ArticlePDF Available

Late-stage evolution of hypogene caves at Tyuya-Muyun (Kyrgyzstan): Quantitative insights from mineral deposits

October 2024
Chemical Geology 670(12):122454

DOI:10.1016/j.chemgeo.2024.122454

License
CC BY 4.0

Authors:

Gabriella Koltai

University of Innsbruck

Christoph Spötl

University of Innsbruck

Laszlo Rinyu

Institute for Nuclear Research (ATOMKI)

Charlotte Honiat

University of Oxford

Show all 8 authorsHide

Content uploaded by Gabriella Koltai

Content may be subject to copyright.

Content uploaded by Gabriella Koltai

Content may be subject to copyright.

Late-stage evolution of hypogene caves at Tyuya-Muyun (Kyrgyzstan):

Quantitative insights from mineral deposits

Gabriella Koltai

a,*

, Christoph Sp¨

otl

, L´

aszl´

o Rinyu

, Charlotte Honiat

a,c

, Tanguy Racine

a,d

Haiwei Zhang

, Yves Krüger

, Yuri Dublyansky

Institute of Geology, University of Innsbruck, Innrain 52, 6020 Innsbruck, Austria

Isotope Climatology and Environmental Research Centre, Institute for Nuclear Research, Hungarian Research Network, Bem t´

er 18/c, 4026 Debrecen, Hungary

Centre Europ´

een de Recherche et d’Enseignement des G´

eosciences de l’Environnement, Universit´

e Aix-Marseille, cedex 04, BP 80 13545 Aix-en-Provence, France

Center for Hydrogeology and Geothermics, University of Neuchˆ

atel, 11 Rue Emile Argand, 2000 Neuchˆ

atel, Switzerland

Institute of Global Environmental Change, Xi’an Jiaotong University, Xi’an 710054, China

Department of Earth Science, University of Bergen, All´

egaten 41, 5007 Bergen, Norway

ARTICLE INFO

Editor: Christian France-Lanord

Keywords:

Caves

Hypogene speleogenesis

Geothermometry

Calcite spar

Barite

ABSTRACT

The Tyuya-Muyun massif in SW Kyrgyzstan hosts a number of caves some of which contain Ra- and U-bearing

minerals that were extensively mined in the early 20th century. Previous studies have suggested that the caves of

the Tyuya-Muyun have experienced a complex speleogenetic history, including epigene and hypogene processes.

Here, we reconstruct the late stages of hypogene processes by studying subaqueous (calcite and barite spars, cave

clouds), near-water table (rafts, folia) and vadose (owstone) mineral deposits in Great Barite and Surprise caves.

We determined the chronology (

230

Th dating), stable isotope composition (δ

O, δ

C and Δ

), and formation

temperature (uid inclusion microthermometry and Δ

) of these minerals and reconstructed the oxygen isotopic

composition of the paleo-water (δ

). In contrast to previous hypotheses about cave evolution, we found no

evidence of an initial epigenic karst phase. The earliest mineral deposits suggest already hydrothermal condi-

tions, with calcite and barite forming at ~40–50 ◦C prior to 600 ka. Cave clouds in Great Barite Cave mark the

beginning of the lowering of the groundwater table and also record a decrease in water temperature from 29.7 ±

8.3 ◦C to 12.7 ±5.6 ◦C, as shown by Δ

thermometry. The latter temperature remained stable, within analytical

uncertainties, during the past 600 ka, as indicated by Δ

results of near-water table and vadose speleothems in

Surprise Cave. At the same time, the decrease in δ

suggests a reduced contribution of thermal water and an

increased input of colder meteoric water. By around 540–450 ka Surprise Cave emerged from the phreatic zone,

as indicated by

230

Th ages of the highest folia and owstone. This lowering of the water table continued until 84

ka and may have been related to the nal uplift phase of the Tyuya-Muyun massif and the concomitant incision

of the Aravan River forming the Dangi Gorge.

1. Introduction

The Tyuya-Muyun massif in southern Kyrgyzstan is famous for its

uranium, vanadium and copper mineralization associated with gangue

calcite and barite. The ore lodes were extensively mined in the early

20th century, rst for Ra and later for U (Pogodin and Libman, 1971;

Dahlkamp, 2010;Dublyansky et al., 2017). Since the initial phases of

exploration, the origin of the ore-bearing cavities has been reported to

be related to karst processes (Vernadsky, 1914;Scherbakov, 1924;

Fersman and Scherbakov, 1925;Fersman, 1927, 1928;Kirikov, 1929;

Pavlenko, 1933). However, details of their evolution remained elusive

until the concept of hypogene speleogenesis, i.e., the formation of

solution-enlarged permeability structures driven by upwelling ground-

water charged with CO

and/or H

S that originated from deep-seated

sources, was developed decades later (e.g., Klimchouk et al., 2000;

Klimchouk, 2009) and subsequently applied to this site (Dublyansky

et al., 2017).

The mineral inll of the Tyuya-Muyun lodes has been removed by

mining operations, which makes a direct reconstruction of the ore-

forming processes using modern mineralogical and geochemical

methods impossible. However, the Tyuya-Muyun massif hosts several

caves of hypogene origin in proximity to the former ore deposits. These

* Corresponding author.

E-mail address: gabriella.koltai@uibk.ac.at (G. Koltai).

Contents lists available at ScienceDirect

Chemical Geology

journal homepage: www.elsevier.com/locate/chemgeo

https://doi.org/10.1016/j.chemgeo.2024.122454

Received 28 April 2024; Received in revised form 9 October 2024; Accepted 15 October 2024

Chemical Geology 670 (2024) 122454

Available online 18 October 2024

http://creativecommons.org/licenses/by/4.0/ ).

caves host calcite and barite mineralization which are genetically

analogous to the gangue mineralization of the lodes (Smolyaninova,

1970;Kirikov, 1929).

Early studies of the Tyuya-Muyun deposit based on mineralogical

observations inferred that: (i) the ore mineral sequence was deposited in

an epigene karst cave, on top of a normal subaerial owstone (“stalag-

mitic crust”); (ii) the mineral-forming water was low-temperature hy-

drothermal (not exceeding 60 ◦C); and (iii) the temperature (T) of the

water decreased during the course of the mineralization (Fersman and

Scherbakov, 1925;Fersman, 1928). The uid inclusion micro-

thermometry method (FIM), allowing quantication of paleo-T, gained

widespread use during the mid-20th century. Applying this method to

samples from Tyuya-Muyun conrmed that the earliest calcite precipi-

tated at T<60 ◦C (Dublyansky et al., 1989;Dublyansky et al., 2017).

Younger minerals (calcite and barite) contained only single-phase all-

liquid inclusions (Dublyansky et al., 2017) and were thus not amenable

to “classical”FIM.

Recent advancements in geothermometry methods, specically the

introduction of the nucleation-assisted uid inclusion micro-

thermometry (NA-FIM) and clumped isotope thermometry (Δ

), allow

the determination of paleo-T in the low-T realm (Krüger et al., 2007;

Eiler, 2011,Meckler et al., 2015,G´

azquez et al., 2018;Løland et al.,

2022;Koltai et al., 2024;Dem´

eny et al., 2024). In this study, we applied

these two geothermometry methods to minerals from Tyuya-Muyun

caves in an effort to reconstruct T and δ

O of the mineral-forming

uid. Furthermore, we applied

230

Th dating to determine the chrono-

logical sequence of the mineral deposits, as no reliable time constraints

were previously available.

This paper focuses on two caves of the Tyuya-Muyun massif, Great

Barite and Surprise caves, located ca. 1.2 km apart. Both caves bear no

ore minerals but contain deposits of calcite spar and barite. Barite in

these caves appears to be identical to barite in the Main Lode (now

Fersmana Cave) and Academy Lode (Smolyaninova, 1970;Kirikov,

1929). Calcite occurring in caves as large crystals was reported from the

lodes as medium-grained dark-grey calcite (“ore marble”-Fersman,

1927). All of these are subaqueous hydrothermal deposits by our

interpretation. We managed to collect two in situ samples of calcite from

Fersmana Cave that represent the early phases of mineralization. Results

from these samples are included in this study. In addition to subaqueous

hydrothermal minerals (calcite and barite spars), Surprise and Great

Barite caves also host epiphreatic and subaerial carbonate speleothems,

indicative of the transition from a shallow phreatic to a vadose setting, i.

e., cave clouds (also referred to as cave mammillaries), accumulations of

calcite rafts, folia and owstone.

2. Study area

Tyuya-Muyun (40.356◦N, 72.600◦E) is located in SW Kyrgyzstan, at

the foothills of the Alai range of the Pamir-Alai mountain system. Tyuya-

Muyun forms a ridge stretching in east-west direction for ca. 7 km and

reaching 1434 m in elevation.

2.1. Climate

The climate in the area is continental with dry and hot summers (Dsa,

according to the K¨

oppen–Geiger classication). The mean annual air

temperature (MAAT) recorded at the Osh meteorological station (875 m

a.s.l.) between 1959 and 1997 was 15.2 ◦C (Wolff et al., 2017).

Considering that Tyuya-Muyun is at 1100–1200 m altitude and applying

a lapse rate of 1 ◦C/100 m, the MAAT at the study site is close to 13 ◦C.

The mean annual rainfall is 360 mm (Dublyansky et al., 2017).

2.2. Geology

The Tyuya-Muyun massif represents ca. 650–700 m thick slab of

Lower Carboniferous limestone, sandwiched between thick sequences of

Paleozoic deep-marine sediments (coal-bearing shales, sandstones and

quartzose schists, partly bituminous limestones) interstratied with

volcanic rocks (Fig. 1a). The strata have an overall E–W strike and a

near-vertical dip. In the north and south, the limestone unit is separated

from the non‑carbonate rocks by steep normal faults (Kazansky, 1970;

Dublyansky et al., 2017).

The eroded surface of Paleozoic rocks is unconformably mantled by a

sequence of Jurassic lacustrine deposits (sandstones, conglomerates,

coal seams) followed by Cretaceous shallow-marine sediments (con-

glomerates, shales, siltstones and sandstones, limestones and dolomite).

This Mesozoic unit was also affected by denudation. The paleo-relief

resulting from the denudation of the Paleozoic and Mesozoic rocks

was buried by Paleogene and Neogene conglomerates and sandstones, as

well as by Quaternary deposits (e.g., loess).

Uplift and concomitant uvial erosion have resulted in the exhu-

mation and shaping of a ca. 7 km-long, E–W trending ridge of Lower

Carboniferous limestone, the Tyuya-Muyun massif. The predominantly

non‑carbonate Paleozoic units north and south of this ridge form a

lower-lying landscape of the river valley, while Mesozoic and Paleogene

to Quaternary deposits form a plateau, beneath which the massif

plunges towards the east and west.

At the southern side of the massif, the two northward owing rivers,

Kyrgyz-Ata and Hos-Chan, merge to form the Aravan River, which cuts

an up to 300 m deep gorge (Dangi) across the massif (Fig. 1a).

2.3. Hydrogeology

The modern groundwater ow system in the Tyuya-Muyun massif

comprises two end-members: slightly thermal (up to 23 ◦C) upwelling

water and downward and eastward owing water recharged at the

surface. The deep-seated component can be inferred from the constant

discharge, stable chemical composition, and the lukewarm character of

the (now dry) Kok-Bulak spring (Uklonsky, 1927;Zaitsev, 1940). The

massif is recharged by inltration from the Paleogene-Quaternary rocks

forming the plateau adjacent to the Tyuya-Muyun massif to the south

and west, as well as by partial capture of the surface runoff in ravines

incised in the westernmost part of the massif. The recharge area is

estimated at 27 km

(Butov and Zaitsev, 1935). At the time when the

massif was only marginally affected by mining operations, the ground-

water level in the Fersmana Cave (formerly Main Lode) responded to

spring snowmelt (3 to 4 m rise between April and June), but was un-

affected by individual rainfall events (Pavlenko, 1933).

Groundwater discharge previously occurred via three natural luke-

warm springs. The largest of them, Kok-Bulak, was located on the

southern slope of the massif, ca. 200 m west of the mouth of Dangi gorge

and ca. 50 m above the level of the Aravan River. Its mean annual

discharge was 6.4 L/s and T was 20–23 ◦С(i.e., 7 to 10 ◦Сwarmer than

the MAAT). The Southern Tange spring was located at the southern

mouth of the Dangi gorge, ca. 10 m above the Aravan River. The

discharge rate was 0.14 L/s and T was 18–22 ◦С. The Northern Tange

spring was located closer to the northern end of the gorge, 30 to 40 m

above the Aravan River. Its discharge was 2.4 L/s and T was 18–22◦С.

Cumulatively, the three springs discharged about 5–10 L/s (Butov and

Zaitsev, 1935). All three springs deposited massive travertines.

Mines in the Fersmana Cave remained dry until they reached a depth

of −175 m a.s.l. In 1947–1950, a 2200 m-long drainage adit was driven

from the southern entrance of the Dangi gorge westward towards the

Lode (Fig. 1b). The Dangi adit lowered the groundwater level in the

mine by 45 m. It also intercepted the ow of all three natural springs and

they fell dry. The water presently discharging through the adit has a

stable chemical composition (Na =150.5 ±0.8 mg/L; Ca =107.1 ±1.0

mg/L; Mg =38.1 ±0.4 mg/L; K =2.5 ±0.1 mg/L; SO

=334.7 ±1.1

mg/L; HCO

=191.6 ±12.3 mg/L; NO

=50.1 ±1.5 mg/L) and an

isotopic composition which is also stable throughout the year (δ

−12.40 ±0.07 ‰;δ

H= − 90.0 ±1.0 ‰VSMOW; (this study). A more

detailed discussion of the geology and hydrogeology of the area can be

G. Koltai et al. Chemical Geology 670 (2024) 122454

found in Dahlkamp (2010) and Dublyansky et al. (2017).

2.4. Caves and minerals of the Tyuya-Muyun massif

Before mining operations started in the Tyuya-Muyun massif in

1904, it hosted about 50 occurrences of ore (containing Cu-, Ra-, U- and

V-bearing minerals) and gangue minerals (calcite, barite) in tubular and

lens-shaped paleokarst cavities. The two economically important ore

bodies, the 220 m-deep Fersmana Cave and the 60 m-deep Academy

Lode, represented well-developed karst caves, almost entirely lled by

ores. The lodes essentially became caves again after their inll was

removed by mining operations between 1904 and 1951.

Besides karst-hosted ore bodies, there are several ‘proper’caves in

the massif, most of which host smaller amounts of minerals. The dis-

tribution of mineralization in the massif exhibits a symmetry in E-W

direction (Kirikov, 1929): U-V mineralization occurs along a ca. 300 m-

long stretch in the central part of the massif (including the Academy

Lode and Fersmana Cave). Cu-minerals and red barite (colour due to

dispersed hematite) occur over ca. 500 m. Yellow-green barite (tabular

crystals up to 10 cm in size) and calcite (scalenohedral crystals up to 40

cm in size) are present over the entire exposed length of the massif (ca.

2.9 km), with the relative proportion of barite decreasing markedly in

easterly direction (Fig. 1b).

A separate variety of calcite occurs as breccia cement. Irregularly

shaped breccia bodies, in which large blocks of limestone (up to 1 m in

size) “oat”in coarsely-crystalline milky-white calcite cement, are

common in the central part of the massif, in the vicinity of Fersmana

Cave; the bodies become smaller and less abundant in the eastern part of

the massif. These breccias pre-date ore mineralization and may be

related to the metamorphic alteration of the limestone.

2.5. Studied caves

2.5.1. Surprise Cave

Surprise Cave opens in a recess in the eastern wall of the Dangi Gorge

40 m above the Aravan River. The canyon wall hosts several cave en-

trances, one leading to the small Petrova Cave and another to an un-

named chamber. The cavities were truncated during the incision of the

gorge. Fragments of smaller (up to 2 m) cavities partly lined by calcite

spar occur in the canyon wall immediately to the north of the Surprise

Cave entrance, as well as 40 m below the entrance, near the river.

Surprise Cave is 505 m long and has a vertical extent of 80 m. The

cave is comprised of a three-dimensional network of anastomosing

crawlways, chambers, rising tubes and a series of chambers and niches

of different sizes (Fig. 2a). The ascending cave passages end blindly.

Locally, the cave walls are covered by scalenohedral calcite crystals

which reach up to 40 cm in length (Fig. 2b). Euhedral barite is present

only in one small chamber whose walls are lined with up to 20 cm-sized

tabular crystals (Fig. 2h).

2.5.2. Great Barite Cave

Great Barite Cave is a natural “super geode”(Dublyansky et al.,

2017) consisting of a 45◦inclined, ca. 114 m long and 20 m wide

chamber with a vertical extent of 58 m (Fig. 3a). The cave has one

natural (upper) entrance, a shaft of about 20 m followed by an inclined

slope. An articial entrance (mining adit) leads to the lower part of the

chamber. The cave walls are coated with a thick sequence of layers of

calcite and barite crystals. The best exposure is located in the lower part

of the cave (Fig. 3b and c). Cave clouds are abundant in this cave (Fig. 3b

and c). The middle and upper parts of the cave are affected by

condensation corrosion, truncating layers of calcite deposits and

bedrock and resulting in smooth cave wall and ceiling surfaces (Fig. 3d).

Fig. 1. (a) Simplied geological map of the study area (modied after Dublyansky et al., 2017), showing the Paleozoic carbonate slab of the Tyuya-Muyun massif.

The caves discussed in the paper are indicated by red circles (1 - Academy Lode, 2 - Fersmana Cave, 3 - Great Barite Cave, 4 - Surprise Cave). (b) Cross section of the

Tyuya-Muyun ridge along trace A-B (see dashed line in (a)) showing the projection of the studied caves and the extent of mineralization zones. Calcite is present in all

caves and is therefore not shown. Natural cave passages are white, while mines are black. The black dashed line in (b) marks the inaccessible (because of rock

collapses) part of the Dangi adit. (For interpretation of the references to colour in this gure legend, the reader is referred to the web version of this article.)

G. Koltai et al. Chemical Geology 670 (2024) 122454

2.5.3. Fersmana Cave

Fersmana Cave (formerly Main Lode) is located in the western part of

the Tyuya-Muyun massif and is the largest cave in the area. At a depth of

ca. 170 m, the cave tube splits in two, with the North stope terminating

at −180 m and the South stope at −220 m below the cave entrance. The

cave was originally lled with ores and gangue calcite and barite (see

Section 2.4). Only traces of the in-situ mineralization remain in the cave

today.

3. Sampling and methods

3.1. Sampling

Barite and different types of calcite were systematically sampled in

Surprise and Great Barite caves to study the mineralization and the

speleogenetic evolution.

•Calcite spar (also referred to as phreatic spar) consists of cm-size,

well-developed, coarsely-crystalline euhedral calcite crystals that

formed in the phreatic zone (Hill and Forti, 1997) at probably very

slow growth rates (Decker et al., 2018).

•Cave clouds (also referred to as cave mammillaries) also formed

below the water table and consist of densely packed calcite crystals

that coat the cave walls in thick layers, creating a more or less uni-

form surface consisting of concave protrusions, reminiscent of

clouds.

•Isopachous calcite is used in this paper as a working term for ca. 1 cm

thick layers of phreatic calcite consisting of elongated calcite crys-

tals, less than 1 mm wide, oriented normal to the substratum, coating

the cave wall. Isopachous calcite may be interpreted as incipient cave

clouds. In contrast to the latter, however, it does not smooth out the

rugged underlying surface, but simply coats it, so that all protrusions

remain expressed in the shape of the formation.

•Folia is typically a rare formation in caves and resembles inverted

rimstone dams or interlocking wavy ribs forming mostly on over-

hanging ledges and walls, and developed at the water-air interface

(Hill and Forti, 1997).

•Rafts and raft accumulations (e.g., cave cones or “Christmas trees”)

form on lake surfaces and accumulate on the lake bottom. Cave cones

or “Christmas trees”(Hill and Forti, 1997) form in cave pools un-

derneath permanent drip points. Rafts sink in stagnant water,

forming conical deposits on the lake bottom which are subsequently

overgrown and indurated by calcite (G´

azquez and Calaforra, 2013a).

•Flowstones are vadose speleothems that were deposited as layers of

calcite fed by thin water lms.

We present data on seven samples of scalenohedral calcite spar, six

samples of isopachous calcite, two samples of folia, one sample of calcite

rafts and one sample of coarsely-crystalline barite from Surprise Cave

(Table 1). In addition, three small diameter (2 cm) drill cores were ob-

tained: two cores intersected isopachous calcite layers and bedrock; the

third core sampled owstone and an underlying sequence of isopachous

calcite, scalenohedral calcite and bedrock. Three samples of scaleno-

hedral calcite crystals, two samples of cave clouds, one sample of light-

green and two samples of bottle-green barite crystals, and one core

drilled across the mineral sequence are presented from Great Barite

Cave. In Fersmana Cave, we managed to collect two samples of in-situ

calcite at the base of the mineral inll at a depth of 170 m.

In addition to the samples mentioned above, other calcite and

bedrock samples were measured for the stable isotopic composition. We

analyzed ve calcite spar samples, three samples of isopachous calcite

and one sample of folia from Surprise Cave, and two samples of calcite

spar from Great Barite Cave. Stable isotope data were also obtained from

Fig. 2. Surprise Cave. (a) 3D prole and plan view. The point cloud was acquired with hand-held ZEB Horizon LiDAR scanner. Overview showing different types of

calcite present in the cave: large scalenohedral calcite at the top (b) and the bottom (c) of the Crystal Chimney; scalenohedral calcite coated by isopachous calcite and

folia (c, d) and vadose owstone (c); isopachous calcite post-dating ne-grained cemented sediment (d, e); near-water table calcite deposits including cave cones (f)

and calcite rafts (g); large tabular barite in the Barite Chamber (h), where condensation corrosion has removed some of the barite exposing the underlying scale-

nohedral calcite.

G. Koltai et al. Chemical Geology 670 (2024) 122454

limestone on which the calcite spar formed, as well as from pre-ore

calcite cement of (presumably) metamorphic breccias.

3.2. Cave survey

Surprise Cave and Great Barite Cave were surveyed using a cali-

brated electronic cave surveying tool (DistoX2) with an accuracy of 0.5

%. We also documented mineral deposits, sediment lls, passage out-

lines and other features of interest. Further, the caves and mines of the

massif were documented with a hand-held laser scanner (LiDAR ZEB

Horizon, GeoSLAM Inc.).

3.3. Fluid inclusion microthermometry (FIM)

Optical microscopy was used to identify primary uid inclusion as-

semblages (FIAs) in the samples. FIM was performed on single-phase

liquid and two-phase liquid-vapour inclusions (FIs) with two different

analytical setups.

On single-phase liquid FIs nucleation-assisted uid inclusion micro-

thermometry (hereafter referred to as NA-FIM) was applied using a

Linkam THSMG 600 heating/freezing stage mounted on an Olympus

BX53 microscope at the Department of Earth Science, University of

Bergen. The microscope is connected to an amplied Ti:sapphire

femtosecond laser (CPA-2101, Clark-MXR, Inc.). The 775 nm laser beam

is coupled into the light path of the microscope via a dual port inter-

mediate tube equipped with a short-pass dichroic mirror. The laser beam

is focused on the sample through a 100×long working distance objec-

tive (Olympus LMPLFLN). Bubble nucleation was induced in the meta-

stable liquid of the inclusions by means of a single laser pulse (Krüger

et al., 2007). The setup allows for repeated and precise homogenization

temperature (T

) measurements of initially single-phase liquid

inclusions. T

was measured at least twice for each individual inclusion

on 15 to 20 FIs per sample. Typical reproducibility (precision) was

±0.05 ◦C. In addition, vapour bubble radii were measured at known T in

order to correct for the effect of surface tension on liquid-vapour ho-

mogenization using the thermodynamic model proposed by Marti et al.

(2012) and mean values of surface tension-corrected homogenization

temperatures (T

h∞

) were calculated. The overall analytical precision of

measurements of individual FIs was ±0.2 to 0.4 ◦C. Uncertainties of

NA-FIM temperatures are given as standard error (2 SE) of the mean and

also include the analytical uncertainty.

“Classical”FIM on two-phase inclusions were carried out using the

cycling technique (Goldstein and Reynolds, 1994) on a Linkam

THMS600 heating-freezing stage mounted to an OLYMPUS BX41 mi-

croscope and at the Institute of Mineralogy and Petrography, University

of Innsbruck. A synthetic FI standard was used to calibrate the stage on

each measurement day resulting in an accuracy of ±0.2 ◦C at T<100

◦C. T

values were corrected using the calibration data, and the T data

from a given FIA was considered reliable if 90 % of measured values did

not vary by more than 10–15 ◦C (Goldstein and Reynolds, 1994). Un-

certainties of FIM temperatures are given as standard error (2 SE).

We note that T

is a minimum formation T only, and the difference

between T

and the mineral formation T increases with water depth (i.e.,

hydrostatic pressure) at which the host mineral precipitated.

3.4. δ

O, δ

C and Δ

measurements

Calcite spar, rafts, folia and owstones were sampled at 1–5 mm

increments using a handheld drill. Carbonate powders were analyzed

using a Delta V Plus isotope ratio mass spectrometer coupled to a Gas-

bench II (Thermo Fisher Scientic). The results are reported relative to

the Vienna Pee Dee Belemnite standard (VPDB). The long-term precision

Fig. 3. Great Barite Cave. (a) 3D prole and plan view. The point cloud was acquired with hand-held ZEB Horizon LiDAR scanner . Studied mineral sequence. Red

laminated silt and clay cover the cave clouds in the lower part of the cave (b). Tabular barite covered with brous barite and cave cloud calcite (c). Condensation

corrosion truncating cave wall, scalenohedral and palisade calcite (d). Orange dashed rectangles in (a) mark the location of (b), (c) and (d). White lines trace the

boundaries of individual mineral zones. (For interpretation of the references to colour in this gure legend, the reader is referred to the web version of this article.)

G. Koltai et al. Chemical Geology 670 (2024) 122454

of the measurements is 0.06 ‰and 0.08 ‰for δ

C and δ

O, respec-

tively (1 SD) (Sp¨

otl and Vennemann, 2003).

A set of 16 calcite samples was analyzed for their clumped isotope

composition. To compare the performance of the two geothermometers

(FIM and clumped isotope thermometry), six scalenohedral crystals

were analyzed with both methods. The latter measurements were car-

ried out at the Isotope Climatology and Environmental Research Centre

(ICER), Institute for Nuclear Research (ATOMKI), Debrecen on a Thermo

Scientic MAT 253 Plus 10 kV isotope ratio mass spectrometer, after

phosphoric acid digestion at 70 ◦C using a Thermo Scientic Kiel IV

carbonate device. Measurements of 100

g aliquots of each sample were

replicated 12 times and measured alongside carbonate standard

samples. ETH1, ETH2, and ETH3 were used as normalization standards,

and IAEA-C2 was used as monitoring sample to determine the long-term

reproducibility of the instrument (1 SD =0.029 ‰;N=57). Negative

background, which is caused by secondary electrons on higher Faraday

cup detectors, was corrected for by application of the pressure-sensitive

baseline correction (Bernasconi et al., 2013) on all raw beam signals.

Data evaluation, standardization, and analytical error propagation of

clumped-isotope measurements were carried out using the CO

Clumped ETH PBL replicate analysis method, implemented in the

Easotope software (John and Bowen, 2016) using the revised IUPAC

parameters for

O correction (Brand et al., 2010). Δ

results are re-

ported on the I-CDES-90 scale (Bernasconi et al., 2021).

Table 1

Sample locations and descriptions, stable isotope data,

230

Th ages (±2

) (see Table 2 for details) and T estimates using FIM (±2SE) and clumped isotope thermometry

(TΔ

±95 % CI).

Cave Sample Elevation

(m a.s.l.)

Description Environment of formation Age (ka, BP) T based on FIM (◦C) T(Δ

) (◦C)

Surprise Cave SUR-2 1137 Scalenohedral calcite Subaqueous, slow growing >600

39.1 ±1.2

(n=16)

45.2 ±6.0

SUR-4 1130 Scalenohedral calcite Subaqueous, slow growing >600 37.1 ±5.0

based on SUR-5 (same geode)

SUR-5 1130 Scalenohedral calcite Subaqueous, slow growing >600

37.1 ±5.0

SUR-13 1081 Scalenohedral calcite Subaqueous, slow growing >600

41.0 ±1.4

(n=15)

37.7 ±3.6

SUR-22 1115 Scalenohedral calcite Subaqueous, slow growing >600

41.5 ±1.0

(n=30)

46.5 ±1.0

(n=9)

49.8 ±11.6

SUR-25 1140 Scalenohedral calcite Subaqueous, slow growing >600

40.5 ±2.0

(n=33)

41.8 ±2.0

(n=42)

47.8 ±6.2

SUR-26 1140 Scalenohedral calcite Subaqueous, slow growing >600

50.1 ±11.2

SUR-17 1129 Isopachous calcite Subaqueous, slow growing >600

 

SUR-27 1131 Isopachous calcite (pre-dating the folia)

Folia (calcite)

Subaqueous growth

Water table

494 ±48

541 ±47

 

SUR-C3 1120 Isopachous calcite Subaqueous, slow growing 389 ±14 13.4 ±7.3

SUR-C4 1121 Isopachous calcite Subaqueous, slow growing 363 ±14  

SUR-14 1088 Isopachous calcite (coating sediment) Subaqueous, slow growing 99.5 ±1.1 9.1 ±6.0

SUR-18 1115 Isopachous calcite Subaqueous, slow growing 332 ±10 7.0 ±6.8

SUR-21 1118 Isopachous calcite Subaqueous, slow growing 350 ±13  

SUR-1 1130 Folia

(post-dating scalenohedral calcite)

Water table 493 ±19  

SUR-29 1129 Folia Water table 524 ±63  

SUR-16 1102 Raft/platelet (calcite) Water table 203 ±314.7 ±6.6

SUR-28 1088 Raft/platelet (calcite) Water table 86.3 ±0.7

84.1 ±0.4

 

SUR-C1 1130 Flowstone Subaerial 476 ±31

(15 mm)

495 ±39 ka

(57 mm)

14.9 ±4.1

SUR-9 1094 Tabular barite Subaqueous   

Great Barite Cave BAR-7 1227 Scalenohedral calcite Subaqueous, slow growing >600 ka

49.2 ±7.2

BAR-8 1227 Scalenohedral calcite Subaqueous, slow growing >600 ka

44.9 ±2.4

(n=17)

51.4 ±8.7

BAR-9 1227 Palisade calcite Subaqueous, slow growing >600 ka

52.3 ±3.0

(n=20)

44.0 ±7.6

BAR-C1 1227 Cave cloud

(early phase)

Subaqueous >600 ka

29.7 ±8.3

BAR-12 1227 Cave cloud

(late phase)

Subaqueous >600 ka

12.7 ±5.6

BAR-10 1227 Light green tabular barite Subaqueous >600 ka

 

BAR-11 1227 Bottle green tabular barite Subaqueous >600 ka

 

BAR-18 1220 Tabular barite Subaqueous >600 ka

54.8 ±1.0

(n=19)



Fersmana Cave FER-05 1149 Palisade calcite Subaqueous   51.4 ±7.8

FER-06 1149 Coarsely crystalline calcite Subaqueous   54.9 ±8.5

Pre-cave breccia TM2022–1 1130 Breccia cement calcite    79.7 ±4.9

Uncertainties on T estimates also account for the overall analytical precision of used geothermometry methods. For details see Table 2 and Table S1.

>600 ka based on the innite

230

Th age of SUR-4, representing the phase of scalenohedral calcite growth.

>600 ka based on the innite

230

Th age of BAR-12 and BAR-C1, representing the youngest phase of calcite formation in Great Barite Cave.

Evolution curve of

230

Th/

238

U (activity) vs. ẟ

234

U (measured) indicates that the U-Th system of these samples may have been opened.

G. Koltai et al. Chemical Geology 670 (2024) 122454

In contrast to T

,Δ

temperatures are considered as mineral for-

mation temperatures. T (Δ

) were calculated from the measured Δ

values using the unied calibration curve of Anderson et al. (2021). We

chose this calibration because it included two slow-growing subaqueous

cave calcites (DVH-2 from Devils Hole and LGB-2 from Corchia Cave)

that formed in a similar environment as some of the samples of our

study. The Δ

uncertainty of the estimated temperatures is given as 95

% condence interval (95 CI) which combine the analytical uncertainty

and the calibration uncertainty.

We calculated the oxygen isotope composition of the paleo-water

(δ

) based on the T dependence of the calcite-water oxygen

isotope fractionation. We used T(Δ

) and the equation of Da¨

eron et al.

(2019), which is based on slow-growing subaqueous speleothems,

similar to our samples.

3.5.

230

Th dating

A total of 14 and 2 calcite samples were selected for

230

Th dating

from Surprise and Great Barite caves, respectively. 25–30 mg was drilled

from one sample of calcite spar, eight samples of isopachous calcite,

three samples of rafts and two samples of folia using a handheld drill in a

clean air lamina-ow hood. Additionally, a owstone from Surprise

Cave was drilled at two discrete horizons following growth laminae.

Ages were determined by measuring U and Th isotope ratios on a

multi-collector inductively coupled mass spectrometer (Neptune Plus,

Thermo Scientic) after chemical separation following Edwards et al.

(1987) and Cheng et al. (2013). The analyses were carried out at Xian

Jiaotong University (China). Uncertainties (2

) for the U and Th isotope

measurements include corrections for blanks, multiplier dark noise,

abundance sensitivity, and contents of the same nuclides in the spike

solution. Decay constants for

230

Th and

234

U were reported by Cheng

et al. (2013) and corrected

230

Th ages assume an initial

230

Th/

232

atomic ratio of (4.4 ±2.2) ×10

−6

, the value for material at secular

equilibrium with the bulk earth

232

Th/

238

U ratio of 3.8. Final ages are

given as years before present (BP) where present is 1950 CE.

4. Results

4.1. Cave mineralization

Calcite spar occurs as individual crystals or as palisade crystal ag-

gregates, both showing euhedral scalenohedral terminations. In Surprise

Cave, scalenohedral calcite coats the walls (Fig. 2b and c) and represents

the rst phase of mineralization (Fig. 4). The growth sequence typically

starts with mm-sized calcite crystals. In the course of geometrical se-

lection individual crystals became larger, commonly resulting in a

columnar (palisade) fabrics. Euhedral crystals of scalenohedral habit,

having sizes from 2 to 3 up to 40 cm, developed during the latest stages.

These crystals seem to have formed in previously disconnected cham-

bers at all cave levels (Table 1).

Thin layers of isopachous calcite covering yellow ne-grained sedi-

ment attached to the host rock are present at several places (e.g., Upper

Chamber, Main Chamber, Big Crystal Chimney, Fig. 2d- and e), while

thick cave clouds are absent in Surprise Cave. Folia and rafts (Fig. 2c-g)

occur at three (Upper Chamber, bottom of the Crystal Chimney and the

Corkscrew, Fig. 2a) and two (Corkscrew, Big Crystal Chamber) locations

(Fig. 2a), respectively. Accumulations of calcite rafts form a cave cone at

the Corkscrew (Fig. 2f). The cave survey indicates that folia is present,

albeit discontinuously, over almost the entire vertical extent of the cave

(ca. 80 m).

A large owstone is present in the lower part of the Crystal Chimney

where scalenohedral calcite coated by folia covers the opposite

Fig. 4. Schematic presentation of the mineral succession in Great Barite (a) and Surprise Caves (b). Note the occasional presence of red silt/clay deposits. (For

interpretation of the references to colour in this gure legend, the reader is referred to the web version of this article.)

G. Koltai et al. Chemical Geology 670 (2024) 122454

overhanging wall (Fig. 2c). We cored this owstone and found it to be 7

cm thick and underlain by a thin layer (few mm) of whitish isopachous

calcite. Beneath this isopachous calcite was a layer of reddish ne-

grained cemented sediment (Supplementary Material) and nally

partly fractured calcite spar was encountered. The core did not reach

bedrock.

In Surprise Cave, tabular barite was found in an isolated chamber

(Barite Chamber, Fig. 2h). Corrosion processes, presumably under low-

pH, removed some of the barite crystals on the ceiling of this chamber,

exposing the calcite crystals beneath, indicating that barite postdates

calcite (Fig. 4).

The last minerals to form in Surprise Cave are evaporites, including

gypsum and mirabilite. Gypsum occurs as discontinuous thin crusts and

“blisters”on cave walls and calcite speleothems, and as delicate gypsum

needles and “cotton balls”on the surface of dry sediment. Mirabilite was

found at one location in a dry passage with abundant ne-grained

sediment (Sahara, Fig. 2a). It forms patches of long, thin crystals on

the sediment-covered cave oor.

In Great Barite Cave the mineralization sequence is exposed in the

lower part of the cave (Fig. 3). The cave walls are covered by a 40–50 cm

thick, bipartite succession of calcite spar. Initially, large single scale-

nohedral crystals (commonly up to 15 cm in length, but locally reaching

25 cm) formed. They are overgrown by a layer of calcite with a palisade

fabric (Fig. 3b). Most of this younger palisade calcite consists of optically

continuous elongate (5–7 cm, Fig. 4) crystals conned by compromise

boundaries; only favourably oriented crystals developed euhedral ter-

minations, also showing scalenohedral faces.

Two generations of barite, both featuring well-formed tabular crys-

tals, precipitated on top of the calcite. The earlier generation is repre-

sented by light-green or yellow crystals 5–6 cm long; the later generation

consists of elongate bottle-green crystals with individual crystals

reaching 2–3 cm (Fig. 4). In some places, brous barite postdates tabular

barite (Fig. 3c and 4). The thickness of the composite barite layer is

typically 30–50 cm, locally reaching up to 1.5 m. The barite sequence is

covered by calcite that forms cave clouds, typically 30–40 cm thick,

representing the latest phase of mineralization (Fig. 3b and c).

In Fersmana Cave, only traces of the mineralization remain and a

detailed characterisation of mineral sequence is therefore not possible.

We managed to collect two samples of calcite at a depth of 170 m. The

rst calcite with a light green tint occurs as a ca. 2 cm-thin layer showing

palisade crystals at the base of a layered calcite-siltstone and represents

the earliest phase of mineralization. The thickness of the calcite and

reddish siltstone layers varies from a few mm to several cm. The second

calcite postdates a red (hematite-stained) quartz layer, which is present

among red (also hematite-stained) barite layers pre-dating yellow-green

barite at a depth greater than 160 m (Fersman, 1928). The sampled

calcite therefore formed relatively early in the mineral sequence.

4.2. Morphological observations

In Surprise Cave solutional channels and other solutional morphol-

ogies are present that postdate calcite spar but formed prior to iso-

pachous calcite. Similar morphological features were not observed in

Great Barite Cave.

Condensation corrosion affected both caves at a later stage and

resulted in metre-sized rounded niches and cupolas, truncating cave

clouds and cave walls in the upper part of the Great Barite Cave (Fig. 3d).

In Surprise Cave condensation corrosion is evidenced by solution mor-

phologies on the cave walls intersecting bedrock and calcite, as well as

smoothed surfaces of the calcite spar at several locations.

4.3. FIM

FIM was performed on calcite and barite crystals; no suitable FIs

were found in other speleothems. Primary FIs in the calcite (large

euhedral crystals and crystals of palisade aggregates) mostly occur in 3-

dimensional clusters (Fig. S1a) or follow crystallographically dened

directions (e.g., parallel to crystal faces). Primary FIs often show nega-

tive crystal shapes (Fig. S1a), whereby the inclusions have crystal faces

as surfaces (Goldstein and Reynolds, 1994) or irregular morphologies

(Fig. S1c). The inclusion size ranges from 10 to 50

m in scalenohedral

calcite from Surprise Cave and up to 100

m in euhedral and palisade

calcite from Great Barite Cave. In some samples both single-phase and

two-phase liquid-vapour inclusion FIAs are present.

Barite samples are rich in FIAs, but primary FIs are only a few

m in

length, rendering FIM difcult (Fig. S1b). Therefore, T

were measured

in only one sample, BAR-18, which contained larger (10–20

m) pri-

mary single-phase and two-phase liquid-vapour inclusions. These FIs

show negative crystal shapes (Fig. S1d). The vast majority of FIs in the

other barites were single-phase.

values of primary FI provide a minimum estimate of the entrap-

ment T of FIs, unless the trapping pressure (P) is known and can be

corrected for (e.g. Goldstein and Reynolds, 1994) Thus FIM in our study

provides minimum estimates of the temperature at which the mineral

was formed. T

h∞

of primary single-phase FIAs of scalenohedral calcite

from Surprise Cave (SUR-2 and SUR-13) show consistent values of 39.0

±1.2 ◦C and 41.0 ±1.4 ◦C (2 SE). T

measured on two-phase FIAs (SUR-

22 and SUR-25) range from 40.5 ±2.0 ◦C to 46.5 ±2.0 ◦C (Table 1 and

Fig. S2). Primary single-phase FIAs of calcite spar from Great Barite Cave

(BAR-7 to 9) yielded T

h∞

of 44.9 ±2.4 ◦C and 52.3 ±3.0 ◦C. FIs in spar

BAR-8 have a bimodal distribution, nevertheless 90 % of the T

h∞

vary by

less than 15 ◦C and is therefore considered reliable (Fig. S2). T

h∞

within

individual FIAs of the second spar (BAR-9, Fig. S2) show a wide scatter

(>15 ◦C range, Fig. S2) and are thus not considered appropriate to

determine the formation T. FIM on barite sample BAR-18 yielded T

h∞

54.8 ±1.0 ◦C (Table 1, Fig. S2).

4.4. Stable C and O isotopes

Samples of the Lower Carboniferous limestone forming the former

cave wall beneath calcite spar from Surprise Cave yielded C and O

isotope values between 2.0 and 4.4 ‰and from −7.0 to −0.6 ‰,

respectively (Fig. 5a). In Great Barite Cave, δ

O values of the host rock

show an even larger spread of 10.2 ‰(from −17.4 to −7.2 ‰,Fig. 5b),

while δ

C values vary from −0.9 to 1.4 ‰(Fig. 5b). Except for the

outermost approximately 2–3 mm of core SUR-C2, stable isotope tran-

sects across the two short drill cores and hand specimens from Surprise

Cave show no clear alteration trend. Similarly, no clear shift in stable C

and O isotope values is seen in transects of hand specimens from Great

Barite Cave.

Calcite spar, isopachous calcite, rafts, folia and owstones from

Surprise Cave show largely overlapping stable isotopic compositions,

ranging from −13.7 to −5.9 ‰for δ

O and from −5.9 to 0.2 ‰for δ

(Figs. 5a and 6). Data from Great Barite Cave show a different picture

(Figs. 5b and 7). In this cave, spar δ

O values (−17.3 and −10.1 ‰) are

on average lower and their δ

C values (−1.8 to 0.1 ‰) are higher than

those of Surprise Cave. Cave clouds, absent in Surprise Cave, are char-

acterised by δ

O values overlapping with the high end of the spar

values (−12.2 to −9.9 ‰). Their δ

C values are distinctly low (−7.9 to

−3.7 ‰) and reach even lower values than vadose speleothems such as

owstone in Surprise Cave (Figs. 5–7). Transects across large calcite

crystals (up to ~10 cm in length) show no evidence of isotopic zoning

(Fig. S3). Stable oxygen and carbon isotope values of the light green

calcite from Fersmana Cave are −9.6 to −8.2 ‰and −3.0 to −1.4 ‰,

respectively, overlapping with those of calcite spar from Surprise Cave

(Fig. 5). Calcite post-dating red quartz in Fersmana Cave exhibits

somewhat lower δ

O (−10.3 to −10.8 ‰) and higher δ

C (0.5 to 1.3

‰), similar to altered bedrock (Fig. 5).

Clumped isotope data of 16 samples show Δ

values ranging from

0.517 to 0.652 ‰with uncertainties between 0.005 ‰and 0.012 ‰(1

SE) (Table S1) and 0.012 ‰and 0.026 ‰(95 % CI). The calculated T of

calcite spar formation range from 37.1 ±5.0 ◦C to 49.8 ±11.6 ◦C and

G. Koltai et al. Chemical Geology 670 (2024) 122454

from 44.0 ±7.6 to 51.4 ±8.7 ◦C in Surprise and Great Barite caves,

respectively (Table 1, Table S1). The two calcite samples from Fersmana

Cave yielded similar T of 51.4 ±7.8 to 54.9 ±8.5 ◦C. Cave clouds in

Great Barite Cave record a decrease in water T from 29.7 ±8.3 ◦C to

12.7 ±5.6 ◦C. Subaerial owstone in Surprise Cave formed at ~15 ◦C

(Table 1, Table S1). Isopachous calcite and rafts from the same cave

yielded T between 7.0 ±6.8 and 14.7 ±6.6 ◦C (Table 1, Table S1).

4.5.

230

Th dating

The

238

U concentration of calcite samples varies widely from 0.006

to 3.7 ppm in Surprise Cave and is below 1 ppm in the cave clouds from

Great Barite Cave (Table 2). Only one sample indicates detrital

contamination by showing a

230

Th/

232

Th atomic ratio of less than 100

ppm. The 2

precision of the remaining ages ranges from 0.5 % to 12.0

%, with greater uncertainties associated with ages close to the limit of

230

Th dating.

230

Th dating indicates that scalenohedral and palisade

calcite formed before about 600 ka (Table 2).

In Great Barite Cave the youngest carbonate deposit (BAR-12, BAR-

C1, Table 2) represented by cave clouds returned

230

Th ages in secular

equilibrium (Table 2), however the evolution curve of the

230

Th/

238

activity ratio versus δ

234

U indicate a partial opening of the U-Th system.

The owstone in the upper level of Surprise Cave was deposited between

495 ±40 ka and 477 ±31 ka (Table 2).

230

Th ages of isopachous calcite,

folia and rafts from different levels of Surprise Cave (Table 1) indicate

speleothem formation close to the former water table between 571 ±47

and 84 ±0.4 ka (Table 2).

5. Discussion

5.1. Conceptual framework

The Tyuya-Muyun caves contain a suite of calcite and barite that

record changes in T, stable isotopic composition, and the position of the

water table of the paleo-aquifer. Below, we summarise the conceptual

framework on which our interpretations are based.

Fig. 5. Stable oxygen and carbon isotopic composition of carbonate samples from Surprise (a) and Great Barite (b) caves. Data of a pre-ore (and pre-cave) breccia

calcite cement formed at ~80 ◦C, and calcite from Fersmana cave (labelled FER) are included as well. Pristine bedrock (limestone, grey box) is characterised by high

C and O isotope values, while most of the bedrock samples show isotopic alteration likely related to metamorphic processes unrelated to subsequent karstication.

Fig. 6. Box and whisker plots showing the carbon and oxygen isotope composition of the different calcite types in Surprise Cave. IQR stands for interquartile range (i.

e., the distance between the upper (Q

) and lower (Q

) quartiles).

G. Koltai et al. Chemical Geology 670 (2024) 122454

•Calcite spar formed in a subaqueous environment at an unknown

depth, depending on the temperature, pCO

salinity and temperature

of the water as well as on the rate of CO

degassing (Dublyansky,

2000;Hill and Polyak, 2010). Based on geochemical considerations,

the precipitation of low-T hydrothermal calcite likely occurred at

depths up to 300 m, under hydrostatic pressure conditions (Malinin,

1979;Dublyansky, 2000). In a similar environment to Tyuya-Muyun,

the depth of spar formation in caves of the Grand Canyon (Arizona,

USA) was estimated at ~100 m (Hill and Polyak, 2010), while

Decker et al. (2018) found that calcite spar formed in the Guadalupe

Mountains (New Mexico and Texas, USA) at 500 ±250 m depth.

•Euhedral barite crystals formed subaqueously at unknown depth.

Comparable barite deposits associated with calcite spar have been

reported from hypogene caves of Mt. Saint Giovanni (Sardinia; De

Waele et al., 2013;G´

azquez et al., 2018).

•Cave clouds (mammillary calcite) also formed subaqueously (cf. Hill

and Forti, 1997) but likely at shallower depth than calcite spar. Cave

clouds formed relatively close to the water table in Sima de la

Higuera Cave (Spain; G´

azquez and Calaforra, 2013b) while similar

calcite formed within a few metres of the paleo-water surface in

M¨

archenh¨

ohle (Austria) Dublyansky et al., 2016) and in caves of the

Grand Canyon (Arizona, USA, Hill and Polyak, 2010). The deepest

known cave clouds occur in Devils Hole (Nevada, USA) down to a

depth of approximately 140 m (Riggs et al., 1994). However, Devils

Hole should be viewed as an exception among sites of subaqueous

calcite deposition, since it is the only site known today where the

cave (fracture) extends to great depth underwater where the mam-

millaries have been observed to the maximum depth reached by

divers.

•Isopachous calcite (in this study) is interpreted to indicate a shallow

phreatic environment (similar to cave clouds). However, thin iso-

pachous calcite represents a much shorter time span than thick cave

clouds.

•Folia constrains the position of the paleo-water table because these

deposits formed within the range of small-scale water table

uctuations.

•Rafts and raft accumulations may form in vadose settings, however

in Surprise Cave they formed near the water table, because they are

associated with isopachous calcite. Therefore, in this study we refer

to rafts and raft cones as near-water table speleothems.

•Flowstones are evidence of vadose conditions.

5.2. Mineralization patterns

The early stages of cave mineralization in Surprise and Great Barite

caves are similar, represented by scalenohedral calcite spar. The

sequence in Great Barite Cave is somewhat more complex, consisting of

two episodes of calcite deposition (scalenohedral crystals and palisade

aggregates), while Surprise Cave shows evidence of only one episode.

Another difference between the two caves is that Great Barite Cave hosts

abundant barite, while in Surprise Cave barite is restricted to a single

isolated chamber. This is consistent with the overall mineralogical

zonation of the Tyuya-Muyun massif (Fersman, 1928;Dublyansky et al.,

2017;Fig. 1b). This zonation is interpreted as a E–W gradient of the

upwelling mineral-forming paleo-waters, not accompanied by a signif-

icant thermal gradient.

More variability can be observed for the later stages of mineraliza-

tion. First, thick cave clouds cover the walls of Great Barite Cave but are

absent in Surprise Cave. In the latter, isopachous calcite of shallow

phreatic origin is present as a thin layer yielding nite

230

Th ages. Iso-

pachous calcite also shows T similar to those of the latest stage of cave

cloud formation in Great Barite Cave.

Fig. 7. Box and whisker plots showing the carbon and oxygen isotope

composition of the different calcite types in Great Barite Cave. IQR stands for

interquartile range (i.e., the distance between the upper (Q

) and lower

) quartiles).

Table 2

230

Th dating results of calcite samples from Surprise and Great Barite caves.

Sample

238

(ppb)

232

(ppt)

230

Th/

232

Th (atomic

x10

−6

)

ẟ

234

U (measured)

230

Th/

238

(activity)

230

Th age (ka)

(uncorr.)

ẟ

234

initial

(corr.)

230

Th age (ka)

(corr.)

SUR-C1-1-

2788 ±10 422 ±9 134,791 ±2851 183 ±2 1.2387 ±0.0052 477.1 ±31.4 703 ±64 477.0 ±31.4

SUR-C1-1-

1672 ±4 773 ±16 44,045 ±919 178 ±2 1.2354 ±0.0059 495.4 ±39.7 720 ±82 495.3 ±39.7

BAR-12 1260 ±3 29 ±2 734,862 ±43,058 14 ±2 1.0231 ±0.0023   Sec. eq

BAR-C1 141 ±1 25,672 ±527 113 ±2 88 ±9 1.2482 ±0.0093   OS

SUR-C3 1385 ±4 2027 ±41 12,815 ±263 125 ±2 1.0596 ±0.0063 389.7 ±14.2 377 ±17 389.6 ±14.3

SUR-C4 1570 ±7 1399 ±29 21,239 ±443 142 ±3 1.2482 ±0.0093 363.2 ±13.8 395 ±17 363.2 ±13.8

SUR-1 2172 ±3 3005 ±60 13,920 ±280 129 ±2 1.1679 ±0.0019 493.2 ±18.7 521 ±28 493.1 ±18.7

SUR-4 6 ±0 2934 ±59 40 ±2 166 ±35 0.7260 ±0.0021   Sec. eq.

SUR-14 2992 ±18 10,825 ±225 3915 ±82 384 ±5 1.1379 ±0.0040 99.6 ±1.1 509 ±6 99.5 ±1.1

SUR-16 933 ±2 2432 ±49 6122 ±125 80 ±2 1.1485 ±0.0052 230.4 ±2.9 153 ±4 230.3 ±2.9

SUR-18 948 ±3 2526 ±52 6979 ±143 138 ±3 0.8592 ±0.0054 332.1 ±9.8 353 ±12 332.0 ±9.8

SUR-21 1512 ±6 2072 ±43 13,344 ±276 116 ±3 0.9675 ±0.0028 350.4 ±12.8 311 ±14 350.3 ±12.8

SUR-27 3718 ±5 5099 ±103 13,426 ±271 87 ±2 1.1279 ±0.0043 571.5 ±47.3 436 ±60 571.4 ±4.7

SUR-27B 1990 ±7 74,978 ±

1528

469 ±10 60 ±3 1.1090 ±0.0049 495.3 ±48.8 240 ±35 494.4 ±48.3

SUR-28 3421 ±16 2572 ±53 16,192 ±335 314 ±3 1.1169 ±0.0023 86.4 ±0.7 400 ±4 86.3 ±0.7

SUR-28-b 3729 ±9 2500 ±51 17,857 ±362 314 ±3 1.0716 ±0.0045 84.2 ±0.4 398 ±3 84.1 ±0.4

SUR-29 2996 ±14 1344 ±28 43,033 ±905 129 ±3 0.7384 ±0.0038 523.9 ±62.6 564 ±104 523.8 ±62.6

ẟ

234

U=([

234

238

activity

–1)x1000. ẟ

234

initial

was calculated based on

230

Th age (T), i.e. ẟ

234

initial

=ẟ

234

measured

x e

λ234xT

. Ages are reported as BP, i.e. before the

year 1950 CE. The error is 2

. Sec. eq. refers to ages in secular equilibrium. OS indicates open system.

G. Koltai et al. Chemical Geology 670 (2024) 122454

Secondly, near-water table deposits such as folia and rafts (associ-

ated with isopachous calcite) are present in Surprise Cave but absent in

Great Barite Cave. This may be related to the difference in elevation

between the two caves (Fig. 1,Table 1), whereby the near-water table

deposits mark the lowering of the water table in Surprise Cave from ca.

540–474 ka at 1131 m a.s.l., 96 m below the elevation of the lowest

calcite deposit in Great Barite Cave (Table 1). It is also plausible that

Great Barite Cave was drained rather quickly so that there was no time

for near-water table speleothems to form. On the other hand, delicate

raft deposits may not have been preserved in Great Barite Cave, since the

latter was pervasively modied by mining activities. Based on the

elevation difference between the two caves and the

230

Th ages of the

highest isopachous calcite and folia (1131 m a.s.l.) in the upper part of

Surprise Cave we postulate that cave cloud formation in Great Barite

Cave (1227 m a.s.l.) occurred likely before 600 ka.

5.3. T and ẟ

O reconstruction of the paleo-water

5.3.1. Paleo-T estimates: FIM and Δ

thermometry

FIM was performed on six calcite spar samples yielding minimum

water temperatures of 39.0 ±1.2 ◦C to 46.5 ±2.0 ◦C and between 44.9

±2.4 ◦C and 52.3 ±3.0 ◦C in Surprise and Great Barite caves, respec-

tively. These temperature estimates show a good agreement with T(Δ

)

obtained from the same samples (Table 1). Clumped isotope thermom-

etry for BAR-9 yielded slightly lower formation T than FIM (Table 1). As

the T

h∞

from BAR-9 show a wide scatter (>15 ◦C range, Fig. S2) these

data are not considered reliable to determine formation T and we

consider T(Δ

) to be accurate.

data of uid inclusions provide minimum estimates of the for-

mation T, and a P correction is required to compensate for elevated P

associated with the water depth of calcite precipitation. The Tyuya-

Muyun caves formed in competent carbonate rocks, therefore we as-

sume a hydrostatic pressure gradient (~10 MPa/km). As the pressure

correction is ca. 1 ◦C/MPa, the T

of calcite formed in deep phreatic

conditions may underestimate paleo-water T by up to ca. 3 ◦C. Unlike

FIM, Δ

thermometry is not affected by pressure and thus provides an

estimate for the “true”formation T. The good agreement between the

thermometry and the FIM data (Table 1) suggests that calcite for-

mation in Surprise and Great Barite caves indeed occurred at relatively

shallow depths (i.e. less than about 300 m). However, the uncertainties

of the Δ

data (3.6 to 11.6 ◦C, 95 % CI, Table 1, Table S1) do not allow a

more nuanced assessment of the calcite formation depth.

Non-equilibrium isotope effects (e.g., as a result of fast CO

degass-

ing) during carbonate formation may alter the Δ

value of calcite

leading to the overestimation of the formation T (e.g., Guo, 2020;

Dreybrodt, 2019,Hansen et al., 2022). There is growing evidence that

calcite precipitation may occur close to oxygen and clumped isotope

equilibrium in subaqueous settings (Dreybrodt, 2019;Bajnai et al.,

2020;Bajnai et al., 2021;Koltai et al., 2024;Dem´

eny et al., 2024).

Similar studies of rafts and raft cones are still scarce. While Parvez et al.

(2024) observed disequilibrium isotope effects in cave rafts from

ventilated cenotes, Dem´

eny et al. (2024) reported T(Δ

) similar to

modern pool T in hydrothermal settings. Subaerial speleothems present

a similar picture: while some multiproxy (Meckler et al., 2015;Mat-

thews et al., 2021) and dual clumped isotope studies (Fiebig et al., 2023)

reported accurate T(Δ

), others indicate disequilibrium in Δ

leading

to erroneously high T (e.g., Wainer et al., 2011;Kluge and Affek, 2012).

Therefore, T(Δ

) should be regarded as maximum T estimates for cave

rafts and owstones in Surprise Cave.

There are three main reasons for considering these two T(Δ

) of

14.9 ±4.1 ◦C (owstone) and 14.7 ±6.6 ◦C (raft) estimates as likely

unaffected by kinetic isotope processes. First, prior to the incision on the

Aravan River, Surprise Cave had no natural entrance. The resulting

restricted air exchange would have promoted slow CO

degassing.

Second, isopachous calcites from the same cave suggest similar water T

(7.0 ±6.6 to 13.4 ±7.3 ◦C). Third, paleo-water temperatures would not

have been lower than cave air T at time of speleothem formation. Today,

the cave air T is close to 13 ◦C.

5.3.2. Mineral formation T and paleo-water ẟ

Mineral formation T suggest that the paleo-aquifer at Tyuya-Muyun

experienced noticeable cooling over time from the early stages when

hydrothermal calcite and barite crystallized (>600 ka) to the later stages

when shallow-phreatic and vadose speleothems formed (<540 ka; Fig.

S4).

Scalenohedral calcite in Great Barite and Surprise caves is thought to

be an analogue of the “ore marble”of the Fersmana Cave (which con-

tained most of the U- and Ra-bearing minerals; Smolyaninova, 1970).

Our data shows that the water T during this early phreatic stage was

around 40–60 ◦C (Fig. S4, Table 1), and likely below ~55 ◦C based on

FIM. The T remained stable during the subsequent stage of tabular barite

formation (Fig. 8) (note, however, that this T estimate is based on a

single sample, BAR-18).

A pronounced drop in water T is recorded by cave clouds in Great

Barite Cave at some time prior to 600 ka. T(Δ47) data suggest a water T

of 29.7 ±8.3 ◦C during the early phase of cave cloud formation, fol-

lowed by a cooling to 12.7 ±5.6 ◦C during the last phase of cave cloud

formation (Fig. 8,Table 1). Similar T were recorded for the near-water

table speleothems in Surprise Cave. These lukewarm T suggest mixing

of thermal and cooler waters of meteoric origin. We found no evidence

of thermal pulse(s) with T>40 ◦C during the last 600 ka in either cave.

Combining the mineral formation T and the δ

O of calcite, we

calculated the oxygen isotope composition of the paleo-water (δ

Table 3). The δ

of the calcite spar in Great Barite Cave was between

−9.3 ±1.2 and −6.2 ±1.3 ‰VSMOW (Fig. S4, Table 3). Six samples of

calcite spar from Surprise Cave yielded a very similar range of δ

from −9.8 ±1.0 to −5.7 ±1.0 ‰VSMOW (Table 3), while the two

calcite samples from Fersmana Cave suggest slightly higher values (−4.7

±1.3 to −4.4 ±1.3 ‰VSMOW). The δ

values show a marked

decrease in parallel with the T change during the later phases of calcite

formation. Cave clouds show δ

of −9.8 ±1.6 ‰at the beginning of

their formation followed by a decline to −14.4 ±1.2 ‰VSMOW

(Fig. 8). Flowstone, calcite rafts and isopachous calcite from Surprise

Cave precipitated from an isotopically similar water (δ

= − 13.0 ±

1.3 to −11.3 ±1.5 ‰VSMOW, Fig. S4, Table 3). These values are

similar to the stable isotopic composition of the modern spring dis-

charging through the adit that drains the Tyuya-Muyun massif (δ

−12.4 ±0.1 ‰VSMOW).

5.4. Late-stage cave evolution as reected by the cave mineralization

Here, we focus on the thermal and hydrological evolution of Great

Barite and Surprise caves, as inferred from their morphology and

mineralization. We note that the evolution of Fersmana Cave was likely

different as suggested by (i) its different cave geometry (tube-shape;

Fig. 1b) that extends ca. 100 m below the water table (lowered by the

construction of the drainage adit), (ii) the spatial relationship between

pre-ore karst cavity and the “ore channel”whereby the latter is always

displaced upward from the former (Kirikov, 1929), and (iii) its more

complex mineral paragenesis (Fersman, 1928).

5.4.1. Stage 0: Pre-mineral hypogene karst

The earliest speleogenetic event responsible for the creation of the

primary set of cavities was clearly of hypogene origin based on the cave

morphology (Fig. 5). Circulating aggressive water created initial cavities

at different elevations in the limestone of the Tyuya-Muyun massif

(Fig. 8). In Surprise Cave these hypogene waters created a series of so-

lution cavities up to several metres across. The observation that these

cavities host similar mineralization and sediment inll suggests that

these cavities were hydraulically inter-connected.

Based on cave morphology and mineral deposits, we postulate that

was the primary source of aggressiveness. No evidence was found

G. Koltai et al. Chemical Geology 670 (2024) 122454

for the involvement of sulfuric acid, although we note that sulfur was

present in the mineral-forming solutions, as indicated by the local

occurrence of barite (Stage 2).

5.4.2. Stage 1: Hydrothermal calcite (calcite spar and palisade calcite)

As the circulating water changed its character from aggressive to

slightly supersaturated with respect to calcite, the hypogene cavities

were transformed into calcite-lined “geodes”(Fig. 8).

230

Th in secular

equilibrium with

234

U indicates that this second speleogenetic stage

Fig. 8. Late-stage evolution of the hypogene caves at Tyuya-Muyun. Estimates of the paleo-water table are given in relation to the different speleothem types: calcite

spar, barite, cave clouds, isopachous calcite, rafts, folia and owstone. (a) Hydrothermal mineralization including deposition of hydrothermal calcite and barite

(~40–60 ◦C) prior to 600 ka. (c) Lowering of the water table, cave cloud formation in Great Barite Cave. Over time the input of meteoric water increased, as indicated

by decreasing δ

(Fig. 8) and cooler water T. With the gradual lowering of the paleo-water table, the mineral deposits and cave walls of the Great Barite Cave were

affected by condensation corrosion. (c) Surprise Cave started to emerge from phreatic conditions ~540–450 ka. Folia and isopachous calcite precipitated from

lukewarm water, while vadose speleothems formed in already air-lled upper parts. (d) Subaqueous and near-water table speleothems formed in the middle and

lower cave levels, tracing the nal dewatering of the cave (~540 to 80 ka). Condensation corrosion affected air-lled passages.

G. Koltai et al. Chemical Geology 670 (2024) 122454

occurred prior to 600 ka.

Large sizes, high degree of transparency and euhedral shapes of

calcite crystals indicate stagnant hydrodynamic conditions and slow

calcite crystal precipitation (e.g., G´

azquez and Calaforra, 2013b) when

general principles of crystal growth are considered (Sunagawa, 1981,

2005). Based on FIM and Δ

thermometry the crystals precipitated at T

around 40–55 ◦C in Surprise Cave and 41–60 ◦C in Great Barite Cave

(Fig. S4). Similar formation T of ca. 44–63 ◦C were measured in calcite

from Fersmana Cave; this calcite, therefore, may be equivalent to the

spar in Great Barite and Surprise caves.

Most calcite spar is relatively depleted in

O similar to hydrothermal

spar elsewhere, reecting elevated T (e.g., G´

azquez et al., 2018;Sp¨

otl

et al., 2021;Koltai et al., 2024). Their δ

C values are variable, with a

narrower range in Great Barite Cave (from −1.8 to 0.1 ‰) compared to

Surprise Cave (−5.9 to 2.0 ‰). The elevated carbon isotope values

suggest that the carbon source of the dissolved inorganic carbon (DIC) in

the water was mostly the Carboniferous limestone in the case of Great

Barite Cave. In contrast, the low δ

C values in Surprise Cave indicate a

second carbon source (e.g., related to the input of soil-derived C).

5.4.3. Stage 2: Hydrothermal barite

Caves located in the central part of the Tyuya-Muyun massif (e.g.,

Fersmana and Great Barite) host barite (Fig. 1b). FIM data suggest that

the T of the barite-forming water was 54.8 ±1.0 ◦C, suggesting no

change compared to the preceding calcite spar. Unlike Great Barite

Cave, Surprise Cave is located outside the zone of barite mineralization

(Fig. 1) and only contains one geode with barite. Barite formation likely

occurred prior to 600 ka, as indicated by the oldest

230

Th ages of iso-

pachous calcite and folia in Surprise Cave, which postdates barite

(Fig. 8).

5.4.4. Stage 3: Lowering of the water table, formation of cave clouds and

folia, dewatering of the caves, and condensation corrosion

At a certain point in time likely before 600 ka, abundant cave clouds

reaching up to 20 cm in thickness formed in Great Barite Cave (Fig. 8).

The cave cloud sequence shows 2.7 ‰and 2.1 ‰variability in δ

C and

O, respectively, indicating only slightly variable hydrochemical

conditions. Furthermore, the signicantly lower ẟ

C values (−7.9 to

−5.5 ‰) compared to calcite spar (−1.8 to −0.1 ‰), and the lower T

(Fig. 7 and Fig. S4) suggest a change in the water source (i.e., a higher

input of likely soil-derived C), which is also supported by the change in

compared to spar (Fig. S4). The decrease in water T (from 29.7 ±

8.3 ◦C to 12.7 ±5.6 ◦C) combined with the change towards lower δ

between the early and the late phase of cave cloud formation also sup-

port a lower contribution of thermal water and a higher contribution of

colder meteoric water over time.

At around 540–450 ka Surprise Cave emerged from the phreatic

zone, as indicated by the highest elevation of folia occurrence. The

decline of the water table separated the cave into a subaqueous and a

subaerial part and resulted in a suite of characteristic morphological and

depositional features (Fig. 8).

Below the water table, an isopachous calcite crust (incipient cave

cloud-type) uniformly coated the substrate, including bedrock, calcite

spar and/or red silty sediment (Supplementary Material). Paleo-water

Table 3

of the paleo-water calculated using calcite δ

O and temperatures derived from both T(Δ

) and FIM using the equation of Da¨

eron et al. (2019).

Cave Sample Description T (±2SE)

based on FIM

(◦C)

T (±95 % CI) based on

thermometry (◦C)

(‰,

VPDB)

(‰,

VPDB)

Calculated δ

±95 %

CI (‰, VSMOW) based on

T(Δ

)

Calculated δ

±2 SE

(‰, VSMOW) based on T-

FIM

Surprise

Cave

SUR-2

Scalenohedral

calcite

39.1 ±1.2

(n =16) 45.2 ±6.0 −2.3 −10.1 −5.7 ±1.0 −6.8 ±0.2

SUR-5

Scalenohedral

calcite



37.1 ±5.0 −3.9 −9.4 −6.5 ±0.9



SUR-

Scalenohedral

calcite

41.0 ±1.4

(n =15) 37.7 ±3.6 −4.4 −10.9 −7.9 ±0.6 −7.3 ±0.2

SUR-

Scalenohedral

calcite

41.5 ±2.0

(n =30)

46.5 ±2.0

(n =9) 49.8 ±11.6 −0.7 −13.3 −8.2 ±1.9

−9.6 ±0.4

−8.7 ±0.4

SUR-

Scalenohedral

calcite

40.5 ±2.0

(n =33)

41.8 ±2.4

(n =42) 47.8 ±6.2 −0.8 −14.6 −9.8 ±1.0

−11.1 ±0.4

−10.9 ±0.4

SUR-

Scalenohedral

calcite



50.1 ±11.2 −0.6 −13.0 −7.8 ±1.8



SUR-

18 Isopachous calcite



9.1 ±6.8 −1.0 −9.7 −12.3 ±1.4



SUR-

14 Isopachous calcite



7.0 ±6.0 −2.9 −9.5 −12.6 ±1.3



SUR-

C3 Isopachous calcite



13.4 ±7.3 −0.7 −9.6 −11.3 ±1.5



SUR-

16 Rafts



14.7 ±6.6 −4.5 −11.6 −13.0 ±1.3



SUR-

C1 Flowstone



14.9 ±4.1 −3.4 −11.7 −13.1 ±0.8



Great Barite

Cave

BAR-7

Scalenohedral

calcite



49.2 ±7.2 −0.6 −14.3 −9.3 ±1.2



BAR-8

Scalenohedral

calcite

44.9 ±2.4

(n=18) 51.4 ±8.7 −0.6 −12.9 −7.5 ±1.4 −8.6 ±0.4

BAR-9

Scalenohedral

calcite

52.3 ±3.0

(n =20) 44.0 ±7.6 −0.8 −10.3 −6.2 ±1.3 −4.8 ±0.5

BAR-

Cave cloud

(early phase)



29.7 ±8.3 −6.6 −11.3 −9.8 ±1.6



BAR-

Cave cloud

(late phase)



12.7 ±5.6 −3.7 −12.5 −14.4 ±1.2



Fersmana

Cave

FER-05 Palisade calcite 51.4 ±7.8 −2.4 −9.8 −4.4 ±1.3 

FER-06

Coarsely

crystalline calcite



54.9 ±8.5 1.0 −10.7 −4.7 ±1.3



G. Koltai et al. Chemical Geology 670 (2024) 122454

table markers, such as folia and rafts, are present in several locations in

the cave, over almost the entire vertical extent of the cave (ca. 80 m).

thermometry shows an approximately 20–40 ◦C decrease in T (7.0

±6.8 ◦C and 14.7 ±6.6 ◦C) compared to spar formation (Fig. S4).

230

ages of folia, isopachous calcite and rafts record the successive lowering

of the water table from about 600 ka to ca. 84 ka, which may have been

related to the uplift of the Tyuya-Muyun massif and the incision of the

Aravan River forming the Dangi Gorge.

At the same time (495 ±39 to 476 ±31 ka), a owstone formed in

the lower part of the ascending Crystal Chimney. This indicates sub-

aerial conditions following the dewatering of the cave after 493 ±19 ka,

marked by folia formation covering the overhanging wall of this chim-

ney opposite of the owstone. The owstone (14.9 ±4.1 ◦C) formed at a

similar T as isopachous calcite (7.0 ±6.8 ◦C to 13.4 ±7.3 ◦C) that was

deposited at ~350 and 230 ka (Surprise Cave) and the outer layers of

cave clouds in Great Barite Cave (12.7 ±5.6 ◦C, Fig. S4). Furthermore,

owstone calcite yielded similar mean δ

O values (−10.2 ‰) as a Late

Holocene stalagmite from Uluu-too cave (mean −10.8 ‰), located ca.

22 km west of Surprise Cave (Wolff et al., 2017).

Above the water table, bedrock and calcite spar were affected by

condensation corrosion (Fig. 8) as evidenced by the corrosive smoothing

of the surfaces of calcite crystals and tabular barite. This process did not

result in well-developed solutional cupolas and domes, nor did it

massively dissolve the limestone substrate. We attribute the limited

extent of condensation corrosion to the low thermal gradient, as indi-

cated by the close-to-ambient water T. This gradient may have been

larger during glacial times, when the cave air T was likely lower.

Condensation corrosion also affected the cave walls in Great Barite Cave

(Fig. 3d), but likely at earlier times.

6. Conclusions

This study reports systematic geothermometric (Δ

and FIM) and

geochemical analyses of the mineral inll of Surprise and Great Barite

caves in the Tyuya-Muyun massif. These mineral deposits (calcite spar,

barite, cave clouds, rafts, folia and owstone) record changes in the

paleo-water table, T and the δ

O composition of the paleo-water.

Together with morphological observations, these geochemical data

suggest the following late-stage speleogenetic evolution: (1) Subaqueous

calcite spar and barite formed prior to 600 ka at T around ~40–60 ◦C

(likely below 55 ◦C based on FIM) from water with slightly higher δ

than the modern lukewarm springs in the area. (2) Cave cloud formed in

Great Barite Cave in shallow phreatic conditions and record the cooling

of the paleo-water to 12.7 ±5.6 ◦C. Lower δ

indicates an increase in

meteoric water input during the later stage of cave cloud formation. This

is corroborated by a decrease in δ

C as compared to calcite spar, sug-

gesting an increase in soil-derived carbon. (3) During the subsequent

lowering of the water table, the T of the paleo-water remained lukewarm

and the upper chambers of Surprise Cave became partially air-lled, as

indicated by owstone formation following folia deposition at 493 ±19

ka. At the same time, isopachous calcite continued to form subaqueously

in the lower levels of the cave. (4) Near-water table deposits (folia and

rafts) in Surprise Cave trace the dewatering of the Tyuya-Muyun massif

until 84 ka. This process was likely related to the nal phase of uplift of

the massif and the associated incision of the Aravan River.

Author contribution

YD, CS and GK designed the study. YD, CS, GK, CH and TR carried out

eldwork. GK performed FIM with the help of YK and YD. LR performed

clumped isotope analyses. GK and HZ dated the calcite samples using the

230

Th dating method. TR created the 3D cave surveys. GK wrote the

manuscript with major contributions by YD and CS and additional

support from the other co-authors. GK, CH, TR, YD and CS designed the

gures.

Funding sources

Financial support for eldwork was provided by the Austrian

Academy of Science (E. Suess Legacy Program) and for NA-FIM analysed

was provided by the Austrian Academy of Sciences (Hlauscheck Legat

Program). NA-FIM analyses was provided by the European Reserch

Council (grant no. 101001957).

CRediT authorship contribution statement

Gabriella Koltai: Writing –original draft, Methodology, Investiga-

tion, Funding acquisition, Conceptualization. Christoph Sp¨

otl: Writing

–original draft, Visualization, Supervision, Funding acquisition,

Conceptualization. L´

aszl´

o Rinyu: Writing –original draft, Investiga-

tion, Formal analysis. Charlotte Honiat: Writing –original draft,

Visualization, Investigation. Tanguy Racine: Writing –original draft,

Visualization, Investigation. Haiwei Zhang: Writing –original draft,

Formal analysis. Yves Krüger: Writing –original draft, Supervision,

Formal analysis. Yuri Dublyansky: Writing –original draft, Visualiza-

tion, Supervision, Investigation, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing nancial

interests or personal relationships that could have appeared to inuence

the work reported in this paper.

Data availability

Data will be made available on request.

Acknowledgements

A. Dudashvili of the Foundation for the Preservation and Exploration

of Caves (Kyrgyzstan) is acknowledged for his support during eldwork.

The authors are grateful to Fernando G´

azquez and two anonymous re-

viewers who helped to improve the manuscript. We thank M. Steck, D.

Sperlich and M. Wimmer for assistance during stable isotope analyses.

We thank K. Wendt for measuring one

230

Th age. J. A. Friedrich and H.

Mihatsch are acknowledged for their assistance during uid inclusion

petrography.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.

org/10.1016/j.chemgeo.2024.122454.

References

Anderson, N.T., Kelson, J.R., Kele, S., Da¨

eron, M., Bonifacie, M., Horita, J., Mackey, T.J.,

John, C.M., Kluge, T., Petschnig, P., Jost, A.B., Huntington, K.W., Bernasconi, S.M.,

Bergmann, K.D., 2021. A unied clumped isotope thermometer calibration

(0.5–1,100◦C) using carbonate-based standardization. Geophys. Res. Lett. 48.

https://doi.org/10.1029/2020GL092069 e2020GL092069.

Bajnai, D., Guo, W., Sp¨

otl, C., Coplen, T.B., Methner, K., L¨

ofer, N., Krsnik, E.,

Gischler, E., Hansen, M., Henkel, D., Price, G.D., Raddatz, J., Scholz, D., Fiebig, J.,

2020. Dual clumped isotope thermometry resolves kinetic biases in carbonate

formation temperatures. Nat. Commun. 11, 4005. https://doi.org/10.1038/s41467-

020-17501-0.

Bajnai, D., Coplen, T.B., Methner, K., L¨

ofer, N., Krsnik, E., Fiebig, J., 2021. Devils Hole

calcite was precipitated at ±1◦C stable aquifer temperatures during the last half

million years. Geophys. Res. Lett. 48. https://doi.org/10.1029/2021GL093257

e2021GL093257.

Bernasconi, S.M., Hu, B., Wacker, U., Fiebig, J., Breitenbach, S.F.M., Rutz, T., 2013.

Background effects on Faraday collectors in gas-source mass spectrometry and

implications for clumped isotope measurements. Rapid Commun. Mass Spectrom.

27, 603–612. https://doi.org/10.1002/rcm.6490.

Bernasconi, S.M., Da¨

eron, M., Bergmann, K.D., Bonifacie, M., Meckler, A.N., Affek, H.P.,

Anderson, N., Bajnai, D., Barkan, E., Beverly, E., Blamart, D., Burgener, L.,

Calmels, D., Chaduteau, C., Clog, M., Davidheiser-Kroll, B., Davies, A., Dux, F.,

Eiler, J., Elliott, B., Fetrow, A.C., Fiebig, J., Goldberg, S., Hermoso, M.,

G. Koltai et al. Chemical Geology 670 (2024) 122454

Huntington, K.W., Hyland, E., Ingalls, M., Jaggi, M., John, C.M., Jost, A.B., Katz, S.,

Kelson, J., Kluge, T., Kocken, I.J., Laskar, A., Leutert, T.J., Liang, D., Lucarelli, J.,

Mackey, T.J., Mangenot, X., Meinicke, N., Modestou, S.E., Müller, I.A., Murray, S.,

Neary, A., Packard, N., Passey, B.H., Pelletier, E., Petersen, S., Piasecki, A.,

Schauer, A., Snell, K.E., Swart, P.K., Tripati, A., Upadhyay, D., Vennemann, T.,

Winkelstern, I., Yarian, D., Yoshida, N., Zhang, N., Ziegler, M., 2021. InterCarb: a

community effort to improve interlaboratory standardization of the carbonate

clumped isotope thermometer using carbonate standards. Geochem. Geophys.

Geosyst. 22. https://doi.org/10.1029/2020GC009588 e2020GC009588.

Brand, W.A., Assonov, S.S., Coplen, T.B., 2010. Correction for the

O interference in

δ(

C) measurements when analyzing CO

with stable isotope mass spectrometry

(IUPAC Technical Report). Pure Appl. Chem. 82, 1719–1733. https://doi.org/

10.1351/PAC-REP-09-01-05.

Butov, P.I., Zaitsev, I.K., 1935. O gidrogeologii i karste Tyuya-Muyuna (on hydrogeology

and karst of Tyuya-Muyun). In: Notes of All-Union Mineralogical Society, 2, 4, 1,

Leningrad,, pp. 145-131 (in Russian).

Cheng, H., Edwards, L.R., Shen, C.-C., Polyak, V.J., Asmerom, Y., Woodhead, J.,

Hellstrom, J., Wang, Y., Kong, X., Sp¨

otl, C., Wang, X., Calvin Alexander, E., 2013.

Improvements in

230

Th dating,

230

Th and

234

U half-life values, and U-Th isotopic

measurements by multi-collector inductively coupled plasma mass spectrometry.

Earth Planet. Sci. Lett. 371–372, 82–91. https://doi.org/10.1016/j.

epsl.2013.04.006.

Da¨

eron, M., Drysdale, R.N., Peral, M., Huyghe, D., Blamart, D., Coplen, T.B., Lartaud, F.,

Zanchetta, G., 2019. Most Earth-surface calcites precipitate out of isotopic

equilibrium. Nat. Commun. 10, 1–7. https://doi.org/10.1038/s41467-019-08336-5.

Dahlkamp, F.J., 2010. Uranium Deposits of the World: Asia. Springer, Berlin.

De Waele, J., Forti, P., Naseddu, A., 2013. Speleogenesis of an exhumed hydrothermal

sulphuric acid karst in Cambrian carbonates (Mount San Giovanni, Sardinia). Earth

Surf. Process. Landf. 38 (12), 1369–1379.

Decker, D.D., Polyak, V.J., Lachniet, M., 2018. U–Pb dating of cave spar: a new shallow

crust landscape evolution tool. Tectonics 37, 208–233. https://doi.org/10.1002/

2017TC004675.

Dem´

eny, A., Berent´

es, A., Rinyu, L., Kov´

acs, I., Sur´

anyi, G., Vir´

ag, M., 2024. Subaqueous

carbonate speleothems as paleotemperature archives –clumped isotope

thermometry and stable isotope compositions of inclusion-hosted water. Int. J.

Speleol. 53 (1), 25–37. https://doi.org/10.5038/1827-806X.53.1.2480.

Dreybrodt, W., 2019. Kinetic fractionation of the isotope composition of

C, and of

clumped isotope

C in calcite deposited to speleothems. Implications to the

reliability of the

O and Δ

paleothermometers. Acta Carsologica 48 (3). https://

doi.org/10.3986/ac.v48i3.7710.

Dublyansky, Y.V., 2000. Dissolution of Carbonates by Geothermal Waters, Speleogenesis:

Evolution of Karst Aquifers. National Speleological Society, Huntsville, Alabama,

USA.

Dublyansky, Y., Hevesi, A., Hromas, J., Kraus, S., Mahanykova, V., Mikhailov, V.,

Mucke, D., Sanykova, V., Sz´

ekely, K., T.Bolner K., 1989. Tyuya-Muyun (USSR,

Kirghizia - preliminary results of the speleological researches carried out by the

International Expedition in 1989). In: 10

Int. Congr. Speleol. Budapest, Hungary, 3,

pp. 825–830.

Dublyansky, Y., Roncat, A., Sp¨

otl, C., Dorninger, P., 2016. Hypogene cave morphology at

high resolution: Full 3-D survey of M¨

archenh¨

ohle (Austria). In: Chavez, T.,

Reehling, P. (Eds.), Proceedings of DEEPKARST 2016: Origins, Resources, and

Management of Hypogene Karst. National Cave and Karst Research Institute,

Carlsbad, New Mexico, pp. 183–187.

Dublyansky, Y., Michajljow, W., Bolner-Tak´

acs, K., Hromas, J., Sz´

ekely, K., Hevesi, A.,

Kraus, S., 2017. Hypogene Karst in the Tyuya-Muyun and the Kara-Tash Massifs

(Kyrgyzstan). In: Klimchouk, A., Palmer, A.N., De Waele, J., Auler, A.S., Audra, P.

(Eds.), Hypogene Karst Regions and Caves of the World. Springer, pp. 495–507.

https://doi.org/10.1007/978-3-319-53348-3_30.

Edwards, R.L., Chen, J.H., Wasserburg, G.J., 1987.

238

U–

234

U–

230

Th–

232

Th systematics

and the precise measurement of time over the past 500,000 years. Earth Planet. Sci.

Lett. 81, 175–192. https://doi.org/10.1016/0012-821X(87)90154-3.

Eiler, J.M., 2011. Paleoclimate reconstruction using carbonate clumped isotope

thermometry. Quat. Sci. Rev. 30, 3575–3588. https://doi.org/10.1016/j.

quascirev.2011.09.001.

Fersman, A.E., 1927. K Morfologii I Geokhimii Tyuya-Muyuna (on the Morphology and

Geochemistry of Tyuya-Muyun). Works on Studies of Radium and Radioactive Ores,

vol. 3. Academy of Sciences of the USSR, Leningrad, pp. 1–90 (in Russian).

Fersman, A.E., 1928. Uran-vanadieviy rudnik Tyuya-Muyun v Turkestane (Uranium-

vanadium mine Tyuya-Muyun in Turkestan). Works of Geological Committee. (in

Russian).

Fersman, A.E., Scherbakov, L.I., 1925. Tyuya-Muyunskoye mestorozhdeniye radievikh

rud v Fergane (the Tyuya-Muyun deposit of radium ores in Fergana). Works NTO 74

(in Russian).

Fiebig, J., Sp¨

otl, C., Markowska, M., Koltai, G., Vonhof, H., Moreno-Caballud, A.,

Zhang, H., Peral, M., Bernecker, M., Staudigel, P., Guo, W., 2023. Dual clumped

isotope thermometry identies kinetically unbiased speleothems for reliable (paleo)

temperature reconstructions. In: AGU Fall Meeting 2023, San Francisco, CA,

2023AGUFMPP14A..07F.

G´

azquez, F., Calaforra, J.-M., 2013a. Origin of double-tower raft cones in hypogenic

caves. Earth Surf. Process. Landf. 38, 1655–1661. https://doi.org/10.1002/

esp.3399.

G´

azquez, F., Calaforra, J.-M., 2013b. Hypogenic speleogenesis and speleothems of Sima

de la Higuera Cave (Murcia, South-eastern Spain). In: Filippi, M., Bos´

ak, P. (Eds.),

Proceedings 16th Intern. Congress of Speleol, 3, pp. 78–83.

G´

azquez, F., Columbu, A., De Waele, J., Breitenbach, S.F.M., Huang, C.-R., Shen, C.-C.,

Lu, Y., Calaforra, J.-M., Mleneck-Vautravers, M.J., Hodell, D.A., 2018.

Quantication of paleo-aquifer changes using clumped isotopes in subaqueous

carbonate speleothems. Chem. Geol. 493, 246–257. https://doi.org/10.1016/j.

chemgeo.2018.05.046.

Goldstein, R.H., Reynolds, T.J., 1994. Systematics of Fluid Inclusions in Diagenetic

Minerals. SEPM Short Course 31, 1–198. https://doi.org/10.2110/scn.94.31.

Guo, W., 2020. Kinetic clumped isotope fractionation in the DIC-H

O-CO

system:

patterns, controls, and implications. Geochim. Cosmochim. Acta 230-257. https://

doi.org/10.1016/j.gca.2019.07.055.

Hansen, M., Kluge, T., Scholz, D., 2022. Investigating clumped isotope disequilibrium

effects for DIC and CaCO

during formation of (speleothem) CaCO3: insights from

cave analogous laboratory experiments. Geochim. Cosmochim. Acta 321, 244–264.

https://doi.org/10.1016/j.gca.2021.12.019.

Hill, C.A., Forti, P., 1997. Cave Minerals of the World, 2

edition. Alabama (National

Speleological Society), Huntsville.

Hill, C.A., Polyak, V.J., 2010. Karst hydrology of Grand Canyon, Arizona, USA. J. Hydrol.

390, 169–181. https://doi.org/10.1016/j.jhydrol.2010.06.040.

John, C.M., Bowen, D., 2016. Community software for challenging isotope analysis: rst

applications of ‘Easotope’to clumped isotopes. Rapid Commun. Mass Spectrom. 30,

2285–2300. https://doi.org/10.1002/rcm.7720.

Kazansky, V.I., 1970. Geologo-Strukturniye Osobennosti Mestorozhdeniya Tyuya-Muyun

(Geological and Structural Particularities of the Tyuya-Muyun Deposit). Notes on

Geology and Geochemistry of Ore Deposits. Nauka, Moscow, pp. 34–57 (in Russian).

Kirikov, A.P., 1929. Tyuya-Muyunskoje Mestorozhdeniye Ragiya (The Tyuya-Muyun

Deposit of Radium) Works of Geological Committee, New Series, 181 (in Russian).

Klimchouk, A., 2009. Morphogenesis of hypogenic caves. Geomorphology 106, 100–117.

https://doi.org/10.1016/j.geomorph.2008.09.013.

Klimchouk, A.B., Ford, D.C., Palmer, A.N., Dreybrodt, W., 2000. Speleogenesis. Evolution

of Karst Aquifers. Huntsville (National Speleol. Soc.).

Kluge, T., Affek, H.P., 2012. Quantifying kinetic fractionation in Bunker Cave

speleothems using Δ

. Quat. Sci. Rev. 49, 82–94. https://doi.org/10.1016/j.

quascirev.2012.06.013.

Koltai, G., Kluge, T., Krüger, Y.K., Sp¨

otl, C., Rinyu, L., Audra, P.H., Honiat, C., Le´

el-

Ossy, S.Z., Dublyansky, Y., 2024. Geothermometry of calcite spar at 10–50 ◦C. Sci.

Rep. 14, 1553. https://doi.org/10.1038/s41598-024-51937-4.

Krüger, Y., Stoller, P., Riˇ

cka, J., Frenz, M., 2007. Femtosecond lasers in uid-inclusion

analysis: overcoming metastable phase states. Eur. J. Mineral. 19, 693–706. https://

doi.org/10.1127/0935-1221/2007/0019-1762.

Løland, M., Krüger, Y., Fernandez, A., Buckingham, F., Carolin, S.A., Sodemann, H.,

Adkins, J.F., Cobb, K.M., Meckler, A.N., 2022. Evolution of tropical l nature

communications and temperature across the last glacial termination. Nat. Comm. 13,

5158. https://doi.org/10.1038/s41467-022-32712-3.

Malinin, S., 1979. Physical Chemistry of Hydrothermal Systems with Carbon Dioxde.

Nauka, Moscow (in Russian).

Marti, D., Krüger, Y., Fleitmann, D., Frenz, M., Riˇ

cka, J., 2012. The effect of surface

tension on liquid-gas equilibria in isochoric systems and its application to uid

inclusions. Fluid Phase Equilib. 314, 13–21. https://doi.org/10.1016/j.

uid.2011.08.010.

Matthews, A., Affek, H.P., Ayalon, A., Vonhof, H.B., Bar-Matthews, M., 2021. Eastern

Mediterranean climate change deduced from the Soreq Cave uid inclusion stable

isotopes and carbonate clumped isotopes record of the last 160 ka. Quat. Sci. Rev.

272, 107223.

Meckler, A.N., Affolter, S., Dublyansky, Y.V., Krüger, Y., Vogel, N., Bernasconi, S.M.,

Frenz, M., Kipfer, R., Leuenberger, M., Sp¨

otl, C., Carolin, S., Cobb, K.M.,

Moerman, J., Adkins, J.F., Fleitmann, D., 2015. Glacial-interglacial temperature

change in the tropical West Pacic: a comparison of stalagmite-based paleo-

thermometers. Quat. Sci. Rev. 127, 90–116. https://doi.org/10.1016/j.

quascirev.2015.06.015.

Parvez, Z.A., El-Shenawy, M.I., Lucarelli, J.K., Kim, S.-T., Johnson, K.R., Wright, K.,

Gebregiorgis, D., Montanez, I.P., Wortham, B., Asrat, A., Reinhardt, E.,

Christensen, J.N., Matamoros, I.W., Rubi, J., Miguel, K., Elliott, B.M., Flores, R.,

Kovacs, S., Eagle, R.A., Tripati, A., 2024. Dual carbonate clumped isotope (Δ47-Δ48)

measurements constrain different sources of kinetic isotope effects and quasi-

equilibrium signatures in cave carbonates. Geochim. Cosmochim. Acta 366, 95–112.

https://doi.org/10.1016/j.gca.2023.11.017.

Pavlenko, D.M., 1933. Noviye danniye o geologii i genezise mestorozhdeniya Tyuya-

Muyun v Uzbekistane (New data on geology and origin of the Tyuya-Muyun deposit

in Uzbekistan). Probl. Soviet Geol. 10, 123–142 (in Russian).

Pogodin, S.A., Libman, E.P., 1971. Kak Dobyli Sovetskiy Radiy (how Soviet Radium Was

Acquired). Atomizdat, Moscow (in Russian).

Riggs, A.C., Carr, W.J., Kolesar, P.T., Hoffman, R.J., 1994. Tectonic speleogenesis of

Devils Hole, Nevada, and implications for hydrogeology and the development of

long, continuous paleoenvironmental records. Quat. Res. 42 (3), 241–254.

Scherbakov, D.I., 1924. Mestorozhdeniya Radioaktivnikh Rud I Mineralov Fergany I

Zadachi Ikh Dalneyshego Issledovaniya (Deposits of Radioactive Ores and Minerals

in Fergana and Problems of their Further Studies). Materials of Studies of Natural

Productive Forces of Russia. Academy of Sciences of the USSR (in Russian).

Smolyaninova, N.N., 1970- Nekotoriye danniye po mineralogii I genezisu

mestorozhdeniya Tyuya-Muyun (Some data on mineralogy and origin of the Tyuya-

Muyun deposit). Articles on geology and geochemistry of ore deposits. Nauka,

Moscow, pp 58–90. (in Russian).

Smolyaninova, N.N., 1970. Nekotoriye danniye po mineralogii i genezisu

mestorozhdeniya Tyuya-Muyun (some data on mineralogy and origin of the Tyuya-

Muyun deposit). In: Notes on Geology and Geochemistry of Ore Deposits. Nauka,

Moscow, pp. 58–90 (in Russian).

G. Koltai et al. Chemical Geology 670 (2024) 122454

Sp¨

otl, C., Vennemann, T.W., 2003. Continuous-ow isotope ratio mass spectrometric

analysis of carbonate minerals. Rapid Commun. Mass Spectrom. 17, 1004–1006.

https://doi.org/10.1002/rcm.1010.

Sp¨

otl, C., Dublyansky, Y., Koltai, G., Honiat, C., Plan, L., Angerer, T., 2021. Stable isotope

imprint of hypogene speleogenesis: Lessons from Austrian caves. Chem. Geol. 572,

1–19. https://doi.org/10.1016/j.chemgeo.2021.120209.

Sunagawa, I., 1981. Characteristics of crystal growth in nature as seen from the

morphology of mineral crystals. Bull. Mineral. 104, 81–87.

Sunagawa, I., 2005. Crystals: Growth, Morphology and Perfection. Cambridge University

Press.

Uklonsky, A.S., 1927. Geokhimicheskaya Kharakteristika Vod Tyuya-Muyuna

(Geochemical Characteristic of Waters of the Tyuya-Muyun). Works on Studies of

Radium and Radioactive Ores. Academy of Sciences of the USSR, Leningrad,

pp. 111–126 (in Russian).

Vernadsky, V.I., 1914. O neobkhodimosti issledovaniya radioaktivnikh mineralov v

Rossiyskoy Imperii (On the need to study radioactive minerals of the Russian

Empire). Works of Radium expedition of the Russian Academy of Sciences. Imperial

Academy of Sciences, Petrograd. (in Russian).

Wainer, K., Genty, D., Blamart, D., Da¨

eron, M., Bar-Matthews, M., Vonhof, H.,

Dublyansky, Y., Pons-Branchu, E., Thomas, L., van Caltersen, P., Quinif, Y.,

Caillon, N., 2011. Speleothem record of the last 180 ka in Villars cave (SW France):

investigation of a large δ

O shift between MIS6 and MIS5. Quat. Sci. Rev. 30 (1–2),

130–146. https://doi.org/10.1016/j.quascirev.2010.07.004.

Wolff, C., Plessen, B., Dudashvilli, A.S., Breitenbach, S.F., Cheng, H., Edwards, L.R.,

Strecker, M.R., 2017. Precipitation evolution of Central Asia during the last 5000

years. The Holocene 27 (1), 142–154. https://doi.org/10.1177/

0959683616652711.

Zaitsev, I.K., 1940. Voprosy Izucheniya Karsta V SSSR (Issues of Karst studies in the

USSR). Gosgeolizdat, Leningrad-Moscow (in Russian).

G. Koltai et al. Chemical Geology 670 (2024) 122454

ResearchGate has not been able to resolve any citations for this publication.

Geothermometry of calcite spar at 10–50 °C

Article

Full-text available

Jan 2024

Carbonate geothermometry is a fundamental tool for quantitative assessment of the geothermal and geochemical evolution of diagenetic and hydrothermal systems, but it remains difficult to obtain accurate and precise formation temperatures of low-temperature calcite samples (below ~ 40 to 60 °C). Here, we apply three geothermometry methods (∆47-thermometry, nucleation-assisted fluid inclusion microthermometry—hereafter NA-FIM—and oxygen isotope thermometry) to slow-growing subaqueous calcite spar samples to cross-validate these methods down to 10 °C. Temperatures derived by NA-FIM and Δ47-thermometry agree within the 95% confidence interval, except for one sample. Regression analyses suggest that the real uncertainty of ∆47-thermometry exceeds the 1 SE analytical uncertainty and is around ± 6.6 °C for calcite spar that formed at 10–50 °C. The application of δ¹⁸O thermometry was limited to a few samples that contained sufficient primary fluid inclusions. It yielded broadly consistent results for two samples with two other geothermometers, and showed higher temperature for the third spar. We also found that calcite with steep rhombohedral morphologies is characteristic of low temperatures (11–13 °C), whereas blunt rhombohedra prevail in the 10–29 °C domain, and the scalenohedral habit dominates > 30 °C. This suggests that the calcite crystal morphology can be used to qualitatively distinguish between low- and higher-temperature calcite.

Evolution of tropical land temperature across the last glacial termination

Article

Full-text available

Sep 2022

The tropical West Pacific hosts the warmest part of the surface ocean and has a considerable impact on the global climate system. Reconstructions of past temperature in this region can elucidate climate connections between the tropics and poles and the sensitivity of tropical temperature to greenhouse forcing. However, existing data are equivocal and reliable information from terrestrial archives is particularly sparse. Here we constrain the magnitude and timing of land temperature change in the tropical West Pacific across the last deglaciation using an exceptionally precise paleothermometer applied to a well-dated stalagmite from Northern Borneo. We show that the cave temperature increased by 4.4 ± 0.3 °C (2 SEM) from the Last Glacial Maximum to the Holocene, amounting to 3.6 ± 0.3 °C (2 SEM) when correcting for sea-level induced cave altitude change. The warming closely follows atmospheric CO 2 and Southern Hemisphere warming. This contrasts with hydroclimate, as reflected by drip water δ ¹⁸ O, which responds to Northern Hemisphere cooling events in the form of prominent drying, while temperature was rising. Our results thus show a close response of tropical temperature to greenhouse forcing, independent of shifts in the tropical circulation patterns.

Devils Hole Calcite Was Precipitated at ±1°C Stable Aquifer Temperatures During the Last Half Million Years

Article

Full-text available

Jun 2021
GEOPHYS RES LETT

Subaqueous carbonates from the Devils Hole caves (southwestern USA) provide a continuous Holocene to Pleistocene North American paleoclimate record. The accuracy of this record relies on two assumptions: That carbonates precipitated close to isotope equilibrium and that groundwater temperature did not change significantly in the last 570 thousand years. Here, we investigate these assumptions using dual clumped isotope thermometry. This method relies on simultaneous analyses of carbonate ∆47 and ∆48 values and provides information on the existence and extent of kinetic isotope fractionation. Our results confirm the hypothesis that calcite precipitation occurred close to oxygen and clumped isotope equilibrium during the last half million years in Devils Hole. In addition, we provide evidence that aquifer temperatures varied by less than ±1°C during this interval. Thus, the Devils Hole calcite δ18O time series exclusively represents changes in groundwater δ18O values.

InterCarb: A Community Effort to Improve Interlaboratory Standardization of the Carbonate Clumped Isotope Thermometer Using Carbonate Standards

Article

Full-text available

May 2021
GEOCHEM GEOPHY GEOSY

Increased use and improved methodology of carbonate clumped isotope thermometry has greatly enhanced our ability to interrogate a suite of Earth‐system processes. However, inter‐laboratory discrepancies in quantifying carbonate clumped isotope (Δ₄₇) measurements persist, and their specific sources remain unclear. To address inter‐laboratory differences, we first provide consensus values from the clumped isotope community for four carbonate standards relative to heated and equilibrated gases with 1,819 individual analyses from 10 laboratories. Then we analyzed the four carbonate standards along with three additional standards, spanning a broad range of δ⁴⁷ and Δ₄₇ values, for a total of 5,329 analyses on 25 individual mass spectrometers from 22 different laboratories. Treating three of the materials as known standards and the other four as unknowns, we find that the use of carbonate reference materials is a robust method for standardization that yields inter‐laboratory discrepancies entirely consistent with intra‐laboratory analytical uncertainties. Carbonate reference materials, along with measurement and data processing practices described herein, provide the carbonate clumped isotope community with a robust approach to achieve inter‐laboratory agreement as we continue to use and improve this powerful geochemical tool. We propose that carbonate clumped isotope data normalized to the carbonate reference materials described in this publication should be reported as Δ₄₇ (I‐CDES) values for Intercarb‐Carbon Dioxide Equilibrium Scale.

Stable isotope imprint of hypogene speleogenesis: Lessons from Austrian caves

Article

Full-text available

Mar 2021
CHEM GEOL

Hypogene speleogenesis involves the formation of solution-enlarged permeability structures driven by ascending water acquiring its acidity from deep-seated sources. This process occurs in a wide range of geological settings and in different lithologies. Although hypogene speleogenesis has received increasing attention in the last two decades, the identification of former hypogene activity in caves remains challenging, unless unambiguous morphological, mineralogical and/or hydrogeological evidence is available. Here, we document that geochemical analyses (mainly stable oxygen and carbon isotopes) of cave wall rock samples may fill this methodological gap and provide a useful tool to identify hypogene karstification through geochemical fingerprinting of water-rock interaction. Stable isotope profiles of a comprehensive set of wall rock cores from eight hypogene caves in Austria in both limestone and dolomite settings were investigated. Alteration halos typically a few centimeters wide and characterized by up to 12‰ and 15‰ depletion in δ¹³C and δ¹⁸O, respectively, were identified in limestone and calcite marble caves. Besides isotopic alteration, bleaching and/or staining of the wall rock was observed due to oxidizing conditions of the paleowater. No isotopic change of the wall rock was identified in hypogene caves carved in dolomite, which may be attributed to low Ca/Mg ratios of the paleowater and/or aggressive fluids leading to fast wall rock retreat. On the other hand, dedolomitization was observed in some wall rocks implying the presence paleowaters with elevated Ca/Mg ratios. Similarly to dolomitic caves, no isotopic alteration front was observed in a cave formed by sulfuric acid speleogenesis. A likely explanation is the high rate of wall rock retreat outpacing the rate of isotope exchange in the wall rock. Our study also shows that water-rock interaction during epigene speleogenesis does not lead to isotopic halos. This geochemical fingerprinting approach offers a powerful tool to test the hypogene origin of caves and may provide additional insights into the temperature and provenance of the paleowater and the source(s) of its dissolved inorganic carbon.

A Unified Clumped Isotope Thermometer Calibration (0.5–1,100°C) Using Carbonate‐Based Standardization

Article

Full-text available

Apr 2021
GEOPHYS RES LETT

Plain Language Summary Carbonate clumped isotope thermometry is a geochemical tool used to determine the formation temperature of carbonate minerals. In contrast to previous carbonate thermometers, clumped isotope thermometry requires no assumptions about the isotopic composition of the fluid from which the carbonate precipitated. By measuring the clumped isotope composition (Δ47) of carbonate minerals with a known formation temperature, we can construct an empirical calibration for the clumped isotope thermometer that is necessary to convert from a Δ47 value to formation temperature. Many previous studies have created Δ47 temperature calibrations, but differences between calibrations have led to large uncertainty in final Δ47 temperatures. This study measures a large number of samples that span a wide range of temperatures (0.5–1,100°C) and include many different types of carbonates. These data show that a single calibration equation can describe many sample types and that when data are carefully standardized to a common set of carbonate materials, calibrations performed at different laboratories agree almost identically. We combine these data to present a carbonate clumped isotope thermometer calibration with broad applicability across the geosciences.

Dual clumped isotope thermometry resolves kinetic biases in carbonate formation temperatures

Article

Full-text available

Aug 2020

Surface temperature is a fundamental parameter of Earth's climate. Its evolution through time is commonly reconstructed using the oxygen isotope and the clumped isotope compositions of carbonate archives. However, reaction kinetics involved in the precipitation of carbonates can introduce inaccuracies in the derived temperatures. Here, we show that dual clumped isotope analyses, i.e., simultaneous Δ 47 and Δ 48 measurements on the single carbonate phase, can identify the origin and quantify the extent of these kinetic biases. Our results verify theoretical predictions and evidence that the isotopic disequilibrium commonly observed in speleothems and scleractinian coral skeletons is inherited from the dissolved inorganic carbon pool of their parent solutions. Further, we show that dual clumped isotope thermometry can achieve reliable palaeotemperature reconstructions, devoid of kinetic bias. Analysis of a belemnite rostrum implies that it precipitated near isotopic equilibrium and confirms the warmer-than-present temperatures during the Early Cretaceous at southern high latitudes.

Subaqueous carbonate speleothems as paleotemperature archives – clumped isotope thermometry and stable isotope compositions of inclusion-hosted water

Article

Feb 2024

Clumped isotope measurements of carbonates and stable isotope analyses of water trapped in fluid inclusions are both promising techniques to determine carbonate formation temperatures. Cave-hosted carbonate deposits (speleothems) are excellent targets for such studies, but kinetic fractionations and diagenetic influences frequently deteriorate the temperature data obtained from these methods. However, subaqueous carbonate deposits may provide reliable data, as kinetic fractionations are less significant in underwater environments. In this study, subaqueous speleothems, whose formation temperatures were directly measured in the water, were investigated. Additionally, temperatures calculated from the oxygen isotope fractionations between calcite and fluid inclusion-hosted water were compared with clumped isotope temperatures obtained for subaqueous carbonate formations in cave-hosted lakes. The clumped isotope temperatures fit the measured and calculated fluid inclusion temperatures within the analytical precisions. Carbonate deposits formed at elevated temperatures (~50°C or above) may undergo post-formational calcite-water oxygen isotope exchange, altering the composition of the inclusion-hosted water. In contrast, subaqueous speleothems formed at about 20-25°C appear to preserve the primary isotopic compositions. Our study shows that subaqueous carbonate speleothems are useful targets for clumped isotope and inclusion water analyses, making them valuable paleotemperature archives.

Paired Δ47-Δ48 measurements constrain different sources of kinetic isotope effects and quasi-equilibrium signatures in cave carbonates

Article

Nov 2023
GEOCHIM COSMOCHIM AC

Eastern Mediterranean climate change deduced from the Soreq Cave fluid inclusion stable isotopes and carbonate clumped isotopes record of the last 160 ka

Article

Nov 2021
QUATERNARY SCI REV

A fundamental issue in the interpretation of speleothem calcite δ¹⁸Occ records is the correct partitioning of the effects of temperature and water δ¹⁸O variations. This study explores the paleo-environmental evolution of the Eastern Mediterranean Sea (EMS) region using Soreq Cave speleothems in the last 160 ka, a period covering glacial MIS6 to the present, through the independent determinations of temperatures via carbonate clumped isotopes (Δ47) and fluid inclusion δ¹⁸Ow, δDw and d-excess values of cave water entrapped in the speleothems. The observed general agreement between the trends in water δ¹⁸Ow composition directly measured from fluid inclusions and calculated from Δ47 temperatures and δ¹⁸Occ support the robustness of the data. Δ47 temperatures, derived using a revised thermometer calibration that includes recent modifications in ¹⁷O correction, vary from high values of 20–22 °C during interglacial periods to low values of 11–12 °C in cold glacial and stadial periods. Temperatures of 19–22 °C are observed during the time of deposition in the EMS of sapropel S5 in the last interglacial MIS5e period and sapropel S1 in the early Holocene, immediately following Terminations II and I, respectively. The Soreq Cave Δ47 temperatures generally agree with previously estimated EM sea surface temperatures, in support of the absence of a significant land-sea temperature gradient. Cave water δ¹⁸Ow values show large variations (from ∼-3 to −7.5‰), with high glacial and low interglacial values. These variations are largely explained by changes in Mediterranean Sea water composition due to glacial ice melting and freshwater input during sapropel periods, combined with the effect of rainfall amounts during interglacial periods. An analysis of relative timing shows that temperature rise precedes hydrological changes in the initial stages of Termination events. Fluid inclusion d-excess values show clear variations between values similar to the Global Meteoric Water Line during cold periods and values similar to the local Eastern Mediterranean Meteoric Water Line (EMMWL) in warm periods.

Late-stage evolution of hypogene caves at Tyuya-Muyun (Kyrgyzstan): Quantitative insights from mineral deposits

Recommended publications

Geothermometry of calcite spar at 10–50 °C

Stable isotope imprint of hypogene speleogenesis: Lessons from Austrian caves

Late Cretaceous to early Cenozoic hydrothermal fluid migration and red calcite formation in the Tran...

The Obir caves adjacent to the Periadriatic Fault in southern Austria: Uplifted hypogene caves forme...