2002-07-10KDL 16657



Past Vegetation Changes in Karst Areas as Revealed by a Comparative Isotopic Study of Two Holocene Speleothems from Romania

Silviu Constantin 1,Bogdan P.Onac 2, Dominik Fleitman 3&Tudor Tamas
1Institutul de Speologie “Emil Racovita ”,str.Frumoasa 11,78114 Bucuresti  ,Romania 
2Department of Mineralogy ,University of Cluj ,Kogalniceanu 1,and Speological Institute 
“Emil Racovita ”,Clinicilor 5,3400 Cluj. Romania. 
3Geologisches Institut.Universitat Bern ,Baltzerstrasse 1,3012 Bern ,Switzerland .

Extended Abstract

The isotopic composition (δ18O and δ13C) of speleothem calcite is now commonly accepted as bearing paleoclimatic signals although the complex processes controlling the isotopic variations are not always fully understood. This is especially true for the carbon profiles because their variations depend on many processes .The photosynthetic (C3 vs .C4) pathways are often discussed as a possible cause of the carbon shifts, although these shifts may also depend on other factors such as the biological activity, rainfall (which controls the drip rate in the cave) or the water-bedrock interaction (LAURITZEN &LUNDBERG, 1999).

In this paper we compare the isotopic profiles for oxygen and carbon of two Holocene stalagmites collected from two Romanian caves located within regions with similar karst landscape settings, but with different topoclimatic influences and, accordingly, different vegetation association .The first cave, Pestera Ursilor (=Bears Cave) is located at the outskirts of the Western Carpathians (Fig .1) under mild and humid climate conditions that are mainly influenced by a WNW (Atlantic) air-circulation .The second , Pestera Poleva is located in the Banat Mountains ,close to the western termination of the Southern Carpathians, in the  vicinity of the Danube Gorge. The climate here is temperate with Mediterranean influences, favoring the predominance of xerophytic vegetation.

Samples and results

PU2.a 72.5 cm-long candlestick stalagmite was collected from the upper (fossil) level of Pestera Ursilor; its base was dated by U-Th alpha spectrometry to 19.19(+2.22; -2.19) ka (ONAC& LAURITZEN, 1996). Five high precision U-series TIMS dates were performed for the top 17.5 cm; these were used for calculating the ages of 40 isotopes samples analyzed for oxygen and carbon along the same sector of the stalagmite. The profile spans the time-interval between 7.1 ka and ~1.05 ka including a hiatus of ~1000 years framed between 3.2 and 2.2 ka BP.The resolution of the isotopic sampling averages ~142 years below the hiatus but decreases to ~100 years above it (ONAC et al., submitted).

PP9, a very thin (2-3 cm) candlestick stalagmite, 32 cm-long, was removed from a fossil meander of Pestera Poleva.Four TIMS dates were performed between its base and cm 24;dating of a fifth subsample collected from the top of the stalagmite failed due to the very low U content. However, the ages show a fairly constant growth rate between 9.9 ka (±0.1) and 4.17(±0.05) ka at 24 cm; taking into account the homogeneous structure of the speleothem we considered acceptable to extrapolate the calculated growth rate in order to calculate the ages of the isotopic samples for the last 8 cm .In total , 64 isotopic samples have been collected every 0.5 cm along this stalagmite ;according to the interpolated ages , the time resolution of this profile varies between ~130 and ~150 years/samples .

For both stalagmites, the modern values for oxygen and carbon have been assessed by sampling several tips of nearby soda-straws and averaging the determined values. Since the  18O values  of these soda straws are considered to reflect the present-day , interglacial, temperatures, the relationship between the isotopic values of the profile and the temperature changes may be established: in both cases, this relation was found to be positive, that is, the “lighter ”the oxygen, the lower temperature .

Since the two isotopic profiles partially overlap each other, between 7.1 ka and 1.4ka, we plotted them on the same graph in order to see whether or not a correlation could be established .The oxygen profiles values are plotted against age  in Figure 2;for comparison ,a graph of  GRIP oxygen values (moving average with a period of 25 ) is also plotted . A strikingly good ccorrelation may be noticed between the isotopic variations recorded by the two speleothems. Several cold and warm episodes known from other speleothem records from northern Europe can easily be recognized. The profile of PP9 shows a trend of abrupt temperature increase between 10 ka and ~8 ka ,corresponding to the general climate warming at the beginning of the Holocene ;afterwards ,the general warming tendency becomes less evident until a clearly marked “warm spike ” at ca 3.5 ka that may also be recognized in the PU2 record .Subsequently ,two cold episodes may be noticed ,one at ~2.8 ka and one at ca. 2 ka, separated by a warm period, which probably corresponds to the warm ,  “Roman Empire ” period .

If we consider strictly the time interval along which the two records overlap, it may be also noticed that the oxygen isotopic values lie within the same range, that is, between -8.53 and -6.97‰ for PU2 and between -8.45 and -6.94‰ for PP9. Assuming that, in both cases ,the calcite has precipitated under isotopic  equilibrium ,this coincidence may be interpreted as representing roughly equal annual mean temperature ranges for both regions.

In contrast with oxygen ,the plot of carbon profiles against age shows very distinct ranges of variation (Fig.2).For PP9 ,the minimal value of  δ13C is slightly above -10‰ vs. PDB, but in general, the graph shows a tendency of increase up to ca 8 ka .Afterwards ,it remains at high values ,within the range of -9 up to -6.4‰ vs .PDB .On the contrary .the variation range of PU2 lies within the extreme values of -11.2 and - 9.5‰.However ,the  successions of peaks and troughs in both profiles show a good correlation .

Discussion and conclusions

The well-marked difference between the ranges of values for carbon corroborated with the similarity observed for the values of the oxygen suggests that the carbon variations may be less controlled by temperature (which, in turn, comtrols the biogenic activity in the soil) and, more dependent on the different vegetation assemblage of the two areas.

The present-day topoclimatic conditions of the two caves are similar in many ways (Table 1). Both caves are located in hilly areas, at low altitudes and have roughly similar position with respect to the limestone massifs .The multi-annual mean values of precipitation recorded for both areas are fairly close, the Ursilor are receiving slightly more rainfall (on the order of 100-200mm/y). Further on the multi-annual mean temperatures within the two areas do not differ by more than 1℃; the temperatures inside the caves are roughly equal ,at least at the speleothem collection sites.

The main differences between the two studied areas are given by the direction of the air circulation .In the western part of Romania, the climate is mainly influenced by the North Atlantic Oscillation, while in the south-west, on the other side of the Carpathians, it is mostly controlled by a Mediterranean circulation .As a result, although the two regions receive roughly the same amount of rainfall throughout the year, their monthly reparation is different; precipitation is more evenly distributed in time for the western regions while south-west of the country is reputed for its arid summers and mild ,rainy ,winters.

Table 1: Topoclimatic conditions of the two areas discussed in the paper


Pestera Ursilor

Pestera Poleva



massive reef limestone

Altitude m a. s. l.



Average thickness of

the limestone above cave



Mean annual precipitation mm/y



Mean multiannual temperature




- surface



- cave



Mean monthly temperature

for July (ºC)



Main atmosphere circulation

 and source

WNW (Atlantic)

ESE (Mediterranean)


mild winters,

rainy summers

mild winters, dry summers

Vegetation above the cave


mainly shrubs

deciduous; mainly forest

& xerophytes

calculated from a 2-years record from Pietroasa meteorological station (I. ORASEANU. pers comm. )

b interpolated from values quoted for Moldova Noua by MUNTEANU & BALANESCU (1999) and records of Girnic station, A. IURKIEWICZ (pers. Comm. )

c source: RACOVITA et al., 1999

d source: TISTEA et al., 1974.

e source: MIHALCA & STANCIU, 1999.

The different climatic regimes are well reflected into the composition of the vegetation .In the northwestern site, the karst landscape is covered by C3-deciduous association, including Quercus ,Carpinus and Juglans(ONAC et al., submitted ), while in the southwestern part the xerophytic and thermophylous  species are prevalent. Within the area around Poleva cave, the surface is covered by a decidous forest that includes submediterranean oak trees, such  as Quercus Pubescens and Q.ceris, and  thermophylous shrubs including species like Syringa vulgaris, Padus mahaleb, Cotinus coggygria and Fraxinus ornus (IANA & PETCU, 1976). The xerophytes are generally known to follow the C4 photosynthetic cycle (CLARK & FRITZ, 1997), therefore we may consider that the present plant association in southern Banat includes more C4 species than in the Western Carpathians although it is not clear yet if in Banat the C4 plants actually prevail over the C3 ones. The present-day value δ13C in Banat is close to -9‰.

Having established the present-day difference in vegetal composition of the two regions, it seems rational to assume that the discrepancy between the Holocene values of δ13C in the two speleothems is mainly due to the different photosynthetic pathways of the plants growing over the two karst areas.

The other principal factors that may influence the carbon variations, namely the biogenic activity, directly dependent on the temperature, and the drip rate, generally controlled by the rainfall, may be, in our opinion, safely ruled out. The comparative oxygen profiles showed that the temperatures in the two caves (and, by extrapolation, within the surrounding areas) were similar, the largest variations being less than 1‰. In order to verify that the variations in carbon content are not simply related to the variations of temperature we plotted in Fig. 2 the solar output curve based on a model elaborated by Perry & Hsu (2000): in some cases the correlation is good, suggesting a depletion of carbon linked to the higher temperature and thus biological activity (e.g. the peak at 6.5 ka or the trough at ca 2.2 ka), but in general it is poor. Therefore, we presume that the overall changes in temperature alone cannot explain the major shifts of δ13C recorded, at least in the case of stalagmite PP9.

On the other hand, since the calculated growth rates of the two speleothems are very similar (3~4 cm/ka), we may assume that, in general, the drip rate has not change such as to fundamentally influence the carbon isotopic composition. The common role of drip rate variation is the higher the drip rate, the more depleted the carbon. This may explain the accentuate decrease of δ13C along the topmost section of PU2(ca2-1 ka, see Fig.2), for which a higher growth rate(5.7cm/ka) has been calculated. As a result, δ13C droped almost 0.5‰ below the present-day value.

Taking into account the above considerations, we consider that the different vlues of δ13C determined for PP9 as compared with PU2 are mainly caused by the different vegetal associations within the two karst areas. Since the present-day values of δ13C (-8.8‰) corresponds to a mixed plant association including C3 and C4 species, we may assume that higher values (around -6.5‰) are indicative for a dominant C4 association, while values below -9.5‰ would correspond to C3 plants. We may therefore conclude that the last 7000 years. The moderate variations recorded in the isotopic profile may be caused by other factors controlling the carbon cycle, such as the variation of the drip rate. On the contrary, in the case

Figure 2: Top: variations of the δ18O for stalagmite PP9 (solid line) vs. PU2 (dashed, grey line) and a 25× moving average plot of GRIP isotopic values. The TIMS ages for PP9 are also plotted vs. stratigraphy (error bars, 2σ). Bottom: variations of the δ13C for PP9 vs. PU2 and the solar output model realized by Perry & Hsu (2000).


of southern Banat, we believe that the large carbon shifts (for example, from ca.-10‰ at ~8.8ka to -6.5‰ at ~6.5ka ) can only be explained by a major change in vegetal composition, from dominant C3 to dominant C4 . This might have been triggered by the set up of a continental, arid atmospheric circulation of African origin which generated a drier, meridional climate in southeastern Europe – possibly the same that has caused the first “Sahara Aridity” event approximately 6900 years ago.


Clark, I. & Fritz, P. (1997), Environmental Isotopes in Hydrogeology. Boca Raton, Lewis Publishers, 328p.

Lauritzen, S.-E. & Lundberg, J. (1999a) Speleothems and climate: a special issue of The Holocene, The Holocene, 9, 6, pp. 643-647.

Iana, S. & Petcu, A (1976) Caracterizare biogeografica. In Geografia. Grupul de Cercetari complexe “Portile de Fier” (Iancu, M, Ed.), Ed. Academiei RSR, Bucuresti, pp.125-137.

Mihalca, D. & Stanciu, E. (1999) Study of the parameters of thermic regime in the Romanian Banat region. Proc. Regional Conf. of Geography “Danube-Cris-Mures-Tisa Euroregion—Geoeconomical space of auatainable development” Timisoara, 1998, West University, Timisoara, pp.131-140.

Munteanu, R. & Balanescu, D. (1999) Air and rainfall temperature regime in the space of Banat in the year 1997. In Proc. Regional Conf. of Geography “Danube-Cris-Mures-Tisa Euroregion –Geoeconomical space of sustainable development”, Timisoara, 1998, West University, Timisoara, pp.141-15.

Onac, B. P. & Lauritzen, S.-E. (1996) The climate of the last 150,000 years recorded in speleothems: preliminary results from north-western Romania. Theoretical and Applied Karstology, 9, pp.9-21.

Onac, B. P, Constantin, S., Lundberg, J. & Lauritzen, S.-E. (submitted) Isotope climate record in a Holocene Stalagmite from Ursilor cave (Romania)

Perry, C. A. &Hsu, K. J. (2000) Geophysical, archaeological, and historical evidence support a solar-output model for climate change. Proc. Natl. Acad. Sci. USA, 97, 23, pp.12433-12438.

Racovita, G., Moldovan, O. & Rajka, G. (1999) Données préliminaires sur l`environnement de la Grotte des Ours en régime d`exploitation touristique. Theoretical and Applied Karstology, 11-12, pp. 61-74.

Tistea, D., Dinca, I., Cazacu, G., Sirbu, V., Calinescu, N., Neamu, G. & Teodoreanu, E. (1974) Temperatura aerului. In Atlas Republica Socialista Romania, Map IV-2, Ed. Academiei RS Romania.


Comparision of Tufa from Three Locations in Croatia in Terms of Mineralogy, 
Major and Trace Elements and Specific Activity

S. Frančišković-Bilinski, D. Barišić, A. Vertačnik, H.Bilinski
( Institute “Ruder Bošković”, Zagreb, Croatia)

E. Prohic
(Faculty of Science, Division of Mineralogy and Petrology, Zagreb, Croatia)


The process of tufa formation in Croatia has been studied the most intensively in Plitvice Lakes and Krka River by Srdoč et al .(1985), Horvatinčić et al.(1989)and many others. In the recent paper by Horvatinčić et al.(2000), ages were reported by U-Th and 14C dating methods. It was concluded that the growth of tufa is restricted to the warm periods of global climate and was interrupted during glacial , stadial and interstadial periods. In the present paper selected samples of tufa from three different locations (Plitvice Lakes, Krka River and Mrežnica River) in Dinaric Karst of Croatia were characterized in terms of mineralogical composition , major and trace elements and specific activity. All tufa deposits contained calcite as the major mineral. From minor minerals dolomite was found in tufa from Krka River and quartz in tufa from Mrežnica River.

The content of calcium carbonate varied form 71.5 to 97.5%, of magnesium carbonate from 0 to 6.6%. From other available elements the most abundant is iron, from 0.016 to 2.7%. Many other trace elements are also present.

Specific activity of radionuclides 40K, 232Th,137Cs,226Ra and 238U showed different concentrations in all samples . In most samples, particularly in Plitvice Lakes, the ratio 238U/236Ra deviates from theoretical value 1, suggesting different transport of this elements into tufa. Specific activity of 137Cs does neither correlate with mineralogical composition, nor with the age determined by Horvatinčić et al. (2000). Relatively high 137Cs concentration in some samples can be explained by pollution after accident in Tchernobyl in 1986.


Horvatičić N. Srdoč D. Šilar J. and Tvrdikova H.(1989) Radiocarbon,31, 884-892

Horvatičić N. Srdoč R. and Geyh M.A.(2000)Quaternary Research ,53, 185-195.

Srdoc D, Horvatinčić N, Obelić B , Krajcar Bronić I . and Slijepčević A(1985) Carsus Iugoslaviae, 11, 101-204.


Research of Selected Tufa from Guangxi Province (China) in Terms of Mineralogy, 
Major and Trace Elements, Specific Activity and Age

Stanislav Frančišković-Bilinski, Halka Bilinski, Delko Barišić, Nada Horvatinčić
(“Rudjer Bošković”Institute, Zagreb, Croatia)

Yuan Daoxian
(Institute of karst geology, Guilin, Guangxi, P.R. China)


It was reported by Wang et al. (1999) that South China is the largest continuous distribution area of carbonate rock in the world.

Yuan(1997) and Jiang and Yuan(1999) initiated karst research in this region, which could contribute to global climate studies and also to paleoenvironmental change. In the present joint work selected tufa samples were chosen in Guangxi province, which represent characteristic examples of subtropical karst tufa and have not been described before.

In terms of mineralogy the major mineral is calcite. Other minor minerals are quartz and dolomite, depending on location. The content of calcium carbonate varied from 65 to 92%, of magnesium carbonate from 0.03 to 1.77%.

All samples contain traces of sodium and potassium(%Na>%K), of iron , titanium, manganese and many other elements.

Specific activity of redionuclides 40K,232Th,137Cs, 226Ra and 238U showed different concentrations in all samples . Low concetration of 137Cs suggests that this part of the world was not exposed to Tzernobyl accident. The ratio 238U/226Ra in samples from Mashan county deviates from theoretical value 1, suggesting different transport of these elements into tufa.

14C age was performed only on 5 out of 11 studies samples. These results can be described as preliminary, due to the unknown initial activity (A0). One sample originates from Pleistocene and other samples from Holocene. From low K+ and Na+ value and K+<Na+, according to classification of Liu and He (1994), it was suggested that studied tufa from Guangxi province belong to the CO2-outgassing “N” type.


Jiang Z.C. and Yuan D.X.(1999) Episodes,22,33-35.

Liu Zai-hua and He Dian-bin (1994) Chinese Science Bulletin, 39,1468-1472.

Wang S.J.,Ji H.B., Ziyuan O.Y., Zhou D.Q., Zhen L.P. and Li T.Y.(1999) Sci. China Ser. D-Earth Sci.,42, 572-581.

Yuan D.X.(1997) Quaternary International, 37,105-113.


Influence of the 14C and 3H Global Atmospheric Contamination on the Karst Region of Croatia

Nada Horvatinčić, Ines Krajcar Bronić and Bogomil Obelić
Ruder Bošković Institute, Radiocarbon and Tritium Laboratory,
Zagreb , Croatia


Karst region of Croatia is widespread covering approximately one half of the whole country . The most part of the area is scarcely populated and environmentally clean area rich with groundwater, and represents favorable groundwater storage reservoir.

Environmental isotope (radioactive isotopes 3H and 14C) in karst region of Croatia with application in hydrology and ecology will be presented. 3H activity in monthly precipitation for 1 year period at 5 different stations in karst area of the Adriatic coast are used as a datd base for hydrological investigations, as well as for hydrometeological studies of the area .The study is a part of the International Atomic Energy Agency Co-ordinated Resesrch Project: “Isotopic composition of precipitation in the Mediterranean Basin in relation to air circulation patterns and climate”. The data are used also as the background of the 3H distribution in the atmosphere for controlling any contamination by nuclear sources, on local or global level.

The global 3H and 14C contamination of the atmosphere (produced by nuclear bomb-tests in Sixties) was recorded in the karst region of Croatia, Increase of 3H activity in a 45m long ice deposit found in a cave in Mt Velebit, as well as increase of 14C activity in the recent lake sediments, recent tufa and surface water in the Plitvice Lake area, reflects the global contamination. Comparison with 3H activity in precipitation and with 14C activity in atmospheric CO2, respectively, showed that global atmospheric contamination was reflected in the karst environment but with damped response and with some delay.


A 4000-year Palaeoclimate Record from
Scarisoara Glacier Cave (Romania)

Gheorghe RACOVITA1 & Bogdan P. ONAC1,2 
(1 Speleological Institute, Clinicilor 5, 3400 Cluj, Romania) 
(2 University of Cluj, Dept. of Mineralogy, Kogalniceanu1, 3400 Cluj, Romania)

Scarisoara Glacier Cave hosts a massive fossil ice deposit with a volume of 75,000 m3 is present. This ice has a stratified structure, consisting of alternating layers of ice formed during winters and layers containing impurities deposited during summers, This structure is dependent on seasonal variation of the underground temperature, determined by the exterior climate. Monthly measurements of the ice level between 1982-1992, allowed determination of the quantitative relation between subterranean and external temperature and of the time step, whose periodicity, noticeable in the stratigraphic profile, can be correlated with climatic oscillations. The stratigraphic profile thus reproduces climate oscillations that occurred ill the second half of the Holocene. A relatively warmer and rainier climatic episode that caused the Histrian transgression in the first millennium of our era was identified. Three shorter phases of climate warming took place during the fast 300 years.

The ice block hosted by the Scarisoara Glacier Cave was subjected to various observations during the last 30 years. These include readings of the relative humidity, temperature (in the air within the cave, and at different depths in bedrock and ice), evapo-condensation, ventilation, ice limits variation, pollen investigations, 14C dates on organic material extracted from the bottom of the ice block, and also few isotope analyses on different ice layers.  In addition, observations on the multi-annual dynamics of the ice speleothems (mostly stalagmites), U/Th dating of flowstones and stalagmites, as well as some paleokarst observations were also done.

The ice that constitutes this block is not homogeneous, but has specific stratified structure due to the seasonal variation of air temperature. During the summer the floor of Great Hall covers with a water film a few centimetres thick, supplied by both percolation and partial thawing of the ice and snow from the bottom of the entrance shaft. In this film limestone dust from frost-weathering of the cave roof accumulates together with other materials (soil particles and vegetal remains, including allochtonous pollen grains), The next winter the film freezes, resulting in a layer of new ice, under which the layer of sediment impurities remains. The whole block has an alternation of ice and impurity layers. Layers appeared from the moment in which the ice began to deposit into the cave, each pair representing an elementary, stratigraphic unit, corresponding to one year.

Given the specific mechanism of genesis, the morphological and structural parameters of each of0rose units (thickness, nature of material, the presence of solid or gaseous inclusions, isotopic composition etc) arc closely related to the topoclimatic conditions under which were formed, which in turn depend un external meteorological factors. Long term observations showed that the optimum topoclimate for formation of an ice layer does not correspond to the most severe winters, but to milder and longer winters with alternations periods of' freezing and melting. The characters of the impurity layers depends only on the length of the summer season and the abundance of precipitation, but not on external temperatures which do not influence the temperature of the underground atmosphere during the summer.

The second reason why the ice block has palaeoclimatic importance is the age of this deposit. A pollen analysis of a 759 cm long profile from the base of the northern flank of the block showed that the ice deposit was formed about 3,000 years ago, in the postglacial beech phase, with a colder and moister climate than present. There are arguments that suggest the age of the ice block must be even greater. The true base of the ice is at least 5 m lower than the lowest limit of the profile that had been analysed. As this thickness represents a third of the 15 m height of tile northern flank of the block and assuming a linear growth of the ice, the age of the first ice layers could be as old as 4,000 years.