where at least three rectilinear segments are observed.


Fig.1 Characteristic chronological graphs o f the karst spring yield regime 
in different hydrodynamic zones of the Crimea: 
(a) Spring 303/82, (b) Spring 304/83, (c) Spring 541/142.

Fig.2 Yield depletion coefficients of kar st springs of the first type (a,b), 
second type (c,d) and third type (e,f) versus time: (a) Spring 303/82-1974, 
(b) Spring 303/82-1984, (c) Spring 308/79-1967, (d ) Spring 308/79-1968, 
(e) Spring 541/142-1979, (f) Spring 534-1987.

The first of them corresponds to the depletion of the largest channels and hollows of the karst massif, the second one is that of the large cracks, and the third one is that of the small cracks and pores. In years of high moisture availability the first rectilinear segment is longer and in years of low moisture availability it can, on the contrary, be invisible.

In the springs of the second type, more often only two rectilinear segments on such graphs are observed (see Fig.2). The intensity of yield, as it was marked, depends both on precipitation of the present year and on the character of moistening of the previous period. For example, for the Ai-Iori Spring, the coefficient of depletion in years of equal moisture availability and an identical value of recharge, i.e., with equal values of maximum yield, after a series of water-deficit years becomes 0.0025, and after a series of water-abundant years 0.0041, i.e., almost twice as greater.

In the springs of the third type (the third zone), the coefficients of depletion in years of different moisture availability are approximately equal (see Fig.2) and even in individual anomalously water-abundant years two rectilinear segments are observed on the graphs, similar to Fig.2. In some cases the character of the intra annual regime of spring yields is rather similar, even if the change in the intensity of precipitation is essential because the contribution of precipitation to karst water recharge of this zone is not great.

The different degree of the regulation of karst water discharge of individual hydrodynamic zones defined a different degree of intra-series relation between annual average values of ground water discharge. In springs of the first type, this interrelation (autocorrelation) is, as a rule, insignificant (Rt =1 < 0.1). In springs of the second type it is also not great (Rt =1 =0.1-0.2), and in springs of the third type it can reach 0.4-0.5. The long-term cyclicity or tendency to alternation of grouping of years of increased and decreased moisture availability manifest themselves most clearly in this regime of such springs. The most characteristic cycles for springs of the Russian Platform, including karst ones, are 2-3 years, 5-6 years and 10-15 years. The most reliable cycling in the Crimea according to spectrum analysis is 5-year (5.3 years) with splashes of spectrogrammes, essentially falling outside the limits of the confidence interval of 95%. The tendencies to 10 and 15-year cycling are defined obviously by their multiplicity with a 5-year one. On the Izhorskoe Plateau the most reliable cycling is 2-3 years, on Ufimskoe Plateau it is 6 and 12-years with tendencies to 2-4 years. The relative shortness of series of observation and character of the cycling do not permit us to speak about its reliability. As a rule, such cycles are not repeated systematically, this does not permit us to use such a phenomenon for forecasts. However, it is impossible not to take into consideration this tendency for practical solutions.

In years with high moisture availability annual average values of spring yields are 2-5 times greater than in years with low moisture availability; this is essential for solving questions of water supply, based (especially in the southern part of the Crimea) only on karst waters. For practical estimation of the degree of intra-annual changeability of karst springs discharges and rivers formed by springs, the notion about the ratio (norm) of spring seasonal depletion in a form of an inter-relation between minimal annual (month average) and annual average discharges should be introduced. For example, an average norm of depletion for two selected rivers of the Izhorskoe Plateau (rivers Yachera and Plussa) is 0 .2, i.e., during a dry season river discharges are only 20% of the annual flow. The average norm of depletion for the Ufimskoe Plateau (also on two characteristic rivers -Vogulka and Ufa) is 0.17, i.e., it is even smaller because the discharge of this territory is larger. For River Salgir in the Crimea, this norm becomes even smaller (0.13), i.e., here during dry season only 13% of the annual disc harge of the river remains because the drainage of the Crimea karst is the greatest among those of the comparable regions. The character of the intra-annual changeability karst water regime of all zones and regions essentially changes in years of different moisture availability. This manifests itself not only in changes of amplitudes of karst water level and discharge fluctuations, rates of their rises and falls but also in the time of the beginning of extremal positions of levels and discharges, their number in a year and also in the portion of precipitation, participating in ground water recharge, that is very important for predicting ground water regime in connection with human-induced transformations of the climate (Fig.3).



Fig.3 Characteristic karst water regime g raphs for the Izhorskoe plateau (a,b) ,
Ufimskoe plateau (c,d) , and the Mountai nous Crimea (e,f),
for high-water, medium-water, and low-water years.

The character of the dependency of intra-annual karst water regime on the degree of moisture availability of a year can be estimated in two ways:

(a) having constructed a probability graph of annual average values of ground water yields or levels, years corresponding to average long term values (50% of probability) of discharges and water-short and water-abundant years of given probability can be estimated by this graph. The character of intra-annual ground water regime in selected years can be considered as a characteristic intra-annual regime for years of corresponding moisture availability (see Fig.1);

(b) according to tables of long-term observations, made, for example, from monthly average or ten-day data, the probability graphs for each month must be constructed; values of given probabilities (1, 5, 10, 20%, etc.) should be taken; calculation graphs of probable intra-annual distribution of ground water levels or yields for years of corresponding moisture availability can be made. An example of such graphs is given in Fig.4.



Fig.4 Characteristic calculation graphs o f the intra-annual karst water regime in years 
with different water availability . Mountainous Crimea:(a) Spring 474, (b) Spring 485, 
Izhorskoe Plateau: (c): Wel l 1043. Ufimskoe plateau: (d) Well 270.

The comparison of these two ways of estimation shows that in the first way cases where extremal (maximal or minimal) values of ground water levels or discharges in years of different moisture availability can coincide and may appear irregular. Such a situation should be considered accidental, appeared due to the fact that the degree of moisture availability of some seasons of a year (precipitation) may be not coincide; therefore, the probability of annual average values cannot correspond to probabilities of extremal values of the same year. That is why we prefer the second way of estimation. Considering the results of this estimation in comparable regions of the Russian Platform, it must be emphasized that in some cases(see Fig.4) with the decrease in the degree of moisture availabil ity of a year amplitudes of levels also decrease, and in the others they, on the contrary, increase. This fact is connected with the character of the change in karstified deposit permeability in vertical direction. In cases where cracking of limestones essentially attenuates with depth, that is characteristic for the Izhorskoe Plateau and some areas of the Ufimskoe Plateau, ground water levels durin g water-short years lower and ground water circulates essentially in the zone with a water yield 2-3 times lesser on the average than in years of high moisture availability. Amplitudes of fluctuations of ground water levels, as it is known, are inversely proportional to the water yield of water-enclosing rocks. In conditions of relatively uniformity of fracture distribution in the vertical direction or even with some increase in deposit karstification near the base of drainage, that is characteristic for the Crimea and some areas of the Ufimskoe Plateau, amp litudes of fluctuations of spring yields are proportional to the degree of moisture availability of a year.

The mentioned high degree of interrelation between karst water regime and resources and precipitation allows us to use it as the basis for forecasting the probable changeability of karst water resources under the effect of human-induced transformations of the climate. The solution of this problem can be made with the help of the established variation of coefficients of ground or surface flow depending on the degree of moisture availability of a year. As it is known, the coefficient of ground water flow is the ratio of ground water flow to precipitation expressed in percentage, i.e. K=[(Q mm/year) /(P mm/year)]´EÀ 100%. The presence of direct and sufficiently high statistically significant correlations be tween precipitation and river discharge allows us to establish the dependence of values of ground water flow coefficients on the degree of moisture availability of a year, expressed in percentage of probability. An example of such dependencies is shown in Fig.5 for three rivers of the Mountainous Crimea-the Uchan-Su, situated on the southern slope of the mountains, and also for the Kokkozka and the Belbek, situated on the northern slope of the mountains. One can see in Fig.5 that with an increase in the degree of moisture availability of a year this coefficient increases, i.e., during years of high moisture availability part of precipitation, partcipating in recharge of karst waters, increases. The dependence of ground water flow coefficients on moisture availability of a year for water-shortage periods can be established in a similar way (Fig.5b). Thus, using predictive changes of precipitation, it is possible to establish the degree of moisture availability of a year from the probability plot of precipitation. Then from the graph of Fig.5 the value of ground water flow coefficient, corresponding to the given degree of moisture availability of a year for precipitation, can be obtained, and then using the formula, mentioned above, it is possible to calculate the predictive value of ground water flow, corresponding to obtained values of K and P. For example, in correspondence with the scenario of the predicted change of the climate of the State Hydrological Institute (GGI) /1/, in this region by the end of the century the decrease in the precipitation norm 10% is planned, i.e., the predictive average long-term norm of precipitation will reach 1090 mm/year, this will correspond to 60% probability of precipitation of the present period. The data of long-term observations at the meteorological station Ai-Petri, situated in the mountainous part of the massif, are used for calculations; here the main karst water recharge occurs. The coefficient of annual ground water flow (practically defining the whole surface flow in a basin) will reach 41% (see Fig.5), which for predicting the precipitation norm will define the flow equal to 446.9 l/sec or 14.2 l/sec.sq km. In comparison with the mean present flow module (16 l/sec.sq km) its decrease by 22% will occur. The minimum flow will also reduce to 63.22 l/sec or 2 l/sec.sq km, which for the mean present dry seasonal module 2.9 l/sec.sq km will make a decrease of up to 32%.

Fig.5 Coefficient relations for the average annual (a,c,e ) 
and low-water (b,d,f) runoff in basins of karst rivers: 
(1) Uch an-Su, (2) Belbek, (3) Kokkozka in the Mountainous Crimea, (4) Yachera, 
(5) Plus sa in the Izhorskoe plateau and (6) Vogulka (7) Ufa in the Ufimskoe Plateau.

During the first decades of the next century (until 202 0), according to the scenario of climate change of GGI, some increase in precipitation to 1250 mm/year is expected that corresponds to 35% of precipitation probability. The coefficient of annual flow for River Uchan-Su for this precipitation probability will reach 49%, the coefficient of minimal flow will be 54.5%. The annual predictive flow will be 614 mm/year or 19.5 l/sec.sq km, the minimal flow will be 137 mm/year or 4.34 l/sec.sq km. In percentage proportion to the present long-term flow norm this change will make correspondingly +21% and +49%.

Probable predictive changes in karst water resources for two other c haracteristic drainage basins of the northern slope of the Mountainous Crimea have been calculated in a similar way. The results of these estimations can be illustrated by Table 1.

In a similar way the predictive estimations of changeability of karst wa ter resources of the Izhorskoe and Ufimskoe plateau were made. According to the sce nario of climatic change of GGI /1/ in the region of the Izhorskoe Plateau the de crease of precipitaion norm to 5% is expected till the end of the century, that corresponds to 508 mm/year or almost 60% of precipitation probability of the present period. The coefficient of annual flow for River Plussa will reach 37.9% that will define flow to 8900 l/sec or 6.14 l/sec.sq km for the predictive norm of precipitation. In comparison with the present module of the flow of 50%-probability, its decrease by 7% is expected. Dry season flow will also reduce to 2500 l/sec or 1.72 l/sec.sq km, i.e., by 25% in comparison with mean long-term present one.

                                                                                                                           Table 1

Drainage basin

Area of Water-shed (km2)

Present modules of flow of 50% p robability l/sec.sqkm

Predictive modules for periods X)

till 2000

Till 2020

Annual average

Dry season

Annual average

Dry season

Annual average

Dry season

% of norm

% of norm

% of norm

% of norm





















11.4 /-9




x) the value of module is in the numerator, its deviation from present one, expressed in percentage, is in the denominator.

For the River Yaschera basin for this period a decrease of annual flow of up to 3880 l/sec or 7.73 l/sec.sq km is expected, i.e., by 8% in comparison with the present module of 50%-probability. The dry season flow will reduce in this watershed to 750 l/sec or 1.49 l/sec.sq km, i.e., by 25%.

In the period till 2020 for this territory, according to the scenario of climatic change of GGI /1/, some increase in precipitation norm by 50 mm/year, i.e., to 585 mm/year is expected that corresponds to 30% of its probability.

For the River Plussa basin the coefficient of annual flow for this period will be 40.8%, that will define flow to 11100 l/sec or flow module to 7.65 l/s ec.sq km for the predictive norm of precipitation. Thus, in comparison with the present flow module of 50%-probability its increase by 15% will occur. A considerable (by 41%) increase of dry seasonal flow is expected, its module will be 2.83 l/sec.sq km. For the same period for the River Yaschera basin the value 52.5% of annual flow coefficient is predicted, that corresponds to the value 4940 l/sec of the flow or 9.84 l/sec.sq km of the flow module. Its comparison with the value of the present module of 50%-probability gives its increase by 17%. The dry season flow will also increase to 1600 l/sec or 3.19 l/sec. sq km that considerably (by 60%) exceeds the present module of 50%-probability.

For the territory of the Ufimskoe Plato, according to the scenario of climatic change of GGI /1/ by the end of the century the increase in the precipitation norm by 10% is predicted, it will make up 464 mm/year of 70% of precipitation probability of the present period. In correspondence with that, for the River Ufa basin the annual flow coefficient will be 39%, for the predictive norm of precipitation it will define the value of a flow equal to 82500 l/sec or 5.69 l/sec.sq km. In comparison with the present flow module of 50%-probability its decrease by 14% is expected. Also, the decrease of dry season flow to 17800 l/sec or 1.23 l/sec.sq km is predicted, that is by 19% less in comparison with the present dry season flow module of 50%-probability.

For the River Vogulka basin during this period the annual flow coefficient will reach 51.8%, that will define flow value equal to 7800 l/sec or module value equal to 7.57 l/sec.sq km, i.e., by 14% less than the present one. Dry season flow will also reduce to 1250 l/sec or 1.21 l/sec.sq km, by 24% less than the present one of 50%-probability.

For the period till 2020 for the territory of the Ufimskoe Plato, according to the scenario of climatic change of GGI, the decrease in the precipitation norm by 20 mm/year is expected, i.e., to 495 mm/year, which corresponds to its 60%-probability. For the River Ufa basin the annual flow coefficient will be 39.9%, that will define the value of the flow equal to 90000 l/sec or 6.21 l/sec.sq km. In comparison with the present flow module of 50%-probability its insignificant (by 6%) decrease will occur. Dry season flow will be 19800 l/sec or 1.36 l/sec.sq km, it is by 10% less than the present one.

The predictive annual flow coefficient for the River Vogulka will reach 53%. It will define annual flow equal to 8500 l/sec or 8.25 l/sec.sq km, i.e., by 7% less than the present module of 50%-probability. The dry season flow will also reduce to 1450 l/sec or 1.41l/sec.sq km, it is by 12% less than the present one.

The result of the calculations, presented above, are given in the Table 2.

Table 2

Drainage basin

Area of Water-shed km2

Modules of flow of 50% probability

Predictive estimations for flow modules for periods

till 2000

till 2020

Annual average

Dry season

Annual average

Dry season

Annual average

Dry season

% of norm

% of norm

% of norm

% of norm

Izhorskoe Plateau








2.83/+ 41








3 .19/+60

Ufimskoe Plateau





5.69/- 14












The analysis of obtained results shows that climatic changes, caused by the human impact, will lead in the near future to insignificant changes in precipitation (only ±E 10%). However, these changes in turn almost proportionally, but more considerably (to ±E 20-40%), will change annual yields of karst springs and total karst water resources. In different karst regions of the Russian Platform these changes will be opposite and at different scales. Moreover, it should be considered that minimal karst water discharge rates, which on the whole not rarely limit the development of water supply, can change more essentially to 20-50% of long-term norms; this obliges us to study this problem in more detail as applied to predicting the intra-annual distribution of karst water discharge in every region and drainage basin. Both extremal and average values of karst water resources and peculiarities of their intra-annual changeability, i.e., their seasonal regime, will be exposed to changes. The character of the intra-annual changeability of karst water regime and resources can be assessed in each concrete region by calculation graphs (the type of Fig.3) for each corresponding predictive level of probability. The transformations of intra-annual regime will be especially intensive for the springs of the first type and on watersheds with high modules of flow.

Considering the performed predictive estimations in the general dynamics of changeability of karst water resources in the Crimea, it should be mentioned that in most cases during the current century the common predominating tendency to the decrease of spring yields has been observed here. Among 58 observed springs, an obvious tendency to decrease of yields is observed in 32 springs, in 18 springs the regime was relatively stationary and only in 8 springs some yield rise in time was observed. The rates of such changeability are not great-from +0.01 to -0.01 -0.08 l/sec per sq km per year. The character of the most typical changeability of spring yields during long periods of time can be illustrated by the combinatory graph of the regime of two springs with the most prolonged periods of observation (Fig.6). Comparing the given graph with predictive values of karst river discharge, expressed in percentage of probability, it can be supposed that starting in 1985 the increase in spring yields will continue further but will not reach their long-term average values to the end of the century, however, it will correspond approximately to 60% of the probability level. However, in the beginning of the next century this increase of karst spring yields will continue and reach 30-40% of probability. Thus, predictive changes in karst water regime and resources will not go far outside the limits observed before. However, these changes, especially in minimal values of spring and small karst river discharge will be nevertheless considerable in comparison with present ones that must be taken into consideration for solving practical problems on regulation of water resources, planning of their use, etc.

For the territories of the Izhorsk oe and Ufimskoe plateau till the end of the current century a certain (to 15-20%) decrease in ground water resources is expected, which for the Ufimskoe Plato will last till 2020, but with a smaller rate (about 10%). In the region of the Izhorskoe Plato in the future (2020) a rather considerable increase in ground water resources is predicted. Dry season modules, which are used for estimation of ground water supplies, will increase by 40-60% that will allow us to improve the conditions of public water supply.

It can be foreseen that with the change in karst water recharge, some change in karstification processes will inevitably occur. Now, the removal of carbonates with ground water in the Crimea during dry seasons is effected at rate of 60-70 kg/day sq km and during the flood season, i.e., with recharge increase, it will reach 1900-2000 kg/day. The predictive increase both of maximal and minimal spring yields will lead in the future to a corresponding increase in the carbonate removal to 20-30% on the average that can cause definite changes in the environment and will require a separate discussion. In the region of the Izhorskoe Plato, at the present time, in the River Plussa basin the carbonate removal during the dry season is 56 kg/day.sq km, during the flood season it is 216 kg/day.sq km. Till the end of the current century some decrease of it can be expected, in the period to 2020 carbonate removal should be approximately 1.5 times greater that will lead to intensification of karstified processes, formation of new sinkholes, primarily in areas with absence or small thicknesses of overlapping covering glacial formations.


Fig.6 Combined chronological graphs of av erage annual yields of karst springs; 
(1) Biuk-Su (18901993), (2) Babu-Koryto ( 1934-1987) and their trends.

In the basin of the River Ufa (Ufimskoe Plato), ionic flow at the present time during the dry season is 62.5 kg/day.sq km (78.6 kg/day.sq km together with the sulphates),

during the flood period it is respectively 389 kg/day.sq km and 552 kg/day.sq km. The calculations show that during these periods (till 2000 and till 2020) an insignificant decrease in ionic flow is expected, therefore, for the territory of the Ufimskoe Plato activiztion of kartified processes is hardly probable. It is interesting to note that the rate of karstified processes in these rather different (by climatic and geological conditions) karst regions during the dry season is approximately equal and varies in the limits 50-70 kg/day.sq km of ionic flow. However, during the flood season in the semiarid zone (in the Crimea), ionic flow is 5-10 times greater than that in karst areas of the humid zone. Considering the fact that human-induced climatic changes will cause primarily changes in the precipitation of the winter period and will affect the determination of the flood volume, it can be supposed that it should affect karstified processes. In arid and semiarid zones, these processes will be especially active and that is why ecologically dangerous.

In conclusion, it must be once more mentioned that the investigations carried out permitted us to estimate the probable direction, scale and character of the karst water regime and resources changeability of three sufficiently large, di fferent by natural conditions, typical karst regions of the Russian Platform that, on the one hand, allows us to use the estimations for extrapolation over adjac

ent numerous karst regions, on the other hand, these estimations will promote the most rational planning of water resources use and management of these vast regions in the near future. And, finally, the analysis shows the importance and necessity of such estimates for all main karst regions of the world.


1. Budyko,M.I., 1988, The climate of the end of XX century. Me teorologia i Gidrologiya, No.10, P.5-25.

2. Gvozdeczky,N.A., 1954, Karst: Moscow, Geografgiz, P.352 (in Russian ).

3. Gorbunova,K.A., 1977, Karst of the gypsum of the USSR: Perm: Publ. House of Perm University, P.84 (in Russian).

4. Dublyanskii,K.N. and Kiknadze,T.Z., 1984, Hydrogeology of the karst of the Alpine folding area of the southern part of the USSR: Moscow, Nauka, P.128 (in Russian).

5. Kruber,A.A., 1915, The karst area of the Mountainous Crimea: Moscow , P.319 (in Russian).

6. Maksimovich,G.A., 1969, The basis of the karst science, Vols.I,II: Perm (in Russian).

7. Ground water flow of the area of the Central and Eastern Europe, 1982, Edited by A.A.Konoplyantsev, Moscow:VSEGINGEO, P.288 (in Russian).

8. Popov,V.G., 1976, The formation of ground water of the north-wester n Bashkiriya, Moscow: Nauka, P.160 (in Russian).

9. Sokolov,D.S., 1962, The main conditions of karst development, Mosco w: Gosgeoltechizdat, P.320 (in Russian).



Ezatolah Raeissi, Parsa Pezeshkpour and Farid Moore
Geology Department, College of Sciences, Shiraz University

Abstract: Gar and Barm Firooz mountains are located nor thwest of Shiraz in the Zagros. The two Mounts are mainly composed of calcareous Sarvak Formation, sandwiched between two impermeable shaly formations, namely Pabdeh-Gurpi at the top and Kazhdomi at the bottom. The main karst feature of the area is the presence of 259 sinkholes in the Sarvak Formation, which along with other superimposed secondary structures such as joints and fissures direct the rain and snow meltwater into the aquifer. The main discharge point is Sheshpeer spring. In order to study the characteristics of the aquifer, the major ions, electrical conductivity, temperature, and discharge were measured for a period of eighteen months. Moreover, carbon dioxide partial pressure, and saturation index of calcite and dolomite are calculated using a hydrochemical model.

The recession curve of the spring is indicative of two different flow regimes. In the first regime the flow is of conduit-diffused type, while in the second, the flow is mainly diffused and the larger channelways act as a drainage system.


The quality of karst water is essentially determined by the processes and reactions that it goes through from the time of precipitation until it is discharged from the aquifer. Rain water dissolves the atmospheric carbon dioxide during precipitation and while passing through the pore-spaces of the soil, the amount of the dissolved gas is increased due to biological activities. The dissolved carbon dioxide interact with limestones and dolomites and as result dissolution of latter occurs.

In the last three decades, numerous studies were carried out to determine the characteristics of karst aquifers using the physico-chemical properties of the discharging springs. Zolt[1] , Smith and Mead[2], Gams[3], and Pitty [4,5] were the forerunners in this field of research. Garrels and Christ[6] based on the amount of carbon dioxide of spring water classified the karst aquifers into two open systems namely diffused and conduit flow regimes, and concluded that total hardness variation is a suitable criteria in distinguishing between the two regimes. Their conclusion was reaffirmed by the results of Jacobson and Longmuir [7]. Atkinson [8] suggested that karst systems are networks of diffused and conduit flow regimes in which narrow fissures and joints play the main role in storing water. Scanlon and Thrailkill [9] using the suggested criteria by the previous workers have carried out a comprehensive study of conduit and diffuse flow regimes. Their results were not consistent with those of previous workers and they refuted the proposed criteria suggested for classification. Novak [10] , Ede [11] and Cowell and Ford [12] have suggested that diffused flow is characterised by small time temperature variations while big variations are indicative of conduit flow.


The study area which is a part of the Zagros Thrust Zone(Nabavi [13] ) is located 80 km west of Shiraz, Fars province Iran (51 50'-52 20'E and 30-30 30'N) (Fig.1). Gar and Barm-Firooz mountains make the heights of the study area with the maximum and minimum elevation of 3714 m and 2110 m respectively. The mean annual rainfall in Berghan station (Fig.1) with an elevation of 2110 m is 750 mm. The precipitation in winter is mainly in the form of snow, but most of the snow cover is melted by early April.

Gar and Barm-Firooz Mounts constitute the northern flanks of two big anticlines that extend in the general direction of Zagros Mountain Range. The exposed core of the anticline is dominantly made of calcareous Sarvak Formation (Albian-Turonian) which is underlaid and overlaid by impermeable shales of kazhdomi (Aptian-Cenemonian) and Pabdeh-Gurpi (Santonian-Oligocene) Formations respectively. The main tectonic feature is the presence of a major thrust fault (Fig.1). The northern flank of the anticline has been brought up by the tectonic forces and southern flank is so shattered that it is either completely removed or can be seen as large slide blocks. Several normal and strike slip faults also occur and the overall tectonic setting of the area has produced suitable conditions for extensive karstification. The most important karst feature is the presence of 160 sinkholes in Gar, and 99 sinkholes in Barm-Firooz Mounts, (Fig.1). As can be seen the sinkholes follow the same trend, which evidently coincides with the direction of longitudinal faults.

Fig.1 Geological map of the study area


The physico-chemical parameters of Sheshpeer spring as the main discharge point of Gar and Barm-Firooz Mounts were measured every three weeks between April 1990 and July 1991. The temperature, electrical conductivity, carbon dioxide and dissolved oxygen were measured at the site. Magnesium was measured by atomic absorption, calcium, sodium and potassium were measured by flame photometric methods. The carbonate and bicarbonate anions were measured using standard titration methods. Chlorine and sulfate were determined using Mohr and turbidimetry methods respectively. Carbon dioxide partial pressure (PCO2), calcium saturation index (SIC) and dolomite saturation index (SID) is calculated using WATEQF model (plummer et al.[14] ). The discharge was measured by current meter and the stage-discharge curved is prepared.

The catchment area of Sheshpeer spring is 81 km2. According to Raeissi et al.[15] the sinkholes seem to be connected by rather large channelways at depth.


The hydrograph of Sheshpeer spring in the period 1990-1991 is presented is Fig.2(a). Using this hydrograph, the recession curves of the spring is presented in Fig.2(b). The recession curves indicate two different discharge coefficients and thus two different a 1 and a2 discharge regimes(Table 1).



Fig.2 The hydrograph (a) and recession curves (b) of Xheshpeer spring
in period 1990- 1991.


Table1. Discharge coefficient and percentage of base flow and quick flow of Sheshpeer spring.






199 0

a 1




a 2





a 1




a 2




The base flow and quick flow of both regimes were determined from th e hydrograph and are presented in Table 1. As it can be seen almost 40 percent of a regime is produced by the quick flow and the remaining is contributed by the base flow, whereas in the a 2, the base flow constitutes the whole flow. The coincid ence of the a regime with snow melting period suggests that the meltwater enter s the conduits througth sinkholes, joints and fissures and quickly discharges from Sheshpeer spring. In a regime which coincides with the dry season, and no re charge from the rain or snow melting occurs, the stored water in the pore spaces, small joints and fissures gradually discharges into the large conduits and makes up the base flow. The large conduit in the second regime does not act as a reservoir for the stored water, and merely provides a water transportation medium.


The minimum and maximum electrical conductivity is 243 and 295 m/cm respectively. The type of water is Ca carbonatic. The relation between time and various physical and chemical parameter are depicted in Fig.3. Time can be divided into three distinct periods, namely a1a and the wet season period. The elec trical conductivity, total hardness, calcium and bicarbonate ions in the a period ar e lower compared with those of a period. The reason for this observation is that in the a regime as the water enters the conduits through sinkholes and quic kly discharges from Sheshpeer spring the contact surface and the residence time of water is appreciably lower than that of a regime in which slow passage of wa ter increases the surface contact and residence time. The rather quick discharge of water in a does not allow it to become saturated in calcium; while in a 2, the calculated SIc indicates the water is saturated in this ion. At the beginning of wet season with increasing discharge, electrical conductivity, total hardmess and the amount of calcium and magnesium increase, but towards the end of this period these parameters begin to decline again. The initial increase is probably caused by the accumulation of evaporitic salts at the surface in dry season and then quick dissolution and infiltration of these salts into the aquifer at the beginning of the wet season. As the wet season proceeds, the amount of surfacial salts and hence the amount of dissolved ions decrease.

Fig.3 Time variation of physical and chemical parameters 
of Sheshpeer spring in period 1990-1991.

The maximum and minimum temperatures of the spring during the study, was 9.2 and 8.1°E respectively with an average of 8.4°E . The comparison of the spring and the atmospheric temperature suggest that no correlation exist between the two. It seems that the temperature of the aquifer has equilibrated with the temperature of the calcarious rock masses. Moreover, the water flows at the depth of more than 100 meters (Raeissi et al [15] ) and at this depth where an appreciable part of the flow is supplied by the base flow, the temperature can not be affected by surfacial temperature fluctuations and remains virtually constant throughout the year. The minor changes of temperature which well correlate with discharge are presented in Fig.3. As mentioned earlier, the increasing discharge is directly related to the infiltration of snow meltwater through the sinkholes, the temperature of the meltwater is more than that of the aquifer and thus the observed slight temperature increase is probably caused by the mixing of quick flow.


The mean, standard deviation and coefficient of variation of discharge, temperature, total hardness, electrical conductivity, total dissolved salts and dissolved ions of Sheshpeer spring are presented in Table 2. Attempts to determine Sheshpeer spring flow system using the criteria proposed by various authors has produced contradictory results. For instance, using Shuster and White [16] criteria, Sheshpeer flow system must be classified as conduit, while according to the criteria of Novak [10] , Ede [11] , Cowell & Ford [12] it is diffused flow. Jacobson and Longmuir [7] scheme can not be used at all since some parameters are indicative of diffused and others of conduit flow. There is no doubt that none of the available classifications gives the full answer to the problem of classification of the flow systems, as they seem to be based on insignificant or nonuseful variants. Probably many other variables must be taken into account if a comprehensive and useful classification is to be presented. Some of the parameters to be considered include the percentage of quick and base flows, presence or absence of sinkholes and large fissures, shape and surface area of drainage , lag time, standard deviation of temperature etc.. Disregarding the available classifications, an attempt is made here to propose a likely model for Sheshpeer spring flow regime using the determined physical and chemical characteristics.

Table 2 Mean, standard deviation and coefficients of variation of physical
and chemical parameters of Sheshpeer Spring in period 1990-1991.

Dissolved Ions(epm)














0.0 25




Standard deviation








0.5 3

Coeff. of varaition









Physical and chemical parameters

Discharge m3/s

Spring Temp°E C

D.O mg/l

CO2 mg/l

Hardness PPM







8 .54








Standard deviation










Coeff. of varaition










As mentioned Sheshpeer hydrograph is made of two distinct flow regimes. The hydrograph indicates that 40 percent of the flow is contributed by quick flow through sinkholes and large fissures. The infiltrated water in small joints and fissures produces 60 percent of the flow. Hence in the first regime, the conduit to diffused flow ratio is 40:60. The quick passage of water has resulted in the total hardness, electrical conductivity, total dissolved solids and calcium and bicarbonate ions to be less than that of the second regime. In addition to transporting water, the large channelways also play a storing role in this regime. In the second regime where there is no snow meltwater and rain during the dry season, the sinkholes do not have a significant role in contributing water to the system. In these circumstances, the stored water in small fissures in winter and spring gradually enters the large conduits and keeps the spring running, though with a lower discharge rate. Hence, the storing role of large conduits in this regime is totally eliminated. The flow regime is diffused and the base flow is the only contributing agent. Electrical conductivity, total hardness, total dissolved solids and the amount of calcium and bicarbonate ions are higher due to diffused flow.


The autors would like to thank the research council of Shiraz University for their finnancial support.


1. Zolt J., Die hydrographic des nordost alpinem karsts, Steirische Beitrage Hydrogeologie (1960/1961) 2, 183 (1960).

2. Smith D.I. and Mead D.G., The solution of limestone, Proc. Univ. Br istol Seleo. Soc. 9(3), 188-211 (1962).

3. Gams J., Factors and dynamics of corrosion of the carbonate rocks i n the Dinaric and Alpian karst of Slovenia (Yugoslavia), Geografski Vestink 38, 11-68 (1960).

4. Pitty A.F., An approach of discharge from a karst rising by natural salt dilution, J.Hydrol. 4, 63-69(1966).

5. Pitty A.F., Some features of calcium hardness fluctuations in two karst streams and their possible value in geohydrological studies, J.Hydrol. 5, 202-208 (1968).

6. Garrels R.M. and Christ C.L. Solutions, minerals and Equilibria, Ha rper and Row, New York, N.Y., 2nd ed., 450(1965).

7. Jacobson R.L. and Longmuir D., Controls on the quality variations o f some carbonate spring waters, J.Hydrol. 23, 247-265(1974).

8. Atkinson T.C., Diffuse flow in limestone terrain in the Mendlip hil ls, Somerset (G.B.), J.Hydrol.35,93-110(1977).

9. Scanlon B.R. and Thrailkill J., Chemical similarities among physica lly distinct spring types in a karst terrain, J.Hydrol. 89, 259-279(1978).

10. Novak D., A contribution to the knowledge of physical and chemical properties of the ground water in the Slovenian karst, Krs Jugoslavije (Cars us Jugoslavie) 715, 171-188(1971).

11. Ede D.P., Comment on "Seasonal fluctuations in the chemistry of limestone springs "By: Evan T. Shuster and William B.White,J.Hydrol.16, 53-55(1972).

12. Cowell D.W. and Ford D.C., Karst hydrology of the Bruce Peninsula, Ontario, Canada, In:Back W., La Moreax P.E. (Guest-Eds) V.T. Stringfield Symposium processes in karst hydrology, J.Hydrol. 61, 163-168(1983).

13. Nabavi H., Brief description of geology of Iran, Geological Societ y of Iran, Tehran(1975).

14. Plummer L.N. et al, WATEQF, A FORTRAN IV version of WATEQ, A compu ter program for calculating chemical equilibrium of natural waters, International Groundwater Modeling Centre, US. Geol. Sur., Reston(1984).

15. Raeissi E., Samani n. and Moore F., Karst hydrogeology of the Sepi dan Region, Fars Province of Iran, International Symposium and field seminar on hydrogeological processes in karst terranes, Antalya, Turkey, 7-17 Oct. (1990).

16. Shuster E.T. and White W.B., Seasonal fluctuations in the chemistr y of limestone springs: A possible means for characterizing carbonate aquifers, J. Hydrol. 14, 93-128)(1971).



Petar Milanovic, Energoprojekt, Beograd, Yugoslavia
Veljko Jokanovic, Karst Institute, Trebinje, Yugoslavia


The question concerning the hydrogeological characteristics of conglomerates has been recently raised several times. It relates, first of all, to the karstification of these rock masses, i.e. whether they are liable to karstification and which karst shapes can be formed in them. The problem of watertightmess of conglomerates has been studied from the point of view of surface storages construction, as well as types of aquifers formed in these rocks and possibilities of intake of this water.

large scale investigation works, which have been carried out during about twenty past years in the Nevesinje field for the requirements of the Trebisnjica Hydrosystem Project, made it possible to define rather exactly the hydrogeological characteristics of, so-called, Promina series conglomerates. Beside the investigations for the purpose of defining the future storages watertightness, some aquifer zones have also been investigated with aim to determine their yield and find out a possibility of an artificial discharge regulation.


Thick Promia series sediments have been deposited at several locations in the dinarides karst zone. Beside their classical locality - Promina mountain near Drnis Town, large masses of these sediments have been deposited in the zones near Posusje, in the Nevesinje field, and the north-western rims of Dabar field are built-up of these sediments as well. Such isolated phenomena are not particularly important or characteristic from the hydrogeological point of view.

The Promina series constitute the Paleogene molasse type sediments. In the Nevesinje field these sediments thickness exceeds 700 m. Conglomerates represent a predominant lithological member, and therefore the term Promina series often means exclusively this kind of sediments. The lithologic column indicates a non-uniformity of sedimented material. In deeper zones, pebbles and boulders prevail as well as slightly rounded cretaceous particles ranging from some milimetres to more than one meter in size, which are cemented together by means of carbonate matrix. Alveoline-nummulite limestones are, also present but to a less significant extent. Pebbles, boulders and cretaceous particles are cemented together by carbonate and sandy substances. These sediments are overlain by thick conglomerate masses whose pebbles and boulders originate from sediments of different age varying from Werfenian to Eocene. Volcanic rocks' boulders and pebbles (gabro and diabase) are often encountered in this zone. The final sequence consists of conglomerates with carbonate pebbles and bouldes of uniform composition. Both in the middle and in the upper sequences these sediments are cemented by carbonate and sandy matrix.

In the few localities the Promina series include, beside conglomerate, the marlstone zone as well. As a result of exploratory boring in this localites, in the geological column of Promina four separate marly zones of the total thickness of 44 m have been encountered down to the depth of 100 m. Along other locality reach, the existance of six marly zones of total thickness amounting to 75 m, wa s established to the depth of 162 m.

The geophysical investigation works confirmed that the carbonate paleorelief of the depression, in which a rapid sedimentation of these molasses has happened, has an expressly hilly topography.

Some parts of this relief represent the depressions which are deep to the elevation +50m, according to the geoelectrical soundings, while in the other area the limestone reefs are outcropping on the terrain surface. The surface altitude is about 850 m.

These layers, in majority of cases, are concordant over Eocene flysch sediments and discordant over the older deposits, although their lateral boundaries are undoubtedly of tectonic origin. Subvertical to vertical tectonic boundary separates the Promina conglomerate mass from the Velez carbonate massif. The location of this contact has been confirmed by boring in the immediate vicinity of the fault zone, at the foothill of Velez, mountain.


The hydrogeological characteristics of the Promina series and its hydrogeological role within a macro-plan depend on the percentage in the composition of marly, marly-clayey and sandy interbeds, or on their thickness and spreading continuity. In the parts of this series where these sediments often occur in zones,thick-bedded and in layers, such rock masses form the basic hydrogeological characteristics of this series and represent, as a whole, a watertight complex, i.e. have the role of hydrogeological barrier. On the contrary, the parts of the Promina series which are composed of conglomerates exclusively are liable to karstification as any other soluble rock. Bearing in mind that the rock base is composed of carbonate (limestone) boulders and pebbles, which are cemented by carbonate matrix, it is logical that all typical karstic shapes -lapies, swallowholes (ponors), pits and caves-are encountered in these rock masses. The karstification process development in these rocks is considerably accelerated due to their intensive fissuration.

A great number of karstic forms has been observed in the Nevesinje polje conglomerates. One dry cave (in the aquifer zone) and two caves in the medium part of the stream, are registred. Wherefrom in the Drazanj brook valley a large quantities of water are outflowing in the rainy period. During the investigations of the Drazanj brook valley watertightness, it was found out that one of the cave channels mentioned exceeded by far the length of 200 m. The source of this stream itself is in shape of a typical karstic "spring eye" formed in conglomerates. Neither of these phenomena has been speleologically investigated.

The Zovidolka river valley, which is cut in the Promina series conglomerates, has all the characteristics of a typical karst valley too. This river takes its rise from the temporary spring Jama. The morphological, hydrogeological and hydrological characteristics of this spring are typical ones for all karstic springs. This spring is formed at the contact of Eocene flysch and the conglomerates. An aquifer with typical karstic circulation is discharged through this spring.

Two developed springs in the Nevesinje field - Jezdos and Jedres - represent typical karstic phenomena in conglomerates as well. In the Jezdos spring only the prospection of the inlet part of the channel system has been done, while the Jedres spring channel was speleologically investigated at a length of 230 m.

In the south east part of Nevesinje polje, a great number of swallow holes (ponors) was discovered in conglomerates at more than 100 m far away from the Zalomka river. Swallow holles and the zones in which water sinking occurs have been found out at several places in the Zovidolka and the Zalomka river valleys.

All the examples mentioned indicate the fact that the Promina series conglomerates are liable to karstification and to creation of aquifers having all the hydrogeological characteristics of typical karstic aquifers.

The results of hydrogeolgical investigation works performed in the Jama spring as well as those of an experimental plugging of the Jedres spring karstic channel are presented here.


The Jama spring surrounding area is built-up of Cretaceous, Paleogene and Quaternary sediments. The spring itself is formed in the south-western limb of the anticline structure whose core is composed of grey to dark grey thick-bedded, layered and sheet limestone deposits with dolomite interbeds. These sediments are concordantly overlain by light grey bedded and massive limestones with hypurites. Their thickness ranges from 200 to 300 m. Cretaceous limestones are discordantly overlain by Eocene-Oligocene molasse-conglomerates of the Promina series. This series is concordantly underlain by flych zone which is composed of sandstones, marlstones, claystones, conglomerates and limestone interbeds. The older sequence of the flysh sediments is build-up of sandy limestones, coarse-grained sandstones and conglomerates. The aquifer is formed at the contact of the Promina conglomerates and the sandy-marly facies.

The prevailing structural features of this area are: dinaric strike of beds, imbricated structure, numerous reverse faults and local folds whose axes most frequently sink in north-western direction.

The hydrogeological investigation works carried out so far indicate the fact that the karstic aquifer of the Jama spring comprises the Slato polje broader area and a part of Bjelasnica mountain massif. This aquifer is not discharged in a concentrated way, i.e. through a single spring, but along the aquifer zone at the contact of Eocene flush - Promina series, and the Jama spring is the one of the highest capacity through which approximately 80% of water from this aquifer is discharged. The aquifer zone is extended along the Zovidolka river bed, downstream of the Jama spring at a length of approximately 3.5 km and it consists of the many springs indicated at Fig.1. In the dry season there is no surface outflow at either of these springs.


Fig.1 Geological map of spring Jama area. 
(1). Quarternary sediments; (2). Conglomerates Promina series; (3). Flysch; 
(4). Cretaceous limestone; (5). Established underground link; (6). Temporary spring; 
(7). Cave with water; (8). Ponor (swallow hole); (9).Erosion of tectonic - erosion boundary; (10). Syncline axis; (11).Anticline axis; (12). Fault; (13). Overthrust.

On the occasion of dye test of Mlinica swallow hole in Slato polje (Fig.1) the tracer appeared at all the springs mentioned.

From Mlinica to Jama (approx. 4.8 km) the tracer wave travelled 142 h. whi le to the most downstream spring this travelling took 214 hours.

One dye test made in area about 18 km eastwards from spring indicated the fact that the Jama spring catchment area comprised a part of the Bjelasnica massif. The presence of the tracer was stated by laboratory analyses in only one water sample taken from the Zovidolka river. The trace of the colour was estimated on the sample taken a day before. Bearing in mind the fact that the tracer was stated in the last sample taken, the results of this dye test can not be considered as absolutely reliable. According to these data, the tracer has travelled 436 h. along the distance of 17,900 m at the wave's speed of 1.14 cm/sec. If this information is accepted as exact, a conclusion issues that the tracered swallow hole is located in the bifurcation zone where from the majority of water outflows to the Trebisnjica river catchment area. No dye test made in the Zalomka river bed downstream of this swallow hole has shown any connection with the Zovidolka river.


The aquifer zone of the Zovidolka river is situated approx. 6.5 km upstream of its mouth in the Zalomka river. A dry valley extends upstream of this aquifer to the Lukavac polje. In the period preceding the karstic process development, the water from the Lukavac polje has been discharged through this valley. When the fluvial erosion has been replaced by the karstic one, and the underground flow having become predominant, the Lukavac water was directed to Dabar polje as groundwater flow. As a result of the evolution of the karstic aquifer appertaining to the catchment area of this polje, the water has been redistributed. A portion of it was directed to the lower erosion basis instead to the Lukavac polje, and it outflows in the zone of the existing Jama spring. Therefore a dry valley remains between the Lukavac polje and Jama, and a temporary flow is formed downstream.

The spring zone is formed at the contact of the Promina conglomerates and flysch, the contact slope ranging from 30°E to 40°E . This is a normal lithostrati graphic contact which occured due to changed sedimentation conditions in the basin. The transition is gradual so that final flysch sequences consist of sandstones with portions of conglomerate. The portion of conglomerates is increasing, and the contact zone designates a definite change of sedimentation conditions.

Due to its position in the geological structure, the flysch complex separated hydrogeologically the Lukavac polje from the aquifer zone and prevented the underground water circulation, e.e. the karstic process development towards the south (toward the Dabar polje basis).

The Zovidolka river flow varies from 0.0 to more than 20 m3/s. Mean annu al discharge ranges from 0.63 to 1.19 m3/s. The surface flow with discharges excee ding 10 l/s exists from 276 to 351 days/year.

When the groundwater flow through the aquifer zone toward the spring is reduced to a minimum, the surface flow is dried up but the underground circulation is continued along the contact of flysch and Promina series sediments. This water flows along the Zovidolka valley, through the fissured and karstified conglomerates beneath the river bed bottom. Having reached the swallow hole (ponor) zone in the downstream part of the Zovidolka river, this water is sinking and flowing in direction to the Buna spring at elevation 36 m.

Subvertical fissures system predominates over all fractured elements and it is approximately perpendicular to the strike of the structure. Undoubtedly, this fissures system plays a considerable part in predisposing of the karstification process and formation of preferential ways of the groundwater circulation.

With the objective of a more detailed study of hydrogeological characteristics of the aquifer zone a part of submerged channels has been investigated by divers. They confirmed the assumption that the cave part of the spring was connected to a channel whose opening was located in a hillside above the spring. The submerged part of the channel has been investigated at a length of approx. 60 m (Fig.2) its width varying from 0.6 to 7 m and its height ranging most frequently from 1 to 3.5 m. The first 45 m of this part of the channel is slightly sloping (to 8°E on the average), while the remaining 15 m of the channel are very steep (t o 32°E slope).

During diving works the depth of water column in the lowest investigated part of the channel syphon was 16.5 m. The channel width on that place was approx. 7 m and its height approx. one meter. According to the divers' statement, starti ng from the last measured point the channel runs further on at the same slope. The natural spring sill is situated approx. 20 m above the deepest part of the investigated syphon.

At the horizontal direction the channel direction is often changed. Along the first 30 m the channel strikes to south-east, then its direction is changed to north-east, while the final investigated part of the channel is striking northwards.

Three exploratory boreholes were carried out in order to determine the location of the waterbearing channel with security. The boreholes were so located as to enter the karstic channel.


The exploratory pumping was performed in September and October 1986 Bearing in mind that during the previous two months there was neither rainfall in this region nor natural water outflow from the spring, it is sure that the drilling results provided the relevant data about capacity of this aquifer in the dry period.


Fig.2 Karstic channel of Jama spring
A - Horizontal view     B - Cross-section

The pumped water was conveyed throuth the pipes of 150 mm and 200 mm dia downstream of the pumping place to approx. 120 m. In order to eliminate the possibility of flowing the pumped water back to the aquifer, such water was coloured by using one kg of Na-fluoresceine. The presence of the tracer was not establihed in the samples taken or by direct divers' insight in the channel. These observations have been made till the end of pumping.

In the course of the investigations made in the Jama aquifer the total volume of 26,647 m3 of water was pumped out.

If three steady-state drawdowns are analyzed, it will be obvious that the relation between the increased quantity of pumping and the drawdown is linear. In such a case the dynamic reserves of this aquifer will be pumped, i.e. the quantities of water permanently inflowing in the driest period to the aquifer zone and flowing further on underground along the Promina-flysch contact to the west, or beyond the Zovidolka river bed in direction to Zalomka river.

This experiment has proved the fact that approx. 25 l/s of water, which represent a part of dynamic reserves, could be pumped out of this aquifer by drawdown of 44 cm.

If the pumping capacity exceeds 30 l/s approximately, the water level permanently decreases. It means that, beside the dynamic reserves, a part of static reserves of this aquifer is pumped.

Recovery is rather uniform and lasting 8-10 hours for each meter of the water level rise. An interesting phenomenon is observed on that occasion. During recovery the water table continued to rise above the starting position of water level, so that it exceeded the initial one by 130 cm during approx. 24 hours (Fig.3).

Fig.3 Graph of pumping test

Then the drawdown towards the initial water level started. At the beginning, this drawdown has been rapid (during the first 10 h. for approx. 60 cm) but later on it got considerably slower (6-8 cm daily). This fluctuation of water level at the initial level requires a ten-day period.

The explanation of this phenomenon of water mass fluctuation in the karst conduits system should be asked for in the analogy with behaviour of water in the tunnel and in the surge tand, i.e. in the kinetec energy of water mass in the k arst aquifer as a result of creation of a depression in the part of the karst channel (karst syphon) being pumped out.

Several possible concepts of developed intake have been analyzed: 1) pumps installation through the natural channel; 2) construction of a gallery with a shaft to the syphon part of the channel, and 3) well development from the terrain surface down to the investigated part of the syphon.

The drilled well intake type is selected as a most acceptable solution from technical and economic points of view. It consists of two wells approx. 45 m deep and with diameter of 444 mm. Such a well diameter enables the erection of the well installation of Ø\XJ300 mm, i.e. the submerged deep pump of capacity up to 50 l/s.


The karstic channel of the Jedres spring is formed in the Promina se ries sed iments. The spring is located on the periphery of Nevesinje urban area at elevat ion 904 m. The exploratory part of the channel represents a part of the karst aq uifer base flow. The catchment area of this spring is also built-up of the Promi na series conglomerates.

The spring's yield varies from 1.0 to 350 l/s. It is so developed that the c onnection between the place of discharge and the reservoir was achieved through a covered canal.

The aquifer zone has been geologically investigated in detail; speleogical i nvestigations were made, and there after three exploratory boreholes were drille d, out of which two holes in the karstic channel. A part of this channel situate d at approx. 60 m from the outlet (Fig.4) was

selected as the most favourable location for construction of the concret e partit ions. At that place the karstic channel penetrates the relatively compact conglo merates. An access heading has been excavated up to this location and the part o f the karstic channel was widened. The constructed concrete plug is approx. of circular shape with diameters of approx. 3 m and 1.5 m thick. In order to avoid leakage along the concrete-conglomerate contact as well as through the conglomerate, a circular grouting was performed around the plug to the depth of 1.5-2 m. T he grouting effects were tested by local dye tests behind the concrete plug.

Four steel pipes are erected in the concrete plug. The pipe at the bottom thereof is used as a bottom outlet, two pipes in the middle of the plug serve for water intake, while the one of large diameter near the top of this plug make pos sible the access to the reservoir area.


Fig.4 The Jedres spring plugging (horizontal and vertical presentation).
1.Karst channel; 2.Boreholes; 3.Concrete pl ug; 4.Auxiliary adit;
5.Max. water table level in natural conditions, and 6.water table level after plugging.

This plug made it possible the storage of water in the widened part of the cave system and its controlled use in the dry period. As a result of several test fillings, it was found out that the maximum elevation of the stored water must not exceed 12 m above the spring level. It was, also, established that conglomer ates at this elevation, or pressure, play the role of a watertight reservoir are a. Bearing in mind the sporadiclly thin overburden and the intensive fissuration of the surface conglomerate zone, further increase of the normal water level could cause a local disturbance of the slope stability above the spring. Therefore, the control of the underground resevoir is very important in the period of int ensive rainfall.


1.Milanovic P., 1988, Artificial Underground Reservoirs in Karst, Experimental and Project Examples, IAH 21st Congress, Guilin, China.

2.Mojicevic M., 1978, Geological Composition and Tectonic Relations Between Sarajevo and Nevesinje, Special Edition of Geoloski Glasnik, Vol.XIV, Saraj evo, Yugoslavia.

Technical Documentation for Trebisnjica HPP.



J.R.Fagundo(1), J.E.Rodriguez(2), J.de la Torre(1),
J.A.Arencibia(1), P.Forti(3)

(1)- Speleological Society of Cuba
(2)- Institute of Geography, Cuban Academy of Sciences
(3)- Italian Institute of Speleology


An hydrologic and hydrochemical characterization of the gypsum area of Punta Alegre, Ciego de Avila province, Cuba was carried out during the joint Italo-Cuban expedition in 1991. Water circulation dynamics, water chemical evolution, a nd gypsum concentration and their relationships with cloride, which comes from t he marine intrusion into the aquifer, and perhaps also from the underlying halit e deposits have been investigated. A chemical laboratory simulation, using gypsu m saturated waters with different NaCl content, was taken as reference. Subseque ntly the Punta Alegre gypsum area was subdivided into 4 geochemical zones according to water cloride concentration.


Papers concerning geomorphology, hydrogeology, and hydrochemistry o f gypsum karst are scarce in the international literature with respect to those devoted to calcareous terrains, even in the countries of Spain, ex-USSR, USA, and Italy in where there is more field research in (CALAFORRA & BOSCH, 1989). In Cuba only a few geological reports (LUKAC, 1969; ITURRALDE-VINENT, et al., 1982) exists b ut none concerning hydrological or hydrochemical research where made where such karst involves a negligible area.

It is well known that dissolution of evaporitic minerals (gypsum and anhydrite) occurs through a process of physical dissolution in which only the molecules of the mineral and the water are involved. According to the thermodynamical low s, the dissolution of 2.41 g/l of gypsum is possible at 25 º\@DC and its solubilit y decreases at higher temperatures.

In many gypsum karsts water reaches concentrations higher than 2.0 g/l, while in carbonate karst calcium carbonate concentrations higher than 0.35 g/l are due to the influence of sulphides and chlorides (FORD & WILLIAMS, 1989). Usually, natural water shows a sulphate content, which depends on the rock type and /or the presence of some pollutants. PERCHORKIN (1986) reported calcium sulphate con centrations ranging from 0.06 g/l, up to 2.4 g/l in a Permian gypsum-anhydrite a quifer, while FORTI (1988) found contents between 2.58 and 2.7 g/l in the waters of Poiano spring, in the Triassic gypsum-anhydrite of Reggio Emilia: according to historic data, NaCl contents of this spring decreased from 15 g/l in 1862 to 3.1 g/l in 1984.

The results of the hydrologic and hydrochemical characterization carried out during the joint Italo-Cuban Expedition to the gypsum area of Punta Alegre, Cie go de Avila province, Cuba (July 1991) are here presented. Our aim is to contrib ute to improve the knowledge about the dynamics of such hydrochemical processes, as well as to compare them, in the future, with the data caming from similar en vironment in other climates.


The gyspum outcrop of Punta Alegre is the cap rock of a larger diapiritic structure present underground. It shows the classical concetric structure with the Miocene gypsum in the center and, along the border, the Olocenic colluvial cove r, which has been pierced by the uplifting diapir (CHIESI et al., 1992).

The gypsum outcrop has an extention of about 20 square kms and consists mainly of detritic aggregates of cristals of different sizes: the rock is not pure but often includes several limestone, sandstone or marl clasts and gravels.

Small relicts of the limestone cover (up to 5-10 m thick) often overlie the gypsum, thus locally hindering its dissolution.

Only scarce news in this karst area were available before the Cuban-Italian expedition of 1991 (JIMENEZ et al., 1988).

The whole gypsum area is characterized by well developed karst forms (dolines, sinkholes, karren etc.) and the deep karst consists of small vertical and hor izontal caves which often reach the groundwater level (CHIESI et al. 1992).


During the period within July 5th and 9th, 1991, many samples of waters from wells, caves, springs, and lagoons located in and around the gypsum outcrop were collected (Fig.1), also the water of the aqueduct supplying two small towns (Maximo Gomez and Punta Alegre) were analyzed.

Some measurements (temperature, pH, electric conductivity and dissolved carbon dioxide) were made in the field (table 1). Moreover in Maximo Gomez settlement a laboratory was set up to determine the concentration of the bicarbonate, carbonate, sulphate, calcium and magnesium ions. The analytic techniques used were those recommended by MARKOWICZ & PULINA (1979) for field conditions. Na+ and K+ were determined by flame photometry.

Table 1 Hydrochemical parameters of the waters sampled during the Italo-Cuban Expedition 1991.

Samples:1. Sugar cane froth polluted cave; 2. Fresh water well; 3. Rivero Well 1; 4. Rivero Well 2; 5. Los Moya Well; 6. Heradio Ochoa Well; 7. Punta Alegresalt minewell; 8.Well close to the cementery; 9. Correa Well; 10. Brackish lagoon; 11. Rock Salt Cave; 12. Cueva del Agua Well 1; 13. Cue'va del Agua Well 2; 14.Cueva del Agua; 15.Los Corrales Well; 16. Santiago Ferrer Well; 17. Isidro Fajardo Well; 18. Salin spring 19. 'Lvlaboa Well; 20. Aqueduct intake.










































































































































































42. 0


















3 192



































































































Waters polluted by sugar cane wastes were analyzed (see table 2) using special techniques (AA.VV., 1985).

The artificial gypsum waters, with different NaCl concentrations, were obtained by mixing solutions of natural gypsum from Punta Alegre and sea waters: the final NaCl content ranging from 25.9 g/l up to 1548 g/l (table 3).

The aggressivity of the natural samples with respect to calcite (CSR), dolomite (DSR) and gypsum (GSR) were determined by a computer algorithm (FAGUNDO et a l., 1985) based on the method proposed by BACK et al., (1966).



The whole massif has an underground drainage: only one superficial small stream exists in its northeastern part and it is presently used for dumping the sug ar cane mill wastes.

Relicts of an ancient superficial water network are evident on the upper part of the massif. Due to the fragmentation and partial demolition of the carbonat e cover, the gypsum outcrops almost the whole area, thus, an accelerated dissolu tion process has occurred, as testyfied by widespread macro- and micro-forms. Th erefore rain waters rapidly percolate to the groundwater through many karst depr essions, ponors and sinkholes. Cavities originated by this process are small and their vertical development is controlled by the piezometric level.

Several superficial streams, such as the Esteron and Chambas rivers, seem to have springs originating at least partially from the gypsum area discharge.


Fig.1 Index map for the collected samples

Table 2 Physico-chemical parameters (mg/l ) of the sample 1 
(polluted with sugar cane froth) 
obtained by using waste chara cterization techniques



























ST- Total solids; STF- total fixed solids; STV- total volatil solids; SST- total suspended solids; SSF- fixed suspended solids; SSV- Volatil suspended solids.

Anyway most of the karst discharge is seaward, mainly through spring s below the sea, the largest of which, northward to the area, may be easily detected in the aerial photography an it is well known by the local fishermen. There are als o several seasonal springs about 10 m a.s.l., which are active during the rain periods, but their waters result normally brackish, probably due to the sea water intrusion into the aquifer (see sample B). In fact the groundwater is in dynami c contact with the sea which floods the gypsum aquifer probably trough karst gal leries.

Another source for the clorides may be the halite disperded into the gypsum formation (as it was proved for the Poiano spring in Italy (FORTI et al., 1988)) : in fact during the geological studies carried out on this area within the periods 1957-1958 and 1963-1965 (LUKAC, 1959; ITURRALDE-VINENT et al., 1982), the existence of halite beds at depths from 150 and 450 m was proven.


The physico-chemical parameters of the collected samples are listed in Tab.1 . Sample 1 corresponds to a sugar cane froth polluted cave, which was flooded due to an accidental overflow of the pipeline bringing the wastes from the sugar cane mill to an oxidizing pond in the cave surroundings. In Tab.2 a more detailed characterization of this sample is given.

All the samples were collected in the Punta Alegre gypsum area, except sample 20, which comes from the aqueduct supplying the region's settlements, whose in take is directly in the Chambas River.

According to the Table 1, the presence of six different hydrochemical faci es may be established as follow:

HCO­3/Ca++: 0 samples
SO=4 / Ca++ 13 samples(1,3,5,7 ,8,9,11,12,13,14,15,16, 19)
SO=4 > Cl- / Ca++> Na+ 2 samples (1, 2)
Cl- > SO=4 / Na+ > Ca++ 1 samples (10 )
Cl- / Na2 samples (17,18)
HCO3- > Cl- / Ca++ > Na+ > Mg++ : 1 sample s (20)

STIFF (1951) graphic diagrams for some of the samples are reported in Fig. 2, where they are listed on the basis of the increase of their clorine content (in meq/l). The chemical composition of sample 20 (the drinking water) is also repo rted.

The acheived data show the existence of a vertical hydrochemical evolution i nside the gypsum karst, from the surface down to the groundwater, which may be e xpressed as:

HCO3- / Ca++ --> SO=4 /Ca++ --> SO=4 > Cl/ Ca++
> Na+ --> Cl- > SO=4/ Na+ > Ca++ --> Cl- / Na+.


Fig.2 Stiff diagrams for some of the waters collected in Punta Alegre area

The first (calcium-hydrocarbonate) facies corresponds to waters associated with the seepage of metheoric waters in the aereate zone of the limestones overlying the diapire: they are present mainly in the eastern part of the massif, but also, as relicts, in small areas of the central and western zones. This facies, unfortunately, could not be identified during the expedition because the dry sea son avoided the presence of water inside the small limestone caves.

The second facies correspond to waters of sulphate-calcium type. One sample of this type (n.12) is quite interesting having a very low mineralization: its salts content (TSS) was 1.16 g/l, CaSO4 being only 0.83 g/l. The origin of this water may be referred to the interaction of carbonate waters associated to a tran sit zone with a soil reach in calcium carbonate with some gypsum: probably the well, from which this sample has been taken, is only drilled in the cover sediments formed by mixed carbonates and gypsum in which the first largely prevail.

Waters of the calcium sulphate faces are, of course, the more abundant. In m any cases, these are originated by direct infiltration in the gypsum outcrop, as it was the case for the samples 11 and 12 collected inside gypsum caves: both a re characterized by low cloride concentration (0.06 and 0.10 g/l respectively), medium to high gypsum content (2.20 and 2.02 g/l) and low total mineralization ( 2.25 g/l TSS).

Owing to the marine intrusion and perhaps also to a possible interaction with the underlying halite beds, the metheoric seeping water mixes in a different ratio with waters of higher NaCl content when reaching deeper zones of the diapir .

As a result of the ionic strenth effect, the gypsum saturated waters in th e karst aquifer, mixing with the clorine ones increase their CaSO4 content: in f act the waters of sulfate-calcic type have Cl-, CaSO4, and TSS values of 0.49, 2.53 and 3.82 g/l respectively.

Waters of the sulphate-chloride-calcium-sodium facies are represented by o nl y two samples, one of which was polluted with sugar froth. Their cloride contents ranges between 0.58 and 0.87 g/l; calcium sulphate between 2.40 and 2.55 g/l and TSS of 3.90.

The chloride-sulphate-sodium-calcium type sample has 4.26 g/l of Cloride, 4.66 g/l of calcium sulphate and 12.22 g/l of TSS.

Laboratory experiments shown that distilled water is able to dissolve abou t 2.0 g/l of pure gypsum, but this quantity increases progressively reaching 7.5 g /l with the addition of 100 g/l of NaCl (FORD & WILLIAMS, 1989). Then, the solub ility decreases if more NaCl is added: this is largely consistent with our labor atory results obtained when pure gypsum from Punta Alegre deposits were dissolve d until saturation using waters with different NaCl content (see Tab.3).

In the Fig.3 the diagrams of dissolved gypsum with respect to cloride concentration for collected samples (N.2 and 12 were excluded due to their anomalous b ehaviour) and those prepared in the laboratory are presented. Both show the same trend: the higher the cloride concentration, the greater the calcium sulphate c ontents for the interval in which anionic facies SO=4, SO4=>Cl- and Cl ->SO4= are prevailing, while in the interval where cloride facies largely prevails the trend for the gypsum dissolution is either to decrease or to mantain its ratio cons tant.


Fig.3 Dissolved gypsum and cloride content in the Punta Alegre sampled waters
and in those prepared in the laboratory tests

Table 3 Chemical composition of the gypsum (from Punta Alegre) saturated waters 
prepared in the laboratory at different NaCl concentrations




Ionic co ncentration(mg/l)







CaSO 4



















































































































Figure 4 provides a scheme for the cloride content in the collected waters : the studied area may be subdivided into four geochemical zones controlled by the ir altitude: 1) a low Cl content (50-110 mg/l) zone associated to areas located, in general, over 50 m a.s.l.; 2) a middle Cl content (111-360 mg/l) zone, mainly related to an altitude between 50-20 m a.s.l.; 3) a high Cl content (361-1000 mg/l) zone between 20-10 m a.s.l.; and 4) a highest Cl content zone located bene ath the 10 m a.s.l.



Fig.4 Zonation of the Punta Alegre area with respect to the
clorine content of its waters

The anomalous high chloride concentration in sample 1 must be due to pollution. The samples 3 and 11 also showed a relatively high chloride contents in rela tion to their altitude, but presently we have no explaination for their behaviou r.

In the carbonate karst the calcite dissolution process is controlled by the carbon dioxide - water - calcium carbonate equilibrium, therefore a constant ratio exists between the partial pressure of CO2, the pH and the dissolved CaCO3. A lthough the above equilibrium do not act directly over the gypsum dissolution process, a sufficent constant ratio between the carbon dioxide and the pH have been experimentally proved for the Punta Alegre waters: the data of Tab.1, in fact, shows that samples with higher CO2 content (52-184 mg/l) present lower pH (6. 4- 7.4). While the lower CO2 (6.7-7.7) corresponds to the higher pH (7.0-7.7). Anyway these two parameters are shown to have no control over the dissolved gypsum.

Finally, the very high CO2 concentration and very low pH values of the sample 1 are due to the additional carbon dioxide, and hydrogen sulphide generated by the decomposition of the organic matter coming from the sugar cane mill:

CH2O + O2 ---> CO2 + H2O
2CH2O + SO=4 ---> S= + 2CO2 + 2H 2O

It is important to point out that these man-induced processes accelerate the dissolution of gypsum firstly increasing the ionic strength and secondly loweri ng the sulphate ion concentration and, therefore, this kind of pollution of the karst area must be avoided.


Generally speaking, the waters of this region are prone to supersaturation regarding calcite and dolomite and nearby the saturation with respect to gypsum. Only sample 1 (polluted with sugar cane froth) and sample 11 (obtained from direct seepage inside the gypsum outcrop) proved to be aggressive toward calcite and dolomite. Sample 20 (aqueduct water) has a gypsum concentration relatively low. The samples 12 and 17 are those which present the lower calcium sulphate content. Finally sample 10 (brackish lagoon) shows a relatively high supersaturation value with respect to gypsum and a high NaCl content.

In conclusion all the available data confirms that in the tropical climate of Punta Alegre, the hydrochemical equilibria controlling the karst process are quickly reached and therefore the karst evolution of the area is very rapid.

It would be interesting to compare these data, which are only referred to a dry period, with other taken in the rainy season, in order to have more complete information on the hydrogeochemistry of such a karst region.


AA.VV., 1985, Standard methods for the examina tion of water and wastewater, XVI Ed., Ed. ALPHA, AWWA, WPCF, P.1268.

Back, W., Cherry, R.N., Handshaw, B.B., 1966, Chemical equilibrium between t he waters and minerals of carbonate aquifer. Nat. Speleol. Soc. Bull., 28 (3 ): 119-126.

Calaforra, J.M., Pulido Bosch, A., 1989, Principales sistemas k rsticos en y eso en Espasa. En: El karst en Espasa. Ed. J.J. Duran J. L"pez Mart!nez. Mo nograf!a de la Soc. de Geomorfolog!a, 9: 277 - 294.

Chiesi, M., Forti, P., Panzica La Nanna, M., Scagliarini, E., 1992, I1 diapiro gessoso di Punta Alegre, Speleologia 27, P.68-73.

Fagundo, J.R., Valdes, J.J., Cardoso, M.E., De La Cruz,A., 1986, Algoritmo p ara el c lculo de par metros e !ndices qu!mico-f!sicos y geoqu!micos en agua s altamente mineralizadas, Revista CNIC Ciencias Qu!micas, 12 (1/2): 72-76.

Ford, D., Williams, P., 1989, Karst Geomorphology and Hydrology. Ed. Univ. Hyman, London, P.601.

Forti, P., 1988, La Fonti di Poiano. In: Guida Alla Speleologia nel Reggia no . Ed. M. Chesi, Gruppo Speleologico Paleotnologico "Gaetano Chierici": 41 -49.

Iturralde-Vinent, M.A., Roque Marrero, F.D., 1982, Nuevos datos sobre las es tructuras diap!ricas de Punta Alegre y Turiguano, en la provincia de Ciego d e Avila. Ciencias de la Tierra y del Espacio, ACC, No4.

Jimenez, A.N., Bayes, N.V., Gonzales, M.A., 1988, Cuevas y Carsos, La Haba na, P.431.

Lukac, M., 1969, Estratigraf!a y g]nesis de la sal gema en Punta Alegre y en Loma Cunagua, provincia de Camaguey. Revista Tecnol"gica, Vol. VII (5-5): 20-42.

Markowicz, M., Pulina, M., 1979, Ilosciowa pomicroanalizna wod w obsarach krasy welanowego. E. Sielesian Universitet, Katowice, P.67.

Perchorkin, A.I., Perchorkin, J.A., 1986, Chemical composition of frac tu re-karst water of sulphate karsted massifs. IV Congreso Int. de Espeleolog!a , Barcelona: 79-82.

Stiff, H.A., 1975, The interpretation of chemical analysis by means of pat terns. Jour. Petroleum Technology, 3 (10): 15-17.





Viles, Heather A.(1), Burger, Dieter(2), Goudie, Andrew S. (3), Pentecost. Allan(4), Pfeffer, Karl-Heinz(2)
(1)St Catherine's College, University of Oxford, England;
(2)Geographisches Institut, Universitat Tubingen, Germany;
(3)School of Geography, University of Oxford, England;
(4)Division of Biosphere Sciences, King's College London, England;

Holocene tufa deposits are found in a range of limestone areas in Germany and Britain, but many are relict features and tufa deposition today is limited. Field and laboratory studies are being carried out to investigate the reason for such a 'tufa decline'. Climatic changes and human impacts may both be important causal factors. Initial findings from sites in the Swabian and Franconian Alb in Germany; Derbyshire and Yorkshire in England; and North and South Wales reveal a great variety of depositional histories. A suite of tufa types has been found associated with a range of environmental conditions. In most places it seems that a crucial factor reducing recent tufa deposition has been a change in valley geo metry and sedimentation, which has restricted the area available for tufa growth. Water quality changes brought about by human activities may be providing an additional limiting factor in some places.



D.Postpischl a, S.Agostini b, P.Forti c and Y.Quinif d
a Istituto di Topografia, Geodesia e Geofisica Mineraria, Italy
b Soprintendenza Archeologica dell'Abruzzo,Villa Comunale, Italy
c Istituto Italiano di Speleologia, Italy
d Politecnic of Mons, Belgium

Postpischl, D., Agostini, S., Forti, P. and Quinif, Y., 1991. Palaeoseis micity from karst sediments: the "Grotta del Cervo" cave case study (Central Italy). In: M. Stucchi, D. Postpischl and D. Slejko (Editor), Investigation of Historic al Earthquakes in Europe. Tectonophysics, 193: 33-44.

Karst speleothems can be used for tectonic and palaeoseismic analyses; in particular, stalagmites can be treated as the records of a natural pendulum.

Samples of stalagmites from the "Grotta del Cervo" and the "Grotta a Male" caves (Central Italy) have been dated using 14C and U/Th radiometric methods. The present paper shows the limits and validity of such methods for dating strong ea rthquakes of the past.

In particular, radiometric 14C dating shows that the youngest general stalagmitic collapse observed inside the "Grotta del Cervo" cave must be related to the December 1456 earthquake of Central Italy.

*See the Fig.1 in next page


Fig.1 Sections of the stalagmites taken from the "Buco dei Buoi" cave in 1978, and from the "Spipola" cave in 1950 showing the growth axes. For each stalagmite, the dates(rounded boxes) indicate the relative dating resulting from the assumption that the marketed discontinuities are a consequence of the January 3, 1117, earthquakes. The corresponding earthquake catalogue dates are also indicateed(retangular boxes).



Georgeta Ionescu, C. Ionescu

Based on surface data and more than 300 boreholes investigations in Farcu Hill area, in the southern part of Padurea Craiului Mountains, the present paper deals with some features of the Jurassic fossil karst relief sunk now under limy Cretaceous deposits. Our purpose is both theoretical: to enrich the image about some factors and their evolution in modelating this paleorelief and practical: to detect bauxite bodies in this region by use of reconstructing methods, as is t he isobathic map and geostatistical analysis of these data, in correlation with geological and morphological observations.



Chafetz, Henry S.
Department of Geosciences, University of Houston, USA

The waters, precipitates on artificial substrates placed within the waters, and naturally formed precipitates, have been analyzed from active ambient (Turner Falls, Oklahoma, U.S.A., and Plitvice, Yugoslavia) and hot water (Mammoth Hot Springs, Yellowstone National Park, Wyoming, and Durango and Pagosa Springs, Col orado, U.S.A., and Bagnaccio and Le Zitelle Springs, near Viterbo, Italy) traver tine systems. Deposition does not begin until the waters are substantially super saturated with respect to calcite (commonly in excess of 4 times saturation). Wh ereas ambient temperature systems precipitate essentially only calcite, hot water systems precipitate both aragonite (at elevated temperatures and saturation states, e.g., some systems exceed 60 times saturation) and calcite. Shrubs (bacter ially induced precipitates), ray crystals, rafts (thin floating crusts), and carbonate-encrusted bubbles are common constituents within hot water deposits where as ambient accumulations commonly consist of carbonate-encrusted algae, mosses, and higher taxa of plants. Exotic crystal and crystal aggregate forms include planar stellate clusters and dumbbells of aragonite, and trigonal prisms and skeletal (spiky) calcite.

Geochemically, the hot water systems have a much greater range in composit io n than ambient water deposits. The waters commonly exhibit a large downflow incr ease in d13C values (e.g., Durango 8.5‰) and modest increases in d18O values, ge nerally about 1‰. In addition to overall downflow trends, paired samples common l y display signigicant compositional differences within precipitates that formed only centimeters from each other, for example, crusts which formed floating on the water surfaces behind rimstone dams were compared to precipitates which forme d on the very shallow pool bottoms immediately subjacent to the crusts. The crusts consistently have greater d13C and d18O values th an the precipitates which formed on the pool bottoms. These, as well as other examples, clearly demonstrate that chemical analyses of travertine deposits, such as to evaluate climatic changes, must take into account the overall type of travertine (ambient vs hot wate r), location within the travertine system, and also the origin of the specific materials to be analyzed. Analyses of bulk grab samples can readily lead to erroneous interpretations.



Williams Paul W.
Department of Geography, University of Auckland, New Zealand

The paleoclimatic significance of speleothems became clear from research on their oxygen isotope record by Hendy & Wilson (1968). The probable significance of variations in Mg concentration as another proxy for paleo-temperatures was al so recognized by Gascoyne (1983). Most colour variations in speleothems are now known to be attributable to organic matter carried down in feed water from the soil (Gascoyne 1977: White and Lauritzen et al. 1986): thus, potentially, colour should have paleo-botanical significance. The organic matter gives rise to fluor escence when irradiated with UV light, and thus, is readily measured. Shopov et al (1989) have claimed this fluorescence to have paleoclimatic signigicance as a surrogate for solar activity. The research reported here focuses on interpretat ion of the fluorescence record. Stalagmites cut down their long-axes were irradi ated by UV light. Emitted fluorescence was recorded on film and measures by scan ning laser densitometer. Several matters are being investigated, including: (1) the reproducibility of signals from speleothems of the same age from the same ca ve; (2) autocrrelation and spectral characteristic of the fluorescene data; and (3) the relationship between fluorescence and other variables such as oxygen and carbon isotopes, and independent paleo-environmental records (from pollen, lake levels, etc.). Speleothems analyzed range from those entirely deposited this century to those with semi-continuous deposition for 80 ka. Since research is still in progress, results are provisional.

(1) Similiar fluorescence signals can sometimes be obtained from speleothems of comparable age from neighbouring sites, but matching is difficult because of (a) variable growth rates and (b) short-term local 'noise' which is superimposed on general trends. There is a clear need tight age control.

(2) Cyclicity is revealed in the fluorescence data by autocorrelation and spectral analysis, but signigicance levels are difficult to establish, especia lly because rates of growth within individual speleothems are variable. No significant cross-correlation was revealed between the fluorescence from a sp ecimen growing this century and sun-spot data.

(3) When Mg/Sr ratio and oxygen isotopes are used as paleo-temperature surrogates, similar conclusions are reached regarding warm and cool phases in the last 15 ka, but no relationship is found these proxies for temperat ure and fluorescence.

(4) By a process of elimination, it appears likely that, for a given speleothem, variations in fluorescence are more closely related to water balance changes than to temperature variations at least during the Holocene. But major temperature changes, as during a glacial cycle, are likely to be associated with fluorescence changes because of alterations both to vegetation feed w ater quality and water balance.




A.Farrant, P.L.Smart and F.Whitaker
Department of Geography, University of Bristol, England

Extensive limestone caves in the Gunong Mulu National Park, Sarawak, have been affected by repeated incursions of coarse fluvial gravels derived from the ad jacent clastic uplands. These gravels are associated with laterally continuous horizontal 'notches' incised into the walls and roof of the passages. The notches are formed during aggradational events by the deposition of fluvial gravels whi ch protect the floor of the passage from further dissolution but permit incision laterally into the side of the passage. The notch elevation, is related to the aggradation limit of the surface alluvial fan which controls the resurgence leve l. In excess of 15 aggradational events are preserved in the caves by uplift of the limestone which form a vertical suite of high level fossil caves. In contras t, only the remnants of two fans are preserved on the surface. A chronology for these aggradation events can be established over the last 1.5 Ma using uranium series, ESR and particularly palaeomagnetic dating techniques. This record enables an average base level lowering rate of 0.19m per ka to be calculated, which is similar to present day estimates of erosion rates derived from dissolution experiments on limstone pills and river solute budgets. Comparison of this record w ith the astronomically calibrated deep sea oxygen isotope record of core ODP 677 shows aggradation is not deriven by episodic uplift, but occurs during times of decreased ice volume and is thus climatically controlled. This suggestion is confirmed by observations of current fluvial activity in the area.





Folk Robert L.
Dept. of Geological Sciences, Univ. of Texas at Austin, USA

A complex of inorganic and organic factors control precipitation of carb onat es in hot springs of Lazio, central Italy. A plot of data from this area collected by Malesnnt and Vannucci (1975) and Duchi et al. (1978) shows that the main i norganic controls are temperature (Pursell, 1985) and Mg/Ca ratio (Leitmeier, 19 10, 1915) of the spring waters. Virtually all springs with waters hotter than 40º\@D precipitate aragonite, and cooler ones form calcite. Furthermore, even cold-w ater springs precipitate aragonite if the Mg/Ca ratio exceeds 1:1. There are two exceptions to these rules: (1) if super-rapid loss of CO2 occurs at highly turbulent local sites (Kitano, 1963), or at air/water interfaces, or at bubble walls (Chafetz et al., 1991), then aragonite will form regardless of temperature of Mg/Ca ratio; and (2) if organic slimes retard movement of ions, then calcite form s regardless of temperature or Mg/Ca ratio (Buczynski and Chafetz, 1990). There is little consistent relationship of mineralogy to Sr++, SO=4, or carb onate saturation.

To what extent is the precipitation of travertine inorganic vs. biochemica l? Surely, conditions in diverse localities can vary between both end-points, but Le Zitelle springs, at the north flank of the caldera of Viterbo, provide a bioc hemical extreme. Waters are hot (60¡;Ð@æ;È@), with Mg/Ca of 0.2, and are highly sulfur ou s. Carbonate precipitation rates can exceed 2 mm/day. Nonethed samples of carbon ate crusts, only minutes to a few hours old, exhibit aragonite, calcite, and 1ª;À@­ -t o 5ª;À@­-micron euhedral rhombs of probable dolomite. Aragonite forms spherical "pi nc ushions" of radial needles, each needle tipped with a nanno-bacterial body of th e same diameter as the needles, 0.1 to 0.4 microns. Each nannobacterium precipit ated its own needle, and was propelled outward by needle growth. As little or no later "fattening"of the needle occurred, inorganic precipitation must have been insignigicant here. Nonetched calcite crystals are composed of 0.05 micron nann obacterial spheres that were incorporated into each layer of the crystal as it g rew. No evidence of bacteria was found on the ? dolomite rhomb surfaces. Ironica lly, aragonite, calcite, and euhedral?dolomite rhombs all grew within minutes to an hour of each other in the same solution under the same conditions, savaging all the rules in the first paragraph; they remain a baffling problem unresolved by chemistry, physics, or microbiology.




Sahabi.F.1, Ghazban,F.2
1 Geological Dept., Faculty of science, university of Tehran , Iran.
2 Geological Dept. Memaster University, Canada.


In the central Alborz mountains of Iran, the upper Jurassic is repre sented b y a sequence of deep marine limestones nefkrred to as Lar Formation. Karstificat ion and to alesser extent cementation are the most important diagenetic evextin this formation.

Karstification in the Polour area, 85km northeast of Tehran,close to the D am avand volcano has a dual nature and varging magnitude. Near surface karstificati on is limited to sulaerial weathering within the zone of meteopic water diagenes is, whereas in deeper parts(7400 m) large and numerous caverns have been formed indicating the presence of more chemically aggressive waters. Pervasive faults a nd fractures in the area act as a conduit network which facilitates surface wate r movement to the phreatic zone.

Based on stable isotope determination of mafnix limestone and fracture fil li ng calcite cement,it a years that meteoric waters and warmer waters of depth wer e responsible for karstification and subsequent comentation in the Lar Formation .

A geothermal spring earrying CO2 and H2S reaces the surface in the vicinity of the Damavand Volcano. The stable isotopic composition of the hydrothermal spr ing replects mafor unput from local meteoric souru. The acidification mechanism to develop karst appears to be associated with volcanic fumarales. The CO2 and carbonic acid dissolution is balanad by subseqent calcite cementation.

Significant pyrite formation during fracture filling cementation is a stro ng indication of the presena of H2S within this karstic wustem at depth below 400m . Uppon H2S arrival to the oxygenated zone, sulpunic acid is formed which subs equently couses major dissolution and karstification.