S7 - - CD 2- OO/l S10200001
Nova Gorica Polytechnic, Slovenia
AN APPLICATION TO THE STUDY OF ENVIRONMENTAL CHANGES AND PALEOCLIWATIC RECONSTRUCTION
Doctoral thesis
Barbara Vokal
Nova Gorica, November 1999
© 9 PLEASE BE AWARE THAT ALL OF THE MISSING PAGES IN THIS DOCUMENT WERE ORIGINALLY BLANK CIP - Katalozni zapis o publikaciji Narodna in univerzitetna knjiznica, Ljubljana
550.4:551.44(043.3)
VOKAL, Barbara The carbon transfer in karst areas - an application to the study of environmental changes and paleoclimatic reconstruction : doctoral thesis / Barbara Vokal. - Nova Gorica : Polytechnic, 1999
ISBN 961-6311-03-4
104048384 This doctoral thesis was supported by the Ministry of Science and Technology of Slovenia. Co- operation with the French University Paris-Sud was supported by the Ministry of Foreign Affairs of France.
ACKNOWLEDGEMENTS
The author is grateful to the supervisors Dr. Ivan Kobal, Dr. Bogomil Obelic and Dr. Dominique Genty for their guidance in the formulation of this doctoral work. I wish to acknowledge and thank the members of my committee for their evaluation of my doctoral thesis: Prof. Dr. Tadej Dolenec, Prof. Dr. Andrej Kranjc, Prof. Dr. Nikola Kallay and Prof. Dr. Laurent Dever. Permission for my experimental work in Postojna cave was granted by Mr. Alojz Cernigoj. For assistance during sampling in the cave I am grateful to Mr. Stanislav Glaiar, Mr. Vilijem Cerne and Mr. Stojan 2igon, who also helped me with the sample measurements. Thanks also to Dr. Sonja Lojen, Dr. Nives Ogrinc and Mag. Polona Vreda who helped me with the sample measurements. Many thanks go to Dr. Ede Hertelendi and Mrs. Magdolna Sandor for their help with the measurements and in the development of this doctoral work. Mr. Mark Massault was of great assistance in preparation of samples and measurements with Accelerator Mass Spectrometry. Mr. Dan Doctor corrected the English grammar of my manuscript and I am grateful to him. Thanks to my parents for offering me moral support throughout my work on this thesis. Tomaz, thank you for everything. SUMMARY
Karst areas constitute a large part of Slovenia, with several thousands of caves located in the limestone formations. The karstic caves provide valuable resources for reconstruction of environmental conditions on the continent in the past. This is possible due to the great stability of climatic conditions within a cave. Secondary minerals deposited in caves, known as speleothems, preserve records of long-term climatic and environmental changes at the site of their deposition and in the vicinity. The purity of speleothems and their chemical and physical stability make them exceptionally well suited for detailed geochemical and isotopic analysis.
To understand the processes influencing the speleothem isotopic composition, monitoring of cave waters as well as springs and underground rivers is very important for at least one hydrological cycle. In this way we can observe the influence of seasonal changes on the cave waters feeding stalagmite growth laminae. Chemical and environmental parameters which influence speleothem formation are the temperature of air and water, as well as the properties 2+ 2+ of the water such as pH, electrical conductivity, Ca , HCO3" and Mg ion concentrations and drip rate. The thickness of the roof above the cave and the types of cave water may also determine the water properties. During one year monthly water samples were collected at three locations in Postojna cave to characterize different types of cave waters (pool, fast and stalactite drip water), and also from the river Pivka and spring Mo6ilnik. Rainwater samples were also collected and analysed.
Dissolved CO2 and carbonate in cave seepage waters originate from various sources: atmospheric CO2, organic matter that decomposes in soil, CO2 from respiration of plants in the soil, and the dissolution of old layers of limestone. The isotopic composition of speleothems is directly linked with the isotopic composition of CO2 produced in the soil above the cave. Production of CO2 in the soil is connected to climatic changes and the isotopic composition of the CO2 is connected to the type of vegetation growing above the cave, according to the photosynthetic mechanism employed by the plants. The influence of human activity on carbon isotopic values of atmospheric CO2 due to fossil fuel combustion and nuclear experiments is also observed in speleothem carbonate composition. For this study, the isotopic values of carbon, hydrogen and oxygen isotopes were determined in air, various soil components, different types of cave waters, and in the growth laminae of a stalagmite.
The growth rate of speleothems is partly influenced by climatic variations. When speleothems grow in isotopic equilibrium with their parent seepage waters, the isotopic composition of ancient speleothems gives a record of the paleotemperatures of the cave in which these are formed. When the temperature in a cave is stable throughout the year, this temperature is equal to the mean annual temperature on the surface and so we can estimate the temperature above the cave for that year. By analysis of the stalagmite from Postojna cave we determined that the stalagmite deposition was not in isotopic equilibrium with its parent seepage water, thus making determination of paleotemperature impossible.
This research represents a significant contribution to the improvement of knowledge about the karstic world, especially regarding the influence of climatic and environmental factors on karstic phenomena. This is the first interdisciplinary investigation in the Postojna karst system of this kind, combining the determination of environmental factors together with chemical and isotopic measurements. This research enabled the determination of the processes of carbon transfer from the surface to the cave, where speleothems are formed. This research was performed in co-operation with scientists from the Rudjer Boskovi6 Institute, Zagreb, Croatia and the University of Paris-Sud, Orsay, France. CONTENTS
Introduction 5 Theory 8 The carbon in the nature 8 Carbonate equilibria 10 Isotope study 15 Stable isotopes 15 Radioactive isotopes 19 The formation of speleothems 22 Speleothem growth rate theory 25 Speleothems and climate 29 Modelling 29 Site description 33 Methods 37 Sampling 37 Measurements 41 Results and discussion 44 Chemical measurements 44 Temperature 44 pH 45 Calcium and bicarbonate ion concentrations and electrical conductivity 46 Mg/Ca ratio 57 Saturation index 60 Growth rate 62 Isotope measurements 65 Atmosphere 65
SoilCO2 65 Soil organic matter 67 Water. 69 Speleothems 76 Modelling 79 Conclusions 82 References 85 INTRODUCTION
The word karst can be traced back to pre-lndo-European origins (Gams, 1973, 1981). It stems from karra, meaning stone, and its derivatives are found in many languages of Europe and the Middle East. In Slovenia the word evolved via kars to kras, which in addition to meaning stony, barren ground is also a regional name for a district on the Slovenian/Italian border. This district is sometimes referred to as the "classical karst', being the type locality where its natural characteristic first received intensive scientific investigations. Caves in Karst region provide unique and potentially valuable resources for the reconstruction of environmental conditions on the continent over the last few hundreds of thousands to millions of years of earth history, mostly due to the great stability of climatic conditions within a cave.
Several thousands of caves of different dimensions and with prevailing limestone and dolomite deposits have been found in Slovenia (Habic, 1982). In the karst region of the underground Ljubljanica River, the Postojna Karst represents the most speleologically interesting region. Since its discovery in 1818, several works more or less completely present the history of exploration and research in the Postojna Cave system (Gospodaric, 1976; Kranjc, 1997). With the exception of chronostratographic measurements performed in the last 20 years, some employing 14C and 23OTh/234U dating methods, little work has been done on isotopic investigations of deposited calcite (Brodar, 1952; Brodar, 1956; Gospodaric, 1972; Gospodaric, 1977; Gospodaric, 1981; Gospodaric, 1985; Urbane, et. al, 1985; Gospodaric, 1988; Zupan, 1991; Zupan-Hajna, 1993).
Karst terrain is complex, and it is difficult to understand the processes that are occurring in such systems composed of distinctive hydrology and specific landforms. Karst landforms are produced as a result of combined rock solubility and well developed secondary porosity. Secondary minerals deposited in caves, known as speleothems, consist of several forms such as stalagmites, stalactites and flowstones. The purity and chemical and physical stability of the speleothems make them exceptionally well suited for detailed geochemical and isotopic analysis. Speleothems preserve a record of long-term climatic and environmental changes at the site of their deposition and within the regional vicinity. The oxygen isotopic composition of speleothems ideally reflects the isotopic composition of the regional annual precipitation at the time of deposition. However, it is difficult to reconstruct the paleoclimatic conditions based solely upon oxygen because the oxygen isotopic content of dripping water in the past is not known. (Yonge et al., 1985; Ford and Williams, 1989; Hendy, 1971; Issar era/., 1991; Talma and Vogel, 1992). Speleothems are also of great importance for Quaternary studies, because they may be dated by uranium-series and radiocarbon methods (Lauritzen and Gascoyne, 1980; Gewelt, 1985, 1986; Gewelt and Ek, 1988; Gascoyne, et al., 1983; Papastefanou and Charalambous, 1989; Pazdur et al., 1995) and because we can get useful paleoclimatic information (Hendy and Wilson, 1968; Wigley and Brown, 1976; Harmon and Crul, 1978; Harmon, era/., 1979; Lauritzen et al., 1990; Goede, 1994). In most speleothems there are alternations of white porous laminae (WPL) and dark compact laminae (DCL), which are deposited annually (Allison, 1926; Orr, 1952; Broeckeref al., 1960; Blanc, 1972, 1992; Genty 1992, 1993). There may also be present the luminescent laminations formed from soil organic matter trapped into calcite (Shopov et al., 1990, 1994; Baker et al, 1993, 1998), whose process of formation is not very well known. Other important factors in the study of speleothem formation are the dissolution processes in the soil and unsaturated zone, which are dependent on vegetation type and meteorological variations (KogovSek and Habic, 1981; Kogovsek, 1983; Kogovsek and Kranjc, 1989; Kogovsek, 1996).
Speleothem deposition is typically caused by degassing of groundwater which contains elevated CO2 concentrations (Holland, et al., 1964; Ford and Williams, 1989). In the first step, dissolution of the parent limestone by dissolved CO2 in the percolating water occurs to form bicarbonate ions. The CO2 is almost always derived from the soil, where the partial pressure of CO2 exceeds that in the atmosphere by two orders of magnitude. When the solution containing the calcium and bicarbonate ions enters the cave atmosphere, precipitation of calcium carbonate takes place either due to the degassing of CO2 from the solution or by the evaporation of the water leading to calcium carbonate supersaturation. The high humidity (near 100 %) in the cave atmosphere generally limits calcium precipitation to CO2 degassing. The carbon isotopic composition of deposited carbonates is closely related to the isotopic composition of the soil CO2. Soil CO2 is produced by decomposition of organic matter and by root respiration (Dorr and Munnich, 1986). Approximately 85% of the dissolved carbon in seepage waters that precipitate speleothems is derived from soil CO2 (Bastin and Gewelt, 1986, Brook et al., 1990). CO2 production in the soil above the cave is a function of climate changes and of the type of vegetation above the cave.
The aim of my doctoral dissertation is a detailed study of carbon isotope transfer to speleothems in a Karst system. Results of geochemical and isotopic measurements will be explained by seasonal climatic changes. Processes of isotope fractionation that occur in the path of CO2 transfer from surface to the cave will be described by a model. This will enable us to see how environmental changes are recorded in Karst areas. The site of Postojna Cave in Slovenia was chosen for sample collection.
Environmental factors influencing speleothem formation, such as the type of vegetation that is growing above the cave, data on surface temperature and precipitation, and drip water temperature have been measured. The chemical factors were measured only in water samples. The sampling locations for waters in the cave were chosen to cover all types of cave waters such as pool, fast and stalactite drip waters and at positions with different thicknesses of rock cover above the cave. Sampled surface waters are the river Pivka and the spring Mocilnik. 2+ Water properties such as pH, electrical conductivity, Ca and HCO3" ion concentrations were measured. Seasonal changes of these properties were followed during an entire year to determine correlations between them, such as the correlation between chemical properties and thickness of rock cover above the cave and between chemical properties and the type of cave 2+ waters. I also compared electrical conductivity, Ca and HCO3" ion concentrations to find out how the different types of cave water samples are correlated. Additionally, Mg/Ca ratio have been determined and the calculated saturation index values compared with the drip rate of cave waters and growth rate of stalagmites.
The carbon cycle plays a dominant role in controlling many of the chemical and biological processes in different ecosystems. The knowledge of 813C content and the concentration of natural radioactive isotope C and its global and local distribution is of great importance for ecology, climatology and hydrology as well as for monitoring the carbon cycle in the 14 environment. A significant increase in the C concentration of the atmospheric CO2 reservoir was caused by intensive above-ground thermonuclear bomb tests in this century (Levin and Kramer, 1997; Krajcar Bronic et al., 1998). This effect was observed in the whole atmosphere and biosphere in the 1960's, with the maximum atmospheric 14C activity measured in 1964. Various nuclear facilities may also contribute to the increase of 14C in the atmosphere, while the combustion of great amount of fossil fuels causes a decrease of 14C concentration in the atmosphere.
For a better understanding of the isotope fractionation processes the 513C content and 14C concentration were measured in the following samples: - surface air - plants - soil CO2, - soil organic matter - cave air - dripping water at different depths - cave water (pool, fast and stalactite drip waters) - recent cave deposits (speleothems).
Outdoor atmospheric and cave air 813C and 14C concentrations were determined. Comparison of the carbon content of outdoor atmospheric and cave air was determined to evaluate any influences of biogenic carbon inside the cave. The carbon content of 513C and 14C 13 concentrations in soil CO2 and soil organic matter were determined. Variations in 8 C values are the result of differences between the two types of photosynthetic pathways C3 and C4 type since the degree of 13C enrichment in plants depends on the photosynthetic mechanism used. Dissolved carbon in cave seepage water originates from various sources: atmospheric CO2, decomposing organic matter in soil, respiring plants in the soil and by dissolution of old limestone deposits. The 14C activity time series in the cave was determined by measuring 14C in growth laminae of a speleothem. This stalagmite was damaged at its base by an explosion of known age and 14C activity was measured by accelerator mass spectrometry techniques. 14C activity measurements on the investigated stalagmite permit us to construct C time series over the last 50 years. Because 80 - 90% of the carbon found in the speleothem comes from soil CO2 the speleothem 14 composition reflects the atmospheric CO2 composition. This means that the speleothem C 14 14 activity will be close to the atmospheric C activity, or at least to the soil CO2 activity. Such C chronicles can improve our understanding of the soil carbon dynamics, which is of great importance since response of soil carbon reservoirs to climate change and/or human activity is still not well known (Trumbore, 1993). In order to explain the differences between stalagmite 14C time series and their relation to the atmospheric activity, a model has been developed considering the possible carbon sources.
This research will contribute significantly to our knowledge of geochemical cycling in Karst system, especially in regards to the influence of climatic and environmental factors on karstic phenomena. This research will be the first such interdisciplinary complex investigation of the Postojna Karst system and it will help us to understand the processes occurring in the Slovene Karst. The research is performed in cooperation with the scientists from Rudjer Bo§kovi6 Institute from Zagreb (Croatia) and from University of Paris-Sud from Orsay (France). THEORY
Carbon in the nature
Many natural chemical substances circulate in the environment and are important to the chemistry and biology of the earth (Wollast et a/., 1993). The circulation of a particular substance (as defined by its reservoir, the processes affecting it, and its fluxes) is called a "biogeochemical cycle". The global biogeochemical cycle of carbon is shown in Figure 1 where it is seen that the major form of carbon in the atmosphere is CO2, constituting over 99% of atmospheric carbon. The mass of CO2 in the atmosphere represents only 2% of the mass of total inorganic carbon in the ocean. However, both of these carbon masses are small compared to the mass of carbon bound in sediments and sedimentary rocks. Furthermore, there is presently 3-4 times more carbon stored on land in living plants and humus than resides in the atmosphere. Therefore, small changes in the mass of the ocean, land and sediment carbon reservoirs can greatly affect atmospheric CO2 contents. The rate of CO2 changes in the atmosphere depends principally on the time scale for carbon changes in the large carbon reservoirs, and on the rates and direction of the feedback mechanisms that operate between the atmospheric carbon reservoir and other reservoirs.
Changes in atmospheric CO2 caused by fluxes involving weathering or carbonate precipitation are considerably smaller than those caused by fluxes involving biosphere, ocean exchange, fossil fuel burning and land use practices. The carbon mass of sediments is huge, and is dominated by inorganic C in carbonate minerals. The non-living organic carbon masses in the land and ocean reservoirs are greater than the living, and represents large sources of CO2 for the atmosphere.
Methane is principally produced at the Earth's surface, by decomposition of organic matter under reducing conditions. The total flux of CH4 to the atmosphere today is estimated to be about 34x1012 moles C per year, of which about 20% results from fermentation processes in animals, termites and humans. About 60% of the remaining flux comes from the processes of methanogenesis in sediments of various types - primarily within mud of rice paddy fields, swamps and marshes. The main sink for methane in the atmosphere is via oxidation to carbon monoxide by OH radicals, and subsequent conversion to CO2.
The release of CO from fossil fuel burning, primarily from transportation sources, is currently about 28x1012 moles C per year. Decomposition of organic matter in soils and generation of CO from respiration of bacteria and algae in the ocean are sources of CO for the atmosphere via the reaction: 2CH2O + O2 -> 2CO + 2H2O. The residence time of CO in the atmosphere is short due to its rapid conversion to CO2. Thus, most of the CH4 and CO that enters the troposphere ends up as CO2. In the natural carbon cycle, this CO2 is bound by photosynthesis into organic material on land and, less importantly, in the ocean. Decomposition of this organic material then produces more CH4 and CO.
Exchange of carbon between the land and ocean is mainly via streams, whereas that between sediments in the ocean and land is by sedimentation and rising of sea-floor.
Land, ocean and sediment reservoirs are not at steady state because of anthropogenic activities. For the carbon cycle, the impact of society's activities stems mainly from the release of CO2 to the atmosphere by fossil fuel burning (about 5% of the total CO2 released by natural processes of decay and respiration), cement manufacturing (<2%) and land use practices such as deforestation and cultivation.
The net gain of carbon in the organic reservoir on land (Figure 1) depends primarily on the relative magnitudes of carbon fluxes owing to land use, storage and stream transport of organic C. ATMOSPHERE
119 (?) 200 EXCE3S 319 x 10" MOLES YR"<
STRATOSPHERE STRATOSPHERE 3.8 4 CARBON MONOXIDE (CO) CARBON DIOXIDE (CO,) 19x1D12 61000 x 101* MOLES MOLES MOLES 1.7 ppm 0.11 ppm 350 ppm R.T. - 8.6 YR R.T.- 0.16 YR R.T.{pro, prod.) - 6 YR ACCUMULATION: ACCUMULATION: ACCUMULATION: 3.7 x10" MOLES YR (some storage 250x10" MOLES YR1 20 pptvyr Hemisphere) or 1.4 ppm
LAND
LIVING BiOMASS 5x10'* MOLES C LIVING BIOMASS 250*101'MOLES C LITTER 0.8 X1016 MOLES C R.f. (photosynthesis) 0.07YEARS 18 R.T, (photosynthesis) 9.S YEARS STREAMS PARTICULATE CH2O 0.04 ppm, 0.5 x 10 MOLES C R.T. (titter fall) 2YEARS DISSOLVED ORGANIC C 0.7 ppm, 8.3 x1016 MOLES C TOTAL INORG. C 1i SOIL HUMUS 17 x10'« MOLES C DISSOLVED INORG. C 33 3.2x10 MGLESC R,T. (rivers) R.T. 32 YEARS DISSOLVED ORG.C 18 \«.,. 96.000YEARS ACCUMULATION 1 PARTICULATE ORO: C tsl(b6J 200 x 10" MOLES C YR- PARTlCOUTE INORG. C 14
SEDIMENTS RESERVOIR CARBONATE 52 x 10" MOLES C RESERVOIR CHjO 13 x 1020 MOLES C R.T, (input) ISO X106 YEARS "POSSIBLE MAXIMUM FLUXSS TODAY NET LOSS (fossil fuels) 444 x1012 MOLES C Figure 1: Global biogeochemical cycle of carbon. Fluxes in units of 1012 moles C per year (from Wollast et al, 1993). Carbonate equilibria
Inorganic substances dissolved in fresh water and seawater have their origin in minerals and the atmosphere (Stumm and Morgan, 1981). Carbon dioxide from the atmosphere reacts with water to form carbonic acid that subsequently reacts with minerals in rocks. The water may also lose dissolved carbon to the sediments by precipitation reactions. Carbon dioxide is added to the atmosphere by volcanic activity and through the combustion of fossil fuels. Atmospheric carbon dioxide is removed in the course of photosynthesis and released during the respiration and oxidation of organic matter. This makes CO2 a unique substance in biochemical exchange between water and biomass. Dissolved carbonate species participate in homogeneous and heterogeneous acid-base and exchange reactions with the lithosphere and atmosphere. Such reactions are significant in regulating the pH and composition of natural waters.
Weathering of limestone rocks, at the surface and underground, can be described by the reaction:
2+ CaCO3 + CO2 + H2O -> Ca + 2HCO3" [1 ]
This is a global reaction equation and according to Plummer et al. (1978), it comprises three specific reactions of calcite dissolution by the H2O - CO2 system:
+ 2+ CaCOs + H -> Ca + HCO3" [2]
2+ CaCO3 + H2CO3 -* Ca + 2HCO3" [3]
2+ 2 CaCO3 + H2O -> Ca + CO3 " + H2O [4]
From these equations it is evident that the CaCO3 dissolution reactions are dependent on the + 2 2+ concentration of the species H , HCO3", CO3 ", H2CO3 and Ca at the calcite surface.
When examining the chemical equilibrium between these species, we have to first consider the pure carbonic acid system before CaCO3 dissolution.
By transfer of CO2 from the gas phase into the solution, the aqueous species CO2(aq) is formed. The equilibrium between CO2(g)and CO2(aq)is determined by the constant KD:
CO2(g)->CO2(aq) K
Carbonic acid is formed by the reaction of CO2 with water:
CO2(aq) + H2O -> H2CO3° [6]
The total analytical concentration of dissolved CO2 is given by the equation:
[H2CO3°] = [CO2(aq)] + [H2CO3] [7]
The dissociation of carbonic acid is determined by the constant K^
+ HCQ H+ H2CO3°->H + HCO3- K .. l 3j[ J [8] 1 [H2CO3]
The second dissociation step, with constant K2l produces the carbonate ion:
+ 2 H COg HCO3" -* H + CO3 " K2 = l I l [9] [[HCO;] ]
10 The dissociation of water is described by Kw:
+ + H2O -• H + OH" Kw = [H ] [OH] [10]
Dissociation constant equations must be written in terms of ion activities, which are calculated as the product of ion concentrations and activity coefficients. These depend on the ionic strength of the solution. In natural karstic water the ionic strength is well below 0.1 and the Debye-Huckel theory is used for calculation of the individual ion activity coefficients.
Instead of thermodynamic equilibrium constants we can use the stoichiometric equilibrium constants, defined at temperature T as:
In K, = A, + B/T + QlnT [11] where A,, B, and Cj are empirical constants. Equilibrium constants for the carbonate solution system have been experimentally determined at different temperatures by many authors (Garrelsand Christ, 1965; Langmuir, 1971; Plummerand Busenberg, 1982).
2 The dissociation of carbonic acid into bicarbonate (HCO3~) and carbonate (CO3 ") ions depends on pH. The total amount of carbon (CT) in a CO2 - H2O system is calculated by the equation:
2 CT = H2CO3 + HCO3" + CO3 " [12]
A more quantitative approach to determine the activity of a single species of given total concentration is by fractionation factor (a). Molar fractions are defined as:
[13]
[HCO3] ^-^ [14]
[cor]
In Figure 2 these ionic fractions are presented as a function of pH for a very dilute solution. In 2 the region pH<4 virtually no HCO3" and CO3 " are present and only H2CO3° exists. With 2 increasing pH, carbonic acid dissociates to HCO3" and there is practically no CO3 ' present 2 below pH=8.3. Above pH=8.3 HCO3" starts to dissociate, until pH 12.0 where only CO3 ' ions exist. The data in Figure 2 represent with sufficient accuracy the conditions in karstic water systems. One important conclusion from this diagram is that in most of karstic waters the 2 species CO3 " can be neglected, since pH values higher than 8.3 are very rarely found in karstic waters.
Dissolution of CaCO3 is determined by the solubility constant Kc:
2+ 2 Ko = [Ca ]eq [CO3 -]eq [16]
2+ 2 where [Ca ]eq and [CO3 "]eq are the activities at equilibrium.
The saturation index (Q) of Langmuir (1971) is used for examining the evolution of chemical measurements in karst. The saturation index for calcite is:
[17]
11 Less than 1% of the total carbon is present as CO3 at pH values less than 8.3; therefore, for 2 carbonate saturation calculations, CO3 " must be calculated from measured pH, K2, and the proper activity coefficients.
From Equation 17 a solution will be in one of three conditions with respect to a given mineral solid phase: - forward reaction predominates, so there is net dissolution of the mineral. The solution is "undersaturated" or "aggressive" with respect to the mineral; Q<1 - there is a dynamic equilibrium, the solution is "saturated" with the mineral; Q=1 - back-reaction predominates, so there may be net precipitation of the mineral. The solution is "supersaturated"; Q>1
2 Figure 2: Ionic fractionation factors a0 for H2CO3°, a-i for HCO3" and a2 for CO3 " presented as a function of pH in the solution (from Dreybrodt, 1980).
2+ 2 Reaction of Ca with HCO3" and CO3 " produces ion pairs. In the case of natural karst water the + presence of CaCO3° and CaHCO3 molecules can be neglected because of their low concentration in the solution, usually <1% of the total amount of all other components (Langmuir, 1984). Electric neutrality of the solution has to be conserved in the calculation of equilibria of an electrolitic system. For electroneutrality, the sum of positive charges is balanced by the sum of negative charges. Electroneutrality of the pure H2O - CO2 - CaCO3 system (neglecting ion pairs) is defined by equation:
2+2+ 22-1 -i 2[Caa ] - [OHT] - [HCO3"] - 2[CO3 -] = 0 [18] where values in [] represents the concentration of the given ion in the solution (in mol dm"3).
In the region from pH = 4.0 to pH = 8.4, we can simplify this equation with an error of 2% to:
2+ 2[Ca ] = [HCO3"] [19]
12 Data obtained from carbonate aquifer waters show that Ca2+concentrations range from 1 to 5 10"3 mol dm"3, which is almost a hundred times higher than those predicted by the dissolution reaction [1] (Ford and Williams, 1989). The higher Ca2+ concentrations observed in the field result from carbonate mineral reaction with carbonic acid (H2CO3) that is derived from 2 + the respiration of organic matter. Calcite is dissolved and CO3 " is bound with H3O of carbonic acid to form bicarbonate ions HCO3". Therefore, in solubility calculations of calcite under natural conditions, the various forms of dissolved carbonate in water need to be taken into account. Concentrations of individual carbonate species depend on the pH of water.
The alkalinity of a water sample is determined by the total amount of titratable bases present. Total alkalinity (TA) of the carbonate system can be calculated using equation:
2 + TA = [HCOsi + 2 [CO3 ] + [OH] + [H ] [20]
Carbonate alkalinity (CA) is calculated as the sum of the contributions to alkalinity from 2 bicarbonate and carbonate ions, HCO3" and CO3 ":
2 CA = [HCO3~] + 2 [CO3 "] [21]
2 Dissolved inorganic carbon (Skirrow, 1975) is contained in CO2{aq), H2CO3, HCO3", and CO3 ". Relative proportions between these carbon species is influenced by temperature, pressure and pH. Bicarbonate ion concentration can be calculated by combining the equations [9] and [21]:
[HCO3~ ] = CA 1—- [22]
Using Equations [21] and [22] we can then calculate carbonate ion concentration:
K, 1 [23] K 2
and total CO2 concentration from Equation [8] and [22]:
fH+] L J [CO2 ]T = CA n [24] K, 1 + 2
The content of dissolved inorganic carbon [DIC] is obtained as the sum of the concentrations of the individual carbon species. Combining Equations [22] to [24] we get:
[DIC] = CA L * [25] 1 + 2 J^, H+ Figure 3 shows how these equations can be applied to solutions originating from calcite and dolomite dissolution. In the case of open system conditions, the solution is in direct contact with air. Therefore CO2 is able to dissolve from the air and replenish the H2CO3 whenever this is consumed. In this way equilibrium is maintained. An example of such an open system would be an open pool of water on limestone. On the other hand, in an ideal closed system there is no possibility of direct contact between air and the solution. Air and water react at open system conditions until the solution is saturated with dissolved CO2 and with H2CO3 and HCO3". The
13 solutions then migrate away from air and contact the carbonate minerals where no air is present, at closed conditions. By dissolution of CaCO3 the carbonic acid is consumed and cannot be replenished.
A real karst system is expected to perform as a combination of these extreme cases. Ideal open system conditions cannot be assumed in soils with low air volume and at low temperatures, when the rate of CO2 dissolution is greater than the CO2 production rate (Drake, 1984). However, detailed studies have obtained results which show that karst waters at equilibrium behave very much like one or the other of these ideal extremes.
CLOSED - AIR OPEN or or SEQUENTIAL- Cgas) CO, -CaMg(CO3) 2 7 t dolomite
\ Mg2* H*
WATER -^H,CO—- •i2wv33 ° H2CO3 / OH' HCO; I / H* HCO" ;
HCO3" CaHCO/ HCO, / 2 CO 2- Ca * 2.
•V"—r JL •V—"T ROCK- I CaCOk" i calcite X i L i T^ I I 1 I ' 1
Figure 3: Schematic presentation of karst waters system with the dissolved species and reactions leading to the dissolution of calcite and dolomite (from Ford and Williams, 1989). Ideal cases of open system and close system conditions are shown.
14 Isotope study
Isotopes are atoms of the same element whose nuclei contain different number of neutrons. Isotopes are either stable or radioactive. The number of known stable isotopes is about 300, while over 1200 unstable isotopes have been discovered or produced so far. Carbon has two stable isotopes: 12C and 13C. These isotopes are present in nature in the following abundance: 12C = 98.89%, C = 1.11% (Nier 1950). The most frequently used radioactive isotope of carbon is 14C (14C « 10"10% ) although short-lived 11'C has recently become useful as tracer in nuclear medicine.
Stable isotopes
Differences in chemical and physical properties arising from differences in atomic masses of an element are called isotope effects (Urey 1947; Hoefs, 1987). Chemical behaviour is determined by the extra nuclear structure of an element whereas the nucleus is responsible for its physical properties. The partitioning of isotopes between two substances with different isotope ratios (ratios of the isotopic concentrations) is called isotopic fractionation.
Isotope fractionation occurs in any reaction due to differences in the rates of reaction for different molecular species (Clark and Fritz, 1997). There are several ways to fractionate isotopes, such as physicochemical reaction under equilibrium or non-equilibrium (kinetic) conditions, or fractionation by molecular diffusion. The foundations of isotope fractionation were first laid by Urey (1947). The basis of physicochemical fractionation is the difference in the strength of bonds formed by the light versus the heavier isotope. Differences in bond strength for isotopes of the same element determine the differences in their reaction rates, as is illustrated in Figure 4. The zero-point energy is essentially the minimum potential energy of a molecular bond in a vibrating atom. Energy is consumed whenever atoms are pushed closer together or to separate them from this "comfort zone", or ideal interatomic distance. The energy required to break atoms apart differs for isotopically different molecules. The heavy isotope has a stronger bond, and thus requires more energy to dissociate than the light isotope. Consequently, lighter nuclei react faster.
Dissociation energy 0) light heavy Isotope Isotope Molecule +3 Dissociated CD
Q. • Energy minimum - light isotope
Energy minimum - heavy isotope
Interatomic distance ->
Figure 4: The potential energy-interatomic distance relationship for heavy and light isotopes of a molecule with different dissociation energies for two isotopes (from Clark and Fritz, 1997).The main phenomena producing isotope fractionations are: isotope exchange reactions and kinetic processes, which depend mainly on differences in reaction rates between molecules.
15 Isotope exchange includes all processes in which no chemical changes occur in the system but in which the isotope distribution changes between different chemical substances, between different phases, or between individual molecules. Isotope fractionation is generally considered under equilibrium conditions for which reliable fractionation factors can be calculated or experimentally measured. However, the condition of isotopic equilibrium requires that: - Chemical equilibrium exists, such that forward and backward reaction rates are the same. - The chemical reactions have advanced in both forward and backward directions sufficiently to ensure mixing of the isotopes between reactant and product reservoirs. - Both reactant and product reservoirs must be well mixed. If this is not the case then isotopic equilibrium exists only between reactants and products in the immediate vicinity of the reaction side. Isotope exchange reactions are a special case of general chemical equilibrium and can be written as:
aA1+bB2 •H>aA2+bB1 [26]
where the subscripts indicate that species A and B contain either the light (1) or heavy (2) isotope. The equilibrium constant for this reaction is:
[27] B
where the terms in parentheses may be, for example the molar ratios.
Usually, we are interested in the fractionation factor, rather than in the equilibrium constant. The fractionation factor (a) is defined as the ratio of the numbers of any two isotopes in the compound A divided by the corresponding ratio for the compound B:
aA-B=^- [28]
If the isotopes are randomly distributed over all possible positions in the compounds A and B, a is related to the equilibrium constant K by
a = K^ [29] where n is the number of exchanged atoms. For simplicity, isotope exchange reactions are written in such a way that only one atom is exchanged (Equation [27]). In these cases, the equilibrium constant is identical with the fractionation factor, K=ct.
For example, speleothems are inorganic precipitates of calcium carbonate, generally calcite, formed from dilute aqueous solutions (Schwarcz, 1986). They are formed under conditions which ensure that oxygen atoms are in isotopic equilibrium with water from which they are precipitated. This is represented by the following exchange equation:
16 18 18 16 -CaC O3+H2 O<^-CaC O3+H2 O [30]
18 16 The fractionation factor for the exchange of O and O between water and CaCO3 is given by:
16 18 7o16/ o a CaC03-H,0 -
H20
This equilibrium constant depends on temperature and is independent of all other environmental variables (Epstein et.ai, 1953). Therefore, if the calcite is deposited at equilibrium, its isotopic composition provides a record of the temperature of deposition.
Kinetic effects are the second main phenomena producing fractionations. Knowledge of kinetic isotopic effects is very important, because it can provide information about details of reaction pathways.
The accepted unit of isotope ratio measurements is the delta value (8), expressed as deviation in permill (%o) from a standard. The 8-value is defined as:
R §[%„] = (sample)-R(standard)1000 [32] R (standard)
where R represents the isotope ratio. If 8A>8B, we speak of A being enriched in the rare isotope or being "heavier" than B.
Worldwide standards in use for the determination of isotopic composition in many natural materials are recommended by the International Atomic Energy Agency (IAEA). The standard samples should have an isotope ratio within the range of natural variations. The most widely used carbon isotope standard is Belemnitella Americana from the Cretaceous Peedee formation, South Carolina, USA (PDB - Pee Dee Belemnite) with the value R(standard)=0.0112372 (Craig, 1953). Standard for oxygen and hydrogen isotopes is mean ocean water at depth 1m and temperature 298K (SMOW- Standard Mean Ocean Water).
The two main carbon reservoirs, organic matter and sedimentary carbonates, are isotopically quite different from each other because of the operation of two different reaction mechanisms: 12 1. kinetic effects during photosynthesis, leading to the depletion of C in the remaining CO2, and concentrating the light 12C in the synthesized organic material; 2. chemical exchange effects in the system: atmospheric CO2-dissolved HCO3~, which leads to the enrichment of 13C in the bicarbonate.
Carbon-13 is an excellent tracer of carbonate evolution because of the large variations in the various carbon reservoirs. In Figure 5 the 513C-values of some important carbon compounds are 13 schematically compared. The evolution of DIC and 8 C of DIC begins with atmospheric CO2 13 (5 C « -7%o PDB). Photosynthetic uptake of atmospheric CO2 is accompanied by a significant 13 depletion in C. This occurs during CO2 diffusion into the leaf stomata and dissolution in the cell sap and during carboxylation (carbon fixation) by the leafs chloroplast, where CO2 is converted to carbohydrate. The combination of these fractionating steps results in a 5%o to 25%o depletion in 13C. The amount of fractionation depends on the pathway followed. Three principal photosynthetic cycles are recognised: the Calvin (C3) cycle, the Hatch-Slack (C4) cycle, and the Crassulacean acid metabolism (CAM) cycle. The C3 pathway operates in about 85% of plant species and dominates in most terrestrial ecosystems. C3 plants fix CO2 with the Rubisco enzyme, which also catalyses respiration through the reaction with oxygen. Most C3 plants 13 have 5 C values that range from -24%0 to -30%o with an average value of about -27%o (Vogel, 1993). The natural vegetation in temperate and high latitude regions almost exclusively follow the C3 cycle. Plants using C4 cycle add an initial step where the enzyme PEP carboxylase delivers more carbon to Rubisco for fixation. The result is smaller C fractionation during 13 carboxylation. C4 plants have 5 C values that range from -10%o to -16%o with the mean value of about -12.5%o (Vogel, 1993). The C4 cycle is present in less than 5% of floral species, but is dominant in warm-climate open systems such as tropical and temperate grasslands. CAM photosynthesis is favoured by about 10% of plants and dominates in desert ecosystems with plant species such as cacti. They have the ability to switch from C3 photosynthesis during the
17 day to the C4 pathway for fixing CO2 during the night. Their isotopic composition can span the full range of both C3 and C4 plants, but usually it is intermediate.
As plants die and vegetative material accumulates within the soil, decay by aerobic bacteria converts much of it back to CO2. Soil CO2 concentrations are 10 to 100 times higher than that of 13 the atmosphere. Microbially respired CO2 has similar S C as the vegetation itself. However, a diffusive fractionation occurs by outgassing of CO2 along this steep concentration gradient. The 13 8 C of the soil CO2 in most C3 landscapes is generally about -23%o, while in soils hosting C4 vegetation it is about -9%o.
Marine carbonates have 13C contents similar to the reference PDB with 813C « 0%o, although values can deviate from PDB by several permill. The S13C-values of carbonate minerals and metamorphic rocks can also vary over a large range (Figure 5).
The 813C value of carbonate minerals is about 15%o more enriched than the DIC from the soil zone. As carbonate is dissolved, 813C DIC evolves to more enriched values. How far it evolves and how much carbonate is dissolved depends on the "openness" of the system. If dissolution takes place under fully open system conditions, then the S13C DIC will be controlled by the soil CO2. This is because of the continual exchange of CO2 between DIC and the soil atmosphere, which is a much larger reservoir than DIC in the non-saturated zone. For open system 13 conditions, final 8 C DIC values are enriched by about 7%o compared to the original soil CO2. If the dissolution takes place under fully closed conditions, then the 813C will simply be diluted by DIC originating from the carbonate mineral source (S13C » 0%o).
-60 -50 -40 -30 -20 -10 10 20 Atmospheric CO,
Plants C3 • Soil COz -
Mantle CO2 -<«miIIII>
Metamorphic CO2 •
ty Atmospheric CH4
i (nTOTTTTTTTnrrmi i Biogenic CH4
Thermogenic CH4 Meteorite graphite <2)> Criondrite carbonate -4- -60 -50 -40 -30 -20 -10 10 20 513C %o VPDB
Figure 5: Ranges for 813C values in selected natural compounds (from Clark and Fritz, 1997).
18 Radioactive isotopes
Radioactive isotopes are unstable and undergo spontaneous disintegration, which ultimately leads to stable isotopes.
14C is created in the upper atmosphere when cosmic radiation interacts with 14N. The 14C is then 14 oxidised to CO2 (radioactive carbon dioxide) and distributed evenly through out the global atmosphere. About 100 14C atoms are being generated over each square centimetre of the earth's surface each minute. The 14C production is not uniformly distributed over all latitudes. The highest production occurs at mid-latitudes. In addition, the output of CO2 from fossil fuels, which contain no 14C, has tended to dilute the natural level (Suess, 1955).However, recent injections of nuclear bomb produced 14C have enhanced the atmospheric 14C level in the opposite direction.
Radioactive decay is a completely random process. Decay is spontaneous and unaffected by external influences. It occurs at a unique rate for each radioisotope, defined by its half-life T^, which is defined as the time required for one half of the radioactive atoms to decay. The decay follows the exponential law:
M N=Noe- [33] where N is the number of radioactive atoms present at time t, No is the number present at the start of decay, and X is the decay constant.
Radiocarbon decays as shown in the Figure 6.14C activity is measured by detecting p" particles. Its half-life is 5730 + 40 years.
Dissolved inorganic carbon in surface water is derived partly from atmospheric CO2, but principally from decaying organic matter, either through bacterial decomposition or respiration. Both sources contribute a component of H2CO3, which is labelled with an atmospheric level of 14 2+ C activity. Carbonic acid then reacts with calcite in the soil to produce HCO3" and Ca ions as shown in equation:
2+ CaCO3 + H2O + CO2 = Ca + 2HCO3" [34]
Subsequently, the water percolates through limestone before reaching a cave and in this way the 14C activity of the water is lowered via dissolved inorganic carbon exchange with the limestone.
It is assumed that when calcite precipitates in a cave the system is isolated, i.e. carbon atoms in the crystal lattice cannot exchange with those in HCO3" solution passing over or through the calcite. Radioisotope decay thus can be used to measure the time elapsed since precipitation occurred.
Half of the carbon atoms in the HCO3" should derive from bedrock CaCO3 (which contains no 14 C) and half from soil air or other sources of modern CO2. However, during deposition heavier isotopes such as 14C may pass preferentially into the solid phase, enriching it slightly.
The Primary Modern Reference Standard for radiocarbon dating is 95% of the count rate in 1950 of a batch of oxalic acid from the National Bureau of Standards of USA (Stuvier and Polach, 1977). The activity of the primary reference standard corresponds to wood growth in AD 14 1890 (therefore uncontaminated by fossil fuel CO2) whose activity was corrected for C decay to correspond to AD 1950. By international agreement the isotopic ratio of 13C/12C of the oxalic acid, as determined mass by spectrometer, is taken to be 813C = -19%o against PDB limestone, the internationally accepted 513C reference standard (Craig, 1953).
19 \^:<:#P*1: '"•'•-.
Figure 6: The radioactive decay of 14C.
20 Rain water dissolves carbon doxide forming carbonic acid
-!:yT .1;sy li^l I»UEIU
,„ ... c>onSoil water aisowesdisoh/es more carboncaroon ••> - .. ,. , ... •<• .. •. • •• .. *• „ ••
Solution of CaCO3 at bedrock surface ±y
Carbon dioxide loss in cave atmosphere and evaporation lead to CaCOi deposition
Figure 7: Schematic presentation of types of caicite and aragonite deposits in the cave system (from Hill and Forti, 1986). The formation of speleothems
The different types of speleothems found in limestone caves are formed as deposits of calcium carbonate (Moore, 1952). The shapes of speleothems are influenced by five basic hydrologic mechanisms: dripping, flowing, and seeping as well as pooled and condensation water. Other factors that influence speleothem growth type and shape are bedding, limestone porosity and evaporation. The principal types of speleothems are shown in Figure 7.
Speleothems growing primarily from dripping water are enlogated in the vertical direction of dripping, and examples are stalactites and stalagmites. Flowstone is deposited from water flowing over cave walls or floors in layer upon layer. Coralloids grow from thin films of splashing or seeping water. Helictites grow by capillary water seeping through tiny canals and are twisted in every direction due to ventilation in the cave or irregularities in crystal structure. Cave rafts formed on the surface of pools are flat, planar speleothems. Rims lining bedrock or other speleothems are deposited in places where water condenses due to temperature and/or humidity changes. Compound speleothem forms are found in places where two or more of these basic mechanisms are present simultaneously.
Stalagmites are the most common objects used in isotopic studies. In Figure 8, a schematic cross-sectional view of a typical stalagmite, consisting of a series of stacked, inverted cup-like deposits is shown (Schwartz, 1986). These growth layers represent successive generations of mineral precipitate with more or less uniform morphology. The axial cross section through the stalagmite intersects the planar tips of growth layers and thus provides a continuous stratigraphic sequence of the stalagmite. Each growth layer can be distinguished from older/younger deposits by colour bands or some textural differences. The colour is due to the presence of organic matter bound to the calcite crystals, and also due to traces of Fe and Mn (Gascoyne, 1977). This enables us to sample a single growth layer using a fine drill, as is shown in Figure 8. From the growth layer textures we can recognise possible hiatuses in growth, which are usually marked by dissolution and corrosion in older layers. Accumulation of insoluble residues can be detected, consisting of clay minerals and hydrous oxides of Fe and Al.
_ sample sites f for axial profile ill
crystal boundaries single crystal domain
_ sample drilled from growth layer
single growth layer
Figure 8: Cross-sectional schematic view of a typical stalagmite, with marked growth layers (from Schwarcz, 1986).
22 Speleothem formation can be described in more detail according to Hendy's model (Hendy, 1971). Precipitation of CaCO3 on speleothems is the result of various processes, as shown in Figure 9. In the first stage, CO2 is taken up from the atmosphere via photosynthesis. CO2 is then released into the soil by respiration of plant roots and by oxidation of dead plant material. The partial pressure of CO2 in the soil atmosphere can be 100 times greater than that of outside air. Soil CO2 is dissolved within rain water infiltrating through the soil, and this solution dissolves CaCO3 when it contacts the limestone bedrock. The solubility of CaCO3 depends on the partial pressure of CO2 in the solution, as well as upon temperature and pressure conditions.
Several models deal with the carbon isotope composition of precipitating carbonates (Hendy, 1971; Salamons and Mook, 1986; Dulinski and Rozanski, 1990), the most comprehensive being that of Hendy. He analysed the CaCO3 precipitation process using a mathematical description of precipitation under equilibrium conditions for both open and closed system. Hendy introduced the kinetic treatment of carbon isotopic exchange between CO2 in cave air, CO2 in the solution and HCO3" ions in the solution. He also determined the necessary physical and chemical conditions for equilibrium fractionation of carbon and oxygen isotopes during speleothem growth. Hendy's model assumes simultaneous outgassing of the solution and precipitation of CaCO3 from the saturated solution.
There are two alternative processes affecting the isotopic composition of precipitated calcite in speleothems; these are defined as open or closed system conditions (Figure 9). In an open system the solution dissolving the carbonate is always in contact with an excess of soil CO2, while in a closed system there is no contact between the solution and soil CO2 prior to emergence of the solution into the cave (Hendy, 1971). A real situation is probably an intermediate between these two extreme cases.
In the closed system, carbon in solution originates from the soil atmosphere and from dissolved limestone. In the open system, limestone carbonate is dissolved in the presence of CO2 gas and isotopic equilibrium is established between CO2 gas in the soil and the saturated calcite solution.
When a saturated solution of calcite eventually seeps into a cave, and if the partial pressure of CO2 in the solution is greater than that in the cave air, CO2 will be lost from the solution. This solution will became supersaturated with respect to calcite and the deposition of calcite or aragonite follows. The calcite isotopic composition is determined by three processes according to the CaCO3 deposition (Hendy 1971): equilibrium fractionated deposition, kinetically fractionated deposition and deposition by evaporation of water (Figure 9).
In first process, the rate of CO2 loss from the solution is slow enough to maintain isotopic equilibrium between the carbon species in the solution. The slowest reaction step is the dehydration of HC03~-ions and the hydration of aqueous CO2. Only in this step can oxygen be exchanged between water and carbon compounds in solution, so equilibrium between these species is necessary to obtain calcite deposition in oxygen isotopic equilibrium with the water. The species remaining in solution will be enriched in 13C and 18O. Isotopes are exchanged also between CO2 of cave air and carbon species in the solution.
In the second process, kinetically fractionated deposition, chemical and isotopical equilibrium between aqueous CO2 and HCO3" ions is not maintained due to the rapid loss of CO2 from solution. Deposited calcite will be enriched in 13C and 18O isotopes.
In the third process water evaporates from the solution, the solution becomes supersaturated, and calcium carbonate deposition occurs. When evaporation of water from a solution flowing over a speleothem occurs, precipitation will continue even after the partial pressure of CO2 in the solution will reach that of the cave air. In Postojna Cave the evaporation of water is not important for speleothem growth because of the high relative humidity and low ventilation. Evaporation of water is significant in speleothem formation only near the cave entrance.
23 CO; RAIN
PHOTOSYNTHESIS
CO2 FROM RESPIRATION
OPEN SYSTEM CLOSED SYSTEM 2
SATURATED SOLUTION OF CALCITE ENTERING CAVE
KINETIC FRACTIONATION
'EVAPORATION EQUILIBRIUM OF WATER FRACTIONATION AND' EXCHANGE WITH ATMOSPHERE
APPROACHES ISOTOPIC EQUILIBRIUM WITH THE ATMOSPHERE
Figure 9: A schematic diagram showing the dissolution of limestone and the formation of speleothems (from Hendy, 1971).
24 Speleothem growth rate theory
The most important mechanism for speleothem deposition is the degassing of ground waters containing elevated CO2 concentrations (Holland et al., 1964; White 1976; Ford and Wiliams, 1989). The elevated CO2 is derived from the high partial pressure of CO2 in the soil atmosphere, where CO2 is generated by organic matter decay and root respiration (Dorr and Munnich, 1986). Other possible growth mechanisms include common ion effects (Atkinson, 1985), evaporation processes (Harmon, 1981) and temperature gradient effects (Dreybrodt, 1981).
Dreybrodt was first to describe stalagmite growth rate as a function of the CaCO3 and CO2 content in the solution as well as of the quantity of water dripping onto the stalagmite. His theory was used to analyse our samples and is presented in more detail below. Using Dreybrodt's model, speleothem growth rates were determined for those that were sampled in Postojna Cave.
Dreybrodt's theory (Dreybrodt, 1980) first introduces the basic mechanism of carbonate deposition: CaCO3 is dissolved by water percolating through small fissures in the rock, and CaCO3 is then deposited when the solution enters the cave. Rainwater absorbs CO2 from the air and by passing through the soil absorbs more CO2 from the soil atmosphere. Since the partial pressure of CO2 in the soil is high, up to 0.1 atm, the solution can contain a high amount of CO2. The solution therefore readily dissolves CaCO3 during infiltration. Upon reaching the cave where the CO2 partial pressure in the cave air is low, degassing of CO2 occurs producing a supersaturated CaCO3 solution. Deposition of CaCO3 is determined by the behaviour of this solution. When a drip of this solution falls onto a stalagmite, part of the drop flows quickly down the stalagmite and the rest forms a thin film of water. This film remains stable and slowly descends the stalagmite. A stagnant thin layer of supersaturated CaCO3 solution covers the planar stalagmite surface with another layer on top, where CO2 degassing takes place (Figure 10).
Calcite precipitation is summarized by the reaction:
2+ Ca + 2HCO3" -> CaCO3 + CO2 + H2O [35] following several individual steps (Figure 11).
There is a dynamic equilibrium between CO2 in the solution and in the cave air, so whenever the CO2 concentration in the solution changes, this affects the concentrations of the carbon 2 species: H2CO3, HCO3" and CO3 '. Conversion of these ions between species following the reactions in Figure 11 is fast compared to the diffusion of these ions. Therefore the rate of CO2 degassing is controlled by diffusion of CO2 molecules in solution. Another important process is 2+ 2 the deposition of Ca and CO3 ~ at the crystal surface (or within the solution of CaCO3). By 2+ 2 deposition of Ca and CO3 " ions, their concentration decreases and the deposition rate slows down. Deposition is controlled by diffusion of these ions from the bulk solution to the crystal surface. From the deposition of 1 mole of CaCO3, 1 mole of CO2 is produced at the crystal surface (Franke, 1975). This effectively reduces the supersaturation of the solution and slows down the deposition rate at the surface of the crystal.
Deposition rate is determined by these independent processes (Dreybrodt 1980): 1. diffusion of CO2 in the solution 2 2. diffusion of Ca , CO3 " in the solution 3. deposition of CaCO3 on the crystal surface 4. production of CO2 at the crystal surface
2+ Carbonate deposition is determined by the concentration of Ca and CO2 at the crystal surface. Therefore, the transport of CO2 to the air is important for crystal growth rate and can be divided into three steps: 1. diffusion of CO2 through the solution to the surface 2. transfer of CO2 from the solution surface into the air 3. diffusion of CO2 through the cave air.
25 CAVE AIR
( co2 j x=R (film thickness) SOLUTION
X=0
Figure 10: Model of a stagnant water film at the stalagmite crystal surface, developed by Dreybrodt (1980).
AIR CO2
H20 SOLUTION (water film) H2CO3
+ -H30
Ca(HCO3)2 HCO3"
+ -H30 ||-H20
Ca,2: + 2 CaCO3 *= * CO3 "
Ca2+
CRYSTAL
Figure 11: System of reactions in the CaCO3- CO2 - H2O - air system leading to deposition of CaCO3 on a speleothem (from Dreybrodt, 1980).
26 The first step is much slower compared to the other two (Picknett et al., 1976), so we can assume an equilibrium between the CO2 concentration at the air-water boundary and the CO2 partial pressure in the cave air. The production of CO2 in the solution results from the 2 conversion between the carbon species H2CO3l HCO3" and CO3 ", where the H2CO3 to CO2 conversion is the slowest step. However, diffusion of CO2 through the solution is even slower and it controls the degassing process. Another important mechanism is the transport of Ca2+ 2 ions to the crystal surface. For precipitation to occur CO3 " ions are required and these are produced by fast chemical conversion from HCO3". The concentration of HCO3' ions is normally 2 2+ much higher than the CO3 " concentration in solution, so this makes the Ca migration the controlling mechanism for the CaCO3 deposition rate.
Dreybrodt derived a linear rate equation to describe the precipitation and dissolution of calcite (Dreybrodt, 1980):
2+ 2+ 2 1 growth rate = a ([Ca ]eq - [Ca ]) (mmol cm' s" ) [36]
where a includes parameters such as film thickness, cave air CO2 concentration (pCO2), temperature and flow regime. This equation is valid when the calcium concentration [Ca2+] is 2+ more than 20% of the equilibrium concentration of calcium, [Ca ]eq.
The conversion reaction of CO2 to H2CO3 is strongly dependent upon temperature, and therefore so is the growth rate. The pCO2 in cave air controls the diffusion of CO2 across the water film. Variations of water film thickness are important as well. The final development of this model was presented in paper by Dreybrodt (1988). He developed the equation for growth rate determination: d[ca2+] R([Ca-]eq-[ca-])l-e^ [37] dt where R is the water film thickness (Figure 10) and t is the time interval between individual drips. Parameter a is defined in the same way as in Equation 36.
The authors found out that for high drip rates this equation simplifies to Equation 36.
27 Speleothems and climate
Air temperature in a cave is related to average annual temperature on the surface above the cave. Isotopic paleotemperatures derived from speleothems reflect surface temperature variations throughout the period over which the speleothem was deposited.
From the oxygen isotope composition of carbonate deposits we can calculate the temperature at which the calcium carbonate was precipitated (Hendy, 1971; Harmon and Schwarcz, 1981; Talma and Vogel, 1992). This is because the isotopic fractionation that occurs during mineral precipitation from an aqueous solution is temperature dependent. This temperature dependence was first determined by O'Neil et al. (1969). The temperature of deposition can be calculated from the temperature dependence of the calcite-water 18O fractionation by the equation (Friedman and O'Neil, 1977):
1000lnac_w = 2 — 2.89 [38]
where ac.w is the fractionation factor for oxygen between calcite and water
18 ac-w = (S^CUHB + 1000) / (8 Owater + 1000) [39]
However, a number of conditions need to be fulfiled before meaningful paleotemperatures can be derived from such measurements (Hendy, 1971; Schwartz et al., 1976): 1. The carbonate precipitated needs to be in isotopic equilibrium with the water at the time of formation. 2. The oxygen isotopic content of the water must be known. 3. No diagenetic alternation of the isotope record must have occurred. 4. The deposit must be properly datable.
In a cave with poor ventilation, the release of CO2 from cave drip water is slow enough that the precipitation of calcite occurs under equilibrium conditions. In this case, the cave temperature is equal to the average annual temperature on the surface. If the stalagmite is compact and nonporous the probability of subsequent isotopic alternation is small. Such speleothems can also be dated with uranium and radiocarbon dating methods. The main problem one faces in attempting temperature calculation from the 18O content of a stalagmite is the determination of the isotopic composition of the water during the calcite precipitation. The 18O content of present cave drip water can be readily measured. Repeated sampling is necessary to establish any possible seasonal variations which, on average, are equal to the mean annual 18O content of precipitation in the area (Goede et al., 1982; Younge et al., 1985). These mean values can be used to estimate whether isotopic equilibrium exists during calcite precipitation at present. The 18O value of drip water in the past is usually unknown and it must be estimated using a model which describes the evaporation-transport-precipitation processes of moisture in the water cycle (Gascoyne et al., 1980), or calculated from the deuterium content of fluid inclusions in stalagmites (Schwartz etal., 1976; Schwartz and Younge, 1983).
28 Modeling
The impact of atmospheric nuclear bomb tests of the late 1950s and 1960s on atmospheric CO2 activity was first reported by Rafter (1957). Since then 14C has been employed as a tracer in studies of different reservoirs of carbon, e.g. in the atmosphere, the biosphere and in the ocean. 14 By C tracking it is possible to trace the atmospheric CO2 in the environment and to determine CO2 sources and sinks.
To determine the provenence of CaCO3 in the growth laminae of a speleothem Genty developed a model for the determination of CaCO3 contributions from various sources (Genty and Massault, 1997). This model was also employed in our analysis of the stalagmite from Postojna cave, in an attempt to explain the difference between the stalagmite 14C time series and the modern atmospheric 14C activity due to nuclear bomb tests (Genty et. al., 1998). Details of the model are explained below.
There are two principal sources of carbon bound in a speleothem: carbon derived from the soil CO2 above the cave, and carbon derived from the dissolution of carbonates both from the limestone bedrock where the cave is developed and from secondary carbonate deposits within the unsaturated zone. The CO2 contained in the soil is produced by root respiration and decomposition of organic matter by microbes. Soil CO2 is of recent origin and is influenced by atmospheric CO2 (Dorr and Mtinnich, 1986) and the type of vegetation. The carbon contained in limestone is generally of much older provenence, and due to radioactive decay the 14C content in the bedrock is negligible. On the other hand, the 14C activity of secondary carbonate deposits can be significant.
By measurement of 14C in growth laminae of a speleothem, the age of the deposits can be estimated. The results need to be crosschecked with the U/Th method for age determination because the 14C method has an error due to the unknown dead carbon proportion (dcp). When the age of a speleothem is determined using U/Th method, we can then track the changes in 14C content in a speleothem through the years, and so determine the 14C atmospheric activity in the past.
Uncertainties in dead carbon proportions are the cause of major problems in these measurements. One reason is that the initial 14C activity in a given speleothem is not known. This means that we can only estimate the dcp or the dilution factor q (Gewelt, 1986). Another uncertainty is introduced by possible variations of dcp with time. A list of dilution factors determined by various authors at different sites is presented in a paper by Genty (Genty and Massault, 1997).
The dead carbon proportion can be calculated using two measured values of 14C activity (Genty and Massault, 1997). The first value must be determined in a calcite speleothem deposit of 14 14 known age (a Cm), and the second value measured in the atmosphere (a Catm)-
Atmosphere 14C activity has varied greatly in the years after the nuclear bomb tests (Levin et al., 1992), therefore the age of individual layers in a stalagmite need to be determined accurately. This is possible with speleothems that are laminated annually.
Estimation of karst water age is difficult because the water can be composed of differentwaters with very different retention times. Rapid transit often occurs shortly after a storm event (few hours to few days). On the other hand, in dry periods karst water is coming mostly out of storage in porous media with undetermined age (Ford and Williams, 1989). Storage characteristics of an aquifer can be determined by spring hydrograph analysis.
Several conceptual models have been developed including different characteristics of water flow: porous (or diffuse), fissure and conduit. These flow models allow for a schematics description of a karst aquifer (Attkinson, 1985; Hobbs and Smart, 1986).
Flow characteristics of a stalactite can be described by both fast and slow components (Pitty, 1966; Baker et al., 1997, Destombes, 1997, Reicher and Trimborn, 1995). For quantitative
29 determination of fast and slow components it is necessary to measure the chemistry of several stalactite drips with continous dripping rate.
Water age can be assumed as of the date of sampling since cave roof thickness is low enough and because flow rate increases after a short delay (few hours to few days) following a storm. Dilution with older water deposits is still possible.
The dead carbon proportion is calculated from the measured 14C activities of a deposited lamina 14 14 (a Cm) and of the atmosphere (a Catm) using the equation (Genty and Massault, 1997):
a m dcp = [i- u° -100% [40]
For groundwaters no 813C normalization was done.
Dead carbon can also be expressed in values of the dilution factor (q), which is connected to the "hard water effect" (Thrope et. al., 1980; Vogel, 1983; Gewelt 1985; Bastin and Gewelt, 1986; Fontes, 1992).
q dcp Dead carbon calculation by equation [38] does not take into account fractionation processes and so it can be considered only as an apparent proportion.
The dead carbon effect is the dilution of the 14C concentration in seepage water due to limestone dissolution. The 14C activity in DIC decreases as dissolution proceeds because the 14 DIC C activity is in equilibrium with the soil CO2 before dissolution, while after dissolution DIC 14C activity approaches that of the calcite. There may also be some discrepancies due to possible fractionation processes.
In Figure 12, the pathways of 14C from the atmosphere to the cave are shown. The transit of CO2 can be divided into several stages (Genty et al., 1998):
1. Production of CO2 in the soil and epikarst from slow and fast cycling reservoirs with contributions Ci and C2, respectively.
In the first reservoir, CO2 is produced by the slow decomposition of soil organic matter (dead roots, leaves, wood). Because of the mixture of organic matter of different ages, the resulting 14C activity represents an average of the atmospheric 14C activity over a period of Y^ears.
In the fast cycling reservoir, CO2 is produced by respiration of plant roots as well as very fast 14 decomposition of soil organic matter. The CO2 C activity is similar to the activity of the atmosphere over the last few years.
14 Soil CO2 activity (a Cg) is calculated as sum of both slow and fast contributions (with C1+C2=1):
a °atmi 14 14 a Cg = C, ^— + C2a Catmn [42] Y1
In the first term of equation [40] the mean atmospheric activity is calculated as the integrated 14 activities of Yi years (a Catm0 divided by Yi. The second term is the activity in the year of sampling.
30 Figure 12: Schematic presentation of C transit from the atmosphere to the cave (after Genty and Massault, 1997).
31 2. Dissolution of soil CO2 and the formation of bicarbonate
Assuming an equilibrium between the soil CO2 and the dissolved inorganic carbon, we can calculate DIC activity by the equation (Saliege, 1984):
14 14 13 13 a CDICo = a Cg +0.23(5 CDIC -5 Cg) [43] and fractionation of 813C by the equation (Mook, 1974):
13 13 3 5 CDlCo = 5 Cg +(-l9i§ - + 23.89) [44] Co
In these equations DIC0 represents dissolved inorganic carbon (mainly in the form of HCO3") 14 prior to the dissolution of limestone causing dead carbon dilution. Activity a CD|Co is expressed 13 as percent of modern carbon (pMC), 5 CDico in permil of PDB standard (%o PDB), and T is absolute temperature (°K).
3. Dissolution of old limestone deposits, which do not contain any 14C
This results in the reduction of 14C activity of DIC:
DICO+tiooJa Cc [45]
In the equation, DIC is the dissolved inorganic carbon (mainly in the form of HCO3" after the 14 dilution with dead carbon. Activity a Cc stands for limestone activity, which is by definition 0 pMC.
Dissolution of limestone also affects the 813C of the DIC:
'-°+llWj5i3Cc 1461
14 4. Mixing effect is employed as an apparent quantity used to smooth the yearly signal a CD|Ci.
This mixing can be attributed to mixing of waters with different residence times as well as to secondary dissolution of calcite in limestone cavities above the cave. Smoothing by mixing effect is given by the equation:
Y2 14 * CDICi •v [47]
where the parameter determining the mixing effect is the period Y2, given in the number of years before the measurement over which the DIC 14C activity is averaged.
5. The precipitation of CaCO3 in the cave
The fractionation of 14C and 13C in the calcite of a speleothem (csp) is given by the ^equations:
14 14 13 13 a Ccsp =a CDIC +0.23(5 Ccsp -5 CDIC) [48]
[49]
32 SITE DESCRIPTION
Postojna Karst
The Postojna Cave system is composed of 23 km of accessible channels and galleries and is a part of the karst region of the underground Ljubljanica river. The Postojna Karst is located at the northwestern part of the Dinaric Karst (L=45°46'N, I=14°12'E, Z=529m). It is limited to the north by the promontory of the Julian Alps, and to the south by the Adriatic Sea (Gospodaric, 1988). Postojna Cave is located between ponor (sinking) channels near Postojna and spring channels near Planina, which are hydrologically connected. The karst region has a variable climate under the Mediterranean, Alpine and Pannonian influences.
The hill under which Postojna Cave is situated is covered by a pine tree forest and grass. Average annual temperature is 8°C and mean annual rainfall is 1500 mm.
Geology of Postojna Karst
The Postojna Karst (Figure 13) is composed of Cretaceous carbonate rocks (Gospodari6, 1976). On the northern and northeastern part it is covered by Triassic dolomite and Jurassic dolomite and limestone of Hrusica Mts., while on the western and south-western part the Karst is buried by Eocene flysch of Pivka Basin. Postojna Cave lies within the Upper Cretaceous rocks.
Lower Cretaceous is composed by bedded and thickbedded limestones and dolomites and limestone breccia nests. The transition to the Upper Cretaceous strata is seen in the Planina Cave where limestone with chert passes to a massive Cenomanian limestone, and where Caprinidae, Chondrodonatae and corresponding microfauna have been found. Turanian is composed by bedded limestone. Senonian rocks have similar lithological structure. On the Cretaceous rocks only a few erosional remnants of Paleocene rocks have remained followed by marl, sandstone and flysch conglomerate of Eocene age (Figure 14).
The tectonic structure of the Postojna Karst is expressed in northwest-southwest oriented Postojna anticline, Studeno syncline and Bukovje folds. The first two folds have been found in the galleries of Postojna Cave. In Planina Cave the Lower Cretaceus beds are gently inclined towards the west, southwest and northwest. The fault systems trending northwest-southeast, northeast-southwest, and north-south are considerably more frequent on the crests than on the limbs of the mentioned folds. Numerous faults and wrench-faults have the same direction. All these structures are characteristic of the tectonic structure of the High Karst, to which the Postojna Karst belongs. The Predjama over-thrust can be considered too; there the dolomitic Hrusica Mts. are over-thrusted from the north towards the Postojna Karst.
Speleological characteristics of the Postojna Karst
In the Postojna Karst region more than 60 caves of different types and size are known. The majority are situated under the surface of the Postojna morphological plateau (600 - 630 m) including a 4 km wide belt of Postojna Karst near the Pivka Basin.
Numerous cave sections in Postojna Cave represent the remains of old sinking channels, which are collapsed in or filled up by sediments today. They are accessible across smaller corrosive potholes or collapsed ceilings (Figure 15). The passages are composed of one or more levels, in many places connected by inclined, fissured galleries of younger corrosive origin. Some vertical potholes, ending at the altitude of the horizontal galleries, have been noted separately. The transitive and connecting galleries in both Postojna Cave and Planina Cave are illustrated by single columns. The depths of present collapsed dolines and the ceiling heights of some collapsed halls have been plotted.
33 On the ponor side the horizontal caves are developed in the same altitude span (550 - 510 m) as the high galleries in Planina Cave (450 - 490 m above the sea level). On the ponor side, there are several shorter caves with more expressive central galleries between 540 - 525 m, while they are connected on the spring side in the uniform high channel. On the ponor side the caves are situated along an 8 km long slope between Postojna and Studeno on the contact between the flysch and limestone. On the spring side they are united in single gallery reaching the surface of Planina Polje near the contact between the limestone and dolomite. Therefore, we can expect in Planina Cave more and better facts about the geological history of the whole underground system than in the dissociated caves on the Postojna part, where the shapes and sediments of former water channels are not so easy to determine due to breakdowns and concretions.
Figure 13: The Ljubljanica River Basin including the Postojna Karst between the Pivka Basin and Polje of Planina (from Gospodaric, 1976).
34 (7 \—~ v V i //
GEOLOSKA KARTA POSTOJNSKEGA KRASA
3 ESB E » R^, K, B
* )«ESS?a fj 9 Y-r-'-A \i" U
Figure 14: Geological map of the Postojna Karst with the Postojna Cave System.Legend: 1-alluvial deposits, 2-Quaternary sediments in Pivka Basin, 3-marI sandstone, conglomerate and breccia, flysch rocks-Eocene, 4-limestone, limestone breccia, red and grey marl-Paleocene, 5-thick bedded limestone with Cheramospherinae, Sabiniae and Hippurites- Senonian and Maastrichtian, 6- bedded and non bedded limestone with chert and radiolitid fauna-Turonian, 7-non bedded limestone and zoogene breccia with Caprinidae and Chondrodontae- Cenomanian, 8-bedded and thick bedded limestone, limestone with chert and limestone breccia-Lower Cretaceous, 9-dolomite-Upper Triassic, 10-geological boundaries, 11-strike and dip of beds and isostrata, 12-folds, 13-over-thrust fault, 14-faults (wrench faults), 15-underground river direction (from Gospodaric, 1976). JAME POSTOJNSKEGA KRASA Shemalski prikaz nivojev po obs. visiiwh PLANIN5KA STOPNJA
POSTOJNSKA SIOPNJA
Figure 15: The Caves of the Postojna Karst presentated in schematic levels with absolute altitudes: 1-surface relief with humus and depressions, 2- collapsed dolines, 3-coIlapsed halls in the underground, 4-single level caves, 5-potholes or inclined caves with reached depth, 6-multiple levels caves, 7-galleries with communicating levels, 8-channels and sumps of underground Pivka, 9-caves cadastre numbers, (from Gospodaric, 1976). METHODS
This chapter describes the sampling and experimental procedures we used in this work. In the first part we will introduce the sampling sites, after which we will describe the chemical measurements we used, and at the end the isotope measurements will be presented.
Sampling
Atmosphere
Samples for the determination of 14C activity in the atmosphere were collected at two different locations: outside the cave (Direkcija), located above the entrance of Postojna Cave, and at another location inside the cave in the Colourful Gallery (Pisani Rov). Samples of atmospheric and of cave air CO2 were collected each month. A saturated carbonate-free sodium hydroxide solution was exposed in a tray to the open atmosphere (Obelic et al., 1986). Trapped atmospheric CO2 was then released by dissolving the sodium carbonate in dilute hydrochloric acid. After purification and removal of water vapour the evolved CO2 was measured using the standard method for determination of the 14C concentration (Srdo6 et. al., 1979).
Samples for the determination of 813C in the air were collected outside the cave at the sampling points Direkcija and Nem6ji Vrh, as well as inside the cave in the Colourful Gallery near the stalagmite Brilliant, and in the Biospeleological Gallery. I took the sample with a 10 cm3 syringe and put into an evacuated vial. Afterward the sample was prepared for isotopic measurements.
Soil
Soil samples were collected at two locations with different thicknesses of rock ceiling above the cave. The first location, called Direkcija, is above the entrance of Postojna Cave where the thickness of the cave roof is about 10 m. The second location, called Nemcji Vrh, is above the stalagmite Brilliant, where the roof thickness is about 120 m (see the map of sampling points in Figure 16). We took a soil profile at the sampling site Direkcija from 0 to 35 cm, and at the sampling site Nem6ji Vrh from 0 to 70 cm.
13 Samples for the determination of 8 C in soil CO2 were obtained at the same locations as the soil samples. I put a copper capillary with a diameter of 1 mm in to the soil. At the sampling site Direkcija carbon dioxide in the soil was sampled at a depth of 35 cm, while at the sampling site Nemcji vrh the samples were collected at a depth of 70 cm.
Water
Monthly water samples from June 1996 to June 1997 were collected from three locations within the cave. Figure 17 shows the sampling of water in the cave. The sampling locations were chosen to cover all types of cave water, as well as positions with different thicknesses of rock ceiling. The cave waters are solely derived from rainfall and several basic water-types are defined (Fornaca-Rinaldi et al., 1968). A stalactite-drip is water that drips slowly from the tips of stalactites throughout the year and is mainly independent of precipitation on the surface. A fast- drip is water that mainly issues out of and along fissures in the cave roof during the rainy winter seasons. A third water type termed film water is that which streams along the ceiling of the cave before accumulating at a point and slowly dripping to the floor of the cave throughout the year. The dripping waters may accumulate in pools. Some pools receive their water from slowly dripping water sources (stalactite-drip, film-water), while others receive most of their water from fast drips.
37 Tourist Centre PIVKAJAMA :
MAGDALENA JAMA
UMETNI ROV VlW'-f'ftfdi1 ti \ PERKOVHOV Call, n MARTELOVA DVORANA \Um,ls.tlilll
CARDBNI VRT W,JJ,'« liwiten PODZEMEUSKA VELIKAGORA PIVKA <»ViW BRILLIANT URFUL GALLERY KONCERTNA PISANI ROV OVORAhtA ftr Hall 2GORNJI TARTAR
OTOSKA JAMA SPODNJ! TARTAR
POOZEMEUSKA PIVKA
KONGflESNA DVORANA 'v Hull ptma panls
Dry Tourist centre E POSTOJNSKA
My1**" Cctvti Ruifway W Sialion JAMA
Figure 16: Sampling points within the Postojna Cave.
38 Figure 17: Sampling in the cave.
Three samples (pool, fast-drip and stalactite-drip waters) were collected in the Colourful Gallery where the thickness of the cave roof is about 35-40 m. Two other fast-drip water samples were collected at the entrance of Postojna Cave, where the thickness of the cave roof is about 10 m, and in the cave near the Brilliant stalagmite, where the roof thickness is about 120 m. Additional samples were also collected from the river Pivka at the point outside Postojna Cave, and at the border of the karst area from the spring of Mocilnik. Sampling points within the cave are shown in Figure 16. Rainwater samples were also collected monthly above the Entrance sampling point at the location called Direkcija.
Stalagmite
The 5 cm high stalagmite (Figure 18) was collected in the Biospeleological Gallery, about 200 m from the entrance. There the cave roof consists of 30 meters of upper Cretaceous limestone and the soil thickness is about 4-10 cm. Visible growth laminae are difficult to count. A polished vertical thin section reveals a thin (<0.1 mm) remarkable black layer at 24 mm from the stalagmite top, whose date of deposition was well known. During the Second World War, Postojna Cave was used as a storage site for gasoline, and on the 23rd of April 1944 it exploded as the result of sabotage action by Slovene Partisans. A large part of the cave is still covered by a black layer of carbon soot, but in some places stalagmites have continued their growth and appear now as white calcite caps on a black surface.
39 Table 1: Sampling points
Sample Location Date Measurements N Atmosphere Direkcija June 96-June 97 14C, 813C
Colourful Gallery C O
Rainfall Direkcija June 96-June 97 S18O, 5D 13 Soil Direkcija 2. October 1996 14C, 513C 6 Nemcji Vrh 10 13 Soil CO2 Direkcija June 96-June 97 8 C 13 Nemcji Vrh 10 Colourful Gallery, chemistry Water Entrance, Brilliant June 96-June 97 813C ,S18O, 8D 91 Pivka, Mocilnik Stalagmite Biospeleological Gallery 27. October 96 14C, S13C, 818O 1
Black layer 1944
Figure 18: Photograph of a stalagmite sample taken in the Biospeleological Gallery of Postojna Cave.
40 Measurements
Chemical measurements
Chemical measurements were performed only on the water samples. The temperature, pH and conductivity of the water samples were measured in situ. All water samples were filtered in the laboratory through a 0.45 |am filter before all other measurements were done. By filtering, most of the bacteria and almost all suspended clays were removed. Filtering is required to assure that the laboratory analysis represents only the dissolved species, which are involved in geochemical reactions and are used in chemical equilibrium reactions. Measurements of alkalinity were performed in the laboratory 24 hours after sampling.
Temperature measurements
Water temperature was determined by a temperature probe that is a component of a pH meter (Norell Unifet Microprocessor) and has an accuracy of ± 0.5 degrees Celsius. pH measurements
The pH measurements were performed using a portable Norell Unifet Microprocessor pH meter with ISFET microelectrode and glass-electrode (HEC 0102 - Mertel) having a measurement error of ± 0.05. The instrument was calibrated with pH 7.0 and 10.0 buffer solutions.
Conductivity measurements
Electrical conductivity was measured using a portable conductivity meter (HATCH) with an accuracy of ± 5 (xScm"1.
Alkalinity analysis
Alkalinity was determined by the Gran titration method (Gran, 1952; Edmond 1970; Stumm and Morgan, 1981). Past the equivalent point at pH 4.3, when all HCO3' has been converted to + + H2CO3, the increase of H concentration is directly proportional to the addition of H . The HCI volume added is plotted against the Gran function:
pH F = (V+V0)-10' [48]
In this equation V is the volume of acid added and Vo is the starting volume of the sample. The equivalence point is then obtained by back extrapolation from the linear part of the curve as shown in the Figure 19. Water samples of 25 cm3 were titrated with 0.02 mol dm"3 HCI until a pH of ~4.5 was reached. Hydrochloric acid was prepared from a commercial 0.1 mol dm'3 HCI solution (Merck). The starting volume of the sample is defined as the volume of the titrated solution. Six points of Gran function were determines by addition of 0, 1, 1, 5, 1, and 1 cm3 of acid (Digitrate T5165, 0-25 cm3) to the solution. Analytical error was estimated to be ±2%.
Ca2+ and Mgz+ measurements
The determination of Ca2+ and Mg2+ ions was done in the Laboratory for Analytical Chemistry, Department of Environmental Sciences at the Jozef Stefan Institute. Measurements were performed on a flame atomic absorption spectrometer VARIAN AA5. Results are reported with an analytical error of ±4% for Ca2+ ions and ±1% for Mg2+ ions, respectively.
41 -0.17155494U B = 0.01980204 • 0.175- R = 0.99990957 /
0.150-
X o. 0.125- O 0.100-
5V+A ) 0.075 - /
0.050 -
0.025 - equivalent point /
nnnn- 10 11 12 13 14 15 16 17 18 19 V(ml)
Figure 19: Illustration of the Gran plot method for alkalinity determination by titration with acid.
Stable isotope analysis
All stable isotope analyses were performed in the Laboratory of Geochemistry, Department of Environmental Sciences at the Jozef Stefan Institute, Ljubljana, Slovenia, with the exception of the sedimentary carbonates, for which analyses were performed at the Institute of Nuclear Research, ATOMKI, Debrecen, Hungary, and at the Laboratory for Hydrology and Isotope Geochemistry at University of Paris- Sud, Orsay, France.
Carbon
The soil samples for the determination of 813C in organic carbon were first dried, ground and passed through a 2 mm sieve, and then carefully examined to remove any residues of plant material. Then samples were acidified with 3 mol dm'3 HCI to remove any carbonates and filtered a cross a GF/C Whatman filter. The re-dried sample was then combusted in pure oxygen flow, and the isotopic composition of the resulting CO2 was determined on a Europa 20-20 ANCA-SL mass spectrometer.
8 C of DIC was measured following the modified procedure developed by Mook et al. (1980). Samples for S13C-DIC were introduced into an evacuated septum vial containing 100% phosphoric acid. The isotopic ratio of the collected CO2 was measured on a Europa 20-20 ANCA-TG mass spectrometer.
All results of S13C values of organic carbon in soil and 813C values of DIC are reported as deviation in per mill (%o) from the PBD standard. The precision of the analysis is ±0.2 %o.
Sedimentary carbonates were reacted with 100% phosphoric acid at 50°C for 3 hours following the method of McCrea (1950). CO2 released during acid treatment was analysed using a Varian MAT 250 mass spectrometer. Results are reported as deviations in per mill (%0) from the PDB standard with an analytical error of ±0.05 %o, calculated as the average standard deviation of at least three measurements.
42 Oxygen
The method for the determination of 818O in water is based on the exchange of oxygen isotopes between water and CO2 following the method of Epstein and Mayeda (1953). Water samples of 3 5 cm were introduced into a special glass vial and bubbled with CO2 for about 2 minutes. Samples were then put into a water bath with a constant temperature of 20°C for 12 hours, and then the equilibrated CO2 was analysed using a Varian MAT 250 mass spectrometer. Results are reported as deviations in per mill (%o) from the SMOW standard with an analytical error of ±0.05%o, calculated as the average standard deviation of at least three measurements.
Deuterium
Chromium reduces water to hydrogen gas at a temperature >700°C, simultaneously binding other elements of the sample to thermally stable compounds, i.e., nitride (N), carbide (C), oxide (O), and halogenide:
2Cr + 3H2O -» Cr2O3 + 3H2
The principle of the technique involves a very fast evaporation and reduction of the sample upon contact with the hot chromium catalyst (Gehre et al., 1996). The specially designed reaction tube is directly coupled to the dual-inlet system of the mass spectrometer. A liquid sample of 4 |uL is injected into a small inlet volume above the hot reaction zone. The sample evaporates very fast and it is reduced quickly upon chromium contact; it is then analysed using a Varian MAT 250 mass spectrometer. Results are reported as deviations in per mill (%o) from the SMOW standard with an analytical error of ±0.1 %o, calculated as the average standard deviation of at least three measurements.
Radiocarbon measurements
Samples of atmospheric CO2 were collected continuously each month. Saturated carbonate- free sodium hydroxide solutions were exposed to the atmosphere in an open tray (Obelic et al., 1986). The absorbed CO2 was collected in the form of sodium carbonate. CO2 is then released by dissolving the sodium carbonate containing trapped atmospheric CO2 in dilute hydrochloric acid. After purification and removal of water vapour the evolved CO2 was measured using the standard method for determination of 14C concentration (Srdoc ef al., 1979) in the 14C and 3H laboratory at Rudjer Boskovic Institute.
Soil samples were measured using the standard method for the determination of 14C concentration (Srdoc etal., 1979) in the 14C and 3H laboratory at Rudjer Bo§kovic Institute.
As compared to conventional techniques such as liquid scintilation counting (LSC) and gas- proportional counting, much less carbon about only 1 mg is required for accelerator mass spectrometry (AMS). This important advantage in smaller sample size allows for measurements 14 of C activity in speleothems. Between 10-36 mg of CaCO3 were sampled with a microdriller from a polished stalagmite cross-section. This represents between 1 and 2 years of calcite deposition. Calcite powders were reacted with H3PO4 in order to obtain CO2. The gas was reduced to graphite on iron with hydrogen at 650°C for 100 minutes; the residual gas was then used for stable isotope measurements on a SIRA spectrometer at Orsay (Laboratory for hydrology and isotope geochemistry at University of Paris- Sud, France). Carbon atoms were counted with an accelerated mass spectrometer (TANDETRON) at Gif-sur-Yvette. Analytical errors including laboratory errors are ±0.1%o for 813C and between 0.6 and 0.9 pMC for 14C.
43 RESULTS AND DISCUSSION
In this chapter, the results of the chemical measurements which influence the formation of speleothems will be presented first, followed by the results of the isotopic measurements. At the end, the model that we have developed will be presented.
Chemical measurements
Environmental factors influencing speleothem formation such as air and water temperatures as 2+ well as water properties at the surface and in the cave (electrical conductivity, Ca , HCO3" and Mg2+ ion concentrations and drip rates), were measured during an entire year in order to characterize seasonal variations. The influence of the cave roof thickness on various water properties was examined and the differences between various cave waters (pool, stalactite drip and fast drip waters) were studied. Chemical measurements were made for drip water samples inside the cave and for the surface waters, Pivka river and Mocilnik spring.
Temperature
The variation of water temperatures during the 1-year period is shown in Figure 20. The temperature of the cave waters is more or less stable through out the year, while the surface water temperatures show seasonal dependence.
In the Colourful Gallery, the temperature of the waters ranges from 7.8 to 8.4°C, with mean annual temperatures for pool, fast drips, and stalagmite drip waters of 7.8 ±0.1, 8.2 ± 0.3 and 8.2 ± 0.4 °C, respectively. The temperature of pool water is, on average, lower by 0.4°C than the drip water temperatures. We assume the pool temperature is equal to the cave air temperature within the limit of error, since there is no ventilation or contact with the outside air in the Colourful Gallery. It was assumed that the cave air temperature would be equal to the mean annual temperature on the surface, but our measurements show a difference between temperature in the Colourful Gallery (7.8 °C, 45 m below the surface) and on the surface (9.1± 0.1 °C average for the 1-year period of measurements). This can be explained by the unusually high temperatures during that year when the measurements were performed, causing an elevated average surface temperature. Fast drips and stalagmite drip waters show stable temperatures during the 1-year period, with the mean temperatures (8.2 °C) lower than the mean annual surface temperature (9.1 °C) and higher than the temperature of cave air in the Colourful Gallery (7.8 °C). The temperature of the drip waters is associated with the temperature of the rock above the cave.
The sampling point, Entrance, is near the Postojna cave entrance in the hall through which the river Pivka flows. The cave roof is only about 10 m thick at this sampling point. It was expected that the temperature of the fast drip water at Entrance depends upon seasonal changes. This is confirmed by the measurements; the drip temperature follows the seasonal trends of the river Pivka and the surface temperature, with a minimum in the winter months (January to April). The mean annual temperature of the Entrance fast drip water is 9.3 ±1.4 °C, which is equivalent to the mean annual surface temperature (9.1 °C) within the margin of measurement error.
The sampling point, Brilliant, is in the deepest part of Postojna cave, 120 m below the surface. The mean annual temperature is 10.4 ± 0.7 °C, significantly higher than the surface temperature average. A logical explanation for this high temperature of the drip water is the depth of this sampling point. The temperature of the overlying rock formation rises with depth; on average the geothermal gradient (temperature increase with depth) is 33m for 1 °C (however, this depends on several geological factors). Comparing the temperature of the drip waters in the Colourful Gallery (35-40 m deep, 8.2 °C) and at the Brilliant site (120m, 10.4 °C) gives a geothermal gradient of 36 m per 1°C in our case.
44 -•—COLOURFULGALLERY pool -•- COLOURFUL GALLERY fast drip water -+- COLOURFUL GALLERY stalactite drip water 20- -*- ENTRANCE fast drip water -T- BRILLIANT fast drip water - X- RIVER PIVKA - X- SPRING MOCILNIK 15- —D— surface temperature o annual average surface temperature 2 10-
Q. 5- Q)
0-
fe & S
«s id to 4 11 I 5 date
Figure 20: Seasonal variations of temperature in cave water samples, in the river Pivka and in the spring Mocilnik. pH
Accurate pH measurements of water samples are important for the calculation of the saturation index (Q) and the dissolved inorganic carbon (DIC) content of the samples.
The pH values are shown in Figure 21. The pH values for cave waters vary between 7.3 and 8.3 and in this range the HCO3" ion is the predominant chemical species.
The mean annual pH values for the individual sampling points are: 7.9 ± 0.2 for pool water in the Colourful Gallery, 7.8 ± 0.2 for fast drip water in the Colourful Gallery, 7.8 ± 0.1 for stalactite drip water in the Colourful Gallery, 7.9 ± 0.2 for the sampling point Entrance and for the sampling point Brilliant the mean pH value is 7.8 ± 0.2. We observe a general trend of increasing pH throughout the year for all sampling points, which we cannot explain by environmental parameters and is not attributed to seasonal changes. These variations could be the consequence of the errors in sampling and measurement because the drip rate was sometimes very low, and it is probable that degassing of the water occurred during sample collection. There is no correlation between the pH values of the water samples and the thickness of the cave roof nor the type of cave waters. No correlation between pH values and drip rate can be observed either.
45 8.4-j -•-COLOURFUL GALLERY poo! -•- COLOURFU L GALLERY fast drip water 8.3- -A- COLOURFUL GALLERY stalactite drip water -•- ENTRANCEfast drip water 8.2- -T- BRILLIANT fast drip water - +• • RIVER PIVKA
8.1- X SPRING MOCILNIK + 8.0- 7.9- 7.8- CL 7.7- 7.6- 7.5- 7.4- 7.3-
date
Figure 21: Variations of pH values in cave water samples and in the river Pivka and the spring Mocilnik during one year.
Calcium and bicarbonate ion concentrations and electrical conductivity
The calcium ion (Ca2+) and bicarbonate ion (HCO3") concentrations were determined in the cave waters, the river Pivka and the spring Mo5ilnik, together with electrical conductivity (s). In Table 2 the mean, maximum and minimum values of Ca2+, HCO3" ion concentrations and electrical conductivity are presented. The variations of Ca2+, HCO3" ion concentrations and electrical conductivity values during 1-year are shown in Figures 22, 23 and 24, respectively. Seasonal variations are observed for samples of fast drip waters at Entrance and Brilliant as well as for pool water in the Colourful Gallery. Less pronounced is the variation of these values for the fast drip water in the Colourful Gallery, while stalagmite drip water in the Colourful Gallery is more or less stable throughout the year. Comparing this behaviour with the seasonal variations of drip rates we can see that the maximum values of Ca2+ and HCO3" ion concentrations and electrical conductivity occur when precipitation is high. This is in agreement with results published in the literature (Genty and Deflandre, 1998), which were obtained from a cave in Belgium. Genty determined the dependence between drip rate and the water excess value. Water excess, calculated as the difference between rainfall and evapotranspiration (Thornthwaite, 1954), depends on the annual rainfall and temperature (Figure 25). It reaches a maximum in late autumn, which coincides with the drip rate maximum, however the drip rate maximum is observed with a 1-2 month delay. In autumn, the water excess reaches a maximum and the water stored in rock pores and fissures is flushed out. The hydraulic pressure due to the water excess increases with depth, and is the highest in the rock of the cave ceiling just above the stalagmite. This water was presumably stored in the rock for a few weeks to a few months, dissolving the limestone, and thus contains high concentrations of ions. The other cave water component is the water that was stored in the soil. During spring and summer, biological activity is the most intense and the CO2 content in soil water is very high. This water with high ion content is flushed out from the soil with the autumn water excess. The increased conductivity value is also partly due to organic matter (e.g. fulvic and humic acids) dissolved in the soil water (Genty and Deflandre, 1998). Drip water may be composed partly of water of the same year's rainfall, and partly from an older water reservoir. Results from a cave in Belgium (Genty and Deflandre, 1998) show that dripping water in 1 hydrological year was composed from fast seepage water (aged few hours to few days) and reservoir water (mixed recent and over 1-year- old water).
46 2+ Table 2: The results of chemical analysis (mean, maximum and minimum values) for concentration of Ca and HCO3' ions, and the electrical conductivity of water samples.
Ca2+ (mgdm"3) ' HCO " (mgdm"3) ' electrical conductivity (uScrrf1) Sampling point 3 mean max min mean max min mean max min
COLOURFUL GALLERY pool 65.3 76.8 54.3 194.0 221.8 163.2 334 379 299
COLOURFUL GALLERY fast drip water 69.3 75.3 65.0 209.7 226.8 197.7 352 384 305
COLOURFUL GALLERY stalactite drip water 71.5 74.4 64.4 214.4 230.5 204.6 365 383 343
ENTRANCE fast drip water 79.9 94.0 61.7 232.8 285.3 184.6 397 488 220
BRILLIANT fast drip water 55.5 69.4 40.3 165.2 210.5 129.8 301 418 238
RIVER PIVKA 65.3 78.1 52.0 211.3 251.4 145.6 424 522 368
SPRING MOCILNIK 56.8 64.5 50.0 212.3 236.8 179.8 366 414 311 —•- COLOURFUL GALLERY pool 100-i -•-COLOURFUL GALLERY fast drip water 95- -A -COLOURFUL GALLERY stalactite drip water -•- ENTRANCE fast drip water 90- -V- BRILLIANT fast drip water + RIVER PIVKA 85- X SPRING MOCILNIK 80-
? 75- •a 70- 65- 60- O 55- 50- 45- 40-
I ! 2 3 rt i- T- iri in date
Figure 22: Seasonal variations of Ca2+ ion concentration in cave water samples, in the river Pivka and in the spring Mocilnik.
-•- COLOURFUL GALLERY pool -•- COLOURFUL GALLERY fast drip water 290- -A- COLOURFUL GALLERY stalactite drip water 280- -•- ENTRANCE fast drip water 270- -T— BRILLIANT fast drip water •+•- RIVER PIVKA 260- X SPRING MOCILNIK 250- 240- 230- 220- •u 210- 200- 190- 180- o 170- 160- 150- 140- 130- 120
3 ? 3 i n s; m T^ o> date
Figure 23: Seasonal variations of HCO3" ion concentration in cave water samples, in the river Pivka and in the spring MoCilnik.
48 —•-COLOURFUL GALLERY pool -•- COLOURFUL GALLERY fast drip water -A- COLOURFUL GALLERY stalactite drip water 550- -•- ENTRANCE fast drip water - V- BRILLIANT fast drip water • - + RIVER PIVKA 500- X SPRING MOCILNIK
450- o 400-
350- rity/
300-
conduct s 250-
200-
I I I I | 4 | | | | § date
Figure 24: Seasonal variations of electrical conductivity in cave water samples, in the river Pivka and in the spring Mocilnik.
-•- COLOURFUL GALLERY fast-drip water -A- COLOURFUL GALLERY stalactite-drip water 1.74- -•- ENTRANCE fast-drip water -•- BRILLIANT fast-drip water -•- rainfall 1.72- water excess 1.70 i
0.5-
5 Q.
T3
•^ Io -> •r- CO t- date
Figure 25: Seasonal variations of drip rate for stalactite and fast drip waters in comparison with precipitation and water excess values.
49 The dependence of the concentrations of Ca2+ and HCO3" and of electrical conductivity on the thickness of the rock cover above the cave is presented in Figures 26, 27 and 28, respectively. At the sampling point Entrance, where the thickness of the rock cover is about 10m, the highest values were observed. In Colourful Gallery these values were lower and the lowest values were measured at the sampling point Brilliant, where the thickness of the cave roof is about 120m. Observed variations in water chemistry are correlated with thickness of the overlying rock for stalactite and fast drip water, but not for pool water in Colourful Gallery. It was expected that ion concentrations and electrical conductivity would increase with rock cover thickness but we observed the opposite relationship (Figures 26-28). These variations are caused by processes of ion exchange during passage of the solution through cavities. There precipitation and solution of carbonate or other minerals takes place and enables changes in the Ca ion and dissolved CO2 concentrations. Linear regression curves in the Figures 26, 27 and 28 are drawn through data points for fast and stalactite drip waters only. Pool water data are shown only for comparison.
• COLOURFUL GALLERY fast drip water 80- A COLOURFUL GALLERY stalactite drip water • ENTRANCE fast drip water T BRILLIANT fast drip water 75- linear regression curve for drip waters R2=0.96
CO E 65-1 (•) pool water
55-
50 0 10 20 30 40 50 60 70 80 90 100 110 120 130 thickness of cave roof /m
Figure 26: Influence of cave roof thickness on Ca ion concentrations for fast and stalactite drip waters.
50 250-1 • COLOURFUL GALLERY fast drip water A COLOURFUL GALLERY stalactite drip water 240- ENTRANCE fast drip water BRILLIANT fast drip water linear regression curve for drip waters ^\ 220- R2=0.99
210- T3 -*> 200- •f (•) o" 190- pool water o 180- X 170-
160-
150- i i i i i 1 1 1 1 1 1 1 1 0 10 20 30 40 50 60 70 80 90 100 110 120 130 thickness of cave roof /m
Figure 27: Influence of cave roof thickness on HCO3" ion concentrations for fast and stalactite drip waters.
• COLOURFUL GALLERY fast drip water 400- A COLOURFUL GALLERY stalactite drip water • ENTRANCE fast drip water 390- V BRILLIANT fast drip water linear regression curve for drip waters 380- R2=0.94 T 370- I 360- -5- 350- "1 340- (•) '"§ 330- pool water O 320~ 310- 300- 290 0 10 20 30 40 50 60 70 80 90 100 110 120 130 thickness of cave roof /m
Figure 28: Influence of cave roof thickness on electrical conductivity for fast and stalactite drip waters.
51 The linear regressions through the data were calculated by the statistical method called least 2+ squares fit. The HCO3"vs. Ca correlation plots are shown in Figure 29 for individual sampling points and also separately for all different cave waters. The correlation coefficient, R2, between 2+ HCO3' and Ca ion concentrations is a measure of stability of conditions at individual sampling points. From the individual diagrams it can be seen that very good correlations (R2 > 0.80) exist 2+ between HCO3" and Ca concentrations for the Entrance and Brilliant sampling points, while ion concentrations are less correlated for pool and fast drip waters in the Colourful Gallery. No correlation at all (R2 < 0.01) was found for the stalactite drip water in Colourful Gallery. From the correlation coefficients we can conclude that conditions are stable throughout the entire year for the cave waters of Entrance and Brilliant, and less for the Colourful Gallery pool and fast drip waters. Nothing can be deduced from the randomly scattered points of the stalactite drip water in Colourful Gallery. Ion concentrations in the river Pivka are very well correlated (R2 = 0.83) showing stable conditions, but much less so for the spring Mocilnik (R2 = 0.48). Good correlation was found for those cave waters, which are sensitive to seasonal changes (see Figure 25). The stalactite drip water in the Colourful Gallery is probably fed by a water reservoir in which concentrations are not sensitive to seasonal variations and stay in the same narrow range during the year. This is confirmed also by the drip rate, which is more or less constant throughout the year (Figure 25).
2+ The ratio of HCO3" to Ca concentrations is given by the slope of the linear regression curve through the data points. The dissociation of 1 mole of calcium bicarbonate yields 1 mole of Ca2+ 2+ and 2 moles of HCO3~ ions. For a solution containing only Ca and HCO3" species, the slope of 2+ the curve should give the ratio [HCO3"]/[Ca ] = 2 (molar ratio) or mHco3- / rnCa2+ = 3.05 (mass 2+ ratio). From the diagrams in the Figure 29 it can be seen that HCO37Ca mass ratio is below 3.05, (between 2.14 and 2.61) for the sampling points Entrance, Brilliant and Colourful Gallery pool and fast drip waters. The Colourful Gallery stalactite drip water data show no correlation, so information about the slope (-0.07) is neglected. Ratios below the stoichiometric value of 3.05 are due to degassing of the solution, when concentrations of HCO3' decrease very quickly 2 with loss of CO2 to the cave air: 2HCO3"H> CO3 " + H2O + CO2{g). The river Pivka shows a high 2+ ratio of HCO37Ca = 4.24, which can be explained by other cation species dissolved in the river (e.g. Mg2+ and others), that are competing with the Ca2+ ions.
The electrical conductivity vs. HCO3" correlation diagrams are shown in Figure 30 for individual sampling points and separately for all cave waters. Very good correlation can be seen in plots for the Entrance (R2 = 0.79) and Brilliant (R2 = 0.86) fast drip waters. The Colourful Gallery samples are less correlated, with correlation coefficients 0.64 for pool water and 0.61 for fast drip water. No correlation can be observed for the stalactite drip water of Colourful Gallery, which shows randomly scattered points (R2 = 0.13).
Similar behaviour can be seen in the diagrams of electrical conductivity correlation with Ca2+ ion concentration (Figure 31). Good correlation is observed for pool water in the Colourful Gallery (R2 = 0.77), and less for Entrance (R2 = 0.69), Brilliant (R2 = 0.64) and Colourful Gallery fast drip waters (R = 0.58). Data for the stalactite drip water are scattered, and the correlation coefficient is 0.46.
2+ 2 Correlation of electrical conductivity with HCO3" and Ca is very poor for the river Pivka (R = 2 2+ 0.30 for HCO3" and R = 0.11 for Ca ). For the spring Mocilnik, the correlation coefficient is 0.79 2 for electrical conductivity vs. HCO3" plot and very low (R =0.28) for the electrical conductivity vs. 2+ 2+ Ca plot. Correlation coefficients of electrical conductivity vs. HCO3" or Ca ion concentrations represent the degree of dependence of conductivity on these ionic species. Where good correlation is observed (fast drip waters, pool in Colourful Gallery), conductivity is almost entirely 2+ explained by the presence of Ca and HCO3" ions. In the stalactite drip water in Colourful 2+ Gallery, the most prevalent ions in solution are obviously those other than Ca and HCO3". A similar explanation can be given for the river Pivka. The spring Mocilnik shows a good 2+ correlation of electrical conductivity with HCO3", but no correlation for Ca ion.
Figures 30 and 31 are diagrams of chemical parameters that are correlated in separate plots for all cave water samples. Correlation is quite good: 0.64 for electrical conductivity vs. HCO3" and 0.77 for conductivity vs. Ca2+ plot. We can conclude that in cave waters the ions Ca2+ and HCO3" are the principal conducting ion species.
52 230 -, • COLOURFULQiUXERYpool 220- T BRZUANT &tt drip mttr 210- ! 220- R =0.88 200- slope=2.61 210- 190- . 180- j 200- ? pl70- ! 160- ! 190 I I ISO - ! 180 1 140 - 170 130- 120- 160- 65 70 110 Ca!*Angdm'
* COLOURFULOAIlZRY<dripwiUr / 260- * RIVER PKKA R'=0.58 2S0- R'=0.83 240- 220- slop«=2.14 230-
220- 210-
200- «
160- 190- 150- 140- 180- 130-
CaMAngdin"! Ca^/mgdm
•< SPRING MOftlMK 240- *• COLQUSFUL GA11£RY italtctite iaf water ] R <0.01 A 230-
210-
200- O 190-
180-
70 Ca2* /mgdnT
Ca!* Angdm'1
• PB/NIROVpool • PCAN1ROV bit drip witer 270- 290- A PBANIROV »tah«tih: drip«*tCT 260- 4 ENTRANCE bf t drip nter 2S0- R'=0.S3 * / 250- / stepo-2.60 •/ T BRIIIMNT6«tdrip-water 270- 240- 230- R2=0.72 260- i 220- * *«P J 210- S|OP 226 ?200- "' • J5
210- 140 200- 130 ^/4 • 120 190- £30
180- 85 90 95 100 85 90 95 100 Ca:
2+ Figure 29: The HCO3' vs. Ca correlation plots for all individual sampling points, and separately for cave waters.
53 • COLOURFUL 0AU2RY pool / r ERJLUANTiitdripvwftr 380- ! R =0.64 y^t R:=0.86
380- 360-
•* 340- 340- 1^320-
320- 5. SCO- • yS 'S 280- 300- yS B
280-
260-
1T0 180 180 20) 210 220 230 120 130 140 1S0 160 170 180 190 200 210 220 HCOj/mgdm' J HCOr/ragdm
540 -n • COLOURFULOALLHRY&itdripimttr + K1VERPIVKA 520- R!=0.61 500- R3=0.30 460-j 460- 440-
380- 360^ 340" 320 300-
210 220 140 150 160 170 180 190 200 210 220 230 240 250 HCO /mgdm'! J r HCOTtegdm'
* COLOURFUL GALLERYibltetitedrfc - SPRJNO MOfllMK 420- 400- R3=0.13 R!=0.79
400-
A 380- ! 380- A A
•t A 360- /^
340-
210 220 200 210 220
3 3 HCOj/mgdm" HCO3./mgdm'
500- * ENTRANCE Ait drip water 500- • COLOURFULGALLERY-pooI y' • COLOURFDLQALLERY&ftfcpw«ter ^y* 480- R1=0.79 480- 480- * COLOURFUL QAUZRYtttltctka dip witer y^ 460- 440- • ENTRANCE frftdrip witer + y^ 440- 420- T BRUIIANT&itdripwiter * y^ ! 420- 400- 360- L 400- 360- j* 380- 340- ! 380- 320- • */*• • | 340- 300- • ">^ •• 280- yS ! 320- 260- ^ ^ yS • 300- 240- 280- 220- yS 260 200- 1S0 200 210 220 230 240 2S0 280 270 280 290 120 130 140 150 160 170 180 190 200 210 220 230 240 250 260 270 280 290
3 3 HCOj-Angdm' HCOr /mgdm'
Figure 30: The conductivity vs. HCO3" correlation plots for all individual sampling points, and separately for cave waters.
54 • COLOURFUL GALLERY peel R^O.77 T BRILLWJT fast drip water R^O.64
55 60 65 70 75 40 45 50 55 60 65 Ca^/mgdm3 Ca^/mgdm3
• COLOURFUL GALLERY last drip tlaler RJ=0.88 * RIVER PIVKA R?=0.11
-S- 450-
70 72 74 78 50 55 W 65 70 75 80 Ca2*7mg dm3
•KJU" * COLOURFUL GALLERY stalactite drip waler 430- SPRING MOClLNIK 380- 1^=0.46 \s^ 420- R!=O.2S 370- 400-
360- j^ A 390-
350- j^ 4
340- 350- 330- 340 330- 320- 320- 310- 310- 84 68 70 72 74 76 300
CaJ*/mgdm'3
* ENTRANCE fast drip water 500- • COLOURFUL GALLERY pool R^O.69 500- • COLOURFUL GALLERY fast drip water 490- A COLOURFULGALLERYstalectitedripwater 450- 460^ * ENTRANCE tastdrfpweler 440- T BRILLIANT fast drip water i 400- 420- l = 400- E,
? 350- ;340- % 320- i 300- <300-
1 250- ' 260- 240- 200- 220- 200- 40 45 50 55 60 65 70 75 80 85 90 95 100
Figure 31: The conductivity vs. Ca2+ correlation plots for all individual water points, and separately for cave waters.
55 85- • COLOURFUL GALLERY fast drip water • ENTRANCE fast drip water • BRILLIANT fast drip water 80- linear regression curve for fast drip waters R2=0.99 75-
•E70^ COLOURFUL GALLERY stalactite drip water E 65-
O 60-
55-
50 0.00 0.05 0.10 0.15 0.20 0.25 0.30 drip rate /cmV1
Figure 32: Correlation between mean annual values of drip rate and Ca2+ ion concentration in fast drip waters.
250-i • COLOURFUL GALLERY fast drip water • ENTRANCE fast drip water 240- • BRILLIANT fast drip water linear regression curve for fast drip waters 230- R2=0.98 220-
210- COLOURFUL GALLERY 200- stalactite drip water E O 180-
170-
160-
150- 0.00 0.05 0.10 0.15 0.20 0.25 0.30 drip rate /cmV1
Figure 33: Correlation between mean annual values of drip rate and HCO3" ion concentration in fast drip waters.
56 420- • COLOURFUL GALLERY fast drip water 4 ENTRANCE fast drip water
Ann •y BRILLIANT fast drip water / 4UU " linear regression curve for fast drip waters fir R2=0.99 / 380- o (A) / 360- COLOURFUL GALLERY / _>» stalactite drip water /^ 340- / DtlV I
T5 / 320-
co n /
300- /
280- 1 1 1 1 1 0.00 0.05 0.10 0.15 0.20 0.25 0.30 drip rate /cmV1
Figure 34: Correlation between mean annual values of drip rate and electrical conductivity in fast drip waters.
2+ The dependence of Ca , of HCO3" concentration, and of electrical conductivity on the drip rate of cave waters is presented in the Figures 32, 33 and 34, respectively. Observed variations in water chemistry are correlated with the drip rate only for fast drip waters, but not for stalactite drip water. Linear regression curves in the Figures 32, 33 and 34 are drawn through average annual data points for fast drip waters, however three points are not enough to characterize general behaviour for all fast drip waters in Postojna Cave. Additional analysis of other sampling points under changing conditions are necessary to obtain more information about the dependence of these chemical quantities on drip rate. Stalactite drip water data are shown only 2+ for comparison. The calculated correlation coefficient for Ca and drip rate is 0.99, for HCO3" and drip rate is 0.98, and for electrical conductivity and drip rate is 0.99. Drip rates for fast drip 2+ waters are very well correlated with Ca , HCO3" concentrations, and electrical conductivity. These quantities increase with drip rate for fast drip waters. This can be explained by flushing of stored water with increased hydraulic pressure, as was also observed in Pere Noel Cave (Genty and Deflandre, 1998).
Mg/Ca ratio
In principle, the chemistry of waters from the karst system and the overlying soils is strongly dependent upon water residence time as well as on the original composition of the solid carbonates. Mg/Ca ratios in percolating waters are expected to be higher for longer water retention times during the summer, and lower when the water residence time is longer during the winter (Roberts era/., 1998).
In Figure 35 the Mg/Ca ratios are presented for the entire year. Seasonal variations are very pronounced for the Brilliant sampling point, with maximum values in periods from August to October 96 and from April to June 97. Similar behaviour, but with smaller amplitude, is observed for the Entrance drips and Colourful Gallery pool water samples. A comparison between seasonal variations of drip rate and water excess values is shown in Figure 36. We can see that the minimum of Mg/Ca ratio occurs during the rainfall water excess. Elevated values of Mg/Ca
57 ratio in summer and spring are due to longer retention times in these dry periods. When water retention is high, the water dissolves not only Ca2+ but also some Mg2+ from the rock formations. The water sample at Brilliant shows the highest values of Mg/Ca ratio. This is explained by the great thickness of the rock ceiling above the sampling point (120m). In the thick cave roof the carbonate solution is slowly dissolving both the limestone and dolomite, and in solution Mg2+ substitutes for Ca2+. Water sampled at Entrance on January 7, 1997 was collected under conditions of very high drip rate, however this did not influence the Mg/Ca ratio much. For samples from the Colourful Gallery, we can see in Figure 35 that pool water Mg/Ca ratio values are higher than the fast and stalactite drip waters. Pool water shows also a slight dependence on seasonal variations of water excess. The dependence of Mg/Ca ratio upon drip rate is presented in Figure 37. From the plot we can see an inverse relationship between drip rate and Mg/Ca ratio for fast drip waters, with a correlation coefficient of 0.77. Stalactite drip water is shown only for comparison. Figure 38 shows the influence of cave roof thickness on the Mg/Ca ratio in drip waters. A good correlation between Mg/Ca and rock cover thickness is observed for all types of cave waters, with the correlation coefficient 0.85. Figures 37 and 38 show that Mg2 concentration is higher for thicker rock cover above the cave where water retention time is high, and where the water dissolves not only Ca2+ ions but also some Mg2+ ions.
0.055-] -•—COLOURFUL GALLERY pool -•- COLOURFUL GALLERY fast drip water 0.050- -A— COLOURFUL GALLERY stalactite drip water • • •• • ENTRANCE fast drip water •V- BRILLIANT fast drip water 0.045-
0.040-
0.035-
0.030-
0.025 - :;—1=:—j-j; 0.020-
in date
Figure 35: Mg/Ca ratio values in cave waters during 1 year.
58 —•— COLOURFUL GALLERY Mg/Ca - ENTRANCE Mg/Ca -T— BRILLIANT Mg/Ca ••••••• COLOURFUL GALLERY drjo rate 0.05- 200 ••••••• ENTRANCE drip fate T BRILLIANT drip rate . '. •••*•-- water excess /mm 0.04- 150.
3
50
0.01
date
Figure 36: Comparison of the Mg/Ca ratio with drip rate and water excess values for the fast drip waters at sampling points Colourful Gallery, Entrance and Brilliant.
0.04-1 • COLOURFUL GALLERY fast drip water • ENTRANCE fast drip water T BRILLIANT fast drip water linear regression curve for fast drip waters 2 0.03- R =0.77
(0 O 0.02 COLOURFUL GALLERY stalactite drip water
0.01-
0.00' 0.00 0.05 0.10 0.15 0.20 0.25 0.30 drip rate /cmV1
Figure 37: Correlation between mean annual values of drip rate and Mg/Ca ratio in drip waters.
59 0.05-1 COLOURFUL GALLERY pool COLOURFUL GALLERY fast drip water COLOURFUL GALLERY stalactite drip water ENTRANCE fast drip water 0.04- BRILLIANT fast drip water - linear regression curve through cave waters R2=0.85 0.03-
O 5 0.02 H
0.01-
0.00 0 10 20 30 40 50 60 70 80 90 100 110 120 130 thickness of cave roof /m
Figure 38: Influence of cave roof thickness on Mg/Ca ratio in drip waters.
Saturation index
The saturation index was calculated using Equation 17. The values are plotted in Figure 39. From the results it can be seen that the saturation index is higher than 1 for most of the samples, which means that cave waters are supersaturated with respect to CaCC>3. The highest values were again observed at the sampling point Entrance, where the cave roof thickness is the least thick (10 m), and Ca2+concentration is highest. At the sampling point Brilliant, were the cave roof thickness is about 120 m, lower values were observed (0.5 to 2.7). The water was undersaturated except during autumn water excess.
From the results plotted in the Figure 40 it can be observed that there is a very good correlation (R2=0.95) between saturation index and drip rate for fast drip cave waters collected at the sampling points in Colourful Gallery, Entrance and Brilliant. A linear regression curve is drawn through the data points for the fast drip waters. For sampling point with stalactite drip water in Colourful Gallery the drip rate is not correlated with the saturation index; this sample is also not dependent upon drip rate variations during the year.
60 -•- COLOURFUL GALLERY pool 6- -•-COLOURFUL GALLERY fast drip water • -&••- COLOURFUL G ALERY stalactite drip water \ -•-ENTRANCE fast drip water / 5- -•- BRILLIANT fast drip water saturation index =1 X
c •2 3 I 2 CO
1 -
\ i i r i i i I I I | 4 date
Figure 39: Variations of saturation index during 1 year.
• COLOURFUL GALLERY fast drip water 3.5- * ENTRANCE fast drip water • BRILLIANT fast drip water • / linear rcyression curve Lnrougn laoi a rip waters / 3.0- R2=0.95 /
2.5-
nde x / c (A) / Q CO 2.0- COLOURFUL GALLERY / • stalactite drip water / "Jo CO 1.5- y
1.0- /
i i i i \ 0.00 0.05 0.10 0.15 0.20 0.25 0.30 drip rate /cmV1
Figure 40: Correlation between mean annual values of drip rate and saturation index in fast drip waters.
61 Growth rate
Two extremes are possible as the source of the supply of carbonate solution to a growing stalagmite (Dreybrodt, 1980). In the first case the supply is continuous, either by a continuous flow of solution or by drops with very short time intervals between each other. The other extreme is a supply in single drops with prolonged time delay. Our sampling points belong to the last case. This means that after the drop has fallen on the top of the speieothem it is distributed over its surface and a stagnant water film builds up in a few seconds. This film begins to degass its dissolved CO2 and carbonate deposition comences. The next drop replaces the old film and the process is repeated. Several variables influence the growth rate: Ca2+ concentration, temperature, cave air CO2 content and hydrological conditions (water film thickness, turbulent or laminar flow conditions and drip rate). The modeled growth rates were calculated using Equation 37 of the Dreybrodt model (1988) described in the section Speieothem Growth Rate Theory (page 25, Chapter The formation of speleothems).
The mean Ca2+ concentrations for each sampling site are used in the calculation of model growth rates. The water temperature is taken as the mean annual temperature for each sampling site. The partial pressure of carbon dioxide in the cave atmosphere is assumed to be constant at 3x10"4 atm, and for stalagmite growth a constant film thickness of 0.1 mm is assumed. This approximation does not have a significant effect on the growth rate. The value of a was interpolated from data presented in Dreybrodt (1988). The growth rate was converted 2 1 1 7 from mmol cm" s" to mm yr" by multiplying by 1.174x10 , with molecular mass of CaCO3 of 100.09 g mol"1 and calcite density of 2.689 g cm'-z"
The modeled growth rates and the mean annual calcium concentration are presented in the Table 3. It is seen that growth rates are of similar order of magnitude (0.13 - 0.20 mm yr" for all sampling sites.
Table 3: Modelled growth rates
Sample Ca'"' mmol dm"3 growth rate mm yr'
COLOURFUL GALLERY fast drip water 1.7 0.14
COLOURFUL GALLERY stalactite drip water 1.8 0.16
ENTRANCE fast drip water 2.0 0.20
BRILLIANT fast drip water 1.4 0.13
62 • COLOURFUL GALLERY fast drip water 0.20- A COLOURFUL GALLERY stalactite drip water • ENTRANCE fast drip water • BRILLIANT fast drip water linear regression curve for drip waters R2=0.84 CO 5. E i 0.15- I 2
0.10- 1.0 1.5 2.0 Ca2+ /mmoldm"3
Figure 41: Dependence of calculated growth rate of stalagmite (Dreybrodt, 1980) on Ca2+ concentration.
0.25-i • COLOURFUL GALLERY fast drip water • ENTRANCE fast drip water V BRILLIANT fast drip water linear regression curve for fast drip waters R=0.82 \_ 0.20 • as
E
(A)
J COLOURFUL GALLERY I 0.15- stalactite drip water CO
0.10' 0.00 0.05 0.10 0.15 0.20 0.25 0.30 drip rate /cmV1
Figure 42: Dependence of calculated growth rate of stalagmite (Dreybrodt, 1980) on drip rate of cave waters.
63 In Figure 41, the dependence of growth rate on Ca2+ ion concentration is plotted. The Ca2+ ion concentration is very important and controls the growth rate. Very good correlation was found between growth rate and Ca2+ ion concentration for all cave waters (R2=0.82). Growth rate theory suggests that drip rate is also an important variable. Dependence of growth rate on drip rate is shown in Figure 42. Good correlation was observed between growth rate and drip rate only for fast dripping water samples, but not for stalactite drip water.
From Figure 43 we can see the linear dependence between growth rate of speleothems and the saturation index for all types of cave waters (R2 = 0.94). This behaviour is explained by the fact that deposition of calcite occurs when the solution is supersaturated and increases with the degree of supersaturation, as described by the saturation index.
0.21 -| • COLOURFUL GALLERY fast drip water 0.20- A COLOURFUL GALLERY stalactite drip water ENTRANCE fast drip water 0.19- • BRILLIANT fast drip water linear regression for drip waters 0.18- R2=0.94 ^ea r 0.17- E 0.16- rat e 0.15-
2 0.14-
0.13-
0.12- 1.0 1.5 2.0 2.5 3.0 3.5 saturation index
Figure 43: Dependence of calculated growth rate of stalagmite (Dreybrodt, 1980) and saturation index in cave waters.
All of the results demonstrate that the Ca2+ion concentration controls the saturation index and growth rate of speleothems. This is normal since both equations for the calculation of saturation index (Equation 17) and growth rate (Equation 36) depend upon the Ca2+concentration. It is also observed that the Car ion concentration can be controlled by the drip rate. A comparison of measured growth rate and Ca2+ concentration at different places would give more information, but growth rate was not measured in this research.
64 Isotope measurements
The isotopic composition of carbon compounds was determined in the atmosphere (outdoor and in the cave), in the soil atmosphere and organic matter, in the cave waters, and in the growth laminae of a speleothem. Values of 813C were determined by isotopic mass spectrometry and the 14C activity was measured both by standard methods and by accelerator mass spectrometry (AMS). Isotopic values of 8D and 818O were determined only in cave waters.
Atmosphere
The isotopic composition of carbon in CO2 was measured in samples of the outdoor atmosphere and of cave air. Early measurements of the carbon isotopic composition of atmospheric CO2 were carried out by Craig (1953) and ranged from -9.9%o to -7.4 %o PDB. Craig assumed that a reasonable value for atmospheric CO2 might be -7%0 PDB. Based on the rate of industrial CO2 addition to the atmosphere since the middle of the last century a decline of -2.6 %o PDB is 13 13 expected in 8 C of present atmospheric CO2. The mean measured value for 8 C of atmospheric CO2 at the sampling points Direkcija and Nemcj'i Vrh is -8.0 %o PDB. The mean 813C values measured inside the cave are -9.3 %o PDB. Lower 813C values inside the cave 13 reflect the larger influence of biogenic CO2 from the soil above the cave, where the 8 C value of soil CO2 is expected to be about -22 ± 2 %o PDB (DOrr and Munnich, 1986; Becker-Heidmann, 1996; Becker-Heidmann et al., 1996). Our measured values of soil CO2 above the cave were - 21.7%0 and -2O.5%o PDB (see section Soil CO2)
14 Fluctuations in the C activity of atmospheric CO2 outside and inside the cave during 1 year are shown in Figure 37. The highest 14C activities outside the cave, at sampling point Direkcija, were observed during the summer months (June 96, May-June 97). Data for some months are missing because the content of CO2 precipitated in NaOH solution was not sufficient enough for measurement of 14C activity using the standard method. The variation between summer and winter months is small for the outside 14C activity. The difference between the maximum in May- June 97 and the minimum in October 1996 is 2.7 pMC. Such a small difference (within the limit of error) indicates that the contribution of fossil fuel carbon (Suess effect) is small enough to be neglected for this 1-year sampling period. This is because in the Postojna region there is no major heavy industry that would cause significant changes in atmospheric CO2 content. The mean 14C activity for the 1-year period outside of the cave was 109.2 ±1.3 pMC. Inside the cave, at the sampling point in the Colourful Gallery, the mean 14C activity was 102.2 ±1.5 pMC. The difference between the maximum 14C values in October 96 and the minimum values in May 96 is 5.3 pMC. The fluctuations of 14C activity inside the cave do not show any seasonal trend, however measurements at only one sampling point are not sufficient to derive conclusions about seasonal dependence. The 14C activity inside the cave is 7 pMC lower than outdoor 14C 14 activity. These lower values of C activity of cave air can be explained by an influx of soil CO2 to the cave through micro-fissures and pores of overlying rock, and by a mixing with the outside air, thus causing a dilution of the 14C activity in the cave air.
Soil CO2
Factors that affect soil CO2 concentration are the soil type, texture and horizon, type of vegetation, and soil flora and fauna (Ford and Williams, 1989). Soil CO2 is produced mainly by decomposition of organic matter and by root respiration. The isotopic composition of carbon dioxide in the soil atmosphere is lower by 2-4%o than the plant material from which it is produced due to isotopic fractionation effects. Large variations in soil organic matter 813C values result from differences between the C3 and C4 photosynthetic pathways utilized by plants. C3 plants 13 have 8 C values ranging from -32 to -22%o, and averaging about -27%o, whereas C4 plants have values ranging from -16 to -9%o, and averaging about -12%o (Dorr and Miinnich, 1986). C3 plants are mostly trees, cool-season grasses, whereas C4 plants are typically warm-season grasses found in tropical and temperate grasslands. It has been observed that the carbon isotopic composition of the plant material is highly correlated with the type of photosynthetic pathway followed by the plant.
65 112- —•— cave air 111 - • outdoor air 110- 109- o 108- :> 107- Q. 106- ^. 105- 104- tiv i A 103 - a d / \ O 102- / 101- / 100- • 99- 98-
c t> CO o date
Figure 37: Radiocarbon activity in the CO2 in the cave air and outdoor atmosphere.
-15 -DIREKCIJA - NEMCJI VRH -16
-17
-18
CO -19- Q Q. -20- P -21- -22 -23 -24- -25-
9- < date
Figure 38: Monthly variations of 8 C values of CO2 in soil atmosphere.
66 Measured monthly 813C values of carbon dioxide in the soil atmosphere are shown in Figure 38. By measurement of the 813C of the carbon dioxide in the soil atmosphere above Postojna Cave we determined the mean 813C values for sampling points Direkcija and Nemcji Vrh to be -21.7%o PDB and -20.5%o PDB, respectively. At the sampling point Direkcija, carbon dioxide in the soil was sampled at a depth of 35 cm, while at the sampling point Nemcji Vrh the samples were collected at a depth of 70 cm. The difference of 1.2%0 PDB can be the consequence of different types of soil present at these sampling points. At the sampling point Direkcija, there is a mixture of dark brown and light brown soil, while at the sampling point Nemcji Vrh only the dark brown soil is present.
Soil CO2 production is strongly dependent upon the season of the year (Dorr and Munnich, 1986; Tieszen and Boutton, 1989). During the summer, when the soil temperature is high, the 813C values are higher due to higher respiration rates. During the winter months the influence of root respiration is decreased, while the contribution of microbial decomposition of soil organic 13 matter relatively increases and shifts the 5 C of soil CO2 to lower values. The isotopic composition of soil CO2 shows a seasonal variation of 3.4%o PDB for the sampling point Direkcija, and 7.3%o PDB for the sampling point Nemcji Vrh (Figure 38). The lowest 813C values were observed at the sampling point Direkcija in the period from October 96 to January 97, and at the sampling point Nemcji Vrh in October 96. The observed seasonal variability can be 13 explained either by mixing with atmospheric CO2 of higher 5 C values, or by decomposition of organic matter with different isotopic compositions.
Some data are missing for sampling point Nemcji Vrh. This is due to difficult access through thick snow cover during the winter months, when measurements were not performed. The 1 C activity of the soil CO2 was not measured in this period.
Soil organic matter
Soil organic matter (SOM) is not homogeneous but consists of at least two components (Dorr and Munnich, 1986): a fast degradable component with a short lifetime (T) of »1 year and slow degradable component with T«1 00 years.
Figure 39 shows the 813C values of soil organic matter as a function of soil depth for the two 13 different sampling sites. Mean 8 C values for both soil profiles are -26.7%0 PDB, and indicate that the organic carbon was produced by plants with the C3 (Calvin-Benson) photosynthetic pathway having a characteristic 513C value of about -27 %o PDB (Dorr and Munnich, 1986). We can divide our soil profile into two parts the topsoil and the subsoil horizon. The topsoil horizon (depth of 0-15 cm) for both sampling sites (Direkcija and Nemfiji Vrh) is dark brown soil with very frequent roots, while the subsoil horizon (depth >15 cm) is dark brown at the sampling site Nemcji Vrh and light brown, clayey soil at the sampling site Direkcija. From Figure 39 it can also be seen that the topsoil horizon shows lower 813C values, with the mean 813C value of -27.1%o PDB compared to the subsoil horizons, where the mean 813C values is -26.5%o PDB. The 513C values of topsoil horizons reflect the input of organic material from modern vegetation via different pathways: a) by return of plant residues to the soil where they are subjected to animal and microbial decomposition, b) by the distribution of the rooting system which determines the depth to which living plant tissues can directly affect the soil carbon cycle. The enrichment of about 0.6 %o PDB between the topsoil and subsoil horizon is probably due to the progressive decomposition of soil organic matter in the subsoil horizon.
67 -•- DIREKCIJA -•— NEMCJI VRH 5-
10- TOPSOIL 15 SUBSOIL o 20-
•B. 0) 25- 30-
35-
40- -27.5 -27.4 -27.3 -27.2 -27.1 -27.0 -26.9 -26.8-26.7 -26.6 -26.5 -26.4 -26.3 -26.2 -26.1 -26.0 /%oPDB
Figure 39: 5,1 3C values determined in soil profiles on October 2, 1996.
The radiocarbon content of the atmosphere was increased due to the nuclear bomb testing after 1954, which also caused changes in the 14C content of soil organic matter. The enrichment of radiocarbon in soil organic matter primarily due to the addition of plant residues growing in the 14C enriched atmosphere. Such residues include dead leaves and twigs, the soluble fractions carried to the soil via leaf drip and stem flow, root exudates, the carbon dioxide respired by the roots and dead root material. Enrichment may also occur by binding of atmospheric carbon dioxide in the soil biomass. The turnover rate also depends upon the type of plant residues, the type of biological population, upon climate and upon soil type. The most important factor affecting soil C enrichment is the additional of modern vegetation material to the soil.
Measured radiocarbon activities in the soil above the cave are presented in Table 3. The radiocarbon activity just below the surface was 117.0 pMC, while at a soil depth of 22-30 cm it was 97.6 pMC. Pine forest and grass grow at both investigated sampling points. Results show that the C enrichment in the topsoil horizon was similar at both sites.being around 118 pMC for these sampling points, (the variations are within the experimental error). The data show that bomb produced 14C is included in the soil organic matter (SOM) of the first 10 cm. We can divide our soil profile into three parts. In the first part (depth < 5 cm) the 14C activity is a mixture of the recent atmospheric 14C activity and shows the average 14C activity of the atmosphere and modern vegetation for last few years. The second part of the soil, at depths between 5 to 10 cm, is a mixture between pre- and post-bomb 14C activity. At a depth below 20 cm, the activity of 14C is attributed to pre-bomb 14C activity only.
In the soil profile at the sampling point Nemcji Vrh it was impossible to determine the 14C activity below 10 cm depth with the standard method, since the 14C values were lower than the detection limit of our method. The soil was not sampled at the same depth intervals for both sampling points (Direkcija and Nemcji Vrh). This does not allow us to compare both sampling points for the depth dependence of C activity and makes it impossible to determine the depth at which C from bomb tests has penetrated.
68 Table 3: Radiocarbon activity in the soil at different depths
SAMPLE MC activity pMC
Direkcija 0-2 cm 117.1 ±1.7 Direkcija 2-7 cm 110.5 + 1.6 Direkcija 22-30 cm 99.2+1.5 Nemcji Vrh 0-2 cm 118.1 ±1.7 Nemcji Vrh 3-5 cm 117.7 ± 1.7 Nemcji Vrh 5-10 cm 110.2 ± 1.7
We also determined the 14C activity in the leaves from plants at Direkcija and Nemcji Vrh. The 14C activity was 115.2 ± 1.7 pMC for the sampling point Direkcija and 112.9 ± 1.7 pMC for the sampling point Nemcji Vrh. From these results it can be seen that upper part of the analysed soil organic matter reflects the input of organic materials from modern vegetation. Since we measured the 14C activity in the atmosphere to be only 109 pMC, there is a large discrepancy between these values (over 115 pMC in the upper soil profile and leaves, and 109 pMC in air). Data in the literature (Levin and Kromer, 1997) show an atmospheric 14C activity in this region to be 114- 115 pMC in the years 1996 -1997. This would suggest that there was an error in our determination of the 14C activity in the atmosphere. To prove this some additional measurements must be done in the future.
Water
The isotopic composition of carbon, oxygen and hydrogen were determined in cave waters. The mean, maximum and minimum values of 813C, 818O and 8D are presented in Table 4, together with values for the surface waters. The 513C values vary with the type of the cave water. Fast drip waters show higher 813C values than stalactite drip waters. For the fast drip sampling point 13 in the Colourful Gallery the mean S C value is -14.4 %0 PDB. For the sampling point at the Entrance it increases to -12.7 %o PDB, while for the sampling point Brilliant this value is even higher (-10.2 %o PDB). Fast drip waters show variations in 813C values at the sampling point Brilliant from -12.6 to -7.3 %o PDB (5.3%o PDB difference), at the sampling point Entrance from -14.2 to -9.9 %0 PDB (4.3 %0 PDB difference) and for the fast drip sampling point in Colourful Gallery from -15.6 to -13.4 %o PDB (2.2 %o PDB difference). From the Figure 40 we can observe two maximums for the sampling points Brilliant and Entrance. The first one falls between September and November 96 and the second one is in April 97. The maximum peak in April 1997 is the consequence of snow melting since snow contains atmospheric CO2 with a higher amount of 13C, while the first maximum is probably due to the autumn water excess. As 13 we determined in the previous chapter the values of 8 C of the soil CO2 reached a maximum during the summer months, in the dry period. This enriched CO2 is then flushed out in the autumn months when precipitation is high. The 813C values of pool water range from -14.5 to - 10.3%o PDB (difference 4.2%0 PDB), with a mean value of-13.1%o PDB, and the values for stalactite drip waters range from -15.6 to -13.4%o PDB (difference 2.2%o PDB), with a mean 813C value of-14.7%o PDB. In the Colourful Gallery, the seasonal variations for drip waters are not pronounced, while for the pool water the maximum occurs in November 96. Lower values of 13 8 C determined in cave waters indicate that these waters contain dissolved CO2 from soil. To 13 test this, we calculated the theoretical 8 C of DIC (mainly HCO3") using the Mook equation 13 (Equation 49, page 32) (Mook, 1974). Since we measured the 8 C values of soil CO2 only above the sampling points Entrance and Brilliant during a 1-year period we can only compare these values with the calculated S13C values. The data are compared to the calculated values in Figure 41. The comparison of calculated and measured values of dissolved inorganic carbon shows a general linear trend, but the points are very scattered. We can conclude that our measurements are described well enough by the Mook equation for these sampling points. The 813C values are not dependent on the thickness of the roof above the cave.
69 Table 4: The mean, maximum and minimum values of 513C, 818O and 8D
. ' ' • 513C %oPDB 518O- %o SMOW Sampling point 8D - %i'SM0W '
mean min max mean min max mean min max
COLOURFUL GALLERY pool -13.1 -14.5 -10.3 -8.9 -9.3 -8.6 -58 -62 -47
COLOURFUL GALLERY fast drip water -14.4 -15.6 -13.4 -8.8 -9.0 -8.6 -57 -61 -48
COLOURFUL GALLERY stalactite drip water -14.7 -15.6 -13.4 -8.4 -8.6 -8.3 -55 -59 -52
ENTRANCE fast drip water -12.7 -14.2 -9.9 -8.8 -9.4 -8.4 -57 -63 -47
BRILLIANT fast drip water -10.2 -12.6 -7.3 -8.3 -8.7 -7.7 -54 -56 -52
RIVER PIVKA -13.7 -15.7 -10.1 -8.2 -9.2 -7.1 -54 -63 -43
SPRING MOCILNIK -13.8 -14.9 -12.7 -9.1 -9.4 -8.7 -60 -65 -48 -•—COLOURFUL GALLERY pool -7-1 -•- COLOURFUL GALLERY fast drip water -A- COLOURFUL GALLERY stalactite drip water -8- -•- ENTRANCE fast drip water -T— BRILLIANT fast drip water -9- +• • RIVER PIVKA X SPRING MOCILNIK -10-
-11 - Q Q. -12- P -13- •""to -14-
-15-
-16-
-17 i r S5 —• ^ £$• X ^ S* M a f- T^ oi N *- \ci 10 I 3 s date
Figure 40: Seasonal variations of 8,13 ,C in cave water samples, in the river Pivka and in the spring Mocilnik.
-6- • ENTRANCE fast drip water -7- T BRILLIANT fast drip water
-8-
-9- m Q o. -10- -11 -
-12-
P -13- -14-
-15-
-16 i i -16 -15 -14 -13 -12 -11 -10 -9 -7 -6 513C cal /%oPDB DIC
Figure 41: Comparison between the measured (813Cmes) and calculated (813Ccai) values of dissolved inorganic carbon for the sampling points Entrance and Brilliant.
71 The 818O values for cave waters are given in Table 4 and their seasonal dependence is shown in Figure 42. For all cave waters the 818O values vary from -9.4 to -7.7 %o SMOW. The mean 518O values range from -8.9 to -8.3%o SMOW. At the sampling point Entrance, where the roof thickness is about 10 m, we can see two maximums, one between October and January 97 and second in April 97. We can conclude that first maximum is caused by the autumn water excess and in April the higher value is due to the snow melting. In the Colourful Gallery we can see slight seasonal variations for the pool water sample. The lower values of 518O are observed during the summer, and higher values during the winter and spring months. That trend is even more pronounced at the sampling point Brilliant. We could not explain this increasing trend since our measurements did not continue after June 1997. 818O values are not dependent upon the type of cave waters, nor upon the thickness of the roof above the cave. The results of 818O measurements in precipitation are plotted in Figure 43.
More or less stable values throughout the year, with some variations, are observed for 8D values, which are plotted in Figure 44. Values for cave waters are in the range from -63 to -47%o SMOW with mean 8D values from -58 to -54%o SMOW. Data are missing for December 96 because those results were anomalous due to a systematic error in the measurements. Some seasonal dependence can be seen for the samples at Entrance and the Colourful Gallery pool, with maximums from October 96 to January 97. The values for the stalactite drip water in the Colourful Gallery decrease in the winter and spring months, but since the measurements were stopped in June 97 we could not find an explanation for this trend. The SD values do not depend on the cave roof thickness.
-7.5-, -•—COLOURFULGALLERY pool -•- COLOURFUL GALLERY fast drip water -A- COLOURFUL GALLERY stalactite drip water -•- ENTRANCE fast drip water -V- BRILLIANT fast drip water -8.0-
O 3 -8.5- P '"to -9.0-
-9.5' 8 8 iri £ 2 date
Figure 42: Seasonal variations of 8 0 in cave water samples during 1 year.
72 -A- •X RIVER PIVKA 18 3K- SPRING M0&LNIK -5 o— • • 3K precipitation -O~ outside temperature 16 -6 14 12 I" -7 o 10I -8 X 8- CO x X . X 8 w Sf . -9 x • 6 o '•* ••*• 9-
p -10 *••• 4 CO O 2 -11 0 -12 -2 I
date
Figure 43: Variations of 818O in the river Pivka and the spring Mocilnik compared to seasonal variations of surface temperature and precipitation during 1 year.
-•—COLOURFUL GALLERY pool -•- COLOURFUL GALLERY fast drip water -A- COLOURFUL GALLERY stalactite drip water -•— ENTRANCE fast drip water —T— BRILLIANT fast drip water -52- >\ -53- -54- -55- -56- o -57- -58- -59- Q -60- to -61 - -62- -63- 8 rJ t» a o > •j a. c T <; » ^ Z 3 •^ o> r4 »- in ? 3 date
Figure 44: Variations of 8D in the cave water samples during 1 year.
73 74 1318 18 river PivkaandthespringMoCilnik. Figure 45:Correlationbetwee5CB)and8OMCWvaluesithcavwatersamples, (PD(S comparison. river Pivkaandthesprin g Mocilnik.TheWMWL(worldmeteoricwaterline ) isshownfor Figure 46:Correlationbetwee n SDMOW)and8Ovaluesforthecavwate r samples,the (S Q CO o to o PDB -20- -u -60- -35- -30- -25- -65- -55- -50- -45- -40- -70- -85- -75- -80- -12- -13- -11 - -10- -16- -15- -14- -8- ~l -9- 7 - -12 10-9-8-7-6 A • • • X + • -9.4 20-8.86-7. A X' X + • • > COLOURFUL GALLERYpool COLOURFUL GALLERYstalactitedripwater COLOURFUL GALLERYfastdripwater PRECIPITATION y SPRING MOCILNIK RIVER PIVKA BRILLIANT fastdripwater ENTRANCE fastdripwater s m y/ COLOURFUL GALLERYstalactitedripwater ENTRANCE fastdripwater COLOURFUL GALLERYfastdripwater COLOURFUL GALLERYpool SPRING MOCILNIK BRILLIANT fastdripwater RIVER PIVKA T * • 1 I • 3*^A*++ A x •m&^K • T • jr 8 O/%oSMOW 18 5O /%oSMOW + yS + + / T • / / m 18 a 8D=7.88O+9 R=0.98 w + 13 18 In Figure 45 the correlation between 5 C(PDB> and S O(SMOW) is presented. From these results we can see that the measured values for the different cave water samples are grouped in well- defined areas of the plot. These areas can be distinguished by the coordinates of their centre. It is possible to distinguish between different types of cave waters as well as between the river Pivka and spring MoSilnik. The Colourful Gallery data are centred around a point with coordinates 518O -8.8 %o SMOW and 513C -14 %o PDB, while the sampling point Entrance has coordinates S18O -8.8 %o SMOW and S13C -13 %o PDB and the sampling point Briiliant has coordinates 518O -8.2 %o SMOW and 813C -10 %o PDB. The River Pivka and spring Mocilnik are centred around points with coordinates 818O -8.1 %o SMOW and S13C -14 %o PDB and 818O -9.2 %o SMOW and S13C -13 %o PDB, respectively. No correlation was found by plotting these coordinates against drip rate or cave roof thickness.
In Figure 46 the correlation between 8D and 818O in the cave waters, river Pivka, and spring Mocilnik are shown together with values for precipitation. The straight line through the points represents the world meteoric water line (WMWL) for Postojna region. Correlation is defined by linear regression through the data points:
8D = 7.8 818O + 9
It can be seen that all of the cave waters, river Pivka, and spring Mocilnik fit quite well with the natural meteoric water data. It is again possible to distinguish between different types of cave waters as well as between the river Pivka and spring Mocilnik, which occupy distinct areas of this plot.
^H amount of precipitation /dm3m'2 -B-818O/%oSMOW -•-5D/%oSMOW temperature/°C
Figure 47: Monthly values of amount of precipitation, surface temperature, and stable isotope contents 818O and 8D for the sampling area of Postojna cave.
75 In Figure 47 are presented the monthly values of the amount of precipitation, surface temperature, and stable isotopic values of precipitation (518O and 8D) in the area of Postojna cave during 1 year. Precipitation values show the seasonal maximum between September and November 1996. The mean monthly values of temperature and stable isotopic values show distinct seasonal behavior. No correlation is observed between the amount of precipitation and stable isotope data. Such a situation is typical for continental stations in the northern hemisphere (Rozanski et a/., 1993, Krajcar-Bronid et al., 1998). Both 818O and 8D data follow the same trend as the temperature. The maximum air temperature during the sampling period coincides with highest 818O and 8D values in August 1996, while the lowest values were in December 1996. There is a discrepancy between the isotope data and temperature in March 97, with a sharp maximum occuring in the isotopic values. This may be caused by an error in sampling or perhaps an anomalous rain event.
Speleothems
13 Speleothem 8 C values reflect the vegetation conditions above the cave and atmospheric CO2 composition, while 518O values are linked to the meteoric water composition (Goede, 1994). Furthermore, the isotopic composition of deposited calcium carbonate depends on the temperature of deposition in the cave, which is generally equal to the annual average temperature of the region.
In Figure 48, the isotopic data for a stalagmite from the Biospeleological Gallery in Postojna Cave are presented. We found S13C values ranging from -9.3 to -10.2 %o PDB. Slightly higher S13C values were observed in the growth laminae of the stalagmite for the period after the gasoline explosion in the cave in 1944, but the variations are less than 0.5%o PDB. We can explain these values with equilibrium isotopic fractionation. If we take into consideration that the 13 average cave temperature is 8.0°C, and the average 8 C value of the soil CO2 is -22%0 PDB, 13 then the calculated 8 C value of the CaCO3 using the Mook equation (Mook, 1974) is -12.1 %o PDB. This differs from the measured values by 1.9 to 2.8%o PDB. This diffence could be explained by kinetic fractionation processes occuring during fast degassing, or by calcite precipitation from parent seepage water prior to its arrival in the cave above the stalagmite. We also observed from the measured data that the fractionation process was quite constant in the time scale studied (50 years). In this time of speleothem growth the climate and average composition of precipitation did not change significantly. Even though atmospheric 813C is theoretically decreasing due to the burning of fossil fuels (Suess effect (Suess, 1955)) as was observed by Baskaran (1993), no decrease can be seen in our 813C time series.
For the investigated stalagmite we also wanted to determine the temperature of the water from which the speleothem was deposited. We found 518O values in the speleothem ranging from -6.05 to -4.49 %o PDB. Using Equation 38, we calculated the temperature in the cave in past 50 years. For the temperature calculation, we needed to know the isotopic composition of S18O in the dripping water, which feeds the stalagmite. The dripping water isotopic composition varies during the year, as we can see from the previous chapter. There are some variations during the year due to seasonal changes in the amount of precipitation and vegetation above the cave. Because speleothems grow throughout the year, for the calculation we used the average S18O values for the sampling point Entrance, which is nearest to the sampling point where the speleothem was taken. The difference between the calculated temperature (12.7 °C) and the average cave temperature (8 °C) is 4.7 °C. Such a large difference can be due to fact that the speleothem did not grow in isotopic equilibrium with the seepage water. The difference in temperature could be also influenced by the fact that the speleothem was taken 200 m away from the Entrance. There is some ventilation, so when water is passing over the surface of the growing speleothem it could be enriched in 18O due to evaporation. This would yield more 16O enriched calcite to be deposited on the speleothem, and would appear as a wamer temperature of deposition.
76 • s13r • O \j
-.-618O -6.2-1 -9.2 -•—813C
-6.4- -9.4
m Q -6.6- -9.6 °-L DL O -6.8- -9.8 "0 o CO -7.0- -10.0
-7.2-
-10.2 1940 1950 1960 1970 1980 1990 2000 Interpolated laminae date
Figure 48: Oxygen and carbon isotopic composition in growth laminae of the stalagmite (A) and in interpolated laminae data (B).
77 120-, 180 —•— speleothem / ; 170 160 110- 150 o 140
100- - 130
••§ 120 CO O 110 90- • } 100 ] 10years,' 90 aimospnere 80- i i i i 80 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 Interpellated laminae date
Figure 49 : Radiocarbon activity in the stalagmite and the atmosphere
Radiocarbon activities in the stalagmite are shown in Figure 49. A sharp peak was expected in the years 1963 - 1964, due to the nuclear tests in the atmosphere (Levin et a/., 1992). Based on the two known time markers, the gasoline explosion mark (April 23, 1944) and the top layer (October 24, 1996), the time scale in the speleothem was determined. The maximum C activity in the speleothem carbonate occurred in 1974, which was 10 years after the atmospheric peak. The difference of 14C activity between pre- and post-bomb periods is 27.4 pMC between the years 1950 and 1978. Compared with the atmospheric variations of 14C activity from 1950 to 1964 (100 pMC and 190 pMC, respectively), the stalagmite activity variations are small, which means that the 14C signal has been damped. Atmospheric 14C activity shows a well pronounced peak in 1964, and since then decreases exponentially, while radiocarbon activities in the stalagmite only slightly decrease after the year 1974 (from 115.4 pMC in 1974 to 107.2 pMC in 1996). The pronounced 14C peak together with the variation between pre- and post-bomb 14C activities demonstrate that Postojna Cave is sensitive to atmospheric changes. By comparison of the activity curves of the atmosphere and in the speleothem, we can gain information about the processes of carbon transfer from the surface to the cave.
Additionally, two measurements made on the 14C in DIC of modern dripping water showed values 105.3 pMC and 106.0 pMC, which are within the limits of error of the activity measured in the stalagmite tip (107.2 pMC).
78 Modelling
Comparison of 14C measurements in the growth laminae of the stalagmite with atmospheric 14C activity data (Levin and Kromer, 1997) showed clear differences in the time scale and in the amplitude of the peak due to nuclear bomb tests. To explain these differences a model considering different carbon sources was developed (Genty et a/., 1998).
Among the sources of DIC in seepage water, the most important are the CO2 in soil (which is the main source with an 80-90% share of the carbon), and the carbon derived from limestone dissolution (with a14C activity of 0 pMC).
In Figure 50 we present the different parameters employed in the model to obtain the best fit of the measured activities. Parameter Ci (%) represents the contribution of CO2 produced by the slow decay of soil organic matter to the total CO2 in the soil. The slow-decay carbon reservoir contains soil carbon older than 1 year. The remaining soil CO2 is produced by root respiration and very fast decay of organic matter; its share is C2 = 1- Ci.
Ci
By variation of parameter d we control the response of the model to atmospheric changes. The effect of d is seen in the slope of the 14C activity after the nuclear bomb tests peak, which decreases with an increasing value of d. Through modelling we obtain the value of Ci = 60%. This high value of d indicates that soil organic matter contribution to the soil CO2 is high in the winter, the season of the highest growth rate of the speleothem.
Y,
The second parameter determined by modelling is Yi, the number of years over which slow- turnover soil organic matter is averaged. The data were corrected for the radioactive decay of 14C, which could also be neglected because of the young age of the speleothem. This parameter affects the amplitude variations of the 14C activity before and after the 14C peak due to the nuclear bomb tests in atmosphere.
With lower Yi values, the peak of the modelled curve is more pronounced, while higher Y-i values flattens the curve. Through modelling we obtained the value of Yi = 50 years. We assumed that the new organic matter activity produced in 1 year is similar to the atmospheric activity in that year. dcp
The dead carbon proportion (dcp) gives us the fraction of CO2 that is derived by the dissolution of the limestone in the rock. The dcp controls the total 14C activity of the modelled curve, thus defining the range of the curve.
By variation of dcp the modelled curve is fit to the measured activities in the period before the nuclear bomb test 14C input. We obtain the value of dcp = 15%.
Y2
The mixing effect is described by Y2, the number of years over which the DIC activity is 14 averaged. Parameter Y2 is employed for smoothinsmooth g of the yearly signal a CD|C. Through the modelling we obtained the value of Y2 = 11 years.
Instead of using this parameter we tried to change parameter Yi and the ratio Ci/C2l but the model did not fit the measured points.
79 Atmosphere CO2 ATMOSPHERE
SOIL CO2 from root respiration and very CO2from SOM decompositon fast OM time residence >1 year decomposition Yi=50 years
C2=(1-C1) C2=40% Ci=60%
Soil/epikarst CO2 carbon EPI KARST
LIMESTONE Carbon from limestone CaCO3 dissolution (1-dcp) %
dcp=15%
Dissolved Inorganic Carbon Dissolved Inorganic Carbon HCO3' of seepage water HCO3" of older seepage water and/or secondary calcite deposit dissolution
Y2=11 years Disolved Inorganic Carbon HCO3' of seepage water dripping in the cave
t CAVE CaCO3 formation: stalagmite formation
Figure 50: Model for carbon infiltration in a karst system (from Genty et a/., 1998)
80 Correction by adjusting the mixing effect could be far from real, because there is no proof of any water mixing or of secondary calcite dissolution, by which such a mixing effect is explained. A non-smoothed curve cannot be explained only by the mixing effect, because eventual short peaks are masked by the sparse sampling (every 5 years). Fractionation must also have occurred at the temperature in the cave (T = 8°C), which is assumed to be equal to the mean annual temperature on the surface.
Comparison of modelled and measured activities is presented in Figure 51, which shows a good fit of the model to the measured values. There are some discrepancies which can be explained by the errors in dating, and by disturbances in the drip rate that occured after the 1944 gasoline explosion. We believe that the growth rate of the stalagmite was slower for some years after 1944 due to morphological perturbations of the flow paths and because of the breakage of the stalactite above the sampled stalagmite.
120 Model parameters:
Fast SOM reservoir: C2=40% 115 - Slow SOM reservoir: 0^60% SOM averaging: Y^SO years 110 dcp=15% Mixing Y2=11 years O 105
100
TO O 95
90
85 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 AD
Figure 51: Comparison of measured and modelled activities for the stalagmite.
81 CONCLUSIONS
Karst terrains are complex, and it is difficult to understand the processes occurring in these systems, as each have their own distinctive hydrology and typical landforms. In the framework of my doctoral thesis I have studied environmental factors and chemical properties influencing the growth of speleothems. The influence of human activity on the atmosphere and the soil, such as the depletion of the carbon stable isotopic ratio in the atmosphere due to fossil fuel combustion and the increase in 14C activity due to nuclear experiments, is observed in the carbon content in the studied speleothem. To better understand the carbon transfer in this karst system, a model for Postojna cave was developed.
The results presented in the thesis and the conclusions thus derived are summarized below.
1. Environmental factors and general parameters that were determined: Postojna Cave is situated under a hill covered by a pine tree forest and grass. Mean annual surface temperature was 9.1 °C for the period from June 96 to June 97 (data obtained from Hydrometeorological Institute, Ljubljana), which is unusually high. Average annual rainfall was 1500 mm (data obtained from Hydrometeorological Institute) in the sampling year. - Water temperature measurements showed that the temperatures of the cave waters at the sampling point Brilliant and in the Colourful Gallery (pool, fast drip and stalactite drip water) were stable through the one-year period. Fast-drip cave water at the sampling point Entrance together with the river Pivka and spring Mocilnik showed seasonal temperature dependence.
2. Chemical measurements that were performed and the conclusions obtained: The mean pH values for all cave and surface waters was 7.9 ± 0.2, so we can conclude that in these karstic waters HCO3" ion is the predominant chemical species. All other variations of pH values during the one-year period are random and are not correlated with any type of cave water, thickness of cave roof, or drip rate. 2+ - Determination of Ca and HCO3~ ion concentrations as well as electrical conductivity showed the following behaviour: - Seasonal variations during the one-year period of these chemical parameters are observed for fast drip waters at the sampling points Brilliant and Entrance, and for pool water in the Colourful Gallery. Stalactite drip water values in the Colourful Gallery were 2+ more or less stable throughout the year. Maximums of Ca and HCO3" ion concentration and electrical conductivity occur when precipitation on the surface and the water excess values were high. Water excess was defined as the difference between rainfall and evapotranspiration. During water excess maximum, water stored in pores and fissures in the rock above the cave is flushed out. Correlation of these chemical parameters with the thickness of rock cover above the cave was also examined. For dripping water samples, we observed an inverse 2+ proportional dependence of Ca , HCO3" ion concentrations and electrical conductivity with the thickness of rock cover above the cave. This relationship can be attributed to ion exchange processes occuring during passage of the water through micro-fissures and pores. 2+ Correlation plots between Ca and HCO3' ion concentrations showed that fast-drip waters at the sampling points Entrance, Brilliant and in the Colourful Gallery were well correlated, while the pool sampling point in the Colourful Gallery showed less correlation and no correlation was found for the stalactite water sample in the Colourful Gallery. By determining the slope of the linear regression curve it was observed that for all examined fast-drip waters the slope is below the stoichiometric ratio of HCO3" and Ca2+. This means that equilibrium conditions are not maintained due to degassing of the calcite solution, and the concentration of HCO3' decreases very quickly with a loss of CO2 to the cave air. Stalactite drip water in the Colourful Gallery is not sensitive to seasonal variations, and also does not show any correlation between different chemical parameters. - Similar behaviour was observed between the correlation plots of electrical conductivity 2+ vs. concentration of HCO3" and Ca ions.
82 The Mg/Ca ratio was determined. Seasonal variations during the one-year sampling period are well defined for the sampling point Brilliant and with smaller amplitude for the pool sampling point in the Colourful Gallery, and for the fast-drip water at Entrance. It was observed that the minimum Mg/Ca ratio occurs during the autumn water excess. Higher values of Mg/Ca ratio during the summer and spring are due to the longer retention times in these dry periods, so water dissolves more Mg + ions from the overlying rock formations. The saturation index was calculated: For most of the samples the saturation index is higher than 1, which means that cave waters are supersaturated with respect to CaCO3. The highest values were observed at the sampling point Entrance. The dependence of saturation index upon the drip rate was also compared. It was observed that the saturation index is very well correlated with drip rate for fast-drip waters, while for stalagmite drip water no correlation between saturation index and drip rate was observed. - The growth rate of speleothems depends strongly on saturation index, which means that deposition of calcite occurs only when the solution is supersaturated.
3. Stable carbon isotopic analysis yielded the following results: - 513C values measured in the outdoor atmosphere were higher than 513C values of air inside the cave. This reflects the influence of biogenic CO2 from the soil above the cave on the cave air. 13 The 8 C values of soil CO2 were determined at two different locations and some seasonal variations of 813C were observed. During the summer the 813C values were higher than during the winter months, because of the contribution of CO2 from root respiration. Root respiration greatly decreases in winter, and the contribution of CO2 from organic matter decomposition with lower 813C increases relatively. Variations in results may also be explained also by mixing with atmospheric CO2. 813C values of soil organic matter were determined at different soil depths. Mean 813C values for both soil profiles indicate that organic carbon was produced predominantly by plants with a C3 (Calvin- Benson) photosynthetic pathway. Soil profiles were divided into two parts: topsoil (0-15 cm) and subsoil horizon (> 15 cm). The results indicate that the topsoil horizon shows lower S13C values in comparison to the subsoil horizon. One can conclude that the topsoil horizon reflects the input of organic material from modern vegetation, and that the subsoil horizon shows enrichment of about 0.6 %o PDB due to the progressive decomposition of soil organic matter. In water samples, the S13C values varied with the type of cave water. Fast-drip waters showed higher 813C values than stalagmite drip water. Two maximums peaks were observed in the 813C values of the fast drip waters. The first one falls between September- November 96 and is explained by the autumn water excess, and the second one in April 97 is the consequence of the spring snowmelt. The 813C values in the cave waters are not dependent on the thickness of the cave roof above the cave. The studied speleothem was not in isotopic equilibrium with its feeding water. 813C values of the speleothem indicated that kinetic fractionation process occurred during fast degassing or calcite precipitation from the solution.
In water samples we determined the 818O and 8D values, and in the speleothem carbonate the 818O value was also measured. The 818O values in the cave water samples were lower during the summer and higher during the winter and spring months, but only for the sampling point Entrance with fast-drip water and at the pool sampling point in the Colourful Gallery. The 818O values are not dependent upon the type of cave water, nor upon the thickness of the rock cover above the cave. Relatively stable values were observed for SD in the cave water samples throughout the year, and there were no seasonal variations. The 8D values are not dependent upon the type of cave water, nor upon the thickness of the rock cover above the cave. - We also determined the 818O and 8D values in precipitation. From these results we observed that the cave waters, the river Pivka and the spring Mocilnik fit quite well to the natural meteoric water data.
83 Using the 818O values of the speleothem we attempted to calculate the cave temperature in the past. Due to the fact that the speleothem was not in isotopic equilibrium with its seepage water, it was impossible to determine past climate changes.
4. The 14C activity of the studied samples showed the following results: 14 - The comparison of the C activity of atmospheric CO2 outdoors and in the cave shows 14 lower C activity for cave air, that can be explained by an influx of soil CO2 to the cave through micro-fissures and pores of overlying rock and by mixing with the outside air. 14 The C activity of the soil CO2 was not measured. Determination of 14C activity in the soil organic matter shows that the soil profile can be divided into three parts. In the top part (0-5 cm) the 14C activity is the mixture of recent soil organic matter, and shows the average 14C activity of the atmosphere intake and modern vegetation for last few years. The second part of the soil, at depth 5 to 10 cm, is the mixture between pre- and post-bomb 14C activity. At the depth below 20 cm the activity of 14C is attributed to pre-bomb 14C activity only. By the AMS technique an atmospheric 14C bomb peak was detected in the carbonate deposits of the stalagmite growth laminae, with a delay of 10 ± 5 years.
5. A model was developed to explain the difference between the stalagmite 14C time series and the atmospheric 14C activity data. The possible carbon sources taken into consideration are: soil CO2, soil organic matter and seepage water. Different parameters of the model were determined: Parameter C-i represents the contribution of CO2 produced by slow decay of soil organic matter to the total CO2 in the soil. With this parameter the response of the model to atmospheric changes is controlled. A value for C-i value of 60% was determined. Parameter Yi represents the number of years over which slow-turnover soil organic matter is averaged. This parameter affects the variations of the 14C activity before and after the 14C peak due to the nuclear bomb tests in atmosphere. The calculated value for Y, is 50 years. The next parameter used in our model is the dead carbon proportion (dcp). This parameter gives the fraction of CO2 that is derived from the dissolution of the limestone rock formation. The dcp controls the total 14C activity of the modelled curve, thus defining the range of the curve. The determined value for dcp is 15%.
Parameter Y2 represents the number of years over which the DIC activity is averaged. 14 Parameter Y2 is employed for smoothing of the yearly signal of a CDic- In our model, the value for Y2 is 11 years. Comparison of modelled and measured activities showed a good fit of the model to the measured values.
84 REFERENCES
Allison V. C. (1926) The antiquity of the deposit in Jacob's cavern, American Museum of Natural History, 19, 294-225 Atkinson T.C. (1985) Present and future directions in karst hydrogeology, Annales de la Soci6t6 Geologique de Belgique, 108, 293-296 Baker A., Smart P. L, Edwards R. L. and Richards D. A. (1993) Annual growth banding in a cave stalagmite, Nature, 364, 518-520. Baker A. and Smart P.L. (1995) Recent flowstone growth rates: Field measurements in comparison to theoretical predictions, Chemical Geology, 122, 121-128 Baker A., Ito E., Smart P.L. and McEvan R. (1997) Elevated 13C in speleothem and implications for paleovegetation studies, Chemical Geology (Isotope Geoscience), Baker A., Genty D., Dreybrodt W., Barnes W.L. Mockler N.J. and Grapes J. (1998) Testing theoretically predicted stalagmite growth rate with recent annually laminated samples: implications for past stalagmite deposition, Geochimica et Cosmochimica Ada, 64, 393-404 Baskaran M. and Krishnamurthy R.V. (1993) Speleothems as a proxy for the carbon isotope composition of atmospheric CO2, Geophysical Research Letters, 20, 2905-2908 Bastin B. and Gewelt M. (1986) Analyse pollinique et datation 14C de concretions stalagmitiques holocenes: Apports complementaires des deux methodes, Geographie Physique et Quaternaires, 15(2), 185-196 Becker-Heidman P. (1996) Requirements for an International radiocarbon soil database, Radiocarbon, 38,177-180 Becker-Heidman P., Scharpenseel H.W. and Wiechmann H. (1996) Hamburg radiocarbon thin layer soils database, Radiocarbon, 38, 295-345 Blanc J. J. (1972) Recherches sur les planchers stalagmitiques et remplissages anciens des grottes du massif de Marseilleveyre (Marseille, B.-du-Rh.), Buletin du Mus6um d'Anthropologie Prehistorique de Monaco, 18, 5-43 Brocker W.S. and Olson E.A. (1960) Radiocarbon measurements and annual rings in cave formations, Nature, 185, 93-94 Brodar S. (1952) Prispevek k stratigrafiji kraskih jam Pivske kotline, posebej Parske globine, Geografski vestnik, 24, 43-76 Brodar S. (1956) Ein Beitrag zum Karstpalaolithicum in Norwestliche Yugoslavien, Actes du IV. Congres International du Quaternaire, Roma, Italy, 737-742 Brook G.A., Burney D.A. and Cowart J.B. (1990) Desert paleoenvironmental data from cave speleothems with examples from the Chihuahuan, Somali-Chalbi, and Kalahari deserts, Pelogeography, Paleoclimatology, Paleoecology, 76, 311-329 Clark I.D. and Fritz P. (1997) Environmental isotopes in Hydrology, Lewis Publishers, New York, 311 Craig H. (1953) The geochemistry of the stable carbon isotopes, Geochimica et Cosmochimica Ada, 3, 53-92 Destombes J.L. Cordonnier M., Gadar J.Y. and Delannoy J.J. (1997) Periodic and aperiodic forcing of water flow through sodastraw stalactites (Choranche, Vercors, France), in Union International de Speleologie, eds., Proceedings of the 12th International Speleological Congress of Chaud-Les-Fonds, Neuchatel, Switzerland 14 Dorr H. and Munich K.O. (1986) Annual variations of the C content of soil CO2, Radiocarbon, 28, 338-345 Drake J.J. (1984) Theory and model for global carbonate solution by groundwater, In Groundwater as a Geomorphic Agent, LaFleur R.G. (ed), London: Allen & Unwin, 210-226 Dreybrodt W., (1980) Deposition of calcite from thin films of natural calcareous solutions and the growth of speleothems, Chemical Geology, 29, 89-105
85 Dreybrodt W., (1981) The kinetics of calcite precipitation from thin films of natural calcareous solutions and the growth of speleothems: revisited, Chemical Geology, 32, 237-245 Dreybrodt W., (1988) Processes in Karst systems, Springer, Berlin, 288p. Dulinski M. and Rozanski K. (1990) Formation of 13C/12C isotope ratios in speleothems: a semi- dynamic model, Radiocarbon, 32, 7-16 Epstein S., Lowenstam H., Buchsbaum R. and Urey H.C. (1953) A revised carbonate-water isotopic temperature scale, Geol. Soc. Am. Bull., 64, 1315-1326 Epstein S. and Mayeda T (1953) Variations of 18O contents of water from natural sources, Geochim. Cosmochim. Ada, 4, 213-224 Ford D. and Williams P. (1989) Karst Geomorphology and Hydrology, London, Chapman & Hall, 601 p. Fontes J.C. (1992) Chemical and isotopic constraints on 14C dating of groundwater, In Taylor, R.E., Long A. and Kra R.S., eds, Radiocarbon after four decades, New York, Springer-Verlag, 242-261 Fornaca-Rinaldi G., Panichi C. and Tongiorgi E. (1968) Some causes of the variation of isotopic composition of carbon and oxygen in cave concretions, Earth and Planetary Sciences Letters 4, 321-324 Franke H.W. (1975) Corespondence between sintering and corrosion, Ann. Speleol., 1,89 Friedman I. and O'Neil J.R. (1977) Compilation of stable isotope fractionation factors of geochemical interest, U.S. Geol. Surv, Prof. Pap., 440-KK:3 Gams I. (1973) The terminology of the types of polje, Slovenska terminologija Krasa, Zveza geografskih institucij, Ljubljana Gams I. (1981) Contribution to morphometrics of stalagmites, Proc. Int. Congr. Speleol., 8, 279- 278 Garres R.M. and Christ C.L. (1965) Solutions, minerals and equilibria, New York: Harper & Row Gascoyne M. (1977) Trace elements in calcite - the only cause of spelothem colour?, Natl. Speleol. Soc. Annu. Conv., Proc, 39-42 Gascoyne M., Schwarcz H.P. and Ford D.C. (1980) A paleotemperature record for the mid Wisconsin in Vancuver Island, Nature, 285, 474-476 Gascoyne M., Ford D.C. and Schwarcz H.P. (1983) Rates of cave and landform development in the Yorkshire dales from speleothem age data, Earth Surface Processes and Landforms, 8, 557-568 Genre M., Hoefling R., Kowski P. amd Strauch G. (1996) Sample preparation device for quantitative hydrogen isotope analysis using chromium metal, Anal. Chem., 68, 4414-4417 Genty D. (1992) Les speleothemes du tunnel de Godarville (Belgique) - un exemple exceptionnel de concretionnement moderne - interet pour I'etude de la cinetique de precipitation de la calcite et de sa relation avec les variations d'environnement, Sp6l6ochronos, 4, 3-29. Genty D. (1993) Mise en evidence d'alternances saisonnieres dans la structure interne des stalagmites, Intkret pour la reconstitution des pal6oenvironnements continentaux. Comptes Rendus de I' Academie des Sciences de Paris, s6rie II, 317, 1229-1236. Genty D. and Massault M. (1997) Bomb 14C recorded in laminated speleothems-part I.: dead carbon proportion calculation, Radiocarbon, 39, 33-48 Genty D., Vokal B., Obelic B. and Massault M. (1998) Bomb 14C time history recorded in two modern stalagmites-importance for soil organic matter dynamics and bomb 14C distribution over continents, Earth and Planetary Science Letters, 160, 795-809 Genty D. and Deflandre G. (1998) Drip flow variations under a stalactite of the Pere Noel Cave (Belgium). Evidence of seasonal variations and air pressure constraints. Journal of Hydrology, 211, 208-232 Gewelt M. (1985) Cintique du concretionnement dans quelques grottes beiges: Apport des datations 14C et Th/234U, Annales de la SociGte Ge~ologique de Belgique, 108, 267-273
86 Gewelt M., (1986) Datation 14C des concrétions de grottes belges:Vitesses de croissance durant l'Holocène et implications paléoclimatiques, In Paterson, K. and Sweeting, M.M. New Directions in Karst: Proceedings of the Anglo-French Karst Symposium, September 1983, Norwich, geo Books: 293-322 Gewelt M. and Ek C. (1988) les concretions carbonatees des grottes: aperçu syntetique, Annales de la Société géologique de Belgique, 111, 9-19 Goede A., Green D.C. and Harmon R.S. (1982) Isotopic composition of precipitation, cave drips and actively forming speleothems at three Tasmanian cave sites, Helictite, 20, 17-27 Goede A. (1994) Continuous early last glacial paleoenvironmental record from a Tasmanian speleothem based on stable isotope and minor element variations, Quaternary Science Rewiews, 13,283-291 Gospodariè R. (1972) Prvi podatki o absolutni starosti sige v Postojnskl jami na podlagi 14C, Naäejame, 13, 91-98 Gospodaric R. (1976) The quaternary cave development between the Pivka Basin and Polje of Planina, Acta Carsologica, VII/1, 5-121 Gospodaric R. (1977) Generacije sige v nekaterih kraëkih jamah, Geografija Krasa IZRK SAZU Postojna, 1-31 Gospodaric R. (1981) Die Entwicklung der Karsthöhlen in der Umgebung von Postojna im jüngeren Pleistozän, International Speleological Meeting, Athens, Greece, 219-229 Gospodaric R. (1985) O speleogenezi Divaske jame in Trhlovce, Acta Carsologica, 13, 5-32 Gospodaric R. (1988), Paleoclimatic record of cave sediments from Postojna Karst, Annales de la Société géologique de Belgique, T.111, 91-95 Gran G. (1952) Determination of the equivalence point in potentiometric titrations, Analyst, 77, 661-670 Hable P. (1982) Pregledna speleoloèka karta Slovenije, Acta Carsologica, 10, 5-22 Harmon R.S. and Curl R.L. (1978) Preliminary results on growth rate and paleoclimate studies of a stalagmite from Ogle Cave, New Mexico, A/SS Buletin, 40, 25-26 Harmon R.S., Schwarcz H.P. Ford D.C. and Koch D.L. (1979) An isotopic paleotemperature record for late Wisconsinan time in northeast Iowa, Geology, 7, 430-433 Harmon R.S. and Schwarcz H.P. (1981) Changes of 2D and 18O enrichment of meteoric water and Pleistocene glaciation , Nature, 290, 125-128 Hendy C.H. and Wilson AT. (1968) Paleoclimatic data from speleothems, Nature, 219, 48-51 Hendy C.H. (1971) The isotopic geochemistry of speleothems -I. The calculation of the effects of different modes of formation on the isotopic composition of speleothems and their applicability as palaeoclimatic indicators, Geochimica et Cosmochimica Acta, 35, 801-824 Hill C.A. and Forti P. (1986) Cave minerals of the world, Huntsville, USA, Nat. Speleo. Soc. Hobbs S.L. and Smart P.L. (1986) Characterization of carbonate aquifers: A conceptual base, In Union International de Spéléologie, ed., Proceedings of the 9 International Speleological Congress of Barcelona 1, 43-46 Hoefes J. (1987) Stable Isotope Geochemistry, Minerals and rocks, 9, Ed. Wyllie P.J., El Goresy A. and von Engelhardt W., Springer-Verlag New York, Berlin, Heidelberg, p. 241 Holland H.D. Kirsipu T.W. Huebner J.S. and Oxburgh U.M. (1964) On some aspects of the chemical evolution of cave waters, Journal of Geology, 72, 36-67 Issar A.S., Govrin Y., Geyh M.E., Wakshal E. and Wolf M. (1991) Climate changes during the Upper Holocene in Israel, Isr. J. Earth Sei., 40, 219-223 Kogovsek J. and Habiö P. (1981) The study of vertical water percolation in the case of Postojna and Planina Caves, Acta Carsologica, 9, 129-148 Kogovsek J. (1983) Prenikanje vode in ¡zloöanje sige v Pisanem rovu Postojnske jame , Acta Carsologica, 11, 59-76
87 Kogovšek J. and Kranjc A. (1989) Vpliv kislih padavin na procese v Postojnski jami Acta Carsologica, 18, 222-238 Kogovšek J. (1996) Vertikální odtok padavin skozi vrtače. Primer vrtač nad Planisnko in Pivko jámo, Acta Carsologica, 25, 158-168 Krajcar Bronič I., Horvatinčič N. and Obelič B. (1998) Two decades of environmental isotope records in Croatia: reconstruction of the past and prediction of future levels, Radiocarbon, 40, 399-416 Kranjc A. (1997) Brief account of the development of speleology in Slovenia (to 1914), Acta Carsologica, 26, 99-107 Lauritzen S.E. and Gascoyne M. (1980) The first radiometric dating of Norwegian stalagmites - evidence of pre-Weichselian karst caves, Norsk. Geogr. Tidsskr, 34, 77-82 Lauritzen S.E., Lovlie R., Мое D. and Ostbye E. (1990) Paleoclimate deduced from a multidisciplinary study of a half-million-year-old stalagmite from Rana, northern Norway, Quaternary Research, 34, 306-316 Lauritzen S.E., Haugen J.E., Lovlie R. and Gilje-Nelsem (1994) Geochronological potential of Isoleucine epimerization in calcite speleotheims, Quaternary Research, 41, 52-58 Langmuir D. (1971) The geochemistry of some carbonate groundwaters in central Pennsylvania, Geochim et Cosmochim Acta, 35, 1023-1045 Langmuir D. (1984) Physical and chemical characteristics of carbonate water, In Guide to the hydrology of carbonate rocks, UNESCO Studies and Reports in Hydrology ser, 69-131 Levin I., Bösiger R., Bonani G., Francey R.J., Kramer В., Münnich K.O., Suter M, Trivett N.B.A. and Wölfli W. (1992) radiocarbon in atmospheric carbon dioxide and methane: Global distributions and trends, In Taylor, R.E., Long A. and Kra R.S., eds, Radiocarbon after four decades, New York, Springer-Verlag, 503-518
14 Levin I. and Kramer B. (1997) Twenty years of high precision atmospheric CO2 observations at Schauinsland station, Germany, Radiocarbon, 39, 205-215 McCrea J.M. (1950) On the isotopic chemistry of carbonates and a paleotemperature scale, Jour, of Chemical physics, 18, 849-857 Mook W.G. (1980) Carbon-14 in hydrogeological studies, In Fritz P. and Fontes J. Ch., eds., Handbook of Environmental Geochemistry 1A, Amsterdam, Elsevier Scientific Publishing Co.: 49-74 Mook W.G., Brommerson J.C. and Staverman W. H. (1974) Carbon isotope fractionation between dissolved bicarbonate and gaseous carbon dioxide, Earth Planet. Sei. Lett, 22, 169- 176 Moore G. W. (1952) Speleothem- A new cave term, Nat. Speleol. Soc. News, 10 (6) 2 Nier A.O. (1950) A redetermination of the relative abudances of the isotopes of carbon, nitrogen, oxygen, argon and potassium, Physical Review, 77, 789 O'Neil J.R., Clayton R.N. and Mayeda T. (1969) Oxygen isotope fractionation in divalent metal carbonates, J. Chem. Phys., 30, 5547-5558 Papastefanou С and Charalambous (1989) Dating of spelothems by radiological methods, Nuclear Instruments and Methods in Physics research, A281, 406-409 Pazdur A., Pazdur M.F., Pawlyta J. Gorny A. and Olszewski M. (1995) Paleoclimatic implications of radiocarbon dating of speleothems from the Cracow-Wielun Upland, Southern Polanf, Radiocarbon, 37, 103-110 Picknett, R.G., Bray L.G. and Stenner R.D. (1976) The chemistry of cave waters, In. T. D. Ford and C.H.D. Cullingford (ed), The Science of Speleology, Academic press, London Pitty A.F. (1966) An Approach to the Study of Karst Water, University of Hull, Ocasional Papers in Geography, 5, 70p. Plummer L.N., Wigley T.M.L. and Parkhurst D.L. (1978) The kinetics of calcite dissolution in
CO2-water system at 5 to 60 °C and 0.0 to 1.0 atm CO2, Am. J. Sei., 278, 179-216
88 Plummer L.N. and Busenberg E. (1982) The solubilities of calcite, aragonite and vaterite in CO2- H2O solutions between 0 and 90 °C, and an evaluation of the aqueous model for the system CaCO3-CO2-H2O, Geochim et Cosmochim Acta, 46, 1011-1040 Rafter T.A. and Fergusson G.J. (1957) Atom bomb effect-Recent increase of carbon-14 content of the atmosphere and biosphere, Science, 126, 557-558
Reichter B. and Trimborn P. (1995) Application of tracer techniques to characterize hydraulics and solute transport of the epikarst zone of a karst aquifer, In. Proceedings of the International symposium on Isotopes in Water Resource Management, Vienna, IAEA, 303-305 Roberts M.S., Smart P.L. and Baker A. (1998) Annual trace element variations in a Holocene speleothem, Earth and Planetary Science Letters, 154, 237-246 Rozanski K., Araguas-Araguas L. and Gonfiantini R. (1993) Isotopic patterns in modern global precipitation, Geophysical Monograph, 78, 1-36 Obelic B., Krajcar-Bronic I., Srdoc D. and Horvatincic N. (1986), Environmental 14C levels around the 632 Mwe nuclear power plant Krsko in Yugoslavia, Radiocarbon, 26, 644-648 Orr P. C. (1956) Excavations in Moaning Cave, Santa Barbara Museum of Natural History Bulletin, 1, 1-19 Saliege J.F. and Fontes J. Ch. (1984) Essai de determination experimental du fractionnement des isotopes 13C et 14Cdu carbone au cours de processus naturals, Int. J. Apply. Radioact. Isot, 35, 55-62 Salomons W. and Mook W.G. (1986) Isotope geochemistry of carbonates in the weathering zone, In: Handbook of Environmental Isotope Geochemistry Vol 2.: The Terrestial environment, B, ed. Fritz P. and Fontes J.Ch., Amsterdam, Elsevier, 239-269 Schwarcz H.P., Harmon R.S., Thompson P. and Ford D.C. (1976) Stable isotope studies of fluid inclusions in speleothems and their paleoclimatic significance, Geochimica et Cosmochimica Acta, 40, 657-665 Schwarcz H.P., (1986) Geochronology and isotopic geochemistry of spleothems, In Handbook of Environmental Isotope Geochemistry Vol 2.: The Terrestial environment, B, ed. Fritz P. and Fontes J.Ch., Amsterdam, Elsevier, 271-303 Shopov Y.Y. and Dermendjiev V. (1990) Microzonality of luminescence of cave flowstones as a new indirect index of solar activity, Comptes Rendus de l'Acad6mie Buigare des Sciences, 43, 9-12. Shopov Y.Y., Ford D. C. and Schwarcz H. P. (1994) Luminescent microbanding in speleothems: High-resolution chronology and paleoclimate. Geology, 22, 407-410 Skirrow G. (1975) The dissolved gases-carbon dioxide, Chemical Oceanography, 2nd ed. (Riley J.P. and Skirrow G., Eds.), Academic Press, London, 1-104 Srdoc D., Slipeevic A., Obelic B. and Horvatincid N. (1979) Rudjer Boskovic institute radiocarbon measurements V, Radiocarbon, 21, 131-137 Stumm W. and Morgan J.J. (1981) Aquatic Chemistry, Chemical Equilibria and Rates in Natural Waters, Wiley & Sons edt., New York, 148-204 Stuvier M. and Polach H. (1977) Discussion: Reporting of 14C Data, Radiocarbon, 19(3), 355- 363 Suess H.E. (1955) Radiocarbon concentration in modern wood, Science, 122, 415-417 Talma A.S. and Vogel J.C. (1992) Late Quaternary paleotemperatures derived from a speleothem from cango Caves, cape Province, South Africa, Quaternary Research, 37,203-213 Thornthwaite C.W. (1954) The measurement of potential evapotranspiration, Mather, Seabroock, NJ, 225p. Thrope P.M., Otlet R.L. and Sweeting M.M. (1980) Hydrological implications from 14C profiling of UK Tufa, In. Stuvier M. and Kra R.S., eds., Proceedings of the 10th International 14C Conference, Radiocarbon, 22(3), 897-908 Trumbore S.E. (1993) Comparison of carbon dynamics in tropical and temperate soils using radiocarbon measurements, Global Biogeochemical Cycles, 7, 275-290
89 Urbane J., Pezdic J. Dolenec T. and Perko S. (1985) Isotopic composition of oxygen and carbon in cave water and speleothems of Slovenia, Acta Carsologica, 13, 99-112 Urey H.C. (1947) The thermodynamic properties of isotopic substances, J Chem. Soc, 562 Vogel J.C. (1983) 14C variations during the Upper Pleistocene, In. Stuiver M. and Kra R.S., eds., Proceedings of the 11th International C Conference, Radiocarbon, 25(2), 213-218 White W.D. (1976) Cave minerals and speleothems, In. T.D. Ford and C.M.D. Cullingford (Ed.), The science of speleology, Academic Press, New York, 329-358 Wigley T.M.L. and Brown M.C. (1976) Cave physics, In. T.D. Ford and C.M.D. Cullingford (Ed.), The science of speleology, Academic Press, New York, 329-358 Wollast R. (1993) Interactions of C, N, P and S Biogeochemical Cycles and Global Change, Ed. Wollast R., Mackenzie FT. and Chou L, NATO AS I Series, Series I, Global Environmental Change, Vol. 4., Springer-Verlag Berlin Heidelberg Zupan N. (1991) Flowstone datations in Slovenia, Acta carsologica, 20, 187-204 Zupan-Hajna N. (1993) Survey of published flowstone datations from Slovenia, NaSejame, 35, 83-87 Yonge C.J., Ford D.C., Gray J. and Schwarcz H.P. (1985) Stable isotope studies of cave seepage water, Chemical Geology (isotope geoscience section), 58, 91-106
90