1. Carbon Cycle in Epigenic Karst Systems

ABSTRACT

The balance of the world carbon exchanges shows an 1.4 GtC/yr unknown sink in the continental biosphere. The aim of this article is to estimate the contribution of karstic dissolution processes to this sink. To calculate the karstic dissolution in every part of the world, a new parameter has been created, called «W maximal potential dissolution »E (MPD or DMP in French). It calculates the theoretical dissolution rate in an idealized karstic system reduced to a simple pure carbonated block crossed by a flux of CO₂ enriched water. Correlations between DMP and other karstic dissolution calculations are good. The DMP can be calculated everywhere from a simple couple of mean annual temperature and rainfall values. In this paper, climatic data from 266 meteorological stations all over the world have been treated. They furnished a mean DMP value for each main climatic type. The calculation have been made by 10°E square grids, each grid assigned to a climatic type i. e. to a mean value of DMP. By this way, the total consumed carbon mass all around the world is around 0.3 GtC/yr which represents 21% of the unknown carbon sink. More precise calculations are in progress, based on a thousand of climatical data.

KEYWORDS

karst, climate, karstic dissolution, carbon cycle, carbon sink, limestone, CO₂

INTRODUCTION

Most of the studies concerning global carbon cycle indicate an unknown sink of carbon in the continental biosphere (AcadéWmie des Sciences, 1994). Karstic dissolution processes are implicated in this phenomenon. Karstic features development is under the conjoined influences of petrographic, climatic, tectonic and geomorphologic factors. However, the atmospheric (or pedologic) carbon consumption results from the action of chemical laws governing carbonated rocks dissolution. It can therefore be considered in function of two main factors: petrography which delimits the outcrops of carbonated rocks around the world and climatology which determines the importance of chemical dissolution processes.

It exists many methods to calculate the karstic dissolution rate:

• specific dissolution of Corbel (1957) and derivated formulas (as Williams’, 1963),

• geochemical balance of Bakalowicz (1979),

• DBL theoretical model (Liu et al., 1998),

• solution ratio of limestone tablets (Gams, 1985; Urushibara-Yoshino et al., 1997 ; Yuan, 1997).

The first method needs the knowledge of parameters that are often bad known: catchment area, discharge flow, mineralization of water, number of springs. The two following need an important survey of the karstic system: 365 flow measurements and mineralization analysis for Bakalowicz (1979) in Baget (France), 200 for Salado & Marjolet (1975) at Lez’ spring (France) and 90 for Denizman (1998) in Florida (USA). The last one needs the equipment of a precise karstic area and don’t often furnish the total amount of dissolved limestone (Gams, 1985; Gerome-Kupper, 1984).

The result of these different methods depends also from the internal structure of the karstic system, its age, its hierarchization degree, the presence of a significant saturated zone, etc... So their results cannot be easily compared from a karst to another. In the goal to compare karstic system all over the world, Gombert (1988) created a new parameter called «W maximal potential dissolution»E (MPD or DMP in French). 

This parameter corresponds to the theoretical dissolution rate of a karstic system reduced to its simpler expression (Gombert, 1988, 1994, 1997): a pure limestone block crossed by a flux of water saturated with CO₂. This water dissolves limestone up to saturation according to mean annual atmospheric temperature. Calculation is done at annual step with the following assumptions:

• entering water flux results from efficient rainfall, i.e. annual rainfall minus real evapotranspiration calculated from Coutagne’s formula (in RéWméWniéWras, 1986) ;

• CO₂ acts through the fictive P_CO2 of water supposed to be in equilibrium with the pedologic P_CO2 calculated from Brook’s formula (1983) and with an atmospheric P_CO2 constant of 3.5µ910^-4 atm. ;

• limestone dissolution is done in reference with Tillmans’ curves (Trombe, 1952) at the mean annual temperature of the site.

Mathematical formulation of DMP is then as (Gombert, 1994, 1997) :

DMP=25.2Ie(K₁K₂K₀P_co2 K_s)^1/3

where Ie is the efficient infiltration (mm), p_CO2 the fictive partial pressure of gaseous phase associated with the solution (atm.), K₀, K₁ and K₂ the equilibrium constants of, respectively, CO₂ dissolution in water, hydratation and formation of bicarbonate ions, formation of carbonate ions and K_S the solubility constant of calcite.

DMP furnishes similar values than Corbel’s specific dissolution or geochemical balance of Bakalowicz (Tab. 1).

WORLD REPARTITION OF MAXIMAL POTENTIAL DISSOLUTION

We’ve calculated a mean DMP for each climatical type from 266 meteorological stations all over the world (Miller, 1966). Mean world DMP value, weighted by the area of each climate zone, is estimated to 35.6 mm/kyr (Tab. 2).

DMP appears to be:

• important under very wet climate, as well as hot (tropical, equatorial) or cold,

• average under cold or cold temperate climate,

• weak under warm or warm tempered climate with a dry season (mediterranean or subtropical),

• insignificant to nil under desertic climate.

This behaviour comes from the main sensibility of DMP to the entering flux of water, i. e. the importance of efficient rainfall. This parameter is weighted by the contradictory influences of temperature : it increases the speed of calco-carbonic equilibrium reactions but decreases the maximum CO₂ and CaCO₃ concentrations of saturated water. For example, the DMP can be comparable under cold and wet climate (45 mm/ky in Trondheim, Norway) or tropical one (44 mm/ky in Calcutta, India).

WORLD CARBON CYCLE

For the 1980-1989 period, the balance of the biosphere-atmosphere carbon exchanges is 1.4 ±E  1.6 GtC/year (AcadéWmie des sciences, 1994). This important imprecision results from the substraction of gross carbon fluxes exchanged between main reservoirs: gross fluxes are about 100 GtC/year when net resulting fluxes are only a few GtC/year. Despite this imprecision, it would exist an unknown sink of carbon, probably situated in the continental biosphere. This sink is able to reduce the carbon excess in the biosphere-atmosphere exchanges.

KARST ROLE

The equation of calco-carbonic equilibrium shows that one mole of dissolved CaCO₃ consumes one mole of atmospheric (or pedologic) CO₂ :

Dissolved calcium bicarbonate is first transported by continental waters and represents the main part of their mineralization. Its final target is marine sedimentation after a more or less long way in the continental biosphere. Above the CCD limit depth, calcium bicarbonate remains in solution in sea water and is substracted for a generally long period from the atmospheric carbon cycle.

Karstic dissolution processes appears to play a significant role in the CO₂ world sink. The amount of carbon substracted from its atmospheric cycle by karstic dissolution can be estimated by the knowledge of karstic dissolution rate and by the world repartition of karstic areas. This amount corresponds to a certain volume of dissolved limestone we can calculate with the DMP on the following basis.

Volume V of limestone dissolved by karstic dissolution is equal to the action of karstic dissolution (expressed by DMP) on a superficies S :

With an average specific gravity µ9 = 2500 kg/m³ (Williams,1963), this volume corresponds to a dissolved limestone masse M :

With the molar mass of calcium carbonate (Mm = 0.1 kg), it comes the number N of dissolved CaCO₃ moles:

With one CO₂ mole for each dissolved CaCO₃ mole, this number corresponds also to the consumed carbon moles. With the carbon atomic mass (Ma = 0.012 kg), we obtain the total mass of consumed carbon TMC  :

Location of karstic outcrops come from the map of Ford & Williams (1989). Calculation has been made by 10°E square grids. Each grid is affected to a climatic type (Tab. 3) and consequently to a mean DMP value (from Gombert, 1994). The total karstic area submitted to each climate leads to calculate a world consumed carbon mass of about 0.3 GtC/year (Tab. 2). This value is about the same order as other estimations (from 0.10 to 0.61 GtC/year (Tab. 4). It represents 21% of the unknown carbon sink of the continental biosphere (AcadéWmie des sciences, 1994).

Table 3 Summary of climatical division of world by 10°E square grids

CONCLUSION

The balance of carbon net fluxes exchanged between the main reservoirs (ocean, atmosphere, continents) shows the existence of an unknown carbon sink in the terrestrial biosphere. The aim of this article is to estimate the contribution of the «W world karst »E to this carbon sink.

The estimation of the actual role of karstic processes in global carbon cycle needs the knowledge of dissolution rate in each karstic area of the world. This rate can be calculated with a new method called maximal potential dissolution (DMP). This method studies the case of a simple karstic system where dissolved limestone exportation results from entering water and CO₂ fluxes. DMP can be reduced to the influence of two climatical parameters : mean annual rainfall and atmospheric temperature. Correlation is good between DMP and other methods of karstic dissolution calculation (Corbel’ formula or Bakalowicz‘s geochemical balance). So DMP can be considered as a good representation of real karstic dissolution rate under different climates.

The calculation of mean DMP value has been made for each main climatic zone in the world. Integrating the karstic area submitted to each climate, we can calculat the total mass of carbon yearly consumed by karstic dissolution processes. We obtain a total amount of 0.3 GtC/year which is in the same order as other estimations. This value can explain about 21% of the unknown continental carbon sink.

More precise calculations are in progress. They’ll be realized at a 10°E square grid step. They lead to improve the calculation by taking into account more climatologic data (about a thousand). They can be used later to estimate the «W world karst »E response to anthropogenic climatic perturbations.

REFERENCES

Carbon dioxide sampled from soils in the Marble Mountains subalpine karst of northern California show clear seasonal trends that relate to the timing of snowmelt, soil temperature, and soil moisture fluxes, through their influence on patterns of root and bacterial respiration. Soil gas was sampled using a method developed by Kiefer (1992) and modified for the study area: gas was extracted with syringes from septum-capped 3-mm (I.D.) wells at 20-cm depth, and transported via vacutainers to a Beckman infrared CO₂. Sampling environments included a north-south transect through soils developed in marble and other metamorphic rocks, including north and south aspects at 1700 to 1900m elevation in red-fir forest, dry meadow, wet meadow, and rock fissure environments. Seasonal CO₂ peaks were approximately 1%, although wet meadow and other moist environments produced levels as high as 4%. Links with an ongoing investigation of karst solution processes monitored through hydrochemical analysis of surface and subsurface waters provide insights into the modern development of an erosion system with significant glacial and karst interaction over at least the past 350,000 years.

Carbon dioxide is considered as the main important agent in limestone solution (Ek and Gewelt, 1985). Classical karst results on solution by soil CO₂ dissolved in water, flowing under the control of gravity (Bakalowicz, 1996). Wood (1985) proposed that CO₂ is generated by the oxidation of particulate and/or dissolved organic carbon that is transported from the land surface deep into the undersaturated zone (cave) by infiltrational water. According to Wood's model, the fluctuation of CO₂ concentration in caves are directly controlled by rainfall rates at the recharge zone of karst system.

Previous studies carried out in several caves located in northern Spain (e.g. Hoyos et al., 1998) show the existence of other factors, in addition to rainfall, influencing in the variation of CO₂ contents in water, as well as, in both cave and soil atmosphere. These factors include the atmospheric pressure and the temperature differences between the exterior and the interior of the cave (ventilation of the cave). In order to build up a model for CO₂ fluxes in near-surface karstic caves (Altamira and Tito Bustillo caves), a system for measuring microenvironmental parameters in caves have been designed and set up. Parameters such as air CO₂, air temperature, relative humidity and ²²²Rn concentration, are measured automatically in several points of the cave atmosphere. CO₂ and ²²²Rn contents in soil atmosphere are also automatically measured. Likewise, CO₂ content is also measured in water that is periodically sampled (dripping points and standing water in pools). The first results obtained reveal two types of CO₂ content variation: seasonal variations mainly influenced by the rainfall regime and by the temperature differences between the exterior and the interior of the caves, and short-term variations controlled by the fluctuation of atmospheric pressure. The increase of atmospheric pressure provokes CO₂ diffusion phenomena from the soil cover towards the cave atmosphere. The behavior of ²²²Rn variations also reflects this phenomenon. During extremely dry periods (e.g., January-April 1997) high CO₂ (up to 6000ppm) and ²²²Rn concentration have been measured in the interior of the caves, while the CO₂ and ²²²Rn concentration of soil atmosphere decrease. When the atmospheric pressure decreases a fall in the air CO₂ (less than 1500ppm) and ²²²Rn cave is observed, together with an increase in the values of these parameters in the soil atmosphere. These data seems to indicate that infiltrational waters are not always the main vehicles for transporting CO₂ into the caves.

This work is supported by the EU CONTRACT N.ENV4 CT95 0104, and is part of the IGCP project 379 (Karst Processes and the Carbon Cycle).

References

Bakalowicz, M. (1996). Les processus de karstification et les differents types de karst associes. Mem. Soc. Geol. Fr.,169: 363-371.

Ek, C. and Gewelt, M. ( 1985). Carbon dioxide in cave atmosphere. New results in Belgium and comparison with some other countries. Earth Surface Proc. and Landforms, 10: 173-187.

Hoyos, M.; Soler, V.; Canaveras, J. C.; Sanchez-Moral, S. and Sanz-Rubio, E.(1998). Microclimatic characterization of a karstic cave: human impact on microenvironmental parameters of a prehistoric rock art cave (Candamo Cave, Northern Spain). Environmental Geology, 33: 231-242.

Woods, W. W. (1985). Origin of caves and other solution openings in the unsaturated (vadose) zone of carbonate rocks: A model for CO₂ generation. Geology, 13: 822-824.

A series of forty soil carbon dioxide measurements were taken in hillslope soils at a single 0.5 hectare site in the karst of the Cayo District, Belize immediately prior to and following the burning of the secondary vegetation during March of 1984. All measurements were made with a Draeger Probe at a soil depth of apporximately twenty centimeters.

The secondary growth had been cut, but not burned, prior to the first set of measurements, which indicated a mean soil CO₂ concentration of 1. 6% (n= 10). The plot was burned the following day, on which no measurements were possible.

Measurements the day after the burn produced a drastically-reduced mean soil CO₂ level of 0.1% (n=10), reflecting the virtual cessation of soil biological activity. Measurements on the following two days indicated a gradual recovery of biological activity, with mean soil CO₂ levels of 0.3% (n=10) and 0.9% (n=10) respectively, the latter following a brief rain shower.

These measurements indicate that CO₂ levels in tropical karst areas may vary considerably over short time periods, particularly as a result of human activities, and they suggest that studies of carbonate rock dissolution in tropical karst should take into account the potential for both short-and long-term variations in soil CO₂ levels.

Many an active streams conduits within karst aquifers effectively transport and deposit non-carbonate, clastic sediment. However, little is known about how these sediments impact conduit development and enlargement rates. For example, can dissolution take place at the sediment/bedrock interface beneath a flowing stream? If not, cavern enlargement might be dominated by flood conditions when the bare rock of the walls and ceiling are in contact with the dissolving fluids.

An approach using limestone tablet weight loss experiments, along with water sampling and geochemical modeling, has been undertaken to understand the nature of fluid movement and chemistry within the sediment beneath an active flowing cave stream within the Kentucky's Mammoth Cave System. Fluid flow and carbonate chemistry were compared between the active stream and within the sediment at 16, 31, 61, and 91 cm below the stream bed. We found that carbon dioxide pressures within the interstitial fluids were elevated an order of magnitude above that of the stream waters, having levels as much as 25 times that of atmospheric background, presumably from microbial decomposition of organic material. The fluids were all undersaturated with respect to calcite (SI= -0. 4 to -0. 9), and limestone blocks buried at these levels all dissolved (rates from 0.8 to 22. 7 g m^-2yr^-1). These results suggest that some locations the limestone bedrock may be dissolving beneath clastic sediment deposits, this in turn has important implications for understanding rates and geometries of conduit evolution within karst aquifers.

Most of carbonate dissolution in aquifers and hence karst formation depends on the high CO₂ pressures generated in soils by decay of organic material and by root respiration. Production of CO₂ within the aquifer is also known, however, by oxidation of organic matter contained in the rock or infiltrated from the unsaturated zone. Alternatively, organic matter rich river waters recharging karst aquifers through swallets (sinkholes) may also contribute to rock dissolution and to maintain the drainage system. The Val d'OrléWans hydrogeological unit in France, is a site of special interest for studying such a feature. At that site, the survey of the variation of the chemical composition of both recharge and discharge waters for no-reactive elements (chlorides) during flood periods have lead to recognize an about 3-days interval time for the underground transit of the water in the system. In summer we must consider that all of the recharge is from the sinking River Loire waters. During this period, the calcite dissolving capacity of the river waters is either null (low water levels in eutrophic conditions) or very small (during flood episodes), but carbonate dissolution actually occurs as indicated by the calcium concentration of the spring discharge waters, which also are anoxic and contain much less organic matter than the River Loire waters.

Computation of alkalinity and total carbonates differences between recharge and discharge waters, based on aerobic consumption of organic matter plus denitrification and calcium carbonate dissolution, compares well to the differences actually observed. The conclusion is that at least during the summer period, the dissolution of the carbonate bedrock is due to the oxidation of the organic matter introduced into the karst system together with the sinking stream waters. The dissolved carbonate load acquired by the water during its underground transit is, however, still small (10-30 mg CaCO₃/L) compared to other sites where dissolution proceeds via infiltration through the unsaturated zone. It is the high value of the total discharge, which accounts for the somewhat high dissolution rates found per km² of the system. All riverine organic fractions do not participate equally in the process. Dissolved organic matter as a whole probably reacts weakly than particulate, the no-fluorescent dissolved pool being the more reactive. Comparison between the pyrolysis characteristics of the particulate organic matter found in the recharge and the discharge waters makes insights into some more reactive fractions.

Karst chemical process occurs in a CO₂-H₂O-carbonate system. The purpose of this study is to model seasonal chemical denudation in karst terrain's of Cumberland Plateau, south central Tennessee. Carbon denudation is also modeled in the study area. Karst denudation is calculated from the equation: D = 0.001* R* H/p, where D is karst denudation rate, R is runoff, H is hardness, and p is density of calcite and dolomite. Runoff is estimated from gagging stations and climatic station. Hardness for equilibrium condition is estimated from chemical equilibrium equations. Soil P_CO2 is estimated from Kiefer equation (1990). Carbon denudation is calculated from the equation, C = 0.001* R* h, where C is carbon denudation rate, R is runoff, and h is carbon hardness. In this study two hydrogeologic settings (open and closed) and two petrologic settings (calcite and dolomite) are examined. Results show that in calcite, open system karst denudation rate is 31 mm/ 1000 years (36% in winter, 43% in spring, 15% in summer, and 6% in fall). In calcite closed system karst denudation rate is 4 mm/1000 years. In dolomite open system karst denudation rate is 36 mm/1000 years, and in closed system 5 mm/ 1000 years. Results also show that in calcite open system carbon denudation rate is 77g/m² year and in closed system 8g/m² year. In dolomitic open system carbon denudation rate is 99g/m² year, and in closed system 9g/m² year. It concludes that most karst chemical denudation occurs in wetter months in the study area.

The corrosion of carbonate rocks is caused by the vadose water which contains a great quantity of carbon dioxide. It is a key link of carbon cycle of coupling zone under the sedimentary lithosphere, atmosphere and hydrosphere in karst area. Both the karstologists and speleologists are paying a good deal of attention to it and hold some different views. There is a world model of soil CO₂ in which CO₂ concentration is predicted from actual evapotranspiration. As the CO₂ is derived by microbial activity in the soil, which is itself climatically controlled. The roots of green plants are a CO₂ pump, but no sufficient data of measuring CO₂ in soil air and loose deposits in the cave air have been collected.

More than eighty measurements of CO₂ were carried out in Slovenia karst area from the soil air in the first half of November, 1996. Other measurements of CO₂ have been carried out in China since 1988. The analyses were done at Alpine karst, Dinaric karst, low littoral karst, high karst and low inland karst of Slovenia.

All the measurements were done with a gas pump detector (GASTEC). Gastec tubes (2LL or 2L) give a direct reading of the CO₂ concentration in the air (ppmv).

The depths of measurements in the soil were 50cm, 30cm and 10-20cm partly. The values of CO₂ in the soil air vary from 1500ppm to 64000ppm in Slovenia and from 1200ppm to 42500ppm in China respectively. CO₂ in loose deposits air in the caves have also been measured. Some influences of the factors of moisture content, type of sediments, climate and so on upon the CO₂ level and vegetal debris carried by the clastic sediments of river caves upon the corrosional enlargement of caves have been discussed too.

Processes of chemical denudation in Spitsbergen have been investigated by University of Silesia polar team since 1972. Catchments in different landscapes were chosen: situated on uplifted marine terraces, non-glaciated but with permafrost (e.g. Fugleberget basin, 1.28 km²), glacier-covered (e.g. Werenskiold Glacier basin, 44 km²). Research was performed not only in polar summers (July-August) but also in the complete hydrological year, including polar winter. Chemical denudation rates were obtained from continuous recordings of water stages and detailed chemistry of water. Those data enabled calculations of CO₂ removal during weathering processes. In Fugleberget catchment chemical denudation rates were 6-7 m³/km²yr, the total amount of carbon used for dissolution of calcite and marble was in the range 20.4 - 23.5 * 10⁵ gC/km² year. In 18 years (from 1978 to 1995) chemical denudation rate was 108.5 m³/km² and carbon removal 450*10⁵ gC. In Wydrzyca catchment estimated carbon removal was 42.12-48.60 * 10⁵ gC/km² yr (value typical for most of the karstic, permafrost catchments).

In the Werenskiold Glacier basin chemical denudation rate of carbonate rocks in 1986 was 17.7 m³/km² yr. For dissolution of carbonates about 53.09µ910⁵ g C/km² yr was removed. Estimated carbonate rocks removal from smaller Kvisla basin was 1167.96-1362.62 tonnes and removal of carbon was in the range 94.43 - 109.00 µ9 10⁵ gC/km² yr.

Models of chemical denudation processes obtained enable attempts to estimate chemical denudation rates and carbon removal in all (46) catchments of Hornsund Fjord covering an area of 1270 km².

Dissolution and precipitation of calcium carbonate minerals in aqueous solutions with turbulent motion are controlled by a diffusion boundary layer (DBL) adjacent to the surface of the mineral, across which mass transfer is effected by molecular diffusion. To investigate the influence of such DBL to the dissolution rates of CaCO₃ a rotating disk technique was used. This technique allows an exact adjustment of the thickness of the DBL by control of the rotating speed of a circular sample of CaCO₃. Measurements of the dissolution rates in CO₂-H₂O-Ca²⁺-solutions in equilibrium with various partial pressures of CO₂ from 10^-3 up to 1 atm showed a dependence of the rates R on the rotation frequency, given by Rⁿ. The exponent n varies from 0.25 at low P_CO2 to about 0.01 at a P_CO2 of 1 atm. This reveals that the rates are not controlled by mass transport only, which would require n=0.5. The experimental data can be explained employing a theoretical model, which takes into account also the slow reaction CO₂+H₂O¬ áW H₂CO₃ and the chemical reactions at the surface (Dreybrodt and Buhmann, Chem. Geol. 90, 1991).

Interpretation of the experimental data in view of this model reveals that conversion of CO₂ plays an important role in the control of the rates. At high P_CO2 and large thickness >0.00lcm of the DBL conversion of CO₂ occurs mainly in the DBL and therefore becomes rate limiting. This is corroborated by the observation, that upon addition of the enzyme carbonic anhydrase, which catalyzes CO₂-conversion, the dissolution rates are enhanced by one order of magnitude. From our experimental observations we conclude that the theoretical model above enables one to predict dissolution rates with satisfactory precision. Since the precipitation rates are determined by the same mechanisms as dissolution, we infer that this model is also valid to predict precipitation rates. The predicted rates for both dissolution and precipitation can be approximated by a linear rate law R=(c_eq-c), where c_eq is the equilibrium concentration with respect to calcite and a rate constant, dependent on temperature, P_CO2, DBL thickness and the thickness of the water sheet flowing on the mineral. Values of CO₂ are listed that can be used for a variety of geologically relevant conditions.

The geochemical cycle of CO₂ in a carbonate rock area, Akiyoshi-dai Plateau (Yamaguchi Prefecture, Southwestern Japan), one of the biggest karst plateaus in Japan, has been studied in connection with the global carbon cycle. From the daily data of the calcium concentration and runoff obtained by continuous measurements of the stream issuing from Akiyoshi-do Cave, which has the biggest catchment in Akiyoshi-dai Plateau, 18.5 km², the following results were obtained: the calcium concentration of the baseflow showed seasonal fluctuations, following changes in CO₂ concentration in the soil. The calcium concentration in the groundwater is controlled by the water-limestone dissolution equilibrium, under open system conditions depending on the meadow's soil CO₂ concentration. At the runoff peak, the calcium concentration increases, because water is flushed out from the epikarst zone. During 1983 - 1986, a yearly average of 2,100 tons of limestone was dissolved in 2.1µ910⁷ m³ of groundwater issuing from Akiyoshi-do Cave, whose catchment basin includes 16.5 km² of a limestone area.The mean solutional denudation rate is 47 mm/ka, if the average specific gravity of carbonate rocks is 2.7 g/cm³. The total amount of CO₂ utilized by chemical weathering in carbonate rock areas all over the world, corresponding to the same amounts of chemically weathered carbonate rocks in mol, was estimated by using a limestone denudation rate of 50 mm/ka and found to be 8.9µ910¹¹ kg/y. The contribution of biological activities in surface streams in carbonate areas and deep source CO₂ in subsiding areas to the geochemical cycle of CO₂ will also be discussed.

Travertine deposition is an important short-term influence on carbon cycling and an interesting component of many karst landscapes. Nash Brook, in South Wales is a small, travertine-depositing stream located in a small valley choked with older travertine deposits. Today, travertine is deposited along a suite of small barrages within the river channel. Simple studies have been made of the morphometry and arrangement of these barrages, and short-term (8 months) measurements of travertine deposition made across a sample of the barrages. The morphometric studies indicate a complex series of relationships between barrage height, spacing and slope gradient and allow the testing of hypotheses about hydrological controls on barrage formation. Large woody debris seems to play an important role in the initiation of barrages, and various microflora (including the algae Vaucheria geminate) aid the small scale accumulation of travertine. The highest short-term rates of travertine deposition occur on barrage crests towards the headward end of the barrage system. Travertine formation today at Nash Brook appears to reflect the interaction of hydrological and biological controls. Although small in magnitude compared with the fossil deposits in this valley, travertine barrages play an important role in the karst geoecosystem here.

The ice core record from Greenland and Antarctica indicates that atmospheric CO₂ levels were elevated during glacio-eustatic sea-level highstands, and were at a minimum during sea-level lowstands. This CO₂ flux may be influenced by carbonate bank submergence cycles, based on the formula:

Ca⁺⁺ + 2HCO₃^- = CaCO₃ + H₂O + CO₂

The global area of carbonate bank subject to glacio-eustatic submergence/emergence cycles is 6 x 10⁵ km². During submergence, precipitation of CaCO₃ on these banks releases bicarbonate CO₂ at the rate of 200g/m²/yr of carbon. Average submergence cycles last 10⁴ years, releasing 1.2µ910¹⁸ grams of carbon as CO₂. Conversely, during emergence, CO₂ is sequestered as bicarbonate by dissolution of CaCO₃. Carbonate dissolution rates through rock volumes yield 75 m³/m³/yr of CaCO₃ dissolved. Given Quaternary limestone densities of 2.0 g/cm³, this is equivalent to 18 g/m³/yr of carbon sequestered. Average bank emergent cycles last 10⁵ years, sequestering 1.1µ910¹⁸ g of carbon. These values exceed the CO₂ flux (1.69µ910¹⁷) observed in the ice core record. The amount of carbon as CO₂ released during submergent cycles (1.2µ910¹⁸ grams of carbon) approximates that sequestered during emergent cycles (1.1µ910¹⁸ grams of carbon), indicated no net loss or gain of carbonate bank CO₂ during the Quaternary. During submergence, deposition occurs on the bank surface, whereas during emergence, dissolution takes place throughout the volume of the exposed bank. This difference in the location of carbon flux allows bank tops to keep up with subsidence.

Review of CO₂ measurements of my own and co-workers under various climates, from arctic (Labrador, Lapland) to warm climate (South-Central China) through alpine and temperate climates, including continental and humid ones. The presentation will not be statistical, but will be a geographical selection of case studies.

A cave studied during two years in the Grand-Duchy of Luxembourg will be considered with more detail, particularly regarding the relationship between CO₂ concentration and cave air wind, rainfall and outside temperature.

The traditional CO₂-H₂O-CaCO₃ system could be regarded as a Karst Dynamic System (KDS) because it can drive the formation of karst features; contribute to the regulation of greenhouse gas in the atmosphere and help mitigate environmental acidification; drive the migration, enrichment and precipitation of a number of elements and thus influence both the formation of mineral deposits and the biodiversity of life in karst areas; and record the course of environmental change. All this functions are played by shuttling of CO₂ and water into or out of the system. The principal function of KDS is to drive karstification. This activity and its close relation with environmental change is shown clearly and repeatedly from many KDS monitoring sites. For example, monitoring was carried out in Yudong underground stream, Zhen'an county, Shaanxi province in 1993-1994. The HCO₃^- data in water shows intensive dissolution in the system but also great seasonal variation from 150-160 mg/l in Winter to 210-220 mg/l in Summer. It also appears to show positive correlation with soil CO₂ and negative correlation with pH value of water. However, the time of greatest HCO₃^- concentration in August is not coincident with the period of highest soil CO₂ content in June, but rather appears one month after the highest rainfall in July. The findings show that the transfer of CO₂ from air into water rather than its absolute concentration in air plays a most important role in karstification. On the basis of published world carbonate rock denudation data the contribution of karst denudation to global CO₂ uptake from the atmosphere can be estimated by simply correlating the amount of CO₂ to the world denudation of carbonate rocks, because 120 kg of carbon is removed from the atmosphere for every tonne of limestone dissolved. It makes 6.08µ910⁸ t/a, and accounts for about 15.9% of the "missing carbon sink" (38.12µ910⁸ t/a) in the global carbon cycle. So far as IGCP299 and IGCP379 are concerned, epigenic karst processes are considered to act more often as a sink of atmospheric CO₂ than a source. In contrast, in karst areas with active faults and geothermal or volcanic activity, CO₂ is commonly measured emitting into the atmosphere at concentration of 23-90%, and large amounts of calcareous tufa are often deposited. It is widespread in the world, especially in the Pacific Rim and the Tethys Belt, extending from Tibet westward through Iran, Turkey, Italy and southeast France. This is an important source of atmospheric CO₂.

Carbon dioxide has a major role in both the solution and precipitation of calcite. As such the levels of carbon dioxide in the Jenolan Caves were examined with a view to predicting the effect on the calcite speleothems of some 250,000+ visitors annually. Jenolan Caves are exceptionally well decorated and preservation and conservation of the speleothems is of prime importance. Grab air samples showed that throughout the caves carbon dioxide levels were elevated and preliminary studies indicated that these in part were associated with guided tours. A three-year study consisted of the following: Potential sources of CO₂ were identified and quantified. Temporal variations of CO₂ concentration were measured continuously in three different cave microclimates using a portable gas chromatograph. Cave water analysis and modeling were used to determine if levels of CO₂ were high enough to dissolve in cave water causing calcite dissolution. Thresholds of CO₂ corrosion were subsequently calculated using modeling of depositing waters. Surface examination of calcite from the speleothems was performed using scanning electron microscopy (SEM).

The CO₂ levels in the caves were found to be markedly affected by tourist activity with the subsequent return to background levels were elevated above the value of the surface atmosphere and the stable isotope investigations revealed microbial activity to be a major contributor to background CO₂. Further measurement of CO₂ levels and their application in modeling of cave waters showed that these levels were not high enough to cause aggressive solutions to calcite. All waters sampled would degas CO₂ leading to calcite saturated solutions. Calculated corrosion thresholds varied over an order of magnitude for the analyzed waters. In all cases the threshold was not being exceeded based on maximum CO₂ at the site measured for the year. However, the SEM examination showed that etching of the inactive speleothems was occuring throughout the caves. This has led to an investigation of the relationship between visitor produced carbon dioxide and condensation in the Jenolan Tourist Caves.

Frasassi karst area is located in Central Italy where develops more than 100 caves with tens of kilometers in length and more than 2µ910⁶ m³ in volume. The caves consist of a maze systems with solutional galleries of hypogenic origin that reach the phreatic zone where both meteoric carbonate and sodium-chloride-sulfur ground water circulation occurs.

The main karst processes take place close the water table where the oxidation of hydrogen sulfide to sulfate occurs in the upper phreatic zone, in the presence of oxygen coming from dripping waters and from the cave atmosphere. Limestone dissolution takes place both in phreatic and vadose conditions and producing CO₂. The rate of sulfuric acid speleogenesis was determined during an experiment with limestone tablets that exhibited surface corrosion and the weight loss with values of up to 20 mg. cm^-2 . year^-1.

Hourly records of CO₂ concentration in the air of the caves shows values from 600 to 1,400 ppm related to the rising of sulfidic water. From July up to November 1997 the record of CO₂ above the sulfidic lake could be well related to the seismic activity of the area. In the touristic part of the cave the natural background of CO₂ concentration in the air is about 400 ppm and reach 1100 ppm with the anthropogenic contribution.

Certain uncommon speleothems such as folia, raft cones, and euhedral calcite pool linings are often assumed to be precipitated by hot water because many examples occur in hypogenic caves. However, these features are more appropriately linked to high rates of CO₂ degassing, since they are also found in normal cold-water caves. For example, Mystery Cave, Minnesota, a distinctly non-thermal cave at the southern edge of Pleistocene glaciation, contains a wide variety of these deposits. Raft cones over 2.5 meters tall, spar-lined pools, and folia are common in this cave, which was filled several times with glacial sediment. Modern speleothems resemble those of past interglacial periods, so it can be assumed that the water chemistries are similar. Cave lakes are recharged by water of high P_CO2 (up to 0.01 atm) and moderate ionic strength (up to 0.014), which infiltrates through thick, organic-rich soil a short distance above the cave. Much of this water enters the cave during flashy recharge events. Raft cones form where infiltrating water drips into cave pools, losing CO₂ as it falls and typically reaching about 10 times supersaturation before it hits. Calcite precipitates almost immediately at the surface. Folia form water-level rings around pools fed by underwater inlets for CO₂-rich water. During floods up to half the CO₂ in the inlet stream is lost by degassing at the water surface. Near the inlets, pool crystals consist of layered, non-touching, long and narrow blades. Chenille spar lines nearby pools. Pools remote from water inputs remain saturated at low ionic strength, and the morphology of pool-lining crystals changes slowly by recrystallization, producing large crystals that resemble those of thermal deposits.

The origin of deep solution conduits in karst depends upon closed-system dissolution, but this process has many limitations. Dissolution rates diminish rapidly downflow from recharge sites, and aggressiveness from oxidation of organic compounds or sulfides is restricted by the generally low oxidation potential of deep groundwater. Mixing solution involving contrasts in CO₂ is limited in waters derived from similar recharge sources. However, where one water type is isolated from its initial CO₂ source (e.g. recharge through an insoluble caprock), large contrasts in P_CO2 can be achieved, and mixing solution is far more effective. Under these conditions, dissolution rates are determined mainly by groundwater flux, rather than by chemical kinetics. H₂S can enhance carbonate solubility, but mixing of waters of contrasting H₂S can boost aggressiveness only if the P_CO2 is low, because otherwise the solubility curves are nearly linear. Yet even at high H₂S concentrations, the mixing of waters of contrasting CO₂ can still produce considerable dissolution. Many deep karst waters contain much sulfate and magnesium, indicating dissolution of dolomite and sulfates far beyond their normal solubilities (especially if low salinity shows that connate water is not the source). In carbonate-sulfate systems, calcite precipitates while dolomite and gypsum dissolve -- up to 5 times more dolomite than in simple dolomite systems, and at least 1.5 times more gypsum, with a large net gain in porosity. The slow dolomite reactions prolong the solutional history, fostering the development of lengthy conduits under closed conditions. The nature of their evolution to shallow accessible caves is uncertain.

The Taroko Limestone gorge area is covered with limestone over 1,000m in thickness and an exceptionally beautiful, narrow ravine created by a river which has cut through mountains of marble: Taiwan's most popular scenic spot. The Taroko area is sub-tropical and humid. The precipitation is over 2,000mm a year, and the rainwater is discharged via a lot of springs as well as river. But underground drainage systems and caves have not been enough developed in this area.

Taiwan Island is located along the boundary of two tectonic plates (the Philippine Sea Plate and the Eurasian continent Plate), and one of the fastest uplift areas in the world. Therefore, an extensive downward erosion is expected to contribute to karst landform development in the Taroko area.

On the other hand, in a short time scale, karst landform development can closely be related to local climate and CO₂ movement via groundwater. Groundwater containing CO₂ produced by biological activity in the soil, dissolves carbonate rocks such as marble in the Taroko area more than water free from dissolved CO₂. Furthermore, in the Taroko area, CO₂ may be provided from a deep source, and also contribute to the limestone solution. Therefore, it is important to estimate quantitatively the effect of CO₂ movement on karst landform development.

In this study, the distribution of caves and groundwater drainage systems, the water chemistry and discharge of groundwater, and the soil CO₂ concentration have been surveyed in the Taroko area. On the basis of the results obtained, the catchment of underground drainage systems and the development of karst landform in this area will be clarified in connection with CO₂ movement.

The investigations of thermomineral water and geothermal energy in Vranjska Spa have been done with the aim to establish ground water reserves and quality. The results of previous investigations proved the reserves of 100 l/s, with temperatures between 85 and 111°E C, which presents a geothermal potential of about 33 M. The geothermal water is of hydrocarbonate-sulphate- sodium type, with low TDS (1.2g/1), carbon-dioxide and hydrogen- sulfide in gas composition and many trace elements. The predicted temperature on higher depths is up to 180 °E C, on the basis of chemical hydrogeothermometers. Hydrochemical investigations suggests that geothermal water have great scaling and corrosion potential.

Alan L. Mayo^a, Anthony B. Muller^b

^aDepartment of Geology, Brigham Young University, P.O. Box 25111, Provo, UT 84602-4646,USA
^bScience Applications International, Mclean,VA,USA

Journal of Hydrology 194 (1997) 286-304

A small (< 15 km²), low temperature (< 20íWC), CO₂ gas overpressured, gently dipping Paleozoic carbonate aquifer rests on Precambrian granite in a narrow canyon along the eastern edge of the Rocky Mountain Front Range, Colorado. The carbonate aquifer is bounded on three sides by granite and in the down dip direction by the Front Range fault. The fault, in a major continental intraplate zone of weakness along which magmatic fluids and gases could migrate to the surface and near surface from great depths. The fault has also overthrusted slices of carbonate and clastic rocks several kilometres below the granitic basement.

Stable isotopic (d ²H and d ¹⁸O) and discharge temperature data suggests that carbonate aquifer ground waters are of meteoric origin and have not circulated to depths greater than the base of the carbonate aquifer (»E 650m). Elevated CO₂ and the d ¹³C of HCO₃^- in the carbonate aquifer suggest an external crustal source of CO₂ gas. ³He/⁴He, O₂/N₂ and Ar/N₂ gas ratios indicate gas contributions from both magmatic and atmospheric sources. Atmosphere contributions account for about 25% of` the exsolving gas, whereas magmatic CO₂ accounts for 7 to 14%. Possible external CO₂ sources, which are consistent with the mean HCO₃^- d ¹³C = - 2.4‰ (PDB), are clay-carbonate mineral diagenesis or low temperature metamorphism of siliceous-carbonate rocks that have been overthrust by 3 to 6 km of granite. Diagenetic or metamorphic CO₂, mixed with some magmatic gas, appears to have migrated from the source rock area upward along the Ute Pass thrust fault until it encountered the shallow carbonate aquifer ground water system where it was further diluted with atmospheric gas.

We demonstrated the potential of the quantitative theory of solubility of karst rocks (Shopov et. al, 1989,1991) in dependence of the temperature and other thermodynamic parameters to make reconstructions of past carbonate denudation rates. For this purpose we obtained a stacked 66000 data points paleotemperature record from Rats Nest cave, Kananaskis karst region, Alberta, Canada. It covers last 1450 yrs with resolution of about 8 days for most of the time span. Paleoclimatic records has been derived from speleothem luminescence, calibrated by actual climatic records from near climatic station in Banff, Alberta. The sample was dated by two 14-C dates, an U/Th age estimate, autocalibration and annual bands counting dating. All produced consistent age, best estimated as 1450 +/- 80 years. A reconstruction of the past annual precipitation rates for the last 280 years has been obtained from speleothem annual growth rates, derived from the distance between annual speleothem luminescence bands ("Shopov bands"), calibrated by actual precipitation record from near climatic station in Banff, Alberta.

The carbon isotopic composition of stalagmite reflects the d ¹³C of soil CO₂ above the limestone cave, which varies as a function of the relative proportion of C3 and C4 plants grown at that locality, and of the CO₂ exchange between soil and the atmosphere. The d ¹³C values of a stalagmite collected from Shihua Cave near Beijing have shown a decrease from -6.7‰ to -8.1‰, with a rate of -0.028‰/year since 1940. This d ¹³C decrease in excellent agreement with the previous studies indicates the isotopic dilution of the d ¹³C in the atmospheric CO₂ and soil CO₂ by consumption of fossil fuel. The d ¹³C values of the same stalagmite was elevated by 3‰ from 1270 A.D. when Beijing became the capital of China to 1550 A.D. During this period, lots of trees from Xi Shan Mountain where Shihua Cave is located were consumed for the construction of Beijing and other purposes. This serious anthropogenic deforestation reduces the C3/C4 plant ratio, which had never returned in the region. The d ¹³C of a stalagmite from Fengyu Cave near Guilin exhibited 6‰ increase from -12‰ in 1780 A.D. to -5.6‰ in 1900 A,D., due to an anthropogenic deforestation. Carbon cycles on the Earth's surface involve the variability of carbon budget related to vegetation change and fossil fuel burning which are strongly influenced by human activities. High- resolution d ¹³C records in speleothem appear to be an excellent proxy for understanding of such influences and reconstructing regional carbon budgets.

New cosmogenic nuclide techniques can be used to date quartz sediment deposition within caves, by measuring the radioactive decay of ²⁶Al (half-life 0.71 million yr) and ¹⁰Be (half-life 1.5 million yr). These radionuclides are produced within quartz near the ground surface by nuclear reactions with cosmic rays. Sediment burial within a cave shields the quartz from further cosmic radiation, burial times may be inferred from differential rates of radioactive decay.

Results from the Mammoth Cave system indicate that sediments within the uppermost 40 meters of the cave date to approximately 2.5 million years ago, coincident with the onset of northern hemisphere glaciation. These sediments may have been laid down as a result of massive aggradation of the nearby Green River. Sediment dates from lower passages indicate that Green River incision has lowered the regional water table at a rate of roughly 30 m/m.y.

The Lithophagus Cave is located in the Middle Basin of Iada Valley, within the easternmost part of the Padurea Craiului Massif (Apuseni Mts., Transylvania).

LFG-2, a 39.5 cm tall stalagmite from north-western Romania has been dated by U-series a-spectrometric dating, and analyzed for stable isotope variations (d ¹⁸O, d ¹³C) along its growth axis. The sample grew all the way through oxygen isotope stage 5(a-e), and perhaps for some time into stage 4. In spite of a rather low uranium content and therefore imprecise chronology, the sample provides an interesting stable isotope record with high temporal resolution that correlate favorable with other speleothems and with the deep-sea record. Termination II is well defined in the record as a rapid shift from light (cold) to heavier (warm) d ¹⁸O values, when C3 vegetation seemed to dominate. The d ¹³C in a slow growth zone, corresponding to oxygen isotope stage 5d, as well after the stage 5/4 transition, suggests that C4 plants possible dominated the surface environment. Also, the d ¹⁸O record correlate quite well with the a-dated FM-2 record from northern Norway.

In order to understand karst dripwater evolution, it is necessary to determine the main factors responsible for modifying the geochemical signal of rainwater during flow through soil and aquifer zones.

Monitoring has been performed at two near-surface sites. At Crag Cave, County Kerry, Ireland, daily sampling occurred in summer and winter, to allow investigation of geochemical subsurface water response to short time-scale variations in surface water input. At P8, Derbyshire, England, monthly sampling occurred over one year, in order to ascertain seasonally influenced changes in karstic chemistries. Dissolution and ion exchange laboratory experiments have been performed in order to constrain the weathering behaviour of solids system, and their contribution to karst water trace element and Sr isotope composition.

Variations in soil gas P_CO2, soil and aquifer residence time and hydrological routing are controlling factors, influencing both spatial and temporal changes in cation yield in soil and aquifer zones and hence the mode of evolution of speleothem-forming karstic drip waters. At different sites, trace element composition reflects either dilution effects along well-connected flow routes, or prior calcite precipitation along the flowpath under slow flow regimes and along poorly-connected pathways. Some sites are both hydrologically and/or chemically unresponsive to surface rainfall events.

An understanding of the influences of dominant controlling factors, coupled with information concerning stalagmite growth rate, is essential for interpreting the geochemical signal of banded speleothems. The postulated time-scale of the variation in surface conditions preserved can be used to determing the continuity of the palaeoclimatic record.

The possibility of using routinely measured d ¹³C for climate reconstruction on Quaternary speleothems is investigated. Recently there has been a great deal of interest in the coupling of cloud cover to solar wind and cosmic ray fluxes, and hence to climate change. Luminescent studies of speleothems hint that the concentration of organic acids dissolved in groundwater is linked to solar insolation, which has a direct link to cloud cover. Because organic carbon is carried by these acids, there is a further suggestion that luminescence and the d ¹³C may be coupled, but only in some favourable cases as the latter involves both the inorganic and organic carbon components.

We present d ¹³C and d ¹⁸O of speleothems from a western Canadian Cave site where we find a link between the d ¹³C and luminescence during some periods, and a degree of correlation between the d ¹⁸O isotopic temperatures. Two other interesting features, which we do not fully understand, is the tendency for d ¹³C to decrease back in time and the d ¹³C for stalagmites and flowstone to be offset from one another.

Strontianite crystals can be found in many vugs in carbonate rocks (e.g. Silurian Rondout Dolomite of NewYork State). These are often found with evaporite mineral assemblages that may include celestite. Celestite is usually precipitated in a sabkha environment under extreme evaporative conditions. Strontianite is generally a low-temperature hydrothermal precipitate, but this paper offers an alternate model in which the mixing of low-temperature groundwater solutions undersaturated with celestite and carbonate minerals precipitates strontianite. A phase diagram for calcite-celestite solutions was modeled to examine this hypothesis. A strontianite domain does occur in solutions undersaturated for both celestite and calcite. It also overlaps with those of both other minerals. The concentration of either original solute necessary for strontianite precipitation varies little with the abundance of the other solute. If P_CO2 is increased in a calcite-celestite system at low temperatures, the strontianite domain diminishes until, at a P_CO2 of 0.1 atm, the boundaries of all three mineral phases meet at a single point. If dolomite is substituted for calcite, the calcite and strontianite boundaries shift toward lower concentrations and the triple intersection is present only above the highest P_CO2 modeled (0.2 atm). With increasing temperature the concentration of the solutes and the P_CO2 needed to reach the triple point of the three phases decreases. Mixing of solutions undersaturated with celestite and either calcite or dolomite is theoretically able to precipitate strontianite. If the carbonate is dolomite, strontianite can precipitate over a broader range of P_CO2 and temperature.

DNA from excrements can be amplified by means of the polymerase chain reaction. However, this has not been possible with ancient feces. Cross-links between reducing sugars and amino groups were shown to exist in a Pleistocene coprolite from Gypsum Cave, Nevada. A chemical agent, N-phenacylthiazolium bromide,that cleaves such cross-links made it possible to amplify DNA sequences. Analyses of these DNA sequences showed that the coprolite is derived from an extinct sloth, presumably the Shasta ground sloth Nothrotheriops shastensis. Plant DNA sequences from seven groups of plants were identified in the coprolite. The plant assemblage that formed part of the sloth's diet exists today at elevations about 800 meters higher than the cave.

(17 JULY 1998, VOL 281, SCIENCE, P.402)

In the movie Jurassic Park, a collector snapped up hundreds of thousands of mosquitoes preserved in amber for DNA they had sucked from dinosaurs. In the real world, however, amber has been a disappointment, yielding no reproducible traces of ancient genetic material. Now researchers report that the treasure of ancient DNA can instead be gleaned from a less glamorous material: fossil feces.

On page 402, a team led by molecular biologist Hendrik Poinar and geneticist Svante P±E±Ebo of the University of Munich demonstrates a way to unlock DNA trapped inside ancient feces. The dung they studied, a firm lump left by an extinct ground sloth about 20,000 years ago, offers clues to that species' ecology. Applied to other droppings, the method may be able to provide a wealth of clues about the ecology and relationships of extinct animals--and perhaps even about early humans. "This adds several new dimensions to the study of ancient animals," says Bob Wayne, an evolutionary biologist at the University of California, Los Angeles.

The P±E±Ebo lab is one of the few to have successfully extracted DNA from ancient bones (Science 11 July 1997, P.176). But the team wasn't having any luck with the well-preserved samples of fossilized dung, called coprolites, collected from Gypsum Cave near Las Vegas, Nevada, a gathering place for ice age animals. Then the researchers chemically analyzed the samples and found several compounds that indicated the presence of Maillard products-sugar-rich tangles of proteins and nucleic acids that prevent DNA amplification, "Everyone looks at the Maillard product as evil," says Poinar. But he realized that the tight cross-links might protect DNA by keeping out damaging water and microbes. The question was how to crack open that coat.

In 1996, the team spotte a possible answer in a Nature paper on a chemical called N-phenacylthiazolium bromide (PTB), which when given to diabetic rats cleaves the bonds between sugars and proteins--the same kind of bonds that may entangle DNA in the Maillard products. "We thought: 'Wouldn't it be great if PTB would release DNA?'But it was still a complete shot in the dark," recalls Poinar.

The shot hit home. Extracts from the sloth coprolite treated with PTB yielded sequences of mitochondrial DNA, presumably from intestinal cells shed into the feces. It probably came from an extinct ground sloth, Nothrotheriops shastensis, because the bones of that animal are scattered throughout the cave and because the DNA is a good match to that of a related extinct ground sloth, Mylodon darwinii, whose DNA was derived from bone and soft tissue.

The team was also able to extract a wide variety of plant DNA from the coproliteclues to the vegetarian sloth's diet. They identified sequences from eight plant families, including grasses, yucca, grapes, and mint. The coprolite had identifiable fragments of only five families, so DNA analysis may help identify plants chewed beyond recognition, says Poinar.

The team hopes to study more sloth dung to help answer the question of why these and other large animals vanished from North America about 10,000 years ago. "We'd like to ... see if there’s a change in diet before they go extinct," says P±E±Ebo. Climate change, a possible agent of extraction, might show up as a change of diet, he says.

Right now, P±E±Ebo is analyzing samples of what could be Neandertal feces, from 45,000-year--old cave deposits in Gibraltar, "They look human, but it's hard to be sure that they're not jackal," he cautions. If the samples do contain Neandertal DNA, they would be the second such sequence ever and could offer additional evidence in the continuing debate over this extinct human's kinship to our own species, "A second sequence would give a real window on Neandertal variation," says paleoanthropologist Christopher Stringer of The Natural History Museum in London, who discovered the feces last summer. They could also reveal what Neandertals dined on and what parasites may have plagued them. "Five years ago, we wouldn't have thought we would have the possibility of reconstructing Neandertal diet in this way," says Stringer.

Still, some paleontologists caution that DNA from dung may not reveal everything its proponents hope for. Changes in coprolite contents could simply reflect seasonal shifts rather than pointing to causes of extinction, says Russ Graham, a paleontologist at the Denver Museum of Natural History. The technique may not work on coprolites found in warmer or wetter conditions, or on very ancient samples, as most DNA is thought to degrade within 100,000 years, says Poinar.

Despite such caveats, "I'm gathering as much poop as I can," Poinar says,"There's going to be a run on feces."

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