Quaternary International 91 (2002) 101–122
A high-resolution record of Late-Glacial and Early-Holocene climatic
and environmental change in the Czech Republic
Petr Pokorn!ya,b,
*
a
Institute of Botany, Academy of Sciences of the Czech Republic, Dukelsk!a 143, CZ-37982 T$rebo$n, Czech Republic
b
Institute of Archaeology, Academy of Sciences of the Czech Republic, Letensk!a 4, 11801-Praha, Czech Republic
Abstract
Recent discovery of thick buried lake sediments in the T$rebo$n Basin, South Bohemia, presents an exceptional opportunity to
study the Late-Glacial history of eastern Central Europe. High-resolution investigation of pollen, plant macrofossils, algal remains,
and lithology for the Late-Glacial and Early-Holocene sediments of former Lake $Svarcenberk yielded a well-founded palaeoclimatic
and palaeovegetational data that can be compared with the results from other parts of a west–east European transect, taking into
account the oceanic/continental gradient and its inﬂuence for palaeoenvironmental conditions. The results demonstrate that the
effect of North Atlantic oceanic changes during the last glacial–interglacial transition extended to the investigated area.
Nevertheless, signiﬁcant differences in timing, intensity, and character of vegetational response to these climatic changes have been
found between the area under study and the western part of Central Europe. These differences can be ascribed to increased
seasonality, speciﬁc regional mesoclimatic and soil conditions, and possible local glacial refugia for pine. The results of sediment
chemical analyses indicate a close correspondence between climatic, vegetational, and soil development in lake catchments. r 2002
Elsevier Science Ltd and INQUA. All rights reserved.
1. Introduction
Organic deposits suitable for palaeoecological research
usually began to form about 13 ka BP in western
and northwestern Europe, classiﬁed into different LateGlacial
phases by pollen analysis. In non-glaciated,
continental regions of Central Europe, this subdivision
is usually not possible, either because minerogenic
sediments contain no pollen or because sediments
simply did not accumulate at this time. Particularly,
lacustrine sequences are very rare in these regions. The
proﬁle under study in the Czech Republic is a unique
example with an extensive and well-stratiﬁed LateGlacial
record. High sediment-accumulation rates
permit the detection of brief Late-Glacial climatic oscillations,
so that comparisons can be made with numerous
results from western and northwestern Europe,
where the basic biostratigraphic and climatostratigraphic
concepts have been developed (Iversen, 1954;
Mangerud et al., 1974; Watts, 1979). As postulated by
Ruddiman and McIntyre (1981) and later recognised in
terrestrial records within the areas adjacent to the North
Atlantic (e.g. Lowe et al., 1994; Walker, 1995), the rapid
climatic changes during the last glacial–interglacial
transition can be ascribed to large-scale shifts in the
position of the oceanic Polar Front, which have amphiAtlantic
or even global effects (e.g. Peteet, 1995).
The goal of the present study is to test the validity of
some of these concepts as applied to the eastern part of
Central Europe. The great distance from the North
Atlantic and from major ice sheets, as well as the high
degree of regional environmental diversity, are the most
important factors that could cause certain differences in
climate and provoke distinct biotic responses to climatic
changes in the area under study. It is well known that
the response of populations (e.g. plant populations) to
climatic change is likely to be greatest near the margin of
their distributional limits (Watts, 1979). Rapid climatic
changes during the last glacial–interglacial transition
usually affected only local populations and did not
permit long-distance migrations (Ammann, 1989). This
caused a high degree of inter-regional biological
diversity, depending on local availability of species.
For example, in the intermontane basins of the Western
Carpathians, some 250 km east of the study area, local
*Address for correspondence: Institute of Botany, Academy of
Sciences of the Czech Republic, Dukelsk!a 143, CZ-37982 T$rebo$n,
Czech Republic.
E-mail address: ppokorny@butbn.cas.cz (P. Pokorn!y).
1040-6182/02/$ - 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 5 - 7
populations of Pinus cembra, P. sylvestris, Larix
decidua, and Picea abies responded to the onset of
Late-Glacial climatic amelioration by expansion to
higher altitudes from their locally present glacial refugia
(Ralska-Jasiewiczova, 1980; Jankovsk!a, 1984; Rybn!ı$cek
and Rybn!ı$ckov!a, 1994). This pattern is completely
different from any other recognised in Europe north of
the Alps, which probably had no local glacial refugia.
In spite of the problems with absolute dating of LateGlacial
events due to the occurrence of 14
C plateaus, the
biostratigraphic subdivision of NW European LateGlacial
is today well ﬁxed on an absolute time scale
(Ammann and Lotter, 1989; Wohlfarth, 1996, Hoek,
1997). In this paper, detailed correlation of recognised
biostratigraphic boundaries with events described from
other parts of Europe is not possible on an absolute
chronological basis due to the lack of sufﬁcent AMS
radiocarbon dates of terrestrial plant remains. Biostratigraphical
terminology is therefore followed in this
paper. Local pollen assemblage zones (PAZ) are
correlated with regional climatostratigraphical units
according to Mangerud et al. (1974) and Ammann and
Lotter (1989). This correlation is only roughly conﬁrmed
by available 14
C dates.
2. The study area and site
The study site is situated in South Bohemia in the ﬂat
landscape of the T$rebo$n Basin (which has an area of
about 700 km2
and maximum relief undulation of
20–40 m). Sandy and clayey Cretaceous sediments with
locally superimposed Tertiary sediments constitute the
principal geological substratum. Depressions are ﬁlled
with Quaternary alluvial silt and gravel, aeolian sands,
and particularly peat bogs. The content of clay in soils
generally increases with depth, and soil aeration is
reduced accordingly. The soil nutrient content is
generally poor. Calcium carbonate deﬁciency is common,
potassium is sufﬁcient only in deep soil horizons,
nitrogen content is low, and concentrations of phosphate
are moderate (Hus!ak and Hejn!y, 1978). Most soils
are leached and show a tendency towards podzolization.
The soil reaction is mostly highly acidic (pH up to 3.3).
Various types of podzols and sandy or peaty gleysols
prevail.
The present climate is suboceanic and is determined
by prevailing westerly air masses, already signiﬁcantly
reduced in moisture by passage across central Europe.
The region is somewhat sheltered by the $SumavaBavarian
Forest highlands. Macroclimatic conditions
are strongly modiﬁed by presence of extensive wetlands:
Frequent occurrence of fog is typical for the region.
Annual mean precipitation is 622 mm (January being the
driest month), annual mean temperature is 7.41C (see
Fig. 1b).
The potential vegetation of the T$rebo$n Basin would
generally be silver ﬁr-oak woodlands (Abieti-Quercetum),
in waterlogged areas bird cherry-pendulate oak
and -alder woodlands (Quercus robur-Padus avium and
Alnus glutinosa-Padus avium communities), alder carrs
(Carici elongatae-Alnetum), and reed swamps and tallsedge
communities (Phragmito-Magnocaricetea) (Neuh!auslov!a,
1998). Vast transitional peat bogs are still
dominated by forests of Bog-pine (Pinus rotundata).
Acidophillous pine (Pinus sylvestris) forests with oak
(Quercus robur), birch (Betula pubescens), and spruce
(Picea abies) prevail on dry soils of river terraces and
aeolian sands.
The T$rebo$n Basin, originally an inaccessible swampy
area, remained largely a wilderness until the 13th
century. During Late Medieval time, it developed into
a cultural landscape of ﬁsh culture and forest plantations,
and ﬁshponds still constitute a characteristic
element of the landscape.
The ﬁrst palynological investigation of the area was
carried out by Rudolph (1917). While his primary focus
was the investigation of plant macrofossils in several
peat bogs, he supplemented his results by analysis of
some types of arboreal pollen. According to his results,
the basal age of some investigated deposits was later
established to be of ‘‘Kiefern-Zeit’’ (Rudolph and
Firbas, 1922). The early postglacial age of most peat
deposits in T$rebo$n basin was later conﬁrmed by Kle$cka
(1926, 1928) and $St$ep!anov!a (1930). Small pollen counts
and exclusive focus on arboreal pollen were the
disadvantages of these early palynological investigations.
In the early 1960s, Jankovsk!a started her
palaeoecological investigations of T$rebo$n Basin, using
modern approaches. Her work (Jankovsk!a, 1980) forms
the basis of our knowledge of the area.
The former lake $Svarcenberk is situated 4 km south of
the city of Vesel!ı nad Lu$znic!ı (49190
N, 141420
E) at 412 m
a.s.l. Limnic sediments are overlain by peat, which
formed after the natural inﬁlling of the lake at approx.
5500 BP according to 14
C dating. Nowadays, the site is
heavily inﬂuenced by intensive management. Between
1698 and 1701 a dammed ﬁshpond was constructed
directly on the site, and its waters almost completely
ﬂooded over the peat and the underlying lake sediments.
The only presumed remnants of the original vegetation
cover are the small patches of tall sedge and Sphagnum
communities (Eriophorion gracilis and Rhynchosporion
albae) in western part of the locality.
For local hydrology the presence of several strong
artesian springs is characteristic. The underground
water ascends along a deep tectonic fault and is rich in
iron oxides. The former lake was presumably supplied
almost exclusively by this artesian water. The activity of
underground water sources was apparently independent
of general climatic ﬂuctuations in the past. As a result,
lake level remained almost constant over the millennia
P. Pokorn!y / Quaternary International 91 (2002) 101–122102
and cannot be used as a climatic indicator. The lake
drained into nearby Lu$znice River. For evaluating the
regional pollen rain on the site, the vicinity to the river
ﬂoodplain is important.
Numerous aeolian deposits are situated along the
river ﬂoodplain. One of the biggest and most prominent
sand dunes, ‘‘Vlkovsk!y p$resyp’’, is situated 1200 m from
the former lake basin. Its relative height over surrounding
terrain (6 m), its size (60 Â 80 m), and the unvegetated
character make it a prominent structure on the
landscape. According to its morphology, ‘‘Vlkovsk!y
p$resyp’’ sand dune was formed by easterly winds. The
material constituting it is derived directly from the
nearby river ﬂoodplain (carried by wind for tens or
hundreds of meters) according to mineralogical analyses
(Ch!abera, 1982).
The existence of the former Lake $Svarcenberk was
noted for the ﬁrst time by Vlasta Jankovsk!a in the late
1970s. In her study, which focused on the vegetational
development of T$rebo$n Basin (Jankovsk!a, 1980), she
presents a pollen diagram and macrofossil analysis
obtained from an open pit. Her proﬁle comprised about
1.5 m of lake sediments, and she correctly assumed that
she dealt with the littoral facies of a larger lake.
Unfortunately, no stratigraphic data were obtained.
The present study fully conﬁrms Jankovsk!a’s original
assumption.
3. Methods
3.1. Field methods, sediment description, and
subsampling
During the pilot study, the extension and stratigraphy
of the former lake basin was studied by coring in a
100m Â 100m grid (sampling distances were reduced
along the shores). For subaquatic coring, a boat was
used. Two right-angle transects across the basin were
chosen as reference sections (shown in Fig. 1c and 2).
They have their crossing point in the centre of the basin,
where the ‘‘main proﬁle’’ is situated. The distances in
metres and the geographical position (N, S, E, W) in
relation to this zero point is given by the core labels. The
cores were levelled according to ﬁshpond water level.
The coring was performed with Russian-type corer
Fig. 1. The study site. (a) Location of the study area within the Czech Republic. (b) Climatic diagram for T$rebo$n (ca. 15 km south from the site)
derived from 50 yr of observations. (c) Quaternary geology and topography of the site. 1F‘‘Vlkovsk!y p$resyp’’ aeolian sand dune, a,b,c,dFtwo
cross-sections used for stratigraphic investigation. Their crossing point is in the centre of the basin, where the ‘‘main proﬁle’’ is situated.
P. Pokorn!y / Quaternary International 91 (2002) 101–122 103
(Jowsey, 1966) 5 cm in diameter. The sampling point
S500 was opened by hand-made pit. Correlation of the
individual cores across the basin was achieved by visual
stratigraphic description and further conﬁrmed by
pollen analysis in some cases.
The core in the centre of former lake was selected as
the standard proﬁle. The central core will most probably
show a continuous record without hiatuses, and it is
more likely to give an ‘‘average’’ picture of the events in
the basin and its catchment (without the background of
local ‘‘noise’’, which is assumed to be greatest along the
shores). This ‘‘main proﬁle’’ actually consists of seven
separate parallel cores taken close together in order to
obtain enough material for all kind of analyses. The
coring was performed using the Russian-type corer 5 cm
in diameter. This type of device permits complete
Fig. 2. Two selected orthogonal stratigraphic cross-sections (their position described on Fig. 1) through the $Svarcenberk lake basin.
P. Pokorn!y / Quaternary International 91 (2002) 101–122104
recovery of the section penetrated. All cores comprising
the ‘‘main proﬁle’’ were correlated according to their
visual lithostratigraphy. This correlation only conﬁrmed
the satisfactory accuracy of parallel sampling.
Sediment description follows the system of TroelsSmith
(1955) as modiﬁed by Aaby and Berglund (1986).
The colours were determined according to standard
(Munsell) soil colour charts. The subsampling intervals
depended on the required temporal resolution and the
sample volume needed for each analysis. Closer
subsampling was undertaken in order to get enough
material for additional analyses if needed (as advised by
Moore et al., 1991). The stratigraphy of ‘‘Vlkovsk!y
p$resyp’’ sand dune was studied in an open ditch.
3.2. Macrofossil analysis
After subsampling for other analyses, the remaining
material was used for macrofossil analysis. Contiguous
samples 10 cm long were cut, and the volume of each
was determined (approximately 250 ml in all cases).
Macrofossils were extracted by heating each sample for
5 min in a 5% potassium hydroxide (KOH) solution and
sieved with running water. Sieves with mesh sizes of 200,
300 and 700 mm were used. The residues were examined
under a dissecting stereomicroscope. The absolute
number of each kind of macrofossil was recalculated
to a standard volume of 500 cm3
fresh sediment. For
determination of the seeds/fruits, a reference collection
and the atlas of macrofossils (Kac et al., 1965) were
used.
3.3. Pollen analysis
The samples used for pollen and other microfossil
analyses were prepared by a modiﬁed acetolysis method.
As the Late-Glacial part of the core had a more or less
mineral character, the samples were pre-treated with
concentrated (35%) cold hydroﬂuoric acid (HF) for 24 h
(Faegri and Iversen, 1989; Moore et al., 1991). The
extracted microfossils were lightly stained by 0.3%
safranin and mounted in a liquid glycerol-water (1:1)
mixture. In each sample at least 1500 pollen grains were
counted, except in the lowermost samples poor in
pollen, where only 500–700 grains were identiﬁed. For
pollen identiﬁcation, the following keys were used in
addition to a reference collection: Faegri and Iversen
(1989); Moore et al., 1991; Punt (1976–1996). Pollen
nomenclature follows ALPADABA (Alpine Palynological
Data-Base, housed at the Geobotanical Institute,
Bern).
An attempt was made to subdivide Alnus viridis and
Betula nana in the pollen diagram. Alnus viridis pollen
was found to be relatively easily separable. More
difﬁculties were connected with Betula nana pollen
identiﬁcation. As the difference in pollen morphology
between Betula nana and arboreal birch species is largely
quantitative (Birks, 1968; Usinger, 1975), exact separation
was impossible in all cases. When uncertainties
occurred, problematic pollen was ranked as ‘‘Betula
indet.’’ This category represents typically about 30% of
total Betula pollen in the Late-Glacial sequence. It is
included in the pollen sum but is not shown in pollen
diagram. Algae and other microfossils were identiﬁed
with the help of the publications of Van Geel et al.
(1981, 1983, 1989), Jankovsk!a (1983), and Jankovsk!a
and Kom!arek (1982).
Absolute pollen concentrations were estimated by the
volumetric method (Davis, 1965). As the volumetric
method has been proved by Walker et al. (1994) to give
results with a relatively high standard deviation, pollenconcentration
determinations were not used for polleninﬂux
calculations. The pollen-concentration curve has
been included only in the main pollen diagram (Fig. 3)
to give a rough idea about absolute pollen content of the
sediment.
The lowermost Late-Glacial and particularly Pleniglacial
sediments contained reworked Tertiary pollen.
An attempt at reliable separation of these pollen grains
was made in order to avoid possible misinterpretations
of Late-Glacial pollen spectra and to obtain a tool for
assessment of erosion rates within the lake basin and its
catchment. In this approach it is assumed that reworked
pollen grains embedded in Late-Glacial and Holocene
sediments originated primarily from eroded Tertiary
subsoil sediments in the lake catchment, including lake
shores. Some of the Tertiary taxa are easily recognised
according to morphology (e.g. Engelhardtia, Carya,
Liquidambar), but some are difﬁcult or even impossible
to separate in a morphological basis only. Fortunately,
the state of preservation of the Tertiary pollen was
relatively poor, giving pollen grains a ‘‘ghostly’’
appearance. It has been assumed that different degree
of fossilisation resulted in different chemical composition
and physical properties of the exine. Considering
this assumption, a simple method was developed to
conﬁrm the Tertiary origin of some pollen grains. Pollen
preparations, already stained by 0.3% safranin, were
bleached in 40% ethanol for 1 min. Then they were
centrifuged and transferred to glycerol-water mounting
medium again. As a result of this procedure, all pollen
grains were bleached to different degrees, but the
Tertiary ones became almost completely transparent
and hence easily separable from the Late-Glacial ones.
3.4. Standard pollen diagram
The selection of types included in the pollen sum is
always an important stage in the interpretation of
palynological results. Excluded from the sum, therefore,
are types potentially produced by the local aquatic and
marsh vegetation of the lake. In case of the Late-Glacial
P. Pokorn!y / Quaternary International 91 (2002) 101–122 105
Fig. 3. Late-Glacial and Early-Holocene pollen diagram of the ‘‘main proﬁle’’. Only selected types are included.
P.Pokorn!y/QuaternaryInternational91(2002)101–122106
however, it is difﬁcult to evaluate exactly the role of
certain wetland taxa (e.g. Cyperaceae, Equisetum, Salix)
in the regional vegetation cover, as we do not have exact
modern analogues to the Late-Glacial communities
(Watts, 1979; Moore, 1979). Percentage values were
therefore calculated on the basis of the AP+NAP pollen
sum, excluding only submerged and ﬂoating-leaf aquatics
but including monolette and Equisetum spores (in
many plant communities, these taxa usually have an
ecological role equivalent to that of higher plants).
Concealed, corroded, degraded, and well preserved but
indeterminable pollen grains were put together and
labelled ‘‘varia’’ in the pollen diagram. Printing of the
diagrams was made with the TILIA computer program,
written by E. C. Grimm (Springﬁeld, Illinois).
The pollen diagram was zoned visually, on the basis
of both presence and abundance of taxa. A more
formalised approach to delimit the local pollen assemblage
zones was also applied on the basis of three
different constrained classiﬁcation procedures implemented
in the computer program ZONE (Lotter and
Juggins, 1991). Consistency among results from these
three different zonation procedures provided the basis
for conﬁrmation and further speciﬁcation of visually
delimited local PAZ (Figs. 4–6).
3.5. Sediment chemistry and 14
C analyses
Numerous palaeolimnological studies have shown
how sediment chemical properties may be interpreted
in terms of processes acting within the lake and the
surrounding catchment (e.g. Na, K, and Mg content in
sediments directly reﬂect the intensity of weathering and
erosion). These processes are often directly or indirectly
related to climatic parameters, and their understanding
may enable further climatic reconstructions (Engstr.om
and Wright, 1984; Dearing, 1991).
Total carbon and nitrogen content was determined by
combustion at 9501C in pure oxygen, with subsequent
conductivity detection of C and N oxides (in Heraeus
CHN-Rapid Analyser). Carbonate content was measured
by sodium hydroxide titration to neutral pH after
dissolution of 0.5 g sample in 0.5 M hydrochloride
acid and boiling for 20 min (after Hammarlund and
Buchardt, 1996). Total organic carbon content was
calculated from the difference between total carbon and
carbonate carbon. The elements Ca, Mg, K, and Fe
were analysed by atomic emission spectrometry in the
Analytical Laboratory of the Institute of Botany,
Academy of Sciences of the Czech Republic, with a
Unicam 9200X AAS instrument.
Radiocarbon dates used in the present study are all
AMS dates determined from bulk sediment samples and
individual plant macrofossils (for sample description see
Table 1). The disadvantage for the Late-Glacial
sequence under the study is the absence of sufﬁcient
terrestrial plant macrofossils for dating purposes. The
dates from gyttjas, clayey lake sediments, and aquatic
plant macroremains are known to give ages that usually
exceed those obtained from terrestrial macrofossils
(T.ornqvist et al., 1992). This effect is often ascribed to
the hardwater error. In the present study a hardwater
error is expected to be relatively small, as the sediments
contain negligible amounts of carbonates. Radiocarbon
analyses were carried out by the Radiocarbon Dating
Laboratory, Department of Quaternary Geology, Lund,
Sweden. The samples have been pretreated with HCl
and NaOH. Age calculations are based on a 14
C half-life
of 5568 yr. For the purposes of simplicity and
comparability, dates are expressed in uncalibrated 14
C
years before present (BP) unless otherwise stated.
3.6. Multivariate data analysis
In an attempt to study the similarities among
individual pollen spectra, a detrended correspondence
analysis (DCA) was made for a complete Late-Glacial
and Early-Holocene data set. Pollen percentages of
individual taxa were used as input data for the
CANOCO 3.10 (ter Braak, 1990) program. DCA
results, plotted as ordination diagrams on DCA axes 1
and 2, show the overall trends in the data (Fig. 7).
Similar samples are close together and dissimilar
samples are far from each other.
To study the relationship between the composition of
individual pollen spectra and sediment chemical composition,
a canonical correspondence analysis (CCA)
(ter Braak, 1986) was made for the combined data set,
including pollen percentages and chemical data. Program
CANOCO 3.10 was again used. Plot of the
centroid of sediment-chemistry variables in the intersection
of CCA axes 1 and 2 was constructed (Fig. 8) in
order to summarise the results graphically.
4. Results
4.1. Lake basin, its origin and stratigraphy
The extent of lake deposits within the basin was
mapped in detail from approximately 120 hand borings.
The altitude of individual stratigraphic transitions was
obtained by relative levelling. The former lake was
found to have a maximum surface of 0.51 km2
, and the
ratio of the surface to drainage basin to be about 1:8
(Fig. 1a). Two lithological cross-sections (Fig. 2) show
the morphometry and the inﬁlling of the depression.
Correlation of the individual cores across the basin was
achieved by visual stratigraphy, which reﬂects well the
environmental conditions during the time of sedimentation
(e.g. Younger Dryas sediments are characterised
over the entire basin by the presence of three distinct
P. Pokorn!y / Quaternary International 91 (2002) 101–122 107
Fig. 4. Late-Glacial and Early-Holocene pollen diagram of the ‘‘main proﬁle’’. Rare pollen types ordered according to results of weighted averaging.
P.Pokorn!y/QuaternaryInternational91(2002)101–122108
Fig. 5. Late-Glacial and Early-Holocene macrofossil diagram of the ‘‘main proﬁle’’. Absolute numbers of the ﬁnds were recalculated to standard volume of 500 cm3
of fresh sediment.
If not stated differently, all ﬁnds represents seeds.
P.Pokorn!y/QuaternaryInternational91(2002)101–122109
yellowish layers of aeolian material, each about 1 cm
thick). The borders distinguished between several
litostratigraphic units can therefore be assumed to be
roughly time-parallel. In two littoral sampling points,
S500 and JC-7B (the latter made and studied in SW part
of the basin by Jankovsk!a, 1980), pollen analysis
conﬁrmed this assumption. Only the limno-thelmatic
contact, marking the time of ﬁnal inﬁlling of the lake,
was expected to be metachronous over the basin.
Radiocarbon dating in the littoral part (sampling point
S500, with a date of 63407110 BP directly at the limnothelmatic
contact) and the central part of the lake (the
‘‘main proﬁle’’, with an almost identical date of
63507100 BP at a level signiﬁcantly under the limnothelmatic
contact) conﬁrmed this expectation.
The striking features of the basin morphometry are its
kidney-shaped form, surprising depth and declivity (the
presence of unusually steep slopes), and relatively great
age of its inﬁlling. Unfortunately, no radiocarbon date
exists from the basal sediments, but their age is
estimated around 16,000 BP from the pollen-analytical
results. On the basis of these ﬁndings, the origin of such
structure can be best explained as the remnant of a huge
Pleniglacial ground-ice lensFan open-system pingo.
Fig. 6. Sediment composition of the ‘‘main proﬁle’’ correlated with local pollen zonation. Simpliﬁed stratigraphic column shows different proportion
of sand (dots and circles) and silt (angles) components in organic sediments.
Table 1
AMS radiocarbon dates from $Svarcenberk littoral (S500) and central (‘‘main proﬁle’’) cores and ‘‘Vlkovsk!y p$resyp’’ sand dune (LuA-4645)
Lab. no. Core label/depth Method Type of material Measured 14
C age
LuA-4297 S500: 200 cm AMS Trapa natans nut 63407110 BP
LuA-4589 ‘‘Main p.’’: 324–327 cm AMS Trapa natans nut 63507100 BP
LuA-4590 ‘‘Main p.’’: 390–393 cm AMS Woody stem fragment 96407115 BP
LuA-4591 ‘‘Main p.’’: 520–523 cm AMS Bulk gyttja sample 10,7807115 BP
LuA-4738 ‘‘Main p.’’: 680–683 cm AMS Alkali-soluble fraction from gyttja 11,7507120 BP
LuA-4645 Surface of a fossil soil AMS Pinus charcoal fragments 11,2607120 BP
P. Pokorn!y / Quaternary International 91 (2002) 101–122110
The absence of a distinct rampart around the former
lake is not surprising, as the original geomorphology of
the site was completely disturbed by human action,
especially during the construction of the ﬁshpond in the
17th century. Moreover, some pingos do not dome very
high over the terrain, although they are relatively large
in diameter (Washburn, 1980). These structures do not
form a distinct rampart after they collapse. What is
Fig. 7. DCA ordination results of percentage pollen data from the ‘‘main proﬁle’’. The use of the symbols follows pollen zonation (LPAZ): Full
circlesFzone S6, full squaresFzone S5, open squaresFzone S4, open dotted circlesFsubzone S3a, open circlesFsubzone S3b, starsFzone S2,
crossesFzone S1.
Fig. 8. CCA ordination results. Percentage pollen data together with sediment chemistry data (as environmental variables) have been used. The use
of the symbols follows pollen zonation (LPAZ): Full circlesFzone S6, full squaresFzone S5, open squaresFzone S4, open circlesFzone S3,
starsFzone S2, crossesFzone S1.
P. Pokorn!y / Quaternary International 91 (2002) 101–122 111
more surprising in case of former lake $Svarcenberk is the
unusually big size of its depression. Considering this fact
and the observation of the ‘‘ridges’’ (Fig. 2) dividing the
basin into three main parts, the origin of the lake can be
best viewed as the remnant of some kind of a compound
pingo structure.
In the littoral parts of the former lake basin, only a
thin layer of Late-Glacial sediments is present, completely
lacking deposits older than Younger Dryas. This
can be explained by intensive reworking of the shores
during Late-Glacial rather being the result of lower lake
levels during this period. As the lake was fed almost
exclusively by artesian water, water-level remained
constant over the entire period of its existence, and
water-level reconstructions cannot be used as a climatic
indicator, unfortunately. After the ﬁnal inﬁlling of the
lake (dated to approx. 5500 BP), oligotrophic peat
started to accumulate.
4.2. Main biostratigraphic events and geomorphic
processes
Pollen stratigraphy of the ‘‘main proﬁle’’ has been
subdivided into six local pollen assemblage zones (PAZ)
and eleven subzones, which are used as a framework for
the discussion of the results. Because of terminological
problems (e.g. Ammann and Lotter, 1989; Walker,
1995) I have decided to subdivide the diagram in this
way rather than into Firbas pollen zones (Firbas, 1949),
as is traditionally done in Central Europe. The absence
of analogous results over a wide region discourages the
use of regional pollen zonation. The local PAZ are later
compared (Fig. 9) with European climatostratigraphical
units according to Mangerud et al. (1974) and Ammann
and Lotter (1989) and with the d18
O curve of the
Greenland ice core GISP2 (Stuiver et al., 1995).
4.2.1. Zone S1
The lowest sediments of the lake-basin sequence
consist of ﬁne silt with coarser sand. Absolute pollen
concentrations in the sediment are low. The zone is
characterised by high NAP values, suggesting open
herbaceous vegetation. Grasses, Cyperaceae, Chenopodiaceae,
Betula nana, Alnus viridis, Salix (most likely
some dwarf willow species), Thalictrum, and Artemisia
were important components of the vegetation. Macroscopic
stem fragments of Salix sp. were found in the
sediment. Sporadic pollen ﬁnds of Ephedra (both
Ephedra distachya and E. fragilis types) are difﬁcult to
interpret, as this type of pollen can be dispersed over
long distances (Huntley and Birks, 1983; Lang, 1994).
Pinus values are below 30%, and Betula values do not
exceed 5%; both can be ascribed to longdistance
transport. Helianthemum is indicative of bare,
Fig. 9. Local PAZ compared with bidecadal d18
O curve of the Greenland ice core GISP2 (data measured by hand by W.O. van der Knaap from
graph presented at Stuiver et al., 1995). This cross-correlation should be considered as a suggested scheme only. Absolute time scale (cal yr BP; yearly
ice-layer counts before A.D. 1950) and chronozones follow Stuiver et al. (1995) with exception of the ‘‘Preboreal oscilation’’ derived from Ammann
and Lotter (1989). A synoptic table of the main events for the site of present study is attached from the left.
P. Pokorn!y / Quaternary International 91 (2002) 101–122112
calcareous substratum (Hoek, 1997). The abundance of
Cyperaceae pollen in relation to Artemisia pollen
indicates prevailing moist conditions for zone S1. This
could be the result of the presence of permafrost and its
progressive decay from the surface. Pollen of submerged
water plants is absent, and the seeds of Ranunculus
subgen. Batrachium and Potamogeton cf. gramineus are
rare, while Charophyta oospores (cf. Chara strigosa, a
pioneer species with subarctic distribution) are exceptionally
abundant. Ranunculus subgen. Batrachium is a
climatic indicator for maximum July temperatures of at
least 101C, while Hippopha rhamnoides presence (found
as pollen) suggests at least 111C (Huizer and Izarin,
1997). Due to the absence of plant macrofossils suitable
for 14
C dating and the mineral character of the sediment,
absolute dating of this zone unfortunately was not
possible. The age is only roughly estimated to be around
16,000 BP.
Very low organic content of the sediments suggests
low productivity in the lake as well as in its catchment.
The lower Mg, Ca, and K content, if compared to the S2
zone, is most likely the result of high proportion of
coarse sediment (and high accumulation rates), rather
than lower allogenic sediment input to the lake basin.
All the results point to severe climatic conditions, the
absence of well-developed soils, and an open, treeless
vegetation of steppe and tundra during S1. A signiﬁcant
difference in the character of the vegetation between this
and the higher zones was conﬁrmed also by DCA results
(Fig. 7), for the samples of the S1 zone are plotted in the
ordination diagram far from the main population of
samples, indicating signiﬁcantly different composition
of their pollen spectra.
4.2.2. Zone S2
This zone starts with a rise in Pinus pollen percentages
up to 50%, together with a decrease in Cyperaceae and a
prominent Artemisia peak. Pollen of submerged water
plants occur for the ﬁrst time. Betula pollen curve is
continuous during the entire zone, but is too low (under
5%) to suggest the local occurrence of tree birch. As in
the previous zone, grasses, Cyperaceae, Chenopodiaceae,
Alnus viridis, dwarf willow, Thalictrum, Artemisia,
and Betula nana were important components of the
vegetation. Seeds and catkin-scales of dwarf birch
(Betula nana) were found in the sediment. Empetrum
pollen percentages around 1% suggest local presence
but not the development of extensive Empetrumdominated
heaths (according to quantitative criteria
given by Huntley and Birks, 1983). Relatively high
Helianthemum percentages (up to 3%) point to the
presence of calcareous substratum. This accords well
with the results of sediment chemical analyses, showing
high Ca content. This ﬁnding is in sharp contrast with
present-day conditions in the area under study, for most
soils are leached, highly acidic, and almost completely
lacking calcium carbonate. The absolute pollen concentrations
in the sediment are still low during this zone.
The silty FeS-coloured sediments in the centre of the
basin suggest anoxic conditions, as iron-sulphide deposition
usually occurs under prolonged or permanent
stratiﬁcation of the lake (Engstr.om and Wright, 1984).
Sediment organic content is only around 3% and reﬂects
low productivity in the lake and its catchment. High
content of Na and K and of reworked Tertiary pollen
suggest high erosion rates and the absence of stable soils
in the catchment.
4.2.3. Subzone S2a
The rise in Pinus percentages up to values around
60% suggests the onset of pine expansion in the area as
the result of some climatic amelioration. Individual pine
trees must have been scattered more or less sporadically
in the landscape of steppe or tundra-like character. The
rise in Artemisia shortly before this time can be also
considered as the ﬁrst signal of warming. The ﬁnd of
Urtica dioica seed (together with Urtica pollen) is
particularly interesting from the point of environmental
reconstruction, for this nitrophillous plant, requiring
mean July temperatures at least 81C (Bos, 1998), is
distributed in eutrophic habitats today. This is in sharp
contrast to the general picture of the S2 subzone, which
points to prevailing oligotrophic and distrophic conditions
in the lake and its catchment (see also the low
sedimentary N content). There must have existed some
favourable, nutrient-rich microhabitats (bird-manured
patches, for example), where Urtica dioica could have
prospered.
4.2.4. Subzone S2b
The signs of progressive climatic amelioration are
traceable from the pollen record during the onset of S2b
subzone as Pinus values rise up to 75%. The local
presence of pine is conﬁrmed also by the ﬁnd of Pinus
stomata. Pine became a successful competitor for light
with the heliophyllous vegetation. Overshading occurred
in moist habitats, as seen from the decline of Salix and
Cyperaceae pollen percentages, but also in drier
habitats, as Artemisia and Gramineae declined during
Pinus maxima as well. If we compare the pollen
concentration curve (attached to the pollen diagram in
Fig. 3) for S2b subzone with the others, particularly S3
and S4, zones, it become obvious, that the rise in Pinus is
not due to the percentage effect caused by generally low
regional pollen production (Moore et al., 1991). The
dating of this subzone is unfortunately uncertain, due to
the absence of appropriate material for 14
C analysis
again. It is only roughly estimated to range between
15,000 and 14,000 BP.
P. Pokorn!y / Quaternary International 91 (2002) 101–122 113
4.2.5. Subzone S2c
The decline of pine and new expansion of heliophyllous
herbs in this subzone can be correlated with the
Oldest Dryas chronozone (sensu Stuiver et al., 1995).
Unfortunately, exact 14
C dating of this event is again
missing. In the upper half of S2c subzone, a new pine
expansion anticipates the transition to zone S3. An
interesting macrofossil ﬁnd, a shrew (Sorex cf. araneus,
det. by I. Hor!a$cek) mandible, was made within this
subzone. The mandible comes from an adult specimen,
but its size is unusually small. Its phenotypic appearance
signiﬁcantly differs from contemporary populations.
Shrews were in general rare constituents of late
Pleistocene faunal assemblages. This ﬁnd is particularly
interesting from the palaeoecological point of view, for
shrews are characteristic of sparsely wooded landscapes
(I. Hor!a$cek, pers. comm.), consistent with the palaeobotanical
results that suggest sparse tree cover during
that time.
4.2.6. Zone S3
Marked vegetation changes are characteristic for this
zone. Reforestation of the landscape by birch and pine
resulted in decline in heliophyllous herbs. Absolute
pollen concentrations in the sediment were about three
times higher than in the underlying zone. Soil development
under forested conditions led to the decrease in
erosion rates (see progressive decline in sedimentary Mg
and K and sharp decline in reworked Tertiary pollen).
Decalciﬁcation of the substratum continued up to the
maximum extent (see the decline in Ca down to the
values comparable with those in the Holocene). Increased
organic production resulted in gyttja sedimentation.
Abundant perch (Perca ﬂuviatilis) scales were
found in the sediment. Perch is an anambitious ﬁsh
genus that can survive even in subarctic lakes. The fry
are produced in large quantities and feed on planktonic
or benthic organisms. Adults feed mostly on their own
young and reach high population densities. Such an
interesting cannibalic food chain has been described
from several contemporary Siberian lakes (Hol$c!ık, 1977;
Karasev, 1987).
4.2.7. Subzone S3a
This subzone starts with a rise in Salix pollen
percentages, followed by Betula increase. Salix probably
formed the shrub belt in front of the Betula forest-line
(Gaillard, 1985; Hoek, 1997), so high values of Salix
may be expected in advance of a Betula expansion. In
the second half of S3a subzone, tree birch became
dominant in the pollen spectra in place of NAP and
Pinus. Betula cf. pubescens seeds and catkin-scales were
found in the sediment. This assemblage reﬂects the
development of open boreal birch woodland with
dispersed pine trees. The lower limit of mean July
temperature needed for tree birch colonisation is usually
taken as 101C, but 121C is the optimum for the
development of Betula pubescens woodland (Birks,
1993). An open character of the forest can be inferred
from the pollen diagram, for most of the heliophyllous
herbs typical for steppe and tundra communities are still
abundant. In the aquatic environment, abrupt climatic
amelioration caused the expansion of submerged macrophytes,
including Ceratophyllum demersum, which appears
for the ﬁrst time in this subzone. The S3a subzone
is correlated with the B.olling chronozone.
4.2.8. Subzone S3b
The basal spectrum of this subzone coincides with a
strong decrease in Betula from 40% to about 20% and a
new increase in Pinus up to values about 60%. The
forest cover became more closed, as reﬂected by a
decrease in all open-communities indicators. Helianthemum
and Plantago maritima-type disappeared from the
spectra completely. Filipendula, Typha latifolia, Nymphaea,
and Nuphar appeared in and around the lake,
pointing to minimum July temperatures, at least 121C
(Huizer and Izarin, 1997). Ceratophyllum demersum
became the dominant submerged aquatic. Gradual
Betula decrease is recorded during the upper half of
this subzone. S3b is correlated with the second half of
the Late-Glacial interstadialFthe Aller.od chronozone.
This correlation has been conﬁrmed by the radiocarbon
date 11 7507120 BP.
4.2.9. Zone S4
The values of Betula decrease to about 10%, whereas
Pinus percentages are generally the same as in the
preceding zone. Alnus viridis, Salix, Betula nana,
Chenopodiaceae, and Artemisia values increase again,
suggesting a new climatic deterioration. A similar
increase, but even more prominent, is recorded in the
Juniperus curve. Juniper percentages reach a maximum
in this zone. Vegetation change reﬂecting climatic
deterioration is recorded also in the lake: Ceratophyllum
spines values decrease, and Nympheaceae trichoblasts
and Nuphar pollen are completely lacking in favour of a
large amount of Myriophyllum verticillatum and Ranunculus
subgen. Batrachium (both recorded as macrofossils).
The presence of Typha latifolia pollen in the entire
zone can be considered the proxy for minimum July
temperatures of at least 121C (Iversen, 1954; Ammann,
1989), i.e. at the same range as those inferred for the
preceding zone. This suggests that climatic deterioration
was the result of an increase in continentality rather
than a decrease in summer temperatures. Absolute
pollen concentrations are lowered relatively to S3 zone,
and the sedimentation character changes to more
minerogenic again (with lower organic carbon content).
Slight increases in erosion indicators (Mg, K, reworked
Tertiary pollen) are observed as well.
P. Pokorn!y / Quaternary International 91 (2002) 101–122114
Although clear evidence of climatic deterioration
(manifested most likely as the increase in continentality)
is found in this zone, this climatic oscillation did not
result in complete deforestation. The pine woodland
only became somewhat more open. Zone S4 is
correlated with the Younger Dryas chronozone, and
this correlation is conﬁrmed by the 14
C date 10
7807115 BP obtained from the position slightly below
the upper zone limit. It is obvious from DCA ordination
diagram (Fig. 7) that climatic deterioration led to
vegetational reversion, as can be concluded from the
shifted position of the samples along the main ordination
axis (x-axis in the diagram).
The stratigraphical investigation of ‘‘Vlkovsk!y p$resyp’’
sand dune has revealed a fossil soil buried under
more than 5 m of aeolian sands. A distinctive layer of
pine charcoal fragments buried by aeolian sands is dated
11,2607120 BP and implies that the formation of
‘‘Vlkovsk!y p$resyp’’ sand dune dates to very beginning
of theYounger Dryas period.
4.2.10. Subzone S4a
This subzone is quite uniform, with only slight
ﬂuctuations in Pinus values and some more or less
synchronous ﬂuctuations in Cyperaceae and Gramineae
that are difﬁcult to interpret.
4.2.11. Subzone S4b
Slightly higher values of Betula nana, Artemisia, and
Ranunculus subgen. Batrachium mark the lower limit of
the subzone. When these herbs subsequently decrease,
the ﬁrst possible evidence of climatic improvement is
recorded by the increase in Filipendula and onset of
Populus rational limit.
4.2.12. Zone S5
The transition from S4 to S5 zone is characterised by
a decrease in all NAP taxa, especially Alnus viridis,
Betula nana, Chenopodiaceae, and Artemisia among the
most important. At the same time, Betula, Populus,
Filipendula, and Equisetum percentages rise, while those
of Salix decrease. Most of the thermophyllous trees have
their empirical limits (sensu Faegri and Iversen, 1989) at
the beginning of this zone, and the development of
mixed deciduous forest is observed during its second
half. Picea abies already occurs in low percentages much
earlier (during the entire Late-Glacial), but these early
ﬁnds can be ascribed to long-distance transport of easily
dispersed pollen rather than local occurrence. Organic
production generally increases, and the erosion rate is
low. Organic sediment (gyttja), rich in macrofossils,
accumulated again in the basin. This zone is correlated
with the beginning of the Holocene, the Preboreal
chronozone.
4.2.13. Subzone S5a
The peak of Populus and Equisetum and relatively
high Filipendula values are characteristic for this zone,
pointing to the development of wet meadows. Populus
tremula may have been favoured as a pioneer tree in
areas that were left open during the preceding period
(Ammann et al., 1994). Alnus viridis and Betula nana, the
most characteristic tundra elements surviving from the
preceding zone, were gradually outcompeted by developing
forests, and their values fall to zero. As in the
preceding period, pine was still dominant in the regional
forest cover, but during the transition to S5b subzone
birch started expanding in place of Pinus. The short
decrease in organic production at 480 cm cannot be
attributed to any climatic oscillation without exact time
control.
4.2.14. Subzone S5b
Boreal forest dominated by birch was slowly replaced
by mixed deciduous forest in this subzone. This change
is reﬂected by the gradual decrease in the Betula curve to
10%. On numerous sites with unfavourable sandy soils,
Pinus stands still occurred, being favoured in competition
with arriving thermophyllous trees. Pine forest
persistence could also be attributed to prevailing
continental character of the climate. Numerous macrofossil
ﬁnds of Najas marina, Najas minor, and Trapa
natans are dated to about 9800 BP (calculated by linear
extrapolation from the two adjacent dates), permitting
climatic reconstruction for this period. Najas marina
suggests a mean July temperature above 151C (Lotter,
1988), and Trapa natans even more. According to Gams
(1926) and Jorga et al. (1982), water chestnut requires
mean July water temperature not below 201C and in
May, when the ﬂowers develop, at least 121C. The rapid
change to warmer climatic conditions is also evidenced
by appearance of macroscopic colonies of the thermophyllous
blue-green alga Gloeotrichia pisum (Van Geel
et al., 1989).
Possible palynological evidence for the short Preboreal
cold climatic oscillation was found, but it is weak: a
short Pinus peak accompanied by the fall in Corylus,
Ulmus, and Quercus percentages is dated slightly before
96407115 BP. Another explanation may be the opening
of the vegetation by ﬁre, but this hypothesis was not
conﬁrmed by higher presence of charcoal particles in the
corresponding layer.
4.2.15. Zone S6
Full expansion of mixed deciduous forest started in
this zone. The pollen curve of Pinus, hitherto very high,
started decreasing as the result of competition with
deciduous trees even on poor substrates, where soils had
developed since the beginning of the Holocene. Alder
carr developed on inﬁlling margins of the lake, as
reﬂected in the strong rise of Alnus glutinosa pollen
P. Pokorn!y / Quaternary International 91 (2002) 101–122 115
curve and the occurrence of alder wood and seeds in the
littoral sediments (correlated with the main proﬁle by
pollen analysis and 14
C dating). The development of
highly productive alder stands around the lake probably
contributed to increased organic deposition (increased
organic carbon content).
At the beginning of the zone, a prominent Fe peak
dated to about 8600 BP may be explained as the
reﬂection of intensive leaching caused by a more humid
climate in conjunction with the build-up of raw humus
on the soil surface (Engstr.om and Wright, 1984; Starkel,
1990). In the pollen diagram, the same period is
characterised by Picea abies expansion. The lower
boundary of zone S6 can be correlated with the
beginning of the Boreal period.
5. Discussion
5.1. Initial warming
The origin of the former lake $Svarcenberk can be best
explained as the remnant of a huge Pleniglacial groundice
lens, an open system pingo. A similar thermokarst
origin has been suggested for several semicircular
depressions in The Netherlands, Belgium, France,
Germany, and Poland (Washburn, 1980; de Gans,
1988; Hoek, 1997). The sandy geological substratum,
the presence of several strong artesian springs on the
site, and its location close to the river are factors known
to be favourable for pingo formation (de Gans, 1988;
Pissart, 1988). The occurrence of such thermokarst
phenomena has certain climatic signiﬁcance and may be
used for mean annual air temperature reconstruction,
suggesting this to be À11C or lower (Mackay, 1988).
The presence of a pingo remnant with a depth of almost
12 m indicates the minimum permafrost thickness. By
ca. 16 ka BP, advanced warming is recorded in the
eastern Alps, for at least some glaciers had receded more
than two-thirds of their original LGM length (Lundqvist
and Saarnisto, 1995). In the Swiss Alps, a very fast decay
of glacier ice must have occurred between about 18 and
15ka BP (Ammann et al., 1994). At about that time,
North American and Scandinavian ice sheets were in full
retreat (Lundqvist and Saarnisto, 1995; Tyr!a$cek, 1995).
Thermokarst lakes are usually the ﬁrst indicators of
climatic amelioration in the periglacial zone (Lundqvist
and Saarnisto, 1995). The basal age, estimated to be
around 16,000 BP for the bottommost sediments of the
$Svarcenberk central core, represents the minimum
possible age of the lake that originated by permafrost
thawing. This suggests the change from high-arctic to
somewhat warmer conditions. The vegetation cover
during this time can be reconstructed from the pollen
spectra. Treeless vegetation of steppe and tundra
character prevailed in the area.
Pinus percentages ranging between 60% and 75% are
characteristic for S2 local PAZ. Relatively straightforward
evidence of climatic amelioration is present,
especially in the S2b subzone, when Pinus percentages
reach their maximum. Pine by then was a successful
competitor for light against the heliophyllous
vegetation, for indicators of certain open communities
declined. Reconstructed vegetation cover during the
S2b subzone can be characterised as a mixture of
shrub-heath and steppe patches with scattered
pine trees. From this point of view, the macrofossil
ﬁnd of a shrew (Sorex cf. araneus) mandible in the lake
sediments is particularly interesting, for shrews are
characteristic of sparsely wooded landscapes (I. Hor!a$cek,
pers. comm.).
High Pinus percentages recorded at $Svarcenberk
during the pre-B.olling period notably exceeded those
found elsewhere in Europe at that time and suggest the
local occurrence of Pinus stands. Huntley and Birks
(1983) conclude that pollen values >50% indicate local
dominance of pine. Ammann et al. (1994) suggested the
Oldest Dryas Pinus percentages around 20% originated
from long-distance transport. The increase of Pinus
from 20% to 65% is deﬁned as the rational limit by
Ammann and Lotter (1989) and Lotter et al. (1992).
Surface pollen samples of the woodland-steppe ecotone
in Inner Mongolia suggest Pinus pollen percentages
exceeding 70% indicate dense local pine woodland (Liu
et al., 1999). In the ecotone itself and in the edge of the
steppe zone these values decrease to 30% or lower.
According to Poser (1948) the poleward limit of forest in
Europe approximates the 101C isotherm for July
(inferred minimum July temperatures for S2 zone are
111C). In the Alpine region, and also elsewhere in
western Central Europe, pine expands about 13,000 BP
or even later in the Aller.od chronozone (e.g. Watts,
1979; Hoek, 1997; Bos, 1998). This pine expansion is not
accompanied by other indicators of climatic warming
and therefore is considered to be caused by lagged
immigration (Gaillard, 1984, 1985). The early pine
expansion can be ascribed to the rapid response of
Pinus populations expanding from locally present glacial
refugia shortly after climatic amelioration. Favourable
mesoclimatic conditions in the marshy area of T$rebo$n
basin (high local humidity) might have played a
signiﬁcant role in early reforestation as well. In
continental parts of Central Europe, Pinus sylvestris
and possibly some other demanding species might have
persisted locally through the entire glacial maximum.
Pine is known to tolerate and even reproduce under
extremely severe climatic conditions (usually in dwarf
forms), either dry, windy, or cold. Unfortunately, it is
usually not possible to discriminate between a pollen
peak caused by a population invading from another
region and one caused by local expansion of the species
previously present but climatically little favoured
P. Pokorn!y / Quaternary International 91 (2002) 101–122116
(Watts, 1979). This is true mainly in the case of pine
pollen transported from a long-distance.
Unfortunately, exact dating of the period of ﬁrst pine
expansion is missing, although it certainly antedates the
B.olling chronozone and ranges approximately to the
period between 15,000 and 14,000 BP. This result is
consistent with similar evidence from southwestern
European pollen records (de Beaulieu and Reille, 1984;
Jalut et al., 1992; de Beaulieu et al., 1994) and the
ﬁndings from northwestern Norway (Vorren et al., 1988;
Alm and Birks, 1991), where early signs of climatic
amelioration occur around 15 ka BP. In northern
Germany (Meiendorf Interval) and The Netherlands
(Epe Interval), there is also evidence for some shortlived
warmer oscillation around 15 ka BP (Menke, 1968;
Kolstrup, 1980). At about that time, weak traces of
initial pedogenesis in the eastern part of Central Europe
indicate a short break in loess deposition during the
period of slightly warmer and wetter climate (Tyr!a$cek,
1995).
The terminal phase of S2 zone, reﬂected in pollen
diagram from $Svarcenberk as the period of new Pinus
decline and a new expansion of open-communities
indicators, can be interpreted as the episode of certain
climatic deterioration and can be correlated with the
Oldest Dryas chronozone (sensu Ammann et al., 1994;
Stuiver et al., 1995). In the Alps as well as in
Scandinavia and North America, the Oldest Dryas is
known to be a period of temporary glacier readvance
(Lundqvist and Saarnisto, 1995).
High values of sedimentary Mg, K, and Ca during the
entire pre-B.olling period may be explained as derived
from eroding, unstable soils. Low nutrient status of the
lake and its catchment together with low productivity
(N and organic C values are very low during that time)
were primarily caused by low energy input into the
ecosystem. The local PAZ boundaries recognised in the
pre-B.olling record from lake $Svarcenberk are compared
(in Fig. 9) with d18
O curve of the Greenland ice core
GISP2 (Stuiver et al., 1995). A certain degree of
correspondence exist between these two.
5.2. Late-Glacial interstadial
An abrupt warming is recorded in the areas adjacent
to the North Atlantic around 13 ka BP, during the
Oldest Dryas–B.olling transition (Lowe et al., 1994).
Reforestation by birch and later by pine is recorded over
most of NW and Central Europe during that time.
Closing of the immigrating forest canopy is usually
anticipated by the period of Juniperus expansion.
Juniper was an important pioneer shrub after a period
of prevailing herbaceous vegetation, and its phase is
prominent especially in the Alpine region and in Britain.
Farther to the east, this shrub appears to be much less
signiﬁcant in the Late-Glacial vegetation (Huntley and
Birks, 1983). In the area under study, juniper plays only
a minor role, for no Juniperus peak (or only an
ambiguous one) occurs during the Oldest Dryas–B.olling
transition. Instead, a prominent Salix peak characterises
this transitional phase. Local edaphic conditions (prevailing
waterlogged soils) may have played a role,
decreasing the importance of juniper and favouring
moisture-demanding pioneer willow shrubs. The species
represented by the pollen are not known due to the lack
of Salix macrofossils. Gaillard (1985) suggested that
Salix caprea may have grown as a pioneer species before
expansion of birch woodland. The Betula nana peak
accompanying that of Salix is most likely also the result
of climatic improvement, reﬂecting shrub-heath development
prior to the closing of forest canopy. This
resembles the Betula nana phase during the very
beginning of the Late-Glacial Insterstadial in the Swiss
Alps (Lotter et al., 1992; Ammann et al., 1994). The
response of locally present species to climatic amelioration
precedes the immigration of species and shows no
lag-phase.
In the second half of the S3a subzone, tree birch
became dominant in the pollen spectra in favour of
NAP and Pinus. Betula-dominated forest developed,
outcompeting most of the heliophyllous plant communities.
Betula and Pinus pollen percentages are almost
the same, indicating the diminished role of pine in fully
developed Interstadial forests. Shortly after, a marked
transition is recorded in the pollen diagram: Betula
percentages suddenly decrease again in favour of Pinus,
pointing to the change of proportion between these two
in the regional forest cover. Heliophyllous herbs
generally decrease to their Late-Glacial minima during
this Pinus phase, indicating that this vegetation change
represents a progressive closing of the forest canopy.
Pinus expansion may indicate increasingly severe conditions,
particularly in winter, i.e. an increase in
continentality (Walker, 1995). This event can be
correlated with the B.olling–Aller.od transition, which
appears to be abrupt in the pollen record. No distinct
transitional phase attributable to Older Dryas is present.
If this oscillation occurred in the area under study, it
had only a small impact on the vegetation. A short-lived
climatic deterioration for the Older Dryas is generally
recognised in central and western Europe (e.g. Walker,
1995), although in some sites its recognition is out of
temporal resolution of the analyses or out of the climatic
threshold of plant communities involved. It is well
known, that the response of a plant population to
climatic change is likely to be greatest near the margin of
its tolerance (Watts, 1979). For example, in Switzerland
the Older Dryas (often correlated with the ‘‘Aegelsee
oscillation’’ in this region) is not detectable at low-lying
sites but is more apparent at higher altitudes (above ca
600 m a.s.l.), where the climatic limits of indicator
species have been crossed (Lotter et al., 1992; Ammann
P. Pokorn!y / Quaternary International 91 (2002) 101–122 117
et al., 1993). During the Late-Glacial Interstadial,
marked climatic gradients developed over Europe, with
temperature differences as much as 6–71C within a few
hundred kilometres (Lowe et al., 1994). Under these
conditions, the response of vegetation to climatic
changes must have been very diverse in relation to
geographical position of the site.
The Late-Glacial Interstadial appears to be a period
with signiﬁcantly increased organic production, as
reﬂected in the sharp transition from minerogenic to
organic sedimentation (with high organic carbon and
nitrogen content in the sediment). The declining values
of Mg and K in the Interstadial sediments are the result
of the formation of clay minerals in the soil horizons, as
soils developed progressively in the lake catchment.
During episodes of relatively stable soils, deep weathering
of mature soil proﬁles should diminish the base
content of mineral material prior to its erosive removal
and sedimentation in lake basins (Engstr.om and Wright,
1984). The same process of soil development is recorded
also in lowland loess plateaus of the Czech Republic.
Loess formation, which is characteristic of late Pleniglacial,
generally terminates during the B.olling phase,
and initial pedogenesis takes place during that time
(Lo$zek and C!ılek, 1995). The decalciﬁcation of soil
horizons, together with expanding forest, were responsible
for ultimate decrease in Helianthemum percentages.
5.3. Younger Dryas
The Younger Dryas as a biozone is widely recognised
over most of Europe. Concerning the duration and
amplitude, this climatic oscillation is the most important
during the whole Late-Glacial period (Lotter et al.,
1992). YD climatic deterioration, dated roughly between
11 and 10 ka BP, is correlated with a readvance of polar
waters into the North Atlantic. Although the consequences
of this event are registered more strongly at the
sites near the ocean fringes of northern Europe, it is
apparent today that regional changes in climatic regime
may have been just as great in southern and eastern
regions of Europe as in the northern part of this
continent (Lowe and Watson, 1993; de Beaulieu et al.,
1994; Khotinsky and Klimanov, 1997). Although
problems of absolute dating accompany the recognition
of Younger Dryas event (the ‘‘14
C plateau’’; Ammann
and Lotter, 1989), it has been described from many sites
in the world and today is believed to be a global event
(Peteet, 1995). For the territory of the Czech Republic,
almost no reliable between-site comparison has been
possible for the Younger Dryas. Only at Vracov,
southern Moravia (Rybn!ı$ckov!a and Rybn!ı$cek, 1972),
is it marked by a small Juniperus and Salix rise after
10,765 BP, but its identiﬁcation is difﬁcult.
At the present study site, clear evidence of climatic
deterioration is ascribed to the Younger Dryas chronozone.
The values of Betula decrease, whereas those of
Alnus viridis, Salix, Betula nana, Chenopodiaceae, and
Artemisia increase. Proxy evidence suggests that climatic
deterioration was an increase in continentality rather
than a decrease in summer temperatures (see also
Ammann, 1989), for reconstructed minimum July
temperatures are at least 121C. The same values are
reconstructed for western Poland (Walker, 1995). The
sedimentation character changes back to more minerogenic,
and erosion intensity rises. The increase in erosion
indicators is only indistinct, suggesting that soil development
was not interrupted completely during this time.
The Younger Dryas can be subdivided in the
$Svarcenberk standard proﬁle into two phases, indicating
climatic amelioration (increase in humidity?) some time
before the onset of Holocene warming. The subdivision
of Younger Dryas into an older phase with colder and
more arid climate and the younger phase with warmer
and wetter climate has also been reported from some
other sites in Europe (from Norway and Poland; Goslar
et al., 1993; Birks et al., 1994), while in the Alps and in
most of Western Europe the younger phase has been
suggested to be somewhat drier (Walker, 1995).
The formation of extensive aeolian deposits in the
region under study is dated to the beginning of Younger
Dryas chronozone: Stratigraphic investigation of one of
the most prominent sand dunes (‘‘Vlkovsk!y p$resyp’’,
situated near $Svarcenberk lake basin) has revealed a
fossil soil buried under aeolian sands with a distinctive
layer of pine charcoal fragments on its surface. This
situation resembles conditions in The Netherlands,
where the ‘‘Usselo soil layer’’ formed during the
Aller.od period of lower aeolian activity. ‘‘Usselo-layer’’
has been dated from surface charcoal fragments to
between 11,400 and 10,300 BP, with average date
around 11,000 BP (Hoek, 1997). These results resemble
those from ‘‘Vlkovsk!y p$resyp’’, where a radiocarbon
date 11,2607120 BP has been obtained from very
similar stratigraphical situation. The formation of soils
during the Aller.od period required stable climatic
conditions with less aeolian activity and relatively dense
vegetation cover. On the other hand, the formation of
aeolian sand dunes requires severe climatic conditions
and sparse vegetation cover. Acceleration of aeolian
activity during the Younger Dryas accords well with the
results of pollen analysis, which point to a certain
opening of the forest relatively to the preceding Aller.od
period. The morphology of sand dunes points to
prevailing easterly winds in time of their formation.
5.4. Early-Holocene
There is abundant evidence throughout Europe for a
rapid rise in temperature at around 10,000 BP, although
precise dating of this event is difﬁcult because of another
‘‘radiocarbon plateau’’ at about that time. Over many
P. Pokorn!y / Quaternary International 91 (2002) 101–122118
areas of central and northwestern Europe, Younger
Dryas open communities were replaced within less than
500 yr by Betula/Pinus/Corylus woodland (Walker,
1995). The preservation of Pinus-dominated forest
during the entire Early-Holocene and the relatively late
development of deciduous forest was connected with the
persistence of a continental climate during that time and
generally low nutrient status together with the sandy
character of soils. Pine-dominated forests persisted in
the area until the Boreal increase in humidity, although
deciduous forests started to develop in favourable
locations somewhat earlier. The rapid temperature rise
during the Preboreal is indicated in the lake environment
by the early occurrence of Najas marina, Najas
minor, and Trapa natans macrofossils. Najas marina
suggests a mean July temperature not below 151C
(Lotter, 1988), and Trapa natans even more. This proxy
evidence suggests that the present-day values were
reached as early as about 9800 BP.
In a number of proxy records from mainland Europe,
there are indications of a cold climatic oscillation during
the ﬁrst millennium of the Holocene: the ‘‘Preboreal
oscillation’’ of the Swiss Plateau (Lotter et al., 1992) or
the ‘‘Youngest Dryas’’ of northern Germany (Behre,
1978). In western Norway, a readvance of the Josteldalsbreen
ice cap has been dated to ca. 9100 BP (Nesje
et al., 1991), while an abrupt fall in snow accumulation
(associated with a fall in North Atlantic sea surface
temperatures) is recorded in the GISP2 Greenland ice
core some 400 yr after the end of the Younger Dryas
(Alley et al., 1993). Equivalents of the European
Preboreal oscillation are supposed to be found also in
North America (Lowe et al., 1994). Possible palynological
evidence for Preboreal climatic oscillation has been
found also at $Svarcenberk, but it is relatively weak: a
short Pinus peak accompanied by the fall in Corylus,
Ulmus, and Quercus percentages and dated slightly
before 96407115 BP can be the result of some short
cooling episode.
The Early-Holocene sediment record from
$Svarcenberk lake comprises a prominent Fe peak dated
to around 8600 BP. It may be best explained as the
reﬂection of intensive leaching caused by sudden
climatic humiﬁcation (Engstr.om and Wright, 1984;
Starkel, 1991). In the pollen diagram, the same period
is characterised by Picea abies expansion. There is
probably some connection between these two phenomena,
as spruce grows preferably on waterlogged soils and
is able to produce highly acidic, raw humus, promoting
intensive leaching. The gradual development of nutrientpoor
acid soils was an important factor in the Holocene
vegetation development, as emphasised by Iversen
(1964). The building-up of raw humus on the soil
surface and resulting reducing conditions may have
released Fe from the soil, and it travelled to the lake in
solution or bound in organic complexes. A similar peak
in Fe has been described from lowland areas of the
Czech Republic, where Early-Holocene debris is cemented
by limonite and goethite (Lo$zek and C!ılek, 1995).
Also in Poland, the beginning of the Holocene is
characterised by inwash of dissolved iron into the lakes,
and this is interpreted as the ﬁrst stage of intensive soil
leaching (Pawlikowski et al., 1982). In southern Sweden,
the Fe content of several Early-Holocene lake sediments
is very high. Digerfeld (1972, 1975) attributed this peak
to early leaching from Late-Glacial soils in the catchment
and subsequent transport by groundwater to the
lake.
6. Conclusions
The results of this study permit the following
conclusions:
1. The Late-Glacial biostratigraphy of the investigated
site can be subdivided into different local pollen
assemblage zones (PAZ). These biostratigraphic units
can be correlated with general chronostratigraphic
and climatostratigraphic subdivision of the last
glacial–interglacial transition as established in
Europe.
2. Three distinct cold events interrupted the LateGlacial
climatic amelioration: The ﬁrst occurred
before 13,000 and is correlated with the Oldest
Dryas. Another regressive phase left only a weak
signal and subdivides the Late-Glacial Interstadial
into distinct Betula and Pinus phases. It is correlated
with the Older Dryas. The third regression, corresponding
to the Younger Dryas chronozone, is the
most prominent one. It resulted in the reduction of
the regional forest cover and a new expansion of open
herbaceous communities. Aeolian activity accelerated
during this time. The continental character of the
climate continued for about 500 yr after the onset of
Holocene warming. It is likely that the observed LateGlacial
and Early-Holocene climatic oscillations can
be attributed to the same processes that acted in the
western parts of Europe, i.e. the large-scale shifts in
the position of the North Atlantic Polar Front
(Ruddiman and McIntyre’s model).
3. The response of regional vegetation to the LateGlacial
climatic changes was different from much of
NW and western part of Central Europe. The main
difference is the early development of primeval pine
forest during climatic amelioration antedating the
onset of the Late-Glacial Interstadial. Pine dominance
is then characteristic for the entire Late-Glacial
and Early-Holocene in the investigated area, while
juniper is only of minor importance. The role of
Juniperus as a pioneer shrub was therefore adopted
by willow species.
P. Pokorn!y / Quaternary International 91 (2002) 101–122 119
4. There is strong correspondence between vegetational
development, soil development, and the intensity of
geomorphic processes acting within lake and its
surroundings. The synchroneity of these processes
has been primarily provoked by external climatic
forcing.
Acknowledgements
I would like to take the opportunity to dedicate this
study to my teacher Vlasta Jankovsk!a. Her advice
concerning the site selection and her enthusiasm were
the ﬁrst impulses for this work. This paper would also
fail much without extraordinary support from Herbert
E. Wright Jr., Brigitta Ammann, Pim van der Knaap
and Jacqueline van Leeuwen during my work. Their
helpful language and factual suggestions also further
improved the text. My sincere thanks belong also to
Andrea Kolmanov!a, L!ad’a Rektoris, Ji$r!ı $Setl!ık, and
other colleagues and friends who assisted in the
ﬁeldwork. I am especially grateful to Jan Pokorn!y and
Jan Kv$et for their practical as well as moral support. I
would also like to thank to Leo$s Klime$s for the
preparation of DCA and CCA diagrams. Grant Agency
of the Czech Republic supported this study through
grant project No. 206/98/0727.
References
Aaby, B., Berglund, B.E., 1986. Characterisation of peat and lake
deposits. In: Berglund, B.E. (Ed.), Handbook of Holocene Palaeoecology
and Palaeohydrology. Willey, Chichester, pp. 231–246.
Alley, R.B., Meese, D.A., Shuman, C.A., Grow, A.J., Taylor, K.C.,
Grootes, P.M., White, J.W.C., Ram, M., Waddington, E.D.,
Mayewski, P.A., Zielinski, G.A., 1993. Abrupt increase in Greenland
snow accumulation at the end of the Younger Dryas event.
Nature 362, 527–529.
Alm, T., Birks, H.H., 1991. Late Weichselian ﬂora and vegetation of
And#oya, northern Norway. Boreas 22, 171–188.
Ammann, B., 1989. Response times in bio- and isotope-stratigraphies
to Late-Glacial climatic shiftsFan example from lake deposits.
Ecologae Geologia Helvetica 82/1, 183–190.
Ammann, B., Lotter, A.F., 1989. Late-Glacial radiocarbon and
palynostratigraphy of the Swiss Plateau. Boreas 18, 109–126.
Ammann, B., Birks, H.J.B., Drescher-Schneider, R., Juggins, S., Lang,
G., Lotter, A.F., 1993. Patterns of variation in Late-Glacial pollen
stratigraphy along a northwest-southeast transect through SwitzerlandFa
numerical analysis. Quaternary Science Reviews 12,
277–286.
Ammann, B., Lotter, A.F., Eicher, U., Gaillard, M.J., Wohlfarth, B.,
Haeberli, W., Lister, G., Maisch, M., Niessen, F., Schlchter, Ch.,
1994. The W.urmian Late-Glacial in lowland Switzerland. Journal
of Quaternary Science 9 (2), 119–125.
de Beaulieu, J.L., Reille, M., 1984. A long Upper Pleistocene pollen
record from Les Echets, near lyon, France. Boreas 13, 111–132.
de Beaulieu, J.L., Andrieu, V., Ponel, P., Reille, M., Lowe, J.J., 1994.
The Weichselian Late-Glacial in soutwestern Europe (Iberian
peninsula, Pyrenees, massif Central, northern Apennines). Journal
of Quaternary Science 9, 101–107.
Behre, K.E., 1978. Die Klimaschwankungen im europ.aischen
Pr.aboreal. Petermanns Geographisches Mitteilungen 2, 97–102.
Birks, H.J.B., 1968. The identiﬁcation of Betula nana pollen. New
Phytologist 6, 309–314.
Birks, H.H., 1993. The importance of plant macrofossils in LateGlacial
climatic reconstructions: an example from western norway.
Quaternary Science Reviews 12, 719–726.
Birks, H.H., Paus, A., Svendsen, J.L., Alm, T., Mangerud, J., Landvik,
J.Y., 1994. Late Weichselian environmental change in Norway,
including Svalbard. Journal of Quaternary Science 9, 133–146.
Bos, H., 1998. Aspects of the Late glacialFearly Holocene vegetation
development in western Europe (Palynological and palaeobotanical
investigations in brabant, The Netherlands and Hessen, Germany).
LPP Contributions Series No.10, LPP Foundation, Utrecht,
240pp.
ter Braak, C.J.F., 1990. CANOCOFa FORTRAN program for
CANOnical Community Ordination by (partial) (detrnded) canonocal
correspondence analysis, principal component analysis and
redundancy analysis. Version 3.10. Microcomputer Power, Ithaca.
ter Braak, C.J.F., $Smilauer P., 1998. CANOCO reference manual and
user’s guide to ‘‘Canoco for Windows’’ software for canonical
community ordination, version 4. Microcomputer Power, Ithaca.
Ch!abera, S., 1982. Geologick!e zaj!ımavosti ji$zn!ıch ech (Geological
attractions of Southern Bohemia. In Czech.). Jiho"eesk!a Vlastiv"ıda
1, "Eesk!e Bud"ıjovice, 158pp.
Davis, M.B., 1965. A method for determination of absolute pollen
frequency. In: Kummel, B., Raup, D. (Eds.), Handbook of
Palaeontological Techniques. W.H. Freeman, San Francisco, pp.
674–686.
Dearing, J.A., 1991. Lake sediment record of erosional processes.
Hydrobiologia 214, 99–106.
Diggerfeld, G., 1972. The post-glacial development of Lake Trummen.
Folia Limnologica Scandinavica 16, 1–104.
Diggerfeld, H., 1975. The post-glacial development of Ranvinken bay
in Lake Immeln. III. Palaeolimnology. Geologiska F.oreningens i
Stockholm F.orhandingar 97, 13–28.
Engstr.om, D.R., Wright Jr., H.E., 1984. Chemical stratigraphy of lake
sediments as record of environmental change. In: Haworth, E.Y.,
Lund, J.G.W. (Eds.), Lake Sediments and Environmental History.
Leicester University Press, U.K, pp. 11–67.
Faegri, K., Iversen, J., 1989. Textbook of Pollen Analysis. Wiley,
Chichester.
Firbas, F., 1949. Sp+at-und nacheizeitliche Waldgeschichte von
Mitteleuropa n+ordlich der Alpen. S-G. Fisher, Jena.
Gaillard, M.J., 1984. Etude palynologique de l’!evolution tardi- et
postglaciaire de la v!eg!etation du Moyen-Pays Romand (Suisse).
Dissertationes Botanicae 77, 1–322.
Gaillard, M.J., 1985. Late.glacial and Holocene environments of some
antient lakes in western Swiss Plateau. Dissertationes Botanicae 87,
236–273.
Gams, H., 1926. Hydrocaryaceae. In: Heggi, M. (Ed.), Flora von
Mitteleuropa. Band V/2. Mnchen, pp. 882–2894.
de Gans, W., 1988. Pingo scars and their identiﬁcation. In: Clark, M.J.
(Ed.), Advances in Periglacial Geomorphology. Willey, Chichester,
pp. 299–324.
van Geel, B., Bohncke, S.J.P., Dee, H., 1981. A palaeoecological study
of an upper Lateglacial and Holocene sequence from ‘‘De
Borchert’’, The Netherlands. Review of Palaeobotany Palynology
31, 367–448.
van Geel, B., Hallewas, D.P., Pals, J.P., 1983. A Late Holocene deposit
under the Westfriese dijk near Enkhuizen (Prov. of NoordHolland,
The Netherlands) palaeoecological and archaeological
aspects. Review of Palaeobotany and Palynology 38, 269–335.
van Geel, B., Coope, G.R., van der Hammen, Th., 1989. Palaeoecology
and stratigraphy of the Lateglacial type section at Usselo (The
Netherlands). Review of Palaeobotany and Palynology 39, 25–129.
P. Pokorn!y / Quaternary International 91 (2002) 101–122120
Goslar, T., Kuc, T., Ralska-Jasiewiczowa, M., R!oz!anski, K., Arnold,
M., Bard, E., Pazdur, M.F., Szeroczy*nska, K., Wicik, B.,
Wieckowski, K., Walanus, A., 1993. High-resolution lacustrine
record of the Late Glacial/Holocene transition in central Europe.
Quaternary Science Reviews 12, 287–294.
Hammarlund, D., Buchardt, B., 1996. Composite stable isotope
records from a Late Weichselian lacustrine sequence at Graenge,
Lolland, Denmark: evidence of Aller.odand Younger Dryas
environments. Boreas 25, 8–22.
Hoek, W., 1997. Palaeogeography of Lateglacial vegetations. Aspects
of Lateglacial and Early Holocene vegetation, abiotic landscape,
and climate in The Netherlands. Netherlandse Geograﬁsche
Studies 230, 1–147.
Hol$c!ık, J., 1977. Changes in ﬁsh community of Klicava reservoir with
particular reference to Eurasian perch (Perca ﬂuviatiois). 1857–
1972. Journal of the Fisheries Research Board of Canada 34,
1734–1747.
Huizer, A.S., Izarin, R.F.B., 1997. The reconstruction of past climates
using multi-proxy evidence: an example of the Wichselian
Pleniglacial in northwest and central Europe. Quaternary Science
Reviews 16, 513–533.
Huntley, B., Birks, H.J.B., 1983. An Atlas of Past and Present Pollen
Maps for Europe: 0-13,000 yr ago. Cambridge University press,
Cambridge, 667pp.
Hus!ak, $S., Hejn!y, S., 1978. General characteristics of the T$rebo$nBasin
and Lednice region. In: Dykyjov!a, D. et Kv$et, J. (Eds.), Pond
Litoral Ecosystems, Ecological Studies, Vol. 28, Springer-Verlag,
Berlin, pp. 13–95.
Iversen, J., 1954. The Late-Glacial ﬂora of Denmark and its relation
to climate and soil. Danmarks Geologiske Undersoegelse II/80,
87–119.
Iversen, J., 1964. Retrogressive vegetational succession in the postglacial.
Journal of Ecology 52, 59–70.
Jalut, G., Monserrat Marti, J., Fortugne, M., Delibrias, G., Vilaplana,
J.M., Julia, R., 1992. Glacial to interglacial vegetation changes
in the northern and southern Ryr!en!ees; deglaciation, vegetation
cover and chronology. Quaternary Science Reviews 11,
449–480.
Jankovsk!a, V., 1980. Palogeobotanische Rekonstruktion der Vegetationsentwicklung
im Becken T$rebo$nsk!a p!anev whrend des Sptglazials
und Holozns. Vegetace "ESSR A11, Academia, Praha, 151pp.
Jankovsk!a, V., 1983. Palynologische Forschung am ehemaligen
Komoany-See (Sptglazial bis Subatlanticum). V"ıstn!ık !Ustedn!ıho
!ustavu geologick!eho 58 (2), 99–107.
Jankovsk!a, V., 1984. Late Glacial ﬁnds of Pinus cembra L. in the
Lubovnianska kotlina Basin. Folia Geobotanica et Phytotaxonomica
19, 323–326.
Jankovsk!a, V., Kom!arek, J., 1982. Das Vorkommen einiger Chlorokokalalgen
in B.omischen Sptglazial und Postglazial. Folia
Geobotanica et Phytotaxonomica 17, 165–195.
Jorga, W., Pietsch, W., Weise, G., 1982. Beitr.age zur .Okologie und
Bioindication von Trapa natans L. Limnologica 14 (2), 385–394.
Jowsey, P.C., 1966. An improved peat sampler. New Phytologist 65,
245–248.
Kac, J.N., Kac, C.V., Kipiani, M.G., 1965. An Atlas and a Key to
Fruits and Seeds Found in Quaternary Deposits of the Soviet
Union. Nauka, Moscow, 366pp (in Russian).
Karasev, G.L., 1987. Fish of Zabaikal’e Nauka, Novosibirsk, 321pp
(in Russian).
Khotinsky, N.A., Klimanov, V.A., 1997. Aller#od, Younger Dryas and
Early Holocene palaeo-environmental stratigraphy. Quaternary
International 41/42, 67–70.
Kle$cka, A., 1926. On the age of Bohemian peat bogs (in Czech). V"ıda
p!ırodn!ı 7, 305–313.
Kle$cka, A., 1928. Agrobotanical study on Rokytsk!e peat bog. Sborn!ık
"Eeskoslovensk!e Akademie Zem"ıd"ılsk!e 3, 195–267.
Kolstrup, E., 1980. Climate and Stratigraphy in Northwestern Europe
between 30,000 and 13,000 B.P., with special reference to The
Netherlands. Mededelingen van de Rijks Geologische Dienst 32/15,
181–253.
Lang, G., 1994. Quart.are Vegetationsgeschichte Europas. G. Fisher,
Jena, 462pp.
Liu, H., Cui, H., Pott, R., Speier, M., 1999. The surface pollen of the
woodland-steppe ecotone in southeastern Inner Mongolia, China.
Review of Palaeobotany Palynology 105, 237–250.
Lotter, A.F., 1988. Pal.aeo.okologische und pal.aolimnologische
studie des rotsees bei luzern. Pollen-, grossrest-, diatomen- und
sedimentanalytische Untersuchungen. Dissertationes Botanicae
124, 1–187.
Lotter, A.F., Juggins, S., 1991. POLPROF, TRAN and ZONE
Programs for plotting, editing and zoning pollen and diatom data.
INQUA-Commission for the Study of the Holocene, Working
group. Data-Handling Methods newsletter 6, 651–660.
Lotter, A.F., Eicher, U., Siegenthaler, U., Birks, H.J.B., 1992. LateGlacial
climatic oscillations as recorded in Swiss lake sediments.
Journal of Quaternary Science 7, 187–204.
Lowe, J.J., Watson, C., 1993. Lateglacial and early Holocene pollen
stratigraphy of the northern Apennines, Italy. Quaternary Science
Reviews 12, 727–738.
Lowe, J.J., Ammann, B., Birks, H.H., Bj.orck, S., Coope, G.R.,
Cwynar, L., de Beaulieu, J.-L., Mott, J.R., Peteet, D.M., Walker,
M.J.C., 1994. Climatic changes in areas adjacent to the North
Atlantic during the last glacial-interglacial transition (14–9 ka BP):
a contribution to IGCP-253. Journal of Quaternary Science 9,
185–198.
Lo$zek, V., C!ılek, V., 1995. Late Weichselian-Holocene sediments and
soils in mid-European calcareous areas. Anthropozoikum 22,
87–112.
Lundqvist, J., Saarnisto, M., 1995. Summary of project IGCP-253.
Quaternary International 28, 9–18.
Mackay, J.R., 1988. Pingo collapse and palaeoclimatic reconstruction.
Canadian Journal of Earth Sciences 25, 495–511.
Mangerud, J., Andersen, S.T., Berglund, B.E., Donner, J.J., 1974.
Quaternary stratigraphy of Norden, a proposal of terminology and
classiﬁcation. Boreas 3, 109–127.
Menke, B., 1968. Das Sp.atzglazial von Glusing. Eiszeitalter und
Gegenwart 19, 73–84.
Moore, P.D., Webb, J.A., Collinson, M.E., 1991. Pollen Analysis.
Blackwell, Oxford, 216pp.
Moore, P.D., 1979. The reconstruction of the Lateglacial environment:
some problems associated with ith interpretation of pollen data. In:
Lowe, J.J., Gray, J.M., Robinson, J.E. (Eds.), Studies in the
Lateglacial of North-West Europe. Pergamon Press, Oxford,
pp. 151–155.
Nesje, A., Kvamme, M., Rye, N., L.ovlie, R., 1991. Holocene glacial
and climatic history of the Josteldalsbreen region, western Norway:
evidence from lake sediments and terrestrial deposits. Quaternary
Science Reviews 10, 87–114.
Neuh.auslov!a, Z. (Ed.), 1998. Map of Potential Natural Vegetation of
the Czech Republic. Academia, Praha.
Pawlikowski, M., Ralska-Jasiewiczowa, M., Sch.onborn, W., Stupnicka,
E., Szeroczy*nska, K., 1982. Woryty near Gietrzwald,
Olsztyn Lake District, northeast PolandFvegetational history
and lake development during the last 1200 yr. Acta Plalaeobotanica
22, 1–120.
Peteet, D., 1995. Global Younger Dryas? Quaternary International 28,
93–104.
Pissart, A., 1988. Pingos: an overview of the present state of
knowledge. In: Clark, M.J. (Ed.), Advances in Periglacial
Geomorphology. Willey, Chichester, pp. 279–298.
Poser, S.C., 1948. Boden- und Klimaverh.altnisse in Mittel- und
Westeuropa w.ahrend der W.urmeiszeit. Erdkunde 2, 53–68.
P. Pokorn!y / Quaternary International 91 (2002) 101–122 121
Punt, W. (Ed.), 1976–1996. The Northwest European Pollen Flora 1–7.
Elsevier, Amsterdam.
Ralska-Jasiewiczova, M., 1980. Late-Glacial and Holocene vegetation
of the Bieszczady mts. (Polish eastern Carpathians). Pa*nstwowe
wydawnictwo naukowe, Warszawa. 202pp.
Ruddiman, W.F., McIntyre, A., 1981. The North Atlantic ocean
during the last deglaciation. Palaeogeography, Palaeoclimatology,
Palaeoecology 35, 145–214.
Rudolph, K., 1917. Untersuchungen .uber den Afbau b.omischer
Moore I. Aufbau und Entwicklungsgeschichte s.udb.ohmischer
Hochmoore. Abhandlungen der K.onige Zoologische und Botanische
Gesellschft, Wien 9/4, 1–116.
Rudolph, K., Firbas, F., 1922. Pollenanalytische Untersuchungen
b.ohmischer Moore. Berichte der Deutschen Botanischen Gesellschaft,
Jena 47, 393–405.
Rybn!ı$cek, K., Rybn!ı$ckov!a, E., 1994. Vegetation histories of the
Panonian, Hercynic, and Carpathian regions of the former
Czechoslovakia. In: Lotter, A.F., Ammann, B. (Eds.), Festschrift
Gerhard Lang. Dissertationes Botanicae, Vol. 234, pp. 473–485.
Rybn!ı$ckov!a, E., Rybn!ı$cek, K., 1972. Erste Ergebnisse pal.aogeobotanischer
Untersuchungen des Moores bei Vracov, S.udm.ahren.
Folia Geobotanica et Phytotaxonomica 7, 285–308.
Starkel, L., 1990. Environmental changes at the Younger DryasPreboreal
transition and during the early Holocene: some
distinctive aspects in central Europe. The Holocene 1, 234–242.
Stuiver, M., Grootes, P.M., Braziunas, T., 1995. The GISP d18
O
climate record of the past 16,500 yr and the role of the sun, ocean
and volcanoes. Quaternary Research 44, 341–354.
$St$ep!anov!a, M., 1930. Une !etude de d!ev!elop!ement des tourbes dans la
Boh!eme de Sud. Spisy P!ırodov"ıdec!e Fakulty KU 109, 1–36.
T.ornqvist, T.E., de Jong, A.F.M., Oosterbaan, W.A., van der Borg,
K., 1992. Accurate dating of organic deposits by AMS 14
C
measurement of macrofossils. Radiocarbon 34, 566–577.
Troels-Smith, J., 1955. Characterization of unconsolidated sediments.
Danmarks Geologiske Undersoegelse 3/10, 73pp.
Tyr!a$cek, J., 1995. Depositonal chnges in non-glaciated regions of
central Europe. Quaternary International 28, 77–81.
Usinger, H., 1975. Pollenanalytische und stratigraphische Untersuchungen
an zwei Sp.atglazial-Vorkommen in Schleswig-Holstein.
Mitteilungen der Arbeitsgemeinschaft Geobotanik in SchleswigHolstein
und Hamburg 25, 1–173.
Vorren, T.O., Vorren, K.D., Alm, T., Gulliksen, S., L.ovlie, R., 1988.
The last deglaciation (20,000–11,000B.P.) on And.oya, northern
Norway. Boreas 17, 41–77.
Walker, M.J.C., 1995. Climatic changes in Europe during the
last glacial/interglacial transition. Quaternary International 28,
63–76.
Walker, D., Atkin, G., Owen, J.A.K., 1994. Uncertainty in the
determination of pollen concentrations. In: Lotter, A.F., Ammann,
B. (Eds.), Festschrift Gerhard Lang, Dissertationes Botanicae Vol.
234, pp. 537–554.
Washburn, A.L., 1980. Geocryology. A Survey of Periglacial Processes
and Environments. Wiley, New York, 406pp.
Watts, W.A., 1979. Regional variations in the response of vegetation
to Lateglacial Climatic Events in Europe. In: Lowe, J.J., Gray,
J.M., Robinson, J.E. (Eds.), Studies in the Lateglacial of NorthWest
Europe. Pergamon Press, Oxford, pp. 1–21.
Wohlfarth, B., 1996. The chronology of the last termination: a review
of radiocarbon-dated, high resolution terrestrial stratigraphies.
Quaternary Science Reviews 15, 267–284.
P. Pokorn!y / Quaternary International 91 (2002) 101–122122