The red bed karst developed in the Jurassic and Cretaceous in Sichuan basin widely distribute in the east of the Longquan Mt., west of the Huaying Mt., south of the Yingshan Mt. and north of the Yibin. Geomorphologically, it presents redbed hills. The red bed rock contains a lot of calcareous cement, calcareous pebbles and a little salt and gypsum rocks, and it mainly characterized by some dissolution pores and fissures. Only some caves with length of 1500m and considerable karst depressions and sinkholes can be seen in a local area. Occurrence of the red bed karst impels, on the one hand, the rock bed to enhance its water absorbing capacity and become the main water supply layer of lacking water red bed area. On the other hand, it may also worsen the engineering geological condition of rock fabric and lead to problems of environmental engineering geology.

Therefore, it is a very important and practical significance to study the development and distribution rule and features of the red bed karst.

                                                                                  (extracted from paper by Wei Lunwu)




The limestone caves on the Akiyoshi plateau were divided into three groups. They were called high altitude group, middle altitude group and low altitude group in order of age of formation. The relationship between each group and karst landform was obtained as follows. At the age of high altitude group, the characteristic karst landform was not recognized. The Poljes and uvalas existing now In this area originated at the age of middle altitude group. Then the present karst landform was formed at the age of the low altitude group.

The limestone eaves on the Akiyoshi plateau are mostly distributed at three levels of height. The cave group of the same level was considered to be formed at the same age which the base level of erosion was stable, and the geomorphic surface was formed at the same time. Accodingly, cave group was recognized to be correlated to geomorphic surface. (Kawano, 1972) ...

                                                 (extracted from a paper by M. Kawano, 1972)



The Haidigou underground karst reservoir is the earliest one built in karst area of parallel ridge and valley in eastern Sichuan. Considering the main topic of IGCP 299, the basic karst geological environment where the underground reservoir was built is a typical example of -buried karst under elastic rock bed. Limestone with intercalated gypsum bed make up the core of an anticline, and supply by vertical leakage from overlain sandstone aquifer. The karst strongly develops, and huge underground rivers and passage-spaces formed in the mountains make up the underground reservoir. The formation and development of the underground river are related to the paleogeomorph and hydro-environment. It is valuable to study underground reservoir as a part of IGCP 299

                                                           (extracted from a paper by Riao Zhongling)



The speleological investigation, drilling and isotopic age of cave deposits suggest that:

1.there are two karstified periods, the Laohudong period of Middle Pleistocene (158-227 ka) and the modern karstified period.

2. 36.3 m3 of dolomite, 556.7 m3 of calcite and 10 m3 of pyrite p.a can be dissolved in the modern karstified period. The volume of pure dissolution, except corrosion and transportation, is 603 m3 p.a. Three karst subsystems of the modern karstic Period are recognized, they are peak cluster montainous district, peak cluster basin and Caixuhe Valley subsytem.

3. A complex mutilayer karst system controlled by the same base level can simultaneously be developed.

4. In general, a transitive zone, that is a weakly karstified zone formed between two subsystems, is located in front of karst subsystem outlet and there is a deep siphon nearby.

5. Rate of the crust uplift was 0.57 mm.p.a. for Middle Pleistocene in Beishan area.

                                             (extracted from paper by Zhu Yuanfeng and Cui Guangzhong,1988)



In each natural geomorphic zone in eastern China, underground water run-off and corrosional rate are regular in distribution.

The following table shows the annual precipitation, depth of annual run-off and corrosions] quantity measured from some springs located in north China, middle China and south China, which represent respectively different climate zone in eastern China.

It shows that the factors of climate and hydrology influence distinctly the modern corrosion. The quantity of corrosion in the semi-arid area of the warm temperate zone is only about 10% of that in the humid area of the middle subtropical zone. In the north Shanxi and middle Shandong, both in warm temperate zone, the difference of corrosional quantity is as much as three times due to the difference of aridity.

Measured location

Climate zone

Annual precipitation(mm)

Depth of annual run-off(cm)

Corrosional rate(mm/1000ys)


North Shanxi

Semi-arid area wars temperate zone





Middle Shandong

Sub-humid area of warm temperate zone





west Hubei

humid area of middle subtropical






North Guizhou





east Yunnan








(extracted from a paper by He Yubin)



The Zhujiang river's watershed covers an area from eatern Yunnan to middle Guangxi and west Guangdong, which lies on plateau of 2000m a.s.l. in east Yunnan to 500-200m a.s.l. in east Guangxi and west Guangdong.

The karst evolution of the Zhujiang river's watershed is:

Karst plain Negative landforms Karst


Karst basin Karst plain
Isolated peak,

peak forest

Positive landform Peak cluster,

Peak forest

Peak forest Isolated peak
Subterranean river system ---------- > Subsurface form Isolated subterranean


Subterranean ---- >


Subterranean river system


Disintegration of paleokarst plain caused by headward erosion   Recent karst plain
Horizontal process (stable earth crust) Vertical process

(earth crust uplift)

Horizontal process strengthen transition area Horizontal process(stable earth crust)
Eastern Yunnan Geomorphic slope zone in west Guangxi Middle Guangxi eastern Guangxi- western Guangdong

Neotectonics controls the geomorphic structural frameworks, and the latter controls condition of hydrodynamic through potential energy. In the different geomorphic portion, the evolution trend of distinct karst is different due to different hydrodynamic condition, and finally result in the intricate karst feature complex.

                                                                                      (extracted from a paper by He Yubin)




Jewel Cave (118 km of mapped passages beneath an area of 2.7 km2) and Wind Cave (70km beneath 1.8 km2) are, respectively, the fourth and tenth longest known cave systems and the world's foremost examples of three-dimensional, rectilinear networks of solutional passages. Other caves in the Black Hills are similar. They occur in 90-140 m of well-bedded Mississippian limestone and dolomite. Walls throughout Jewel Cave are lined with euhedral calcite spar as much as 15 cm thick. Wind Cave displays lesser encrustrations and remarkable calcite boxwork. Since 1938, opinion has favored cave excavation by slowly circulating meteoric waters in artesian confinement similar to that surrounding the Black Hills.

We believe that the caves were developed by regional thermal waters focusing on paleospring outlets in outlying sandstones. Four sets of criteria are evaluated: (1) morphological--the three-dimensional, one-phase maze form having convectional features is similar to known and supposed thermal caves in Europe; (2) petrographic and mineralogical study of the chief precipitates shows a record of carbonate solution-->calcite precipitation consonant with a model of cooling, then degassing, waters; (3) a thermal anomaly at regional hot springs is shown to extend beneath Wind Cave, where basal lake-water samples show chemical and isotopic affinities with the thermal waters; and (4)d 13C and d 18O measurements place all suspected paleothermal water precipitates in the domain of thermal calcites reported by others and being deposited at the modern hot springs. Finally, U-series dates show that the Wind Cave deposits are Quaternary and that the cave is still draining. Jewel Cave is truly relict and divorced from the modern thermal ground water system; its great calcite spar sheets are probably older than 1.25-1.5 Ma.

The evidence we have presented suggests that the large network caves of the Black Hills were formed by CO2-rich waters that were heated and that ascended through the Pahasapa Formation. Most or all of the thermal precipitates in Wind Cave are of Quaternary age. It is possible that the dissolutional enlargement of this cave was also limited to the Quaternary, but we suspect that it probably began during the Miocene or Pliocene. Formation of Wind Cave is compatible with the geothermal and hydrogeological conditions that exist in its local region today. Waters rose and converged through what is now the cave zone en route to spring points in an adjoining valley. As a consequence of continued surface entrenchment, spring positions have shifted down dip. The valley is now a dissected relict, and the known cave has become a drained backwater. It retains a strong thermal gradient, and final-stage warm-water precipitates (the "calcite ice") are still forming at the modern water table in it, where exploration is terminated.

Jewel Cave is significantly older and more comletely relict. It is not related to the hydrogeological conditions prevailing in its region today. Nevertheless, we suggest that it was formed in the same mode as Wind Cave, by thermal waters rising and converging through it toward spring positions in an earlier level of Hell Canyon. The culminating morphological event in Jewel Cave was the deposition of the calcite spar sheets. They are among the greatest known in any cave. We have shown that they are deposited from thermal waters probably at some time before about 1.0 Ma.

The caves display a broad tendency to desend in stratigraphic elevation in the direction of stratal dip. This suggests that the thermal plumes were rising with an updip component. An igneous heat source within the Precam-brian rocks is envisioned, related to the igneous activity occurring else where in the Black Hills throughout much of the Tertiary. Recharge to the systems was probably from infiltration of meteoric waters over wide areas, a pattern of circulation that has been documented in many geothermal systems. The position of the Pahasapa Formation close to the base of the sedimentary sequence plus the greater elevation of the Black Hills make it unlikely that any significant part of the cave discharge consisted of basinal fluids from strata beneath the surrounding plains.

There are three principal modes of cave development: (1)solution of limestone and dolomite at nearly equal rates by water considerably undersaturated with respect to both carbonates, which mode was quantitatively predominant, (2) selective solution of dolomite only, by water near to saturation with calcite, and (3) deposition of calcite from supersaturated water. In an ideal thermal model, all three stages will occur simultaneously in a vertical sequence. At any fixed Pco2, the saturation concentration of dissolved carbonate in water increases as the temperature decreases. As the thermal water rises and cools, it acquires or retains solutional aggressiveness with respect to both calcite and dolomite, regardless of the initial dissolved carbonate content. If cooling of water is very gradual, however, the system can hover near the saturation value of calcite and dolomite, preferentially dissolving the species that is more soluble under prevailing geochemical conditions.

Decreasing hydrostatic pressure in the rising water may allow partial degassing of CO2, which sharply reduces the saturation concentrations and causes precipitation of the secondary carbonates. At most sites observed by Ford in the Budapest thermal caves, precipitation was intense down to 2 m below the paleo-water tables and reduced to zero at depths greater than ~10m. Rapid degassing in well-ventilated hills best explains such sharp zonation. Wind Cave is not so well ventilated (despite its name); slower degassing probably explains its poorer zonation of precipitates. Simultaneous deposition of the spar crust throughout Jewel Cave can be explained by a phase of warming, including degassing within the cave zone, or by a protracted backwater phase marked by very slow circulation and degassing.

This model is highly simplified and must be modified to account for details in the history of the individual caves. Water levels appear to have fluctuated in response to local aggradation at springs or to wetter spells during the later Tertiary and Quaternary, within the over-all lowering induced by regional erosion. As higher springs were abandoned, meteoric floodwaters may have penetrated by way of them, contributing to cave enlargement.

Finally, the three-dimensional network pattern of the caves must be explained. Their origin requires a way of distributing the solutional capacity of the water rather uniformly between major joints over particular areas of several square kilometers or more. We suggest that regional and diffuse, heated discharge converged upon what become the cave zones, flowing up dip in the Pahasapa and ascending through the lower formations. Cooling of these waters simultaneously throughout the joint nets produced the crucial solutional aggressiveness. Fluctuating head within the evolving cave zones (in response to varying recharge) and mixing corrosion probably played subordinate roles.

(extracted from a paper by M.J. Bakalowicz, D.C. Ford, T.E. Miller, A.N.  Palmer and M.V. Palmer, 1987)




The relationship between karst and glaciation in formerly glaciated terrains in Canada is reviewed. The studied examples suggest that karst landforms may be divided into eight types in terms of their temporal relationships to the record of reported Quaternary glaciations. Two types are postglacial, two are subglacial, one type occurs where glacial features are adapted to karstic drainage, and three types display sequences of karstic and glacial action. Glacier effects upon karst landforms and their underlying aquifers display the gamut of possibilities. They may destroy, inhibit, preserve, or stimulate karst development. Where continuous permafrost is maintained when covered by glacier ice, postglacial karst is limited to the active layer epikarst. Where permafrost is thawed beneath ice or during deglaciation there are a variety of postglacial karst developments, depending in part upon climate and in part upon local lithologic and relief conditions.

It is clear that amongst the karstic rock terrains of Canada, there has existed a wide variety of relationships between glacial processes and those of karst system genesis. The range of effects is from complete destruction to perfect preservation of antecedent karst systems, and from prohibition to mostrapid generation of new ones.



Fig.1 Map of the outcrop of karstic rocks in Canada

(extracted from D.C. Ford's paper, 1987)




Diagenetic paleokarst breccia masses occur on the Cambrian carbonate rock strata, where two key water conservancy projects located in Gaubazhou and Geheyan-at lower reach of, Qingjiang River, Hubei.

Karst breccia is characteristicly similar to normal sedimentary rock, which consists of particle, filling, cement and pores.

According to the relationship between breccia and filling, feature of filling, supported types of breccia and property of cement, karst breccia can be divided into four rock types: travertine cementing broccia, argillaceous matrix supported breccia, calcareous clast filling breccia and calcite cementing breecia. .

After detailed microscopic analyses, it is clear that the calcareous clast filling breccia was formed in subsurface vadose zone, and the calcite cementing breecia was formed in subsurface saturation zone.

Paleokarst breccia can be assigned to three kinds of genetic categories, i.e. karst collapse breccia, which can be subdivided into calcareous clast filling subdivision and calcite cementing subdivision. karst modified fault breccia and karst filling clastic rock.

Based on studies of petrological feature of karst breccia, and relationship between space-distributive position of breccia and relative dissection of gorges combining the history of regional tectonic movement and few dating data, it is concluded that the purplish red calcareous clast filling karst collapse breccia was formed In late Paleogene, the giant crystal calcite cementing karst collapse breccia in late Neogene and the karst fault breccia in Pleistocene.


(extracted from a paper by Shen Jifang, Wang Liangchen and Wang Zengyin,1990)





Sediments as natural tracers have different value, depending on the karst, on the underground and on normal relief in vicinity. Flysch pebbles in Kacna jama do not prove the underground Reka in this cave just because they take their origin in Ecocene flysch sandstones but that Reka comes from somewhere on flysch. In such case the sediments as a natural tracer cannot be considered from hydrological studies point of view but from the point of view of speleogenesis. For the statements on speleogenesis it is very important to know from where the water flows or has flown into the cave. Of course this point of view could be important for other branches too, in such case the sediments give lateral results being very important for the practice.

More in detail I quote some examples from karst of Dolenjska, where the distances and flows are in general smaller and at the same time the connections less researched. To the question from where comes the water appearing in Podpeska jama on Dobrepolje the most different answers are cited in the literature (see the biliography in Kranjc 1981b; Kranjc 1982c). In the riverbed of cave brook Bic there is great rate of coarse well rounded chert pebbles (Kranjce 1981a). The only known source of such material lies in the hinterland of Paleozoic rocks of Ortnek horst where there are the layers of chert conglomerate with coarse pebbles (Ramovs & Kochansky-Devide 1965). From this area towards the Mala gora foot three bigger brooks are following, the biggest among them is Locica transporting corresponding chert gravel. Thus the origin of chert pebbles in Podpeska jama lies in this area and so is confirmed the connection between brook Locica behind Dvorska vas with the brook in Podpeska jama on the other side of Mala gora ridge.

On northwestward side of Ribniska Mala goralies several times mentioned dolomitic karst in the region of Lasce, well bordered from the near geomorphological units. On south and southwestward side it is bordered by non-carbonate rocks from where several brooks are following, sinking in the karst underground on the contact with carbonates. The biggest among them is brook Cereja sinking close to Velike Lasce. Water reappears in several karst springs, the biggest among them is spring Peci on the western border of Lasce karst and on the northern side there is small spring Bec. Approximatively in the middle of these two springs and ponors for some meters only appears on the surface water in doline Zajcjak.

From sediments point of view the spring Bec can be immediately eliminated because it does not transport sediments, neither allochthonous nor autochthonous. But the spring Peci draws the attention bringing sand from the underground from the first sight of allochthonous origin. The detailed review has shown that the gravel is annexed too, composed mostly by sandstone and shale pebbles. Such or similar pebbles I've found in the riverbed in the bottom of doline zajcjak. Itremained only to find the suitable sinking stream transporting such material that could be the origin for this gravel. Firstly I've overlooked Cereja as the biggest flow but it does not transport the gravel neither to ponors. But Predvratnica, sinking into cave Vratnica, does. The fact that if flows in big cave for some distance does not disappear in bottom ponors or in alluvium as Cereja supports the supposition that it transports the gravel through the underground. In the riverbed and in the cave I've found the debris from the same rocks as appear in Zajcjak and Peci. That could be enough for proof. Predvratnica transports into cave Vratnica sandstone and shale debris and sand and gravel of the same material appears in Zajcjak and in spring Peci. By additional study the supposition was confirmed. Morphometrical properties (size,roundness and flatnessin particular)confirm that the debris had originated in the Predvratnica ponor, the doline Zajcjak and Peci are genetically connected, so it is the same sediment deposited on different places along the underground flow and therefore suitably transformed. It is not only the same petrographic composition, and not only thefact that the sediment in spring contains the gravel from the same material brought by the sinking stream into the underground, but the gravel properties confirm that they are genetically directly connected. These properties leave the possibility that the gravel in Zajcjak is transformed debris from Vratnica, andgravel in Peci is transformed sediment from Zajcjak, inversely is not possible. On the base of sediments it is not possible to ascertain that in Peci appearsonly the water from Predvratnica but that the water from Predvratnica surely appears in spring Peci too.

Later we've traced the water from Predvratnica brook and it was confirmed that it is flowing as was previously shown by the sediments: from the cave Vratnica through the doline Zajcjak into spring Puske Peci. According to the time used by tracer for this way and according to the concentration curve it was shown that the underground flow is relatively branch work and winding (Kogovsek & Kranjc 1987).


(extracted from Andrej Kranjc, 1989)




The fact that the underground flow transports the sediments, f.e. gravel is very important for speleogenesis and thus the data about gravel existence are substantial and important argument from speleogenetical point of view.

We know that the gravel could be transported by the flow with velocity bigger than is critical velocity for a particular sediment. The pebbles with diameter of 2 mm (the finest gravel) is transported by the flow with velocity over 0.2 ms-2 (Hjulstrom 1935; White & White 1968). Such flow is turbulent requiring beside the defined minimal velocity defined minimal channel dimensions too. It could be either gravitation or flow under pressure but in any case the gravel presence indicates the syngenetical type of channel (Renault 1967-69) where the erosion contributes to channel formation. Until fresh pebbles are present in the riverbed, until the water transports the pebbles, it is a proof of erosion, mechanical weathering respectively of the underground. If the carbonate pebbles are recent it is the best shown by their surface - if it is smooth and polished, shining even it prooves that they are "fresh" and the water transforms them still.

Nice example was found in Babja jama mear Most Soci where both types of carbonate pebbles are represented. the recent and subfossil ones (Kranjc 1982). The pebbles from the layer deposited some 200 years ago, from this time water did not transport neither move them. have in the former smooth and rounded surface (mean roundness index  is 192) cut up to 3mm deep and to some mm wide "channels", the corrosion result in the time until the water flow stopped to transform them(Kranjc 1988). About corrosion tranformation of carbonate debris and gravel in fine grained sediment on the surface we have detailed studies of Radinja(1967.1968).

The water flow could transport the gravel along the whole system,, through the whole karst mass  respectively. from Ponor to spring.,or in particular system parts or in particular part of one cave only. On this base the quick, direct through flow system, channels respectively can be distinguished from the others in the sense of Bakalowicz(1979) statements.

                                    (Extracted from a paper by Andrej Kranjc, 1989)



In carbonate karst basins, adequate denudation rates are difficult to estimate when there are allogenic inputs. It is shown that the common practice of correcting for allogenic input by Corbel's modified formula leads to erro-neous results unless the karst ratio is very high.

Mass balance of solutes in drainage basins is currently used to estimate area-specific rates of denudation in carbonate karst terrains. The observed regional differences in rates have, apart from obvious lithologic variation, been primarily ascribed to variation of runoff, secondarily to Pco2 effects. However, the determination of denudation rates in karst involves both conceptual and practical problems.

Conclusions of the study of samples from Svartisen, North Norway are as follows:

1, Provided that the assumptions of spatial uniformity (runoff and lithology) can be satisfied, the apparent solutional denudation rate of an allogenic karst basin may be expressed by a linear mixing model, equation 3.

2, Autogenic and allogenic denudation rates may then be deduced from twoor more separate determinations of apparent denudation rates in subbasins having different autogenic/allogenic area ratios.

3, In the Glomdal basin, Northern Norway, the autogenic chemical denuda-tion rate was determined to be 32.5 10.2 mm ka-1. The allogenic contribution converts to 8.2 1.5mm ka-1.

4, This relatively low rate contrasts with results of earlier regional studies (i.e. Smith and Atkinson, 1976), but is closely comparable to other resent estimates in the region (Hellden,1974; Lauritzen, 1984) and with direct measurements of surface corrosion rates. 42-72 per cent of the total chemical denudation appears to occur on karst surfaces. Only 26-58 percent of the total corrosion potential as given by the environmental Pco2, is utilized in dissolution. This is tentatively explained by the low residence time for the groundwater which does not allow the system to attain chemical equilibrium before it leaves the karst.


(extracted from a paper by Stein-Erik Lauritzen, 1990)




The karren is the vastest feature in the carbonate area in South Mengzi, Yunnan Province. The karrens distinguishably developed on the surface with the different slope, of limestone, dolomitic limestone and dolomite. Under the different condition of lithology microgeomorphology and micro-geologic structures, the scale, forms and the development of karrens are great variations, and they are specially controlled by the micro geo-structures and geomorphology.



The lithology is the prerequisite condition for the karren development and also the win factor to control the feature of karren. During the deformation of dolomite, there were a lot of X fractures originated, the edges of opening would be dissolved at first. The X karrens with 1-10mm wide, 1-5mm deep and several tens cm long generally developed on the dolomite surface. Their cross-section took 'V' shape, while, the large scale of karren commonly developed on the pure and thick limestone surface. The dimensions greatly fall in the range of 1-50 cm wide, 0.1 cm-1 m deep, and tens centimetres to several metres long and they shaped the 'U' form of the cross-sections. The parallel or dentritic karren systems are usually produced.



The slope of limestone surface generally affects the structure, geometry and development of karrens. On the gentle surface of pure and thick limestone, the solutional, pots 55-160mm long, 26-65mm deep and 24-47mm wide are developed near Shalima, Their floors with lichen and other organic materials are quite flate. The solutional channels are developed very well on the surface sloped 10-50 . When the slope of the surface steep than 60 , the parallel channels 40 mm wide and 4 mm deep are originated. Where the limestone is very pure and massive (some more than 30 a thick), the vertical solutional trough or balabash-shaped are perfectly developed in some places such as in Shalima stone forest.



The micro-fractures developed during the deformation of limestone is visiblly controlling the development and evolution of karrens. Near Shalima the karrens about 1-10mm wide and 10mm deep are well developed on the white grey pure limestone. It is evident that all the solutional gully produced along the micro-fractures, the larger gully forms in the zone of dense fractures, and the cross-section Is controlled by the dipping and density of micro-fracture.

(extracted from a paper by Song Linhua)





Stratigraphic-paleogeographic research shows that there are three periods of paleokarst development in the Eastern Hubei region. The first period corresponds to the end of Sinian and the second the end of Carboniferous. The paleokarst formed in these two periods belongs to the seashore type and has a large distribution area but a shallow development depth.

The third period, the Indo-Chinese-Yanshanian period, is the chief period of paleokarst development in the region. The Indo-Chinese movement (T3/T2 and J1/T3) causes a widespread regression in Southern China and the beginning of continental sedimentation. The strata folding and Crust uplift make pre-middle Triassic carbonate rocks exposed to the atmosphere and the result is the development of continental type paleokarst. The paleokarst of this period has the following characteristics.

1. The paleokarst has a multistage and impulsive development history, which is controlled by the alternation of humid-arid climates. The four humid climate time spans--late Triassic early Jurassic, early Cretaceous, early Paleocene and early Eocene correspond to the strongest paleokarst development in the region.

2. The paleokarst formation is located in the Jurassic and Cretaceous-late Tertiary river-lake basins and the paleoerosional areas near them.

3. The carbonate strata involved in paleokarst development include the Sinian, Cambrian, Ordovicion, Carboniferous, Permian and the lower Triassic. The paleokarst is most strongly developed in the upper part of the Daye Group (T1) which contains gypsum-anhydrite layers.

The Eastern Hubei region is one of the most important polymetallic ore belts in China. Close relationship between paleokarst and ore genesis is as foll-ows.

1). Almost all the ore deposits are also formed in the Indo-Chinese-Yanshanian period. The multitage magma intrusions and strong tectonic movements in the Yanshanian period favour the development of hydrothermal karst and the inv-olvement of the latter in ore-forming processes.

2). Isotopic and fluid inclusion data indicate that groundwater takes part in the formation of ore-forming solution and the migration and concentration of ore elements.

3). Paleokarst breccias have been identified in many ore deposits, including pre-ore and during-ore (recrystallized, skarnized and mineralized) as well as post-ore paleokarst breccias. In some ore fields, paleokarst breccia belt stretches nearly a thousand meters in length and several hundred meters in width along the margins of orebodies. In some ore deposits, as many as five stages paleokarst sediments are found.

The role of paleokarst in ore genesis has 4 aspects. (1).The paleokarst formation is the favorable hostrock for hydrothermal alteration.

(2).The paleokarst formation contains some ore elements. Chemical analysis data show the contents of Fe and Mn in the paleokarst formation are 5-20 times higher than its carbonate hostrock, and the Cu, Pb and Zn contents about 10 times higher. It can be inferred that the paleokarst formation is an important source rock for ore genesis.

(3).The place where the paleokarst is developed is an important flow channel for ore-forming fluid and a beneficial segment for water-rock interaction to form ore deposits. The karst water is an important agent for dissolving, transporting and enriching ore elements.

(4).The paleokarst cave-fracture system provides an ideal place for magmatic emplacement.

(extracted from a paper by Wang Liangchen and Wang Yanxin, 1990)





At the south terminal of Taihang Mountain in north Henan and southeast Shanxi, more than 30 caves have been investigated, which are mainly in the middle Ordovician vackestone and packstone and the middle Cambrian grainstone, and distribute concentratedly in the southern flank of Liushukou--Duohuo anticline and its southwest plunging end. Caves outcrop mainly on both sides of river valley and dry valley with elevation all higher than local water level of present river.

Caves generated mainly in saturation zone, but larger caves generated mainly in shallow saturation zone under high hydrodynamic Pressure. At later period they were under water level stage. Simll number are early Period water level caves; very few are continuously developing at present in aeration zone condition.

Since Cenozoic, main Pattern of tectonic movement in north China has been synchronous oscillatory uplift, appearing as the alternation of denudational stage and depositional stage. Ancient Physiographic surface on which old depression, dolines, sinkholes, peak forests etc. remain are well preserved in Yangcheng County and other places (Qian Xuefu, 1984).

Tab.1 The elevation of denudational surfaces and cave layers of the studied area and neighbouring regions


Geological   Denudation stage    Elevation of the                Elevation of
   epoch                                        denudation surface (m)    cave layers
                                                                                            in the area
--------------------- ---------------------------------------------------------   
                   Shanxi     Huabei        Shanxi        the area
----------- --------------------- ----------- ----------- --------------------
Q2           Huangshui Huangshui
-------------------------------------------------- 400-600------ 450
Q1              Fenhe        Fenhe         500-650
----------- -------------------------------------- 700-900------ 750-850
N2           Tangxian                     850-1350
----------- ----------- Tangxian ----------- ----1000-1400---1100-1200
Nl             Taihang                     1600-1700
----------- ---------- Luliang ---------------------------------------------
-- - ------------------------------------------------------------------------
E2             Beitai                       2200-3058

According to the statistics on cave elevations, there are roughly three cave layers in the area: 1000-1200 m, 750-850 m and about 450 m,while the most intensive karstification was happened in the layer of 750-850 m. The three cave layers roughly correspond to three denudational surfaces of the area, i.e. 1000-1400 m. 700-900 m and 400-600m. corresponding to denudational stages of Shanxi  plateau as well as the other regions of north China (Tab.1).

It can be seen from the table that the earlist cave formation stage in the area roughly corresponds to Tangxian epoch of Pliocene, and the major, cave development stage my be Fenhe epoch of Pleistocene.

The general trend of climate change in north China has been toward temperate cool and arid environment since late Tertiary, the temperature may even such lover than today at that time,so brought about peculiar karstic landscape and huge thickness of loess deposit in the area. However, the changes of the climate is provided with cycles, i.e. there were some warm-humid and rainy periods. The repeatedly overlapping of loess and paleosoil represent repeated fluctuation of deposit in dry state and rain stage (Liu Dongsheng et al,1985); different cave layers represent polyphase of karst development, recording similarly polycycle of the climatic change. Three cave layers of the area all enjoy major horizontal caves. with corrosional forms formed in saturation zone, such as ceiling Potholes, scallops etc. well decorated. Especially the ceiling potholes we the most typical. and their size, density, corrosional depth, vertical karren etc. are very much like those seen in South China. It can be seen that climate environment at that time was quite warm-humid, rich in rain- fall, remarkably different from that of present in the area. It is clear that after late Tertiary, there were several stages of warm-humid and rainy climate, when subterranean river might develop and water circulation was quite strong.

(extracted from a paper by Weng Jintao, Tao Youliang & Tong Changshui)




1, Glacial marginal karst

It can be seen that some geomorphic features which are similar to "stoneforest", "peak forest" and uneven forms of stone back, stone wall, individual peak and stone column etc. occur over certain altitude in the plateau every-where. Their combination type consists of rock waste slope and rock waste cones, shaped a unique geomorphic type. It is evident that the "stone forest" and "peak forest" are in the fact production of contemporary glacial marginal climate. Because these forms don't possess the characteristics of latitude zoning, it mainly distributes in the frigid zones of high mountain especiallyfrom Baima snow mountain (west Deqin county of Yunnan, 28 30' N) to Aqige-kule lake of Kunlun Mt. (37 N). The lower distributive limit of stone forest and peak forest is 4000-4500 m in altitude that only in parts of an area can reduce to 3700 m, which is the same as the height of isotherm with annual mean temperature of 2 -3 . The daily variation is very large in this area, whether in winter or summer, fine day's temperature reaches 50 to 70 in general and difference of violent daily temperature make rock surface rapidly breaking up and produced rock waste which moving toward to the foot of the moutain under the action of gravity and freeze melt landslide and formed rock waste slope and rock waste cone. It follows that, the formation mechanism and feature complex of this kind of limestone forms is different in essence from the stone forest and peak forest in South China. Their formation and occurrence are related to glacial climate, so it is called glacial marginal karst.

2, Environment of buried karst and karst planation surface

Red soil weathering crust with various thickness often develops on lime-stone planation surface of the Hengduan mountain area, e.g. there are red soil weathering crust of a few tens cm to more than 3 m thick on the planation surface (ca 3000 m a.s.l.) of the eastern part of the Baihanchang, Yunnan and under the weathering crust buried stone teeth or stone forest. The stone teeth and stone forest usually possess any slight round small solutional cavities and dissolutional pores. After the red soil weathering crust was eroded, stone teeth and stone forest exposed on the surface. According to those seen in the Yanyuan basin (Sichuan Province), the red soil weathering crust are overlain by the Yanyuan lakebog sediment of Upper Pliocene. So the karst planation surface was formed during Miocene of Eogene.

(extracted from Wang Fubao's report, 1990)




Friars Hole is a narrow dry valley in Greenbrier and Pocahontas Counties, West Virginia. Along the 11 km length of its floor are a series of inliers of Union Limestone of the Mississippian Greenbrier Group. However, almost all of the catchment (some 97%) is underlain by siliclastic strata of the Pocono and Pottsville Groups. These all drain to the limestone inliers, beneath which some 70 km of cave passages have been explored.

The streams which flow off the impermeable strata sink soon after reaching the limestone. Occasionally, a surface stream flows into an open cave en-trance, such as Snedegar North Entrance or Bruffey Creek Cave. In these cases, stream-borne sediments are carried away through cave passages. However, in most instance the streams sink in their alluvial beds and the water drains away into underlying cave passages. Consequently, only fine sediments are carried away, while the coarser fractions remain on the surface to form a mantle of alluvium which may reach 10 m or more in thickness. This alluvium has a mean grain size of 15-20 mm, with over 90% being coarser than 2 mm.

Sinkhole development has taken place along Friars Hole on the inliers of Union Limestone. In many instances it is possible to follow cave passages to terminations in blockages of sandstone cobbles and boulders lying below surface depressions. However, many former stream sinkpoints have no surface expression. For example, at the Friars Hole Cave inlier, 18 active sinkpoints have been identified from within the cave, yet only four of these are visible on the surface as depressions or stream sinks. There are also 29 inactive sinkpoints, recognized in the cave as abandoned inlet passages. None of theseis now identifiable on the surface. Sinkpoints are often underlain by vertical shafts, with typical widths ranging from a few centimetres to several metres. All these shafts became blocked by alluvium before the surface flood-plain was abandoned. Subsequently, these blocked shafts formed favourable sites for sinkhole development.

Two populations of sinkholes may be discerned at Friars Hole. Suffosion sinkholes develop relatively rapidly along the margins of the valley where there is some surface runoff down the valley slopes, and in some cases a sink-ing stream may be present. Typically, the sinkholes are 5-10 m deep, are steep-sided (20 -35 ), and are irregularly spaced along the valley margin at intervals of 20-300 m.

In contrast, solution sinkholes have developed away from the valley sides, are broader, and rarely have slopes greater than 10 . They too have an irregular spatial distribution. The shallow sides reflect the predominance of solution on the bedrock by percolation water.

The evolution of Friars Hole has been dominated by Hills Creek, which has a catchment of 50 km2. There has been a 4 km retreat of the main sinkpoint of the creek since it was first captured underground. The alluviated flood-plains abandoned as a result of this retreat have been progressively dissected by the development of sinkholes. Nearly fifty uranium/thorium dates have been obtained from speleothems from Friars Hole System, which is the principal cave underlying the valley. These show that almost all the cave is >350,000 years old. The analysis of 234U/238U ratios has indicated that three of the sampled speleothems are probably >1.24 Ma old. Two of these three speleothems have reversed magnetic polarity, and thus predate the Brunhes Chron (apparoximately >730,000years B.P.). The third of these speleothems has normal polarity and is thus probably >1.67 Ma old, that date representing the termination of the previous normal-polarity sub-chron.

It has been possible to correlate the retreat of the sink-point of Hills Creek with the development of the cave passages of Friars Hole System. It is thus possible to use the speleothem age results to date the antiquity of the abandoned alluviated flood-plains and gain an insight into the rate of sinkhole development in the area. Results suggest that it has taken more than two million years for the destruction of the alluvial flood-plains at Friars Holeto take place, while the oldest alluviated limestone surface at Friars Hole may be more than four million years old.

The spatial distribution of sinkholes at Friars Hole is thus controlled bytwo criteria. The availability of surface runoff dictates the type of sinkhole which will develop. Slow-developing shallow solution sinkholes develop where only percolation water is present, while fast-developing, steep-sided suffosion sinkholes result when surface runoff from impermeable rocks is avail-able. Secondly, the maturity of sinkhole development in the valley floor is dependent upon the time that has elapsed since the alluvial flood-plain was last active.

(extracted from a paper by S.R.H. Worthington and D.C. Ford, 1984)





In 1989 the precipitation measurements and analyses have shown that inPostojna "acid" precipitations were fallen. The mean value of precipitation pH evidences that in 1988 the precipitations were still more "acid" (mean annual pH value=4.56, min 3.2, max 6.8) than in 1987, but the sulfates (mean6.2, min 0.0, max 28), nitrates (mean 1.43, min 0.0, max 6.0), clorides (mean0.63, mean 0.0, max 1.7) concentrations were lower.

In Postojnska jama the trickles of supersaturated percolating water, de-positing flowstone, prevail. On some places in Pisani rov,in Lepe jame (Rdecaddvorana) and at the end of Carobni vrt we've observed visible signs of flow-stone and rock chemical denudation.

On these places the values of SEC, carbonate and total hardness are the lowest from all the analysed samples in Postojnska jama. Mean annual valueof carbonate hardness oscillates between 104 and 105 mg CaCO3 l-1, 2.07 to 2.1 respectively, the total hardness between 116 and 121 mg CaCO3 l-1, 2.32 to 2.41 respectively mekv l-1. The sulfates, nitrates and chlorides content in these waters is slightly higher than in precipitations.

Supersaturated percolating waters situated in immediate vicinity of aggressive trickles, have higher carbonate hardness, and some of them higher sulfates content at correspondingly higher total hardness. Supersaturated percolating water forming the speleothem Dvojcka in Pisani rov contains extremelymore sulfates than the aggressive water. Also the trickles, depositing flow-stone on Velika goar, are according to sulfates content equal to aggressive water in Lepe jame. The lowest carbonate and at the same time the highest total hardness and higher concentrations of nitrates and sulfates reaches the aggressive water from the chimney in Lepe jame.

From previous measurements we can infer that the corrosion effects of percolating water in Pisani rov and in Lepe jame cound not be the result of increased concentrations of sulfates, nitrates and chlorides as these anions are present in the same amount in supersaturated percolating water which deposits flowstone. It seems that the reason for corrosion lies in periodically undersaturated water with carbonates when it comes to the cave, which is probably due to mode of percolation.

The increased content of sulfates, nitrates and chlorides mostly in the precipitations, combined by vegetation and lithology influences could be reflected in the composition of percolating water in the cave, but on its way through the cave roof the water is neutralized, causing augmented corrosion in the cave roof itself and this is not the reason for corrosion in the cave.

(extracted from a paper by Janja Kogovsek & Andrej Kranjc, 1989)




Niangziguan spring is the biggest karst spring in North China, with average discharge about 12.6 m3/s. The western part of the spring watershed is buried karst, covered by sandstone, shale and coal strata. Whereas its eastern part is bare karst, composed of limestone, dolomite and gypsum.

The recharge to the spring is from infiltration of meteoric waters and surface water in the area, and ground water flows eastward from the north, west and south of the spring watershed, which appears near Niangziguan in the form of spring. The spring consists of ten primary discharge points.

Disruption of coal strata during mining accelerates the oxidation of pyrite which is associated with coal, by exposing greater surface areas of the reactive mineral to the weathering effects of the atmosphere, hydrosphere, and biosphere. Acidic water in a coal mine in the Niangziguan spring watershed is related to this process and is characterized by low pH (min. 2.52), high sulfate (max. 4100 mg/l),iron (max. 257 mg/l),and hardness (max. 2274.45 mg/l). However, it is possible that the kind of acidic coal mine water is subsequently neutralized because of the dissolution of calcite presenting in the coal strata. The hydrochemical characters of the produced alkaline coal mine water are high pH (max. 8.18), sulfate (max. 542 mg/l), and hardness (max. 1183.56 mg/l) and lower iron (min. 0.12 mg/l).

Experiments were conducted to further understand the mechanism of the formation of both acidic and alkaline waters in the spring watershed coal mines by modelling natural conditions with simplification, and the results of the experiments have shown that they are basically successful.

The high sulfate (max. 223.82 mg/l or 4.66 meq/l) and hardness (max. 435.53 mg/l) of the Niangziguan spring are related to the effect of the neutralized acid coal mine water as demonstrated by analysis of water temperature, total dissolved solid, Q-mode cluster analysis and sulfur isotopes (Tab.1). The influence of the neutralized acid coal mine water on the pH and iron in the spring is not obvious because of the neutralization effect of calcite in aquifer, the buffer effect of ground water and the precipitation of iron. Some measures to prevent the formation of acid and alkaline coal mine waters are presented.


TAB.1 The effect of coal mine waters of variable pH on the SO42- and  hardness of the Niangziguan spring groups.


Spring group No.

SO4-- in water


SO4-- from coal mine water

Total Hard. in water(mg/l)

Total Hard. from coal mine water


































































Notes: 1. Spring group No.1 includes three individual discharge points Niangziguan spring, namely Xiwuzhuai sp., Chengjia sp. and Chengxi sp., and spring group No.2 includes the other seven individual discharge points of Niangziguan spring, namely Podi sp., Shibanmo sp., Wulong sp., Gun sp., Qiaojun sp., Shuiliangdong sp. and Weizeguan sp. (Fig.2).
            2. Results in 1994 and 2007 are those forecasted.

(Extracted from a paper by Liu Zaihua, Yuan Daoxian and Shen Zhaoli, 1988)





Cave gravel, recent and subrecent cave sediments respectively are the indicator of particular changes going on the karst surface, frequently caused by man. The sediment are material proof about the history of geomophological events on the earth surface and the advantage of the caves is that the cave sediments have more change to remain untouched than those on the surface.

As an example I quote the results of sediments analyses from two caves:Rupa in Zaloka (Ljubija river basin under Smrekovec) and Babja jama near Socariver under Most na Soci.

In the both caves the sediments profiles are preserved showing that it was, on particular places at least, the cave passage filled up to the roof or almost to the roof. The sediments within such deposit are in both cases composed by alternating layers of fine-grained and coarse-grained material. In Babja jama the profile is composed by four gravel layers mixed with sand and by four fine sand with silt. In Rupa the alternating of fine-grained and coarse-grained layers is not so regular, among ten layers two are composed mostly by gravel only, and in six of them sand fraction prevails, one layer is laminated loam, one is built by mixture of rubble, sand and organic matters. In both cases there is among the fine-grained fraction a lot of organic matter, including wood and char coal remains which has rendered possible the absolute chronology of these layers.

In both cases the evolution was similar -- some hundred years ago the water started to fill up the cave and in short period probably, from some 10to about 150 years, the cave passages were filled up to the roof almost and sediments were deposited in 2-3 m thick layer. It is not suggested that the sedimentation was continous, more probably not, at least in the case of gravel layers were short-termed, periodical processes when the water had deposited relatively big quantities of sediments. About 150 years ago when the uppermost layer was already deposited (in both cases fine-grained) the sedimentation was interrupted. Water strated to downcut in previously deposited sediments, cut them and in great extent washed them off. The proof of the previous sediments deposition are profiles presented in leeward places. In both caves the actual flow transports and partly deposits gravel and little sand only. In Babja it is clearly seen that this is the autoregulation of the profile, the equilibrium of deposition and eroding off the gravel respectively.

Two questions arise -- the reason of quick sedimentation, filling up of the cave passage respectively and the reason of interruption of this sedimentation, erosion of former deposited sediments or excavation of the cave. Bigquantities of wood, in particular charcoal show that the reason of big sedimentation changes in Rupa under Smrekovec could be the remains of burning matter of farming and intensive production of charcoal, which are attested by literature. By cutting and burning the wood on slopes, on igneous rocks of Smrekovec slopes with a lot of weathered material people could cause increase of the runoff of precipitation water and thus increased sheet erosion from steep slopes which were uncovered because of fire or cut cut trees and thus maybe increase the sedimentation in ponor cave Rupa. Stop of this activity and overgrow of vegetation on the slopes the water did not wash off so much debris and started to excavate the sediments in the cave. The alternation of

fine-grained and coarse grained sediments could show bigger changes guring thesedimentation as the granulometry of deposits shows the flow velocity about 0.1 and later 1.5 m/s. It shows periodical filling up the channel, stagnating water depositing fine-grained material (silt-sand fraction) more quickflow transported and depositing gravel sediments.

It was possible to observe directly the influence of human activity on karst surface to loosing of fine-grained material and to washing off through the vertical channels into underground with parallel observation of vertical water percolation through the roof of Pivka jama. Within the frame of the Institute's research work "Study of vertical water percolation in karst" I've periodically analyzed the suspension load transported by trickles flowing through the roofs of karst caves. In observed trickles in Planinska jama and Pivka jama the usual quantities (I cannot speak about the mean value because of too small number of observations) of suspension load were about 20 g/m3 od water, and in extreme cases -- the spring water pulse -- about 100 g/ m3. Maximal values were reached in spring 1978 255 g/ m3 and in autumn 1980 55 g/ m3, both in Planinska jama (Kogovsek 1981, 141; 1982a, 120-121). In Pivka jama they were usually lower. In Pivka jama the observed trickles percolate through 40 m thick cave roof of limestone. In spring 1985 above Pivka jama huge building works were going on -- building of new restaurant and reception, new lavaratories and bungalows in camping. Just above the point where the observed trickles appear in the cave some blastings were done. Soon after in the time of spring snow melting (7 March 1985) we've sampled nine trickles in Pivka jama. Apparently the water was turbid which was previously never the case. It is interesting that 5 trickles contained more or less usual suspension load quantities, under 75 g/ m3. Other four trickles had ectremely big quantities, between 270 and 753 g/ m3. In Timavo springs the biggest ever measured suspension load quantity was 2000 g/ m3, and during my observations the biggest quantities were 1095 g/ m3 in Rupa under Smrekovec and in Lokva, in ponor to Jama 970 g/ m3. In all the cases these are either brooks flowing directly from impermeable surface or springs with huge hinterland in the same rocks. And the quantities of suspension load in trickles in Pivka jama could be compared to those, the biggest quantities in superficial flows.

Obviously the suspension outflow in these trickles is very small because these are small trickles with discharge during the observation between 0.001and 0.180 l/h. During heavy rain the discharge increase, e.g. the discharge of trickle was at the beginning of observation 0.06 l/h and later it increased to 1.2 l/h, it means for 20 times. Corresponding to small discharge the suspension outflow is small, between 0.001 g/d and 1.68 g/d. But I have to stress that some trickles, in Planinska jama, e.g. have essentially bigger discharges, to some 100 l/min, it means from some 10,000 to over 10 millions times bigger.

I am sure that these extremely big quantities of suspension load in some trickles in Pivka jama in spring 1985 are due to human impact on karst surface. Two facts are usually overlooked or not considered are. First, verticalchannels, or better vertical fissures and joints in the rock present the way where big quantities of suspension load could be washed through quickly during suitable conditions which could have all sorts of negative results from re-moving the soil from the surface to pollution or even poisoning of karst water. Second, all sorts of human activities on the karst surface not necessarily "dirty" ones can cause or essentially change the activities in the underground. If I transfer the statements from Pivka jama to the new meliorations on karst surface near Sezana -- blasting and extracting of big rocks with bulldosers, digging the soil out of dolines and shooting it to freshly got flat surfaces - the question arises if it is not a Sisyphean labour. By extraction and blasting of rocks the possibility of fine particles of soil erosion into underground increases, by shooting of fresh soil a new material which could be washed off into the underground is added. It is not necessary but it would be wise to think about such possibilities before the works have started and to foresee the suitable professional attend at such works which could give us a look into processes changes.

(extracted from a paper by Andrej Kranjc, 1989)




Terrestrial freshwater calcareous deposits, including tufa and transvertine, are calcareous deposits in karst regions under certain physicochemical and biological conditions. They may have various kinds of textures and structures as well as impurities or contaminates. However, tufa is thought to be an irregular soft and porous deposit of amorphous CaCO3 with bended layers.

Samples observed in this research were collected from Jiuzhaigou District and Huanglongshi District in NW Sichuan of China and Plitvice and Komatevra of Yugoslavia.

Bio-effects on calcareous deposits can be divided into active and passive(Zhang,1989).Active effects refer to the processes directly related to metabolic biochemistry of organisms, while passive effects with indirect relation. However, active effect of Bio-effect on calcareous deposits may change the micro-environment of water condition and then to cause the precipitate of calcite. Algae may contributes to calcite deposition in several ways like binding, trapping, the calcification of its filaments and nucluation (Schneider, 1977;

Schneider et al., 1983; Pentecost and Bauld, 1988; Pentecost and Lord, 1988).Some of those processes are active, like physiological calcification of algal filament. Passive bio-effects on tufa deposition in Jiuzhaigou have researched

with relation to the dynamics and nucluation of crystal growth (Zhou, 1989).Diatoms can effect calcite encruststion on moss (Emeis at al., 1987) and other

objects in the stream. Such phenomenon has been found both in Plitvice and inJiuzhaigou with SEM observation. This is a complex passive effected process ofboth diatom and moss.

(extracted from paper by Bao Haosheng and ZhangJie)




Biokarst is defined as a landform produced directly by biological corrosion and/or deposition of calcium carbonate. It can be divided into three simple types: erosional forms, depositional forms and mixed erosional-depositional forms. So far, biokarst investigation has been spreaded from coast to inland, salt water to fresh water and surface to cave.

There are very large karst areas and varied karst forms In China, but investigation on biokarst is only a little. With the purpose to develop biokarst research in China, the authors have collected many samples in some place of Guangxi and examined SEM.

Some samples of light-oriented pinnacles from a Tiandeng cave and a lichen covered limestone from Guilin suburb show very distinct honeycomb surface texture, with numerous holes. The holes or pores can morphologically be grouped into two types:

1). small diameter pores with circle and smooth aperture;
2). large diameter pores with irregular aperture and crake around the bottom of some pores.There are some filaments (my be algae) on the surface or in the holes of some samples from the both location, which imply relationship between filaments and formation of the honeycomb texture.

In a sample of light-oriented coral-like stalagmite from Tiandeng, some filamentous algae (?) encrusted and embedded in the rock matrix. In a sample of light oriented pinnacle from a Tiandeng cave, spheroidal and filamentous algae show very distinct close relationship with the rock matrix, in which the better is coming out from a curve "canal" may be done by boring algae.

Although algae taxa is difficult to determine in fossil-type, the Preservation state of the filaments show that they are in situ position and suggest some information of biogenesis of karst corrosion and deposition.

(extracted from a paper by Cao Jianhua and Wang Fuxing)




Calcite deposits in caves, known as speleothem, can provide valuable palaeoenvironmental information. In particular, because speleothem are deposited only in air-filled caves, gaps in deposition in coastal caves record high-stands of sea-level. Here we report the use of isotope-dilution mass spectrometry to date speleothem by the uranium-series method. The speleothem dated here--a flowstone from 15m below modern sea level in a Bahamian cave--records changes in sea level over the past 280,000 years.

Isotope ratios were measured on a VG354 solid-source mass spectrometer with a 27cm radius and 90 magnet sector. Uranium and thorium analyses used a Daly multiplier detector in the analogue mode. Ions of all the isotopes of interest were measured sequentially. Counts were corrected for background (measured at mass 227.3) and for inter peak interference (measured at 0.5 mass unites above and below each peak). Background noise was extremely low and tailing was negligible. A stable ion beam could be maintained for several hours with no significant fractionation: for Th runs of 2-3h, means of each of the sets of 10 ratios deviated from the grand mean of the whole run by less than 0.2%.230Thcould be measured to 0.13% (1s ) and 234U to 0.05%. The errors were propagated through the standard dating equations. The resulting 1s uncertainty in age wasoften less than 1%.

Uranium was analysed in two stages: isotope masses 235,236 and 238 over 60 cycles and then masses 234,235 and 236 over 100 cycles. Each run was monitored

and adjusted to keep the 234U ion beam high and the Ta filament current at the optimum for U ionization, between 2.0 and 2.3 A. This single-spike, two-stage method for 234U measurment differs from the double-spike, single-stage method of Edwards et al.

Thorium was normally run in one stage over 100 cycles(masses 229,230 and 232) but if detrital Th content was high, interference from the high 232Thnecessitated a two-stage run (masses 229,232 and then 229,230). The centre-filament Re current was kept at ~5A to give a Re ion-beam current of 1.5 1011A.Each run was monitored to yield a stable and long-lived `+230!`Th beam with current ~5 1017A, while keeping the side-filament Re current between 2.2 and 2.6A.

The sample shows three prominent, and two lesser, growth layers. Each is terminated by an erosional hiatus. Upon each hiatus, except the final one, is a thin layer of goethite (FeO(OH)) flocculate. This pattern is found to be the predominant flowstone sequence to depths of -20m in caves of the Bahamas Banks.

The sequence of (1)flowstone deposition, (2)dissolution of its surface, (3)flocculate deposition and (4)resumption of flowstone deposition represents

1).calcite deposition in the vadose zone,

2).subaqueous solution, probably at the fresh/saltwater interface,

3).salt-water flocculation of iron compounds and

4).return to vadose conditions.

The hiatuses are products of sea-level rise. The flowstone immediately above each goethite layer dates sea-level fall through -15m or a little more(to allow for a shallow fresh-water lens); however, the date of each sea-level rise cannot be pinpointed so confidently because an unknown amount of flowstone is lost by dissolution during inundation and submergence of the cave.

Except for samples on the hiatuses (which probably experienced U leaching during ensuing dissolution events), all the dates are in chronological order within the error bars. Bearing in mind tectonic subsidence of the Bahamas banks(estimated at ~1m per 50kyr) and the difficulty of determining sea-level rise because of the erosion, the hiatuses indicate periods of sea level higher than-10 to -15m at >280kyr,235-230kyr, 220-212kyr, 133-110kyr,100-97kyr and <39kyr(Fig.3.). The four rather short-lived peaks, at around 230,215,125 and 100kyr Bp, are generally in agreement with the high-sea-stand chronology inferred from oxygen isotope stratigraphy of oceanic foraminiferal cores and do not support the recent suggestion that the chronology of the marine foraminiferal record is questionable.



Fig. 1 Pleistocene sea-level curve for Bahamas from speleothem dates, correlated with the oxygen isotope record. The slope on which the speleothem dates are plotted is the maximum estimated tectonic subsidence rate of 1 m per 50 kyr. The mass-spectrormetric dates from this study are shown as solid circles, alpha-counted dates from Gascoyne et al as open diamonds and those from Harmon et al as open squares.

This study has shown that isotope-dilution mass-spectrometric U-series dating of speleothem is possible to at least 400kyr BP on samples with U contentas low as 0.08 p.p.m. Speleothem with high U content ( >3 p.p.m.)should permit dating approaching 600kyr BP. Where 234U/238U ratios can be shown to have remained constant, mass spectrometry will also give higher precision age estimates beyond 600kyr. As in alpha counting, the presence of detrital Th creates the problem associated with common 230Th. For the samples analysed here, however, with 230Th/232Th activity ratios greater than 100:1, this problem was negligible.

(extracted from paper by W.X.Li, J.Lundberg, A.P.Dickin, D.C.Ford, and D.Williams, 1989)





Ling canal is one of the three famous water conservancy projects built in the Qin Dynasty. Since the First Emperor of Qin unified South China in 221B.C.,Ling canal has been digged for transportation of soldier's logistics. Ever since, this project has been constantly improved for 2200ys.The Ling Canal has become a key of water control between Yangzie River and Zhu River.

The Ling Canal consists of "big smell balance", "mouth of share plough", dams, sluice gate and bridges. After built up, its feature has undergone a great of changes. According to the record of events inscribed on a tablet, the length of "big or small balance" was 391.8m after renovation in the Ming Dynasty, but now 474m. As a matter of record inscribed on the tablet, "the mouth of share plough" was cut down in hundred meters. Now the dam of "the mouth of share plough" is 100m in length, 3.6m in height, being built by stone block with rectangle shape in 5 2 3 chi(one chi-1/3m).

A shallow pit being 71.5m at a distance from "the mouth of share plough" was digged by the Xingan hydroelectric Burea, Guangxi Province in 1988. There were found out many stone blocks and cement lied underground in depth 2-3m.Thisfact has confirmed the record on the tablet and testified that the ancient dam of "the mouth of share plough" is longer than now.

When repairing the tail of "big small balance", four pine takes have been excavated in 1979. Except one was decomposed, the other three stakes are 1.63m, 0.90m, 0.72m in length respectively. One of them is 12cm in diameter. There isa hole(12 8.5cm) being 26cm at a distance from the top of the shake. These pine stakes were very scientific measure for steady foundation of "balance" in ancientry. Based on the record of historical literature, there were three times for repairing "the big small balance" and "the mouth of share plough" on a grand scale during 600 years of the Ming-Qing Dynasties. The first time was in HongWu 29thyr(1396), towards the begin of the Ming Dynasty, the second in Kang Xi 53-54thyrs(1714-1715), the Qing Dynasty and the third in Guang Xu 14thyr., the Qing Dynasty(1988).

In accordance with the C-14 dating, the C-14 age of pine stake is 360+80yr.B.P.(T=5568). We may preliminarily conclude that these pine stakes are the things left behind by Chen Yuanlong, a governor, Guangxi, in Kangxi 53thyr,when he had repaired "the big small balance".

(extracted from a paper by Chen Xian)





1. Two results have been obtained from the study of paleomagnetism:

1) Through the past 6500 years or so, the long-term variation curve of the D and I of the paleomagnetic field in Guilin is a composite activity including different cycles of variation big and small, and the VGP transfer curve center does not coincide with the geographical axis.

2) The long-term variation curve of the relative intensity of the paleomagnetic field in Guilin over the past 6500 years, derived from the application of DRM/ARM and DRM/SIRM, is a composite activity including different cycles of variation big and small.

2, Two results have been obtained from the study of paleotemperature:

1) For the period from about 6500 years to 3000 years past, the 18O result of the stalagmite might serve as the indicator of the longterm variation of paleotemperature in Guilin. The result shows that the paleotemperature during that period does not vary very much.

2) For about 3000 years past, the 18O result could not serve as the indicator of the long-term variation of paleotemperature of that period.

3, Two results have been obtained from the study of elementary content:

1) There might have been three violent attacks of corrosion about 2870 years ago, 900 years ago and 100 years ago respectively, in which the dissoluble CaO was lost and the indissoluble SiO2, Al2O3,FeO were accumulated.

2) These three corrosion phenomena seems to have been caused by the increase of rainfall.

4, From the comprehensive study of the variation pattern with time of such paleoenvironmental factors as paleomagnetism, paleotemperature and elementary content, two conclusions have been drawn:

1) During the period from about 6500 years to about 3000 years past, there was no unusual paleoclimatic variation while over the past 3000 years or so there might have been three paleomagnetic changes characterized by the increase of rainfall.

2) There seems to be a correlation between the relative intensity of paleomagnetic field, paleotemperature and paleoclimate, i.e. when the relative intensity of paleomagnetic field varies slightly,paleotemperature also varies slightly, and when the relative intensity of the paleomagnetic field varies greatly, there would be unusual paleoclimatic variation.

(extracted from Liu Yuyan's reports, 1990)




As part of a feasibility study to see whether stalagmites could provide useful records of secular variation, nine oriented stalagmites were collected from the states of Chiapas and San Luis Potosi, Mexico. Of these, six have yielded measurable natural remanent magnetizations (NRMs) throughout their length. The cleaned magnetizations of one of these samples were shown unequivocally to have recorded the ambient field, and there is no reason to believe that this is not true of the primary magnetizations of other samples. The sequences of paleofield directions, up the samples' length, have varying degrees of resolution and serial correlation, depending on the rate of field change as averaged by the stalagmite growth rate and the thickness of the measured specimens. Ages and growth rates were estimated by the U-Th method. The main problem in stalagmite paleomagnetic analysis is a weak NRM, although this may be avoided by judicious choice of the sample. Samples which possess significant viscous components may be cleaned by alternating field or thermally if there is sufficient magnetic material. Unlike many sediments, stalagmites do not appear to suffer from depositional error problems. Dating problems may include low initial U-content, yielding ages with large errors, and the presence of allogenic 230Th in detritus, which causes older apparent ages. The dating limit is about 350 ka. It is usually not possible to obtain long records comparable with those of most lake sediments, and there area esthetic reasons for not spoiling caves adorded with stalagmites. The method is seen to be complementary to the use of sedimentary sequences to study paleosecular variation. Studies of the rock magnetism of stalagmites are presented to suggest the mineral carriers of the magnetization and the origins of the natural remanence.

Interest in long SV records has increased in the last few years because of the mixed claims for the duration of westward drift, lifetime of anomalies and the relative magnitudes of standing and drifting fields. Observatory data from as far back as 1700 have been used to infer the configuration of the field at the core-mantle boundary and its SV, in turn leading to tests of the forzen flux hypothesis. Stalagmites have the potential to provide SV records, error free relative to sediments, which allows direct comparison with observatory data at the site. DAS2 is an example. It is doubtful whether stalagmites will be able to provide the length of record necessary for full time-spectral analysis, but the U-Th dating method of stalagmites allows for the dating of possible excursions or events with estimation of their duration.

Pure white stalagmites seldom carry a measurable NRM and even some coloured ones, such as DAS1 and DAS2, require near-clinical laboratory conditions if contamination of the weak signal is to be avoided. Their signal also has to accommodate AF demagnetization to remove viscous parts of the NRM. A SQUID

magnetometer is usually a necessity, and sample-holders have to be ultrasonically cleaned and sometimes AF demagnetized. Specimens have to be cleaned periodically and stored in a dust-free environment.

On the other hand, if detritus-laden, more strongly magnetic stalagmites are used, then the problem of allogenic 230Th, causing apparently older ages, may be encountered. Both accuracy and precision are usually lost in the age correlation. Set against this are the 14C dating problems frequently en-countered with sediment records, such as low 14C and 'old' 14C (hard water) effects and problems with post-depositional DRM.

Another disadvantage of stalagmites is that long records comparable with4- or 5-m lake cores are seldom obtainable from northern continental caves. Such stalagmites are an adornment, and cave conservationists and some state laws (in the U.S.A) prohibit extraction of speleothems even for 'scientific' purposes.

More rock magnetic experiments are needed in order to determine more closely the chief carrier of the remanence (at present believed to be magnetite; and also titanomagnetite in the case of the two Mexican samples, SJLS and SJHS), and to determine if the depositional error-free NRM is, as we suspect, largely a CRM.

(Extracted from a paper by A. G. Latham, D.C. Ford, H.P. Schwarcz and T. Birchall, 1989)





The oxygen isotopic analyses of speleothem samples from the 4th cave of Peking Man Site have demonstrated that the formation of speleothem was in oxygen isotopic equilibrium with their parent seepage waters. The cross section of this speleothem can be divided into three zones by oxygen isotopic composition. The d 18O of the inner zone ranges from -6.48 to -7.43. Its formation temperature is 13.2 C and its foundational age is 78 103 yr B.P. by 230Th/234Udating. The average vaalue of d 18O of the mediate zone is -7.58 0.11, and the formation temperature is 18.6 C . Its age is 63 103 yr B.P. That suggests the paleoclimate during that period might be more moderate. The average value of d 18O of the outer zone is -6.43, its formation temperature is 9.2 C, and the age is 51 103 yr B.P. The well-known Peking Man Site is located at the Longgushan(Dragon Bone hill), Zhoukoudian, about 50 km south-west of Beijing. The geochronology of 1st. But up to now there is little reseach on stable isotope for cave deposits. Here we attempt to report some results about stable isotope and 230Th/234U dating of

speleothem from 4th cave (New cave) of Peking Man Site, and also attempt to determine its paleotemperature and paleoclimate.

The New Cave is situated on the south slope of Longgushan and was developed in Majiagou Limestone (middle Ordovicion System) and extended along the gentle stratification. It belongs to a type of horizontal cave. The cave deposits are composed of stratified sandy clay, caliche and travertine, many speleothems have been developed in different size. Fortunately, a speleothem sample was collected from the collapsed roof of this cave. It is composed of calcite, 36cm in length,14cm in bottom diameter and 3cm in top diameter.


Tab.1 Oxygen and Carbon Isotope Composition of Speleothem From 4th Cave of Peking Man Sit



Sampling layer No. Distance from centre(cm) d 18O
PDB ()
d 18C
PDB ( )
Colour of Sampling layer

sampling zone number

A 1
light grey
light grey


-7.64  -7.71
-9.33 -10.21
-10.67 -10.16  -9.37
light yellow
light yellow
light brown
light yellow
light yellow


light grey
light grey
light grey
B 11


-6.31  -7.24 -7.61 -7.46 -7.55
-7.39 -6.72  -6.05 -6.47 -6.21
-8.06   -7.24
light grey
light grey
light yellow
light yellow
light brown
light yellow
light grey
light grey
light grey
light grey
light grey

(extracted from paper by Yan Zheng, Ye Lianfang, Zhao Shusen, Liu Minglin, Liu Rongmo and  Zhao Dinghua, 1984)




Liu Huangfeng, Cave Paleomagnetism.

Sui Guojun, Karst development of the Taizibe depression.

Zhang Jie, A short comunication on Lichen Karst.

Zhou Xulun, Calcified karst feature complex in Huanglong, Sichuan.

Zhou Zhengxian, and Li Xingzhong, Influence factors or the development of peak-cluster depression in Haolan karst forest area.



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