Mid-Holocene bottleneck for central European dry grasslands: Did steppe survive the forest optimum in northern Bohemia,Czech Republic?

Abstract

Revisiting the classical Gradmann’s ‘steppe theory’ for central Europe, we examine whether the early Holocene steppe habitats survived the critical period of maximum Holocene afforestation: the mid-Holocene bottleneck. Despite the undisputable fact that afforestation was a dominant ecological factor in this period, our parallel analyses of pollen and molluscs from sedimentary sequences discovered in the dry lowland area of northern Bohemia, Czech Republic (Zahájí and Suchý potok sites, lower Ohře area) provide strong evidence for uninterrupted local occurrence of steppe grasslands throughout the Holocene. At the onset of the Neolithic agriculture, this area was covered by forest-steppe. Analogously to the present forest-steppe landscapes of eastern Europe and south-western Siberia, dry areas of northern Bohemia were dominated by open-canopy pine–birch forests that enabled continuous survival of many light-demanding plant species from the late Glacial and early Holocene to the Neolithic. Later on, anthropogenic deforestation and livestock grazing created a semi-natural steppe. Our data suggest that this secondary steppe can be viewed as a direct continuation of the late Pleistocene and early Holocene natural steppe rather than a purely cultural steppe developed only after deforestation of a continuously forested mid-Holocene landscape by humans. At the same time, we provide evidence supporting Gradmann’s ‘steppe theory’, assuming that in central Europe, Neolithic farming started in those areas that were not completely forested but contained remnants of natural steppes. This finding has important implications for the interpretation of present biodiversity patterns in central Europe.

Keywords

Czech Republic dry grasslands forest-steppe Holocene refugia molluscs pollen analysis steppe

Introduction

Euro-Siberian steppe and forest-steppe is zonal vegetation occurring in a broad belt from the coastal lowlands near the north-western Black Sea coast through southern Ukraine and Russia to the northern foothills of the Altai–Sayan Mountains in southern Siberia (Bohn et al., 2000–2003; Dengler et al., 2014; Lavrenko and Karamysheva, 1993; Walter, 1974; Werger and van Staalduinen, 2012). It develops in areas with dry continental climate characterized by sharp differences between summer and winter temperatures. Low precipitation and winter frosts limit the growth of trees and shrubs, thus providing habitat for herbs and particularly grasses. Under such conditions, litter decomposition and cation leaching are weak, resulting in the development of base-rich chernozem soils (Eckmeier et al., 2007; Walter and Breckle, 1991).

Westernmost outposts of Euro-Siberian steppe grasslands occur in rain-shadow areas of central Europe, such as dry valleys of the Central Alps (Braun-Blanquet, 1961; Schwabe and Kratochwil, 2004), the upper Rhine valley in south-western Germany and adjacent Alsace (Korneck, 1974), the lower Saale valley in central Germany (Mahn, 1965), northern and central Bohemia (Chytrý, 2012; Chytrý et al., 2007), low-altitude areas of southern Moravia and Lower Austria (Chytrý, 2012; Chytrý et al., 2007; Willner et al., 2013) and the Pannonian Basin in Hungary and adjacent hilly areas (Dúbravková et al., 2010; Illyés and Bölöni, 2007; Zólyomi and Fekete, 1994). Similar to the eastern zonal steppe and forest-steppe landscapes, most of these central European areas are characterized by annual precipitation lower than 500 mm and a continental temperature regime with warm summers and cold winters, although climatic extremes are less pronounced than in eastern Europe and Siberia. Plant species composition of the central European steppe grasslands is very similar to that of eastern zonal steppe, which is reflected in phytosociological classification of vegetation by combining them to a single class, Festuco-Brometea (Royer, 1991).

Steppe was widespread across northern Eurasia during the cold phases of the Pleistocene, supported by dry climate with continental temperature regime and widespread occurrence of calcareous soils (Frenzel, 1964, 1987). There is no doubt that the modern zonal steppes of Ukraine, Russia, Kazakhstan, Mongolia and northern China represent direct continuation of the Pleistocene glacial steppe, although the current steppe zone is markedly restricted in comparison with the presumed extent of the Pleistocene glacial steppe (Tarasov et al., 2001). The Pleistocene steppe also covered large areas of central Europe (Frenzel et al., 1992; Kuneš et al., 2008a; Magyari et al., 2014); however, early- to mid-Holocene climatic amelioration in this region caused the spread of forests and associated retreat of steppe.

The fundamental question related to the history of central European steppe grasslands is whether they are historically continuous since the Pleistocene, i.e., whether they survived the period between the beginning of the presumably humid Atlantic period (traditionally called the Holocene climatic optimum; e.g. Kalis et al., 2003), which supported the spread of closed forest, and the onset of Neolithic agriculture at about 7500 cal. BP (Bogaard, 2004), which maintained the landscape open through human and livestock activities. We will call this critical period the mid-Holocene bottleneck. The classical ‘steppe theory’ (Steppenheidetheorie; Gradmann, 1906, 1933; further advocated, e.g., by Firbas, 1949) proposed that the first Neolithic farmers in central Europe settled in areas with dry climate and calcareous soils, which were still open and covered with steppe, therefore easily exploitable for agriculture. Humans kept these areas open and further extended them at the expense of forest, creating the cultural steppe. Gradmann’s steppe theory has been much debated and frequently opposed by alternative views emphasizing that especially in the more oceanic western part of central Europe (Germany and Bohemia), the mid-Holocene climate was rather humid for several centuries before the introduction of agriculture, thus supporting the spread of continuous forest (for review see Dúbravková and Hajnalová, 2012; Hein, 2010; Hejcman et al., 2013; Poschlod et al., 2009). The debate about historical openness of western and central European landscapes has recently emphasized also the role of large herbivores (Vera, 2000; but see alternative views by Sandom et al., 2014; Birks, 2005 and Mitchell, 2005) and Mesolithic hunter-gatherers (Kuneš et al., 2008b).

Holocene continuity of central European steppe grasslands is partly supported by the high similarity of their species composition to eastern zonal steppes and forest-steppes. However, steppe species in central European landscapes may also be considered as new immigrants from the east or southeast to the landscapes that were forested in the early Atlantic period and subsequently cleared by the Neolithic farmers. The contrasting views of continuous persistence of steppe species since the Pleistocene versus their post-glacial immigration to central Europe can partly be resolved based on genetic analyses of steppe plants that would include both the populations from contiguous eastern steppes and isolated central European steppes. However, such studies are still rare (Durka et al., 2012; Wagner et al., 2011, 2012; Wróblewska, 2008) and do not allow inferences on the history of central European populations. Phylogeographic analysis of common hamster (Cricetus cricetus), a rodent typical of continental steppe, suggested that Pannonian populations have persisted in central Europe over the last 50,000 years, while German populations probably immigrated from the east (Neumann et al., 2005). Pollen-analytical studies from the mid-Holocene sediments seem to contradict the hypothesis of the continuity of central European steppe grasslands (Krippel, 1982; Lang, 1994; Litt, 1992; Rybníček and Rybníčková, 1994; Rybníčková and Rybníček, 1972). However, pollen analysis must be interpreted with caution considering its poor taxonomic and spatial resolution and strong overrepresentation of trees in pollen diagrams. Using a model simulation approach, Sugita et al. (1999) have shown that arboreal to non-arboreal pollen ratio (AP/NAP ratio; the parameter often used in such studies) in pollen diagrams gives only a rough first approximation of the landscape openness. There is a substantially higher pollen production by wind-pollinated trees, especially pine (Pinus spp.) or birch (Betula spp.), than by other tree species or of open grasslands. Another problem is the general absence of sediments that would preserve pollen grains in the key regions of central Europe, i.e., in the warm and dry lowlands. Most suitable sediments for pollen preservation are those from peat bogs and lakes, which are nearly absent in forest-steppe areas. Although some sedimentary sequences from the mid-Holocene can be found in them, they are often rich in calcium-carbonate, which disables good preservation of pollen grains. In such cases, valuable palaeoecological data can be obtained by investigations of subfossil assemblages of molluscs, which reflect both vegetation and soil conditions. Mollusc shells are well preserved in calcareous sediments and can be reliably identified to species level, thus providing very precise fine-scale information about local palaeoenvironments, including detection of small grassland patches in forested landscapes (Ložek, 1964b; Svenning, 2002).

In this paper, we combine evidence from two nearby sedimentary records recently obtained from a dry area of northern Bohemia, which is one of the key regions of central European forest-steppe but so far lacking continuous Holocene fossil records. One of the new records contains detailed pollen analysis and the other mollusc analysis. Both of these records are unique in that they cover the critical period of the presumed mid-Holocene bottleneck for steppe vegetation, therefore providing answers for the traditional but still unresolved questions: (1) has steppe vegetation existed continuously since the late Glacial or early Holocene in the dry lowland region of Bohemia, or was it overgrown by forest in the mid-Holocene? and (2) Did Neolithic farmers settle in a partly open landscape, or did they have to clear forests to obtain areas suitable for agriculture? We believe that answering these questions is crucial for correct interpretation of the present biogeographical patterns observed in central Europe.

Materials and methods

The Zahájí pollen site and its analyses

The valley-bottom fen Zahájí is situated in the lowland area along the lower Ohře river in northern Bohemia, Czech Republic (50°22′44″N, 14°07′04″E, 200 m.a.s.l.; Figure 1). The bedrock of the shallow valley of a small brook is formed of Cretaceous calcareous sandstones and marlstones, while the surrounding plateau is covered by loess. The present climate is relatively warm, dry and continental (mean annual temperature 8.3°C, annual precipitation 496 mm), as a result of low elevation (190 m.a.s.l.) and position under an orographic rain shadow of the Ore Mountains. The site’s surroundings have been densely and continuously populated at least since the Neolithic (Kuna, 1998).

Figure 1.

Location of the study area in the Czech Republic and of the study sites in the archaeological map for the Neolithic linear pottery culture (LBK). Symbol descriptions: white arrow – Zahájí pollen site; grey arrow – Suchý potok mollusc site; black points – important archaeological sites of the LBK culture according to the Czech Archaeological Database (data kindly provided by I Pavlů; Pavlů, 2007).

Pollen analyses from the Zahájí site have already been published (Albert and Pokorný, 2012; Bieniek and Pokorný, 2005; Pokorný et al., 2010a), but new data were obtained after much deeper drilling in October 2012. Unlike the previous record, the new one covers also the key period of the early Holocene and the Mesolithic/Neolithic transition, and we focus on this new record in this paper. This new sedimentary sequence was sampled by a piston corer (50 mm in diameter) in a deep, mostly organic deposit (eutrophic peat with chemical precipitates and admixture of fine silt in some levels) formed near a strong artesian spring in the upper part of the valley. Immediately after coring, the visual stratigraphy of the sedimentary sequence was recorded. Preservation of pollen grains in the entire core was excellent due to low pH and the content of alum (determined as K₂Fe₅(Fe, Al)₄(SO₄)₁₂.32–36 H₂O), a compound that prevents biological degradation of organic materials (Pokorný et al., 2010a). This unusual chemical composition is also the reason why the deposit is exceptionally thick and most likely without any stratigraphic hiatuses, providing a fossil record that is absolutely unique for the dry lowlands of central Europe.

Samples for pollen analyses were treated by standard acetolysis (Moore et al., 1991). In each sample at least 700 (but usually over 1000) pollen grains were counted. For pollen identification, a reference collection and standard keys (Beug, 2004; Fægri and Iversen, 1989; Moore et al., 1991) were used. Pollen taxonomy and nomenclature follows the Czech Quaternary Palynological Database – PALYCZ (Kuneš et al., 2009).

In the pollen diagram (Figure 2), percentage values were calculated on the basis of the total pollen sum (AP + NAP = 100%). Excluded from this sum were aquatic and marshland taxa (incl. Cyperaceae) because of their local indicative value. The pollen diagram (Figure 2) was drawn using the TILIA and TGView 2.0 programs (Grimm, 1990). Local Pollen-Analytical Zones (LPAZ) were distinguished based on the expert judgement, considering both presence and abundance of taxa. Moreover, CONISS cluster analysis was applied to the dataset and the resulting dendrogram was plotted with the pollen diagram. Both the subjective and the objective criteria for diagram zonation can be thus compared.

Figure 2.

Pollen diagram from the site Zahájí, Czech Republic. White color in lithology column represents purely organic sediment (peat).

The Suchý potok mollusc site and its analyses

The Suchý potok site (50°24′51″N, 13°56′15″E, 182 m.a.s.l.; Figure 1) is situated ca. 13 km from the Zahájí pollen site, in a loamy floodplain of a small brook within a shallow valley downstream of the village Vojnice. The site is similar to the former in its topography, bedrock, climate and long history of human impact.

The mollusc succession was sampled using a hand-operated auger soil corer set. Mollusc shells were extracted from the sediments by a combination of floating and sieving. After drying, each sample was disaggregated in water and if necessary, also in hydrogen peroxide. Floating shells and their fragments were repeatedly decanted into a 0.5 mm fraction of the sieve and dried under laboratory conditions. Afterwards, the fraction above the sieve was dried and sorted by sieving into fractions based on the type of resulting sediment. Shells were removed from the sediment and identified under a dissection microscope. The mollusc diagram (Figure 3) was again made in TILIA and TGView 2.0 programs (Grimm, 1990).

Figure 3.

Mollusk diagram from the site Suchý potok, Czech Republic.

Results

Pollen diagram (Figure 2) and mollusc diagram (Figure 3) show a selection of taxa that are relevant for the assessment of the mid-Holocene bottleneck for steppe vegetation. Absolute chronology of both the studied sections is based on radiocarbon dating (Table 1). Hereafter, we refer exclusively to the calibrated radiocarbon ages (cal. BP).

Table 1.

Results of AMS radiocarbon age determination for both the studied profiles. Calibration results from OxCal 4.4 (Ramsey, 2009).

Sample name, depth	Material dated	Measured radiocarbon age	Age (cal. BP; 95.4% interval)	Lab. code
Zahájí, 97–99 cm	Alnus wood (twigs)	665 ± 38 BP	679–554	Erl-3007
Zahájí, 184–186 cm	Wood (twigs)	1005 ± 39 BP	1046–795	Erl-3008
Zahájí, 297–299 cm	Aboveground parts of cf. Molinia	2830 ± 44 BP	3076–2801	Erl-3012
Zahájí, 430–435 cm	Panicum miliaceum seeds	3140 ± 40 BP	3449–3265	Poz-29572
Zahájí, 497–499 cm	Aboveground parts of cf. Molinia	4134 ± v50 BP	4827–4529	Erl-3013
Zahájí, 554–555 cm	Aboveground parts of cf. Molinia and wood (twigs)	4788 ± 49 BP	5606–5329	Erl-3011
Zahájí, 647–749 cm	Betula seeds and mosses	6180 ± 40 BP	7236–6952	Poz-51054
Zahájí, 669–671 cm	Piece of bark and charcoal	7130 ± 40 BP	8021–7866	Poz-51055
Zahájí, 783–785 cm	Betula seeds	8340 ± 50 BP	9482–9145	Poz-51057
Zahájí, 825–827 cm	Betula – bark from a trunk	8360 ± 50 BP	9493–9255	Poz-51058
Suchý potok, layer no. 14, 208–223 cm	Indeterminable fragments of snail shells	5230 ± 30	6176–5917	UGAMS-8733
Suchý potok, layer no. 27, 450–460 cm	Indeterminable fragments of snail shells	7450 ± 30	8348–8192	UGAMS-8734

The pollen record

The Zahájí pollen record covers the last approximately 9000 years. The following biostratigraphic zones (LPAZ) were distinguished (Figure 2).

LPAZ Z1a

Pinus sylvestris-type fluctuates from 30% to 80%. There are fluctuating percentages of broadleaf trees, most common being Corylus avellana, Quercus, Ulmus and Tilia. Birch (Betula sect. Albae) proportion is low. Juniperus, a light-demanding shrub, reaches its Holocene maximum in this zone. Grassland indicators such as Gramineae, Artemisia and Thalictrum dominate among herbs. Most of the other herbal types encountered (Anthemis-type, Compositae Subfam. Cichorioideae, Cruciferae, Rubiaceae) probably also indicate open habitats. The rather high and continuous curve of microscopic charcoal indicates frequent fires. The upper limit of this zone is marked by the fall of Pinus sylvestris-type, the rise in the Betula sect. Albae and a prominent peak in Corylus avellana, known as a relatively light-demanding shrub (Ellenberg, 1988; Vera, 2000).

LPAZ Z1b

With respect to the mid-Holocene bottleneck, the most important feature of this zone is a high sum of AP, fluctuating slightly around 95%. This is the maximum reached over the entire record. However, it does not necessarily imply a total afforestation of the area, because the dominant contribution to the AP sum is from Pinus sylvestris-type and Betula sect. Albae, the taxa with exceptionally high pollen productivity. Each of these two pollen types separately fluctuates from 40% to 60%. Among herbal pollen types, Gramineae decrease considerably, while others stay stable, or decrease only slightly (Artemisia). Finds of Pulsatilla pollen deserve special attention as they clearly indicate presence of open dry grasslands.

LPAZ Z1c

At the lower boundary of this zone, the Pinus sylvestris-type curve decreases sharply. Betula sect. Albae curve strongly fluctuates between 20% and 60%. All broadleaf trees and shrubs (Corylus avellana, Quercus, Tilia, Ulmus, Fraxinus excelsior, Acer) present before and Picea abies reach their absolute maxima in this zone. This is especially true for the relatively light-demanding oak species (Quercus), which dominates this broadleaf assemblage. Beech (Fagus sylvatica), a highly competitive shading tree, appears at the beginning of this zone and gradually increases. In all these indicators of broadleaf forest spread, indicators of open habitats stay high or even increase (e.g. Artemisia, Gramineae and Rubiaceae). Clear indication of human impact starts already at the onset of this zone, and its intensity gradually increases towards its upper limit. Anthropogenic pollen indicators point to the spread of pastures or grazed forests (Calluna vulgaris, Plantago media, Rumex acetosa-type). Pollen grains of cereals (Triticum-type) indicate field cultivation. As the sediment accumulation rate is slow in the lower part of this zone, precise dating of the start of human impact on vegetation is difficult. Nevertheless, the beginning of this zone is roughly synchronous with the beginnings of Neolithic agriculture in central Europe (radiocarbon dating at 648 cm level is 7236–6952 cal. BP; see Table 1).

LPAZ Z2a

In this zone, human impact on vegetation strongly increases, as indicated by the rise in most herbs, especially grazing indicators. Grazing activity around the investigated site most probably resulted in the expansion of secondary dry grasslands and subcontinental heathlands (the latter indicated by the absolute maximum of Calluna vulgaris). For the first time since the start of the pollen record, NAP percentages exceed 30%. Pinus sylvestris-type stays low, while Betula sect. Albae falls sharply. At the same time, Fagus sylvatica becomes fully established and strong expansion of silver fir (Abies alba) and hornbeam (Carpinus betulus) takes place. The start of this expansion is synchronous with the increase in human impact. The decrease in Tilia, Ulmus and Fraxinus excelsior may be caused by competition from these new arboreal immigrants.

LPAZ Z2b

All the processes described for the previous LPAZ are reaching its peak in this zone. The largest difference is the qualitative change in human impact. Indicators of heathlands are much less frequent, but grazing must have been still important (see Plantago lanceolata and Rumex acetosa-type). It seems plausible that heathlands were replaced by grazed grasslands. A large amount of cereal pollen (Triticum-type, Secale cereale) points to an increased importance of arable agriculture at this time. Strong fire activity, indicated by microscopic charcoal particles and Pteridium aquilinum (a fire indicator) can be ascribed to human influence. Radiocarbon measurement at 430–435 cm level (3449–3265 cal. BP) assigns this pronounced anthropogenic phase to the Bronze Age.

LPAZ Z2c

This long and rather complicated period covers the entire post-Bronze Age prehistory. It is characterized by continuous, but fluctuating human impact, which leaves a strong signal fluctuation in most arboreal taxa. However, details of this development are beyond the scope of this article.

LPAZ Z3a

This transitional zone is characterized by a short period of abandonment (see the Betula sect. Albae peak that is synchronous with the decrease in most indicators of human impact) followed by a new increase in agricultural impact.

LPAZ Z3b

This period corresponds to the Holocene maximum of human impact dated to the High Middle Ages and Modern Period. The landscape was almost absolutely deforested and grasslands reached their maximum extent. A single pollen spectrum that was sampled at the surface of the deposit (= 0 cm) reflects the recent state of agricultural abandonment and, in suitable soil and micro-climatic settings, initial spontaneous reforestation.

The lithology record

Four lithological units were distinguished during the core inspection (see ‘lithology’ column in the pollen diagram; Figure 2): (1) fen peat with chemical precipitates, but without any detectable allochthonous mineral material; (2) silt with organic admixture; (3) sand and (4) gravel. The middle section of the core between 175 and 647 cm consists of solely autochthonous organic sediment (peat), while the overlying and underlying strata contain variable amount of allochthonous inorganic compounds.

The mollusc record

Unlike in the pollen record, we could not apply a detailed biostratigraphic zonation to the malacostratigraphic record (Figure 3), because it is rather monotonous, nearly completely dominated by indicators of open habitats. It is characterized by the general absence of woodland species that are typical of most mid-Holocene mollusc records from the prehistorical settlement areas of central Europe (Horáčková et al., 2014), but also by the absence of common generalist species and modern immigrants.

Fruticicola fruticum, Vallonia costata and V. pulchella are present over the entire sequence, reflecting a patchy environment composed of various open habitats, including short and tall-herb vegetation. There is nearly continuous presence of characteristic steppe species Chondrula tridens and Helicopsis striata, which probably lived on the valley slopes.

Increased abundance of Fruticicola fruticum in the lower part of the sedimentary sequence (layers 23–29) and that of Euomphalia strigella in the upper part (layers 10–15) reflect the environments with patches of trees and shrubs. The records of Discus ruderatus near the bottom of the core (layers 28 and 29) are of particular stratigraphic importance, suggesting the early Holocene age of these layers, which is also supported by radiocarbon dating of layer 27 to 8348–8192 cal. BP. The upper part of the record corresponds to the period of agricultural impact (layer 17: 6176–5917 cal. BP). In this period, steppe indicators Chondrula tridens and Helicopsis striata and other open-landscape species such as Vallonia pulchella and V. costata increase in abundance together with the first appearance of non-native species of open habitats (Pupilla muscorum and Vertigo pygmaea). These changes indicate the extension of the area of secondary steppe. According to radiocarbon dating, this phase corresponds to the later phase of the Zahájí Z1c pollen zone, where increased human impact is also recorded. At the same time, appearance of Pseudotrichia rubiginosa, Vallonia enniensis and Vertigo angustior indicates the formation of an open wetland. The uppermost layers (1–5) are characterized by a general impoverishment of the snail fauna that is attributable to increased human impact.

Discussion

Holocene continuity of the central European steppe

There is a general consensus, supported by extensive palaeoecological evidence, that large areas of central Europe were densely forested in the Holocene climatic optimum (Lang, 1994; Rybníčková and Rybníček, 1996; see also literature cited in the ‘Introduction’ section). However, Zoller and Haas (1995) emphasize that this forest could have existed as a mosaic of regeneration states. Moreover, regional macroclimatic differences may have caused a considerable contrast between the precipitation-rich areas with continuous forest cover and drier lowlands with open woodland and scattered patches of steppe grasslands. Based on the extensive sampling of fossil mollusc sequences in the Czech Republic and Slovakia, Ložek (1982, 2005, 2007), Juřičková et al. (2013a, 2013b) and Horáčková et al. (2014) suggested Holocene continuity of dry grasslands in some lowland regions with Neolithic settlements. Mania (1995) arrived at the same conclusion based on the analysis of fossil mollusc records from the dry lowland area along the lower Saale River in central Germany. However, this hypothesis, consistent with Gradmann’s (1906, 1933) ‘steppe theory’, has never been supported by fossil pollen data (see, e.g. Litt, 1992; Rybníčková and Rybníček, 1996).

In this study, we used two new, unique sedimentary records of the mid-Holocene environment in the dry lowlands of northern Bohemia, which have a potential to give a credible answer to the old question of the mid-Holocene bottleneck (as we call this putative process here) and historical continuity of steppe vegetation, and possibly reconcile the contrasting palynological and malacological perspectives. The surroundings of both study sites lack rocky slopes that could act as azonal refuges of steppe biota over the Holocene. Therefore, we can interpret our findings as evidence for the development of ‘average’, zonal landscapes in dry lowland areas of central Europe.

In general, we conclude that patches of steppe grassland persisted in the study region throughout the Holocene (since ca. 9000 cal. yr BP until the present). The rather monotonous mollusc record from the Suchý potok site and also some other mollusc successions from the adjacent regions (lower Ohře floodplain and České středohoří Mts; Juřičková et al., 2013a, 2013b) characterized by the absence of forest species are hardly comparable with the standard central European mollusc successions recorded in precipitation-richer areas, which contain many forest species (Juřičková et al., 2014; Ložek, 1964a, 1982). Holocene continuity of dry grassland vegetation was already supported, although with less quantitative data, by malacological analyses from other lowland sites in the Czech Republic where current annual precipitation is below 500 mm (Ložek, 1964a, 1964b, 2005, 2007) and from the dry lowland area of central Germany (Mania, 1995 Some of the studied sedimentary sequences were characterized by constant presence of steppe molluscs, particularly Chondrula tridens and Helicopsis striata, which are considered unable to survive in closed forests (but see Horsák et al., 2009 for somewhat contrasting evidence on C. tridens).

Our study, for the first time, presents strong palynological evidence related to the mid-Holocene bottleneck that is in agreement with the malacological evidence discussed above. We demonstrate that a landscape mosaic with patches of open pine–birch forests and steppe grasslands existed in the dry region of northern Bohemia until the Neolithic, and the spread of mesophilous broadleaf trees was synchronous with the onset of Neolithic agriculture. Thus, the mid-Holocene bottleneck for steppe vegetation was probably counteracted by increasing human impact. This is in agreement with similar evidence from the more continental regions in the southeast, namely from the Carpathian (Pannonian) Basin, where similar steppe grasslands with many continental and sub-Mediterranean species probably also existed continuously throughout the Holocene (Magyari et al., 2010).

Pre-Neolithic pine–birch forest-steppe: A Holocene refugium?

In the Zahájí pollen dataset, the first direct evidence of agriculture is provided by the appearance of cereal (Triticum-type) pollen. According to radiocarbon dating, this corresponds to the onset of early Neolithic LBK (linear pottery) culture. We can assume anthropogenic deforestation and grazing since that time, both resulting in the promotion and area extension of dry grasslands. For the assessment of the mid-Holocene bottleneck for steppe vegetation, the key period is the one that immediately precedes this first (LBK) agricultural impact, i.e., the period covered by the Z1b zone of the pollen diagram (Figure 2).

The entire Z1b zone is characterized by fluctuating, but generally high curves of Pinus sylvestris-type and Betula sect. Albae. All the other trees are rarer than in the younger zone (Z1c). At the same time, many herbaceous taxa are present and some of them (e.g. Artemisia) have continuous pollen curves. We interpret this finding as the evidence for forest-steppe vegetation, with steppe species growing partly in open grassland patches and partly in the herb layer of open pine or birch forests.

The sedimentary record from Zahájí can be used as another proxy, which also suggests landscape openness. Under local topographic setting, the allochthonous inorganic compounds must have been deposited after being eroded in the catchment. This cycle of erosion and deposition was best possible in open or semi-open surface, where the energy of water was not hampered by dense forest cover. The presence of silt and sand in the lowermost ca. 200 cm of the deposit (from the basal layers to the level of 647 cm, i.e., to the very start of the Z1c pollen zone) thus can be interpreted as indirect evidence of landscape openness.

Currently, pine–birch forest-steppe with local occurrence of broadleaf deciduous trees such as Quercus, Acer, Tilia and Ulmus (similar to that interpreted for the Zahájí Z1c zone) occurs in eastern Europe up to the foothills of the Southern Urals (Chytrý et al., 2010), and its variant (without the above-mentioned broadleaf trees) extends further to the east, forming a distinct vegetation belt in south-western Siberia (Walter, 1974). Detailed comparative studies of the woodland component of the Siberian forest-steppe (Ermakov et al., 2000) revealed that the herb layer of these open-canopy forests contains numerous species which are also typical of open steppe and mesic to wet grasslands, many of them shared with current steppe grasslands of central Europe (like Pulsatilla that is recorded over the critical period at Zahájí pollen site). The herb layer of some of the light pine–birch woodlands of the southern Siberian forest-steppe is also extremely rich in herbaceous species, much richer than any current forests in central Europe (Chytrý et al., 2012). Such forests would act as refuges for many grassland species even though the entire landscape was covered by forests. However, in the Southern Ural foothills and south-western Siberia, species-rich pine–birch forests rarely form a continuous cover over the entire landscapes. They are typically associated with patches of open steppe grasslands which occur on south-facing slopes, elevated landforms or on recently burned sites. In drier areas, steppe occurs on the flatland, whereas pine, birch or aspen forests are confined to topographically wetter depressions (Ermakov et al., 2000; Walter, 1974). These modern analogues, involving the same dominant tree species and herbaceous flora that is very similar to that of central Europe, support the idea that the pre-Neolithic landscape around the Zahájí pollen site was a species-rich forest-steppe with patches of steppe grasslands.

The history of chernozem and phaeozem soils dominating the central European forest-steppe regions can possibly be explained in an analogous way. There are contradictory opinions concerning their genesis and historical interpretation. They are mostly considered zonal steppe soils which developed already before the forest expansion, and the historical continuity of open land is supposed as a precondition of their persistence (Ložek, 1973). However, several central European authors (e.g. Eckmeier et al., 2007; Vysloužilová et al., 2014) suggested that these soils may have survived some periods of Holocene afforestation. We believe that our findings may help reconcile these contrasting views, considering that the formation of these soils is a contextual and dynamic process, not only a passive response to steppe conditions. The persistence of chernozem under the long-term influence of forest is hardly conceivable if we consider closed broadleaf forest, since it causes their irreversible degradation. However, semi-open forest-steppe landscape with light forests that were directly replaced by Neolithic cultural landscape may have been a suitable environment for formation, preservation and further development of the chernozem and phaeozem soils.

Early human impact: A counterbalance of the forest spread?

Central European palaeoecological data show that in all altitudinal zones, early Holocene pine and birch woodlands were replaced by more competitive trees towards the Holocene climatic optimum (Lang, 1994). In the lowlands, mixed oak woodlands developed, characterized by the occurrence of broadleaf trees (Acer, Fraxinus, Quercus, Tilia, Ulmus; Pokorný, 2004). Modern analogues from the Southern Urals indicate that the spread of these trees may have resulted in a dramatic decline in plant species diversity in forest herb layers (Chytrý et al., 2010). How open these new woodlands actually were, and how much energy had to be spent by humans for their initial clearing, is still unclear.

Although the broadleaf trees were present around Zahájí already before the onset of the Neolithic period and although they increased during the early Neolithic, data from both records (pollen and molluscs) suggest that the mid-Holocene forest optimum was not fully attained in the study region. This was probably caused by an early human impact that is dated to the early Neolithic in the Zahájí pollen diagram, i.e., the period corresponding to the very start of agriculture in central Europe (around 7500 cal. BP; Bogaard, 2004). Moreover, recent evidence suggests that already the pre-Neolithic human activity, in particular the use of fire by Mesolithic hunter-gatherers, might have significantly reduced broadleaf forest cover on the territory of the present Czech Republic (Kuneš et al., 2008b; Pokorný et al., 2010b).

In all mollusc successions and recent data available from the old settlement areas of the Czech Republic (Altsiedlungslandschaft according to Firbas, 1949), i.e., from dry lowlands, woodland fauna is consistently poorer than in the precipitation-richer areas with younger settlement history. Interestingly, the same common mollusc woodland species are always missing: Cochlodina orth-ostoma, Daudebardia rufa, D. brevipes, Helicodonta obvoluta, Isognomostoma isogomostomos, Macrogastra plicatula, Petasina unidentata, Ruthenica filograna, Vitrea diaphana and V. subrimata (Juřičková et al., 2013a, 2013b; Ložek, 2007, 2011), which might be the consequence of forest grazing and litter raking since ancient times. Recent and ancient mollusc assemblages thus indicate forests under a long-term use by humans.

Late Holocene cultural steppe

In the Zahájí pollen record, anthropogenic impact sharply intensified at around 5000 cal. BP (Z1c/Z2a zones transition). It is marked by an increase in Calluna vulgaris, an oligotrophic dwarf shrub typical of acidic, nutrient-poor open pastures (Gimingham, 1960). At the same time, nutrient-demanding broadleaf trees such as Fraxinus, Tilia and Ulmus declined while Fagus, Abies and Carpinus increased. The spread of Abies alba in central European forests can be, to a large extent, associated with livestock grazing in forests. Domestic livestock prefers browsing on deciduous trees, and grazed forests usually lack a thick layer of leaf litter which would prevent fir germination (Kozáková et al., 2011; Vrška et al., 2009). This sharp change in anthropogenic impact in Zahájí, indicated also by the increase in Rumex acetosa-type pollen percentages, can probably be interpreted as intensification of livestock grazing (both in forests and in open pastures) and associated nutrient depletion and acidification in soils. From this period, the early Holocene natural steppe probably changed, under anthropogenic impact, into cultural steppe, i.e., semi-natural grassland containing many species of the early Holocene natural steppe (both plants and snails, e.g. Helicopsis striata). At the same time this steppe was affected by management, early livestock grazing and, since the introduction of the grass scythe during the Iron Age (after ca. 2600 cal. BP; Beranová and Kubačák, 2010), also by hay cutting. Studies reporting correlations between prehistoric settlement density and plant species richness in dry grasslands (Hájková et al., 2011; Pärtel et al., 2007) suggest that long-term human management maintained or even enhanced the diversity of the cultural steppe.

Mid-Holocene bottleneck: How long and how narrow?

Most interpretations of the Holocene history of central European steppe grasslands have been inferred from the current landscape processes extrapolated to the Holocene scale. Despite extensive modern deforestation, broadleaf trees clearly demonstrate their ability to establish in most of the current habitats. Except for rock outcrops and some steep south-facing slopes in dry areas, central European steppe grasslands persist today mainly thanks to the livestock grazing or cutting, often practised as a part of nature conservation management (Chytrý et al., 2007; WallisDeVries et al., 1998). In the absence of management, they can be overgrown by scrub and ultimately by shady nutrient-rich forest with Acer, Carpinus, Fraxinus, Quercus, Tilia and Ulmus. This succession also takes place in dry oak or pine forests (Vera, 2000).

These findings, when applied to the Holocene scale, led several authors (e.g. Krippel, 1982; Lang, 1994; Rybníček and Rybníčková, 1994) to the opinion that (a) the natural replacement of steppe by forest started early, as soon as the broadleaf trees spread; (b) most types of steppe, especially those on deeper soils, and many steppe species disappeared in the period between the forest formation and the onset of the Neolithic agriculture, because the forest covered the landscape continuously and the steppe biota was restricted to a few isolated refugia, especially on rock outcrops; (c) current steppe vegetation is of entirely secondary origin, developed at sites that were forested during the mid-Holocene, and cleared by farmers since the Neolithic; and (d) the full renewal of dry grasslands was enabled only by massive Medieval deforestation. This straightforward substitution of ancient processes by the current ones led necessarily to the conclusion that the mid-Holocene bottleneck was rather narrow and that this critical period lasted for a long period.

In contrast, our interpretation supports the perspective of the survival of the early Holocene steppe through the mid-Holocene bottleneck. We prove for the dry area of northern Bohemia that dense broadleaf forest did not cover the entire landscape before the onset of the Neolithic agriculture. Therefore the bottleneck effect was not fatal in the critical period due to the dominance of open pine–birch forests or forest-steppe mosaics, which were directly incorporated into the Neolithic cultural landscape.

Conclusion

Our parallel analyses of pollen and mollusc successions from two sedimentary sequences in the dry lowland area of northern Bohemia provide a strong evidence for the historical continuity of steppe grasslands throughout the Holocene. Although this scenario was previously suggested based on mollusc data, it found no support in previous pollen analyses. For the first time, we provide palynological data which are consistent with the malacological evidence. We show that pine–birch forest-steppe landscape existed in this area until the onset of the Neolithic agriculture. This supports the traditional Gradmann’s ‘steppe theory’, supposing that the first Neolithic farmers in central Europe settled in the still persisting remnants of the late Pleistocene and early Holocene steppes. In our study area, the maximum spread of the mid-Holocene mesophilous broadleaf trees (Acer, Fraxinus, Tilia and Ulmus) was synchronous with the onset of agriculture, which maintained the landscape partly open. Therefore, forest was never able to cover the entire landscape and the mid-Holocene bottleneck did not lead to disappearance of steppe vegetation. Later on, human impact intensified, causing considerable extension of the steppe areas and development of semi-natural grasslands, which harbour many species of the late Pleistocene or early Holocene steppe but at the same time are considerably modified by management, especially livestock grazing. These findings have important implications for the interpretation of temporal and spatial biodiversity patterns in central Europe.

Footnotes

Acknowledgements

Professor Jean Nicolas Haas and also one other anonymous reviewer are greatly acknowledged for many valuable comments to the first version of the manuscript.

Funding

This study was supported by the Czech Science Foundation (GA ČR), project no. 13-08169S.

References

Albert

Pokorný

(2012) Pollen taphonomy and hydrology at Vranský potok versus Zahájí alluvial pollen sites: Methodological implications for cultural landscape reconstruction in the Peruc sandstone area, Czech Republic. Interdisciplinaria Archaeologica: Natural Sciences in Archaeology 3: 85–101.

Beranová

Kubačák

(2010) Dějiny zemědělství v Čechách a na Moravě [The History of Agriculture in Bohemia and Moravia]. Praha: Libri, 430 pp.

Beug

(2004) Leitfaden der Pollenbestimmung für Mitteleuropa und angrenzende Gebiete. München: Friedrich Pfeil, 542 pp.

Bieniek

Pokorný

(2005) A new find of macrofossils of feather grass (Stipa) in an Early Bronze Age storage pit at Vliněves, Czech Republic: Local implications and possible interpretation in a Central European context. Vegetation History and Archaeobotany 14: 295–302.

Birks

HJB

(2005) Mind the gap: How open were European primeval forests? Trends in Ecology and Evolution 20(4): 154–156.

Bogaard

(2004) Neolithic Farming in Central Europe. London: Routledge, 209 pp.

Bohn

Gollub

Hettwer

. (eds) (2000–2003) Karte der natürlichen Vegetation Europas/Map of the Natural Vegetation of Europe. Mastab/Scale 1:2 500 000. Münster: Landwirtschaftsverlag.

Braun-Blanquet

(1961) Die inneralpine Trockenvegetation von der Provence bis zur Steiermark. Stuttgart: Gustav Fischer Verlag, 273 pp.

Chytrý

(2012) Vegetation of the Czech Republic: Diversity, ecology, history and dynamics. Preslia 84: 427–504.

10.

Chytrý

Danihelka

Horsák

. (2010) Modern analogues from the Southern Urals provide insights into biodiversity change in the early Holocene forests of Central Europe. Journal of Biogeography 37: 767–780.

11.

Chytrý

Ermakov

Danihelka

. (2012) High species richness in hemiboreal forests of the northern Russian Altai, southern Siberia. Journal of Vegetation Science 23: 605–616.

12.

Chytrý

Hoffmann

Novák

(2007) Suché trávníky (Festuco-Brometea) [Dry grasslands (Festuco-Brometea)]. In: Chytrý

(ed.) Vegetace České republiky. 1. Travinná a keříčková vegetace [Vegetation of the Czech Republic. 1. Grassland and Heathland Vegetation]. Praha: Academia, pp. 371–470.

13.

Dengler

Janišová

Török

. (2014) Biodiversity of Palaearctic grasslands: A synthesis. Agriculture, Ecosystems and Environment 182: 1–14.

14.

Dúbravková

Hajnalová

(2012) The dry grasslands in Slovakia: History, classification and management. In: Werger

MJA

van Staalduinen

(eds) Eurasian Steppes: Ecological Problems and Livelihoods in a Changing World. Dordrecht: Springer, pp. 253–271.

15.

Dúbravková

Chytrý

Willner

. (2010) Dry grasslands in the Western Carpathians and the northern Pannonian Basin: A numerical classification. Preslia 82: 165–221.

16.

Durka

Nossol

Welk

. (2012) Extreme genetic depauperation and differentiation of both populations and species in Eurasian feather grasses ( Stipa). Plant Systematics and Evolution 299: 259–269.

17.

Eckmeier

Gerlach

Gehrt

. (2007) Pedogenesis of chernozems in Central Europe – A review. Geoderma 139: 288–299.

18.

Ellenberg

(1988) Vegetation Ecology of Central Europe. Cambridge: Cambridge University Press, 731 pp.

19.

Ermakov

Dring

Rodwell

(2000) Classification of continental hemiboreal forests of North Asia. Braun-Blanquetia 28: 1–129.

20.

Fægri

Iversen

(1989) Textbook of Pollen Analysis. 4th Edition. Chichester: John Wiley & Sons, 328 pp.

21.

Firbas

(1949) Spät- und nacheiszeitliche Waldgeschichte Mittel-europas nördlich der Alpen. I. – Allgemeine Waldgeschichte. Jena: Gustav Fischer Verlag, 480 pp.

22.

Frenzel

(1964) Zur Pollenanalyse von Lössen. Untersuchungen der Lössprofile von Oberfellabrunn und Stillfried (Niederösterreich). Eiszeitalter und Gegenwart 15: 5–39.

23.

Frenzel

(1987) Grundprobleme der Vegetationsgeschichte Mitteleuropas während des Eiszeitalters. Mitteilungen der Naturforschenden Gesellschaft in Luzern 29: 99–122.

24.

Frenzel

Pécsi

Velichko

(1992) Atlas of Paleoclimates and Paleoenvironments of the Northern Hemisphere. Late Pleistocene–Holocene. Budapest: Geographical Research Institute, Hungarian Academy of Sciences; Stuttgart: Gustav Fischer Verla, 153 pp.

25.

Gimingham

(1960) Biological flora of the British Isles. Calluna Salisb. A monotypic genus. Calluna vulgaris (L.) Hull. Journal of Ecology 40: 455–483.

26.

Gradmann

(1906) Beziehungen zwischen Pflanzengeographie und Siedlungsgeschichte. Geographische Zeitschrift 12: 305–325.

27.

Gradmann

(1933) Die Steppenheidetheorie. Geographische Zeitschrift 39: 265–278.

28.

Grimm

(1990) TILIA and TILIA GRAPH: PC spreadsheet and graphics software for pollen data (INQUA Working Group on Data-Handling Methods). INQUA Newsletter 4: 5–7.

29.

Hájková

Roleček

Hájek

. (2011) Prehistoric origin of the extremely species-rich semi-dry grasslands in the Bílé Karpaty Mts (Czech Republic and Slovakia). Preslia 83: 185–204.

30.

Hein

(2010) Die Anwendung der Steppenheidetheorie auf das Altsiedelland in Sachsen-Anhalt. München: GRIN Verlag, 76 pp.

31.

Hejcman

Hejcmanová

Pavlů

. (2013) Origin and history of grasslands in Central Europe – A review. Grass and Forage Science 68: 345–363.

32.

Horáčková

Ložek

Juřičková

(2014) List of malacologically treated Holocene sites with brief review of palaeomalacological research in the Czech and Slovak Republics. Quaternary International. Epub ahead of print 16 April. DOI: 10.1016/j.quaint.2014.03.007.

33.

Horsák

Chytrý

Danihelka

. (2009) Snail faunas in the Southern Ural forests and their relations to vegetation: An analogue of the Early Holocene assemblages of Central Europe? Journal of Molluscan Studies 76: 1–10.

34.

Illyés

Bölöni

(eds) (2007) Slope Steppes, Loess Steppes and Forest Steppe Meadows in Hungary. Budapest: MTA ÖBKI, 234 pp.

35.

Juřičková

Horáčková

Jansová

. (2013a) Mollusc succession of the prehistoric settlement area during Holocene: Case study of the České středohoří Mountains (Czech Republic). The Holocene 23: 1809–1821.

36.

Juřičková

Horáčková

Ložek

. (2013b) Impoverishment of recent floodplain forest mollusc fauna in the lower Ohře River (Czech Republic) as a result of prehistoric human impact. Boreas 42: 932–946.

37.

Juřičková

Horsák

Horáčková

. (2014) Patterns of land-snail succession in Central Europe over the last 15,000 years: Main changes along environmental, spatial and temporal gradients. Quaternary Science Reviews 93: 155–166.

38.

Kalis

Merkt

Wunderlich

(2003) Environmental changes during the Holocene climatic optimum in central Europe – Human impact and natural causes. Quaternary Science Reviews 22: 33–79.

39.

Korneck

(1974) Xerothermvegetation in Rheinland-Pfalz und Nachbargebieten. Schriftenreihe für Vegetationskunde 7: 1–196.

40.

Kozáková

Šamonil

Kuneš

. (2011) Contrasting local and regional Holocene histories of Abies alba in the Czech Republic in relation to human impact: Evidence from forestry, pollen and anthracological data. The Holocene 21: 431–444.

41.

Krippel

(1982) Príspevok k pôvodnosti stepi v strednej Európe [Contribution to the origin of steppe in Central Europe]. Geografický časopis 34: 20–33.

42.

Kuna

(1998) The memory of landscapes. In: Neustupný

(ed.) Space in prehistoric Bohemia. Praha: Institute of Archaeology, pp. 77–83.

43.

Kuneš

Abraham

Kovářík

. (2009) Czech Quaternary Palynological Database – PALYCZ: Review and basic statistics of the data. Preslia 81: 209–238.

44.

Kuneš

Pelánková

Chytrý

. (2008a) Interpretation of the last-glacial vegetation of eastern-central Europe using modern analogues from southern Siberia. Journal of Biogeography 35: 2223–2236.

45.

Kuneš

Pokorný

Šída

(2008b) Detection of the impact of early Holocene hunter-gatherers on vegetation in the Czech Republic, using multivariate analysis of pollen data. Vegetation History and Archaeobotany 17: 269–287.

46.

Lang

(1994) Quartäre Vegetationsgeschichte Europas. Jena: Gustav Fischer Verlag, 462 pp.

47.

Lavrenko

Karamysheva

(1993) Steppes of the former Soviet Union and Mongolia. In: Coupland

(ed.) Ecosystems of the World. 8B. Natural Grasslands: Eastern Hemisphere and Résumé. Amsterdam: Elsevier, pp. 3–59.

48.

Litt

(1992) Fresh investigations into the natural and anthropogenically influenced vegetation of the earlier Holocene in the Elbe-Saale Region, Central Germany. Vegetation History and Archaeobotany 1: 69–74.

49.

Ložek

(1964a) Biostratigraphic research of the important Quaternary profiles of the ČSSR. Zprávy o geologických výzkumech 1: 348–350.

50.

Ložek

(1964b) Quartärmollusken der Tschechoslowakei. Rozpravy Ústředního ústavu geologického 31: 1–374.

51.

Ložek

(1973) Příroda ve čtvrtohorách. Praha: Academia, 372 pp.

52.

Ložek

(1982) Faunengeschichtliche Grundlinien zur spät- und nacheiszeitlichen Entwicklung der Molluskenbestände in Mitteleuropa. Rozpravy Československé akademie věd, Řada matematických a přírodních věd 92: 1–106.

53.

Ložek

(2005) Holocene malacofauna from Řisuty and its significance for the environmental history of the north-west Bohemian forest steppe area. Severočeskou Přírodou 36/37: 11–22.

54.

Ložek

(2007) Malakostratigrafie nivy Klentnického potoka u Pavlova (CHKO/BR Pálava). Zprávy o geologických výzkumech 2006: 79–81.

55.

Ložek

(2011) Chronostratigrafic revision and paleoclimatic significance of the talus deposit at the foot of the Martinka cliff in the Protected landscape area and Biosphere reserve Pálava (South Moravia). Geoscience Research Reports for 2010, pp. 66–69.

56.

Magyari

Chapman

Passmore

. (2010) Holocene persistence of wooded steppe in the Great Hungarian Plain. Journal of Biogeography 37: 915–935.

57.

Magyari

Kuneš

Jakab

. (2014) Late Pleniglacial vegetation in eastern-central Europe: Are there modern analogues in Siberia? Quaternary Science Reviews 95: 60–79.

58.

Mahn

E-G

(1965) Vegetationsaufbau und Standortsverhältnisse der kontinental beeinflußten Xerothermrasengesellschaften Mitteldeutschlands. Abhandlungen der Sächsischen Akademie der Wissenschaften Leipzig, Math.-Naturwiss. Klasse 49: 1–138.

59.

Mania

(1995) Zur Paläoökologie des Saalegebietes und Harzvorlandes im Spät- und Postglazial. Mitteilungen der Deutschen Bodenkundlichen Gesellschaft 77: 35–42.

60.

Mitchell

FJG

(2005) How open were European primeval forests? Hypothesis testing using palaeoecological data. Journal of Ecology 93(1):168–177.

61.

Moore

Webb

Collinson

(1991) Pollen Analysis. Oxford: Blackwell, 216 pp.

62.

Neumann

Michaux

Maak

. (2005) Genetic spatial structure of European common hamsters (Cricetus cricetus) – A result of repeated range expansion and demographic bottlenecks. Molecular Ecology 14: 1473–1483.

63.

Pärtel

Helm

Reitalu

. (2007) Grassland diversity related to the Late Iron Age human population density. Journal of Ecology 95: 574–582.

64.

Pavlů

(ed.) (2007) Archeologie pravěkých Čech 3. Neolit [Archaeology of Prehistoric Bohemia 3. Neolithic]. Praha: Archeologický ústav AV ČR, 118 pp.

65.

Pokorný

(2004) Postglacial vegetation distribution in the Czech Republic and its relationship to settlement zones: Review. In: Gojda

(ed.) Ancient Landscape, Settlement Dynamics and Non-Destructive Archaeology. Praha: Academia, pp. 395–414.

66.

Pokorný

Sádlo

Bernardová

(2010a) Holocene history of Cladium mariscus (L.) Pohl in the Czech Republic. Implications for species population dynamics and palaeoecology. Acta Palaeobotanica 50: 65–76.

67.

Pokorný

Šída

Chvojka

. (2010b) Palaeoenvironmental research of the Schwarzenberg Lake, southern Bohemia, and exploratory excavations of this key Mesolithic archaeological area. Památky archeologické 101: 5–38.

68.

Poschlod

Baumann

Karlik

(2009) Origin and development of grasslands in Central Europe. In: Veen

Jefferson

de Smidt

van der Straaten

(eds) Grasslands in Europe of High Nature Value. Zeist: KNNV Publishing, pp. 15–25.

69.

Ramsey

(2009) Bayesian analysis of radiocarbon dates. Radiocarbon 51: 337–360.

70.

Royer

J-M

(1991) Synthèse eurosibérienne, phytosociologique et phytogéographique de la classe des Festuco-Brometea. Dissertationes Botanicae 178: 1–296.

71.

Rybníček

Rybníčková

(1994) Vegetation histories of the Pannonian, Hercynic and Carpathian Regions of the former Czechoslovakia. Dissertationes Botanicae 234: 473–485.

72.

Rybníčková

Rybníček

(1972) Erste Ergebnisse paläogeobotanischer Untersuchungen des Moores bei Vracov, Südmähren. Folia Geobotanica et Phytotaxonomica 7: 285–308.

73.

Rybníčková

Rybníček

(1996) Czech and Slovak Republics. In: Berglund

Birks

HJB

Ralska-Jasiewiczowa

. (eds) Palaeoecological Events During the Last 15 000 Years. Regional Syntheses of Palaeoecological Studies of Lakes and Mires in Europe. Chichester: John Wiley & Sons, pp. 473–505.

74.

Sandom

Ejrnæs

Hansen

MDD

. (2014) High herbivore density associated with vegetation diversity in interglacial ecosystems. Proceedings of the National Academy of Sciences of the USA 111: 4162–4167.

75.

Schwabe

Kratochwil

(2004) Festucetalia valesiacae communities and xerothermic vegetation complexes in the Central Alps related to environmental factors. Phytocoenologia 34: 329–446.

76.

Sugita

Gaillard

M-J

Brostrom

(1999) Landscape openness and pollen records: A simulation approach. The Holocene 9: 409–421.

77.

Svenning

J-C

(2002) A review of natural vegetation openness in north-western Europe. Biological Conservation 104(2):133–148.

78.

Tarasov

Volkova

Webb

. (2001) Last glacial maximum biomes reconstructed from pollen and plant macrofossil data from northern Eurasia. Journal of Biogeography 27: 609–620.

79.

Vera

FWM

(2000) Grazing Ecology and Forest History. Wallingford: CABI Publishing, 506 pp.

80.

Vrška

Adam

Hort

. (2009) European beech (Fagus sylvatica L.) and silver fir (Abies alba Mill.) rotation in the Carpathians – A developmental cycle or a linear trend induced by man? Forest Ecology and Management 258: 347–356.

81.

Vysloužilová

Danková

Ertlen

. (2014) Vegetation history of chernozems in the Czech Republic. Vegetation History and Archaeobotany 23: 97–108.

82.

Wagner

Durka

Hensen

(2011) Increased genetic differentiation but no reduced genetic diversity in peripheral vs. central populations of a steppe grass. American Journal of Botany 98: 1173–1179.

83.

Wagner

Treiber

Danihelka

. (2012) Declining genetic diversity and increasing genetic isolation towards the range periphery of a Eurasian feather grass: A study along a 3000 km longitudinal gradient. International Journal of Plant Sciences 173: 802–811.

84.

WallisDeVries

Bakker

van Wieren

(1998) Grazing and Conservation Management. Dordrecht: Kluwer Academic Publishers, 379 pp.

85.

Walter

(1974) Die Vegetation Osteuropas, Nord- und Zentralasiens. Stuttgart: Gustav Fischer Verlag, 452 pp.

86.

Walter

Breckle

S-W

(1991) Ökologie der Erde, 1: Öko-logische Grundlagen in globaler Sicht. Stuttgart: Gustav Fischer Verlag, 238 pp.

87.

Werger

MJA

van Staalduinen

(2012) Eurasian Steppes. Ecological Problems and Livelihoods in a Changing World. Dordrecht: Springer, 565 pp.

88.

Willner

Sauberer

Staudinger

. (2013) Syntaxonomic revision of the Pannonian grasslands of Austria – Part I: Introduction and general overview. Tuxenia 33: 399–420.

89.

Wróblewska

(2008) From the center to the margins of geographical range: Molecular history of steppe plant Iris aphylla L. in Europe. Plant Systematics and Evolution 272: 49–65.

90.

Zoller

Haas

(1995) War Mitteleuropa ursprünglich eine halboffene Weidelandschaft oder von geschlossenen Wäldern bedeckt? Schweizerische Zeitschrift für Forstwesen 146: 321–354.

91.

Zólyomi

Fekete

(1994) The Pannonian loess steppe: Differentiation in space and time. Abstracta Botanica 18: 29–41.