Impoverishment of recent floodplain forest mollusc fauna in the lower Ohře River (Czech Republic) as a result of prehistoric human impact

Abstract

Analyses of fossil mollusc successions have rarely been used to study the development of floodplain forests during the Holocene. The Ohře River, located in a prehistorically settled chernozem area in the Czech Republic, is partly situated in Cretaceous marlstones, yielding sediments suitable for fossilization directly in floodplain deposits. We analysed five fossil mollusc successions situated in the lower stretch of the Ohře River and compared the results with recent mollusc assemblages studied along the entire 256 km of the river. Fossil samples were composed mostly of open-country species throughout the Holocene or the whole preserved succession. Only some samples also contained woodland assemblages, but these were always greatly impoverished, with a very low frequency of strictly woodland species. Although the natural-looking appearance of the present-day floodplain forests of the lower river stretch has resulted in its being declared a nature reserve, modern floodplain forest mollusc assemblages there are also impoverished. This reduction in the distribution of strictly woodland species compared with modern assemblages in the upper stretch of the river seems to be the result of an ancient human settlement and continuous disturbances of the floodplain forest development since the Neolithic. Thus, fully developed floodplain forest assemblages occur recently only in the upper non-impacted stretch of the river. Based on the studied fossil successions we can conclude that the lower Ohře River floodplain was probably a mosaic of open and disturbed forest habitats throughout the Holocene. This area is part of a central European landscape island, where forests probably never fully developed and open patches from the early Holocene continually developed into an agricultural landscape.

Full text

Impoverishment of recent floodplain forest mollusc fauna in the lower
Ohrˇe River (Czech Republic) as a result of prehistoric human impact
LUCIE JUR
ˇIC
ˇKOVÁ, JITKA HORÁC
ˇKOVÁ, VOJEN LOŽEK AND MICHAL HORSÁK
Jurˇicˇková, L., Horácˇková, J., Ložek, V. & Horsák, M. 2013 (October): Impoverishment of recent floodplain forest
mollusc fauna in the lower Ohrˇe River (Czech Republic) as a result of prehistoric human impact. Boreas, Vol. 42,
pp. 932–946. 10.1111/bor.12006. ISSN 0300-9483.
Analyses of fossil mollusc successions have rarely been used to study the development of floodplain forests during
the Holocene. The Ohrˇe River, located in a prehistorically settled chernozem area in the Czech Republic, is partly
situated in Cretaceous marlstones, yielding sediments suitable for fossilization directly in floodplain deposits. We
analysed five fossil mollusc successions situated in the lower stretch of the Ohrˇe River and compared the results
with recent mollusc assemblages studied along the entire 256 km of the river. Fossil samples were composed
mostly of open-country species throughout the Holocene or the whole preserved succession. Only some samples
also contained woodland assemblages, but these were always greatly impoverished, with a very low frequency of
strictly woodland species. Although the natural-looking appearance of the present-day floodplain forests of the
lower river stretch has resulted in its being declared a nature reserve, modern floodplain forest mollusc assem-
blages there are also impoverished. This reduction in the distribution of strictly woodland species compared with
modern assemblages in the upper stretch of the river seems to be the result of an ancient human settlement and
continuous disturbances of the floodplain forest development since the Neolithic. Thus, fully developed floodplain
forest assemblages occur recently only in the upper non-impacted stretch of the river. Based on the studied fossil
successions we can conclude that the lower Ohrˇe River floodplain was probably a mosaic of open and disturbed
forest habitats throughout the Holocene. This area is part of a central European landscape island, where forests
probably never fully developed and open patches from the early Holocene continually developed into an agri-
cultural landscape.
Lucie Jurˇicˇková (lucie.jurickova@seznam.cz) and Jitka Horácˇková, Department of Zoology, Faculty of Science,
Charles University in Prague, Vinicˇná 7, CZ-128 44 Prague 2, Czech Republic; Vojen Ložek, Laboratory of
Environmental Geology, Institute of Geology AS CR, Rozvojová 269, CZ-165 00 Prague 6, Czech Republic; Michal
Horsák, Department of Botany and Zoology, Masaryk University, Kotlárˇská 2, CZ-611 37 Brno, Czech Republic;
received 13th February 2012, accepted 13th December 2012.
The general scenario of Holocene mollusc succession in
central Europe is one of the best known in the world
(e.g. Ložek 1964a, b, 1982a, b; Mania 1972, 1973; Fuhr-
mann 1973; Piechocki 1977; Alexandrowicz 1987;
Füköh 1993; Füköh et al. 1995; Meyrick 2001; Frank
2006), but focus on particular landscape details is still
lacking. Mollusc successions are a useful proxy for
studies of the development of such landscape details
(e.g. Ložek 1964a; Davies 2008), especially in spatially
and temporally variable environments such as flood-
plain forests. Mollusc assemblages can help us to
understand the succession as well as present-day con-
ditions of floodplain forests. The occurrence of strictly
woodland species, many of which are endangered, can
indicate sites with low human impact in the past. When
shells fossilize in river valley toe slopes immediately
next to a floodplain, their succession provides informa-
tion about the floodplain character during the geologi-
cal past. The mosaic of floodplain malacocoenoses then
reflects the mosaic of floodplain habitats, where river-
drift accumulations of shells are easily distinguishable
from slowly accumulated autochthonous material
(Ložek 1964a; Davies 2008).
The importance of floodplains for landscape func-
tioning and biodiversity has long attracted attention
(e.g. Gurnell 1997; Brown et al. 1997; Schnitzler 1994).
As floodplains have supported humans since Palaeo-
lithic times, the impact of human settlements and agri-
culture on floodplain ecosystems is also critical (e.g.
Gore & Shields 1995; Brown et al. 1997). While many
studies have been concerned with the connectivity of
recent floodplain forests in human-impacted mosaic
landscapes (e.g. Trockner et al. 1999; Ward et al. 1999),
none of them has tackled the question of their age and
historical development. Although the history of soil
erosion can inform us about agricultural impacts on the
fluvial system (Dotterweich 2008), the response of the
biota remains unclear.
Most authors have studied floodplain evolution
during the geological past using geomorphological or
pollen data (e.g. Brown 1999, 2009; Pokorný 2005;
Dotterweich 2008), with molluscs used only rarely.
However, some scattered information is available on
the mollusc succession of floodplain forests in central
Europe (Smolíková & Ložek 1978; Pišút & C
ˇejka 2002;
Brˇízová & Jurˇicˇková 2011). While alluvial malaco-
coenoses exist from NW Europe, including Poland
(Alexandrowicz 1996; Alexandrowicz & Teisseyre
1997), northern France (Limondin-Lozouet & Preece
2004; Lespez et al. 2008; Granai et al. 2011) and the
bs_bs_banner
DOI 10.1111/bor.12006 © 2013 The Authors
Boreas © 2013 The Boreas Collegium

British Isles (Preece & Day 1994; Preece & Bridgland
1999; Davies 2008), these are strongly impacted by
oceanic climate, and the woodland assemblages are dif-
ferent from those of central Europe.
We studied the succession of floodplain mollusc
assemblages in the lower Ohrˇe River, a suitable model
area for mollusc shell fossilization and for answering
general questions about Holocene floodplain develop-
ment. The study area is situated in the NW Bohemian
chernozem region and within the Bohemian Massif
province. It is one of the few areas in river floodplains
with suitable geological conditions for mollusc fossili-
zation. However, only the lower stretch of the Ohrˇe
River flows through the Cretaceous marlstones that
enable favourable conditions for shell preservation. So
far, only two mollusc successions have been analysed
from this region (Ložek 1976; Smolíková & Ložek
1978), with both lacking radiocarbon dating. Both of
these profiles clearly indicate a strong impoverishment
of floodplain forest mollusc faunas throughout the
Holocene, a phenomenon probably associated with
human impacts. The recent terrestrial mollusc faunas
of the Ohrˇe River alluvium also reflect an impoverish-
ment of forest species in the lower stretch of the river, in
contrast to the upper stretch with its association of
many strictly forest species (Horácˇková et al. 2011).
The maximum species diversity and abundance were
both found in the floodplains along the upper stretch of
the river. Differences in the historical development
caused by human impacts during the Holocene might
be one possible explanation for this sharp decrease of
mollusc diversity in the lower river stretch. It is known
that the whole area of the lower Ohrˇe River and the
western part of the C
ˇeské strˇedohorˇí Mountains have
been influenced by human activities since prehistoric
times, because of their natural resources and location in
an ancient settlement area containing a high number of
archaeological sites (Jiránˇ & Venclová 2007–2008).
Using a detailed analysis of five mollusc successions
supplemented by radiocarbon dating and comparisons
with recent assemblages of the river floodplain forests,
we attempt to (i) reconstruct the environmental devel-
opment of the alluvium during the Holocene, (ii) assess
the present condition of the floodplain forests in light of
the past state, and (iii) characterize the human impact on
floodplain forests located in an ancient agriculture area.
Methods and materials
Five mollusc successions were studied, sampled by two
techniques (see below) and excavated between 1964 and
2009. Two new sites were sampled in 2007–2009: a
toe-slope deposit at an undercut slope bank of an
ancient Ohrˇe meander situated at the floodplain edge in
the Nature Reserve Myslivna near Kostelec nad Ohrˇí
(50°23′47″N, 14°5′13″E, 160 m a.s.l.) (Fig. 1), and levee
deposits of an Ohrˇe River cut-off near the Nature
Fig. 1. Location of the study area in the NW part of the Czech Republic, with the positions of 47 samples of modern mollusc assemblages
(black dots; adopted from Horácˇková et al. 2011) and fossil successions (empty circles; A =Šebín; B =Poplze; C =Myslivna; D =Brˇežany,
E=Suchý brook).
Prehistoric human impact on floodplain forest mollusc fauna, Ohrˇe River, Czech Republic 933
BOREAS

Reserve Pístecký luh near Brˇežany nad Ohrˇí, situated
within the floodplain itself (50°25′10″N, 14°7′40″E,
160 m a.s.l.). These mollusc successions were sampled
by standard methods (Ložek 1964a),with 8-dm3dis-
crete sediment samples taken from the central part of
each macroscopically distinguishable layer exposed in
excavation pits (Figs 2, 3). Mollusc shells were
extracted from the sediments by a combination of float-
ing and sieving. After careful drying, each sample was
disaggregated in water. Floating snails were repeatedly
decanted into a 0.5-mm sieve and dried under labora-
tory conditions. Afterwards, the sediment was dried
and sorted by sieving. Shells were systematically
removed from the sediment and examined under a bin-
ocular dissecting microscope. The mollusc succession
from the site Šebín – ve Vodotocˇi (50°23′22″N,
13°59′31″E, 170 m a.s.l.) was sampled using the
(unpublished) method developed by V. Ložek in 1964.
It is a spring tufa deposit situated 3 m upslope of
the alluvium at the boundary of the floodplain and
slope forest. Previously published mollusc succession
samples from Poplze (Smolíková & Ložek 1978) were
also used in this study and supplemented with radio-
carbon dating. The Poplze site was sampled on 16th
April 1970 using the same method as described above.
The fifth succession from the floodplain deposits of
Suchý brook was sampled by a different method. The
whole deposit was sampled in 1994 using a hand-
operated pedological corer set, which provided only
slightly disturbed metre-length cores. Although this
method is much less precise than the aforementioned
sampling in open excavation pits, especially in light of
the highly reduced volume of samples, it provided a
continuous mollusc succession that approximately
reflects the palaeoenvironmental development of the
surrounding countryside as well as the stream-corridor
role of the Suchý brook. It provides the first data from
the plain extending between the C
ˇeské strˇedohorˇí
Mountains and the valley of the Ohrˇe River. This site is
situated on the western bank of the stream near the
road bridge downstream of the village Vojnice
(50°24′51″N, 13°56′15″E, 182 m a.s.l.).
Fig. 2. Myslivna – lithology of the toe-of-slope deposit and MSI histogram, ecological groups sensu Ložek (1964a). Layers: 1 =10 YR 3/2 dark
brown-greyish humic loam, crumb structure, scattered marlstone rubble (topsoil); 2 =10 YR 2/1.5 black humic loam, medium-sized marlstone
fragments (1 clast of indurate tufa); 3 =10YR 3/3 dark brown loam with scattered marlstone slabs; 4 =10 YR 4/3 brown fine-sandy less humic
loam, scattered fine rubble; 5 =10 YR 5/4 yellowish brown fine-sandy loam, rich in marlstone rubble, scattered coarser fragments; 6 =10 YR
4.5/3 brown fine-sandy loam, low in fine marlstone fragments; 7 =10 YR 5/3 pale brown fine sandy loam, fine rubble with scattered coarser
fragments; 8 =10 YR 4/2 dark greyish brown humic loam with finer rubble (<3 cm); 9 =10 YR 3/2 very dark greyish brown humic loam,
medium-sized rubble (⫾3 cm); 10 =very dark brownish grey humic loam, rather rich in coarser rubble; 11 =10 YR 3/2 very dark brown greyish
humic loam, rich in marlstone slabs, coarser fragments; w =groundwater level.
934 Lucie Jurˇicˇková et al. BOREAS

Ecological groups were used sensu Ložek (1964a)
and Alexandrowicz (1987); the nomenclature follows
Horsák et al. (2010). Mollusc diagrams (Figs 2–5)
express absolute and relative proportions of the total
number of species (MSI and MSS malacospectra) in
separate layers. Stratigraphical subdivision of the
Holocene sensu Ložek (1964a) was used.
Radiocarbon dating
Radiocarbon dating was performed in the Poznan´
Radiocarbon Laboratory, Poland, and in the Center
for Applied Isotope Studies of the University of
Georgia, USA. Mollusc shells and charcoal (Table 1)
were measured by the AMS (Accelerator Mass Spec-
trometry) method and calibrated for variable initial 14C
concentration using the OxCal 4.1 calibration
program (Bronk Ramsey 2009). Unfortunately, except
for in one layer at the Poplze site, no types of fossil
remains other than mollusc shells were at our disposal.
Although recent papers have defended using snail
shells for radiocarbon dating (Pigati et al. 2010), it is
known that shells can contain so-called dead carbon
that can lead to an overestimation of their age (Good-
friend & Stipp 1983). For this reason we controlled the
results of radiocarbon dating using relative time esti-
mations. The general stretch of central European
mollusc succession is well known (Ložek 1964a, 1972,
1982a, b), based on more than 300 Holocene succes-
sions. Using these data we compared the absolute
dating with relative dating based on a standard stretch
of mollusc succession correlated at many sites with
archaeological dated horizons, and with more than 100
profiles supplemented by absolute radiocarbon dating.
The lithology was used as another proxy to control the
undisturbed development of particular sites, which is
important in such dynamic environments as flood-
plains. The common chronostratigraphical method is
discussed in Ložek (1964a). Moreover, we also used the
shells of small species for radiocarbon dating, because
78% of them did not contain any dead carbon (Pigati
et al. 2010).
Recent assemblages and data analysis
A total of 70 mollusc species were found at the 53
woodland sites studied in all floodplain forests along
the Ohrˇe River. Recent sites were divided into the
upper and lower sections of the river, strictly separated
by the totally deforested area around the Nechranice
dam (see Fig. 1) (Horácˇková et al. 2011). The above-
mentioned published faunal data enabled us to perform
a direct comparison between fossil and recent assem-
blages. We compared numbers of strictly forest species
(sensu Ložek 1964a) recorded at each site and each
fossil layer with those recorded in the upper or lower
sections of the river. We did not use data from the
Suchý potok profile here owing to the different sam-
pling technique and smaller amount of sediment ana-
lysed per layer. Differences in the numbers of recorded
species were tested using a non-parametric Mann–
Whitney U-test and graphically shown using box-and-
Fig. 3. Brˇežany – lithology of the levee
deposits and MSI histogram, ecological
groups sensu Ložek (1964a). Layers: 1 =
dark grey humic crumbly loam (cher-
nozem A horizon); 2 =brownish grey,
medium humic, crumbly loam with small
pebbles and clasts (chernozem A/C
horizon); 3 =pale brownish grey clayey
loam, silty and admixture; 4 =light olive
brown silty sandy clayey loam; 5 =as 4,
rusty staining; 6 =dark yellowish brown,
partly stained, clay, high in silty sand.
Layers 6–3 are muddy inundation depos-
its; after a lateral erosion of their body,
they were covered by levee deposits (partly
3, 2, 1).
Prehistoric human impact on floodplain forest mollusc fauna, Ohrˇe River, Czech Republic 935
BOREAS

whisker plots. The analysis and plots were performed in
R (version 2.12.2, R Development Core Team 2011).
Below, we use the term alluvial softwood to mean
regularly inundated willow and poplar forests, and the
term hardwood to mean irregularly inundated ash and
oak forests.
Results and interpretations
Successions of fossil mollusc assemblages in the lower
Ohrˇe River floodplain
There are two components of floodplain mollusc fauna
that are important for the interpretations of fossil
records: species strictly confined to floodplain forests,
which can further be divided into hardwood and soft-
wood forest species; and woodland species sensu stricto
penetrating to the floodplain via the toe-slope ecotone,
and providing evidence of the surrounding forests.
Altogether, 58 terrestrial species were found in the
fossil records in the lower Ohrˇe River, belonging to all
ecological groups sensu Ložek (1964a), including 21
woodland species. Ten of them were woodland species
sensu stricto (Acanthinula aculeata,Aegopinella nitidula,
A. pura,Cochlodina laminata,Discus ruderatus,Macro-
gastra plicatula,Merdigera obscura,Monachoides incar-
natus,Sphyradium doliolum,Vertigo pusilla), six were
species preferring hardwood forests (Aegopinella
nitidula,Cochlodina laminata,Alinda biplicata,Vitrea
crystallina,Macrogastra ventricosa,Perforatella biden-
tata) and two preferred softwood forests (Fruticicola
fruticum,Urticicola umbrosus). The majority of
recorded fossil species belong to open-country (16
spp.), indifferent (13 spp.) and wetland (8 spp.) ecologi-
cal groups (Table 2).
The five studied successions covered the whole
Holocene, but only two of them showed a complete or
nearly complete Holocene succession (Poplze and
Suchý brook). The other three profiles contained the
young Holocene mollusc succession only. Table 1
shows radiocarbon dates for selected layers of each
profile using mollusc shells and charcoals obtained
from these layers. The oldest samples were dated back
to 8348–8192 (Suchý potok) and 7939–7793 (Poplze)
cal. a BP; however, the lower layer of the Poplze profile
was older, with its glacial age clearly indicated by the
loess-steppe species Pupilla loessica and Vallonia tenu-
ilabris. Numbers of layers, lithology, malacospectra
MSS and MSI, and species composition are given in
Fig. 4. Šebín – lithology of the tufa deposit and MSI histogram, ecological groups sensu Ložek (1964a). Layers: 1 =very dark brown loam with
admixture of tufa particles increasing towards the base; 2 =pale grey yellowish stained loose granular tufa including scattered indurations with
heavily iron-stained pores; 3 =level of indurate tufa slabs (3–10 cm) with granular tufa matrix; 4 =whitish extremely fine-grained tufa (alm),
becoming increasingly darker towards the base; 5 =dark grey humic alm with small rusty spots and diffuse greenish grey horizons; 6 =reddish
grey slightly humic clayey with minor admixture of tufa particles.
936 Lucie Jurˇicˇková et al. BOREAS

Fig. 5. Suchý brook – lithology of the drill core and MSI diagram, ecological groups sensu Ložek (1964a). The entire sedimentary sequence
represents a mixture of weathering products of Upper Creatceous semisolid marlstones and re-deposited highly humic chernozem of phaeozem
soils, whose percentage alternates in particular strata. Pale grey layers are dominated by clayey marlstone detritus; dark horizons, by humic
soil particles. A number of layers include small limonitic concretions that indicate gleyfication. Traces of iron-staining in the lowermost part
of the sequence suggest water-logged deposits and shallow water bodies. Limonitic concretions are marked with black points; accumulations
of fine marlstone rubble, with white spots.
Table 1. Results of AMS radiocarbon age determination from mollusc successions on the Ohrˇe River floodplain.
Sample name and
number of layer
Depth below
surface (cm)
Dated material Age (uncal.
a BP)
Age (cal. a AD/BC
95.4% interval)
Age (cal. a BP
95.4% interval)
Lab. code
Myslivna 11 220–240 Aegopinella sp. 2485⫾30 -773–(-418) 2723–2368 Poz-28337
Brˇežany 5 120–140 V. pulchella 1910⫾30 21–210 1929–1740 Poz-32520
Suchý potok 14 208–223 Undistinguishable
fragments
5230⫾30 -4226–(-3967) 6176–5917 UGAMS-8733
Suchý potok 27 450–460 Undistinguishable
fragments
7450⫾30 -6398–(-6242) 8348–8192 UGAMS-8734
Šebín 2 30–50 V. pygmaea 1350⫾20 645–688 1305–1262 UGAMS-8735
Šebín 5 80–90 A. minor 2290⫾20 -401–(-235) 2351–2185 UGAMS-8736
Poplze 3 210–220 Undistinguishable
fragments
1730⫾30 245–397 1705–1553 UGAMS-8748
Poplze 7 420 Charcoal 7030⫾30 -5989–(-5843) 7939–7793 UGAMS-8749
Prehistoric human impact on floodplain forest mollusc fauna, Ohrˇe River, Czech Republic 937
BOREAS

Table 2. Relative frequencies of land snail species in the upper and lower stretches of the Ohrˇe River compared with those in particular layers
of fossil sequences. Ecological characteristics: General ecological groups: A =woodland (in general); B =open country; C =woodland/open
country; D =water, wetland. Ecological groups: 1 =woodland (sensu stricto);2=woodland, partly semi-open habitats; 3 =damp woodland;
4=xeric open habitat; 5 =open habitats in general (moist meadows to steppes). Woodland/open country: 6 =predominantly dry; 7 =mesic
or various; 8 =predominantly damp; 9 =wetlands, banks.
Ecogroup Species Frequency of species
Upper stretch Lower stretch Fossil successions
A1Acanthinula aculeata 21.1 0.0 18.3
Aegopinella nitidula115.8 20.0 1.7
Aegopinella pura 42.1 0.0 6.7
Cochlodina laminata17.9 80.0 31.7
Columella aspera 5.3 0.0 0.0
Discus ruderatus 0.0 0.0 3.3
Ena montana 2.4 0.0 0.0
I. isognomostomos 1.1 0.0 0.0
Macrogastra plicatula 13.2 0.0 1.7
Merdigera obscura 0.0 6.7 3.3
Monachoides incarnatus 92.1 86.7 33.3
Oxychilus depressus 2.6 0.0 0.0
Petasina unidentata 2.6 0.0 0.0
Ruthenica filograna 2.6 0.0 0.0
Semilimax semilimax 71.1 0.0 0.0
Sphyradium doliolum 0.0 0.0 1.7
Vertigo pusilla 23.7 0.0 6.7
2Aegopinella minor 5.3 26.7 18.3
Alinda biplicata12.6 26.7 21.7
Arianta arbustorum 89.5 86.7 30.0
Cepaea hortensis 31.6 0.0 16.7
Discus rotundatus 79.0 40.0 10.0
Eucobresia diaphana 79.0 0.0 0.0
Fruticicola fruticum226.3 60.0 50.0
Helix pomatia 23.7 100.0 18.3
Vitrea crystallina181.6 6.7 6.7
3Clausilia pumila 0.0 0.0 6.7
Macrogastra ventricosa12.6 6.7 8.3
Perforatella bidentata12.6 6.7 18.3
Urticicola umbrosus2100.0 100.0 0.0
B4Cecilioides acicula 0.0 13.3 15.0
Cepaea vindobonensis 0.0 0.0 3.3
Granaria frumentum 0.0 0.0 3.3
Helicopsis striata 0.0 0.0 41.7
Chonrula tridens 0.0 0.0 40.0
Oxychilus inopinatus 0.0 0.0 3.3
Pupilla sterrii 0.0 0.0 5.0
Truncatellina claustralis 0.0 0.0 6.7
Xerolenta obvia 0.0 0.0 5.0
5Pupilla loessica 0.0 0.0 1.7
Pupilla muscorum 0.0 0.0 50.0
Truncatellina cylindrica 0.0 0.0 38.3
Vallonia costata 18.4 50.0 80.0
Vallonia pulchella 0.0 13.3 88.3
Vallonia tenuilabris 0.0 0.0 1.7
Vertigo pygmaea 7.9 0.0 42.0
C6Cochlicopa lubricella 0.0 0.0 11.7
Euomphalia strigella 0.0 0.0 56.7
7Cochlicopa lubrica 92.1 60.0 55.0
Euconulus fulvus 52.6 6.7 5.0
Oxychilus alliarius 2.6 0.0 0.0
Oxychilus cellarius 10.5 6.7 3.3
Oxychilus draparnaudi 2.6 0.0 0.0
Perpolita hammonis 97.4 60.0 6.7
Punctum pygmaeum 76.3 40.0 16.7
Trochulus hispidus 76.3 73.3 26.7
Trochulus sericeus 50.0 13.3 0.0
Vitrea contracta 0.0 6.7 0.0
Vitrina pellucida 84.2 46.7 13.3
938 Lucie Jurˇicˇková et al. BOREAS

Figs 2–5 and Tables S1–S4 (Supporting Information),
and further compared in Fig. 7.
In the Nature Reserve Myslivna, impoverished
woodland malacocoenoses occurred during the Iron
Age (2723–2368 cal. a BP) and then again recently
(Fig. 2; Table S1). The important floodplain forest
species Perforatella bidentata lived here during these
two periods, with the common woodland species Alinda
biplicata and Cochlodina laminata. The occurrence of
fully developed floodplain forests has never been docu-
mented here. Softwood species occurred sporadically at
this site during the whole succession, so we can con-
clude that softwood forests never dominated. Open-
country species had an abundance peak approximately
in the Middle Ages, when agriculture probably
restricted forests to a minimum. Most of the historical
period was characterized by predominant open-
country fauna at this site, with the dominant species
Vallonia costata and Truncatellina cylindrica. Wetland
fauna, especially Pseudotrichia rubiginosa,Vallonia
enniensis and Vertigo angustior, occurred here at the
beginning of the Subatlantic, but later became extinct.
The decreasing proportion of wetland species at the
Myslivna site demonstrates the isolation of this old
Ohrˇe meander that lost its connection with the present-
day river channel. However, this succession has been
continuously impacted by humans from the beginning
of the Subatlantic, as is directly documented by prehis-
toric potsherds in layers 10 and 11 and by scattered
pottery fragments throughout the sequence. Ironically,
the site is currently a natural reserve and appears to be
one of the best-preserved floodplain forests along the
entire Ohrˇe River.
Open-country and woodland/open-country species
dominated during the whole late Holocene succession
at Brˇežany (Fig. 7; Table S2) and Poplze (Smolíková &
Ložek 1978); thus, floodplain forests have been totally
lacking at these two sites during the last two thousand
years. The mollusc succession from the levee deposits
at Brˇežany did not even yield any woodland species
sensu stricto (Fig. 3; Table S2). The low proportion of
freshwater species documents a slow accumulation of
these aggradations; initial soil formed only in the upper
layers (Fig. 3). This may suggest that this deposition
area was not affected by floods. If we consider the
well-sorted fossiliferous sediment – silty sand and clay
– we may conclude that the depositional process was
quiet, far from the active stream, with dispersed shells
of meadow communities. Layer 3 represents the tran-
sition to terrestrial conditions, which culminates in the
chernozem-like soil formation in 2 (A/C soil horizon)
and 1 (A soil horizon). Both layers correspond to a
much longer time-span, because soil formation is a
much longer process than sedimentation. This is also
reflected by the accumulation of shells, particularly in
layer 1, that reflect semi-xerophilous grassland. Open-
country species were dominant throughout the whole
succession. Steppe elements, such as Chondrula tridens
and Helicopsis striata, probably lived on adjacent
marlstone slopes. These slopes were almost certainly
treeless, as the surroundings had been settled since the
beginning of the era (1929–1740 cal. a BP). Terrestrial
environments are also reflected by the appearance of
the modern immigrant Cecilioides acicula. Of particu-
lar interest is the record of Xerolenta obvia in the low-
ermost layer 6, which is the earliest radiocarbon-dated
find of this modern immigrant in central Europe. As
far as we know, the Brˇežany site is the first levee
deposit ever studied from a mollusc succession point of
view.
Table 2. Continued
Ecogroup Species Frequency of species
Upper stretch Lower stretch Fossil successions
8Carychium tridentatum 71.1 26.7 25.0
Columella edentula 2.6 0.0 6.7
Perpolita petronella 5.3 0.0 0.0
Succinella oblonga 18.4 13.3 21.7
Vertigo angustior 0.0 0.0 33.7
Vertigo substriata 2.6 0.0 0.0
D9Carychium minimum 71.1 26.7 16.7
Euconulus praticola22.6 0.0 0.0
Oxyloma elegans20.0 0.0 ?
Pseudotrichia rubiginosa22.6 0.0 16.7
Succinea putris286.8 46.7 ?
Vallonia enniensis 0.0 0.0 31.7
Vertigo antivertigo 0.0 0.0 16.7
Vertigo moulinsiana 0.0 0 1.7
Zonitoides nitidus276.3 40.0 13.3
1Hardwood species.
2Softwood species.
Prehistoric human impact on floodplain forest mollusc fauna, Ohrˇe River, Czech Republic 939
BOREAS

The mollusc succession of Šebín (Fig. 4, Table S3)
differs from the others because of its situation at the
boundary of the floodplain and slope of the river valley.
Common woodland species occurred here continuously
throughout the Subatlantic (2351–2205 cal. a BP),
except for a weakly dryer period indicated in layer 3
(possibly at the beginning of the era). Notable is the
only occurrence of Sphyradium doliolum close to the
Ohrˇe floodplain in the lowermost layer. It documents
the occurrence of forest close to the river, although the
woodland fauna was also reduced. Wetlands did not
form here until the Subrecent (layer 1). The occurrence
of a spring is reflected by the higher proportion of
wetland species. However, the rare species Vertigo
moulinsiana,V. angustior and V. antivertigo that had
occurred in this period became extinct recently at the
site.
At first sight, the mollusc succession of Suchý potok
(Fig. 5; Table S4) seems very monotonous and thus
hardly comparable with the standard mollusc succes-
sions of the central European Holocene (Ložek 1964a,
1982b), being almost completely dominated by open-
country or catholic species. Of particular importance is
again the absence of close-woodland species that are
characteristic for particular phases of the Holocene. In
this context, it is worth noting the occurrence of Fru-
ticicola fruticum in the lower part of the sedimentary
sequence (layers 23–29), which may reflect parkland
environments, whereas in the upper part a similar situ-
ation is indicated by Euomphalia strigella (layers
10–15). The sedimentary succession of this site can be
divided into two main mollusc zones: layers 1–19 and
20–30. The lower group of beds (layers 20–30) is char-
acterized by a rather high number of aquatic taxa, with
high numbers of Gyraulus laevis,Pisidium casertanum
and in particular Bithynia tentaculata, implying the
presence of some standing water bodies. It should be
stressed that these species are also associated with
several terrestrial species such as Fruticicola fruticum,
Vallonia costata and V. pulchella in high amounts,
which may reflect a patchy environment composed of
various habitats (water bodies, open habitats with
short vegetation and tall herb vegetation) within the
floodplain. Of particular malacostratigraphic impor-
tance are the records of Discus ruderatus in layers 28
and 29, which suggest that this horizon is of early
Holocene age, as is supported by the 14C-dating of layer
27 (8348–8192 cal. a BP).
The upper group (layers 1–19) is dominated by ter-
restrial species, while aquatic taxa show a decrease both
in species and in individuals, with the exception of
Galba truncatula that culminates in the topsoil. The
records of two fragments of Perforatella bidentata in
layers 1 and 2 probably reflect the occurrence of river-
side woodland. Pseudotrichia rubiginosa appears only
in the upper group, where it occurs in considerable
numbers. This is also true of Vallonia enniensis,
V. pulchella and V. costata and partly also of Pupilla
muscorum, which were found in much higher numbers
than in the lower layers (20–30). The uppermost layers
(1–5) are characterized by a general impoverishment of
the snail fauna. Thus, the upper complex largely corre-
sponds to the onset of agriculture (layer 17: 6176–5917
cal. a BP) and a subsequent open landscape. Both char-
acteristic steppe elements – Helicopsis striata and Chon-
drula tridens – lived on the valley side and indicate open
patches in this landscape since the onset of the
Holocene. The succession at Suchý brook differs from
the other floodplain successions in the absence of
common and generalist species, such as Euconulus
fulvus,Perpolita hammonis,Punctum pygmaeum,
Vitrina pellucida,Carychium minimum,C. tridentatum
and Zonitoides nitidus. This also applies to modern
immigrants, such as Xerolenta obvia or Cecilioides
acicula, although they are currently common in the
surrounding area. These peculiar differences from the
regional developmental pattern considerably hinder
the exact placement of the Suchý brook succession into
the standard Holocene biochronological scale (Ložek
1964a, 1982a). Concerning the function of the Suchý
brook valley as a stream corridor, there are no traces of
the migration of molluscs from the C
ˇeské strˇedohorˇí
Mountains to the valley of the Ohrˇe River.
On the basis of these results we conclude that the
lower Ohrˇe River floodplain was probably a mosaic of
open sites, wetlands and forest habitats with reduced
faunal complexity for the majority of the Holocene.
Discussion
The choice of palaeoecological proxy and material
for dating
We attempted to study the Holocene dynamics and
succession of floodplain habitats based on mollusc
assemblages, using five sites with succession deposits.
Because floodplains are often spatially limited habitats,
we chose molluscs as a suitable proxy capable of reflect-
ing changes on fine spatial scales (Ložek 1964a; Davies
2008). In contrast, the frequently used pollen analyses
provide information on larger spatial scales (e.g. Firbas
1949, 1952). We can clearly demonstrate the differences
between mollusc and pollen evidence from the results of
pollen analysis at the site of Zahájí (Pokorný 2005),
which is situated at a small brook discharging into the
Ohrˇe River, only 3 km away from the Myslivna site.
Although there has been forest here continuously since
the Subboreal period (Eneolith), the pollen diagram
shows important differences between the occurrence of
trees that (i) never grow in alluvium (such as pine,
birch, beech and even yew) and (ii) are characteristic of
floodplain forests (such as ash and alder). The first
group occurs continuously, while the second group
peaks in sub-recent times. Thus, pollen evidence reflects
940 Lucie Jurˇicˇková et al. BOREAS

the character of the whole surrounding landscape,
while mollusc successions show the situation directly in
the alluvium.
Although mollusc shells have been regarded as less
suitable for radiocarbon dating than charcoals or other
plant remains (e.g. Goodfriend & Stipp 1983), our data
do not support this assertion. As mentioned above, we
controlled the absolute dating based on mollusc shells
by comparison with the standard central European
mollusc succession (Ložek 1964a, 1982a). From this
point of view, the dating based on shells was quite
successful and we did not find any obvious discrepan-
cies. We have also had the same experience using
Holocene snail shells for AMS dating of more than 100
samples (L. Jurˇicˇková, unpubl. data).
The possible origin of floodplains and the lithology of
their deposits
The origin and development of floodplains are subjects
still under discussion (e.g. Johnson et al. 1995; Brown
et al. 1997), and more evidence is necessary to under-
stand human impacts and climatic changes in flood-
plain habitat dynamics over the Holocene. Published
data from southern England based on mollusc succes-
sions did not show any evidence of either flooding or
alluviation during the late Mesolithic and early Neo-
lithic periods (Davies 2008). The Holocene alluviation
probably started as late as in the post-Neolithic period
(Evans 1993). Similar evidence of changes in river sedi-
mentology after the Neolithic settlement have been
widely confirmed (e.g. Lespez et al. 2008), and this
probably also applies to the floodplain development in
the lower Ohrˇe River. We have no evidence of the
existence of floodplain forest during the Climatic
Optimum or in earlier times; however, data from the
early Holocene are still rare. The malacospectra assem-
bled from Ložek’s data in the Poplze profile
(Smolíková & Ložek 1978) start in the last glacial loess
steppe with the index species Pupilla loessica and Val-
lonia tenuilabris (Fig. 7). Later, shells occurred very
rarely because fossilization was inhibited during the
development of chernozem soils. It is possible that this
trend is common in the Ohrˇe River floodplain, as also
indicated by the succession at Suchý brook. A similar
hiatus in mollusc succession was recorded during the
Preboreal and Atlantic periods in the Seine valley
(Granai et al. 2011). However, more evidence is needed,
as the other studied profiles in the Ohrˇe River covered
the young Holocene period only. In these younger
deposits, however, sedimentation was mostly continu-
ous, without notable interruptions caused by floods or
erosion. We did not find any flood accumulation in the
lithology sequences (Figs 2–5). In contrast to Davies
(2008), who found that alluvial deposits in southern
England consisted mainly of aquatic snail species, our
assemblages were composed of terrestrial autoch-
thonous species with a surprisingly low freshwater
component (Tables S1–S4).
Present condition of the floodplain forests in light of
their past state
Unfortunately, there are no suitable sites with pre-
served fossil mollusc successions in the upper stretch of
the Ohrˇe River, because of the lack of calcareous
deposits in this area. Thus, we cannot provide a clear
link between the modern mollusc fauna in the upper
stretch of the river and its Holocene development.
However, the frequency and modern distribution of
strictly woodland snail species along the whole Ohrˇe
River (Horácˇková et al. 2011) correspond well with the
evidence of all analysed successions in the lower section
of the river (Table 2). The number of woodland species
in samples was the highest in the upper stretch of the
river, in contrast to both modern and fossil assem-
blages of the lower section (Fig. 6, Table 2). This dif-
ference is particularly noteworthy, as modern
environmental conditions are more favourable for rich
woodland assemblages in the lower stretch of the river.
Likewise, the modern extent of floodplain forests in the
lower Ohrˇe River is notably larger than in the upper
part. This was partly caused by agricultural changes
after the Second World War, when small-scale agricul-
ture was replaced by more intensive and large-scale
operations that neglected some more difficult-to-
cultivate areas such as in the floodplain. One of our
studied profiles, Myslivna, is actually situated in the
Nature Reserve of a preserved floodplain forest,
although the fossil evidence clearly shows that present
conditions at this site are of recent origin. Despite this,
both the whole species diversity and the absence of
Fig. 6. Comparison between numbers of strictly woodland snail
species recorded in recent samples in the upper and lower stretches of
the Ohrˇe River and those found in individual layers of fossil sedi-
ments. All three groups are significantly different from each other at
p=0.05. The central line of the box represents the median, margins of
the box are the interquartile distance, the non-outlier range is 1.5
times the interquartile distances at each fence, and the circle repre-
sents one outlier.
Prehistoric human impact on floodplain forest mollusc fauna, Ohrˇe River, Czech Republic 941
BOREAS

Fig. 7. Comparison of MSS malacospectra of all mollusc successions from the lower Ohrˇe alluvium. Chronozones were used sensu Ložek (1964a). On the left side of the spectra are numbers
of layers, with those dated absolutely marked with a circle; on the right side are numbers of specimens. This figure is available in colour at http://www.boreas.dk.
942 Lucie Jurˇicˇková et al. BOREAS

some anthropophobic woodland species (e.g. Isogno-
mostoma isognomostomos,Ruthenica filograna and
Petasina unidentata) indicate a clear impoverishment of
modern assemblages in the lower section of the river
(Horácˇková et al. 2011). Only half of all woodland
species recorded in the Ohrˇe River alluvium spread to
the lower part of the floodplain. The remarkable con-
trast between the upper and lower Ohrˇe River faunas
may have been further strengthened by the recently
built Nechranice Dam and by extensive deforestation
along the middle stretch of the river (Fig. 1).
The historical impoverishment of woodland faunas
in the lower stretch of the river seems to have been a
continuous process, as indicated by some evidence in
the fossil successions. Some woodland species that were
rare in the lower Ohrˇe in the past currently survive only
in the upper stretch of the river (Columella edentula and
Macrogastra plicatula) or are extinct (Sphyradium doli-
olum). However, some of them also managed to survive
in the lower stretch (Macrogastra ventricosa,Merdigera
obscura and Pseudotrichia rubiginosa), but were very
rare in the past. In contrast, the woodland species
Urticicola umbrosus occurring in all recent assemblages
seems to be a recent immigrant, because it is absent
from all fossil records. This species, which favours tall
herb vegetation, could indicate an increase of nutrients
in the alluvium, resulting in the development of a
denser herb layer.
Human impact on the development of Ohrˇe River
floodplain forests
The reason for the historical impoverishment in the
richness of woodland species is probably also linked to
the location of the area within the intra-Bohemian cher-
nozem area. In this region the first Neolithic farmers
settled in persisting relicts of the early Holocene forest
steppe, converting it to open agricultural landscape and
pastureland, and consequently hindering the expansion
of woodland malacocoenoses and further supporting
the survival or spread of open-country taxa.
The majority of alluvial profiles of southern England
cover later prehistoric and historic periods, when no
woodland species were recorded and open-country
assemblages of different moisture preferences occurred
(Davies 2008). Even when older molluscan successions
were analysed and showed some woodland species
occurring during the Climatic Optimum (e.g. Evans
1993; Preece & Day 1994; Preece & Bridgland 1999),
the woodland species diversity was very low in com-
parison with that in central Europe. The situation in
northern France is similar (Limondin-Lozouet &
Preece 2004). Some strictly forest species occurred there
during the Climatic Optimum (e.g. Acicula fusca,
Cochlodina laminata,D. ruderatus,Helicodonta obvo-
luta,Spermodea lamellata,Vertigo pusilla,V. alpestris)
and then went extinct or survived only in well-
preserved sites. This is most likely linked with the
deforestation of great areas of western Europe that had
occurred by Greek and Roman Antiquity. While in vast
areas of western and central Europe the period of
deforestation was followed by a period of natural refor-
estation re-colonization (Kaplan et al. 2009), there are
some areas where reforestation never fully came about.
One such region is chernozem area in central Bohemia
(e.g. Ložek 1964a, b; Pokorný 2005). Anthropogenic
deforestation in this region started with Neolithic colo-
nization, as supported by numerous archaeological
investigations from the wider region (Jiránˇ & Venclová
2007–2008), although not exactly at the sites of our
mollusc successions. Neolithic and Eneolithic settle-
ments have been documented from the lower Ohrˇe allu-
vium in Brˇezno, ~20 km west of our sites (Pleinerová &
Pavlu˚ 1979; Pleinerová 1990). Although the importance
of human impacts is clearly supported, we also have to
consider the role of climate on the development of
floodplain habitats. The dry sub-continental climate of
this part of central Bohemia makes the study region
sensitive to deforestation and favours the persistence of
open vegetation, and thus the reforestation obvious in
many European regions was partly hindered here. This
finding corresponds with palaeobotanical data from a
nearby site, Zahájí (Pokorný 2005). The evidence there
shows the strong impact of humans since the Neolithic
period, with maximum settlement dated to 4300 cal. a
BP, although the presence of forest habitats was also
confirmed later in this pollen profile.
The lower Ohrˇe as part of a broader area of reduced
woodland fauna in central Europe
Woodland fauna in NW Europe has a reduced charac-
ter. Boycott (1934) even considered no British land
snail as strictly woodland! In contrast, strictly wood-
land species have occurred in central Europe since the
Atlantic (e.g. Ložek 1964a, b; Alexandrowicz 1987).
Nevertheless, an island of reduced but woodland
mollusc fauna has been documented from the broader
area of the C
ˇeské strˇedohorˇí Mountains and some
neighbouring areas in Bohemia (Ložek 1963, 1964a, b,
2005). The profile below Kuzov near Trˇebívlice (Ložek
1976) is situated at the edge of the Ohrˇe lowland cher-
nozem area at the foot of the C
ˇeské strˇedohorˇí Moun-
tains, not far from our study area (Fig. 1). Floodplain
forest species were documented here only during the
Atlantic/Epiatlantic, followed by open-country fauna.
The evidence of human impacts on the woodland fauna
of this area during the Holocene is also demonstrated
by comparison with results from the fully developed
woodland Middle Pleistocene interglacial fauna
(Kovanda 2005). Several strictly woodland species (e.g.
Sphyradium doliolum and Ruthenica filograna) occurred
commonly during the Middle Pleistocene interglacial
but have only rarely been found in the Holocene suc-
Prehistoric human impact on floodplain forest mollusc fauna, Ohrˇe River, Czech Republic 943
BOREAS

cessions. The others went extinct in the region and are
fully absent in the Holocene (Aegopis verticillus,Daude-
bardia rufa,Discus perspectivus,D. ruderatus,Helico-
donta obvoluta,Platyla polita and Vitrea subrimata). On
the basis of these data we infer that the lower Ohrˇe
floodplain is part of a broader area of reduced wood-
land fauna, whose extent corresponds with the area of
ancient settlement (Jiránˇ & Venclová 2007–2008).
Conclusions
We chose molluscs as suitable proxy for the study of
Holocene succession in the spatially limited Ohrˇe River
floodplain. Based on five mollusc successions situated
immediately in the floodplain we reconstructed the
development of malacocoenoses. While fully developed
woodland assemblages occurred in other central Euro-
pean lowlands since the Climatic Optimum, the lower
Ohrˇe alluvium probably consisted of a mosaic of pre-
viously open-country habitats with wetland and impov-
erished woodland habitats throughout the whole
Holocene. The local nature reserves thus protect flood-
plain forests that have originated only recently. The
lower Ohrˇe floodplain is part of a broader prehistoric
settlement area, where forest probably never fully
developed and open patches from the early Holocene
continually developed into an agricultural landscape.
Despite this, a few strictly woodland snail species per-
sisted in this island of impoverished woodland fauna, in
contrast to the situation in NW Europe where
impoverishment was probably caused mainly by cli-
matic conditions.
Acknowledgements. – The research reported here was supported by
the Grant Agency of the Czech Republic P504/10/0688, project no.
40007 of the Grant Agency of Charles University, and project
MSMT 0021620828. We are grateful to Ivan Pavlu˚ from the Institute
of Archeology of the Academy of Science, Czech Republic, Prague
for assessments of potsherds from the Myslivna site, to Mrs Mária
Fapšová for technical collaboration, and to David Hardekopf for
English revision. Our thanks also go to Nicole Limondine-Lozouet
and an anonymous reviewer for their valuable comments, which
helped improve the manuscript.
References
Alexandrowicz, S. W. 1987: Malacological analyses in Quaternary
research. Kwartalnik AGH, Geologia 13, 1–240.
Alexandrowicz, S. W. 1996: Malacofauna of late Holocene overbank
deposits in southern Poland. Bulletin of the Polish Academy of
Sciences, Earth Sciences 44, 235–249.
Alexandrowicz, S. W. & Teisseyre, A. K. 1997: Holocene overbank
deposits and mollusc fauna, Oława River valley (Sudety Foothills,
western Poland). Bulletin of the Polish Academy of Sciences, Earth
Sciences 45, 47–58.
Brˇízová, E. & Jurˇicˇková, L. 2011: Can canopy forests survive agri-
cultural colonization in the Polabí lowland (Czech Republic)? Bul-
letin of Geoscience 86, 283–300.
Boycott, A. E. 1934: The Habitats of Land Mollusca in Britain.
Journal of Ecology 22, 1–38.
Bronk Ramsey, C. 2009: Bayesian analysis of radiocarbon dates.
Radiocarbon 51, 337–360.
Brown, A. G. 1999: Biodiversity and pollen analysis: modern pollen
studies and the recent history of floodplain woodland in S. W.
Ireland. Journal of Biogeography 26, 19–32.
Brown, A. G. 2009: Colluvial and alluvial response to land use
change in Midland England: an integrated geoarcheological
approach. Geomorphology 108, 92–106.
Brown, A. G., Harper, D. & Peterken, G. F. 1997: European flood-
plain forests: structure, functioning and management. Global
Ecology and Biogeography Letters 6, 169–178.
Davies, P. 2008: Snails: Archaeology and Landscape Change, 208 pp.
Oxbow Books, Oxford.
Dotterweich, M. 2008: The history of soil erosion and fluvial deposits
in small catchments of central Europe: deciphering the long term
interaction between humans and the environment – a review. Geo-
morphology 101, 192–208.
Evans, J. G. 1993: The influence of human communities on the
English chalklands from the Mesolithic to the Iron Age: the mol-
luscan evidence. In Chambers, F. M. (ed.): Climate Change and
Human Impact on the Landscape, 147–156. Chapman and Hall,
London.
Firbas, F. 1949: Spät- und Nacheiszeitliche Waldgeschichte Mitteleu-
ropas Nördlich der Alpen. I. 480 pp. Gustav Fischer Verlag, Jena.
Firbas, F. 1952: Spät- und Nacheiszeitliche Waldgeschichte Mitteleu-
ropas Nördlich der Alpen. II. 256 pp. Gustav Fischer Verlag, Jena.
Frank, C. 2006: Plio-Pleistozäne und Holozäne Mollusken Osterreichs
1, 2. 860 pp. Verlag der Osterrachischen Akademie der Wissen-
schaften, Vienna.
Fuhrmann, R. 1973: Die Spätweichselglaziale und Holozäne Mol-
luskenfauna Mittel- und u. Westsachsens. Freiberger Forschung-
shefte C 278, 121 pp.
Füköh, L. 1993: Holocene malacological assemblages in Hungary.
Skripta Geologova, Special Issue 2, 121–125.
Füköh, L., Krolopp, E. & Sümegi, P. 1995: Quaternary malacos-
tratigraphy in Hungary. Malacological Newsletter,Supplement 1,
218 pp.
Goodfriend, G. A. & Stipp, J. J. 1983: Limestone and the problem of
radiocarbon dating of land-snail carbonate. Geology 11/10, 575–
577.
Gore, J. A. & Shields, F. D. 1995: Can large rivers be restored?
BioScience 45, 142–152.
Granai, S., Limondin-Lozouet, N. & Chausee, C. 2011: Evolution
paléoenvironnementale de la vallée de la Seine r Paris (France)
d’après l’étude des malacofaunes. Quaternaire 22, 327–344.
Gurnell, A. 1997: The hydrological and geomorphological signifi-
cance of forested floodplains. Global Ecology and Biogeography
Letters 6, 219–229.
Horácˇková, J., Ložek, V. & Jurˇicˇková, L. 2011: Nivní malakofauna
rˇeky Ohrˇe – její minulost a soucˇasnost. Malacologica Bohemoslo-
vaca 10, 51–64.
Horsák, M., Jurˇicˇková, L., Beran, L., C
ˇejka, T. & Dvorˇák, L. 2010:
Annotated list of mollusc species recorded outdoors in the Czech
and Slovak Republics. Malacologica Bohemoslovaca,Supplement
1, 1–37 (in Czech).
Jiránˇ , L. & Venclová, N. (eds) 2007–2008: Archeology of the Prehis-
toric Czech 1–8. Archeologický ústav AV C
ˇR v.v.i., Prague.
Johnson, B. L., Richordson, W. B. & Naimo, T. 1995: Past, present,
and future concept in large river ecology. Bioscience 45/3, 134–
141.
Kaplan, J. D., Krumhardt, K. M. & Zimmermann, N. 2009: The
prehistoric and preindustrial deforestation of Europe. Quaternary
Science Reviews 28, 3016–3034.
Kovanda, J. 2005: Pátek, Levousy and Chlumcˇany – important fos-
siliferous mid-Pleistocene localities on the right bank of the Ohrˇe
River between Louny and Libochovice. Malacologica Bohemoslo-
vaca 3, 149–172.
Lespez, L., Clet-Pellerin, M., Limondin-Lozouet, N., Pastre, J.-F.,
Fontugne, M. & Marcigny, C. 2008: Fluvial system evolution and
environmental changes during the Holocene in the Mue valley
(Western France). Geomorphology 98, 55–70.
Limondin-Lozouet, N. & Preece, R. C. 2004: Molluscan successions
from the Holocene tufa of St-Germain-le-Vasson in Normandy,
France. Journal of Quaternary Science 19, 55–71.
944 Lucie Jurˇicˇková et al. BOREAS

Ložek, V. 1963: Der altholozäne Dauch von Mrsklesy im Böh-
mischen Mittelgebirge. Antropozoikum 1, 63–74.
Ložek, V. 1964a: Quartärmollusken der Tschechoslowakei. 374 pp.
Czech Academy of Science, Prague.
Ložek, V. 1964b: Biostratigraphic research of the important Quater-
nary profiles of the C
ˇSSR. Zprávy o Geologických Výzkumech 1,
348–350.
Ložek, V. 1972: Holocene Interglacial in Central Europe and its land
snails. Quaternary Research 2/3, 327–334.
Ložek, V. 1976: Molluscan fauna of the alluvium under Kuzov near
Trˇebívlice and its importance for the knowledge of nature devel-
opment of the Milešovské strˇedohorˇí Mts. Severocˇeskou Prˇírodou
7, 1–14.
Ložek, V. 1982a: Contribution of malacology to the chronological
subdivision of the Central European Holocene. Striae 16, 84–87.
Ložek, V. 1982b: Faunengeschichtliche Grundlinien zur spät– und
nacheiszeitlichen Entwicklung der Molluskenbestände in Mitteleu-
ropa. Rozpravy C
ˇeskoslovenské Akademie Veˇd, Rada Matematick-
ýchaPrˇírodních Veˇd 92, 1–106.
Ložek, V. 2005: Holocene malacofauna from Risuty and its signifi-
cance for the environmental history of the north-west Bohemian
forest steppe area. Severocˇeskou Prˇírodou 36/37, 11–22.
Mania, D. 1972: Zur spät- und nacheiszeitlichen Landschaftsges-
chichte des mittelen Elb-Saalegebietes. Hallenser Jahrbuch für Mit-
teldeutsche Erdeschichte 11, 7–36.
Mania, D. 1973: Paläoökologie, Faunenentwicklung und Stratigra-
phie des Eiszeialters im mittleren Elbe-Saalegebiet auf Grund von
Molleskengesellschaften. Geologie, Beiheft 78/79, 175 pp.
Meyrick, R. A. 2001: The development of terrestrial mollusk faunas
in the Rheinland region (Western Germany and Luxembourg)
during the Lateglacial and Holocene. Quaternary Science Review
20, 1667–1675.
Piechocki, A. 1977: The Late Pleistocene and Holocene Mollusca of
the Kun´ ów region (N-E margin of the S
´wie˛tokrzyskie Mts. Folia
Quarternaria 49, 23–36.
Pigati, J. S., Rech, J. A. & Nekola, J. C. 2010: Radiocarbon dating of
small terrestrial gastropod shell in North America. Quaternary
Geochronology 5, 519–532.
Pišút, P. & C
ˇejka, T. 2002: Historical development of floodplain site
using Mollusca and cartographic evidence. Ekológia 21, 378–396.
Pleinerová, I. 1990: Dva eneolitické dlouhé domy u Brˇezna. Památky
Archeologické 81, 266–274.
Pleinerová, I. & Pavlu˚ , I. 1979: Brˇezno. Osada z Mladší doby Kamenné
v Severozápadních C
ˇechách. 139 pp. Severocˇeské nakladatelství,
Ústí nad Labem.
Pokorný, P. 2005: Role of man in development of Holocene vegeta-
tion in Central Bohemia. Preslia 77, 113–127.
Preece, R. C. & Bridgland, D. R. 1999: Holywell Coombe, Folke-
stone: a 13,000 year history of an English chalkland valley. Qua-
ternary Science Reviews 18, 1075–1125.
Preece, R. C. & Day, S. P. 1994: Comparison of post-glacial
molluscan and vegetational successions from a radiocarbon-dated
tufa sequence in Oxfordshire. Journal of Biogeography 21, 463–
478.
R Development Core Team 2011: R:A Language and Environment for
Statistical Computing. R Foundation for Statistical Computing,
Vienna.
Schnitzler, A. 1994: European alluvial hardwood forests of large
floodplains. Journal of Biogeography 14, 97–117.
Smolíková, L. & Ložek, V. 1978: Die nacheiszeitlichen Bodenabfol-
gen von Poplze und Šteˇtí als Beleg der Boden- und Landschaftsen-
twicklung im böhmischen Tschernosemgebiet. In Fink, G. J. (ed.):
Beiträge zur Quartär und Landschaftsforschung Festschrift zum 60,
531–549. Verlag F. Hirt, Vienna.
Trockner, K., Schiemer, F., Baumgartner, G. K., Kum, G., Eeigand
Zweimüller, I. & Ward, J. V. 1999: The Danube restoration
project: species diversity patterns across connectivity gradients in
the floodplain system. Regulated Rivers: Research and Manage-
ment 15, 245–258.
Ward, J. V., Trockner, K. & Schiemer, F. 1999: Biodiversity of
floodplain river ecosystems: ecotones and connectivity. Regulated
Rivers: Research and Management 15, 125–139.
Supporting Information
Additional Supporting Information may be found in
the online version of this article:
Table S1. Mollusc assemblages of separate layers of an
ancient Ohrˇe meander in the floodplain forest of the
Nature Reserve Myslivna near Kostelec nad Ohrˇí.
Layer numbers relate to Fig. 2. Ecological character-
istics: General ecological groups: A =woodland (in
general); B =open country; C =woodland/open
country; D =water, wetland. Ecological groups: 1 =
woodland (sensu stricto); 2 =woodland, partly semi-
open habitats; 3 =damp woodland; 4 =xeric open
habitat; 5 =open habitats in general (moist meadows
to steppes). Woodland/open country: 6 =predomi-
nantly dry; 7 =mesic or various; 8 =predominantly
damp; 9 =wetlands, banks; 10 =aquatic habitats.
Biostratigraphic characteristics: (+)=local or occa-
sional loess species; ! =species of warm phases; (!) =
eurythermic species of warm phases; !! =index species
of warm phases; G =species surviving glacial out of
loess zone; (G) =ditto as relics. Presence in layers: 1 =
number of individuals; 1? =only an approximate
determination. Fragments of Oxyloma elegans and
Succinea putris are hardly distinguishable.
Table S2. Mollusc assemblages of separate layers of
levee deposits of an Ohrˇe River cut-off near Brˇežany
nad Ohrˇí. Layer numbers according to Fig. 3. Eco-
logical characteristics: General ecological groups: A
=woodland (in general); B =open country; C =
woodland/open country; D =water, wetland. Eco-
logical groups: 1 =woodland (sensu stricto);2=
woodland, partly semi-open habitats; 4 =xeric open
habitat; 5 =open habitats in general (moist meadows
to steppes). Woodland/open country: 7 =mesic or
various; 8 =predominantly damp; 9 =wetlands,
banks; 10 =aquatic habitats. Biostratigraphic char-
acteristics: (+)=local or occasional loess species; ! =
species of warm phases; (!) =eurythermic species of
warm phases; !! =index species of warm phases; G =
species surviving glacial out of loess zone; (G) =ditto
as relics. Presence in layers: 1 =number of individu-
als; 1? =only an approximate determination.
Table S3. Mollusc assemblages of separate layers of
the tufa fen Ve vodotocˇi near Šebín. Layer numbers
according to Fig. 4. Ecological characteristics:
General ecological groups: A =woodland (in
general); B =open country; C =woodland/open
country; D =water, wetland. Ecological groups: 1 =
woodland (sensu stricto); 2 =woodland, partly semi-
open habitats; 3 =damp woodland; 4 =xeric open
habitat; 5 =open habitats in general (moist meadows
to steppes). Woodland/open country: 6 =predomi-
nantly dry; 7 =mesic or various; 8 =predominantly
damp; 9 =wetlands, banks; 10 =aquatic habitats.
Biostratigraphic characteristics: (+)=local or occa-
Prehistoric human impact on floodplain forest mollusc fauna, Ohrˇe River, Czech Republic 945
BOREAS

sional loess species; ! =species of warm phases; (!) =
eurythermic species of warm phases; !! =index
species of warm phases; G =species surviving glacial
out of loess zone; (G) =ditto as relics. Presence in
layers: 1 =number of individuals; 1? =only an
approximate determination.
Table S4. Mollusc assemblages of separate layers from
the core to the deposits of the Suchý potok brook
near Vojnice. The depth of particular layers (in cm):
1 (10–29), 2 (20–37), 3 (37–55), 4 (55–70), 5 (70–85),
6 (80–100), 7 (100–123), 8 (123–138), 9 (138–151), 10
(149–165), 11 (165–175), 12 (175–195), 13 (195–205),
14/15 (208–223/223–230), 16 (230–250), 17 (250–
267), 18 (267–290), 19 (290–323), 20 (323–356), 21
(356–370), 22/23 (370–380/380–400), 24/25 (400–410/
410–435), 26 (435–450), 27 (450–460), 28 (460–476),
29/30 (476–485/485–493). Ecological characteristics:
General ecological groups: A =woodland (in
general); B =open country; C =woodland/open
country; D =water, wetland. Ecological groups: 1 =
woodland (sensu stricto); 2 =woodland, partly semi-
open habitats; 3 =damp woodland; 4 =xeric open
habitat; 5 =open habitats in general (moist meadows
to steppes). Woodland/open country: 7 =mesic or
various; 8 =predominantly damp; 9 =wetlands,
banks; 10 =aquatic habitats. Biostratigraphic char-
acteristics: (+)=local or occasional loess species; ! =
species of warm phases; (!) =eurythermic species of
warm phases; !! =index species of warm phases; G =
species surviving glacial out of loess zone; (G) =ditto
as relics. Presence in layers: 1 =number of individu-
als; 1? =only an approximate determination. Frag-
ments of Oxyloma elegans and Succinea putris are
hardly distinguishable.
946 Lucie Jurˇicˇková et al. BOREAS