IV.??? Paleoclimatic Reconstruction with Karst Records
(New Zealand Journal of Geology and Geophysics, 1996, Vol. 39, 225-241)?
PAUL W. WILLIAMS
Department of Geography, University of Auckland, Private Bag 92019 Auckland, New Zealand
Abstract?Caves overrun by glaciers are known to accumulate dateable evidence of past glacial and interglacial events. Results are reported from an investigation of Aurora Cave on the slopes above Lake Te Anau in Fiordland. The cave commenced to form before c.230 ka B.P. Sequences of glacifluvial sediments interbedded with speleothems are evidence of the number and timing of glacial advances and the status of intervals between them. Twenty-six uranium seris dates on speleothems underpin a chronology of seven glacial advances in the last 230 ka, with the peak of the late Otira glaciation. Aurora 3 advances, at c. 19 ka B.P. With five?advances in?the Otiran, the last glaciation is more complex than previously recongnised. Comparison of the record with that recorded offshore from DSDP Site 594 reveals little matching, but the correspondence of the Aurora sequence with that interpreted from other onshore deposits is more convincing. Glacial deposits on slopes above the cave for a further 660 m may be evidence of the "missing" glacial events of the mid-early Pleistocene.
(Paper Presented at Nouakchott Meeting, Jan.,1997)
Mohamed Salem Ould Sabar (Mauritania), Patrick Michel (France)
The karstic fissures of the Doukkala I and Doukkala II quarries, near Rabat, respectively on the right and the left banks of the Yquem Oued,are filled with thin rubefied deposits.These levels contain an abundant and diverse fauna of vertebrates. The Doukkala II deposits are distributed into three stratigraphic levels (" upper level"," middle level" and "lower level"); for Doukkala I only one level(probably same age are the " upper level " of Doukkala II quarry) contains vertebrate remains.
The Anuran faunas consist of only three families which are the BUFONIDAE in Doukkala I and Doukkala II (Bufo mauritanicus and Bufo bufo spinosus), DISCOGLOSSIDAE (Discoglossus pictus cf. scovazzi ) and RANIDAE ( Rana Ridibunda ) in Doukkala II. These Anuran Amphibian faunas could be contemporaneous with more humid and perhaps warmer climatic conditions than those of the present day.
The " lower level "of Doukkala II with Rana ridibunda belongs either to the middle pleistocene period ( Amirian ) or to the beginning of the Upper pleistocene period (Tensiftian ). The landscape was an open steppe with graminacea; a "ripisylve " bordered the Yquem Oued close to the sites.
The "middle level"of Doukkala II with Rana ridibunda, Bufo bufo spinosus and especially Discogiossus pictus cf. scovazzi could be attributed either to the Tensiftian period or to the beginning of the Soltanian period (Upper Pleistocene). There the climate seems to have been more humid with green meadowland and trees. This is evidenced by the dicovery of Dicerorhinus hemitoechus and sylvicol species in this level.
The "upper level" of Doukkala II and the level with vertebrate remains of Doukkala I dated as recent Soltanian period contain only the genus Bufo (mainly Bufo mauritanicus). These two levels seem to have been dryer with Gazelles and Antelopes significantly dominant.
(Paper Presented at Nouakchott Meeting, Jan.,1997)
Klaus, Heine, Department of Geography, University, 93053 Regensburg, Germany
The remarkable phenomenon of the occurrence of coastal deserts along the western shore of most continents has attracted the attention of scientists for a long time. These deserts are of different nature, some are hyper-arid while others are semiarid. The age and origin of these deserts and their Quaternary climatic changes have in recent years been the focus of much research of an interdisciplinary nature. The Narnib Desert in southwest Africa with a length of about 1,400 km and a width varying between 40 and 120 km, is characterized by a climate that has been more or less arid for approximately 40 million years. Marine pollen assemblages from cores off the coast document arid climatic conditions along the coast ever since Pliocene times. The question whether more humid phases occurred in the Namib Desert and surrounding areas during the late Quaternary has been discussed recently by different authors (HEINE 1992, RUST 1989, ,RELLER et al. 1990, BROOK et al. 1996). Paleoecological interpretation of sediments and landforms together with numerous?14?C dates from calcretes, fossil soil horizons, speleothem etc. show that climatic variations did not occur in the central Namib Desert during the late Quaternary. From all available data it can be inferred that during the last glacial maximum at least this part of the Namib Desert was not affected by more than normal rainfall.
In the hyper-arid central Namib Desert, in the area of the Rossin Mountains and the Tinkas Flats, caves have developed in dolomites. These caves show sinter growth. Sinter deposits have been found to be an excellent repository of climatic data for terrestrial envirorunents. In the central Namib Desert the very presence of relict speleothems indicate more moisture at the time of formation. Both,?14Cand?230Th/234?U?are used to date the speleothems and, hence, the climatic conditions. The230Th/234?U?dates from the cave sinters and calcretes, as well as TL dates from dune sands and fluvial silts show that the chronostratigraphies base on?14C dates have to be revised. The last significant pre-Holocene pluvial phase with speleothem formation in the central Desert probably occurred during oxygen isotope stage 6 and/or 7.?230Th/234?U?dates from other areas in southern Africa support this view (BROOK et al. 1990). It seems likely that climate in the arid parts of southern Africa experienced a more hmid phase around and after 200-150 ka BP. As the TL chronostratigrapliies of paleodunes of the Etosha area in northern Namibia together with paleosoil evidence do not show any indication of a significantly more humid phase after 140 ka BP, the results from the central Namib Desert prove continued aridity along the Namibian coast.
Little is known about shifts of the Namib Desert margins in the south and in the north during the last 135,000 yr BP. East of the Namib Desert (area of the Great Escarpment), sediment and soil sequences document late Quaternary moisture fluctuations that become more and more accentuated towards northeast and east. More humid phases compared with todays climatic conditions occur >25 to 19 ka BP and about 10 to 8.5 ka BP (according to soil formation), while speleothems from the northern Namibian Highland indicate drier conditions compared with today around 26-24, 13 -11 and 9-7.5 ka BP.
The interpretation of the proxy data of the late Quaternary time interval is difficult and, at least, ambiguous. Nevertheless, the geochronological evidence suggests a prolonged period of aridity during the late Quaternary that was neither influenced by large climate changes nor the fast and abrupt climate variability recorded in Greenland ice cores (BROECKER 1994, ZAHN 1994) or in long terrestrial sediment sequences (HOOGHIENISTRA 1989).
The late Quaternary climatic development of the Namib Desert reflects the situation of the surface oceanography in the South Atlantic Ocean ,which is characterized by a cyclonic gyre circulation including the northwest-directed Benguela Current, the eastward South Equatorial Counter Current, and the Angola Current which flows southward along the Agola Margin.
BROECKER, W.S. 1994, Nature 372: 421-424.
BROOK, G.A. et al. 1990, Palaeogeogr., Palaeoclim., Palaeoecol. 76: 311-329.
BROOK, G.A. et al. 1996, Palacoecology of Africa 24 (in press).
HEINE, K. 1992, Palaeoecology of Africa 23: 149-164.
HOOGHIEMSTRA, H. 1989, Palaeogeogr., Palaeocilm., Palacoecol. 72: 11-26,
RUST, U. 1989, Munchner. geogr. Abh. Reihe B, B27: 1-158.
TELLER, J.T. et al. 1990, Quat. Sci. Rev. 9: 343-364.
ZAHN, R. 1994, Nature. 372: 621-622.
Department of Physical Geography, Stockholm University, 10691, Stockholm, Sweden
High-resolution, precisely dated proxy data, spanning long periods are needed for understanding regional and temporal patterns of climate change and possible forcing factors. Such data increasingly available for the Northern Hemisphere and high latitudes, but there is a need for high-quality records from the Southern Hemisphere to detemine to what extent climatic change: global or local. Continuous terrestrial high-resolution climate records from southern Africa are scarce, mainly because the region has few suitable palaeoenvironmental sites. The potential tropical and sub-tropical cave speleothems is promising, although not yet fully explored (Brook?et al?1990, Railsback?et al.?1994, Holmgren?et al. 1995).
This on-going project aims at reconstructing southern African Holocene and Pleistocene climatic changes through studies of speleothems from caves in Tanzania (5-10°?S), Botswana and South Africa (20-27°?S) (Fig. 1). The sites in Tanzania are largely influenced by the Inter Tropical Convergence, while the sites further south experience the interaction between tropical and temperate circulation systems. Palaeoenvironmental records from these sites are expected to increase the knowledge about past patterns in the atmospheric circulation.
Figure 1.??? Map of southern Africa showing the location of caves mentioned in the text
Research in the Lobatse II Cave, Botswana, using stable isotope analysis, radiocarbon and uranium-series dating, was initiated in 1991. One problem in semi-arid areas is evaporative cave environments, making the speleothems unsuitable for stable isotopic studies. However, the Lobatse II Cave is a closed cave with only a narrow entrance shaft and high relative humidity in the cave interior. Petrologic and stable isotopic investigation of one stalagmite, LII4, suggests that isotopic equilibrium conditions did exist in the past, although the speleothems today suffer from surface corrosion (Holmgren?et al. 1995).
The stable oxygen isotopic content (?d?18Oc.) in speleothems is determined by the cave temperature and by the stable oxygen isotopic content in meteoric water, which in turn is determined by composition of the sea-water source and the geographic position of the cave. Both positive and negative relationships between the?d?18Oc?value and temperature are possible. By analysing recent seepage water and speleothems, the relative temperature variations back in time can be obtained. This was a primary aim of this study, but since no recent formations were found, comparison was sought from other sites. The nearest comparable sites are Wolkberg Cave in Transvaal (Talma?et al.1974) and Cango Cave along the south coast, South Africa (Talma and Vogel 1992). At both localities a negative relationship between the?d?18Oc?value and the temperature was inferred, i.e. the speleothems are enriched in?18O during colder periods and depleted in?18O during warmer periods. It is proposed that this relationship is also valid at Lobatse.
The stable carbon isotopic content in speleothems,?d?13Cc, reflects changes in vegetation, provided that bedrock conditions can be assumed to have been constant for the period of speleothem growth. A?d13Cc?value of around -13 reflects an environment?dominated by C3?vegetation, while?d?13Cc?values of around +1.2 reflect a pure C4?biomass (Talma and Vogel 1992). C4?plants- grow preferentially in low latitude areas where high temperatures and summer rainfall dominate, whilst C3?plants are common at higher latitudes, where night temperatures are low, winter precipitation high and water stress is limited (Vogel?et al.?1978, Hattersley 1983).
Stalagmite LII4 has been U-series dated to 51,000 yr B.P. at the base, and to 21,000 yr B P. at the top. Two long growth hiatuses, probably due to drier conditions, began at about 43,000 and 35,000 yr B.P., the latter period ended prior to 27,000 yr B.P (Fig. 2). The?d?18Oc?record indicates a gradual temperature decrease of about 2°?C from 50,000 yr B.P. towards the Last Glacial Maximum at 21,000 yr B.P. (Holmgren?et al.?1995), which for southern Africa has been estimated to be about 5-6°?C lower than the present (Heaton?et al.?1986). The?d?13Cc?record suggests periods of increased precipitation due to contribution of winter rainfall at 51-47,000, 27,000, 25,000 and 23,500 yr B.P. The calcite between the two hiatuses is recrystallised, thus the stable isotopic record for that period was not used for palaeoclimatic interpretation. Average growth rates calculated between absolute dated sections indicate that decreasing growth rates coincide with decreasing humidity, suggesting that the amount of precipitation is the major limiting factor for speleothem growth in semiarid areas.
Figure 2.??? Growth rate and stable isotopic variations versus age of stalagmite LII4 Lobatse II Cave, Botswna and inferred palaeoclimate changes.
It is problematic to compare the Lobatse record with other palaeoclimatic records in the region because detailed records from this period of time are few and are mostly based upon uncalibrated radiocarbon age estimations of caleretes, speleothems and organic remains. In a preliminary comparison with?14C-dated sites younger than 30,000 yr B.P., the wetter episodes documented in the Lobatse record at 27,000, 25,000 and 23,500 yr B.P., coincide with drier conditions further north. This pattern supports a palae-oclimatic model proposed by Cockroft?et al.?(1987), where an equatorwards migration of the westerlies increases the winter rainfall over the southern summer rainfall regime, but leads to drier conditions north of the boundary. Further strengthening of the westerlies and lowering of temperatures during the LGM led to dry and cold conditions over the whole region around 20,000 yr B.P.
Preliminary radiocarbon dating of a second stalagmite from Lobatse II Cave, indicate that this speleothem extends the climate record from Lobatse to include the period after the LGM. While it has been inferred from studies in north and east Africa that this period was one of globally drier conditions in presently semi-arid regions (Petit-Maire 1994), geomorphological studies from the Middle Kalahari document a wetter episode between 16-13,000?14C yr B.P. (Shaw and Thomas 1996). The14?C age estimations of LII2 suggest that the growth rate decreased rapidly during the LGM but recovered shortly afterwards, hence supporting the idea of moister conditions after the LGM over large regions in semi-arid southern Africa.
Multiple sampling of fossil speleothems in semi-arid regions is necessary for palaeoclimatic purposes. However, the validity of these initial results from the project is encouraging. Of specific interest is the potential of southern African speleothems to provide high-resolution records documenting regional climatic variability, the relative influence of winter and summer rainfall regimes, and thus contributing to palaeoclimatic modelling.
Brook, G.A., Bumey, D.A. and Cowart, J.B. 1990. Virg. Mus. Nat. Hist. Mem. 1: 49-70.
Cockcroft, MJ., Wilkinson, M.J. and Tyson, P.D. 1987. Clim. Change 10: 161-191.
Hattersley, P.W. 1983. Oecologia 57: 113-128.
Heaton, T.H.E., Talma, A.S. and Vogel, J.C. 1986. Quat. Res. 25: 79-88.
Holmgren, K., Lauritzen, S.-E. and Possnert, G. 1994. Quat. Geochron. (QSR) 13: 111-119.
Holmgren, K., Karlen, W and Shaw, P. 1995. Quat. Res. 43: 320-328.
Petit-Maire, N. 1994. NATO ASI Series 1: 3-26.
Railsback. L.B., Brook, G.A., Chen, J., Kalin, R. et al. 1994. J. Sed. Res. A64: 147-155.
Shaw, P and Thomas, D.S.G. 1996. J. Arid Environ. 32: 9-22.
Talma, A.S.. Vogel, J.C. and Partridge, T.C. 1974. S. Afr. J. Sci. 70: 135-140.
Talma, A.S. and Vogel. J.C. 1992. Quat. res. 37: 203-213.
Vogel, J.C., Fuls. A. and Ellis, R.P. 1978. S. Afr. J. Sci. 74: 209-215.
lGeological Survey of Israel, 30 Malkhei Israel St, 95501 Jerusalem.
2Dept. of Environ. Sci, Weizmann Institute of Science, 76100 Rehovot
Determinin- recent continental climatic changes in the Eastern Mediterranean reaion is one of the major challenges that faces scientists concerned with developing models for climatic evolution. The present-day Eastern Mediterranean region is semi-arid, located in a narrow transition zone between humid and arid climates.The area is located within a wider desert belt which extends from India to Morocco, but possesses a milder and less extreme climate because of the moderate influence of the Mediterranean Sea. Like other climatic transition zones, this region is very sensitive to environmental changes. In this study we show that a continuous stable isotopic record obtained from cave deposits (speteothems) dated from 25,000 to 1,000 yr BP. allows a detailed insight into climatic evolution during the Holocene and late Pleistocene time and that the global events of glaciation, deglaciation, warming etc. are evident in the Eastern Mediterranean area. We particularly focus on determing the evolution of rainfall and vegetation pattern, using constraints provided by our understanding of the present-day system. The Soreq Cave is located in the Judea mountains, near Jerusalem, Israel. The present-day conditions are typical to semi-arid climate with most of the rain occurring from December to March with an average of 500 mm rain.
Cave waters in the Soreq Cave are?18O fractionated relative to rain waters (usually 1 to 1.5‰ higher) and variations in?d18O values reflect fluctuations in the annual amount of rainfall ,wet/dry seasonal changes and evaporation processes in the epikarst zone. In a closed cave system, such as the Soreq cave, low magnesium calcite(LMC) is depositing in isotopic equilibrium from the cave waters whose temperatures range from 18 to 22° C and?d18O SMOW) vary from -6.3 to -3.5‰. Thed13C (PDB) values of dissolved CO2?in waters ary rom -15.6 to -5.4‰?and reflect the contribution of the dolomitic host rock, soil-derived CO2?with C3 type vegetation, CO2?degassing and arbonate precipitation. Thus, the?d18O (PDB) values of most of the present-day LMC speleothems are between -6 to -5‰ and Most?d13C (PDB) Values range from -12 to -8‰ (shaded rectangle in Fig.1).
The?d18O -?d13C ranges of fossil speleothems dated from 25,000 to 7,000 yr. B.P. differ from present-day range, and thus could not have deposited under present-day conditions. Only speleothems younger than 7,000 yr BP exhibit isotopic compositions similar to those of present-day speleothems, On the basis of the secular tends in isotopic compositions a division into six separate age-groups?is?possible (Fig. 1).
25,000± 2,500 to 17,000± 1,500 yr BP. The speleothems are markedly?18O -enriched with?
d13C = -4.0 to -2.39‰ and?d13C=-11.0 to -7.5‰.
17,000 to 15,000± 1,500?yr BP. Isotopic compositions show a decrease, with?d18O =-5.0 to 3.6‰ and?d13C= -1?2.5?to -10.0‰.
15,000 to 12,000± 1,500?yr BP with low values of?d18O = -6.1 to -5.1‰ and?d13C=-13.5 to -12.0‰.
12,000 to 10,000± 1,500 yr BP with low values of?d18O?=?-6.5 to -5.8‰, 700, but with?d13C showing an increase relative to the previous group,?d13C=-12 to - 10.5‰.
10,000 to 7,000± 1,500 yr BP. Speleothems are?markedly?13C-enriched and?18O-depleted?
with?d18O = -6.8 to -5.8‰ and?d13C=-9.5 to -4.0‰, thus defining a clear change in climatic conditions.
7,000 to 1000± 600 yr BP with?d18O = -6.0 to-5.0‰ and?d13C=-12.0 to-9.5‰ (i.e., mostly in the present-day range).
The?d18O -?d13C trends of the various age groups merge into one another (Fig. 1) and clearly demonstrate that the Soreq cave speleothems are an excellent tool for the reconstruction of the continental paleoclimate in the Eastern Mediterranean region.
Figure 1.??????d13C vs.?d18?O plot of Soreq cave speleothems
The time period between 25,000 to 17,000 yr. BP is known as the last glacial, and in the Soreq cave is characterized?by constant growth of speleothems, with an average growth rate in the cross-sectional direction of more than 1 mm/100 yr. Growth habits consist of light-colored concentric laminae with large LMC crystals and low detrial components and are indicative of deposition from slow, constantly dripping water.?d18O and?d13C values are relatively high, and they are positively correlated (Fig.1). Modeling of these ranges using the carbonate paleotemperature scale and criteria based on the present-day variations of?d18O of cave waters with rainfall, suggests that deposition temperatures were 12° C-16° C, cave water?d18O (SMOW) values ranged from -4.5 to -3‰ and the mean annual rainfall was 300-450 mm. The?d13C values are consistent with a soil cover containing both C3?and C4?type vegetation.
Speleothems from the three groups representing the time periods between: 17,000-15,000; 15,000-12,000; 12,000 to 10,000 yr BP. have similar petrographic characteristics and a distinct isotopic compositional trends that grade into one another. The average growth rate in the cross-sectional direction is 1mm/100 yr, slightly lower than the growth rate during the end of the last glacial period. During the time period of 17,000-15,000 yr B.P. there is a gradual decrease in both?d18O and?d13C values. Model calculations show that the drop in?d18O is associated with warming to a temperature range of 14.5° C to 18.5° C and precipitation reaching 375-540 mm. The gradual decrease?d13C values indicates a transition towards a C3?type vegetation cover. Short periods with ‘spikes’ of higher?d18O and?d13C values indicate that climatic conditions were not stable. A continued, but more moderate decrease in isotopic compositions occurred between 15,000 to 12,000 yr BP, leading into an interval characterized by relatively constant, but low,?d18O at around -6.1 to -5. 1‰ and?d13C at -13.5 to - 12‰ (Fig. 1). These latter isotopic values indicate that the vegetation was totally of a C3?type. Temperatures appear to have been the same as in the previous period, but conditions were wetter with a mean annual rainfall of 550-725 mm. Between 12,000 to 10,000 yr BP,?d18O values are the among the lowest observed throughout the whole studied period, but?d13C values begin to show increase. This period is viewed as transitional to the following one (10,000 to 7,000 yr) which is characterized by a sharp increase in?d13C. A notable 'spike' in?d18O values, observed at around?11,500 yr BP, could be correlated with the Younger Dryas Event, which represented an abrupt global reversal toward colder glacial conditions.
The period from 10,000 to about 7,000 yr BP is represented by relatively small numbers of speleothems, and their petrography and isotopic compositions are very different from the rest of the Holocene speleothems. They are characterized by thin, irregular, brown-red laminae with the high concentrations of oxides and detritus. The growth rate in the cross-section direction is relatively slow at 0.5 to 0.6 mm per 100 years. The?d18O values of -6.8‰ to -5.8‰ are the lowest observed and are coupled with increases to the highest?d13C values of -9.5‰ to -4‰ (Fig. 1). The?d18O values indicate that the annual average precipitation was almost twice that of the present-day. The?d13C values of 9.5 to -49‰ indicate a lower dissolved CO2component in cave waters, relative to other periods. The combination of the highest?d13C values with the lowest?d18O values, the slow deposition rate, the rarity of speleothems, and the high concentrations of detritus indicate flooding events causing d removal of the soil.
The time period of 7,000 to 1,000 yr BP is the only one that partially overlaps with the present-day isotopic compositions.?d18O values mainly range between -6 to -5‰, with some samples having values almost as high as -4. Thus, the temperatures estimated for this period are the same as for present-day. The?d18O range is equivalent to 350-580 mm rainfall, a range similar to that of preset- day. The?d13C values range mainly between -12 to -10.5‰, which is smaller than the present-day (-11 to -8). The higher?d13C values of the present-day range could be a result of human activity. Isotopic composition variablity in this time period indicates that short climatic variations occurred. For these we have independent evidence from archeological, geomorphological and climatological data.