Quaternary International 91 (2002) 137–152
Climate-induced changes in Holocene calcareous tufa formations,
Bohemian Karst, Czech Republic
Karel $Z!aka,
*, Vojen Lo$zekb
, Jaroslav Kadlecb
, Jana Hlad!ıkov!aa
, V!aclav C!ılekb
a
Czech Geological Survey, Kl!arov 3, 118 21 Praha 1, Czech Republic
b
Institute of Geology, Czech Academy of Sciences, Rozvojov!a 135, 165 00 Praha 6, Czech Republic
Abstract
About 70 localities where Holocene calcareous tufa is formed recently and/or was formed in the past are known in the Bohemian
Karst, a small karst area located SW of Prague. All known tufa accumulations display a very similar pattern of lithological and
biostratigraphic evolution, reﬂecting climatic changes, erosion events and biota succession. A 17 m thick tufa accumulation at Svat!y
Jan pod Skalou was selected and thoroughly studied as a typical, well developed representative of local tufa deposits. Tufa
formation started at about 9500 BP, on a ﬂat ﬂuvial gravel terrace of Late Glacial/Early Holocene age. Deposition of lithologically
uniform, pure hard porous tufa continued until about 6500 BP. From that time, a more unstable climate with several dry periods
and erosion events produced a lithologically varied complex of loose tufa alternating with embedded soils and scree layers.
Termination of the tufa deposition occurred about 2500 BP, and was followed by partial erosion connected with relocation of the
spring below the tufa body. Holocene climatic changes were recorded in lithology, molluscan assemblages, and oscillations of
oxygen and carbon stable isotope ratios in carbonate. The observed patterns are in good agreement with the evolution of calcareous
tufa deposits throughout Central Europe. r 2002 Elsevier Science Ltd and INQUA. All rights reserved.
1. Introduction
The Bohemian Karst, formed by limestones of
Paleozoic age and located in the central part of Bohemia
between Praha and Zdice, is an area with numerous
deposits of calcareous tufa of Holocene age. In this
karstic area (approx. 200 km2
) about 70 localities where
calcareous tufa precipitates or was deposited in the past
are known (Kovanda, 1971; Lo$zek, 1992; Kadlecov!a
and $Z!ak, 1998). In general, the following three
morphogenetic types of tufa accumulations can be
distinguished (using terminology of Ford and Pedley,
1996):
* valley-side, with perched spring accumulations on
slopes, with small extent but locally large thickness
(a very frequent type, but carbonate precipitation is
usually limited for this type under present-day
conditions),
* ﬂuvial accumulations in surface streams, usually
braided but rarely small barrage types (with less or
more intensive active deposition under present-day
conditions), and
* lacustrine and paludal accumulations (inactive under
present-day conditions, deposited in ﬂat valleys of the
southern part of the area, typically during the
climatic optimum of the Holocene).
Very little information about these calcareous tufa
localities exists in the international literature, though
regional, usually Czech- or German-language literature
is voluminous and lists almost 100 papers (see Kovanda,
1971 or $Z!ak et al., 2001, for a review). Some 20 tufa
bodies of the area were studied using lithological and
especially biostratigraphical methods (assemblages of
Mollusca) over the past decades. Detailed study, based
on application of geological, biostratigraphical and
geochemical methods combined with dating was still
missing. A combination of biostratigraphical study with
stable isotope data can be especially useful, since both
methods can decipher different aspects of past climatic
changes. Holocene tufa accumulations were generally
shown to contain a stable isotope record of paleoclimatic
change (Pazdur et al., 1988; Andrews et al., 1994,
1996).
The calcareous tufa body at Svat!y Jan pod Skalou
was selected as a case section for detailed study of tufa
deposition in the Bohemian Karst. This section exhibits*Corresponding author.
1040-6182/02/$ - see front matter r 2002 Elsevier Science Ltd and INQUA. All rights reserved.
PII: S 1 0 4 0 - 6 1 8 2 ( 0 1 ) 0 0 1 0 7 - 0
the largest thickness of Holocene calcareous sediments
of the whole area and was frequently studied in the past
(e.g., Babor, 1901; Petrbok, 1923, 1956; N$emejc, 1928;
Lo$zek, 1955, 1959, 1964, 1967a; J.ager and Lo$zek, 1968;
Kovanda in $Sibrava et al., 1969; Kovanda, 1971;
Horvatin$ci!c et al., 1989; $Silar et al., 1990; Bouzek,
1993; Lo$zek and C!ılek, 1994, 1995; Benkov!a and
$Ctver!ak, 1998; a full list of all references related to the
locality is given in $Z!ak et al., 2001), is well accessible for
study, and represents a unique archive of Holocene
development of the whole area.
As already shown by Lo$zek (1992), most of the
Holocene calcareous tufa accumulations of the Bohemian
Karst show very similar biostratigraphical and
lithological pattern, reﬂecting Holocene climatic and
biota changes. The tufa section at Svat!y Jan pod Skalou
fulﬁlls requirements for a type locality with record
spanning a major part of the Holocene.
2. Site description
Svat!y Jan pod Skalou (literally St. John under Rock)
is a village located 25 km SW of Prague in a deep valley
of the Ka$c!ak stream incised into Silurian and Devonian
sedimentary sequences. A karstic spring (mean discharge
of about 20 l/s) resurges near the valley bottom
above the boundary between limestone and underlying
volcanic rock. The resurgence site is controlled by
longitudinal and transverse faults. Detailed 3 yr monitoring
of the present-day spring (Buzek et al., 1998; $Z!ak
et al., 2001) concluded that water with long residence
time (approximately 20 yr based on 3
H data) dominates.
The temperature of the spring water oscillated between
10.51C and 11.61C with an average of 11.41C, which is
signiﬁcantly higher than the annual mean temperature
of the area (approximately 8.51C), and indicates deep
and slow groundwater circulation. The spring shows
limited discharge variability. The maximum observed
discharge during local ﬂoods was only about twice of the
observed discharge minimum during an extended
drought period. Present-day mean annual precipitation
of the area is slightly above 500 mm.
The studied calcareous tufa body can be morphologically
classiﬁed as valley-side-sited, perched spring
accumulation with relatively small lateral extent (about
5000 m2
), large thickness (up to 17 m), a steep frontal
cascade, and almost horizontal strata behind the
cascade.
The existence of a permanent source of drinking water
attracted man’s attention to this place very early.
Several caves of the area were inhabited as early as
Late Paleolithic, and a large fortiﬁed settlement of Late
Bronze/Early Iron Age was discovered during ﬁeld
studies on a hill above the spring. According to medieval
and Baroque legends, the hermit St. Ivan lived in a cave
inside the tufa body in the 9th century AD. The tufa
cave is mentioned in a manuscript dated 1205. The hard
tufa was quarried as early as the 12–13th century, as
indicated by the tufa portal of the Romanesque church
in nearby Tet!ın. The spring and tufa became one of most
important places of pilgrimage during the Baroque
period. Some of the early descriptions of the site and
karstic processes date back to Renaissance court of
emperor Rudolf II. Almost half of the tufa body was
extracted for building material, but on the other side
artiﬁcial proﬁles including large cellars made the proﬁles
accessible for detailed studies. The location and extent
of the calcareous tufa body at Svat!y Jan pod Skalou are
shown in Fig. 1.
The internal structure of the accumulation is best seen
in an artiﬁcial longitudinal cut through the tufa body,
about 70 m long. The outcrop starts at the steep frontal
part of the deposit adjacent to the church in the NW,
continues as an 8 m high artiﬁcial cut ca. 30 m long to
the SE, then extends to the E and continues for ca. 45 m
along the shallow gully (Fig. 2). At the spot where this
longitudinal proﬁle changes its direction, an excavated
6 m deep test pit is located. The total thickness of
calcareous Holocene deposits is nearly 17 m here. This
vertical section was selected for an integrated study (see
Fig. 3).
A dry karstic valley, Propadl!e vody (Sunken Waters),
is located east of the tufa body. The ephemeral, episodic
stream transported clastic material through this valley
during strong precipitation events. The layers of clastic
deposits are preserved especially in the upper portion of
the tufa body. An increasing volume of limestone scree
Fig. 1. Sketch-map of the calcareous tufa accumulation at the Svat!y
Jan pod Skalou. (1) anthropogenic deposits, (2) calcareous tufa body
including natural and artiﬁcial cavities, (3) ﬂuvial deposits, (4) slope
deposits, (5) Paleozoic limestones, (6) buildings including the
monastery and church complex in the centre, (7) karst spring, (8)
longitudinal proﬁle through tufa body (see Fig. 2).
K. $Z!ak et al. / Quaternary International 91 (2002) 137–152138
in the sections under the outlet of the Propadl!e vody
gorge was reported by Kovanda in $Sibrava et al. (1969)
(see Fig. 2). However, the massive tufa of the lower part
of section contains almost no scree or fallen boulders
from the high cliff above the tufa body. Rock
disintegration had to be limited during the ﬁrst half of
the Holocene, possibly because the rocks were partly
covered and thus protected by vegetation and because
the mild oceanic-type climate limited the number and
intensity of freeze-thaw cycles.
The spring resurges now at the base of the tufa
accumulation. The position of the original resurgence
point is unknown, but it is generally believed that it
was located either somewhere in the upper part of the
tufa body or not far above in the Propadl!e vody
gorge. It is evident that the position of the resurgence
oriﬁce was not stable during tufa formation. As
indicated by drillings below the church and at the
bottom of the test pit, most of the tufa accumulation is
underlain by the ﬂuvial terrace of the Ka$c!ak stream
formed by pebbles of Upper Proterozoic graywackes,
shales and silicites.
Lithology of the calcareous tufa deposits is variable
both laterally and vertically. The frontal cascade is
formed by hard porous phytohermal (structural) tufa
across the whole thickness of the deposit. Behind the
frontal cascade, where individual sedimentary layers are
horizontal or sub-horizontal, two lithologically different
complexes can be deﬁned, with the boundary between
them roughly equivalent to the test pit top, i.e. level 0 m
in Fig. 3:
(i) the lower lithologically uniform sequence of hard
porous tufa with only a few intercalations of loose
tufa containing a very limited amount of clastic
material, and
(ii) the upper lithologically varied complex of layers of
porous phytohermal tufa, oncoidal tufa and even
lacustrine-like tufa intercalated with several horizons
of buried soils and horizons of limestone
scree.
The uppermost portion of the sequence contains
numerous fragments of prehistoric pottery (especially
beds No. 3–9 in Fig. 3) produced in the Late Bronze or
Early Iron ages (approximately 8th–5th century BC,
Benkov!a and $Ctver!ak, 1998).
Based on micromorphological study, the soil horizons
preserved in the upper part of the sedimentary sequence
are autochthonous (or semi-autochthonous) rendzinatype
soils, with minor admixture of quartz, mica and
feldspar grains, indicating transport of clastic material
on the surface of the tufa accumulation during soil
evolution ( $Zigov!a in $Z!ak et al., 2001). The scree
horizons are dominated by angular limestone fragments
but also contain quartz clasts and pebbles, indicating
transport of sedimentary material derived from Mesozoic
karst inﬁllings and Tertiary ﬂuviatile terraces from
the upper segment of the gorge.
Fig. 2. Longitudinal proﬁle of the calcareous tufa accumulation at the Svat!y Jan pod Skalou: idealized and simpliﬁed proﬁle through tufa
accumulation. The line L2LÃ represents the level of erosional gully bottom and axis A a bend of the proﬁle. Only the left upper part of the proﬁle is
accessible to direct observation. The other parts are either compiled from former unpublished sources (especially Kovanda and Lo$zek in $Z!ak et al.,
2001) or deduced from a pit and bore hole located along A axis, outcrops in nearby cellars of the monastery and in Cave of St. Ivan, and by earth
radar results. The actual stratigraphy is very complex and besides climatic factors it reﬂects different tufa facies such as pool-rim transitions. It can be
simpliﬁed into four basic complexes: (1) Underlying sediments are formed either by rock scree (local Paleozoic limestones and volcanites) or by
ﬂuviatile sediments (gravels). (2) The massive phytohermal tufa used as building material occurs as a ﬂat body in the lower part of the proﬁle. Several
loose to sandy tufa intercalations were observed. (3) The main body of tufa exposed above the gorge bottom consists of highly heterogeneous strata
where massive layers alternate or form facial transitions into loose tufas that sedimented in pools. Notice the ‘‘cascade’’ inclination of layers close to
the left end of proﬁle. The proﬁle is interrupted there by the church wall. The horizontal dark layers represent humic soils, calciﬁed soil sediments and
intercalations of rock scree. (4) The uppermost part is formed by anthropogenic sediments associated with construction of monastery, prehistoric
cultural layers, soils and different slope sediments.
K. $Z!ak et al. / Quaternary International 91 (2002) 137–152 139
K. $Z!ak et al. / Quaternary International 91 (2002) 137–152140
Tufa mineralogy is very uniform, dominated by lowMg
calcite with minor admixture of clay minerals and
quartz grains in the upper portion of the sequence.
Yellow, red and black strips of Fe- and Mn-hydroxides
(MnO2 type) precipitated on the contacts between soils
and tufa due to the inﬂuence of geochemical barriers.
Some authors (e.g., Lo$zek, 1963; Kovanda, 1964)
have suggested diagenetic changes of calcareous tufa
resulting in signiﬁcant solidiﬁcation of the tufa by ﬁlling
of open spaces and recrystallization of calcite (so called
‘‘travertinization of tufa’’). Detailed study of hard tufa
from the lower portion of the sequence has shown that
even small cavities around decomposed roots or stems
remained unﬁlled by secondary carbonate. The diagenetic
processes were therefore of minor importance or
were restricted to partial recrystallisation of the already
precipitated material. Petrological study suggests that
tufa diagenesis which took place probably almost
immediately after tufa deposition should have no
signiﬁcant effect on the isotopic record. We observed
that the hardening and diagenesis of modern tufa
deposits frequently happens within a few years after
deposition, and subsequently the carbonate system
remains ‘‘locked’’ to secondary changes.
3. Methods
3.1. Malacozoology
Samples for the study of molluscan assemblages were
taken in several sampling campaigns from 18 levels in
the section, including the test pit S1 and core obtained
by drilling at the bottom of the test pit. Samples with
volume 5–10 L were dried and wet-sieved before
separation of the molluscan shells. The techniques of
sampling and shell separation were described by Lo$zek
(1964).
3.2. Dating
For radiometric dating of the calcareous tufa, a
combination of several dating methods was applied. The
only method for which a sample can be collected
anywhere in the studied section is 14
C dating of
carbonate. By this method 11 samples covering 14.3 m
of the thickness have been dated by conventional
counting method in the Environmental Isotope Laboratory,
Department of Earth Sciences, University of
Waterloo, Canada. Organic matter, either small pieces
of charcoal or sedimentary organic matter of buried
soils, were (after treatment with hot solutions of acid,
alkali, and acid to remove all carbonate carbon and
easily soluble organics) measured by the AMS method
in the Rafter Radiocarbon Laboratory, Institute of
Geological and Nuclear Sciences, Lower Hutt, New
Zealand. The 14
C data on organic matter have been
calibrated for variable initial concentration of 14
C using
published calibration curves. In the intervals between
dated samples, ages of individual depth levels were
calculated assuming constant accumulation rates. In the
interval between the deepest sample of organic matter
dated by AMS 14
C (À2.95 to À3.05 m) and the base of
tufa accumulation (i.e., À9.65 m) the age difference was
estimated based on carbonate 14
C data, which indicates
an age difference of about 1000 yr between these two
horizons. Several samples of tufa were dated also by
230
Th/234
U (a-particle counting) method in the Centre
d’Etudes et de Recherches Appliqu!ees au Karst, Facult!e
Polytechnique de Mons, Belgium, and in the Institute of
Geological Sciences of the Polish Academy of Sciences,
Warszawa.
3.3. Stable isotope study
Samples for geochemical study were collected by the
channel method, each sample representing an average of
a section segment about 0.2 m thick. Altogether 83
samples were collected in two sampling campaigns.
After drying coarse clastic material was removed
by combination of sieving and hand-picking, with
special care given to the removal of any clasts of
Fig. 3. The studied section. Lithology: (1) dark brown humic soil with
small limestone and tufa clasts with a layer of tufa scree at the base, (2)
brown humic soil with limestone clasts and a layer of limestone scree,
(3) light brown extremely loose tufa mixed with brown soil with
frequent tufa clasts, (4) limestone scree with silty and tufa matrix, (5)
brown sandy soil with tufa scree and limestone clasts at the base, (6)
brown sandy soil with tufa clasts, (7) brown–yellow to yellow loose
tufa, locally irregular relics of solid tufa, (8) yellow solid tufa, (9)
limestone scree with silty and sandy matrix, (10) grey–yellow partly
loose tufa, (11) grey–yellow extremely loose tufa with brown silty and
sandy layers, thin layer of limestone scree at the base, (12) light grey
solid tufa, loose in places, (13) brown sandy soil with tufa clasts, (14)
light grey solid tufa, loose in places, (15) dark brown sandy and clayey
soil, (16) grey to brown–grey strongly loose tufa with rusty limonite
streaks, (17) grey to brown–grey strongly loose tufa with limonite
streaks, (18) light grey solid tufa, loose in places, (19) whitish grey solid
tufa with conspicuous limonite streak at the top, (20) light grey sandy
lacustrine tufa, (21) light grey, locally rusty brown solid tufa, (22) light
grey to yellow, locally rusty brown solid tufa, (23) light grey strongly
loose tufa, (24) light brown loose tufa with lenticular layer of sandy
gravel in the upper part, (25) limestone scree with sandy matrix, (26)
light brown strongly loose tufa, locally with silty matrix, (27) yellowish
white solid tufa, (28) grey–yellow to light grey massive solid tufa, (29)
light grey strongly loose tufa, locally with silty admixture, (30) grey
yellow to light grey massive solid tufa, (31) light grey to brown
strongly loose tufa, locally with silty admixture, (32) grey–yellow loose
tufa, (33) loose grey–brown tufa, (34) grey–yellow to brown–yellow
massive solid tufa, more porous at the base, (35) limestone scree with
tufa fragments and incrustations in the upper part, pebbles of
Paleozoic and Proterozoic rocks in the lower part of the layer, (36)
sandy ﬂuvial gravel with pebbles of Paleozoic and Proterozoic rocks
and minor presence of limestone scree.
3———————————————————————
K. $Z!ak et al. / Quaternary International 91 (2002) 137–152 141
Paleozoic limestones. Samples were then homogenized,
and carbonate content, organic matter content, carbonate
d13
C and d18
O and organic matter d13
C were
determined.
3.4. Other methods
Many other methods were used to determine the tufa
mineralogy and petrology (electron microprobe JEOLJXA-50A,
difractograph Philips), to describe local karst
hydrology (water chemistry, tracing experiments, d18
O
and 3
H study of water, d15
N study of dissolved nitrate),
and the geology of tufa deposit (ground-penetrating
radar). All three important outlets of the karst spring
were monitored for 3 yr with respect to temperature and
discharge. The prehistoric ﬁnds were collected at several
places and examined by an archaeological team.
Analyses of slope sediments, soils, cave inﬁllings was
conducted. The results of these interdisciplinary studies
are described in a monograph ( $Z!ak et al., 2001), but in
this paper we focus on climatic and environmental
analysis only.
4. Biostratigraphical record
Molluscan shells are present in all layers of the studied
section. They are much more abundant in loose
sediments than in solid phytohermal tufa. The malacofauna
from this site was described previously by Petrbok
(1923) and this ﬁrst list of species was repeatedly
supplemented (see Petrbok, 1956, for a review). Petrbok
attributed all molluscan records from Svat!y Jan to the
‘‘Atlantic Littorinian’’, a moist phase in which a great
majority of tufa deposits of the Bohemian Karst were
formed. The malacofauna described by Petrbok consists
predominantly of demanding woodland species and
includes no elements of Early Holocene Discus ruderatus-fauna,
because the Early Holocene layers were not
exposed at that time. Of particular interest is a rather
high number of aquatic species reported by Petrbok, for
instance the presence of Bathyomphalus contortus (L.);
among terrestrial species even the alpine element
Neostyriaca corynodes (Held) is cited, whose occurrence
in the Holocene of the Bohemian Karst is unlikely.
New investigations were initiated in 1960 when K.-D.
J.ager and V. Lo$zek cleaned and documented the section
which was later connected with the excavated research
pit (see Fig. 1 for location, Fig. 3 for a section drawing).
During the recent period earlier studies were complemented
by more precise analysis of molluscan assemblages
from the test pit and research drilling. The results
are given in Table 1 including information on ecology
and biostratigraphy of each species.
The entire section is dominated by a well-developed
woodland malacofauna, which includes a number of
species with high temperature and moisture requirements,
i.e., an assemblage corresponding to the climatic
optimum. The number of shells transported from
various habitats surrounding the tufa deposit, particularly
from xerothermic rock cliffs, remains very low
throughout the sequence. Changes in the occurrence of
several species as well as in the presence of particular
ecological groups enable the tufa sequence to be
subdivided into several units characterized by the
presence/absence of certain elements.
The number of species and specimens obtained from
the lowermost section of the accumulation using the
research drilling (layers 31–36) was very small. The
occurrence of individual species did not differ signiﬁcantly
from the overlying complex of solid tufa studied
in more detail.
The basal complex of solid tufa in the excavated test
pit (layers 30 to 28) is characterized by high numbers of
two open-country speciesFTruncatellina cylindrica and
Vallonia costataFwhich suddenly disappear in horizon
28. This layer is characterized by abrupt decrease in
species richness. Carychium tridentatum is abundant.
Trichia sericea, characteristic of the older half of
Holocene, occurs in layers 30 to 28. The lowermost
horizon 30 contained the demanding thermophilous
element Truncatellina claustralis associated with Early
Holocene elements, such as Discus ruderatus and
Perpolita petronella that occur in very small numbers.
Interesting is the occurrence of Laciniaria plicata and
the appearance of neoendemic Bulgarica nitidosa in the
loose interlayer 29. Some species characteristic of the
younger half of the Holocene, particularly Helicodonta
obvoluta, and aquatic species are absent.
Loose layers between the complex of solid tufa
(28–30) and lower scree horizon (25) include a very rich
malacofauna characterized by the appearance of sensitive
woodland species, such as Platyla polita, Bulgarica
cana, and Cochlodina orthostoma. Early Holocene
elements are represented by Clausilia cruciata, whereas
Trichia sericea, Vallonia costata and Truncatellina
cylindrica disappear. Of interest is the ﬁnd of one
specimen of Chondrina avenacea, a rupestral epilithic
species, and the strong occurrence of aquatic species
Radix ovata and Pisidium spp. The woodland species
show a marked increase in the assemblage. Early
Holocene elements, for example Discus ruderatus and
Perpolita petronella, decline and disappear in the loose
layers 26 and 27. All these data suggest a rather open
forest with denser parklands in some places under very
favorable soil and moisture conditions, which corresponds
to the late Atlantic phase.
The lower scree horizon (25) is characterized by
maximum species richness (46 species), high numbers of
aquatic and wetland elements, as well as by the
appearance of Isognomostoma isognomostomos. The
genus Aegopinella is here probably represented even by
K. $Z!ak et al. / Quaternary International 91 (2002) 137–152142
the hygrophilous species Ae. nitens. This horizon
provided the only specimens of Vertigo alpestris, and
the last occurrence of Discus ruderatus. Even though the
scree indicates transport of clastic material from the
surroundings, the horizon includes no rupestral species
coming from the cliffs situated in the neighborhood of
the tufa deposit.
The thick complex (layers 24 to 10) between the lower
and upper scree horizons is characterized by a wide
variety of lithofacies reﬂecting the alternation of tufa
wetlands, small pools and short dry periods, but its
assemblages are less diverse. Aquatic and wetland
species remain important and their representation
attains its maximum in the intercalation of ﬁne limnic
tufa (layers 21 and 20), largely dominated by Radix
ovata with subordinate Pisidium casertanum. Only this
horizon contained the wetland element Oxyloma elegans.
In layer 17, Arianta arbustorum and Helicigona
lapicida appear, whereas the ﬁrst record of a species
characteristic of the younger part of Holocene, Helicodonta
obvoluta, comes from the layer 12. The richest
woodland assemblage was recorded in layer 15 (32
species).
The upper scree with humic rendzina matrix and
prehistoric pottery (layers 9 to 7), archaeologically dated
to the declining Bronze Age and Early Iron Age at about
800–500 BC, includes again species-rich (34 and 38
species) assemblages, in which woodland snails remain
dominant. Even here, thermophilous Truncatellina
claustralis appears in association with Ena obscura.
Table 1
Assemblages of Mollusca in individual layers of calcareous tufa in Svat"y Jan pod Skalou. Layer numbers relate to Fig. 3
Ecological-biostratigr. Layer No.
characteristics List of species
36
35B
35A
33
31
30
29
28
26/27
25
20/21
17
15
12
9
7
4
2
! Acanthinula aculeata (Müller) - - - - - + + - + + + + + + + + - !
Aegopinella nitens (Michaud) - - - - - - - - - +? - - - - - - - !
Aegopinella pura (Alder) - - - - - + + + + + + - + + + + - +
! Bulgarica cana (Held) - - - - - - - - + + - - - - - - - Clausilia
cruciata (Studer) - - - - - - - - + - - - - - - - - !
Cochlodina laminata (Montagu) - - - - + - + - + + - + + + + + + +
! Cochlodina orthostoma (Menke) - - - - - - - - + - - - - - - - - (G)
Discus ruderatus (Férussac) - - - - - + - - - + - - - - - - - !
Ena montana (Draparnaud) - - - - - + - - - + - - - - - + + +
! Ena obscura (Müller) - - - - - - - - - - - - - - + + - -
1 ! Helicodonta obvoluta (Müller) - - - - - - - - - - - - - + + + + +
! Isognomostoma isognomostomos (Schröter) - - - - - - - - - + - - - - + - - +
! Macrogastra plicatula (Draparnaud) - - - - - + - - + + - - + + + + + !
Monachoides incarnatus (Müller) + + - + + + + + + + + + + + + + + +
(G) Oxychilus depressus (Sterki) - - - - - + - - - + - - - - - - - !
Petasina unidentata bohemica (Lozek) - - - - + + + + + + - - + + + + - +
A ! Platyla polita (Hartmann) - - - - - - - + + + - + + + + - - +
Semilimax semilimax (Férussac) - - - - + + + + + + + + + + - - - !
Sphyradium doliolum (Bruguiere) - - - - - + + + + + - + + + + + + +
(!) Vertigo pusilla Müller - - - - - + + - + + - - + + + + + !
Alinda biplicata (Montagu) - - - - - + + - + + + + + + + + + +
(+) Arianta arbustorum (Linné) - - - - + - - - - - - + + + + + + W
(M) ! Cepaea hortensis (Müller) - - - - + - + - + + - - - - - + - +
! Discus rotundatus (Müller) - - - - + + + + + + + + + + + + + +
2 ! Limax sp. (cf.cinereoniger Wolf) - - - - - - - - - + - - - - - - - !
Aegopinella minor (Stabile) - + - + + + - + + + + + + + + + + +
W (S) (!) Fruticicola fruticum (Müller) + - + + + + + + + + - + + - - + - +
! Helix pomatia Linné - - - +? - + + - + + - + + + - - - +
W (H) (+) Vitrea crystallina (Müller) - - - - + + - + + + + + - + - - - (G)
Clausilia pumila C.Pfeiffer - - - - - - + - + + - - + - + + + -
3
4
! Macrogastra ventricosa (Draparnaud) - - - - + + + + + + + + + + - + + !
Urticicola umbrosus (C.Pfeiffer) - - - - + - - - + + + + + - + + + M
Cecilioides acicula (Müller) - - - - - - - - - - - - - - - - - +
S (+) Granaria frumentum (Draparnaud) - - - - - - - - - - - - - - - + + +
Ca Chondrina avenacea (Bruguiére) - - - - - - - - + - - - - - - - - S
(W) !! Truncatellina claustralis (Gredler) - - - - - + + - + - - - - - + - - B
+ Pupilla muscorum (Linné) - - - - - - - - - - - - - - - - - +
(!) Truncatellina cylindrica (Férussac) - - - - - + + - - - - - - - - - - -
5 (+) Vallonia costata (Müller) - - - + - + + - + + - - - + + + + +
G Vallonia pulchella (Müller) - - - - - - - - - - - - - - - - + -
^
K. $Z!ak et al. / Quaternary International 91 (2002) 137–152 143
Cochlicopa lubricella and Vallonia costata in high
amounts as well as the appearance of Granaria
frumentum (layer 7) reﬂect partial clearance of forests.
Importantly, the woodland character of the malacofauna
is preserved, although higher numbers of such species
as Cochlicopa lubricella, Vallonia costata and the
admixture of Granaria frumentum reﬂect a clearance of
the adjacent forest complex.
In the uppermost, predominantly clastic layers (6 to 2)
the woodland character of malacofauna continues, the
species richness slightly decreases and several elements
indicating further clearance in the surroundings appear
(Vallonia pulchella, Pupilla muscorum, Cecilioides).
However, Platyla polita, a sensitive woodland species
is still present, Laciniaria plicata re-appears (at present
extinct in the Bohemian Karst) and even in the
subsurface layer 2 Helicodonta obvoluta remains
important. By contrast, modern immigrants are
represented only by the terricolous Cecilioides acicula,
which penetrates into deeper soil horizons. The assemblages
of the youngest layers converge to present-day
conditions. Today’s environment is characterized by
fully developed woodland biocenoses including rich
forest snail communities. It is obvious that neither
prehistoric nor historic human activities were sufﬁciently
intensive to lead to depopulation of woodland
ecosystems and expansion of species characteristic of
cultivated areas.
! Bulgarica nitidosa (Ulicný) - - - + - - + - + + - - + + + + + +
(!) Cochlicopa lubricella (Porro) - - - - - + + - - + - - - - + + - +
(!) Euomphalia strigella (Draparnaud) - - - + + + + - - + - - - + + + - +
! Tandonia rustica (Millet) - - - - - - - - - - - - - - - + - (+)
Cochlicopa lubrica (Müller) - - - - + + + + + - - + + + - - - +
(+) Euconulus fulvus (Müller) - - - - - + + + + + + - + + - + - (+)
Limacidae /Agriolimacidae - - - - + - + + - + - + + + + + + +
(+) Perpolita hammonis (Ström) - - - - - + + - - + - - - - - - - Me
(+) Punctum pygmaeum (Draparnaud) - - - - - + + - + + - + + - + + - +
Trichia hispida (Linné) - + - + - - - +? + - - + - - + - - (+)
Trichia sericea (Draparnaud) - - - - - + + + - - - - - - - - - C
7
6
! Vitrea contracta (Westerlund) - - - - - + + - + + - - + - - - - (G)
Vitrina pellucida (Müller) - - - - - - - - - - + - - - + - - (+)
Clausilia dubia (Draparnaud) - - - - - - - - + + - - - - - - + !
Helicigona lapicida (Linné) - - - - - - - - - - - + + + + + + +
Wp ! Laciniaria plicata (Draparnaud) - - - + - + - - - - - - - - - - + G
Vertigo alpestris Alder - - - - - - - - - + - - - - - - - !
Carychium tridentatum (Risso) - - - + + + + + + + + + + + + + + (!)
Columella edentula (Draparnaud) - - - - + + + + + + + + - - - + - -
8 (G) Perpolita petronella (L.Pfeiffer) - - - - - + - - - - - - - - - - - +
Succinella oblonga (Draparnaud) - - - - - - - - - - - - - - - + - G
Carychium minimum Müller - - - - - - - - + + - - - - + - - (G)
Oxyloma elegans (Risso) - - - - - - - - - - + - - - - - - -
9 (+) Succinea putris (Linné) - - - - - + + + + + - - - - + + + +
G Zonitoides nitidus (Müller) - - - - - + + - + + + - + + + + - (+)
Aplexa hypnorum (Linné) - - - - - - - - - - + - - - - - - D
(+) Galba truncatula (Müller) - - - - - - - - - - - + + - - + + (+)
Pisidium casertanum (Poli) - - - - + - - - + + + - + - - - + +
10 (+) Pisidium obtusale (Lamarck) - - - - - - - - + - - - - - - - - Pisidium
personatum Malm - - - - - - - - + + + + + + + + + Pisidium
supinum A. Schmidt - - - - - - - - - - - - - - - - - +
(+) Radix ovata (Draparnaud) - - - - + - - - + + + + + - - - - +
G Radix peregra (Müller) - - - - - - - - - - + - - - + + - number
of species 2 3 1 10 19 35 32 19 42 46 21 25 32 28 34 38 27 29
Ecological-biostratigr. Layer No.
characteristics List of species
36
35B
35A
33
31
30
29
28
26/27
25
20/21
17
15
12
9
7
4
2
Ecological characteristics: General ecological groups: AFwoodland (in general), BFopen country, CFwoodland/open country, DFwater,
wetlands. Ecological groups: 1Fwoodland (sensu stricto); 2Fwoodland, partly semiÀopen to open habitats [W(M)Fmesic, W(S)Fxeric,
W(H)Fdamp]; 3Fdamp woodland; 4Fxeric open habitats [SFin general, CaFlimestone rocks, S(W)Fpartly shaded habitats]; 5Fopen habitats
in general (moist meadows to steppes); woodland/open country: 6Fpredominantly dry; 7Fmesic or various (MeFmesic in general, catholic,
WpFmesic rocks, scree woodland); 8Fpredominantly damp; 9Fwetlands, banks; 10Faquatic habitats.
Biostratigraphic characteristics: +Floess species, (+)Flocal or occasional loess species, !Fspecies characteristic of warm phases, (!)Feurythermic
species of warm phases, !!Findex species of warm phases, GFspecies surviving the Glacial out of the loess zone, (G)Fditto as relics, MFmodern
immigrants (Late Holocene index species)
Presence in layers: +Fpresent, +?Fonly approximate determination, ÀFabsent.
K. $Z!ak et al. / Quaternary International 91 (2002) 137–152144
5. Chronology and the stable isotope record
Since 14
C dating of carbonate is plagued by a number
of uncertainties (unknown initial activity because of
mixing of organic matter-derived carbon and limestonederived
carbon, a possibility of post-depositional
changes because of tufa diagenesis, inﬁlling of pores,
etc.), small pieces of charcoal or sedimentary organic
matter from selected horizons were also 14
C dated using
AMS. Nevertheless, 14
C dating of carbonate is the only
method for which a sample can be collected anywhere in
the section, which was the reason why both methods
were applied. The data are given in Table 2.
In Fig. 4, calibrated 14
C data of organic matter and
14
C model ages of carbonate are plotted against depth in
the section. The carbonate 14
C model ages were
calculated assuming constant initial 14
C activity of
carbonate carbon equal to 80% of modern carbon
(80 pmc). This selected value is slightly lower than the
14
C activity of HCO3
À
of the present-day spring water,
which was determined several times (83 pmc in 1987,
Horvatin$ci!c et al., 1989; between 80.5 and 83.9 pmc
during 1997–1998 period, $Silar and Z!ahrubsk!y, 1999).
This selected value gives an identical age for both
methods at the uppermost level in the section, where
both carbonate and organic matter were dated. The
agreement between model carbonate 14
C ages and
organic matter 14
C ages in the deeper part of the section
is not satisfactory. Application of the calibration curve
to the carbonate 14
C data would improve this discrepancy
only partly. It follows that the initial activity of
deposited carbonate changed signiﬁcantly through time
and/or that the carbonate was partly inﬂuenced by postdepositional
changes. This is why organic matter 14
C
data were taken as a basis for calculation of ages for
individual depth levels. Constant accumulation rates
were assumed between individual dated levels. A
problem arose in the basal portion of the section formed
by pure tufa, where collection of organic matter for
dating was impossible. Based on the difference in
carbonate 14
C activity, it was roughly estimated that
the segment between the deepest dated sample of
organic mater (À3.00 m) and the carbonate layer base
(À9.65 m) corresponds to approximately 1000 yr. Under
this assumption, the base of calcareous tufa layer was set
at 9500 yr BP. Based on calibrated 14
C data of organic
matter and the assumptions mentioned above, a model
age was attributed to each carbonate sample used for
obtaining of the stable isotope record.
The 230
Th/234
U dating method was applied to verify
the 14
C dating. All samples had suitable U contents
(0.263–0.761 ppm U), but an admixture of clastic Th
component adsorbed on clay minerals was too high and
the 230
Th/232
Th was lower then 10. Moreover, variable
234
U/238
U ratio indicated disturbed isotopic systems.
Even the sample with the highest 230
Th/232
Th ratio
(À5.40 to À5.50 m, 0.761 ppm U) indicated an unrealistically
high age of 10.670.4 ka. After ﬁnishing
the drilling into the deepest part of the accumulation,
230
Th/234
U dating was applied once on a selected
sample of solid pure tufa with limited number of pores
(sample location: À8.30 m). The sample contained
0.5070.01 ppm U and the 230
Th/232
Th ratio was again
low. The sample yielded an uncorrected age of
10.3970.24 ka. Since this sample was the most suitable
for the 230
Th/234
U dating, a rough correction to subtract
the detrital thorium was applied on the raw data. The
corrected age of 871 ka is in rough agreement with
other dating methods.
The radiometric dating is in good agreement with
both biostratigraphic dating and with dating of archaeological
ﬁnds. The molluscan assemblages indicate that
the tufa sequence does not cover the initial phases of the
Holocene, since typical faunal assemblages of the Late
Glacial/Early Holocene transition are not well developed
here. The 14
C dating of organic matter derived
from layer No. 6 with pottery fragments of the Knov!ız
culture (generally dated from approximately 8th to 6th
century BC, Benkov!a and $Ctver!ak, 1998) yielded a
corrected age within an interval 2876 to 2483 yr BP
(95% conﬁdence interval).
Carbonate content and carbonate d13
C and d18
O data
plotted against the model age are shown in Figs. 5 and 6.
The highest carbonate contents were recorded in the
proﬁle section with model ages between 8800 and
8600 yr BP. Near the level of around 8500 yr BP, there
are several distinct horizons enriched in clastic material
(mostly clay minerals, organic matter, quartz and
feldspar grains). The most stable climatic conditions
(so-called climatic optimum of Holocene) with very
limited transport of any clastic material and with
deposition of relatively pure calcareous tufa occurred
between depths À3.00 and +0.07 m (corresponding
model ages of 8400 and 6700 yr BP). In the second part
of the Atlantic (since about 6700 yr BP) the carbonate
content decreased and the proportion of clastic material
irregularly increased. The ﬁrst maximum of clastic
material transport into the studied section occurred
around 6200 yr BP. The overlying lithologically variable
sequence contains numerous horizons very rich in clastic
material and also horizons of buried soils. The lowest
carbonate content was recorded in a buried soil horizon
at +3.50 to +3.75 m (layer 15 in Fig. 3). The upper
scree horizon at +5.05 to +5.20 m corresponds roughly
to 3100 yr BP and above it the deposition of calcareous
tufa was discontinued.
The oxygen isotopic composition of freshwater
carbonates is inﬂuenced by several partly interdependent
factors, the two most important being temperaturecontrolled
oxygen isotope fractionation calcite-water
and temperature-controlled variability of oxygen isotopic
composition of meteoric waters (Gascoyne, 1992;
K. $Z!ak et al. / Quaternary International 91 (2002) 137–152 145
Table 2
14
C dates from carbonate and organic matter
Laboratory No. Meters in the
proﬁle (m)
Sample description Preparation and
measurement method
14
C activity %
MC
Conventional
14
C age years
BP
Calibrated 14
C age
(95% conﬁdence
interval) years BP
Rafter Radiocarb.
Lab., New Zealand
NZA 5877
+6.00 to +6.10 Charcoal pieces separated from
layer No. 6 containing Late Bronze
Age pottery (Knov!ız Culture). Dated
materialFcharcoal
AÀAÀA leaching+AMS 71.6170.60 2639767 2876 to 2693
2653 to 2483
Univ. of Waterloo,
Canada 55465
+4.20 to +4.30 Larger massive block of hard pure
tufa within layer No. 12. Dated
materialFcarbonate
Acid decomposition+conventional
counting method
54.0070.47
Rafter Radiocarb.
Lab., New Zealand
NZA 5878
+4.00 to +4.10 Isolated charcoal pieces separated
near boundary of layer No. 13 and
14. Dated materialFcharcoal
AÀAÀA leaching+AMS 64.4170.61 3489777 3935 to 3555
Rafter Radiocarb.
Lab., New Zealand
NZA 5879
+2.60 to +2.65 Isolated charcoal pieces separated
from upper part of layer No. 20.
Dated materialFcharcoal
AÀAÀA leaching+AMS 56.2070.51 4585774 5467 to 4997
Univ. of Waterloo,
Canada 55466
+1.95 to +2.05 Hard porous ‘‘structural’’ tufa,
central part of a large block within
layer No. 22. Dated materialFcar-
bonate
Acid decomposition+conventional
counting method
50.1070.46
Univ. of Waterloo,
Canada 55467
À0.35 to À0.45 Hard porous ‘‘structural’’ tufa,
upper part of layer No. 28. Dated
materialFcarbonate
Acid decomposition+conventional
counting method
45.0070.40
Univ. of Waterloo,
Canada 55468
À2.55 to À2.65 Hard porous ‘‘structural’’ tufa,
lower part of layer No. 28. Dated
materialFcarbonate
Acid decomposition+conventional
counting method
42.7070.41
Rafter Radiocarb.
Lab., New Zealand
NZA 7772
À2.95 to À3.05 Loose light brown tufa with soil
admixture. Dated materialForganic
matter in the soil admixture
AÀAÀA leaching+AMS 38.2670.32 7673768 8551 to 8325
Univ. of Waterloo,
Canada 55469
À5.40 to À5.50 Hard porous ‘‘structural’’ tufa,
lower limit of layer No. 30. Dated
materialFcarbonate
Acid decomposition+conventional
counting method
34.4070.34
Univ. of Waterloo,
Canada 74030
À7.05 to À7.10 Hard porous ‘‘structural’’ tufa,
upper part of layer No. 34 (from
drilling). Dated materialFcarbo-
nate
Acid decomposition+conventional
counting method
30.5470.30
Univ. of Waterloo,
Canada 74031
À8.25 to À8.35 Hard porous ‘‘structural’’ tufa,
lower part of layer No. 34 (from
drilling). Dated materialFcarbo-
nate
Acid decomposition+conventional
counting method
30.0870.30
K.$Z!aketal./QuaternaryInternational91(2002)137–152146
Dorale et al., 1992). These two factors have an opposite
inﬂuence on the carbonate d18
O data, with the effect of
temperature-controlled variability of oxygen isotopic
composition of meteoric waters usually being the more
important factor in the mild climatic zones.
With respect to the large discharge of the spring, long
underground residence time of the spring water and
almost constant spring temperature, and with regard to
the location of the tufa deposit near the spring, any
signiﬁcant water temperature changes and/or isotopic
effects connected with water evaporation at the site of
deposition probably can be neglected. Changes in the
d18
O values of meteoric waters were the most important
factor controlling changes in d18
O of the deposited
carbonate. A similar conclusion about freshwater tufa
deposits of British Isles was drawn by Andrews et al.
(1993). Under these conditions higher carbonate d18
O
values should indicate higher temperatures and vice versa.
The carbonate d18
O record obtained from the Svat!y
Jan section was compared with ice d18
O in the Greenland
GISP 2 (Fig. 7; data for GISP 2 after Grootes et al.,
1993; Stuiver et al., 1995; GISP 2 dating after Sowers
et al., 1993; Meese et al., 1997). With respect to dating
uncertainty at Svat!y Jan, the oscillations of both curves
can be viewed as similar. A signiﬁcant temperature
increase marking the onset of Holocene is clearly not
recorded at the Svat!y Jan. This is in good agreement with
both the results of dating of the tufa section and with the
biostratigraphic analysis of molluscan assemblages.
Carbonate d13
C values at Svat!y Jan show higher
variability than d18
O values. Moreover, it was found
that the carbonate d13
C data are probably controlled by
different mechanisms in the lower series of hard pure
tufa with high carbonate content and in the upper varied
sedimentary series. While in the lower portion of the
section the d13
C and d18
O of carbonate change
independently, in the upper portion there is a signiﬁcant
correlation between oscillations of the d13
C and d18
O
data. With respect to the fact that differences between
the d13
C values of carbonate and organic matter are
almost constant throughout the section (d13
Ccarb are in
the range from À10.54% to À7.92%, d13
Corg in the
range À30.6% to À26.3%, D13
CcarbÀorg in the range
18.8–20.7%), the kinetic effect of CO2 escape from the
water at the site of deposition had probably only a
Fig. 4. 14
C ages based on calibrated data of organic matter and also
the 14
C carbonate ages, calculated assuming identical initial activity of
carbonate carbon of 80 pmc vs. depth.
Fig. 5. Carbonate content vs. model age.
K. $Z!ak et al. / Quaternary International 91 (2002) 137–152 147
minor inﬂuence on the carbonate d13
C systematics. The
environmental and climatic impacts on soil processes in
the recharge area were probably more important. We
can speculate on the nature of these processes. They
could be caused by human activities associated with
more intensive pasture and deforestation, enhanced
sheet erosion, changed microbial conditions of soil
edaphon, gradual decalciﬁcation of soils or oceanity/
continentality shifts. The ﬁeld observations from Central
Bohemian region support all these hypotheses, but
some contradictory evidence appearsFe.g. tufa formation
ceased even in nearby K$rivokl!at region at sites
located at steep slopes which were never affected by
human inﬂuences and where the mollusc assemblages
prove the continuous existence of forest.
6. Discussion
6.1. Onset of tufa depositionFcomparison with other
tufa bodies
The calcareous tufa section at Svat!y Jan pod Skalou is
the one of few sites within the Bohemian Karst where
biostratigraphic analysis was combined with an extensive
application of geochronological methods. During
the Late Glacial any deposits of calcareous tufa at the
Svat!y Jan pod Skalou (if formed) were destroyed due to
increased ﬂuvial erosion. The rock ﬂoor of the Ka$c!ak
Valley was at this time up to 8–9 m deeper than is the
surface level of the recent ﬂoodplain. The tufa accumulation
at the Svat!y Jan pod Skalou is underlain by river
terrace probably of Late Glacial/Early Holocene age.
For the onset of calcareous tufa formation two
controlling factors were of prime importance:
* decrease in ﬂuvial activity of the Ka$c!ak stream, and
* an increase in dissolved load of spring water
connected with expansion of plant communities and
evolution of soil proﬁles in the recharge area.
Both geochronological dating and assemblages of
Mollusca indicate that the basal tufa layer at Svat!y Jan
pod Skalou was deposited around 9500 yr BP; i.e. the
transition Late Glacial/Holocene is not represented by
carbonate sediments nor here. These results are in good
Fig. 6. The carbonate d13
C and d18
O data vs. model age.
Fig. 7. Comparison of the carbonate d18
O data from Svat!y Jan and
GISP 2 d18
O record.
K. $Z!ak et al. / Quaternary International 91 (2002) 137–152148
agreement with studies of ﬂuvial sediments of the largest
Bohemian Labe River (R(u$zi$ckov!a and Zeman, 1994),
where the formation of Early Holocene ﬂoodplain also
was dated around 9500 yr BP.
Other tufa deposits of the area, studied by biostratigraphic
methods only, reﬂect very similar evolutional
patterns (Lo$zek, 1992). Tufas containing molluscan
assemblages typical of the transition from the Late
Glacial to Holocene are rare within the Bohemian
Karst. One of such accumulation was found by Lo$zek
(1967b, 1968) in the valley of $Svarcava Creek, 12 km
east of Svat!y Jan pod Skalou. An approximately 3 m
thick accumulation of tufa of braided ﬂuvial type forms
a low creek terrace incised by younger erosion. The
basal sandy and muddy sediments at $Svarcava contained
three Late Glacial elementsFColumella columella,
Vallonia tenuilabris and Vertigo parcedentataFwhile
the overlying loose tufa layer was rich in typical Early
Holocene Discus ruderatus-assemblage with Perpolita
petronella, Clausilia cruciata, Vertigo substriata and
abundant Vallonia costata (Lo$zek, 1967b, 1968; Lo$zek
in $Z!ak et al., 2001). Surprisingly, carbonate d13
C and
d18
O data from the lower portion of this section were
within the range of data from the lower portion of the
Svat!y Jan pod Skalou tufa. One 14
C AMS date of
sedimentary organic matter of fossil soil from the basal
part of the calcareous tufa layer from $Svarcava yielded a
corrected 14
C age within an interval 8954–8554 yr BP
(sample ref. no. NZA 7702, 37.0570.41 pmc, $Z!ak et al.,
2001), excluding the possibility that intensive carbonate
tufa deposition occurred at this locality at the Late
Glacial/Holocene transition. This 14
C date supported
the idea of almost synchronous onset of intensive tufa
formation in the whole Bohemian Karst. In the case of
the $Svarcava section, either some shells of Mollusca were
re-deposited under ﬂuvial conditions or a relic assemblage
survived for some time in a narrow, inversion
valley of $Svarcava.
6.2. Phase of humid stable climate in the Atlantic
During the phase of humid and stable climate,
intensive calcareous tufa growths are evident at most
sites. Massive accumulations of solid phytohermal tufa
with very limited quantities of clastic material were
formed during this period. Based on results obtained
from the Svat!y Jan pod Skalou section, the maximum
growth of massive structural calcareous tufa occurred
between 8 400 and 6500 yr BP. In the surrounding
areas brown decalciﬁed rendzina-type soils were
formed with some residual corroded clasts of limestone.
Transport of slope scree was very limited. Based on
the d18
O record, the temperature reached its maximum
around 7500 yr BP. Generally, the mean
annual temperatures of this period were only slightly
higher than during the later period. More importantly,
annual precipitation was higher and oceanic-type
climate prevailed with smaller temperature differences
between winters and summers. The landscape was
almost continuously covered by forests inducing increased
evapotranspiration and steady stream discharge
conditions.
6.3. Phase of oscillation between dry and wet climate
The start of next phase is limited approximately to
6500 yr BP in the Bohemian Karst. The phase is
characterized by short rapid oscillations of dry and
wet periods, which resulted in alternation of structural
tufa and loose tufa with intercalations of soil and scree
horizons. In several sections located in the Bohemian
Karst, up to 5 dry oscillations can be identiﬁed. The
duration of these dry oscillations is not precisely known.
The phase spans about 4000 yr. Based on paleopedological
analysis, the buried soils preserved at Svat!y Jan
pod Skalou were classiﬁed as autochthonous or semiautochthonous.
The lithological characteristics of the
more than 6 m thick upper portion of the studied section
indicate that dry periods probably were not longer than
several hundreds of years.
Based on the study of peat proﬁles in Denmark, Aaby
(1976) concluded that climatic oscillations occurred each
260 or, less frequently, each 520 yr. An 800-yr periodicity
was found as dominant in peat sections in N
England (Barber et al., 1994). Similar periodicities with
several hundred-years order also were recorded through
study of glacier movements (Grove, 1979). The d18
O
record from the Svat!y Jan pod Skalou shows that
changes in precipitation quantity were more important
during these climatic oscillations than changes in
temperature. In addition to dry periods, this phase is
characterized also by ﬂash-ﬂoods that transported
clastic material through karstic valleys which are dry
at present.
Local deforestation that can be probably attributed to
both climate and expanding pasture can be documented
biostratigraphically during the younger part of this time
period. This feature, together with irregular precipitation
during the year, produced a distinct ﬂood event
around 3100 yr BP, recorded at Svat!y Jan pod Skalou by
the limestone scree horizon overlain by the calcareous
tufa dated both archaeologically and by 14
C to
approximately 2700 yr BP.
Wet and cooler climate with frequent temperature
oscillations around 01C is generally more effective for
scree formation below the slopes and overhanging cliffs
(Lo$zek and C!ılek, 1995). At the same time, a signiﬁcant
increase in precipitation was recorded biostratigraphically
in lake sediments in the $Sumava Mountains
(Vesel!y, 1998) and by expansion of wetlands in lowelevation
areas (e.g., around 2650 BP in the Netherlands,
Geel et al., 1996).
K. $Z!ak et al. / Quaternary International 91 (2002) 137–152 149
6.4. Termination of tufa formation
Not only the synchronous start of the tufa formation
of the area is expected on the basis of the same
lithological and biostratigraphical pattern, but the
termination as well. The formation of calcareous tufa
bodies was terminated at numerous localities in central
Europe between 2500 and 2200 yr BP. This stage is
characterized by several short dry periods, indicated by
spreading human settlement across river ﬂoodplains
(Bouzek, 1993). During the Subatlantic period, the
deposition of calcareous tufa almost ceased in the
Bohemian Karst. The Late Holocene decline of calcareous
tufa formation in central Europe was probably
connected with a decrease in the quantity of inﬁltrating
precipitation and/or with changes in soil processes as
discussed above.
Thick valley-side-sited spring accumulations were
later typically cut by younger erosion that often incised
them down to the bedrock. Precise dating of these
erosion events is not available: some could occur as late
as in the Middle Ages, when enormous ﬂoods were
recorded on Bohemian rivers. The karstic springs often
moved and started to discharge at the base of individual
tufa bodies. There still exist karst springs that precipitate
present-day tufas, but their total number is
diminished signiﬁcantly when compared with the situation
some 3000 yr ago. We believe that tufa termination
could be caused by a complex interplay of prehistoric
man and natural factors entering a new phase of
environmental co-evolution, but the precise evidence is
lacking.
6.5. Regional comparisons and Holocene climatic
provinces
The biota evolution inﬂuenced by Holocene climatic
changes in central Europe has been reviewed several
times (e.g. Lo$zek and C!ılek, 1995). These studies
concluded that the temporal evolution of calcareous
tufa bodies has numerous similar features in a large
territory limited by the Danube River in the south, by
German and Polish lowlands in the north, by Slovak
Karst in the east, and by Thuringia in the west. These
similarities indicate that regional climatic factors were
the main control on calcareous tufa formation, while
local factors were of only minor importance. The same
roughly synchronous processes were involved in Late
Holocene erosion and destruction of tufa bodies (J.ager
and Lo$zek, 1968; Hennig et al., 1983; Pazdur, 1988;
Lo$zek and C!ılek, 1995; Lo$zek, 1997; C!ılek, 1997). A
similar pattern of calcareous tufa formation has been
observed over much of Europe. It has been postulated
that in the Late Holocene (since ca. 2500 BP) there was a
marked decline in the deposition of tufa (see Goudie
et al., 1993, for a review of available data).
The Central European Holocene strata developed in
tufas, slope sediments and in the cave entrances provide
certain patterns where the individual layers are usually
well deﬁned, with distinct terminations and characteristic
fossil contents. The Central European Holocene
stratigraphy was thus proposed on the basis of this
similar observable stratigraphic pattern. However, it
was proposed for the western part of Europe several
times (see Roberts, 1998) that stratigraphic subdivision
of the Holocene is artiﬁcial, because the Holocene
consists of numerous oscillations. The question deserves
more attention, but we propose according to J.ager
(personal communication, 2000) the possibility that
several Holocene stratigraphic provinces exist. The
oscillating divide between Central and Western European
Holocene Provinces could be indicated by
prevailing continental climate in contrast to more
oceanic western climates. The general features of
Holocene climates seem to be alike throughout the
modern European temperate zone, but regional differences
must be taken into account when we are dealing
with secondary oscillations.
7. Conclusions
The tufa body of Svat!y Jan pod Skalou represents an
exceptionally large body of Holocene fresh water
limestone and probably the site where interdisciplinary
studies of geology, paleoclimatology, malacozoology,
karst hydrography, archaeology and some related
disciplines have provided the most detailed results from
all studied Central European tufa bodies. K.-D. J.ager,
the Coordinator of INQUA Holocene Group proposed
recently (2000) the proﬁle as the International Stratigraphic
Point for the continental Holocene, and called
for publication of basic characteristics, as given in this
paper. The tufa formation started at about 9500 BP, on
a ﬂat ﬂuvial gravel terrace of Late Glacial/Early
Holocene age. Deposition of lithologically uniform,
pure hard phytohermal tufa continued until about 6500
BP. From that time more unstable climate with several
dry periods (see stable isotope results in Figs. 4–6 and
Table 1) and erosion events produced a lithologically
varied complex of loose tufa alternating with embedded
soils and scree layers (Figs. 2 and 3). Termination of
tufa deposition occurred about 2500 BP and was
followed by partial erosion connected with relocation
of the spring below the tufa body.
Acknowledgements
This study was enabled thanks to the support of
International Atomic Energy Agency, coordinated
research project No. 7119/R1/RB ‘‘Stable isotope
K. $Z!ak et al. / Quaternary International 91 (2002) 137–152150
climatic record in the Quaternary sediments of the
Bohemian Massif’’, and Grant Agency of the Czech
Republic, research project No. 205/95/1392 ‘‘Dry
climatic oscillations of the Middle Holocene’’ and
Academic project CEZ: Z3-013-91.
References
Aaby, B., 1976. Cyclic climatic variations in climate over the past 5500
years reﬂected in raised bogs. Nature 263, 281–284.
Andrews, J.E., Riding, R., Dennis, P.E., 1993. Stable isotope
compositions of recent freshwater cyanobacteria carbonates from
the British Isles: local and regional environmental controls.
Sedimentology 40, 303–314.
Andrews, J.E., Pedley, H.M., Dennis, P.F., 1994. Stable isotope record
of palaeoclimatic change in a British Holocene tufa. Holocene 4,
349–355.
Andrews, J.E., Riding, R., Dennis, P.F., 1996. Regional environmental
and climatic signatures from stable isotopes in modern terrestrial
microbial carbonates from Europe. Palaeogeography, Palaeoclimatology,
Palaeoecology 129, 171–189.
Babor, J.F., 1901. Mollusca of the Bohemian Pleistocene and
Holocene. Archiv p$r!ırodov$ed. Prozk. $Cech 11(5–14) (in Czech).
Barber, K.E., Chambers, F.M., Maddy, D., Stoneman, R., Brew, J.S.,
1994. A sensitive high-resolution record of late Holocene climatic
change from a raised bog in northern England. Holocene 4,
198–205.
Benkov!a, I., $Ctver!ak, V., 1998. ‘‘The precinct of Svat!y Jan pod
Skalou’’ (Beroun district). Archeologie ve st$redn!ıch $Cech!ach 2
(1998), 115–144 (in Czech).
Bouzek, J., 1993. Climatic changes: new archaeological evidence from
the Bohemian Karst and other areas. Antiquity 67, 386–393.
Buzek, F., Kadlecov!a, R., $Z!ak, K., 1998. Nitrate pollution of karstic
groundwater system. Isotope techniques in the study of environmental
change. In: Proceedings of an International Symposium on
Isotope Techniques in the Study of Past and Current Environmental
Changes in the Hydrosphere and Atmosphere, Vienna,
Austria 14–18 April 1997, IAEA–Proceeding Series, SM–349/31.
Internat Atomic Energy Agency, Wien, pp. 453–464.
C!ılek, V., 1997. Mid-Holocene dry spells in Bohemia, Central Europe.
In: Dalfes, N., Kukla, G., Weiss, H. (Eds.), Third Millenium BC
Climate Change and Old World collapse, NATO ASI Series 49,
Springer, Berlin, pp. 309–320.
Dorale, J.A., Gonzalez, L.A., Reagan, M.K., Pickett, D.A., Murell,
M.T., Baker, R.G., 1992. A high resolution record of Holocene
climate change in speleothem calcite from Cold Water Cave,
Northeast Iowa. Science 258, 1626–1630.
Ford, D., Pedley, H.M., 1996. A review of tufa and travertine deposits
of the world. Earth Science Review 41, 117–175.
Gascoyne, M., 1992. Paleoclimate determination from cave calcite
deposits. Quaternary Science Review 11, 609–632.
Geel, Van B., Buurman, J., Waterblok, H.T., 1996. Archaeological and
palaeoecological indications of an abrupt climate change in The
Netherlands, and evidence from climatologic teleconnections
around 2650 BP. Journal of Quaternary Science 11, 451–460.
Goudie, A.S., Viles, H.A., Pentecost, A., 1993. The late-Holocene tufa
decline in Europe. Holocene 3, 181–186.
Grootes, P.M., Stuiver, M., White, J.W.C., Johnsen, S., Jouzel, J.,
1993. Comparison of oxygen isotope records from GISP2 and
GRIP Greenland ice cores. Nature 366, 552–554.
Grove, J.M., 1979. The glacial history of Holocene. Progress in
Physical Geography 3, 1–54.
Hennig, G.J., Gr.un, R., Brunnacker, K., 1983. Speleothems,
travertines and paleoclimates. Quaternary Research 20, 1–29.
Horvatin$ci!c, N., Srdo$c, D., $Silar, J., Tvrd!ıkov!a, H., 1989. Comparison
of the 14
C activity of groundwater and recent tufa from karst
areas in Yugoslavia and Czechoslovakia. Radiocarbon 31,
884–892.
J.ager, K.-D., Lo$zek, V., 1968. Beobachtungen zur Geschichte der
Karbonatdynamik in der Holoz.an Warmzeit. $Cesk!y Kras 19,
7–20.
Kadlecov!a, R., $Z!ak, K., 1998. Karst springs of the Bohemian Karst.
$Cesk!y Kras (Beroun) 24, 17–34 (in Czech).
Kovanda, J., 1964. Quaternary freshwater limestones of CzechoslovakiaFraw
materials for agriculture, industry, and export. Geol.
Pr(uzk. 11, 323–324 (in Czech).
Kovanda, J., 1971. Quaternary limestones of Czechoslovakia. Sbornik
geologick!ych V$ed, $R. Antropozoikum 7, 5–236 (in Czech).
Lo$zek, V., 1955. Mollusks of the Czechoslovak Quaternary. Rozprary
!Ust$redniho !Ustavu geologick!eho 17, 75–80 (in Czech).
Lo$zek, V., 1959. Quaternary travertines of Czechoslovakia. $Casopis
Mineralogii Geologii 4, 85–90 (in Czech).
Lo$zek, V., 1963. ‘‘P$enovec’’Fa new Czech term for loose and
partly consolidated calcareous tufa. $Cesk!y Kras 14, 113
(in Czech).
Lo$zek, V., 1964. Quart.armollusken der Tschechoslowakei. Rozpr.
!Ust$redniho !Ustaw geologick!eho 31, 1–1374.
Lo$zek, V., 1967a. Svat!y Jan pod Skalou, okr. Beroun. In: Kliewe, H.,
et al. (Eds.), Probleme und Befunde der Holoz.anstratigraphie in
Th.uringen, Sachsen und B.ohmen, Quart.arkomitee der DDR.
Berlin–Prag, pp. 168–169.
Lo$zek, V., 1967b. Holoz.ane Binnenwasserkalke und klastische
Hangsedimente im B.ohmischen Karst. In: Kliewe, H., et al.
(Eds.), Probleme und Befunde der Holoz.anstratigraphie in
Th.uringen, Sachsen und B.ohmen, Quart.arkomitee der DDR.
Berlin–Prag, pp. 137–178.
Lo$zek, V., 1968. Paleogeographical research of the Czechoslovak
Holocene. Zpr. geol. V!yzk. v Roce 1966, 355–357 (in Czech).
Lo$zek, V., 1992. Netz von St.utzproﬁlen zur Landschaftsgeschichte des
B.ohnisches Karstes. Bohemia centr. 21, 47–67 (in Czech with
German abstract).
Lo$zek, V., 1997. Development of sediments, soils, erosional events,
molluscan and Vertebrate assemblages in connection with human
impact in Central Europe during the time-span 3000–5000 BP. In:
Dalfes, H.N., Kukla, G., Weiss, H. (Eds.), Third millenium BC
Climate Change and Old World Collapse, NATO ASI Series, Series
I: Global Environmental Change, Vol. 49. Springer, Berlin, pp.
551–564.
Lo$zek, V., C!ılek, V., 1994. Humidity and temperature course of the
Mid-European Holocene. In: R(u$zi$ckov!a E., Zeman A. (Eds.),
Manifestation of Climate on Earth’s Surface at the End of
Holocene. PAGES Stream I Proceedings, pp. 79–81.
Lo$zek, V., C!ılek, V., 1995. Late Weichselian–Holocene sediments and
soils in Mid-European calcareous areas. Sbornik geologickycl V$ed,
$R. Antropozoikum 22, 87–112.
Meese, D.A., et al., 1997. The Greenland ice sheet project 2 depth-age
scale: methods and results. Journal of Geophysical Research 102,
19483–19494.
N$emejc, F., 1928. Paleobotanical study of fossil travertins sediments
between Prague and Beroun. Rozpravy $ceskoslovenska Akademie
V$ed a um., II.t$r.36(22), 1–10 (in Czech).
Pazdur, A., 1988. The relation between carbon isotope composition
and apparent age of freshwater tufaceous sediments. Radiocarbon
30, 7–18.
Pazdur, A., Pazdur, M.F., Starkel, L., Szulc, J., 1988. Stable isotopes
of Holocene calcareous tufa in Southern Poland as paleoclimatic
indicators. Quaternary Research 30, 177–189.
Petrbok, J., 1923. Mollusks of travertine (calcareous tufa) of the
Beroun area.. $Casopis N!arodn!ıho Musea, odd!ıl p$r!ırodov$edn!y
XCVII, 12–18 (in Czech).
K. $Z!ak et al. / Quaternary International 91 (2002) 137–152 151
Petrbok, J., 1956. Research in the Bohemian Karst until 1950.
Antropozoikum V (1955), 9–46 (in Czech).
Roberts, N., 1998. The Holocene. An Environmental History, 2nd
Edition. Blackwell Publishers, Oxford, 316pp.
R(u$zi$ckov!a, E., Zeman, A., 1994. Holocene ﬂuvial sediments of the
Labe river. In: R(u$zi$ckov!a, E., Zeman A. (Eds.), Holocene Flood
Plain of the Labe River. Geol. !ust. Akad. v$ed. $Ces. Republ. Praha,
pp. 3–25.
$Sibrava, V., Fejfar, O., Kovanda, J., Valoch, K., 1969. Quaternary in
Czechoslovakia. History of investigations between 1919 and 1969.
!Usted!ni !Ustau geologick!y., Praha.
$Silar, J., Z!ahrubsk!y, K., 1999. Consetraci!on inicial de radiocarbono en
agua subterr!anea en el karst de centro de Bohemia. In: Proceedings
I Simposio de Geolog!ıa. Cueva de Nerja, Spain.
$Silar, J., J!ılek, P., Melkov!a, K., Tvrd!ıkov!a, H., 1990. Radiocarbon
dating of Holocene tufa deposits in the Barrandian region, Central
Bohemia. Freiberger Forschungshefte Hefte, Reihe C 442, 31–41.
Sowers, T., Bender, M., Labeyrie, L., Martinson, D., Jouzel, J.,
Raynaud, D., Pichon, J.J., Korotkevich, Y.S., 1993. A 135 000 year
Vostok–SPACEMAP common temporal framework. Paleoceanography
8, 737–766.
Stuiver, M., Groetes, P.M., Braziunas, T.F., 1995. The GISP2 d18
O
climate record of the past 16,500 years ant he role of the sun, ocean
and volcanoes. Quaternary Research 44, 341–354.
Vesel!y, J., 1998. Changes in vegetation of the $Cern!e lake are inferred
from pollen analysis of lake sediment: period between 3400 BC and
1600 AD. Silva Gabreta 2, 141–153.
$Z!ak, K., Hlad!ıkov!a, J., Buzek, F., Kadlecov!a, R., Lo$zek, V., C!ılek, V.,
Kadlec, J., $Zigov!a, A., Bruthans, J., $St’astn!y, M., 2001. Calcareous
tufa accumulation of Holocene age and karst spring in Svat!y Jan
pod Skalou in the Bohemian Karst. Czech Geol. Survey Special
Paper, 13. 135pp, Czech Geological Survey (in Czech with English
summary).
K. $Z!ak et al. / Quaternary International 91 (2002) 137–152152