STUDIES ON OXYGEN ISOTOPE THERMOMETRY OF
CAVE SEDIMENT AND PALEOCLIMATIC RECORD

Qin Jiaming

To forecast future climatic changes has to understand past climatic change history and study the change regularities on various scales. In view of the climatic change on large scale, the climatic changes in China show no difference from the global climatic changes. But local or regional climatic changes (especially on small scale) may not be identical. The difference includes both phases and changing range. Accordingly, to study the global or the national climatic changes must study the regional change trend. There have been a lot of researches on the ice cores, the lacustrine deposit on the Qing-hai-Tibet Plateau, and the loess profiles in the north China, and some important progresses have been made. In the karst areas of the south China, for lack of similar climatic informa-tion sources, it is necessary to extract climatic record from karst formation or speleothem, but there have not been systematic researches so far. Guilin is a famous karst area in the world, the karst speleothem and formation are well developed. Accordingly, a big stalagmite at Panlong Cave in Guilin, with favourable cave environmental conditions, was selected for the research. On the basis of monitoring the pre-sent carbonate deposit, the drip water and the environmental conditions near the sedimenta-tion, and the detailed sedimentological study on its cut sections a synthetic study including stable isotopes, AMS-C¹⁴ dating and U-series dating was applied for extracting the climatic change information.

1 GENERAL SITUATION OF THE ENVIRONMENT OF PANLONG CAVE AND THE SEDIMENTOLOGICAL CHARACTERISTICS OF THE STALAG-MITE

Panlong Cave is located on the western side of Guilin-Yangsuo highway about 37 km to the south of Guilin City. The elevation of the cave's entrance is about 190m in a. s.1.. The country rock is the Upper Devonian limestone. The cave is about 251 m long, 6-12 m high (only 2 m at the northern entrance), and 11-15 m wide (only 5 m at the southern blocked end). The thickest part of the overlying roof is up to 120 m. The cave's floor is flat. Stalagmites, stalactites, columns and so on are widely distributed. No.1 stalagmite(photo size: 40KB,JEPG) selected for the research is located at the part about 191 m away from its entrance, and 122 cm high, with a diameter of 45 cm at its bottom. Based on one-hydrologic year's monitoring, the low-est air temperature is 18.9^oC , the higest 19.9 ~ 20.0^oC , averagely 19. 5^oC. The tempera-ture of the drip water is more stable, the lowest temperature 19.0^oC and the highest 19.8^oC, averagely 19.5^oC, that is close to the ground annual mean air temperature (19.1^oC) over 10 a of Guilin area. The relative humidity is 100%. Accordingly, the conditions above could be favourable for the reconstruction of paleoenvironment.
The inside growth ring of the stalagmite is clear. The rings consist of the calcites with various grain size and tones. According to the tone rhythem (from light to dark) and the sedimentation hiatus characteristics, it can be layered from bottom to top as follows:

1. Yellow-white and fine grained, with cerebrum-like microlaminae and uniform colour and texture. The layer ranges from 103.2 cm to 122 cm from its top. On the top of this layer there is a brown weathering crust that was caused by a long period of the interaction of solid-gas (and biologic) phase under the interruption of drip water. During this period, the car-bonate was leached out and Fe, Mn, Cu, Zn, etc. were relatively enriched.
2. White and fine grained, 5.1 cm thick and with 4 intercalation layers of dark lami-nae. On the layer top , the boundary of the dark lamina is obvious, which shows a long period of sedimentation hiatus.
3. White and a little coarse grained, with wavy laminae. The layer ranges from 75.8 cm to 101. 5 cm from its top.
4. Pink, grey-white hybridized, with 4 secondary tone rhythems, sitting at 36.2 ~ 75.8 cm from its top.
5. Grey-white, a little coarse grained, with not-clear laminae, sitting at 30.6 ~ 36.2 cm from its top.
6. . Pink, with obvious laminae, sitting at 15. 6 ~ 30.6 cm from its top.
7. I. Grey-white, with many dark laminae, sitting at 0 ~ 15.6 cm from its top.

There are two hiatus surfaces with a long period of duration at the lower part of the section (on the top of layer I and II ). There are some hiatus at the middle and the upper part, but the duration is short. Accordingly, these parts may be thought as continuous deposition.

2 CHECK OF THE ISOTOPE DEPOSITION EQUILIBRIUM

When calcium carbonate (calcite and aragonite) precipitates from water, the exchange eq-uation of oxygen isotope is as follows:

1/3CaC¹⁶O₃+H₂¹⁸O <=> 1/3CaC¹⁸O₃+H₂¹⁶O

where the equilibrium constant of the reaction is a_cw=R_c/R_w.. R_c is the ¹⁸O/¹⁶O ratio of carbonate, R_w is the ¹⁸O/¹⁶O ratio Of water. As the isotopes in two phases are in equilibrium, the equilibrium constant only depends on the temperature during precipitation, and isn't related to other environmental factors, Accordingly, the temperature during precipitation may be inferred from the isotopic composition. The precipitation temperature can be determined by O'neil's oxygen isotope tempera-ture equation, that is, 1000lna_cw=2.78x10⁶T^-2-2.89^[1,2].
In order to check if the isotopes of sediment are in equilibrium with those of drip water, two clear laminae, sitting at 41. 5 cm and 76. 9 cm from the top of the stalagmite respectively, were sampled in a certain distance from the section axis to the two sides for carbon and oxygen iso-tope analysis. The results are shown in Fig. 1.

Fig. 1 shows that the change of d¹⁸O is small, the standard deviation of lamina 24-1 is 0.15%o, that of lamina 35-2 is 0.3%o. If the samples could be restricted in a lamina, the variation range may be much smaller. Since the d¹⁸O values in a growth layer are well identical and haven't simple correlation with d¹³C, it can be thought that there wasn't dynamic isotopic fractionation in the process of sedimentation.

Fig.1 Variation of d¹⁸O(d¹³C) in the same growth lamina of No. 1 stalagmite, Panlong Cave, Guilin Solid line: lamina 24-1; Dotted line: lamina 35-2

Based on monitoring at different time and sites in the cave, the average carbon and oxygen isotope values of 9 present carbonate samples are:d¹³C(PDB)= -8.59%o , d¹⁸O(PDB)= -5.89%o, and those of the relevant present drip water are: d¹⁸O(SMOW)=-5.66%o. Based on O'neil equation and replaced by d¹⁸O values of the drip water and the carbonate, the calculated sedimentation temperature is 19.8^oC , that is completely identical with measured mean temperature. On the basis of the environmental conditions of the cave, and the equilibrium check above, No.1 stalagmite is thought to be formed in isotopic equilibrium with the drip water.

3 RADIOACTIVE DATING AND GROWTH RATE OF THE STALAGMITE

Usually, speleothem is pure CaCO₃ precipitate. For several hundred thousand years old samples, the ideal dating methods are ¹⁴C and U-series dating methods. This study mainly applied AMS-C14 dating, and the results were checked by (beta)-counting ¹⁴C dating method and a--counting ²³⁰Th/²³⁴U dating method.

The selection of the dating samples was mainly based on the changing points of sedimen-tation rate, especially, the hiatus surfaces and the characteristic points of climatic changes.

The reliability of the ¹⁴C ages of speleothemes mainly depends on if the carbon isotope in water and in carbonate reaches full exchange during the precipitation of the carbonate. If the full exchange doesn't reach, ¹⁴C specific radioactivity of the sediment is on the low side, and the age is older. If the full exchange reaches, the age is reliable. In order to prove this, a growing sediment near No.1 stalagmite was collected in 1993 for the determination of ¹⁴C specific radioactivity, and the ratio to the recent carbon is 1.22+0.021 that is close to recent at-mospheric ratio. The result shows that the full exchange has reached.
The AMS-¹⁴C ages of the stalagmite are in normal order from top to bottom. Mean-while, the data at the bottom are well identical with the data of U-series dating method. Accordingly, the ages of No.1 stalagmite are believable.

Fig. 2 shows the relationship between the measured ages and the growth height.
The numhers in the figure are the calculated growth rate(mm/100 a).

Fig. 2 shows that the growth rates at differ-ent time are different. According to the division from large sedimentation cycle to small sedimentation cycle, the stalagmite can be divided in five cycles (A ~ E) from bottom to top. Each cycle can be divided into 2~3 stages based on the rates. The large sedimentation rates are in the 3 stages of C1, D1, E1, reaching up 38 ~ 80 mm/ l00 a. The small sedimentation rates range from zero to several mm/100 a. There are two obvious black-brown hiatus at the bottom, that is, stages A2, B2. A2 is between 11080 a B.P. and 32440 a B.P. , and the interrupted time is 21000 a. B2 ranges from 6140 a B.P. to 6930 a B.P. , and the interrupted time is 790 a. The two hiatus resulted evidently from the interrup-tion of drip water under dry and cold climate.
The changes of the local environment of caves, e.g. earthquake, roof collapse, block-ing-up of drip water passages, displacement of drip water and so on, can also result in the interruption of drip water to form hiatus. In addition, growth rate can reflect regional envi-ronmental changes. For example, a larger sedimentation rate may reflect a relatively warm, humid climate and abundant rainfall, and a smaller sedimentation rate may reflect a relatively dry and cold climate. The changes in the growth rates of No.1 stalagmite are basically iden-tical with the climatic characteristics reflected by oxygen and carbon isotope, which shows that the growth of the stalagmite is mainly controlled by regional environmental factors.

4 RECORD OF STABLE ISOTOPES

During stable isotopic analysis, the carbonate samples react with 100% phosphoric acid to produce CO₂ that is purifed to be determined by mass-spectrograph (MM-903E of England VG Company). The relative standard of d¹⁸O and d¹³C is PDB standard. The systematic er-ror is less than 0.1%o. The oxygen isotope of the water samples is processed by H₂O-CO₂ equilibrium method, and the hydrogen isotope is got by zinc-reduction method. The decrepi-tation and zinc-reduction method is used for the process of inclusion hydrogen. The relative standards of dD, d¹⁸O are SMOW standards. The systematic error of dD is less than 1%o, that of d¹⁸O is less than 0.1%o,and that of dD of inclusion water is less than 2%o ~ 5%o.
The samples of carbon and oxygen isotopes were taken in a certain distance along the axis of the cut section of the stalagmite. The results show that the d¹⁸O during Holocene E-poch ranged from - 5.5 to - 6.8%o , the relevant temperature ranged from 5 ~ 6 ^oC. Many low peak areas (with high values) show the existence of some cold periods in the climatic changes.
Most of sedimentation hiatus are related to glacial periods. The high d¹⁸O value from 32000 a B.P. to 36000 a B. P. and the suddenly high d¹⁸O value round 11000 a B. P. show the low temperature at those periods that are analogous to the Ashihe glacial stage and the Fuping glacial stage of the north China respectively^[3].
During Holocene Epoch, except for the high d¹³C values in several cold periods of the Middle and the Late Holocene Epoch, the d¹³C values in other periods were generally low. The d¹³C generally ranged from -8.0 to -12.0%o , which shows that the general climatic characteristics during Holocene Epoch were warm, and humid, the ground vegetation was luxuriant, and the carbon isotope composition of the stalagmite resulted from the mixture of the isotopes of vegetation, atmosphere and source rock. During the last glacial period, the global climate was dry and cold, the d¹³C of the carbonate was generally between -6.0%oand-8.0%o or more, which shows that vegetation was difficult to grow under dry and cold climate, and the carbon isotopic composition of the stalagmite was formed by the carbon iso-tope exchange between atmospheric CO₂ and source rock.

5 THE CHANGING PATTERN OF THE CLIMATE SINCE 36000 YEARS B.P. IN GUILIN AREA

Because speleothemes are usually composed of pure calcite and there is not hydrogen-bearing mineral in them, the hydrogen exchange between water and rock can't take place and the original hydrogen composition can be kept. Accordingly, we directly measured the hy-drogen isotopic composition of the inclusion water, and calculated the d¹⁸O value of the original drip water by applying the meteoric water isotope equation (dD = 8.39 d¹⁸O + 16.04) of the local area. The results are as follows:
During the normal precipitation period of Holocene Epoch, the average dD was -31.4%o, accordingly the calculated d18O was - 5.65%o, that is equal to the measured average d¹⁸O value ( - 5.66%o) of the present drip water near No.1 stalagmite.
During the last glacial period (32000-36000 a B.P. ), dD was -42.4%o and the calcu-lated
d¹⁸O was - 6.97%o,that is 1.31%o lower than the d¹⁸O(-5.66%o) of the normal pre-cipitation periods, which shows that the sea level greatly descened during that period and Guilin area was much farther from the coast-line at that time than it is today. Because of the influence of continental-oceanic effect, the precipitation enriched light isotopes.
During the periods disturbed by storm effect, the sedimentation rates are larger and the d¹⁸O of the carbonate usually is less than -7.0%o that have to be corrected by synchronous-ly sampling analysis.
On the basis of d¹⁸O of the carbonate and d¹⁸O of the medium water, as well as O'neil e-quation, the temperature during the formation of the sediment can be calculated, and the yearly d¹⁸O changing curve can be transformed into the yearly temperature changing curve.
Fig. 3 shows the climatic record extracted from No.1 stalagmite (photo size: 40KB,JEPG) . The temperature on the ordinate has been corrected by the present mean temperature (19.1^oC) of Guilin. The climatic changes since 36000 a B.P. can be concluded as follows:

Fig.3 Curve showing air temperature variation since 36000 years B. P. in Guilin area M.D.L: The middle Dali sub-glacial period; L.D.L - The late Dali sub-glacial period; Y.D.: Younger dryas; I. II. III: 3 climatic cycles during Holocene Epoch

5.1 The Last Glacial Period

1. Because of the influence of the global dry and cold climate, the karst processes in this karst area were greatly weakened and the growth of stalagmite was slow. During some stages, the drip water might stop and the deep brown weathering front was formed.
   
2. During 36000 a B.P. to 32000 a B.P. that corresponds to the coldest period of the middle Dali sub-glacial period, the lowest ground annual mean temperature was 8 ~ 9^oC, the highest 13^oC .
3. During 32000 a B.P. to 12000 a B.P. that corresponds to the late Dali sub-glacial period, the sedimentation was interrupted. On the basis of the oxygen isotope of the fine car-bonate lamina on the hiatus surface, the lowest annual mean air temperature was 6 ~ 8^oC that was the lowest air temperature in this area during the studied time interval.
4. Around 11000 a B.P. , the expression of Younger Dryas was obvious in this area and the lowest annual mean air temperature was up 9^oC . After that period, the air tempera-ture rapidly went up and was close to present level to 10700 a B.P.. The air temperature was going up 9^oC during 300 a.

5.2 Holocene Epoch

The jump-like rise of the air temperature in Younger Dryas resulted in a distinct leap between the last glacial period and the Holocene Epoch. The period from 9000 a B.P. to now can be divided into 3 climatic cycles from warm to cold, namely, 9000 a B.P. to 6000 a B.P. , 6000 a B.P. to 3000 a B.P. and 3000 a B.P. to now. Each cycle lasts about 3000 a. The highest annual mean air temperature during warm period was up to 22 ~ 23^oC that is about 3 ~ 4^oC higher than that of today. The lowest annual mean air temperature during cold period was 16^oC that is about 3^oC lower than that of today. The distinguishable cold periods can be compared to the phenology records of China. For the modern small glacial pe-riods, the stalagmite records are fully identical with the historic records and the conclusion of the tree ring index of the cypress of Qilian Mountain^[4].
The oxygen and carbon isotopic records of the stalagmite are also fully identical with the climatic information from the growth rates itself, and can be compared to the research re-sults of the loess profiles in north China and the isotopic records of the global ocean.

6 CONCLUSIONS

1. AMS-14C dating method has the distinctive advantage of high resolution and sensi-tivity. Combining with the stable isotopic studies, we can get the records in stalagmites and reveal the changing events of climate. The setting up of the changing pattern since 36000 a B. P. in Guilin area provides a reliable way and method for the studies of the land paleocli-matic changes in south China.
2. The temperature changing records from the oxygen isotope are checked by the infor-mation of the carbon isotope and the growth rates. The study results show that the low d¹⁸O and
   d¹³C and the large growth rates usually reflect warm and humid climatic characteristics. On the contrary, the high d¹⁸O and d¹³C and the small growth rates often reflect cold and dry climatic characteristics.
3. Compared with the similar researches at home and abroad, this study applied a large stalagmite and advanced AMS-¹⁴C dating method, and the sampling interval can reach up millimeter magnitude. Except for the interrupted periods of sedimentation, the resolution of climatic changes has been up 100-200 a, even during cold period, it can be up 500 a. For a time interval of tens of thousand years, other researches can't reach this resolution. A Franch Duplessy et al. (1987)^[5] selected the mono-species foraminiferal fossils from the two, deep - sea borehole cores ( SU81~18 , CH73~139C ) in the North Atlantic Ocean , made AMS-¹⁴C dating and d¹⁸O analysis and achieved the outstanding research results of the oxygen iso-topic records (in the time interval since 20000 a B.P.). But because the selection of the fos-sils was difficult, the sampling interval was large (usually l00 mm) and the resolution of cli-matic changes could only be 500-1000 a.
   There have been achieved many results in the studies of paleoclimatic changes by using stalagmites, e.g. in Cold Water Cave, Iowa of USA^[6]; in Uamhan Tartair Cave , Scotland (cooperation of England and USA)^[7]; in Drotsky Cave of Botswana, Africa (made by Amer-icans)^[8]. These researches have reached up high resolution, but all time intervals are confined to Holocene Epoch because of the small stalagmites (only 16-40 cm high).
4. The karst areas are widely distributed and the cave resources are abundant in the south of China, which provides a superior position for paleo-climatic studies by using large stalagmites. The study in Paniong Cave shows that speleothem is a good carrier of age and paleoctimatic information. Applying the comprehensive method of AMS-¹⁴C dating (TIMS U-series method for the samples older than 40000 a) and stable isotope study, we may get the details of climatic changes with a resolution of 10-100 a, and provide reliable regional data for the studies of the global climatic changes and the forecast of the future climate.

References

1. O'neil, J.R.. Oxygen isotope fractionation in divalent metal carbonates. J. Chem. Phys. , 1969, Vol. 51
2. Zhang Zigang. Stable isotopic geothermometer. In: Application of Stable Isotopes in Geological Science. Shaanxi Science and Technology Publishing House, 1983: 23-26 (in Chinese with English abstract)
3. Sun Jianzhong et al. Paleoclimatic environment of loess plateau during last glacial period. Quaternary of Loess Plateau. Science Publishing House, 1991, 154-185 (in Chinese with English abstract)
4. State Science and Technology Committee of China. Possible climatic changes of future 60 years in China. In: Chinese Science and Technology Blue Book (No. 5, Climate). Scientific and Technological Literature Publishing House, 1990, 128-136 (in Chinese with English abstract)
5. Duplessy, J.C., Bard, E., Arnold, M. , Maurice, P.. AMS-¹⁴C-chronology of the deglacial warming of the North Atlantic Ocean. Ibid, 1987, B29(1,2). 223-227
6. Dorale, J. A. et al. . A high resolution record of Holocene climate change in speleothem calcite from Cold Water Cave, Northest Iowa. Science, Vol. 256, 1626-1630
7. Andy Baker et al.. Annual growth banding in a cave stalagmite. Nature, 364(5): 518-520
8. L. Bruce Railsback et al.. Environment controls on the petrology of a Late Holocene speleothem from Botswana with annual layers of aragonite and calcite. Journal of Sedimentary Research, 1994, Vol. A64, No. 1: 147-155

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