A first chironomid-based summer temperature reconstruction
(13e5 ka BP) around 49
N in inland Europe compared with local lake
development
Petra Hajkova a, b, *
, Petr Paril a
, Libor Petr a
, Barbora Chattova a
, Tomas Matys Grygar c
,
Oliver Heiri d
a
Department of Botany and Zoology, Faculty of Science, Masaryk University, Kotlarska 2, CZ-61137, Brno, Czech Republic
b
Department of Vegetation Ecology, Institute of Botany, Academy of Sciences of the Czech Republic, Lidicka 25/27, CZ-602 00, Brno, Czech Republic
c
Institute of Inorganic Chemistry, Academy of Science of the Czech Republic, CZ-25068, Rez, Czech Republic
d
Institute of Plant Sciences and Oeschger Centre for Climate Change Research, Altenbergrain 21, CH-3013, Bern, Switzerland
a r t i c l e i n f o
Article history:
Received 27 January 2016
Received in revised form
1 April 2016
Accepted 2 April 2016
Available online 23 April 2016
Keywords:
Carpathians
Climate
Diatoms
Geochemistry
Holocene
Lake-productivity
Late Glacial
Pollen
Transfer functions
Water level changes
a b s t r a c t :
Temperature reconstructions for the end of the Pleistocene and the first half of the Holocene based on
biotic proxies are rare for inland Europe around 49
N. We analysed a 7 m long sequence of lake deposits
in the Vihorlat Mts in eastern Slovakia (820 m a.s.l.). Chironomid head capsules were used to reconstruct
mean July temperature (TJuly), other proxies (diatoms, green algae, pollen, geochemistry) were used to
reconstruct local environmental changes that might have affected the climate reconstruction, such as
epilimnetic total phosphorus concentrations (TP), lake level changes and development of surrounding
vegetation. During the Younger Dryas (YD), temperature fluctuated between 7 and 11 
C, with distinct,
decadal to centennial scale variations, that agree with other palaeoclimate records in Europe such as d18
O
content in stalagmites or Greenland ice cores. The results indicate that the site was somewhat colder
than expected from the general south-to-north YD temperature gradient within Europe, possibly because
of north-facing exposition. The warmer phases of the YD were characterised by low water level or even
complete desiccation of the lake (12,200e12,400 cal yr BP). At the Late-Glacial/Holocene transition TJuly
steeply increased from from 11 to 15.5 
C (11,700e11,400 cal yr BP) e the highest TJuly for entire
sequence. This rapid climate change was reflected by all proxies as a compositional change and
increasing species diversity. The open woodlands of Pinus, Betula, Larix and Picea were replaced by broadleaved
temperate forests dominated by Betula, later by Ulmus and finally by Corylus (ca 9700 cal yr BP). At
the same time, input of eroded coarse-grained material into the lake decreased and organic matter (LOI)
and biogenic silica increased. The Early-Holocene climate was rather stable till 8700 cal yr BP, with
temporary decrease in TJuly around 11,200 cal yr BP. The lake was productive with a well-developed
littoral, as indicated by both diatoms and chironomids. A distinct decline of TJuly to 10 
C between
8700 and 8000 cal yr BP was associated with decreasing chironomid diversity and increasing climate
moistening indicated by pollen. Tychoplanktonic and phosphorus-demanding diatoms increased which
might be explained by hydrological and land-cover changes. Later, a gradual warming started after
7000 cal yr BP and representation of macrophytes, periphytic diatoms and littoral chironomids increased.
Our results suggest that the Holocene thermal maximum was taking place unusually early in the Holocene
at our study site, but its timing might be affected by topography and mesoclimate. We further
demonstrated that temperature changes had coincided with variations in local hydrology.
© 2016 Elsevier Ltd. All rights reserved.
1. Introduction
Recent climate changes have stimulated an intense research on
past climatic variations and their impact on both biotic and abiotic
ecosystem processes. Quaternary climate changes have been
reconstructed using isotope composition in long ice-core or marine
Abbreviations: LOI, loss-on-ignition; LG, Late Glacial; TJuly, mean July temperature;
TP, epilimnetic total phosphorus; YD, Younger Dryas.
* Corresponding author. Department of Botany and Zoology, Faculty of Science,
Masaryk University, Kotlarska 2, CZ-61137, Brno, Czech Republic.
E-mail address: buriana@sci.muni.cz (P. Hajkova).
Contents lists available at ScienceDirect
Quaternary Science Reviews
journal homepage: www.elsevier.com/locate/quascirev
http://dx.doi.org/10.1016/j.quascirev.2016.04.001
0277-3791/© 2016 Elsevier Ltd. All rights reserved.
Quaternary Science Reviews 141 (2016) 94e111
sequences (e.g. Blockley et al., 2012; Lowe et al., 2008), or, for
Europe, by climate model runs driven by changes in past climate
forcing factors such as variations in the North Atlantic thermohaline
circulation (e.g. Renssen et al., 2012). Climatic changes, recent
or pre-historic, are, however, never uniform across different regions
(e.g. Heiri et al., 2014a) and spatial variability in climate dynamics
may affect large-scale edaphic processes and species
distribution. Regional and local climate can substantially deviate
from the global models (Mayewski et al., 2004; Feurdean et al.,
2014) because of specific topography and landscape settings. Fossil
remains of different organisms like pollen, macrofossils, diatoms
or chironomids are often used as climate proxies in local and
regional reconstructions (e.g. Davis et al., 2003; Buczko et al., 2013;
Heiri et al., 2014a; V€aliranta et al., 2014). Generally, they have
shown that the Holocene (since ca 11,650 cal yr BP, Walker et al.,
2009) is a warm period with relatively stable climatic conditions
compared to Pleistocene. At the end of the Late Glacial (LG), summer
temperature in Europe increased, partially as a consequence of
orbitally-forced summer insolation, which in the northern Hemisphere
was the highest in the Early Holocene (Laskar et al., 2004),
partially due to changes in other climate forcing and amplifying
factors such as greenhouse gas concentrations, ocean current
changes and melting of large continental ice sheets (e.g. Clark et al.,
2001; Renssen and Isarin, 2001; Menviel et al., 2011). Nevertheless,
in Europe there was some variation in climate during the Holocene,
even if with lower amplitude than observed in the late Pleistocene.
A review of 50 globally-distributed palaeoclimatical records has
shown that Holocene climate variations have been larger and more
frequent than is commonly recognized (Mayewski et al., 2004).
Several periods of rapid climate change (RCC) were revealed, from
which two took place in the Early and Middle Holocene
(9000e8000 cal yr BP, 6000e5000 cal yr BP). Most of the climate
change events in these globally distributed records were characterised
by polar cooling, tropical aridity, and major atmospheric
circulation changes. Several abrupt short-term oscillations during
the Holocene were also recorded by both, oxygen isotopes in icesheet
cores (Blockley et al., 2012) and biotic proxies (e.g. Magny
et al., 2003; Rosen et al., 2001; Davis et al., 2003; Toth et al.,
2012, 2015). The so called 8.2 ka event was the most pronounced
temperature change within the Early and Middle Holocene, which
was reflected by a decrease in Corylus pollen in the fossil record of
North Europe (Sepp€a et al., 2005; Rasmussen et al., 2008) and less
frequently also in Central Europe (Tinner and Lotter, 2001; Dudova
et al., 2014). Contrary, chironomid-based reconstructions captured
this event rarely (Płociennik et al., 2011; but see Sepp€a et al., 2007
and Heiri et al., 2003). It is hence likely that this short-term North
Atlantic cooling triggered by Laurentide ice-sheet collapse
(Wiersma and Renssen, 2006) influenced regional and local summer
temperatures and some types of ecosystems only locally and
moreover, it appears that some biotic proxies do not consistently
reflect this short-term climate oscillation.
A widely used biotic proxy for temperature reconstruction are
fossil chironomids in lake sediment records. Chironomids have a
rather short life-cycle and relatively high dissemination ability and
therefore show a rapid response to changing environment (Brooks
et al., 2007). There are numerous stenotopic species within the
chironomids which can provide reliable reconstructions of the past
environment. Identification is usually possible at the level of genera
or species morphotypes, often with known ecological preferences.
In the last 15 years, several calibration data-sets were developed for
July air temperature (TJuly) reconstruction in Eurasia (e.g. Brooks
and Birks, 2001; Nazarova et al., 2011; Holmes et al., 2011; Heiri
et al., 2011, 2014a). In East-Central Europe, there is a gap in
knowledge on chironomid-inferred climate from the LG and Holocene
periods. Further, even if chironomids are very good
indicators of changes in July temperatures, some autogenic processes
not triggered by climate can influence chironomid species
turnover and thus distort the climate reconstruction. Typically,
there is a general positive correlation between temperature and
productivity, but lake productivity can increase independently of
temperature because of changing nutrient concentrations and it
may be difficult to separate these two influences (Velle et al., 2010).
Interpretation of quantitative reconstructions should be therefore
done with caution and other biotic or abiotic proxies can help to
separate potential independent effects of productivity, oxygen and
water level changes from climate influence (e.g. Heiri and Lotter,
2005). Geochemical analyses may serve as a proxy for catchment
erosion and diatoms and green algae as reliable proxies of trophic
conditions (Battarbee et al., 2001). Regional vegetation composition,
reconstructed by means of fossil pollen, can characterise lake
catchments in terms of potential intensity of erosion, hydrology or
biogeochemistry and in addition may indicate coarse-scale climatic
changes as well (Davis et al., 2003; Mauri et al., 2015). In this study
we covered all these proxies to provide the first chironomid-based
temperature reconstruction for inland Europe around 49N,
covering the end of the Pleistocene and the first half of the Holocene,
and to compare it with reconstructed local development of
the sedimentary environment.
The study site in the Vihorlat Mts (49N) is situated between a
more southerly located site with a chironomid inferred temperature
reconstruction in the Eastern Carpathians (Retezat Mts., 45N;
Toth et al., 2012, 2015) and a more northerly located site in the
Polish lowland (52N; Płociennik et al., 2011). According to a review
by Heiri et al. (2014a), there is a rather high number of sites where
July air temperatures are reconstructed based on chironomids in
the Alps, British islands and NW Europe, but data are almost
missing for the latitude 47e52N in East-Central Europe. Tatosova
et al. (2006) and Hosek et al. (2014) provided some data on fossil
chironomid assemblages, but without quantitative TJuly reconstruction.
Thus, this study fills a gap in our knowledge about past
climate in East-Central Europe. Moreover, the position of the study
site is transitional between oceanic and continental climate influences
and thus shifts in atmospheric circulation and pressure
changes, e.g. associated with variations in the predominance of
North Atlantic oscillation states, may substantially have influence
local climate. Combining different proxies, we aim to separate the
influence of past climate changes in the study region from independent
local processes like autogenic changes in productivity and
lake depth. The main aims of our study were: 1) to reconstruct
mean July temperatures (TJuly) based on chironomid assemblages;
2) to reconstruct local environmental conditions and processes like
lake productivity and water level changes using diatoms and green
algae to control for undesired local effects in climate reconstruction
and 3) to reconstruct changes in the lake surrounding using pollen
and geochemical methods to detect influence of changing vegetation
cover and extent of erosion.
2. Material and methods
2.1. Study site and sediment sampling
The study site named Hypkana is located in the westernmost
part of the Eastern Carpathians, in the Vihorlat Mts in eastern
Slovakia (East-Central Europe; 820 m a.s.l.; 4854.7870 N,
2209.8140 E; see Fig. 1). The geological bedrock is formed by
neogenic andezite. The recent climate of the region is characterised
by mean annual temperatures of 4e6 C (mean in January
e5ee6 C, mean in July 14e16 C) and mean annual precipitation
of 1000e1200 mm (http://geo.enviroportal.sk/atlassr). The daily
mean temperature (long time series of air temperature,1961e1990)
P. Hajkova et al. / Quaternary Science Reviews 141 (2016) 94e111 95
in the nearest meteorological station Kamenica nad Cirochou
(178 m a.s.l., northern foothills of the Vihorlat Mts) is 18.5 C, which
corrected to 820 m a.s.l. based on a July temperature lapse rate of
0.6 C/100 m of altitude would be equivalent to 14.7 C. However,
the real temperature of the study site is probably lower because of
the northern slope position of the Motrogon Mt (1018 m a. s. l.). The
study site has a small catchment area and the present-day mire of
ca 2.1 ha has been a Nature Reserve since 1980. The recent vegetation
is dominated by Eriophorum vaginatum, Oxycoccus palustris,
Vaccinium myrtillus and Molinia caerulea in the herbaceous layer
and Sphagnum recurvum agg., S. magellanicum and Polytrichum
commune in the bryophyte layer. The surrounding landscape is
overgrown by beech forests and the nearest village Zemplínske
Hamre is situated 3.5 km northwards at 400 m a.s.l. The whole
profile was 11.1 m deep, of which almost 8 m consisted of lake
sediments suitable for chironomid and diatom analyses. The profile
was obtained from the central part of the mire in the beginning of
May 2012 using combination of a single gouge auger (6 cm diameter,
100 cm length) for the upper slightly decomposed peat
sequence and a chamber corer (5 cm diameter, 50 cm length) for
limnic sediments analysed in this study. We have sampled two
parallel overlapping cores to avoid incomplete recovery.
2.2. Dating and age-depth modelling
Selected macrofossils of terrestrial plants (seeds of taxa specified
in Table 1, spindles of Eriophorum, bryophytes, Picea needles)
and ephippia of Cladocera were sent for AMS dating to the Centre
for Applied Isotope Studies, University of Georgia, Athens, USA. The
IntCal13 calibration curve was used for calibration of 14
C dates
(Reimer et al., 2013). We obtained altogether 14 radiocarbon dates,
from which 11 were used for the depth-age modelling of the entire
core including the upper peat layer (see Table 1). Two dates (UG-
15694, UG-15690) were excluded because they caused an age
reversal and decreased the quality of the model to zero. One date
(UG-15689) did not disagree with other ages, but excluding of this
date was important for obtaining a reliable Bayesian model with an
agreement value at least around 60% (the recommended level). The
upper two excluded 14
C dates were obtained from macrofossils of
mire vegetation (seeds of Carex rostrata and Menyanthes trifoliata)
found in the lake sediment. Likely these macrofossils were transported
from the upper layers (by coring or by bioturbation processes)
and therefore their 14
C date was younger than expected. An
age-depth model (Fig. 2) with 1 cm resolution based on a
P_Sequence function with the k parameter equal to 0.5 cmÀ1
and
log10(k/k0) equal to 0.3 was calculated using OxCal 4.2.4. (Bronk
Ramsey, 2009). To incorporate potential changes in the sedimentation
rate (e.g., contact of different types of deposits), the command
Boundary was applied. The boundaries were placed at 955 cm
(grey gyttja/brown gyttja) and at 323 cm (gyttja/peat). In the text
below we use mean values of modelled data in the range of 95.4%
and we rounded them to the nearest 50 year step. For the formal
subdivision of the Holocene we followed Walker et al. (2012) with
the Early-Middle Holocene boundary at 8200 cal yr BP and the
Middle-Late Holocene boundary at 4200 cal yr BP.
2.3. Biotic proxies
Samples for pollen analysis (0e800 cm: 1 cm3
, 800e1110 cm:
0.5 cm3
) were treated by acetolysis (Faegri and Iversen, 1989). A
minimum of 500 terrestrial pollen grains were counted and
determined using pollen keys (Beug, 2004; Reille, 1992). The algae
Fig. 1. Position of the study site in Europe and the Carpathian Mts and its geomorphological features and catchment size. The position of some of the sites with TJuly reconstructions
based on chironomids that are discussed in the text are indicated.
P. Hajkova et al. / Quaternary Science Reviews 141 (2016) 94e11196
of the genus Pediastrum and other chlorococcal algae were identified
according to Komarek and Jankovska (2001). The nomenclature
of all identified pollen types follows Beug (2004). Percentage pollen
diagrams were constructed using the total sum (TS) comprising
arboreal and non-arboreal pollen. Aquatic and local wetlands
plants (including Cyperaceae and Alnus), algae and other nonpollen
palynomorphs were excluded from the TS. Using Lycopodium
tablets as a marker we calculated pollen and microcharcoal
(fraction 0.01e0.1 mm) concentration and finally pollen and
microcharcoal influx.
Sediment samples for chironomid analysis (2e3.5 g of wet
weight; 0.8e2.8 g of dry weight) were deflocculated for 20 min in
10% KOH solution (60e75 C) and then passed through 250 and
100 mm sieves. The chironomid capsules were hand sorted under a
stereomicroscope (20e40Â magnification) and only specimens
consisting of more than half of the mentum were counted. The wet
sediments were dried to a constant weight and the number of
chironomid remains was calculated to 1 g of dry sediment and
identified using Brooks et al. (2007). In all sorted layers (excluding
depth 1030e1035 cm) the number of 50 head capsules was
reached, which is recommended as a minimum count for the
calculation of temperature reconstructions (Heiri and Lotter,
2001). Thus this single layer was excluded from TJuly reconstruction.
Reconstructed TJuly was re-calculated to TJuly in 0 m a.s.l. for
better comparison with other reconstructions in the literature
based on July temperature lapse rates of 0.6 C/100 m (see e.g.
Heiri et al., 2014a). Selected chironomid taxa were classified into
ecological categories according to demands on trophic status,
bathymetric distribution and preference of macrophytes using
relevant literature (Wiederholm, 1983; Brooks et al., 2007; see also
Appendix A).
Diatom samples were prepared following the method described
in van der Werf (1955). Small quantities of the samples were
cleaned by adding 37% H2O2 and heating to 80 C for about 1 h. The
reaction was completed by addition of KMnO4. Following digestion
and centrifugation, the resulting clean material was diluted with
distilled water to avoid excessive concentrations of diatom valves
that may hinder reliable observations. Known quantities of Lycopodium
spores were added to estimate diatom concentrations.
Cleaned diatom valves were mounted in Naphrax®
, a highrefractive
index medium. In each sample, 400 diatom valves were
identified and enumerated on random transects at 1,000Â magnification
using an Olympus BÂ50 microscope equipped with Differential
Interference Contrast (Nomarski) optics. Further, diatoms
were classified into five groups according their life form (see also
Buczko et al., 2013 and Appendix B): aerophytic (in subaerial and
terrestrial habitats), benthic (at the bottom and shore of the lake),
planktonic and tychoplanktonic (in the water column) and periphytic
(attached to surfaces).
2.4. Geochemical analyses, LOI, MS
The weight percentage of organic matter was determined by
means of loss-on-ignition (LOI) according to Heiri et al. (2001) and
Holliday (2004) in each sample. The samples were dried at 105 C
for 24 h, and the combustion at 550 C took 3 h. Magnetic susceptibility
(MS) was determined using a Kappabridge KLY-2 device
(Agico, Czech Republic). The results were normalized to get massspecific
magnetic susceptibility in m3
$kgÀ1
.10À9
. Magnetic susceptibility
provides information about input of eroded clastic sediments
(e.g. Shakesby et al., 2007). X-ray fluorescence analysis
(EDXRF) of geochemical properties of rocks and soils was carried
out using a PANalytical MiniPal4.0 spectrometer with a Peltiercooled
silicon drift energy-dispersive detector. The samples were
powdered by agate pestle and mortar and put into measuring cells
with a Mylar foil bottom without any further pre-treatment. The
analyses were not calibrated and signal counts per second (c.p.s.) of
individual elements were evaluated (Grygar et al., 2010). The XRF
analytical signal is proportional to element concentrations (Matys
Grygar et al., 2014), however, the calibration of this simple XRF
setup was not performed: it would depend on matrix effects
(element composition, grain size, mean organic matter content),
especially in the case of light elements (Al and Si). Matrix effects for
such non-destructive XRF analyses are best corrected by using ratios
of element signals, such as Zr/Rb or Al/Si.
Because of variable and mostly very high biogenic silica content,
the Al/Si ratio, otherwise a versatile proxy of sediment grain
size (Grygar et al., 2010; Bouchez et al., 2011) could not be used as
it reflects contributions of biogenic silica as well as the grain size
trends. The Zr/Rb ratio, another grain size proxy (Jones et al.,
2012), can be used to evaluate relative proportions of coarse silt
or the finest sand (the typical grain size of zircons, Bouchez et al.,
2011) relative to other clastic components, but with negligible
influence by autochthonous components. Because in clastic components
Zr and Si usually correlate due to their prevalence in
coarser size fractions (zircons and quartz), we used Si/Zr as proxy
for the relative ratio of biogenic silica to detritic clastics with little
lithogenic influence. The Rb/K ratio was used as a proxy for the
intensity of chemical weathering, because although both elements
are mobilized by chemical weathering, Rb is more strongly
retained in clay minerals (illite and smectite) and, hence, in surface
sediments it is enriched by chemical weathering (Hu and Gao,
2008).
Table 1
Results of 14
C dating (AMS method) from the sediment profile studied. The calibrated ages are median values and intervals of the calibrated 2s range BP. Dates assigned by
asterisk are excluded from the age-depth model.
Samples Depth (cm) Dating method 14
C age in uncal. BP Cal yr BP (interval) Cal yr BP (median) Material
UG-17161 54e56 AMS 630 ± 25 553e662 597 Bryophytes (Sphagnum leaves)
UG-17162 104e106 AMS 2450 ± 25 2361e2701 2526 Spindles (Eriophorum vaginatum) þ sphagna
UG-17163 174e176 AMS 2770 ± 25 2789e2943 2862 Spindles (Eriophorum vaginatum)
UG-19965 214e216 AMS 2870 ± 20 2925e3066 2988 Needles, spindles, bryophytes
UG-15688 265e270 AMS 3380 ± 25 3570e3692 3631 Seeds (Carex rostrata)
UG-15689* 320e325 AMS 3510 ± 25 3700e3855 3778 Seeds (C. rostrata, Menyanthes trifoliata)
UG-15690* 455e460 AMS 3650 ± 30 3888e4084 3986 Seeds (Carex rostrata)
UG-19966 540e545 AMS 6970 ± 50 7689e7930 7802 Ephippia (Cladocera)
UG-15691 700e705 AMS 7930 ± 30 8635e8978 8807 Seed (Acer cf. campestre)
UG-19968 800e805 AMS 8530 ± 35 9480e9545 9518 Ephippia (Cladocera)
UG-15692 835e840 AMS 8830 ± 30 9709e10,147 9928 Seed (Picea abies)
UG-15693 930e935 AMS 9980 ± 45 11,259e11,695 11,477 Ephippia (Cladocera)
UG-15694* 1035e1040 AMS 9780 ± 30 11,181e11,241 11,211 Needles (Pinus), seeds (Betula, Carex sp.)
UG-15695 1090e1095 AMS 11,020 ± 40 12,749e13,010 12,880 Needles (Pinus), ephippia (Cladocera)
P. Hajkova et al. / Quaternary Science Reviews 141 (2016) 94e111 97
2.5. Data analyses
Species stratigraphic diagrams of chironomids and diatoms as
well as a diagram of chemical sediment composition were created
using the C2 software (Juggins, 2007). Both, diatom and chironomid
counts were converted to percentage data. The pollen percentage
diagram was plotted using the Tilia v.1.7.16 (Grimm, 2011) software.
The zonation in pollen, diatom and chironomid diagrams is a result
of Coniss cluster analyses with square root transformation of data.
To analyse changes in the total species composition of pollen, diatoms
and chironomids we used detrended correspondence analyses
(DCA) in the Canoco software (ter Braak and Smilauer, 2002)
with down-weighting of rare species, logarithmic transformation of
species data and detrending by segments. The length of gradient
was 2.14 standard deviation units (SD) for diatoms, 2.08 SD for
chironomids and 2.22 by SD for pollen. The variation explained by
the first axis was 17.2% for diatoms,16.1% for chironomids and 28.6%
for pollen.
We used an inference model for chironomid-based temperature
reconstruction calculated from a Swiss-Norwegian chironomid
calibration dataset (Heiri et al., 2011) in order to infer TJuly. This
calibration dataset has been formed by amalgamating two local
calibration datasets from Switzerland (Heiri and Lotter, 2010) and
Norway (Brooks and Birks, 2001). Altogether, 60 chironomid taxa
from the fossil data were used for TJuly reconstruction. Colddemanding
Derotanypus sp. was abundant in some Late-Glacial
Hypkaňa [Amodel:59]
Bottom
UG-15695 [A:99]
brown/grey gyttja
UG-15693 [A:103]
UG-15692 [A:97]
UG-19968 [A:105]
UG-15691 [A:96]
UG-19966 [A:64]
peat/gyttja
UG-15688 [A:88]
UG-19965 [A:44]
UG-17163 [A:100]
UG-17162 [A:61]
UG-17161 [A:97]
Corring date [A:100]
UG-15694 [A:100]
UG-15690 [A:100]
UG-15689 [A:100]
050001000015000
Modelled date (cal. yrs BP)
0
200
400
600
800
1000
1200
Depth(cm)
0.076 cm/year
0.146 cm/year
0.139 cm/year
0.079 cm/year
0.071 cm/year
0.102 cm/year
Fig. 2. Age-depth model based on 11 radiocarbon dates. Three dates were excluded from the final model to obtain a reliable Bayesian model with the maximal possible agreement
value, which reached 59% between calibrated and modelled dates. The command Boundary was used to incorporate sedimentary boundaries, which might change sedimentation
rate. Sedimentation rate values are given on the right site of the age-depth curve. The horizontal line indicates the gyttja/peat boundary.
P. Hajkova et al. / Quaternary Science Reviews 141 (2016) 94e11198
layers of the sediment record. In the Swiss calibration dataset head
capsules of Derotanypus have not been differentiated from other
Tanypodinae larvae belonging to Macropelopia, Apsectrotanypus
and Psectrotanypus, since all these groups were very rare (max.
abundance 2.6%) and often missed glossae, paraglossae and other
diagnostic features. Therefore these taxa are also grouped in the
Swiss-Norwegian calibration dataset (Heiri et al., 2011). In the
absence of other options we assigned Derotanypus in the Hypkana
record to the category Apsectrotanypus/Derotanypus/Macropelopia/
Psectrotanypus in the calibration data. TJuly estimates were based on
weighted averaging partial least squares (WA-PLS) regression and
calibration of square-root- transformed chironomid percentage
data. A bootstrapped (cross-validated) root mean squared error of
prediction (RMSEP) was 1.4 C and r2
0.87 (Heiri et al., 2011).
125 diatom species from the fossil data, which were present also
in the calibration datasets, were used to infer epilimnetic total
phosphorus (TP) concentrations using diatom-water chemistry
transfer functions (Juggins, 2001) based on a combined European
diatom data-base (EDDI; http://craticula.ncl.ac.uk/Eddi/jsp). The
modern diatom calibration set consists of 477 samples and covers a
range of 2e1189 mg TP LÀ1
. The weighted averaging method (WA)
and log transformed TP values were used for reconstruction. A
jackknifed RMSEP was 0.33 log TP, r2
0.64, mean bias 0.002 and
maximum bias 0.72 log TP. For reconstruction, we used squareroot-transformed
diatom percentage data.
3. Results and interpretations
3.1. Chronology and sediment description
Using 11 radiocarbon dates we obtained a reliable depth-age
model (Fig. 2), which reached the agreement value of 59% between
calibrated and modelled values. The sedimentation rate in
the lake part of the profile was relatively stable and linear, ranging
between 0.07 cm yrÀ1
(in the depth of 935e802 cm and
542e323 cm) and 0.15 cm yrÀ1
(in the depth of 702e542 cm). The
error values varied mostly between 50 and 100 years, being only
higher (120e220 years) in the depth of 280e500 cm.
The lake sediment (gyttja) accumulated from the LG (ca.
13,000 cal yr BP) up to ca. 4800 cal yr BP. The bottom layer
(1115e1037 cm) consisted of light greyish-brown gyttja, the layer
1037e1031 cm of light grey gyttja with admixture of sand, the layer
1031e1002 cm of greyish-brown gyttja with small inorganic
admixture with exception of 1025e1022 cm, which was more dark
and organic. The layer 1002e955 cm was built up by light greyishbrown
gyttja. At 955 cm there was a gradual transition from grey to
brown gyttja. The zone 955e920 cm was characterised by alternation
of dark and light brown layers. The layer 920e720 cm was
built by dark brown gyttja and between 720 and 705 cm there was a
gradual transition to brown gyttja (705e531 cm) and light brown
gyttja (531e323 cm). For more details see Appendix C.
3.2. Reconstruction of mean July temperature
Pollen analysis confirmed the age depth model for the site and
suggested that the record encompassed the entire Younger Dryas
(YD) period. The Allerød/YD transition is characterised by a Betula
pollen decrease, whereas the YD/Early Holocene transition is very
clearly distinguished by distinct increase of Betula pollen and steep
decrease of Pinus pollen. Based on the Swiss-Norwegian calibration
dataset and inference model for chironomid-based temperature
reconstruction, the reconstructed TJuly oscillated between 7 and
11 C in the LG (11.8e15.9 C if corrected to modern sea level;
Fig. 3). The lowest TJuly values were reconstructed at the end of the
Allerød period (ca. 13,050e12,950 cal yr BP; 6.9e7.3 C), at ca.
12,500 cal yr BP (8.5 C) and at ca. 12,000 cal yr BP (7.6 C, last
cooling). Periods of relatively high temperatures were reconstructed
for 12,850e12,600 (9.2e11 C) and 12,200e12,400 cal yr
BP (10.1e10.9 C). At the end of the YD before the LG/Holocene
transition, the first warming up to 10.8 C was reconstructed (dated
to ca. 11,900 cal yr BP in our record). The next warming can be
already attributed to the LG/Holocene transition. At 11,600 cal yr BP
TJuly increased to 13.7 C and at 11,400 cal yr BP to 15.5 C, which
was the highest reconstructed value of TJuly within the whole Early
and Middle Holocene in the study site, although large sections of
the interval 11,000e8700 cal yr BP were characterised by very
similar temperature values. These temperatures were also higher
than recent (1961e1990) mean daily July temperature at the
elevation of the study site (14.7 C). Comparing the course of Holocene
temperatures, a distinctly cooler phase was reconstructed
between 8700 and 8000 cal yr BP (from 13.9 C to 10.1e11.1 C) and
at about 7000 cal yr BP (from 12.2 C to 10.2 C).
3.3. Lake development in the Late Glacial
Radiocarbon dating suggests that the lake originated at the end
of the Allerød interstadial due to landslide activity, which created a
dam on the small brook discharging on the hill slopes. Higher
abundance of Betula, Ulmus and Quercus pollen rather confirm this
age of origin, however, the rest of pollen spectra do not differ
substantially from that typical for YD vegetation. Therefore it may
also be possible that the oldest sediment layers originate from the
earliest section of the YD interval. The Greenland ice core records
indicate that the YD interval started around 12850 cal yr BP which
overlaps with our 14
C age, but the accuracy of age-depth model in
this section is only ca 100 years. In the LG, the lake was a shallow
pond as is indicated by chironomids and diatoms (Figs. 4 and 5,
zones Hd-1a, 1110e1085 cm, Hch-1a, 1115e1080 cm;
13,100e12,750 cal yr BP). Chironomid assemblages were composed
of oligotrophic and cold-demanding taxa such as Tanytarsus lugenstype
and Derotanypus sp., pioneering taxa such as Corynocera
ambigua-type, semiterrestrial taxa (Limnophyes sp.) and chironomids
from nearby streams and terrestrial environments (Georthocladius
sp., Pseudoorthocladius sp.). Diatoms were represented by
small benthic forms of the genera Staurosira, Pseudostaurosira and
Staurosirella, which are pioneer species typical for cold oligotrophic
lakes with frequent ice cover. Planktonic species (Asterionella formosa,
Fragilaria crotonensis, Stephanodiscus dubius) were continually
increasing in this developmental zone (Figs. 5 and 6) possibly
reflecting increasing water level (cf. also increasing Tanytarsus
lugens-type, Chironomus anthracinus-type). Oligotrophic algae taxa
such as Pediastrum integrum and Pediastrum kawraiskyi were
typical for the initial zone (Fig. 6). Around the shallow lake, wetland
vegetation dominated by Cyperaceae and Sphagnum species
developed (Fig. 7, Hp-1). In the next zone (Hd-1b, 1080e1040 cm;
Hch-1b; 1075e1040 cm; 12,700e12,400 cal yr BP) evidence suggests
that the water depth was continually decreasing and planktonic
diatoms were again substituted by small benthic species
(Pseudostaurosira brevistriata, Opephora mutabilis, Staurosira construens
var. venter). At about 1040 cm terrestrial species like Pinnularia
obscura, Pinnularia borealis, P. schoenfelderii and
Microcostatus cf. kraskei indicate dry conditions. Cold-demanding
species with high demands for oxygen which can also colonize
running waters, occupied (but in low abundances) probably the
shallow water of the cold lake or nearby streams (Heterotrissocladius
marcidus-type, Stempellinella sp.). The next sediment
layer (zones Hd-2 þ Hch-2, samples in 1030 cm and 1030e1035 cm,
respectively, ca 12,300 cal yr BP) suggests total lake desiccation
which is indicated by a sole presence of (semi)terrestrial Limnophyes
sp. with a single head capsule, a species-poor terrestrial
P. Hajkova et al. / Quaternary Science Reviews 141 (2016) 94e111 99
diatom assemblage (Hantzschia abundans, H. amphioxys, Pinnularia
borealis) and the almost total absence of green algae (Fig. 6). Hence,
the TJuly reconstruction from this layer was biased and therefore not
used. Later (zones Hd-3a þ b, Hch-3, 1020e970 cm,
12,200e11,700 cal yr BP), the succession typical for a shallow pond
started again with terrestrial chironomid groups (Smittia sp.),
species typical of cold oligotrophic lakes (Corynocera ambigua-type)
and taxa tolerating low temperatures (Derotanypus sp., Hydrobaenus
sp., Stempelinella sp.). Chironomid assemblages suggest that
the water level was gradually increasing, but the lake was probably
still shallower than before desiccation. Diatoms indicate gradual
restoration of the lake-environment as well as a development of
phytoplankton. From ca 1000 cm upwards (12,000 cal yr BP),
phytoplankton strongly decreased and it was substituted by small
flagilaroid species (Staurosirella pinnata, Staurosira construents var.
venter, Opephora mutabilis). The presence of aerophytic and limnoterrestrial
diatoms (Chamaepinnularia aerophila, Caloneis aerophila)
might indicate either water level decrease, development of a
shallow littoral environment or erosion from the surroundings. An
increase of macrophytes in the littoral area is also suggested by
chironomids (Cricotopus intersectus-type, Tanytarsus pallidicornis-
type).
3.4. Lake development in the Early Holocene
The very beginning of the Holocene (from ca 11,600 cal yr BP
onwards) was characterised by the onset or increase of chironomids
requiring higher temperatures and higher trophic states (e.g.
Microtendipes pedellus-type, Paratanytarsus penicillatus-type).
Groups adapted to cold conditions decreased or even disappeared
(Corynocera ambigua-type, Derotanypus sp., Zavrelimyia type A).
Tanytarsus lugens-type also disappeared for a short time, since the
lake had not yet developed an increased water table and deep
profundal. Also phytophilic chironomid groups (e.g. Glyptotendipes
pallens-type, several Cricotopus types) occurred in this zone (Hch-
4a; 960e890 cm, 11,600e10,700 cal yr BP) and overall chironomid
diversity increased steeply (Fig. 3). The green algae assemblage
(Fig. 6) is characterised by the disappearance of cold-demanding
Pediastrum kawraiskyi, decrease of oligotrophic P. integrum and
steep increase of planktonic Tetraedron minimum and Scenedesmus
sp. Diatom assemblages were characterised by the dominance of
planktonic species as well (Asterionella formosa, Stephanodiscus
dubius). Stauroneis smithii was substituted by S. gracilior. Local
wetland vegetation around the lake consisted of Alnus (ca 25% of
terrestrial pollen sum) and Cyperaceae, which reached lower
abundance than in the previous zone (Fig. 7; Hp-2; 965e925 cm; ca
11,700e11,200 cal yr BP). The presence of macrophytes was indicated
by epiphytic diatom taxa (e.g. Cocconeis pediculus, Cocconeis
placentula, Epithemia goeppertiana, E. andata). Running water taxa,
probably coming from a small stream nearby, were present within
both, chironomids (Epoicocladius sp., Chaetocladius sp.) and diatoms
(Planothidium frequentissimum, P. lanceolatum, Reimeria sinnuata).
Even though delimitation of zone 4 is similar for both proxies
(960e705 cm for chironomids and 955e720 cm for diatoms; ca
11,600e8950e(8800) cal yr BP), the subzones are positioned
differently. Chironomids indicate an earlier change (between 890
and 840 cm), characterised by an increase in abundance of Cladotanytarsus
mancus-type and Tanytarsus lugens-type. The highest
5000
5500
6000
6500
7000
7500
8000
8500
9000
9500
10000
10500
11000
11500
12000
12500
13000
)PBrylac(egA
3 6 9 12 15 18
°C
TJuly
(820
m
a.s.l.)
12 14 16 18 20 22
°C
TJuly
(0
m
a.s.l.)350
400
450
500
550
600
650
700
750
800
850
900
950
1000
1050
1100
)mc(htpeD
1.5 2.0
log TP
totalphosphorus
T
1.0 2.0 3.0
1.DC
A
axis
diatom
s
1.0 2.0 3.0
1.DC
A
axis
chironom
ids
6 18 30 42 54 66
sp/400 valves
diatom
diversity
800 1600
Betula
pollen
influx
1.0 2.0 3.0
1.DC
A
axis
pollen
-41 -38 -36 -33
d18O (permil)
N
G
R
IP
G
50 100
hc/1g
chironom
id
productivity
0 4 8 12 16 20
chironom
id
diversity
(rarefacted)
grains.cm-2.yr-1
SD units SD unitsSD unitsδ18O (‰)
Fig. 3. Results of the chironomid-based TJuly reconstruction (at 820 m a.s.l. and adjusted to modern sea level using lapse rates of 0.6 C/100 m), the NGRIP d18
O record, Betula pollen
influx, 1st DCA axis of pollen, diatoms and chironomids, chironomid productivity (head capsules per 1 g of dry sediment), chironomid diversity (rarefacted number of taxa), diatom
diversity (species per 400 counted valves) and diatom-based epilimnetic total phosphorus. The asterisks indicate layers with Derotanypus presence and the arrow indicates the dry
layer. The colder periods discussed in the text (low reconstructed TJuly) are indicated by grey shadings. Data about 18
O concentrations from Greenland ice core were obtain from
http://www.iceandclimate.nbi.ku.dk/data/ This data file accompanies the following two papers: Seierstad et al. (2014) and Rasmussen et al. (2014).
P. Hajkova et al. / Quaternary Science Reviews 141 (2016) 94e111100
300
350
400
450
500
550
600
650
700
750
800
850
900
950
1000
1050
1100
Depth(cm)
Hch-1a
Hch-1b
Hch-2
Hch-3
Hch-4a
Hch-4b
Hch-5a
Hch-5b
Hch-6
0 20
C
hironom
us
anthracinus
t.
0
C
orynoneura
t.A
0
Zavrelim
yia
t.A
0
Parakiefferiella
bathophila
t.
0
H
eterotrissocladius
m
arcidus
t.
0
Stem
pellinella
sp.
0 20
D
erotanypus
sp.
0 20 40 60
C
orynocera
am
bigua
t.
0 40 80
Lim
nophyes
sp./Paralim
nophyes
sp.
0 20
C
hironom
us
plum
osus
t.
0
Pagastiella
sp.
0 20
Paratanytarsus
penicillatus
t.
0 20
M
icrotendipes
pedellus
t.
0 20
C
ladotanytarsus
m
ancus
t.
0 20 40
Procladius
sp.
0
C
ricotopus
spp.
0 20 40 60 80
Tanytarsus
lugens
t.
0 20 40
Tanytarsus
pallidicornis
t.II
0 20
G
lyptotendipes
pallens
t./barbipes
t.
0
Paratendipes
albim
anus
t.
0
C
orynoneura
arctica
t.
0 20
M
icropsectra
contracta
t.
0Endochironom
us
albipennis
t.
0
cf.H
ydrobaenus
sp.
0
Paracricotopus
sp.
0
C
haetocladius
sp.
0
Epoicocladius
sp.
0
C
hironom
us
sp.juv.
0
Param
etriocnem
us
sp./Paraphaenocladius
sp.
0
M
onopelopia
sp.
0
Sm
ittia
sp./Parasm
ittia
sp.
0
G
eorthocladius
sp.
0
Psectrocladius
sordidellus
t.
0
Stenochironom
us
sp.
0
Psectrocladius
flavus
t.
0
C
ladopelm
a
lateralis
t.
0
Polypedilum
sordens
t.
0
Brillia
sp.
0
N
anocladius
branchicolus
t.
0
Phaenopsectra
flavipes
t.
0 20
Polypedilum
nubeculosum
t.
0 20
Einfeldia
natchitocheae
t.
0
C
orynoneura
edw
ardsit.
0 20 40
Tanytarsus
m
endax
t.
0
Pseudorthocladius
sp.
0
Lauterborniella
sp.
5500
6000
6500
7000
7500
8000
8500
9000
9500
10000
10500
11000
11500
12000
12500
13000
Age(cal.yrBP)
YDEarlyHoloceneMiddleH.
%
Hypkaňa (820 m a.s.l.; Vihorlat Mts; Slovakia)
Analyst: P. Pařil
Fig. 4. Stratigraphical diagram of the most important and abundant chironomid morphotypes. The taxa are given in percentages. Rare groups were merged into wider species complexes or to genus level for illustration purposes (but
not for quantitative reconstruction). The local chironomid zones are based on results of Coniss cluster analyses with square root transformation of fossil data. Chironomids were identified by P. Paril (t. ¼ type according Brooks et al.,
2007).
P.Hajkovaetal./QuaternaryScienceReviews141(2016)94e111101
330
380
430
480
530
580
630
680
730
780
830
880
930
980
1030
1080
Depth(cm)
Hd-1a
Hd-1b
Hd-2
Hd-3a
Hd-3b
Hd-4a
Hd-4b
Hd-5a
Hd-5b
Hd-6
0
Stauroneis
sm
ithii
0
C
aloneis
aerophila
0 20
H
antzschia
abundans
0 20 40
Pinnularia
borealis
0 20 40
H
antzschia
am
phioxys
0 20
G
om
phonem
a
parvulum
0
R
eim
eria
sinuata
0 20 40
Asterionella
form
osa
0 20
Encyonem
a
silesiacum
0 20
Fragilaria
crotonensis
0
Achnanthidium
exiguum
0 20 40 60
Stephanodiscus
dubius
0 20
Am
phora
copulata
0 20
O
pephora
m
utabilis
0
Pinnularia
schoenfelderii
0
Epithem
ia
adnata
0
C
aloneis
silicula
0 20 40 60
Staurosira
construens
var.venter
0 20
N
avicula
radiosa
0 20
Pseudostaurosira
brevistriata
0 20 40 60 80
Staurosirella
pinnata
0 20 40 60
Staurosira
construens
var.binodis
0 20
Staurosira
m
utabilis
0
Stauroneis
gracilior
0 20 40
Aulacoseira
granulata
0
Planothidium
frequentissim
um
0 20 40
Staurosira
pseudoconstruens
0
Epithem
ia
goeppertiana
0 20
C
occoneis
placentula
0 20
C
occoneis
pediculus
0
Aulacoseira
am
bigua
0
Planothidium
lanceolatum
0
Tabellaria
flocculosa
5500
6000
6500
7000
7500
8000
8500
9000
9500
10000
10500
11000
11500
12000
12500
13000
Age(cal.yrBP)
YDEarlyHoloceneMiddleH.
%
Hypkaňa (820 m a.s.l.; Vihorlat Mts; Slovakia)
Analyst: B. Chattová
Fig. 5. Stratigraphical diagram of the most important and abundant diatom species. Species are given in percentages. The local diatom zones are based on results of Coniss cluster analyses with square root transformation of fossil data.
Diatoms were identified by B. Chattova.
P.Hajkovaetal./QuaternaryScienceReviews141(2016)94e111102
diversity of chironomid taxa within the whole profile was recorded
at the beginning of the subzone Hch-4b (at about 10,000 cal yr BP).
In contrast, a peak in diatom diversity appeared already at the
beginning of the subzone Hd-4a (at about 11,500e11,000 cal yr BP),
with diversity decreasing thereafter. Diatom species composition
indicates distinct change at 790e780 cm (Hd-4a/4b; ca 9400 cal yr
BP), where planktonic species (Asterionella formosa, Fragilaria crotonensis,
Stephanodiscus dubius) gradually decreased in abundance
(Fig. 6) and mesotrophic Encyonema silesiacum and Achnathidium
exiiguum contrarily increased. Simultaneously, the composition of
planktonic green algae also changed (at 800 cm; ca 9600 cal yr BP)
from an assemblage dominated by Tetraedron minimum to dominance
of Scenedesmus. Abundances of Pediastrum boryanum agg.
increased. At the end of the subzone Hd-4b, planktonic species
almost disappeared and benthic species started to dominate again.
A distinct change in all proxies was apparent at ca 710 cm. In the
local terrestrial vegetation, pollen abundance of Alnus again
increased (Hp-5; 710e430 cm; 8850e6250 cal yr BP). The diatom
assemblages of the zone Hd-5a (710e690 cm; 8850e8750 cal yr
BP) were characterised by dominance of tychoplanktonic species of
the genus Aulacoseira, mostly Aulacoseira granulata. This species
creates hard silicified frustules and therefore it requires an
increased degree of turbulence in order to stay in suspension
(Saunders et al., 2008). Thus, it indicates turbulent, unstable conditions
such as those caused by mixing of water layers by wind (cf.
Buczko et al., 2013). Increased turbulence is almost always associated
with increased nutrient flux from the hypolimnion (Stone
et al., 2011) which corresponds well with the diatom-inferred TP
increase at 710 cm indicating increased trophic state and productivity
of autotrophic organisms (diatoms, algae Pediastrum boryanum
agg., macrophytes). Also chironomids indicate a distinct
change in the depth between 710 and 690 cm, even if the reaction
to increased phosphorus was apparently not so distinct and the
representation of eutrophic (including mesotrophic) and oligotrophic
taxa was similar. From the depth of 690 cm onwards, the
taxonomic diversity was very low and cold-demanding and
profundal-preferring taxa such as Tanytarsus lugens-type and
Procladius sp. started to dominate in the record. Taxa requiring
coarse-grained substratum (Brillia sp., Microtendipes pedellus-type)
also occurred (subzone Hch-5a; 695e590 cm;
8750e8050 cal yr BP). Planktonic green algae disappeared (Tetraedron
minimum) or decreased (Scenedesmus; Fig. 6). Also diatom
assemblages (subzone Hd-5b; 680e550 cm; 8650e7750 cal yr BP)
were characterised by the disappearance of (tycho)-planktonic
species and dominance of benthic species (Amphora copulata,
Navicula radiosa, Pseudostaurosira brevistriata, Staurosira pseudoconstruens,
Staurosira construens var. binodis).
3.5. Lake development in the Middle Holocene
In the chironomid record (subzone Hch-5b; 545e440 cm;
7750e6350 cal yr BP), the more warm-demanding littoral taxa
(Cladotanytarsus mancus-type) and phytophilic taxa (Cricotopus
spp., Glyptotendipes pallens-type) appeared at the beginning of the
Middle Holocene (the beginning of the zone Hch-5b; ca
8000e7600 cal yr BP), but later their abundances again decreased.
Towards the end of this zone the abundance of taxa which can
colonize sediment or aquatic vegetation (e.g, Polypedilum nubeculosum-type)
increased, which suggest that the lake may have been
shallower and probably over-grown by macrophytes and wetland
vegetation (at about 450 cm; 6500 cal yr BP). This is supported also
by an increase of Potamogeton and Cyperaceae pollen and
Sphagnum spores (Figs. 6 and 7). In diatom assemblages (lower part
of the zone Hd-6; ca 540e430 cm; 7700e6250 cal yr BP), benthic
species dominated (e.g. Staurosirella pinnata) and were represented
at a high diversity. Total diatom species diversity steeply increased
(Fig. 3). There was also a higher representation of periphytic (Cocconeis
pediculus, Cocconeis placentula, Gomphonema spp.) and
tychoplanktonic species (Aulacoseira ambigua, Aulacoseira granulata)
as compared to the previous zone. Finally, the last phase
before the complete lake terrestrialization (Hch-6, 395e340 cm,
5750e5050 cal yr BP; Hd-6 e upper part; 390e340 cm;
5700e5050 cal yr BP) was characterised by a decrease in green
algae (Pediastrum boryanum agg., Fig. 6) and chironomids typical of
Fig. 6. Summarizing diagram showing representation of different diatom life forms, species of green algae, pollen of macrophytes and chironomid ecological groups. For information
about taxa included into particular categories see supplementary material S1 (chironomids) and S2 (diatoms).
P. Hajkova et al. / Quaternary Science Reviews 141 (2016) 94e111 103
Fig. 7. Pollen percentage diagram with time scale in calibrated years BP (before 1950). At the end of the diagram, microcharcoal particles influx and total pollen influx are given. The local pollen zones are based on results of Coniss
cluster analyses with square root transformation of fossil data.
P.Hajkovaetal./QuaternaryScienceReviews141(2016)94e111104
cooler and oligo-to mesotrophic lakes (e.g. Tanytarsus lugen-type).
Diversity of chironomids and diatoms distinctly increased (Fig. 3).
In chironomid assemblages, littoral species (Polypedilum nubeculosum-type,
Einfeldia natchitocheae-type) and groups often associated
with vegetation (Cricotopus spp.) increased in abundance and
ubiquistic groups such as Tanytarsus mendax-type started to
dominate. In diatom assemblages, species of epibryon (Tabellaria
flocculosa) and acidophilous species (of the genera Eunotia, Delicata,
Stauroneis, Neidium) appeared. In the pollen record (Hp-6,
425e345 cm, 6150e5100 cal yr BP), the increase of Cyperaceae
pollen and re-appearance of Sphagnum spores also indicate
spreading of local wetland and mire vegetation.
3.6. Regional vegetation development
The LG forests near the lake were mostly composed of Pinus
(40e90%) and Betula (5e30%) according to fossil pollen spectra and
also according to needles and seeds used for AMS dating (Fig. 7; Hp-
1; 1110e975 cm; ca 13,050e11,750 cal yr BP). Pollen of other trees
and shrubs including the temperate ones (Picea, Larix, Ulmus and
Quercus) were present in lower or very low abundances (Fagus,
Fraxinus and Corylus) and came probably from the lower altitude.
Juniperus, and Salix pollen suggested that in the shrub layer these
taxa may have occurred. Artemisia and Gramineae were dominant
pollen taxa suggesting open steppe or tundra vegetation in the
vicinity of the lake, less common were Chenopodiaceae, Thalictrum,
Filipendula, Urtica, Rumex acetosella t. (¼type) and steppe species
Ephedra fragilis t. and E. distachya t. The AP/NAP ratio fluctuated
between 65 and 90%. Microcharcoal was relatively frequent. After
the warming, at the transition between the LG and Early Holocene
(Hp-2; 970e920 cm; ca 11,700e11,150 cal yr BP), the pollen percentages
of Pinus strongly decreased (to 30%), whereas Betula (up to
30e40%) and Alnus (10e20%) increased. At the end of the zone Picea
and Ulmus pollen also increased, whereas other temperate trees
expanded to a lesser extent (Quercus, Tilia, Fraxinus, Corylus and
Fagus). The AP/NAP ratio was very high, around 95%. Larix and
Juniperus pollen almost disappeared together with Ephedra. Openlandscape
pollen taxa, such as Artemisia and Gramineae, remained
dominant among the herbs but their abundances decreased along
with Chenopodiaceae. Microcharcoal also decreased. In contrast,
pollen of Cannabis/Humulus t. appeared. The next zone (Hp-3;
915e820 cm,11,050e9750 cal yr BP) was characterised by the onset
of temperate deciduous forests in the landscape. Pollen of Ulmus
(30e40%) was dominant, pollen of Quercus (5e10%), Fraxinus
(5e10%), Tilia (<5%) and Picea (5e20%) increased substantially,
whereas pollen of Betula and Alnus decreased. This zone is also
characterised by higher abundance of Picea stomata indicating the
presence of this tree near the lake. The AP/NAP ratio was still high.
The zone Hp-4 (815e715 cm; 9700e8900 cal yr BP) was characterised
by a distinct decrease of Ulmus (from ca 30 to 10%) and steep
increase of Corylus pollen (up to 50e55%) indicating rather dry
climate. Pollen of Pinus, Picea and Betula decreased, whereas Acer
and Fagus pollen started to increase even if at low abundances. The
AP/NAP ratio stabilised around 95%. Artemisia and Gramineae
reached their lowest values, but their curves remained continual
and uninterrupted indicating presence of treeless vegetation
somewhere in a wide region, probably at lower altitudes. In the
next zone (Hp-5; 710e430 cm; 8850e6250 cal yr BP), Corylus
decreased slowly (down to 35%), but still remained dominant,
whereas Picea pollen and stomata increased slightly. Fagus and Acer
pollen curves were already continual (closed curves) and, at the end
of the zone, pollen of Carpinus also appeared. Microcharcoal influx
had decreasing values during this zone. Finally, zone Hp-6
(425e345 cm; 6150e5100 cal yr BP), the last zone of the pollen
record in the lake sediments, was characterised by a steep increase
in Fagus (up to 40%) and Carpinus (about 15%) pollen. Pollen of other
tree taxa (Ulmus, Quercus, Tilia, Fraxinus) decreased slightly (to
5e10%) and pollen of Corylus decreased steeply (to 5%). The first
pollen grains of Abies appeared and AP/NAP remained high.
3.7. Results of geochemical analyses
According to geochemical composition, the profile was divided
into six zones (Fig. 8). While the sediments were composed of
mainly mineral components in the zone HG-1, autochthonous
components like organic matter (represented by LOI) and biogenic
silica (Si or Si/Zr in XRF analysis, microscopic observation of diatom
frustules) were present in substantial amounts in all other zones. In
zone HG-1 there are several cycles of changing input of mineral
matter. In the minima of these cycles (minima of lithogenic elements)
there are the first maxima of Si/Zr representing the biogenic
silica to mineral matter ratio (1010e1005 cm; ca
12.100e12.050 cal yr BP). In the zone HG-2 the Si/Zr ratio is much
increased, whereas Zr/Rb and Zr/Ti decreased slightly, which we
interpret as fining of mineral grains. The change between zones
HG-1 and HG-2 (LG/Early Holocene transition) was very abrupt.
From zone HG-1 up to zone HG-4 the element ratios show
increasing weathering intensity, K/Ti decreases, while Rb/K increases.
In zone HG-3, only 10 cm thick but very distinct, there is a
maximum of mineral matter components (increase of all lithogenic
elements and MS) at the expense of biogenic silica, with only a
minor change in the lithogenic element ratios. This suggests that
the nature of the mineral matter did not change much but its input
was enhanced, or, alternatively, productivity suddenly dropped.
Zone HG-3 is an interruption of the system evolution. Subsequent
zone HG-4 is not much different geochemically than zone HG-2.
The Si/Zr ratio in zone HG-4 is even higher than in zone HG-2
indicating the highest proportion of the biogenic silica in the
sediment. The boundary of zones HG-4/5 is abrupt, the transient
strata are only 20 cm thick. Zone HG-5 resembles zone HG-3. The
amount of clastics increases at the expense of the biogenic silica
with the clastics characterised by enhanced proportion of silt/finest
sand content (higher Zr/Rb). Simultaneously MS increases nearly to
the level of the prevalently clastic zone HG-1. Zone HG-5 is terminated
abruptly by a return to enhanced biogenic silica content at
the expense of mineral matter components (beginning of the zone
HG-6). Contrarily to the zones HG-2 and HG-4, MS is enhanced in
the zone HG-5 and 6, although total Fe content remains more or
less the same. Generally, higher MS in zones HG-5 and HG-6 is not
proportional to the rather weakly enhanced Zr/Rb ratio. This,
together with the nearly constant content of total Fe, suggests that
these higher values are probably not related to a higher amount of
coarser lithogenic magnetite grains or other magnetic mineral
grains. Either large amounts of diamagnetic biogenic opal in underlying
zones (high Si/Zr) or post-depositional destruction of
magnetic minerals decreased MS in the zones with high
productivity.
4. Discussion
4.1. Potential effects of productivity and water table changes on the
reconstruction
The TJuly reconstruction based on chironomids corresponds well
with results of other climate proxies in East-Central and Eastern
Europe (pollen: Feurdean et al., 2008a,b; chironomids: Toth et al.,
2012, 2015; Płociennik et al., 2011; stalagmites: Tamas¸ and
Causse, 2001, Demeny et al., 2013). It was recently illustrated that
co-varying factors in calibration datasets may influence reconstructions
based on biotic proxy-indicators (Velle et al., 2010;
P. Hajkova et al. / Quaternary Science Reviews 141 (2016) 94e111 105
300
350
400
450
500
550
600
650
700
750
800
850
900
950
1000
1050
1100
Depth
(cm
)
HG-1
HG-2
HG-3
HG-6
HG-4
HG-5
0 10 20 30
Al
0 200 400
Si
0 5 10 15 20
S
0 200 400
K
0 1000
C
a
0 4000
Fe
0 20 40 60 80
M
n
0.0 0.1 0.2
R
b/K
0.0 0.1 0.2
Zr/Ti
0 12 24 36 48
Si/Zr
0.0 2.0 4.0 6.0
Zr/R
b
0.2 0.4 0.6 0.8
K/Ti
0 20 40 60
LO
I
0 12 24 36
M
S
5000
5500
6000
6500
7000
7500
8000
8500
9000
9500
10000
10500
11000
11500
12000
12500
13000
Age
(cal.yrBP)
YDEarlyHoloceneMiddleH.
m3
/kg.10-9
%
Hypkaňa (820 m a.s.l.; Vihorlat Mts; Slovakia
Analyst: T.M. Grygar
Fig. 8. Results of geochemical analyses, magnetic susceptibility (MS) and loss-on-ignition (LOI at 550 C). Also ratios of some elements are shown.
P.Hajkovaetal./QuaternaryScienceReviews141(2016)94e111106
Juggins, 2013). In the case of reconstructions based on chironomids,
Brodersen and Anderson (2002) demonstrated that taxa with high
temperature optima occur mostly in eutrophic conditions and taxa
with low temperature optima occur in oligotrophic conditions. This
positive correlation between temperature and trophic status was
documented to some extent in most of the published calibration
datasets (Brooks and Birks, 2001; Heiri and Lotter, 2005; Heiri et al.,
2011). If these two environmental factors develop independently,
the TJuly reconstructions may be biased. Therefore we reconstructed
trophic status using diatom-inferred TP to assess whether past
variations in TP may have led to potential problems in the TJuly
reconstruction. Diatom-inferred TP stayed relatively constant in the
profile varying between 23 mg LÀ1
(in 955 cm, 11.500 cal yr BP) and
80 mg LÀ1
(in 700 cm, 8800 cal yr BP). However, an increase (from
39 mg LÀ1
to 79 mg LÀ1
) is registered at 8850 cal yr BP (from 710 cm
up), with fluctuating values and only few short-term decreases
thereafter (most distinct in 640e630 cm; 8300e8400 cal yr BP). In
contrast, chironomid-inferred TJuly decreased at 8700 cal yr BP and
with the exception of 7600e7700 cal yr BP stayed lower than in the
Early Holocene (Fig. 3). As the relationship between temperature
and trophic status in this part of the profile is negative, not positive,
we exclude a bias in TJuly reconstruction caused by a positive
interaction. Further, also minimum values during the YD (at ca
12,000 and 12,500 cal yr BP) were characterised by moderate increases
in diatom-inferred TP (Fig. 3).
Another factor, which might influence TJuly reconstruction is the
changing depth of the lake as water in the shallow lake can be
strongly warmed during the summer, whereas deep stratified lakes
are characterised by profundal environments with optimal conditions
for cold-demanding chironomid taxa (Velle et al., 2010). A
deepening of the lake could therefore, in principle lead to a
decrease in reconstructed Holocene temperatures. However, the
most pronounced decrease in the Holocene part of the record at
8700 cal yr BP was probably associated with a decrease of the lake
depth probably due to terrestrialization, as suggested by a decrease
in planktonic diatoms and algae and by high abundance of Aulacoseira
species (Saunders et al., 2008). An opposite situation was
observed within the YD period. Here, higher reconstructed TJuly
values are related to the lower inferred lake depth (Fig. 3). Thus in
this section of the record the reconstructed TJuly values could be
slightly biased (increased) by lake level variations, although they
still remained in the range of the YD temperatures and were clearly
cooler than reconstructed Holocene temperatures. Finally, the
reconstructed increase of TJuly after 7000 cal yr BP could be called
into question, because the lake started to terrestrialize up to ca
5000 cal yr BP, when the open water surface completely disappeared
and thus the depth of lake was decreasing.
4.2. Younger Dryas TJuly fluctuations and the LG/Holocene transition
Focusing on the reconstructed TJuly for the YD, we recorded very
variable temperatures between 7.2 and 10.8 C at 820 m a.s.l. (i.e. ca
12.1e15.7 C adjusted to modern sea level, ca 49N). These values
are considerably lower (about ca 4.5 C) than those reconstructed at
more southerly locations in the eastern and central parts of
Southern Europe (42e44.5N) or the Alpine region (ca 46N), but
about 3 C higher than in the Baltic region (56.5e58N) or British
Isles (54e55.5N) situated more to the north (Heiri et al., 2014a). It
seems that reconstructed values reflect the latitude, because
reconstructed TJuly is intermediate. As the study site is geographically
closer to the Alps than to the Baltic region, the reconstructed
values are likely slightly lower then would be expected according to
latitude, which might be attributed to the exposition of the study
site (northern slopes). The comparison with the record from central
Poland (almost 52N, 13e16 C corrected to 0 m a.s.l.; Płociennik
et al., 2011) also suggests slightly lower TJuly values at our study
site than would be expected according to latitude. The rather
distinct climatic fluctuations during the YD resemble minor variations
recorded in the oxygen isotope records of the Greenland ice
sheet (e.g. Lowe et al., 2008; Rasmussen et al., 2014) and also by
other non-quantitative climate proxies in Europe (Von Grafenstein
et al., 1999). However, Younger Dryas summer temperatures
reconstructed in other chironomid records (e.g. Toth et al., 2012;
Ilyashuk et al., 2009; Brooks and Birks, 2001) are much more stable
and other YD reconstructions from Europe also do not show
such minor centennial-scale oscillations (e.g. Lauterbach et al.,
2011).
Some previous studies have suggested that YD climate was
unstable. For example, it has been proposed that rapid alternations
between glacial growth and melting may have affected sea-ice
cover and the influx of warm salty water in the Nordic sea (Bakke
et al., 2009). Schwark et al. (2002) suggested a middle Younger
Dryas warming (MYDE; 12,200e12,300 cal yr BP) in southwestern
Germany by increased pollen productivity of Betula and increased
content of nC27-alkanes in the lake sediment which correspond
with increased input of Betula litter to the lake. Such short-term
warmings are also apparent in the NGRIP and GRIP ice core data
(Rasmussen et al., 2014), which show several short positive excursions
of d18
O, although the amplitude of these changes is small
compared with the centennial scale shifts between the YD and
adjacent climate periods. High YD fluctuations were recorded also
in the Eastern Carpathians by means of variability in 18
O content in
stalagmites (Tamas¸ and Causse, 2001). Sediment composition from
the YD period also varied considerably in our record, confirming
that local environmental conditions were affected by climatic
changes at our study site (Fig. 8). In the chironomid-inferred TJuly
record, there are two warmer periods dated to 12,850e12,600 cal yr
BP and, more distinctly, at 12,400e12,200 cal yr BP within the YD.
These warmer phases coincide with distinct water level declines
which are indicated in our record by gradual displacement of
aquatic taxa by terrestrial ones and by changing input of mineral
matter into the lake (Fig. 8). Low lake level episodes have been
recorded also in Central Europe during the YD and were accompanied
by changes in sediment composition as well (cf. Karasiewicz
et al., 2013). Since temperature variations inferred by chironomids
within the YD co-vary with apparent changes in water table, this
may have influenced the results. Furthermore, inferred temperature
variations within the YD are within the prediction error of the
age-depth model (Fig. 3). We conclude that our results suggest that
July air temperatures may have shown centennial-scale variations
at the study site, but that these changes coincided, and may have
been amplified, by local hydrological changes. Additional evidence
(e.g. additional YD temperature reconstructions for the study region)
will have to be developed to resolve whether these variations
represent real air temperature changes or chironomid response to
other environmental variations (e.g. changes in water table).
The LG/Holocene transition was characterised by a rapid increase
of the reconstructed TJuly. The beginning of the warming in
our record is dated to between ca. 11,700 and 11,600 cal yr BP (from
10.3 to 13.7 C), which agrees with the onset of Holocene assessed
as Global Stratotype section in NGRIP (11,700 b2k, Walker et al.,
2009). In the next sample temperatures rise by another 1.8 C
(from 13.7 to 15.5 C) to reach a first maximum at ca 11,400 cal yr
BP. Thus in ca. 300 years the TJuly has apparently increased by ca.
5 C. According to the review of Heiri et al. (2014a), a similar steep
increase of reconstructed TJuly has been recorded in some sites in
Alpine region (by ca 4 C) but especially also at higher latitudes
(British islands by ca 5 C, Norway 4e5 C). Much lower and slower
warming occurred at low latitudes (SW Europe, E and CS Europe)
and altitudes (Baltic region, Heiri et al., 2014a; almost no change in
P. Hajkova et al. / Quaternary Science Reviews 141 (2016) 94e111 107
Holocene onset in middle Poland, Płociennik et al., 2011). The only
chironomid-based TJuly reconstruction from the whole Carpathians
(45N, Retezat Mts; Toth et al., 2012) has also shown a delayed
warming at the beginning of the Holocene by about 2.5e3 C. At the
YD to Holocene transition no major changes in summer temperature
was detected in this record, but summer temperatures then
increased in two steps and reached 12.0e13.3 C during the Preboreal.
The comparison of pollen data from the Western Carpathians
and the adjacent areas indicates that during the LG the
pollen spectrum of the study site was similar to the sites located in
steppe-tundra landscapes of the Inner-Carpathian basins (e.g.,
Hajkova et al., 2015), while just after the Holocene onset it became
similar to samples from Western Slovakia which are rich in
temperate trees (Petr et al., 2013; Hajkova et al., 2013). Such a
conspicuous development towards a temperate landscape is
exceptional within East-Central Europe as is indicated also by
gradient analysis of available pollen samples from the LG and Holocene
onset (Jamrichova et al. unpublished data) and might be
caused by the geographic position of the study site on the boundary
between oceanic and continental climate influences. The LG/Holocene
transition was apparently accompanied by a steep increase
of lake depth indicated by increased representation of planktonic
diatoms and algae (Tetraedron minimum), by a decrease of coarseclastic
input to the lake sediment and a steep decline of magnetic
susceptibility. These changes could be attributed to the climate
moistening which followed climate warming (e.g. Feurdean et al.,
2008a). This climate improvement triggered rapid spread of deciduous
trees, firstly Betula, which has probably spread into the
semi-open forest-tundra above the lake, then Alnus, which has
occupied wet places near the lake and forest springs (ca
11,600 cal yr BP) and later also Ulmus (11,300e11,000 cal yr BP). The
rapid reaction of Betula to climate improvement was recorded also
by macrofossils in the Eastern Carpathians (at ca 11,500 cal yr BP;
Feurdean et al., 2008b) and by pollen and lipid biomarkers in
Central Europe (Schwark et al., 2002).
4.3. Holocene thermal maximum
Another often studied and discussed topic is the timing and
duration of the Holocene thermal maximum (HTM) in summer
temperature, which is mostly positioned in the large interval of 11
and 5 ka BP (Renssen et al., 2009, 2012). Based on global
atmosphere-ocean-vegetation model runs, Renssen et al. (2012)
have revealed that in large sections of Europe the timing of HTM
is expected to be between 7 and 6 ka BP, which agrees with some
proxy-based reconstructions (e.g. Heiri et al., 2014b; Renssen et al.,
2009). For example, based on pollen assemblages Davis et al. (2003)
reconstructed maximum summer temperatures around 6 ka for
Northern Europe and Western Central Europe and around 7e8 ka in
central eastern Europe. However, reconstruction based on other
proxy records placed the HTM to different sections of this interval,
some of them also to the Early Holocene (9500e9100 cal yr BP
(Toth et al., 2015) or 10,000e8600 cal yr BP (Ilyashuk et al., 2011)).
In our study site, the sample with the highest reconstructed TJuly
was at the beginning of the Holocene (15.5 C at 11,400 cal yr BP and
15.0 C at 10,750 cal yr BP), but the whole period
11,600e8850 cal yr BP was relatively warm with temperatures
being slightly higher, the same or slightly lower than today
(13.7e15.5 C; today temperature for the altitude of the lake estimated
to 14.7 C). Temperature was slowly decreasing from ca.
10,500 cal yr BP onwards toward the Early/Middle Holocene transition
with an exception of short cooling event at about 11,200 cal yr
BP (11.9 C). A possible reason for the discrepancy between the
different records is that local orography and the effects of mountain
ranges may have influenced local climates, leading to earlier
Holocene temperature maxima in regions which are downwind
from mountain ranges and, e.g., shielded from the influence of
North Atlantic air masses. In these situations the high summer
insolation values of the Early Holocene may have led to higher
temperatures than later in the Holocene period. Alternatively,
changes in water tables discussed above may have to some extent
affected chironomid assemblages and influenced the trend in
chironomid-inferred temperatures. However, since an early HTM
has also been reported from other Holocene records from European
mountain lakes, we consider orographic effects or regional differences
in the timing of the HTM across Europe the more likely
explanation.
4.4. Early/Middle Holocene cooling
We recorded a distinct TJuly decrease (from 13.9 to 10.1 C) between
8850 and 8750 cal yr BP, which lasted to ca 8000 cal yr BP
(Fig. 3). A distinct cooling event at 8200 yr BP lasting for 100e200
years has been reported from the Greenland ice core oxygen
isotope records but also many other climate records from around
the North Atlantic (Alley and Agústsdottir, 2005). Mayewski et al.
(2004) showed that this cooling event was embedded, or formed
part of, a longer lasting period of climate variations they named the
first Holocene rapid climate change (RCC) placed to
9000e8000 cal yr BP. Short lived cooling events within this period
have been reported from several other chironomid records. Heiri
et al. (2003) described a period of cooler summer temperatures
(ca. 8200e7800 cal yr BP) which coincided with changing meltwater
flux from the American continent to the North Atlantic (Heiri
et al., 2004). A decrease of chironomid-inferred TJuly was also
recorded in the central Poland (Płociennik et al., 2011:
8700e8000 cal yr BP) and from the eastern central Alps (Ilyashuk
et al., 2011: 8200e8000 cal yr BP). A temperature decrease was
also recorded by oxygen isotope composition of diatoms in the
Eastern Carpathians (Magyari et al., 2013) and by oxygen isotope
data from a Hungarian speleothem (Demeny et al., 2013:
9000e8000 yr BP). Based on pollen-inferred temperatures,
Feurdean et al. (2008a) detected a decrease of winter temperatures
(in the coldest month) between 8200 and 8000 cal yr BP, whereas
summer temperatures did not change. Total pollen influx in our
record (mostly of Ulmus, Quercus, Fraxinus and Corylus; data not
shown) decreased at ca 8700 cal yr BP probably due to decreased
pollen productivity as a reaction to climate deterioration (Fig. 7,
Andersen, 1980; Sj€ogren et al., 2006). At the study site, simultaneously
other analysed parameters and proxies also changed at
about 8700 yr BP. The duration of the cold episode at
8700e8000 cal yr BP is distinctly longer in our record than reported
for the 8.2 ka event (e.g. Wiersma and Renssen, 2006). However,
since we have only a limited number of radiocarbon dates in this
section of the record (Fig. 2, Table 1) we cannot rule out that this
event is shorter than dated in our records. Alternatively, the cooling
could be related to the longer first Holocene RCC described by
Mayewski et al. (2004).
Geochemical analyses have shown that the cool period in our
core came after an abrupt short-term input of mineral matter into
the lake indicated by an increase of all lithogenic elements and MS
and decrease of biogenic silica proxy (cf. zone HG-3, Fig. 8), which
was reflected by dominance of tychoplanktonic diatoms in the
sediment (Fig. 6). Both proxies could indicate increased precipitation
resulting in higher erosion and more intense water column
mixing in the lake, such as due to enhanced wind. Enhanced mixing
usually creates an upward flux of nutrients and silica from the
hypolimnion (Stone et al., 2011), which would agree with increased
diatom-inferred TP in our record. Also increased Alnus pollen (both
percentages and influx) might indicate higher wetness and
P. Hajkova et al. / Quaternary Science Reviews 141 (2016) 94e111108
development of wet alder carr in the terrestrialized littoral or
expanding wetlands in the surrounding landscape. Moreover, Alnus
was revealed to have higher pollen productivity under wetter
climate in the research of pollen traps (Van der Knaap et al., 2010).
The climate moistening in this period (8700e8600 cal yr BP) is also
documented by a conspicuous expansion of temperate forests in
our record (decline of Corylus, slight increase of Ulmus, Fagus and
Picea) and also in other records from the deciduous-forest zone of
the Carpathians (Feurdean et al., 2015; Hajek et al., 2016). Increased
precipitation in Central and East-Central Europe was reconstructed
also by testate amoebae (higher water level at 8300e8000 cal yr BP,
Schnitchen et al., 2006), by increased water level of lakes
(8300e8000 and 7500e7000 cal yr BP, Magny, 2004; 8400 cal yr
BP, Buczko et al., 2013) or climate-mediated decline of fire activity
expressed by decrease of charcoal influx at about 8500 cal yr BP
(Feurdean et al., 2012, 2015). Even if the timing in these records is
not completely the same as in our study, the differences are not so
high to clearly exclude synchronicity and might be caused by some
inaccuracy in age-depth models. Nevertheless, the timing of
climate moistening could be also affected by local and regional
factors like position on mountain ranges, distance from the Atlantic
or prevailing wind direction. According to Magny et al. (2003), midlatitudes
between ca 50 and 43 responded to the cooling by
wetter climate, whereas Southern and Northern Europe had
contrarily drier climate. The extent of this wet mid-European zone
could reflect the strength of the Atlantic Westerly Jet in relation to
the thermal gradient between low and high latitudes (Magny et al.,
2003). Moreover, climate moistening was probably associated with
increased cloudiness, which could have prevented warming of the
lake water in summer and thus locally amplified the effect of
cooling and decreased values of reconstructed TJuly.
It is remarkable that indicators of erosion and turbulent water
conditions were present only before the reconstructed cold period
and later no input of eroded inorganic matter from the surrounding
landscape to the lake was detected by geochemical analyses (Fig. 8).
Likely the development of terrestrial vegetation in the lake surrounding
triggered by climate moistening prevented further
erosion, as suggested by increase of tree pollen (Tilia, Fagus, Picea)
and decrease of Corylus pollen. This period is also characterised by
increased diatom-inferred total phosphorus content (Fig. 3).
Increased trophic state and nutrient availability for autotrophic
organisms in the lake is indicated also by increased abundance of
nutrient-demanding algae Pediastrum boryanum agg. (Fig. 6). There
are several possible mechanisms of nutrient enrichment of lake
water. It could have been caused by increased input of litter from
nearby trees, especially Alnus. Another explanation could be flux of
nutrients from the hypolimnion due to enhanced wind activity (e.g.
Stone et al., 2011). Input of phosphorus through the erosion from
the surrounding catchment may have been possible only at the
beginning of the cold phase, later such an increased nutrient input
is not supported by geochemical analyses, as the proxies for clastic
input (Zr/Rb, Zr/Ti or Rb/K) did not change. Another possible
mechanism for changes in nutrient availability in the lake could
have been changes in seasonality connected with cooling of
climate. The prolonged winters could have caused longer ice cover
and thus stronger stratification in the lake, anoxia in the hypolimnion
and consequent internal phosphorus loading (Kirilova
et al., 2009). Released phosphorus would then become available
for organisms after spring mixing of hypo- and epilimnetic layers.
However, increased phosphorus concentrations were not so
distinctly reflected in the species composition of chironomids since
the proportion of eutrophic and oligotrophic taxa was rather equal
in this zone (Fig. 6) and oligotrophic taxa distinctly increased their
abundances compared to the previous zone. It was documented
that the reaction of heterotrophic chironomids to changes in lake
water nutrients is not as predictable as the reaction of autotrophic
organisms because chironomids are influenced by increased productivity
indirectly through the food availability (macrophytes,
algae, but also other invertebrates and organic detritus) and oxygen
concentrations (Brodersen and Lindegaard, 1999; Brodersen and
Quinlan, 2006), which could be influenced also by other factors
than phosphorus availability. Moreover, fossil chironomid records
from the central part of small lakes are composed by species from
both, the littoral and the profundal, where different environmental
conditions can occur (Van Hardenbroek et al., 2011). Jeppesen et al.
(1997) demonstrated that littoral environments can host more
oligotrophic taxa than the profundal and their input to deepest part
of lakes can shift the chironomid signal to more oligotrophic conditions.
Also oxygen availability in the profundal might be an
important factor. If for some reasons (e.g. cold climatic conditions)
the oxygen amount remains high in the profundal, the chironomid
species composition may stay the same (cold-demanding oligotrophic
taxa) and may not react on the increased nutrient input
(Brooks and Birks, 2001; Little et al., 2000).
5. Conclusions
From the analysis of an almost 7 m thick layer of lake sediments
from East-Central Europe (Carpathian Mts.), we inferred considerable
variation in environmental conditions and in particularly
mean July temperatures reconstructed from chironomids during
the latest Glacial and Early Holocene. In the cold YD, TJuly was on
average around 9.5 C with lowest values about 7 C. Considerable
temperature variability within the YD was inferred, with two
warmer events (12,850e12,600 and 12,200e12,400 cal yr BP)
which may have been related to instability of climate recorded in
other records (e.g. the Nordic Seas). Other chironomid records from
Europe typically do not feature within YD temperature variability of
this amplitude. However other proxies (geochemistry, diatoms)
support climatic variations, particularly hydrological changes, at
the study site during this interval and suggest that warmer conditions
were accompanied by a water table decrease. The YD was
followed by distinct warming of up to 5 C between
11,600e11,400 cal yr BP. The warmest Holocene temperatures
(15e15.5 C) were recorded in the earlier part of the Holocene
between 11,400 and 8850 cal yr BP. This agrees with some other
chironomid records from European mountain ranges (Carpathians,
central Alps) but contrasts with other reconstructions from
different parts of Europe. Possibly this difference is related to local
topographic effects on local climates or regional differences in
Holocene climate development within Europe. During the course of
the Holocene, a distinct but short-lived cooling was reconstructed
around 11,200 cal yr BP, possibly corresponding with the 11.4 ka
event and a more pronounced cooling event between 8700 and
8000 cal yr BP. The timing of the latter cooling coincides with cool
episodes recorded in other palaeoclimate records from Europe and
the North Atlantic region. The cooling is longer than reported for
the 8.2 ka event, but similar in duration as the longer lasting cooling
episode referred to as the first Holocene RCC by Mayewski et al.
(2004) that the 8.2 ka event is embedded in. At Hypkana
geochemical evidence and the composition of diatom assemblage
supported a shift to wet and cold conditions during this interval.
Author contributions
PH and LP conceived the research, LP identified pollen and green
algae, prepared samples for geochemical analyses and measured
LOI and MS, PP identified chironomids and prepared data for
quantitative reconstruction of TJuly, BCh identified diatoms and
prepared data for quantitative reconstruction of TP and TMG
P. Hajkova et al. / Quaternary Science Reviews 141 (2016) 94e111 109
carried out geochemical analyses. PH created age-depth model and
prepared samples for AMS dating. OH calculated chironomidinferred
TJuly. All authors provided ecological interpretations of
results and participated on the preparing of manuscript, which was
leaded by PH.
Acknowledgement
The research was financed by the grant project of the Czech
Science Foundation (P504/11/0429), by institutional supports and
project MUNI/A/1048/2015 of Masaryk University and Czech
Academy of Sciences (RVO 67985939). We are grateful to Ondrej
Hajek for creating the map, our colleges M. Hajek, L. Dudova, T.
Duda and S. Rezník for help by coring the profile in the field and
anonymous reviewer for valuable comments. E. Jamrichova helped
us with preparing of pollen diagram.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.quascirev.2016.04.001.
References
Alley, R.B., Agústsdottir, A.M., 2005. The 8k event: cause and consequences of a
major Holocene abrupt climate change. Quat. Sci. Rev. 24, 1123e1149.
Andersen, S.T., 1980. Influence of climatic variation on pollen season severity in
wind-pollinated trees and herbs. Grana 19, 47e52.
Bakke, J., Lie, Ø., Heegaard, E., Dokken, T., Haug, G.H., Birks, H.H., Dulski, P., Nilsen, T.,
2009. Rapid oceanic and atmospheric changes during the Younger Dryas cold
period. Nat. Geosci. 2, 202e205.
Battarbee, R.W., Jones, V.J., Flower, R.J., Cameron, N.G., Bennion, H., Carvalho, L.,
Juggins, S., 2001. Diatoms (Pp. 155e202. Springer, Netherlands.
Beug, H.J., 2004. Leitfaden der Pollenbestimmung: für Mitteleuropa und angrenzende
Gebiete. Pfeil, München.
Blockley, S.P.E., Lane, C.S., Hardiman, M., Rasmussen, S.O., Seierstand, I.K.,
Steffensen, J.P., et al., 2012. Synchronization of palaeoenvironmental records
over the last 60,000 years, and an extended INTIMATE event stratigraphy to
48,000 b2k. Quat. Sci. Rev. 36, 2e10.
Bouchez, J., Gaillardet, J., France-Lanord, C., Maurice, L., Dutra-Maia, P., 2011. Grain
size control of river suspended sediment geochemistry: clues from Amazon
River depth profiles. Geochem. Geophys. Geosyst. 12 (3), 1e24.
Brodersen, K.P., Anderson, N.J., 2002. Distribution of chironomids (Diptera) in low
arctic West Greenland lakes: trophic conditions, temperature and environmental
reconstruction. Freshw. Biol. 47, 1137e1157.
Brodersen, K.P., Lindegaard, C., 1999. Classification, assessment and trophic reconstruction
of Danish lakes using chironomids. Freshw. Biol. 42, 143e157.
Brodersen, K.P., Quinlan, R., 2006. Midges as paleoindicators of lake productivity,
eutrophication and hypolimnetic oxygen. Quat. Sci. Rev. 25, 1995e2012.
Bronk Ramsey, C., 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51,
337e360.
Brooks, S.J., Birks, H.J.B., 2001. Chironomid-inferred air temperatures from lateglacial
and Holocene sites in north-west Europe: progress and problems.
Quat. Sci. Rev. 20, 1723e1741.
Brooks, S.J., Langdon, P.G., Heiri, O., 2007. The Identification and Use of Palaearctic
Chironomidae Larvae in Palaeoecology. QRA Technical Guide No. 10. Quaternary
Research Association, London.
Buczko, K., Magyari, E.K., Braun, M., Balint, M., 2013. Diatom-inferred lateglacial and
Holocene climatic variability in the south Carpathian mountains (Romania).
Quat. Int. 271, 123e135.
Clark, P.U., Marshall, S.J., Clarke, G.K.C., Hostetler, S.W., Licciardi, J.M., Teller, J.T.,
2001. Freshwater forcing of abrupt climate change during the last glaciation.
Science 293 (5528), 283e287.
Davis, B.A.S., Brewer, S., Stevenson, A.C., Guiot, J., Contributors, D., 2003. The temperature
of Europe during the Holocene reconstructed from pollen data. Quat.
Sci. Rev. 22, 1701e1716.
Demeny, A., Czuppon, G., Siklosy, Z., Leel-}Ossy, S., Lin, K., Shen, C.C., et al., 2013. MidHolocene
climate conditions and moisture source variations based on stable H,
C and O isotope compositions of speleothems in Hungary. Quat. Int. 293,
150e156.
Dudova, L., Hajkova, P., Opravilova, V., Hajek, M., 2014. Holocene history and environmental
reconstruction of a Hercynian mire and surrounding mountain
landscape based on multiple proxies. Quat. Res. 82, 107e120.
Faegri, K., Iversen, J., 1989. Textbook of Pollen Analysis, fourth ed. Wiley, Chichester,
p. 328.
Feurdean, A., Klotz, S., Mosbrugger, V., Wohlfahrt, B., 2008a. Pollen-based quantitative
reconstructions of Holocene climate variability in NW Romania.
Palaeogeogr. Palaeoclimatol. Palaeoeol. 260, 494e504.
Feurdean, A., Klotz, S., Brewer, S., Mosbrugger, V., Tamas¸ , T., Wohlfarth, B., 2008b.
Lateglacial climate development in NW Romania e comparative results from
three quantitative pollen-based methods. Palaeogeogr. Palaeoclimatol. Palaeoeol.
265, 121e133.
Feurdean, A., Spessa, A., Magyari, E., Willis, K.J., Veres, D., Hickler, T., 2012. Trends in
biomass burning in the Carpathian region over the last 15,000 years. Quat. Sci.
Rev. 45, 111e125.
Feurdean, A., Pers¸ oiu, T., Stevens, S., Magyari, E.K., Onac, B.P., Markovic, S., et al.,
2014. Climate variability and associated vegetation response throughout Central
and Eastern Europe (CEE) between 60 and 8 ka. Quat. Sci. Rev. 106,
206e224.
Feurdean, A., Marinova, E., Nielsen, A.B., Liakka, J., Veres, D., Hutchinson, S.M., et al.,
2015. Origin of the forest steppe and exceptional grassland diversity in Transylvania
(central-eastern Europe). J. Biogeogr. 42, 951e963.
Grimm, E.C., 2011. TGView Version 1.5.12. Illinois State Museum, Research and
Collections Center, Springfield.
Grygar, T., Svetlík, I., Lisa, L., Koptíkova, L., Bajer, A., Wray, D.S., et al., 2010.
Geochemical tools for the stratigraphic correlation of floodplain deposits of the
Morava River in Straznicke Pomoraví, Czech Republic from the last millennium.
Catena 80 (2), 106e121.
Hajek, M., Dudova, L., Hajkova, P., Rolecek, J., Moutelíkova, J., Jamrichova, E., et al.,
2016. Contrasting Holocene environmental histories may explain patterns of
species richness and rarity in a Central European landscape. Quat. Sci. Rev. 133,
48e61.
Hajkova, P., Jamrichova, E., Horsak, M., Hajek, M., 2013. Holocene history of a Cladium
mariscus-dominated calcareous fen in Slovakia: vegetation stability and
landscape development. Preslia 85, 289e315.
Hajkova, P., Horsak, M., Hajek, M., Jankovska, V., Jamrichova, E., Moutelíkova, J.,
2015. Using multi-proxy palaeoecology to test a relict status of refugial populations
of calcareous-fen species in the Western Carpathians. Holocene 25,
702e715.
Heiri, O., Lotter, A.F., 2001. Effect of low count sums on quantitative environmental
reconstructions: an example using subfossil chironomids. J. Paleolimnol. 26,
343e350.
Heiri, O., Lotter, A.F., 2005. Holocene and Lateglacial summer temperature reconstruction
in the Swiss Alps based on fossil assemblages of aquatic organisms: a
review. Boreas 34, 506e516.
Heiri, O., Lotter, A.F., 2010. How does taxonomic resolution affect chironomid-based
temperature reconstruction? J. Paleolimnol. 44, 589e601.
Heiri, O., Lotter, A.F., Lemcke, G., 2001. Loss on ignition as a method for estimating
organic and carbonate content in sediments: reproducibility and comparability
of results. J. Paleolimnol. 25, 101e110.
Heiri, O., Lotter, A.F., Hausmann, S., Kienast, F., 2003. A chironomid-based Holocene
summer air temperature reconstruction from the Swiss Alps. Holocene 13,
477e484.
Heiri, O., Tinner, W., Lotter, A.F., 2004. Evidence for cooler European summers
during periods of changing meltwater flux to the North Atlantic. PNAS 101 (43),
15285e15288.
Heiri, O., Brooks, S.J., Birks, H.J.B., Lotter, A.F., 2011. A 274-lake calibration data-set
and inference model for chironomid-based summer air temperature reconstruction
in Europe. Quat. Sci. Rev. 30, 3445e3456.
Heiri, O., Brooks, S.J., Renssen, H., Bedford, A., Hazekamp, M., Ilyashuk, B., et al.,
2014a. Validation of climate model-inferred regional temperature change for
late-glacial Europe. Nat. Commun. 5, 4914.
Heiri, O., Koinig, K.A., Sp€otl, C., Barrett, S., Brauer, A., Drescher-Schneider, R., et al.,
2014b. Palaeoclimate records 60-8 ka in the Austrian and Swiss Alps and their
forelands. Quat. Sci. Rev. 106, 186e205.
Holliday, V.T., 2004. Soils in Archaeological Research. Oxford University Press,
Oxford.
Holmes, N., Langdon, P.G., Caseldine, C., Brooks, S.J., Birks, H.J.B., 2011. Merging
chironomid training sets: implications for palaeoclimate reconstructions. Quat.
Sci. Rev. 30, 2793e2804.
Hosek, J., Pokorný, P., Kubovcík, V., Horacek, I., Zackova, P., Kadlec, J., et al., 2014. Late
glacial climatic and environmental changes in eastern-central Europe: correlation
of multiple biotic and abiotic proxies from the Lake Svarcenberg, Czech
Republic. Palaeogeogr. Palaeoclimatol. Palaeoecol. 396, 155e172.
Hu, Z.-C., Gao, S., 2008. Upper crust abundances of trace elements: a revision and
update. Chem. Geol. 253, 205e221.
Ilyashuk, B., Gobet, E., Heiri, O., Lotter, A.F., van Leeuwen, J.F.N., van der Knaap, W.O.,
et al., 2009. Lateglacial environmental and climatic changes at the Maloja Pass,
Central Swiss Alps, as recorded by chironomids and pollen. Quat. Sci. Rev. 28,
1340e1353.
Ilyashuk, E.A., Koinig, K.A., Heiri, O., Ilyashuk, B.P., Psenner, R., 2011. Holocene
temperature variations at a high-altitude site in the Eastern Alps: a chironomid
record from Schwarzsee ob S€olden, Austria. Quat. Sci. Rev. 30, 176e191.
Jeppesen, E., Jensen, J.P., Søndergaard, M., Lauridsen, T., Pedersen, L.J., Jensen, L.,
1997. Top-down control in freshwater lakes: the role of nutrient state, submerged
macrophytes and water depth. Hydrobiologia 342/343, 151e164.
Jones, A.F., Macklin, M.G., Brewer, P.A., 2012. A geochemical record of flooding on
the upper River Severn, UK, during the last 3750 years. Geomorphology 179,
89e105.
Juggins, S., 2001. The European Diatom Database, User Guide,Version 1.0. http://
craticula.ncl.ac.uk/Eddi/docs/EddiGuide.pdf.
Juggins, S., 2007. C2 Version 1.5 User Guide. Software for Ecological and
P. Hajkova et al. / Quaternary Science Reviews 141 (2016) 94e111110
Palaeoecological Data Analysis and Visualisation. University of Newcastle,
Newcastle upon Tyne, UK.
Juggins, S., 2013. Quantitative reconstructions in palaeolimnology: new paradigm or
sick science? Quat. Sci. Rev. 64, 20e32.
Karasiewicz, M.T., Hulisz, P., Noryskiewicz, A.M., Krzeslak, I., Switoniak, M., 2013.
The record of hydroclimatic changes in the sediments of a kettle-hole in a
young glacial landscape (north-centralPoland). Quat. Int. 328e329, 264e276.
Kirilova, E., Heiri, O., Enters, D., Cremer, H., Lotter, A.F., Zolitschka, B., Hübener, T.,
2009. Climate-induced changes in the trophic status of a Central European lake.
J. Limnol. 68 (1), 71e82.
Komarek, J., Jankovska, V., 2001. Review of the green algal genus Pediastrum;
implication for pollen-analytical research. Bibl. Phycol. 108, 1e127.
Laskar, J., Robutel, P., Joutel, F., Gastineau, M., Correia, A.C.M., Levrard, B., 2004.
A long term numerical solution for the insolation quantities of the Earth.
Astron. Astrophy. 428, 261e285.
Lauterbach, S., Brauer, A., Andersen, N., Danielopol, D.L., Dulski, P., Hüls, M., et al.,
2011. Environmental responses to Lateglacial climatic fluctuations recorded in
the sediments of pre-Alpine Lake Mondsee (northeastern Alps). J. Quat. Sci. 26
(3), 253e267.
Little, J.L., Hall, R.I., Quinlan, R., Smol, J.P., 2000. Past trophic status and hypolimnetic
anoxia during eutrophication and remediation of Gravenhurst Bay, Ontario:
comparison of diatoms, chironomids, and historical records. Can. J. Fish. Aquat.
Sci. 57, 333e341.
Lowe, J.J., Rasmussen, S.O., Bj€orck, S., Hoek, W.Z., Steffensen, J.P., Walker, M.J.C.,
Yu, Z.C., Grp, I., 2008. Synchronisation of palaeoenvironmental events in the
North Atlantic region during the Last Termination: a revised protocol recommended
by the INTIMATE group. Quat. Sci. Rev. 27 (1e2), 6e17.
Magny, M., 2004. Holocene climate variability as reflected by mid-European lakelevel
fluctuations and its probable impact on prehistoric human settlements.
The record of human/climate interaction in lake sediments. Quat. Int. 113 (1),
65e79.
Magny, M., Begeot, C., Guiot, J., Peyron, O., 2003. Contrasting patterns of hydrological
changes in Europe in response to Holocene climate cooling phases. Quat.
Sci. Rev. 22, 1589e1596.
Magyari, E., Demeny, A., Buczko, K., Kern, Z., Vennemann, T., Forizs, I., et al., 2013.
A 13,600-year diatom oxygen isotope record from the South Carpathians
(Romania): reflection of winter conditions and possible links with North
Atlantic circulation changes. Quat. Int. 293, 136e149.
Matys Grygar, T., Elznicova, J., Babek, O., Hosek, M., Engel, Z., Kiss, T., 2014. Obtaining
isochrones from pollution signals in a fluvial sediment record: a case study in a
uranium-polluted floodplain of the Ploucnice River, Czech Republic. Appl.
Geochem. 48, 1e15.
Mauri, A., Davis, B.A.S., Collins, P.M., Kaplan, J.O., 2015. The climate of Europe during
the Holocene: a gridded pollen-based reconstruction and its multi-proxy
evaluation. Quat. Sci. Rev. 112, 109e127.
Mayewski, P.A., Rohling, E.E., Stager, J.C., Karlen, W., Maasch, K.A., Meeker, L.D., et al.,
2004. Holocene climate variability. Quat. Res. 62, 243e255.
Menviel, L., Timmermann, A., Timm, O., Mouchet, A., 2011. Deconstructing the last
Glacial Termination: the role of millennial and orbital-scale forcings. Quat. Sci.
Rev. 30 (9), 1155e1172.
Nazarova, L., Herzschuh, U., Wetterich, S., Kumke, T., Petryakova, L., 2011. Chironomid-based
inference models for estimating mean July air temperature and
water depth from lakes in Yakutia, northeastern Russia. J. Paleolimnol. 45,
57e71.
Petr, L., Zackova, P., Grygar, T.M., Pískova, A., Krízek, M., Treml, V., 2013. Súr, a former
late-glacial and Holocene lake at the westernmost margin of the Carpathians.
Preslia 85, 239e263.
Płociennik, M., Self, A., Birks, H.J.B., Brooks, S.J., 2011. Chironomidae (insecta:
Diptera) succession in _Zabieniec and its palaeo-lake (central Poland) through
the late Weichselian and Holocene. Palaeogeogr. Palaeoclimatol. Palaeoecol.
307, 150e167.
Rasmussen, P., Hede, M.U., Noe-Nygaard, N., Clarke, A.L., Vinebrooke, R.D., 2008.
Environmental response to the cold climate event 8200 years ago as recorded at
Højby Sø, Denmark. Geol. Surv. Den. Greenl. Bull. 15, 57e60.
Rasmussen, S.O., Bigler, M., Blockley, S.P., Blunier, T., Buchardt, S.L., Clausen, H.B.,
et al., 2014. A stratigraphic framework for abrupt climatic changes during the
Last Glacial period based on three synchronized Greenland ice-core records:
refining and extending the INTIMATE event stratigraphy. Quat. Sci. Rev. 106,
14e28.
Reille, M., 1992. Pollen et spores d'Europe et d'Afrique du nord. Laboratorie de
Botanique Historique et Palynologie, Marseille.
Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C., et al.,
2013. IntCal13 and Marine13 radiocarbon age calibration curves 0e50,000 years
cal. BP. Radiocarbon 55 (4), 1869e1887.
Renssen, H., Sepp€a, H., Crosta, X., Goosse, H., Roche, D.M., 2012. Global characterization
of the Holocene thermal maximum. Quat. Sci. Rev. 48, 7e19.
Renssen, H., Isarin, R.F.B., 2001. The two major warming phases of the last deglaciation
at 14.7 and 11.5 ka cal BP in Europe: climate reconstructions and AGCM
experiments. Glob. Planet. Change 30, 117e153.
Renssen, H., Sepp€a, H., Heiri, O., Roche, D.M., Goosse, H., Fichefet, T., 2009. The
spatial and temporal complexity of the Holocene thermal maximum. Nat.
Geosci. 2, 410e413.
Rosen, P., Segerstr€om, U., Eriksson, L., Renberg, I., Birks, H.J.B., 2001. Holocene climatic
change reconstructed from diatoms, chironomids, pollen and nearinfrared
spectroscopy at an alpine lake (Sjuodjijaure) in northern Sweden.
Holocene 11 (5), 551e562.
Saunders, K.M., Hodgson, D.A., McMinn, A., 2008. Quantitative relationships between
benthic diatom assemblages and water chemistry in Macquarie Island
lakes and their potential their potential to reconstruct past environmental
changes. Antarct. Sci. 21, 35e49.
Schnitchen, C., Charman, D.J., Magyari, E., Braun, M., Grigorszky, I., Tothmeresz, B.,
et al., 2006. Reconstructing hydrological variability from testate amoebae
analysis in Carpathian peatlands. J. Paleolimnol. 36, 1e17.
Schwark, L., Zink, K., Lechterbeck, J., 2002. Reconstruction of postglacial to early
Holocene vegetation history in terrestrial central Europe via cuticular lipid
biomarkers and pollen records from lake sediments. Geology 30 (5), 463e466.
Seierstad, I.K., Abbott, P.M., Bigler, M., Blunier, T., Bourne, A.J., Brook, E., et al., 2014.
Consistently dated records from the Greenland GRIP, GISP2 and NGRIP ice cores
for the past 104 ka reveal regional millennial-scale d18
O gradients with possible
Heinrich event imprint. Quat. Sci. Rev. 106, 29e46.
Sepp€a, H., Hammarlund, D., Antonsson, K., 2005. Low-frequency and highfrequency
changes in temperature and effective humidity during the Holocene
in southcentral Sweden: implications for atmospheric and oceanic forcings
of climate. Clim. Dyn. 25, 285e297.
Sepp€a, H., Birks, H.J.B., Giesecke, T., Hammarlund, D., Alenius, T., Antonsson, K., et al.,
2007. Spatial structure of the 8200 cal. yr BP event in northern Europe. Clim.
Past. 3, 225e236.
Shakesby, R.A., Smith, J.G., Matthews, J.A., Winkler, S., Dresser, P.Q., Bakke, J., et al.,
2007. Reconstruction of Holocene glacier history from distal sources: glaciofluvial
stream-bank mires and the glaciolacustrine sediment core near Sota
Scter, Breheimen, southern Norway. Holocene 17, 729e745.
Sj€ogren, P., van Leeuwen, J.F.N., van der Knaap, W.O., van der Borg, K., 2006. The
effect of climate variability on pollen productivity, AD 1975/2000, recorded in a
Sphagnum peat hummock. Holocene 16 (2), 277e286.
Stone, J.R., Westover, K.S., Cohen, A.S., 2011a. Late Pleistocene paleohydrography
and diatom paleoecology of the central basin of Lake Malawi, Africa. Palaeogeogr.
Palaeoclimatol. Palaeoecol. 303 (1e4), 51e70.
Stone, J.R., Westover, K.S., Cohen, A.S., 2011b. Late Pleistocene paleohydrography
and diatom paleoecology of the central basin of Lake Malawi, Africa. Palaeogeogr.
Palaeoclimatol. Palaeoecol. 303, 51e70.
Tamas¸ , T., Causse, C., 2001. UeTh TIMS chronology of two stalagmites from V11 Cave
(Bihor Mountains, Romania). Theor. Appl. Karstol. 13e14, 25e32.
Tatosova, J., Veselý, J., Stuchlík, E., 2006. Holocene subfossil chironomid stratigraphy
(Diptera: Chironomidae) in the sediment of Plesne lake (the Bohemian forest,
Czech Republic): palaeoenvironmental implications. Biologia 61 (Suppl. 20),
401e411.
ter Braak, C.J.F., Smilauer, P., 2002. CANOCO Reference Manual and CanoDraw for
Windows User's Guide: Software for Canonical Community Ordination (Version
4.5). Biometrics, Wageningen, p. 500.
Tinner, W., Lotter, A.F., 2001. Central European vegetation response to abrupt
climate change at 8.2 ka. Geology 29, 551e554.
Toth, M., Magyari, E.K., Brooks, S.J., Braun, M., Buzcko, K., Balint, M., Heiri, O., 2012.
A chironomid-based reconstruction of late glacial summer temperatures in the
southern Carpathians (Romania). Quat. Res. 77, 122e131.
Toth, M., Magyari, E.K., Buczko, K., Braun, M., Panagiotopoulos, K., Heiri, O., 2015.
Chironomid-inferred Holocene temperature changes in the south Carpathians
(Romania). Holocene 25, 569e582.
V€aliranta, M., Salonen, J.S., Heikill€a, M., Amon, L., Helmens, K., Klimaschewski, A.,
et al., 2014. Plant macrofossil evidence for an early onset of the Holocene
summer thermal maximum in northernmost Europe. Nat. Commun. 6, 6809.
Van der Knaap, W.O., van Leewen, J.F.N., Svitavska-Svobodova, H., Pidek, I.A.,
Kvavadze, E., Chichinadze, M., et al., 2010. Annual pollen traps reveal the
complexity of climatic control on pollen productivity in Europe and the Caucasus.
Veg. Hist. Archaeobot. 19, 285e307.
Van der Werff, A., 1955. A new method for cleaning and concentrating diatoms and
other organisms. Verhandlungen Int. Ver. für theoretische Angew. Limnol. 12,
276e277.
Van Hardenbroek, M., Heiri, O., Wilhelms, M.F., Lotter, A.F., 2011. How representative
are subfossil assemblages of Chironomidae and common benthic invertebrates for
the living fauna of Lake De Waay, the Netherlands? Aquat. Sci. 73, 247e259.
Velle, G., Brodersen, K.P., Birks, H.J.B., Willassen, E., 2010. Midges as quantitative
temperature indicator species: lessons for palaeoecology. Holocene 20 (6),
989e1002.
Von Grafenstein, U., Erlenkeuser, H., Brauer, A., Jouzel, J., Johnsen, S.J., 1999. A midEuropean
decadal isotope-climate record from 15,500 to 5000 years BP. Science
284 (5420), 1654e1657.
Walker, M., Johnsen, S., Rasmussen, S.O., Popp, T., Steffensen, J.-P., Gibbard, P., et al.,
2009. Formal definition and dating of the GSSP (Global Stratotype Section and
Point) for the base of the Holocene using the Greenland NGRIP ice core, and
selected auxiliary records. J. Quat. Sci. 24, 3e17.
Walker, M.J.C., Berkelhammer, M., Bj€orck, S., Cwynar, L.C., Fisher, D.A., Long, A.J.,
et al., 2012. Formal subdivision of the Holocene series/epoch: a discussion paper
by a working group of INTIMATE (integration of ice-core, marine and
terrestrial records) and the subcommission on quaternary stratigraphy (international
commission on stratigraphy). J. Quat. Sci. 27 (7), 649e659.
Wiederholm, T. (Ed.), 1983. Chironomidae of the Holarctic Region. Keys and Diagnoses.
Part 1- Larvae. Entomologica Scandinavica. Suppl. 19.
Wiersma, A.P., Renssen, H., 2006. Modeledata comparison for the 8.2 ka BP event:
confirmation of a forcing mechanism by catastrophic drainage of Laurentide
Lakes. Quat. Sci. Rev. 25, 63e88.
P. Hajkova et al. / Quaternary Science Reviews 141 (2016) 94e111 111