lan J. Fairchild^a,Andrea Borsato^b,Frank McDermott^c, Silvia Frisla^b, Anna F. Tooth^a,
Christopher J. Hawkesworth^d , Timothy H.E. Heaton^e, Yiming Huang^a,d,
Eddy Keppens^f, Antonio Longinelli^g, Baruch Spiro^e,
Klaas van der Borg^h and Sophie Verheyden^f
aDepartment of Earth Sciences,Keele University,Staffs ST5 5BG,UK
^bMuseo Tridentino di Scienze Naturali,Via Calepina 14,38100 Trento,Italy
^cDepartment of Geology,University College,Belfield,Dublin 4,Ireland
^dDepartment of Earth Sciences,Open University,Milton Keynes,MK76AA,UK
^eNERC Isotope Geology Laboratory,Keyworth,Nottingham NG12 5GG,UK
^fDepartment of Geology(WE-CHRO),Vrije Universiteit Brussel,Pleinlaan 2,1050 Brussel,Belgium
^gIstituto di Mineralogia e Petrografia,Universita di Trieste,Piazzale Europa 1,34100 Trieste,Italy
^hDepartment of sub-Atomic Physics,University of Utrecht,Netherlands
(to be published in Earth & Planetary Science Letter)

Abstract At each of four European caves(Crag,SW Ireland,Père-Nöel,SW Belgium;Clamouse,S France and Ernesto,NE Italy), cave drip and pool waters were collected and sampled at intervals over a 2-3 year period. Mg/Ca and Sr/Ca concentration ratios are compared with those of bedrocks and, in some cases, aqueous leachates of soils and weathered bedrocks. Cave waters do not lie along mixing lines between calcite and dolomite of bedrock carbonate.THis is mainly because of two effects:1)selective leaching during weathering under prolonged water-soild contact times by one or more processes which tend to enhance Mg/Ca and Sr/Ca ratios and 2) prior calcite precipitation in the karstic system which results in characteristic enrichments in Sr/Ca and Mg/Ca as aqueous Ca²⁺ falls.The net effect is a sympathetic Mg/Ca-Sr/Ca covariation at individual sites even where dolomite-calcite mixtures display antipathetic variation. Comparison of the cave sites utilizing a standard form of presentation contrasts the small weathering-related chemical variations at Ernesto with the large spatial variability and influence of calcite precipitation at Père-Nöel. Clamouse displays strongly fractionated waters with less hydrologic variation, whilst Crag shows the influence of calcite precipitation, but has more complex sources of trace elements. Although elemental chemistry tends to vary greatly within a cave system, the patterns of variations are consistent and explicable, and provide a basis for interpretation of trace element trends of fossil speleothems in terms of palaeohydrology and hence variations in the balance of atmospheric precipitation and evaporation in the past.

Corresponding author E-mail:i.j.fairchild@keele.ac.uk

N.Horvatincic, R. Bistrovic and B. Obelic
Ruder Bosk ovic Institute, Zagreb

From Acta Geologica Hungarica, 39, Suppl., 1996 (Isotope Workshop ///) × 77

Travertine or tufa, calcareous deposits precipitated from fresh water, is the result of decomposition of dissolved calcium bicarbonate in cold fresh water. The process of calcium carbonate precipitation depends on physicochemical conditions (temperature, pH, alkalinity, concentration of calcium, magnesium and dissolved CO₂ in water) as well as biological, hydrogeological and climatological conditions.

We investigated the physicochemical conditions for travertine precipitation at Plitvice Lakes area as well as isotopic measurements including ¹⁴C, ¹³C and ¹⁸O in water and travertine samples (Srdoc et al, 1985). d ¹³C values of travertine ranging from -9.6 to -8.7‰ confirmed the biogenic origin of travertine, which makes the radiocarbon dating of travertine samples possible.

During the precipitation of calcium carbonate from fresh water the trace amounts of uranium are incorporated in travertine. Thus, it was possible to determine the age of travertine beyond the ¹⁴C range by the ²³⁰Th/²³⁴U dating. Results of ¹⁴C and ²³⁰Th/²³⁴U dating are reliable if the travertine samples belong to the so called "closed system".

In this paper we present the results of radiocarbon and ²³⁰Th/²³⁴U dating of travertine samples from two karst areas in Croatia: the Plitvice Lakes and the Krka River. Both areas are situated in the Dinaric Karst region and belong to the National Parks. Recent travertine was found along streams and lakes, and its precipitation is presently very intensive, whereas old travertine deposits were found as outcropping above the present water level.

Table 1:    Results of ²³⁰Th/²³⁴U dating of travertine from the Plitvice National Park and the Krka National Park.

Figure 1. Frequency of ¹⁴C results of travertine samples collected in the Plitvice National Park and Krka National Park. The results are expressed in pMC and in years BP ( Calculate by A0=85 pMC )

Kazuko URUSHIBARA-YOSHINO¹ and MASATOSHI YOSHINO²

1Komazawa University, 1-23-1 Komazawa, Setagayaku, Tokyo, 154, Japan
²Aichi University, 1-1 Machihatacho, Toyohashi, 441, Japan

(From J. Quarternary Sci., Vol. 12, 435-442)

ABSTRACT: Palaeoenvironment in Java island and its surrounding areas since the last interglacial stage was reconstructed initially using previous studies by various researchers. The karst area in Gunung Sewu on Java. island was also examined by the authors in order to reconstruct the palaeoeiivironment through the process of karstification. Summarizing the evidence related to palaeoclimate in the various areas made it clear that the palaeoclmiatic condition was cooler and drier at the Last Glacial Maximum 18000yr BP,than today. Formation of dry valleys in Gunung Sewu was conditioned by lowering of sea-level, and establishment of a cool and extremely dry climate. Subsequent to the initial reconstruction, the results of the estimated palaeoclimate at 18000yr BP were plotted at the respective points where the samples were obtained, and maps for northern winter and northern summer were constructed showing the estimated streamlines of monsoon circulations, polar frontal zones and intertropical convergence zones, in order to better explain the distribution of the palaeoclimate conditions of the area.

by K.Holmgren ¹⁾, W.Karlén ²⁾, S.E.lauritzen ³⁾, J. Lee-Thorp ³⁾,T.C.Partridge ⁴⁾, P.A..Shaw ⁵⁾ and P. D. Tyson ⁴⁾
1)Department of Geology, University of Bergen, Allégt. 41, 5007 Bergen, Norway
2)Department of Physical Geography, Stockholm University, S- 10691 Stockholm, Sweden
3)Department of Archaeology, University of Cape Town, Private Bag, Rondebosch, Cape 7700, South Africa
4)Climatology Research Group, University of the Witwatersrand, Wits 2050, Johannesburg, South Africa
5) Department of Environmental Science, University of Luton, Park Square, Luton LU 1 3JU, UK

(paper presented at 12^th ICS, Switzerland, 1997)

Holocene arid Pleistocene speleothems from South Africa, Botswana and Tanzania have been uranium-series dated and analysed on their stable isotopic content. Most of the stalagmites sampled are composed of finely laminated dense calcium carbonate. The structure and periodicity of the growth laminae are presently being analysed. New results from this study are presented and compared with data already available.

Speleothems are often suitable climate archives arid provide an option to obtain terrestrial high-resolution palaeoclimatic records from all over the world (SCHWARCZ, 1986). During the last twenty years several highly resolved palaeorecords from ice cores from Antarctica and Greenland have been presented and inferences have been drawn regarding global Late Quaternary climatic changes, global symmetry and driving forces (LOWELL et al., 1995). There is an urgent need to obtain good quality terrestrial records of equally high resolution from low latitudes in order to test hypotheses about global climate changes arid forcing factors. It is hoped that this recently initiated project will contribute such information through detailed analysis of spelaothems from Tanzania, Botswana and South Africa.

Tanzania is situated just south of the equator and has a tropical climate influenced by the seasonal latitudinal movements of the Intertropical Convergence Zone and the monsoonal winds. Previous theories that climate in the tropics remained stable during the last ice ages have been proven incorrect (GUILDERSON et al, 1994). Very few terrestrial high-resolution climate data have so far been presented from this region, but studies from caves from Somalia arid Zaire (BROOK et al, 1990), glacier studies from Mt Kenya (KARLÉN & ROSQVIST, 1988; MAHANEY, 1990), and lake studies from Kenya and Tanzania (BONNEFILLE et al., 1995) generally suggest wetter conditions during the Eemian interglacial, aridity and cooler temperatures during the last degladation, wetter conditions at the Pleistcene-Holocenc transition, a climate optimum in the early Holocene arid since then a trend toward drier climates.

We have sampled stalagmites from cave systems along Tanzanian coast; the Amboni Caves and the Matumbi Caves, and from caves on a small coral island outside the coast; The Songo-Songo Island (figure 1 ). Uranium-series dating so far has yielded ages from recent to the Last Glacial Maximum and from 38 to 34 ka (1000 years) (figure 2). The stalagmites are well laminated, with growth rates varying from 0.5 mm per 100 years for the Holocene samples, to 7 mm per 100 years for one that grew between 38 and 34 ka. Considering the high growth rate on the latter it is expected that a yearly resolution will be obtained from the laminae analysis and a 10 year resolution from the stable isotope analysis. This work is in progress.

The speleothems studied from Botswana arid South Africa are from cave systems in semi-arid central southern Africa (figure 1). This region is of interest due to its

Transitional position between the summer rainfall zone to the north, the winter rainfall zone to the south, the semi-arid Kalahari Basin to the west and the more humid plateau to the east. Clearly, the climate and the borders of these zones have varied in response to past global climate changes. The limitations of most of the previous palaeoenvironmental data available are poor chronological control, low time resolution, arid poor continuity (Thomas &Shaw, 1991; Partridge, 1995).

Uranium-series dating, close-interval stable oxygen and carbon isotope measurements (d ¹⁸O, d ¹³C), and petrologic analyses carried out on one 33-cm-high stalagmite were reported to provide a palaeoclimatic and palaeoecologic record for the period 50-20 ka (HOLMGREN, 1995; HOLMGREN et al., 1995) (figure 2). The d ¹⁸O record was interpreted as reflecting relative temperature variations, and comparing it with data from other palaeoenvironmenal sites in central southern Africa, a regional decrease in temperatures from 50 ka, to a minimum during the Last Glacial Maximum, was confirmed. The d ¹³C record in speleothems is governed by several factors, among which is the relative influence of C₃ arid C₄ vegetation in the region (SCHWARCZ, 1986). This distribution is climatically controlled. As a working hypothesis we propose that the Lobatse d ¹³C record reflects regional shifts in vegetation characteristics, supporting a model that involves an enhancement of the westerlies and an equatorward displacement of the subtropical convergence, which leads to increased influence of the winter rainfall belt in the Southern Kalahari, and to drier conditions in the Middle and Northern Kalahari (COCKROFT et al., 1987). Subsequent dating and stable isotope analysis of cores from a 1.5m-high stalagmite, LII2, confirm parts of previous interprtations and contradicts others, emphasizing the importance of cross-checking. Further Work on material from Lobatse to test the validity of the data is in progress.

Spelaeothem samples have recently been collected from Ficus Cave arid Cold Air Cave in Transvaal in South Africa. These caves lie in the same climate region and the same bedrock formation as Lobatse Caves. Hence, the results from studies of spelaeothems from these caves are expected to test the robustness and regional validity of the Lobatse data. The first uranium-series datings of the South African speleothems have yielded ages from Late Holocene to the present, from the end of the last glacial and from isotope stage 5a (figure 2). Thus, material is available both for comparison with the Lobatse record and for extending the record in time. The young Holocene-to-recent stalagmite has grown at a rate of 15 mm per 100 years arid will be tested against the meteorological record.

Speleothemss are one of few potential archives for preserving palaeoenvironmental data in semi-arid regions, their validity is not yet fully explored. However, no archives are perfect and several problems are connected with semiarid and tropical speleothems. They often possess parts that have suffered from resolution, this can cause open system conditions and invalidate both the age dating and the stable isotope analyses. They can be composed of aragonite or calcite, or be a mixture of both. It is important to decide whether the structure, where calcite or aragonite, is of primary or secondary origin before any firm conclusions can be drawn regarding past environmental conditions.

These problems are best elucidated and overcome by appropriate selection of samples, by analyzing several speleothems from the same locality and by performing a variety of analyses upon them. With this approach there may be a conflict between scientific interests and conservative interests. However, thanks to development of new techniques, smaller samples are needed.

Uranium-series dating of speleothems from Tanzania, Botswana and South Africa has yielded ages from 90 ka to the present. Most of the speleothems analysed are composed of compact, laminated, pure calcium carbonate. The prospect of obtaining high-resolution palaeoclimatic records from the ongoing analyses is most encouraging.

Andrej Mihevc*, Stein-Erik Lauritzen**

*Inštitute za raziskovanje krasa ZRC SAZU, Titov trg 2, 66230 Postojna, Slovenia **Institute of Geology, University of Bergen, Allegaten 41, N-5007 Bergen,Norway

(Paper presented at 12^th ICS,Switzerland,1997)

Abstract Various speleothems, including tall stalagmites and massive flowstone deposits from the Kras plateau, Western Slovenia have been sampled and dated with the ²³⁰Th/²³⁴U method by a -particle counting. Results of datations on speleothems in two caves, Divaška jama cave and Jazbina v Rovnah are presented. Sampling focused on the morphological evidence of the late Pleistocene and morphological impact of colder climate in caves.

In Divaška jama three periods of growth of flowstone were recorded. The oldest flowstone was deposited between two phases of sedimentation of flood loam, before 350 Ka, and are out of range of dating method. Intensive growth of thick large stalagmites is recorded at 240 -170 Ka. The newest deposition started on gravel after last glaciation, around 16 Ka B.P.

In Jazbina v Rovnah cave some flowstone crusts were deposited at about 240 Ka B.P. Flowstone was later fractured, scattered and arranged in a circle due to cryoturbation. On that stalagmite started to grow at 43 ±2 Ka. Growth was continuous till present, showing change of flowstone colour at 14.9 ±0.9 Ka. Another stalagmite analysed, started to grow at 16 ±0.6 Ka, having hiatuses in the bottom part, most distinct at 10.6±0.2-9.3±0.2 Ka, probably showing climatic changes of younger Dryas.

Dating allows us to connect some morphological evidence from caves with climatic changes of the late Pleistocene.

William D. MacDonald²and Charles J. Yonge^1,2

1EIUDP, Universitas Sam Ratulangi , PO Box 1357, Manado 95013, Sulawesi Utara, Indonesia. E-mail Yonge@manado.wasantara.net.id .
²Alberta Karst Institute, Box 8213, Canmore, Alberta, Canada.

(Paper presented at 12^th ICS Switzerland,1997)

Cave Ice Isotopic Composition Which Does Not Fall on the Global Meteoric Water Line

Stable isotope measurements, d ¹⁸O and d D, were made on perennial ice from several ice caves in western North America. Three striking features of the data emerge, especially for cave sites west of the Great Divide: firstly, all d ¹⁸O and d D points lie on or to the right of the Global Meteoric Water Line (GMWL), secondly, those points lie on positive slopes less than the GMWL, and thirdly, the ice is more enriched in tile "heavy" isotopes than expected. Caves from east of the Great Divide are compared to those on the coastal, eastern side, the latter exhibiting trends more concordant with the GMWL. The cave ice trend away from the GMWL is in general made up of three components. The nearest (and most depleted in ¹⁸O and D) points to the GMWL are from seepage or ground water, followed by those from massive floor ice. Finally, hoar ice formed on the walls and ceiling of the cave are most enriched in ¹⁸O and D , and lie the furthest off the GMWL. The western Caves exhibit isotope enrichments more than that expected for vapour-water-ice fractionation (an enrichment in the ice of around 3‰ in d ¹⁸O and around 14‰ in d D ) over a water condensate), whereas our hoar frost samples enrichments of up to 8‰ in d ¹⁸O and 60‰ in d D are observed. Therefore, while vapour-ice fractionation must occur, there needs to be all additional enriched moisture source. Water condensed from the cave atmosphere is generally a mean annual average of precipitation falling at the site, and is positively correlated with the site's mean annual temperature, and we might expect the cave ice to be derived from this source except that it is too depleted. Hoar ice therefore appears to be growing mostly from warm moist air entering the cave during the summer months. This warm air being enriched in ¹⁸O and D over that of the mean annual average, and being able to carry more moisture than cooler air, supplies the additional moisture source required.

Moist, warm air enriched in ¹⁸O and D enters the cave sublimating as hoar frost on the walls and ceiling. Over the warm part of the year the hoar builds up to such an extent that it becomes overloaded and falls to the floor. Along with seepage, melting and recrystallization, it reforms as massive floor ice. As a result, this massive ice has a mixed composition between mean annual seepage and warm-season derived hoar ice. For example we detemine that 10ºC vapour entering the cave and sublimating as hoar ice would yield a mixing line slope less than the GMWL (around 7.1 ), with enrichments of about 8‰ in d ¹⁸O and 55‰ in d D, which fits our field data very well.

How might these results be applied to paleoclimatology? Serendipity Cave in the Crows Nest Pass contains a 15-meter thick plug of massive ice, where guano trapped at a depth of 14.3 meters yields a carbon- 14 date of 970 years BP. This site clearly offers the opportunity for an extremely detailed study of the last 1000 years or so, assuming that the isotopic composition of the ice can be interpreted climatologically. While the scenarios can be complex, we find that the caves encountering lower mean annual site temperatures contain massive ice that is more enriched in ¹⁸O and D compared to caves sited in a higher temperature regime. It therefore seems to be the extent of hoar ice from warm-season air masses that dominates the changes observed in the composition of massive cave ice. Swings in isotope composition in cave ice sections, interpreted as warming/cooling trends, would then be in the opposite directional to conventional glacial polar or cap cores.

Analysis of ice from Caves west of the Great Divide yields a very different picture, data generally fall close to the GMWL, and unusual isotopic enrichments are not observed, actually the reverse. Here, winter snow falling into the cave entrances is preserved in ponded cold air. Recrystallization enhanced by summer rain results in massive ice, but not a complete homogenization of the layers. This means that a muted record of precipitation is recorded and this offers some potential as a proxy paleoclimate indicator. We would interpret swings in isotopic composition in the same way as for polar and glacial ice cores, and thus opposite to records from the eastern divide caves. That is to say that isotope depletion would be associated with a lowering of temperature and vice versa (although we recognize further complexity when considering glacial/interglacial climate transitions). In essence though it appears that we studying two trends; ice derived directly from a vapour source (east divide regime) and ice derived from a water/ice source(west divide regime). Our investigation of perennial cave as proxy climate indicators continues.

Barredo, S.P. ¹ , A. Balbi ²y G.Ré³

1CONICET. Universidad de Buenos Aires. Facultad de Ciencias Exactas y Naturales. Pabellón II. Grupo Espeleológico Argentino (GEA). E-mail: silvia@tango.gl.fcen.uba.ar
²Universidad de Buenos Aires. Facultad de Ciencias Exactas y Naturales. Pabellón II. Grupo Espeleológico Argentino(GEA)
³Laboratorio de Palcomagnetismo "Ing. Valencio". Universidad de Buenos Aires. Facultad de Ciencias Exactas y Naturales. Pabellón II. E-mail: indio@tango.gl.fcen.uba.ar

(Paper presented at 12^th ICS, Switzerland, 1997)

The Cuchillo Cura System has one of the most important karst records of Argentina. In spite of this, the system is not large and comprises an area of approximately 2 km ². Composed of four caves, the system has exquisite speleothems and important fauna which make of this karst a valuable mean of understanding the related processes.

The origin of the Cuchillo-Cura caves is likely to be due to solution processes that occurred during Pleistocene times under a favourable climate. This climate was drastically changed to arid conditions during the last phases of the Andean Orogeny resulting in the present day inactive karst system. Regarding rock deformation induced by tertiary tectonism, the present upward growing corresponds to the collapse of blocks along the joints. Periodically these caves arc flooded by waters from nearby lakes, but this had no influence on the karst evolution.

The paleomagnetic method has been widely applied in Southamerica for geological studies but rarely for speleological matters, specially in Argentina. Difficulties exist with samples extraction, which modifies essentially the natural features of the fragile environment in in such little caves.

A first study has been conducted in this country with the purpose of analysing the method efficiency for such rocks, the probability of an accurate dating. Furthermore, we wanted to know the needed sampling density and the resulting consequences for the caves.

The cave system under study is located along the western side of the Neuquen basin in the Cuchillio Cura range, 10 km south from Las Lajas town in the Neuquen Province, Argentina (see Fig. 1). The geographical coordinates are: latitude 38° 37' 25" south, and longitude 70° 23'03" west. Situated at 900 m above sea level, the climate is semiarid with a mean annual rainfall between 200 and 250 mm per year, and temperatures that vary from 20° to 6° Celsius for the warmest and coldest months respectively. The cave air temperature is below 10° C, the humidity nearly 100 % and slight wind currents have been detected for some of them.

The geology of this region has been described by GROFBER (1946) and many others. Mainly Jurassic and Cretaceous rocks form part of the sedimentary infill of the basin which covers almost the whole province.

Alternating marine and continental deposits characterise the depositional story of the basin, developed over a Paleozoic to triassic heterogeneous basement; only during tertiary times there were volcanic and pyroclastic events of basaltic and andesitic nature.

Light-coloured limestones, sandstones and gypsum are Los Tábanos, Lotena, La Manga and Auquilco formations, which are surrounded by a basaltic plateau. These units comprise oolitic fossiliferous and medium grained limestones, greenish-grey fine grained sandstones with carbonate cement and important fossils, and finally an evaporate sequence (gypsum) interfingered by thin bedded stromatolitic limestones, All of them range in age from middle Callovian to Kimmeridgian. The structure consists of a huge east-northeast dipping anticline cut by northeast trending faults and a number of minor high angled joints with northeast - southwest and southeast - northwest strike.

The La Manga carbonatic facies (mainly packstones) is the primary cave-bearing unit which holds four caves known as El Gendarme, El Templo, El Arenal and Los Cabritos. They are situated in the northern flank of an anticline dipping 15° .

This work was focused on the evaluation of the Paleomagnetic Method applied to the speleological analysis of the Cuchillo Cura System. This system is composed of a number of different caves partially disconnected and fully developed in the limestones of the Lower Oxfordian (La Manga Formation).

Although the isolated primary components are widely dispersed, the preliminary results suggest a normal polarity for those sites situated in the El Arenal Cave. It was also possible to establish the main responsible for this magnetization: Magnetite; and to determine that the remanent magnetization of cave sediments might include secondary components of a possibly viscous origin.

This method could be essential to understand the origin and the evolution of these caves but it can not be considered as the most appropriate for those partially evolved karsts. The devices used to pick the samples produce irreparable holes and the fact that the stalactites or stalagmites are removed makes this method one of the least advisable in terms of cave protection.

Joyce Lundberg
Dept. of Geography, Carleton University, Ottawa,
Ontario, K IS 5B6, Canada

(paper presented at 12^th IGS, Switzerland, 1997)

TIMS U-Th dating of a Bahamian stalagmite from -45 m ASL yielded a (preliminary) date for sea level rise at the end of the Penultimate Glaciation, past -45 m, at » 131 ka. Sea level rise from 45 in ASL to tile Last interglacial high stand of +6 m ASL (» 130 ka) must have occurred in a short time frame of perhaps 1 kyr. The stable isotope record from this sample, of higher resolution than most other proxy records, reveals high frequency spikes and troughs that mimic the GRIP ice core record. Two preliminary dates place the isotopic curve almost exactly coincidental with GRIP, and almost mid-way between the orbitally-tuned marine foraminiferal record and DH11, the Devil's Hole calcite vein from Nevada.

A stalagmite recovered from 45 m below present sea level in a blue hole on Andros Island, Bahamas, has been U-Th dated by thermal ionization mass spectrometry. The cave was air-filled during glacial low sea levels and stalagmites grew. Calcite deposition was stopped by rapid sea level rise at the end of the penultimate glacial (Termination I I). The sample was then subject to dissolution giving it a broadly cuspate and asymmetric form (Figure 1). After this some boring of the outer surfaces by marine sponges occurred and the sample was encased in 1-5 cm of marine packstone.

A similar sample was dated by alpha counting in the 1970s (GASCOYNE et al 1979). Those dates were unreliable for two reasons: the first is that the large samples size required for alpha Counting reduces the resolution attainable; the second is that the sponge borings on the outer layers introduced a younger aragonitic component. Only the dates on the inner material can be considered reliable (within the large error inherent to the alpha Counting system) and these dates are not of great interest.

For this study the stalagmite (78032) was sampled from close to the core out to the youngest layer still present at the tip of a dissolution cusp. Samples for oxygen and carbon stable isotope analysis were taken approximately every millimeter using a binocular microscope to choose sites free from sponge borings. Uranium content is » 0.2 ppm; for each date 0.2 g material was obtained (enough for 1-2 repeats) by drilling a slice parallel to the growth layers, then splitting this into 1 mm thin cleavage fragments. Each fragment was examined under binocular microscope and any boring sponge holes were drilled out. Most of the holes were empty but some had a reddish filling of younger material. Any slice with more than about 20% borings was rejected. Chemical purification of the dissolved samples was by ferric hydroxide co-precipitation of uranium and thorium and separation on anion exchange columns. Oxygen and carbon isotopic ratios were measured on a VG Isogas SIRA mass spectrometer in McMaster University, Canada. Uranium and thorium isotopic ratios were measured on a Finnigan-Mat 262 thermal ionization mass spectrometer in The University of Bergen, Norway. Dates were calculated using ²³⁰ Th/ ²³⁸ U and ²³⁰ Th/ ²³⁴ U ratios.

Date of sea level rise:

Preliminary test dates gave 142 ±7 ka close to the core and 132 ±3.6 ka on an outer layer; these results place the Termination II sea level rise past -45 m ASL a thousand years or so after 132 ka. TIMS dates on Atlantic coral reefs from isotope substage 5e extend back to » 130-131 ka (eg, BARD et al. 1990; CHEN et al. 1991; GALLUP et al. 1994). Thus there is a narrow window between » 131 ka and » 130-131 ka when sea level rose rapidly from -45 m to +6 m ASL.

Stable Isotope Profile:

The high resolution stable isotope record (» 6 points per kyr) is shown in Figure 2, the heavy line for the two-point running mean and the lighter line for the raw data. The resolution of this record for this time period is rivalled by few other proxy paleoclimatic records. Moreover, this significant interval, from the full glacial (isotope stage 6) through the dramatic climatic changes of Termination II, is covered by few other terrestrial records with such a fine time and signal resolution. Several features are apparent:

(i) as might be expected for this oceanic site, d ¹⁸O is dominated by the meteoric signal, with a positive relationship of d ¹⁸O and temperature;

(ii)there is a double warming reminiscent of the end of the last glaciation (the Younger Dryas stage of termination I): this two step deglaciation has also been documented but with much lower precision from pollen diagrams, speleothem, ocean foraminifera, and pedostratigraphy by SEIDENKRANTZ et al (1996);

(iii) in spite of being in a tropical oceanic environment where climatic changes might be well moderated, the record is spikey, showing rapid oscillations of climate rather than a steady progression;

(iv) the range in d ¹⁸O is high at » 3‰: if the range in the ocean foraminiferal d ¹⁸O is» 2‰ . then the extra » 1‰ can be assumed to represent temperature dependent fractionation in the cave.

In Figure 2,78032 is first compared with FM2 (LAURITZEN 1995), another high resolution speleothem d ¹⁸O record from this time period, but from Northern Norway, at the extreme north of the Atlantic and in an area close to the Fennoscandian ice sheet and sensitive to climatic changes. 78032 actually shows a greater overall range than does FM2. Note that the isotopic signal of FM2 during the period of very rapid growth at the end of Termination II is inhibited because the sea water d ¹⁸O is changing rapidly (see LAURITZEN 1995). The peaks and troughs for the two records call be tentatively correlated but 78032 seems to pre-date FM2 by » 3kyr. However, this discrepancy is probably not real: the seven alpha counted dates on FM2 have 2s errors of 8-17 kyr. In addition, the dates on FM2 can be shifted back by at least 2000 years by correlating the very cold spike at » 114 ka in FM2 with the spike in the GRIP ice core record (DANSGAARD et al 1993, GRIP MEMBERS 1993) at» l16 ka.

In the lower graph of Figure 2, 78032 is positioned in relation to sea level and age. It is superimposed on the GRIP record and DH11, the calcite vein from Nevada (WINOGRAD et al 1992). THe coincidence of 78032 and GRIP is remarkable, as is the extent of the offset of DH11.

The isotopic record of 78032 during Termination II shows rapid oscillations similar to those shown by the GRIP ice core. Overall there is a marked two-step pattern similar to that of Termination I. In terms of timing, 78032 correlates well with the GRIP record and with FM2, post-dates DH 11 by » 5 kyr, and predates the orbitally tuned SPECMAP marine record (and thus the insolation record) by » 4 kyr (MARTINSON et at 1987). Until further dating confirms tile position of 78032), more detailed discussion is premature.

Andrea Borsato
Museo Tridentino di Scienze Naturali, via Calepina 14, I-38100 Trento. Italy

(Paper presented at 12^th ICS, Switzerland,1997)

Abstract Dripwater monitoring at Grotta di Fniesto (NE-Italy) reveals the different hydrological functioning of slow- and fast-dripping stalactites. The conductivity record of fast dripping stalactites is punctuated by several abrupt negative peaks following strong rainfall episodes with a response time of a few hours. During these "infiltration event" the discharge increases immediately, and the dripwater becomes strongly undersaturated. The recovery of the previous condition needs from 10 days up to two months which indicates the progressive increase of the residence time of the water in the aquifer. In low-dripping soda straws no infiltration events are recorded, and the water is near the saturation state throughout the year. Nevertheless, the conductivity trend is similar to that of fast-dripping stalactites, which indicates a direct connection of the slow- and fast-dripping stalactite plumbing systems.

Grotta di Ernesto opens at an elevation of 1167 m a.s.l. (long. 11º39'28", lat. 45º58'37" ) in Valsubgana valley ( Trentino, NE-Italy ) ( DALMERI, 1985) and developed in partially dolomitised oolitic limestone of the Calcari Grigi Formation(Lias),which exhibit a good porosity due to the volume reduction during dolomitisation process.

The cave consists of a single downdipping gallery, 2 to 5 m wide, up to 4 m high and 72 m long, which developed along a NW trending subvertical fault, 10 to 25 m below the surface.The original syngenetic morphology of the cave is now blurred by collapse phenomena, and the whole floor is covered by angular blocks, partially coated by flowstones (BORSATO, 1995).

The cave contains a lot of active and fossil speleothems, the age of which ranges from 150,000 years to the present (BORSATO, 1995; FRISIA et al., 1997) and preserves important Mesolithic findings (DALMERI, 1985). At present, water enters the cave through several drippings, with low (from 0. 1 to 2 ml/minute) to medium discharges (up to 50 ml/minute) feeding small, muddy or concretional pools. Pools show seasonal water level fluctuations: in summer some are dry, in autumn and spring all pools fill up mid several overflow.

The cave lies below the timberline, within the deciduous forest vegetation belt consisting of Fagus,Carpinus,Larix decidua and Abies alba. The climate is temperate humid, with a mean annual precipitation between 1000 and 1500 mm. From December to March most of the precipitation is snowy, and the snow cover remains until March. Between March and the first half of April, snowmelting occurs.

The cave entrance was discovered and enlarged in 1983, and subsequently closd by an iron door which inhibits air flow in the cave. All this characteristics, make the cave particularly interesting for paleoclimatic and hydrogeological studies.

The conductivity of dripwater is a direct function of the total mineralisation, i.e. of the Ion Activity Product (IAP) which, in pure limestone aquifers, represents fairly well the total CaCO₃ in solution (BAKALOWICZ, 1974).

In Grotta di ErnestoG1 and G2 drippings, the IAP is mostly identified by Ca²⁺(2.5 to 2.95 meq/1) and HCO₃^- (2.8 to 3.1meq/1), the sum of which commonly exceeds 90% of the total ion concentration (BORSATO, 1995). This means that strong fluctuations in conductivity can only be explained by differences in the total dissolved CaCO₃. Furthermore, G2 dripping is commonly undersaturated, with a Saturation Index for calcite from -0.4 to +0.2 (before the drop splashes), which indicates that, during several periods, the water can potentially dissolve more carbonate. This is also recorded by the conductivity record of G1 dripping which is generally over 10 m S/cm higher than that of the fast dripping G2 site (Fig.).

The overall trend of G2 conductivity directly reflects the rainfall pattern, which strongly influence the calcite saturation state: during infiltration events the drip-rate increases immediately, and dripwater becomes strongly undersaturated. Water slowly shifts towards the saturation state during dry periods, when conductivity increases towards the asymptote and the drip-rate slows down. The first rapid conductivity increase probably corresponds to the soil and rock drainage after the complete soil imbibition, whereas the slow positive trend towards the asymptote identifies the progressive increase of water residence time in the aquifer.

(paper presented at 12^th ICS, Switzerland, 1997)

Abstract During the last few years, speleothems have appeared to be of increasing interest for high resolution paleo-environmental applications. Among the many parameters which govern the speleothem growth rate, the drip water discharge is of the uppermost importance.

Since May 1996, high temporal resolution flow rate measurements are carried out by the «drop counting » method on five stalactites located in the Choranche caves (Vercors, France). Even if the recording period is still too short for characterising long term evolution, recordings already show interesting periodic variations of the flow rate of 2 sodastraw stalactites. Preliminary Fourier analysis clearly show the existence of at least two major forcings with diurnal and semi-diurnal periods, confirming recordings obtained more than 20 years ago in a small cave of the same area. Possible relations with earth and atmospheric tides are discussed. A preliminary analysis of a spectacular «stalactite swelling » is also presented.

Comments and suggestions to Webmaster,
Copyrights by Karst Dynamics Laboratory and Network Center of Guangxi Normal University,
1995-1999. All rights reserved.